anc african national congress
TRANSCRIPT
COPYRIGHT AND CITATION CONSIDERATIONS FOR THIS THESIS/ DISSERTATION
o Attribution — You must give appropriate credit, provide a link to the license, and indicate if changes were made. You may do so in any reasonable manner, but not in any way that suggests the licensor endorses you or your use.
o NonCommercial — You may not use the material for commercial purposes.
o ShareAlike — If you remix, transform, or build upon the material, you must distribute your contributions under the same license as the original.
How to cite this thesis
Surname, Initial(s). (2012) Title of the thesis or dissertation. PhD. (Chemistry)/ M.Sc. (Physics)/ M.A. (Philosophy)/M.Com. (Finance) etc. [Unpublished]: University of Johannesburg. Retrieved from: https://ujdigispace.uj.ac.za (Accessed: Date).
AN INTEGRATED BENEFICIARY CENTRED
SATISFACTION MODEL FOR PUBLICLY FUNDED
HOUSING SCHEMES IN SOUTH AFRICA
CLINTON OHIS AIGBAVBOA
2013
i
AN INTEGRATED BENEFICIARY CENTRED
SATISFACTION MODEL FOR PUBLICLY FUNDED
HOUSING SCHEMES IN SOUTH AFRICA
A thesis presented
by
CLINTON OHIS AIGBAVBOA
to
THE FACULTY OF ENGINEERING AND BUILT
ENVIRONMENT
In fulfillment of the requirements for the degree of
DOCTOR OF PHILOSOPHY
in the subject of
ENGINEERING MANAGEMENT
UNIVERSITY OF JOHANNESBURG, JOHANNESBURG,
SOUTH AFRICA
PROMOTER: PROF W.D. THWALA
CO-PROMOTER: PROF B. TWALA
2013
ii
DEDICATION
This thesis is dedicated to my Son (Jehoshua Ohi-Williams Aigbavboa), Wife (Sweeta Vivian
Aigbavboa), and the entire Aigbavboa Williams family; most especially to my Mother and
Brother (father figure) Mansfield Ojo Aigbavboa who denied themselves the necessities of life to
ensure I am educated- thank you for the support and love, without your patience, understanding
and encouragement this study would not have been completed.
iii
DECLARATION
I, Aigbavboa Clinton Ohis, declare that “An integrated beneficiary centred satisfaction model for
publicly funded housing schemes in South Africa” is my own work and that all the sources that I
have used or quoted have been indicated and acknowledged by means of complete references.
The thesis is submitted in fulfillment of the requirements of the degree Doctor of Philosophy in
Engineering Management.
________________________ ________________________
Clinton Aigbavboa Date
iv
ACKNOWLEDGEMENTS
My gratitude to Jehovah God, my Lord Jesus Christ and the Holy Spirit without Whose grace and
inspiration neither I nor this study would have been conceived, and this task could not have been
completed. There is a South African saying that: “Muthu ndi muthu mga vhanwe vhathu” (A
person is a person because of others) and the same can be said of this study.
This moment in my educational journey is one that I could not have reached without the support
of so many individuals. First, I would like to thank my academic supervisor, adviser and mentor,
Professor Wellington Didibhuku Thwala, for his wise and discerning guidance throughout the
thesis process and the wealth of support since my master’s degree. Thank you Professor for the
National Research Fund grants, numerous recommendations for sponsorship and exposure. I was
able to collect data, buy books, laptop, attend national and international conferences as a result of
your generous financial assistance. Thank you for the valuable time spent to offer professional
advice, insight and motivation until the task was completed. My thanks and deepest appreciation
also goes to the following people and institutions:
My Co-Promoter, Professor Bhekisipho Twala for the valuable time spent to offer expert
advice and insight until the thesis was completed;
The University of Johannesburg for the financial (UJ Merit Award for three years and
Supervisor Linked Bursary’s) assistance and office facilities;
Statkon for data input and analysis. My special gratitude goes to Mr. Richard Devey for
the countless hours spent working with me on structural equation modeling;
Mr Ferdinand Fester, Ms. Shandler Maxine, my Head of Departments, Late Mr Ferdinand
Christopher Abrahams, my former senior course coordinator and Dr Pauline Machika for
the support and numerous time off so that I was able to conduct research;
My work colleagues; Dr Innocent Musonda, Dr Justus Agumba, Kauzya Siwale, Ms.
Maphefo, Nazeem, Mr George Onatu (Head of Department, Town and Regional
Planning), Mr. Aurobindo Ogra, Ms. Jordaan Corlia, Patrick, Angel Khumalo, Sadi
Seyama and Erastus for their encouragement and for making this task easier and
enjoyable;
Also, special thanks goes to the Delphi Study expects who gave their valuable time;
My Pastor Bernard Igwe and his wife Pastor Kristabel Igwe, my Senior Pastor
Christopher Oarhe, for their encouragement, support and prayers;
v
My friends, Lerato Ngwenya, Michelle Ayanda, Thuli, Kabelo Lesito, Mailula Kagiso,
Bubu Cromwell, Jessica, Latifah, Koposo, Maake Prudence and others whose
contribution and association made this study a reality; and
My family, the Aigbavboa’s- Caroline & family, Comfort & family, Bose & family,
Peterson Roy & family, Gladys, Solomon Aigbavboa & family, Kunle Okelola &
family, Julius Otoikhian & family, Pastor Jude Oarhe & family, Pastor Glory
Ohiomah & family, Dr Folu Farotimi & family, Dr Deolu Arogundade & family,
Ayotunde Akinradewo & family, Mrs. Fadeke Omatseye and family and others who
are too numerous mention.
--------------------
Aigbavboa, C.O.
November 2013
vi
EXECUTIVE SUMMARY
One of the greatest challenges faced by the post-1994 South African democratic government
is an immense backlog and shortage of housing for poverty-stricken South Africans. Since
1994, the government has embarked on aspiring housing programmes in order to engage in
mass delivery of housing, which was done to fulfill the vision of adequate housing for all, as
reflected in the South African National Housing Policy Framework. Over the last seventeen
years, the programmes have delivered more than 3 million houses to families, who had no
proper housing previously, providing more than 13 million people with secure homes; thus
ensuring that essential services were made available to advance the lives of ordinary people.
This research investigated and modeled subsidised low-income resident’s satisfaction. The
primary aim of the research was to model to what extent dwelling unit features, neighbourhood
features, building quality, services provided by government, beneficiary participation, needs
and expectations predict the occupants’ residential satisfaction, which were classified as the
exogenous variables. A conceptual integrated holistic residential satisfaction model was
developed based on the theory developed from the literature review and the Delphi Study
findings. The Questionnaire Survey was conducted for the purpose of validating the conceptual
model. The survey was conducted in three metropolitan municipalities and one district
municipality in the Gauteng Province of South Africa.
Results from the investigation pertained to three broad areas. The first results related to theory
on housing studies. The findings were that the study addressed the lack of theoretical
information about which factors are most significant in predicting resident satisfaction in
subsidised low-income housing. The findings also revealed the theory that low-income housing
occupants’ satisfaction is multi-faceted and that the latent variables thus lead to residential
satisfaction outcome variables which could be used for residential satisfaction measurement.
The second set of findings relates to the Delphi Study. The findings from this study were that
a number of factors (dwelling unit features, neighbourhood features amongst others),
considered to be paramount determinants of residential satisfaction in South Africa low-income
housing are similar to the determinants in other cultural contexts. Further findings from
literature and the Delphi Study indicated that subsidised low-income housing residential
satisfaction could be a six-factor model defined by the influence of dwelling unit features,
vii
neighbourhood features, and the other classified exogenous variables. The last set of results
pertained to the Field Questionnaire Survey. Generally, the findings were that the hypothesis
could not be rejected. Hence, it was found that dwelling unit features, and the other exogenous
variables, predict subsidised low-income occupants’ residential satisfaction. The Structural
Equation Modeling results on the model’s goodness-of-fit and statistical significance of
parameter estimates met the cut-off criteria for the hypothesised model’s fit to the sample data.
The study’s contribution to the body of knowledge is significant because it addresses the lack
of theoretical information (historical literature data) about which factors are most significant
in predicting resident satisfaction in subsidised low-income housing. Also, the study developed
a new holistically-integrated residential satisfaction model for prediction of residents’
satisfaction in subsidised low-income housing. The current integrated model advances that
residential satisfaction is a six-factor construct, with the inclusion of two new variables,
namely: beneficiary participation and needs and expectations. Previous studies have tried to
model satisfaction using other variables without the inclusion of the present two additional
variables. This study has thus shown that there are more than one factor that influences resident
satisfaction with the dwelling unit. Another noteworthy contribution to the body of knowledge
is in the methodology adopted. The literature review revealed lack of evidence, suggesting that
a Mixed Method of using the Delphi Study and SEM had been used before in South African
housing studies. Hence, this study offers a base for other researchers to use as a follow-up for
future studies.
Therefore, the study recommends that governmental, corporate, institutional and community
policy makers should consider the empirically tested constructs as they plan for, and implement
subsidised housing programmes, designed to enhance the quality of life of the poor and low-
income groups. The results of this study should constitute a reference of guidance in
developing countries’ low-income housing policies. The factors that increase residents’
satisfaction should also be taken into consideration in future planning. Consequently, housing
planners, designers and other stakeholders will be able to contribute to the ways of solving and
improving the low-income groups’ quality of life and level of satisfaction by carefully
regarding the factors that determine residents’ satisfaction in housing. Stakeholders and
institutions, who are involved in the planning process, should consider the contemporary
factors revealing residents’ preferences about housing satisfaction as part of the planning input,
so as to increase the level of housing satisfaction.
viii
Likewise, the validated conceptual model of residents’ satisfaction with their houses which has
been formulated in this study will provide a reference for the researchers who will study
housing satisfaction in the future. Furthermore, the future of subsidised low-income housing in
South Africa should be responsive to the six-factor model and especially to the beneficiaries’
participation and the assessment of their needs and expectations as these are considered vital
in the total housing provision. Thus, the development of low-income housing projects should
take into account the needs of the residents more than their effective demand for housing.
ix
LIST OF ABBREVIATIONS
ANC African National Congress
ANOVA Analysis of Variance
BNP Beneficiaries’ Participation
BNG Breaking New Ground
BQF Building Quality Features
CHB Central Housing Board
CIA Central Intelligence Agency
CFI Comparative Fit Index
CFA Confirmatory Factor Analysis
COSATU Congress of South Africa Trade Union
CSIR Council for Scientific Institute for Research
DF Degree of Freedom
DHA Department of Human Settlement
DM District Municipality
DUF Dwelling Unit Features
EPHP Enhanced People’s Housing Process
EQS EQations Software
FCT Federal Capital Territory
FHA Federal Housing Authority
FSP Family Support Programme
GHS General Household Survey
GLM General Linear Modelling
GFI Goodness of Fit Index
GoG Government of Ghana
HSS Housing Subsidy Scheme
IHP Inclusionary Housing Policy
UISP Informal Settlements Upgrading Programme
IDP Integrated Development Plan
IHHSD Integrated Housing and Human Settlement Development
IRDP Integrated Residential Development Programme
IQD Interquartile Deviation
LM Lagrange Multiplier
x
MI Measurement Invariance
MM Metropolitan Municipality
MOWH Ministry of Works and Housing
MAR Missing at Random
MCAR Missing Completely at Random
MM Mixed Method
MANOVA Multivariate Analysis of Variance
NHC National Housing Code
NHF National Housing Forum
NHSS National Housing Subsidy Scheme
NNS National Norms and Standards
NDF Neighbourhood Features
NGOs Non-Governmental Organisations
PHP People’s Housing Process
RDP Reconstruction and Development Programme
RS Residential Satisfaction
RML Robust Maximum Likelihood
RMSEA Root Mean Square Error of Approximation
S – Bχ2 Satorra-Bentler Scaled Chi-square
SPG Services Provided by Government
SSNIT Social Security and National Insurance Trust
SA South Africa
SD Standard Deviation
SRMR Standardised Root Mean Square Residual
SHC State Housing Corporation
SPSS Statistical Analysis Software Package
SAP Structural Adjustment Programme
SIP Sustainable Ibadan Project
SEM Structural Equation Modeling
SCP Sustainable Cities Programme
TDC Tema Development Corporation
UK United Kingdom
UN United Nations
UNCHS United Nations Centre for Human Settlement
xi
UNDP United Nations Development Programme
US United States
UMP Urban Management Programme
xii
TABLE OF CONTENTS
Contents Page No
DEDICATION...................................................................................................................................... II
DECLARATION................................................................................................................................ III
ACKNOWLEDGEMENTS .............................................................................................................. IV
EXECUTIVE SUMMARY ............................................................................................................... VI
LIST OF ABBREVIATIONS ........................................................................................................... IX
TABLE OF CONTENTS ................................................................................................................. XII
LIST OF TABLES ........................................................................................................................... XIX
LIST OF FIGURES ......................................................................................................................XXIII
LIST OF MAPS .............................................................................................................................. XXV
LIST OF APPENDICES .............................................................................................................. XXVI
CHAPTER ONE ................................................................................................................................... 1
INTRODUCTION ................................................................................................................................. 1
1.1 BACKGROUND........................................................................................................................... 1
1.1.1 RESIDENTIAL SATISFACTION ............................................................................................. 1
1.1.2 HOUSING ADEQUACY ISSUES ............................................................................................ 8
1.2 THE RESEARCH PROBLEM STATEMENT ....................................................................... 11
1.3 AIM OF THE STUDY ............................................................................................................... 12
1.4 RESEARCH MOTIVATION .................................................................................................... 13
1.5 SIGNIFICANCE OF THE STUDY .......................................................................................... 13
1.6 THE STUDY ............................................................................................................................... 14
1.6.1 RESEARCH QUESTIONS .................................................................................................... 14
1.6.2 RESEARCH OBJECTIVES ................................................................................................... 15
1.6.3 RESEARCH METHODOLOGY ............................................................................................ 16
1.6.4 RESULTS .......................................................................................................................... 21
1.6.5 DELIMITATION OF THE STUDY ......................................................................................... 22
1.7 ETHICAL STATEMENT ......................................................................................................... 24
1.8 STRUCTURE OF THE THESIS .............................................................................................. 24
1.9 CONCLUSION ........................................................................................................................... 27
CHAPTER TWO ................................................................................................................................ 28
THEORETICAL AND CONCEPTUAL PERSPECTIVES OF RESIDENTIAL
SATISFACTION RESEARCH ......................................................................................................... 28
2.1 INTRODUCTION ...................................................................................................................... 28
2.2 SATISFACTION THEORY AND DEFINITION ................................................................... 28
2.2.1 APPROACHES TO THE STUDY OF SATISFACTION ............................................................. 31
2.2.1.1 Assimilation Theory ................................................................................................... 31
2.2.1.2 Contrast Theory ......................................................................................................... 32
xiii
2.2.1.3 Assimilation-Contrast Theory .................................................................................... 33
2.2.1.4 Negative Theory ......................................................................................................... 34
2.2.2 FURTHER APPROACHES TO THE STUDY OF SATISFACTION ............................................ 35
2.3 ASSESSING RESIDENTIAL SATISFACTION .................................................................... 38
2.3.1 WHAT IS A THEORY? ....................................................................................................... 38
2.3.2 RESIDENTIAL SATISFACTION RESEARCH ........................................................................ 39
2.3.3 PROBLEMS RAISED IN RESIDENTIAL SATISFACTION STUDY ........................................... 46
2.3.3.1 Definition of residential environment ........................................................................ 46
2.3.3.2 Interaction between the Resident’s and their Residential Environment ..................... 47
2.3.4 METHODOLOGICAL ISSUES IN THE STUDY OF RESIDENTIAL SATISFACTION ................... 49
2.3.5 RESIDENTIAL SATISFACTION CONCEPTUAL MODELS ..................................................... 51
2.3.5.1 Michelson’s Integrated Model .................................................................................... 51
2.3.5.2 Onibokun ‘Habitability’ Model ................................................................................... 52
2.3.5.3 Marans-Rodger Model ................................................................................................ 52
2.3.5.4 Path Analysis Model ................................................................................................... 53
2.3.5.5 Housing Adjustment Model ......................................................................................... 54
2.3.5.6 Francescato Model ..................................................................................................... 55
2.3.5.7 Weidemann and Anderson Model ............................................................................... 56
2.3.5.8 Marans and Sprecklemeyer ‘Inclusive’ Model ........................................................... 56
2.3.6 MEASURING RESIDENTIAL SATISFACTION ...................................................................... 58
2.3.7 MEASURING RESIDENTIAL QUALITY AND ADEQUACY (SATISFACTION) ........................ 60
2.3.8 DETERMINANTS OF RESIDENTIAL SATISFACTION ........................................................... 62
2.4 CONCLUSION ........................................................................................................................... 64
CHAPTER THREE ............................................................................................................................ 66
GAPS IN RESIDENTIAL SATISFACTION RESEARCH ............................................................ 66
3.1 INTRODUCTION ...................................................................................................................... 66
3.2 GAPS IN RESIDENTIAL SATISFACTION CONCEPTUAL FRAMEWORK ................. 66
3.3 GAP ONE: UNDERSTANDING BENEFICIARY’S NEEDS AND EXPECTATIONS ..... 67
3.3.1 SATISFYING HOUSING NEEDS AND EXPECTATIONS ........................................................ 75
3.4 GAP TWO: UNDERSTANDING PARTICIPATION OF BENEFICIARY ........................ 77
3.4.1 ORIGIN OF BENEFICIARY PARTICIPATION ....................................................................... 79
3.4.1.1 Participation as Good Development Project Practice ............................................... 79
3.4.1.2 Participation as Good Governance ............................................................................ 80
3.4.1.3 Participation as Political Empowerment .................................................................... 81
3.4.2 BENEFICIARY PARTICIPATION DEFINED ......................................................................... 82
3.4.3 THE LEGISLATIVE AND POLICY FRAMEWORK FOR PARTICIPATION IN SOUTH AFRICA .. 87
3.4.4 BENEFICIARY PARTICIPATORY PROCESS IN SOUTH AFRICA ........................................... 90
3.4.5 LEVELS OF BENEFICIARY PARTICIPATION ...................................................................... 92
3.4.6 BENEFICIARY EMPOWERMENT ........................................................................................ 97
3.4.7 BENEFITS OF BENEFICIARY BARTICIPATION ................................................................... 98
3.4.7.1 Normative Benefits ..................................................................................................... 98
3.4.7.2 Pragmatic Benefits ..................................................................................................... 99
3.4.8 BENEFICIARY PARTICIPATION IN HOUSING DEVELOPMENT ......................................... 100
3.5 CONCLUSION ......................................................................................................................... 100
CHAPTER FOUR ............................................................................................................................. 102
HOUSING RESEARCH THEORY ................................................................................................ 102
4.1 INTRODUCTION .................................................................................................................... 102
xiv
4.2 HOUSING THEORETICAL FRAMEWORK ...................................................................... 102
4.2.1 THEORETICAL PERSPECTIVE OF HOUSING STUDY ........................................................ 104
4.2.1.1 Neo-Classical Perspective ....................................................................................... 106
4.2.1.2 Institutional Perspective .......................................................................................... 108
4.2.1.3 Neo-Marxist Perspective .......................................................................................... 111
4.2.1.4 Economic Perspective .............................................................................................. 114
4.2.1.5 Social Perspective – The right “to Adequate Housing” .......................................... 117
4.2.2 METHODOLOGIES IN HOUSING STUDIES ....................................................................... 118
4.2.2.1 Positivist Methodology ............................................................................................ 120
4.2.2.2 Social Constructionist Methodology ........................................................................ 122
4.3 WHAT IS HOUSING POLICY .............................................................................................. 125
4.4 THE EVOLUTION OF HOUSING POLICY FRAMEWORK .......................................... 128
4.5 FORMS OF HOUSING POLICY .......................................................................................... 131
4.5.1 PUBLIC HOUSING ........................................................................................................... 132
4.5.2 AIDED SELF-HELP .......................................................................................................... 134
4.5.3 MARKET ENABLING STRATEGY .................................................................................... 136
4.6 OBJECTIVES OF HOUSING POLICY ................................................................................ 138
4.7 THE PURPOSE OF HOUSING POLICY ............................................................................. 139
4.8 HOUSING POLICY INSTRUMENTS .................................................................................. 141
4.9 CONCLUSION ......................................................................................................................... 143
CHAPTER FIVE .............................................................................................................................. 144
HOUSING IN DEVELOPING COUNTRIES – AN AFRICAN EXPERIENCE ....................... 144
5.1 INTRODUCTION .................................................................................................................... 144
5.2 HOUSING IN DEVELOPING COUNTRIES ....................................................................... 144
5.3 NIGERIA .................................................................................................................................. 150
5.3.1 BACKGROUND ............................................................................................................... 151
5.3.2 HOUSING IN NIGERIA .................................................................................................... 152
5.3.3 PHILOSOPHICAL BASIS FOR HOUSING DEVELOPMENT IN NIGERIA ............................... 155
5.3.4 THE HISTORY OF HOUSING POLICY IN NIGERIA ........................................................... 156
5.3.5 HOUSING POLICY IN NIGERIA........................................................................................ 162
5.3.6 CHALLENGES FACING THE PROVISION OF HOUSING IN NIGERIA .................................. 165
5.3.7 PROGRAMMES SUPPORTING HOUSING CREATION IN NIGERIA ..................................... 166
5.3.8 HOUSING IN NIGERIA – NEED, DEMAND AND SUPPLY .................................................. 168
5.3.9 LESSONS LEARNT FROM NIGERIAN HOUSING STUDIES ................................................ 171
5.4 GHANA ..................................................................................................................................... 172
5.4.1 BACKGROUND ............................................................................................................... 173
5.4.2 HOUSING IN GHANA ...................................................................................................... 174
5.4.3 PHILOSOPHICAL BASIS FOR HOUSING DEVELOPMENT IN GHANA ................................ 176
5.4.4 HISTORY AND DEVELOPMENT OF HOUSING POLICY IN GHANA ................................... 177
5.4.5 HOUSING POLICY IN GHANA ......................................................................................... 184
5.4.6 CHALLENGES FACING THE PROVISION OF HOUSING IN GHANA ................................... 185
5.4.7 HOUSING IN GHANA – NEEDS, DEMAND AND SUPPLY .................................................. 186
5.4.8 LESSONS LEARNT FROM GHANA HOUSING STUDIES .................................................... 187
5.5 HOUSING POLICY ISSUES: NIGERIA AND GHANA .................................................... 188
5.6 CONCLUSION ......................................................................................................................... 189
xv
CHAPTER SIX ................................................................................................................................. 190
HOUSING IN SOUTH AFRICA ..................................................................................................... 190
6.1 INTRODUCTION .................................................................................................................... 190
6.2 HOUSING POLICY TRENDS IN SOUTH AFRICA .......................................................... 190
6.2.1 THE EVOLUTION OF HOUSING POLICY IN SOUTH AFRICA ............................................ 192
6.2.2 HOUSING STATUTORY AND POLICY FRAMEWORK IN SOUTH AFRICA .......................... 196
6.2.2.1 Constitution of the Republic of South Africa (1996) ................................................ 198
6.2.2.2 The Housing Act (1997) ........................................................................................... 200
6.2.2.3 National Housing Code (2000, revised in 2009) ..................................................... 202
6.2.2.4 National Housing Programmes ............................................................................... 203
6.3 HOUSING POLICY PROGRESS IN SOUTH AFRICA (1994 – 2010) ............................. 216
6.3.1 NATIONAL HOUSING FORUM (1992-1994) .................................................................... 217
6.3.2 WHITE PAPER ON HOUSING (1994) ............................................................................... 220
6.3.3 BREAKING NEW GROUND (2004) .................................................................................. 222
6.4 HOUSING DELIVERY AND BACKLOGS .......................................................................... 226
6.4.1 HOUSING DELIVERY SINCE 1994 ................................................................................... 227
6.5 HOUSING DELIVERY IN SOUTH AFRICA ...................................................................... 229
6.5.1 STATE SUBSIDISED HOUSING IN SOUTH AFRICA .......................................................... 229
6.6 LESSONS LEARNT FROM SOUTH AFRICA HOUSING STUDIES .............................. 234
6.7 CONCLUSION ......................................................................................................................... 236
CHAPTER SEVEN ........................................................................................................................... 238
RESEARCH METHODOLOGY .................................................................................................... 238
7.1 INTRODUCTION .................................................................................................................... 238
7.2 QUANTITATIVE VERSUS QUALITATIVE RESEARCH METHODOLOGY ............. 238
7.3 PHILOSOPHICAL CONSIDERATIONS IN RESEARCH METHODOLOGY .............. 239
7.3.1 ONTOLOGICAL CONSIDERATION ................................................................................... 239
7.3.2 EPISTEMOLOGY ............................................................................................................. 241
7.3.3 QUANTITATIVE METHODOLOGY ................................................................................... 241
7.3.4 QUALITATIVE METHODOLOGY ..................................................................................... 243
7.3.5 COMBINED QUANTITATIVE AND QUALITATIVE METHODS ........................................... 246
7.3.6 MIXED METHOD APPROACH ......................................................................................... 247
7.3.7 JUSTIFICATION OF THE MIXED METHOD APPROACH .................................................... 250
7.4 RESEARCH DESIGN ............................................................................................................. 251
7.4.1 METHODS ...................................................................................................................... 257
7.4.2 LITERATURE REVIEW .................................................................................................... 257
7.4.3 THE DELPHI METHOD .................................................................................................... 259
7.4.3.1 Epistemological Approach towards the Delphi Design ........................................... 264
7.4.3.2 When to use the Delphi Technique ........................................................................... 265
7.4.3.3 Components of the Delphi Technique ...................................................................... 266
7.4.3.4 Designing, Constructing and Executing the Delphi Study ....................................... 268
7.4.3.5 Specific Objectives of the Delphi ............................................................................. 285
7.4.3.6 Computation of Data from Delphi Study ................................................................. 288
7.4.3.7 Determination of Consensus from the Delphi Process ............................................ 289
7.4.3.8 Reliability and Validity of the Delphi Method ......................................................... 291
7.4.4 QUESTIONNAIRE SURVEY .............................................................................................. 292
xvi
7.4.4.1 Questionnaire Survey Instrument ............................................................................ 296
7.4.4.2 Variables .................................................................................................................. 298
7.4.4.3 Population ................................................................................................................ 301
7.4.4.4 Sample Frame .......................................................................................................... 302
7.4.4.5 Sampling Method ..................................................................................................... 303
7.4.4.6 Sample Size .............................................................................................................. 305
7.4.4.7 Sample Selection ...................................................................................................... 307
7.4.4.8 Site Visits.................................................................................................................. 307
7.4.4.9 Fieldworkers ............................................................................................................ 308
7.4.4.10 Pilot Study ................................................................................................................ 309
7.4.4.11 Data Collection ......................................................................................................... 310
7.4.4.12 Data Analysis from Questionnaire Survey ................................................................ 311
7.4.4.13 Ethical Considerations ............................................................................................. 321
7.4.4.14 Reliability and Validity of the Questionnaire Survey ................................................ 321
7.4.4.15 Generalisability ........................................................................................................ 323
7.5 CONCLUSION ......................................................................................................................... 323
CHAPTER EIGHT ........................................................................................................................... 324
RESULTS FROM THE DELPHI STUDY ..................................................................................... 324
8.1 INTRODUCTION .................................................................................................................... 324
8.2 BACKGROUND TO THE DELPHI SURVEY ..................................................................... 324
8.3 FINDINGS FROM THE DELPHI STUDY ........................................................................... 328
8.4 DISCUSSION OF DELPHI RESULTS ................................................................................. 357
8.5 CONCLUSION ......................................................................................................................... 378
CHAPTER NINE .............................................................................................................................. 379
THE CONCEPTUAL INTEGRATED RESIDENTIAL SATISFACTION MODEL ................ 379
9.1 INTRODUCTION .................................................................................................................... 379
9.2 SELECTION OF VARIABLES FOR RESIDENTIAL SATISFACTION ......................... 379
9.2.1 DWELLING UNIT FEATURES (DUF) ............................................................................... 380
9.2.2 NEIGHBOURHOOD FEATURES (NDF) ............................................................................ 383
9.2.3 BUILDING QUALITY FEATURES (BQF) .......................................................................... 385
9.2.4 SERVICES PROVIDED BY GOVERNMENT (SPG) ............................................................. 386
9.3 MODEL SPECIFICATION AND JUSTIFICATION .......................................................... 386
9.4 STRUCTURAL COMPONENT OF THE MODEL ............................................................. 389
9.5 MEASUREMENT COMPONENT OF THE MODEL ........................................................ 390
9.6 CONCLUSION ......................................................................................................................... 391
CHAPTER TEN ................................................................................................................................ 392
SURVEY RESULTS ......................................................................................................................... 392
10.1 INTRODUCTION ............................................................................................................... 392
10.2 DESCRIPTIVE STATISTICS ............................................................................................ 392
10.3 INFERENTIAL STATISTICS ........................................................................................... 398
10.3.1 STRUCTURAL EQUATION MODELLING (SEM) .............................................................. 398
10.3.2 CONFIRMATORY FACTOR ANALYSIS OF THE LATENT CONSTRUCT .............................. 402
10.3.3 FIT STATISTICS ON MEASUREMENT MODELS (CFA) .................................................... 403
10.3.3.1 Measurement Model for Dwelling Unit Features (DUF) Construct ........................ 403
xvii
10.3.3.2 Measurement Model for Need and Expectation (NAE) Construct ............................ 409
10.3.3.3 Measurement Model for Beneficiary Participation (BNP) Construct ...................... 415
10.3.3.4 Measurement Model for Building Quality Feature (BQF) Construct ...................... 420
10.3.3.5 Measurement Model for Neighbourhood Features (NDF) Construct ...................... 426
10.3.3.6 Measurement Model for Service Provided by Government (SPG) Construct .......... 432
10.3.3.7 Measurement Model for Residential Satisfaction (RS) Outcome Variables ............. 438
10.3.4 STRUCTURAL MODEL – TESTING OF THE HYPOTHESISED SEM MODEL ...................... 445
10.3.5 HYPOTHESISED RELATION FOR THE STRUCTURAL MODEL .......................................... 447
10.3.6 FIT STATISTICS ON THE STRUCTURAL MODEL .............................................................. 447
10.3.6.1 Analysis of Residual Covariance Estimate ............................................................... 451
10.3.6.2 Structural Model Goodness-of-Fit statistics – Robust Maximum Likelihood ........... 451
10.3.6.3 Internal Reliability and Construct Validity of the SEM Model ................................. 453
10.3.6.4 Structural Model Hypothesis Testing ........................................................................ 454
10.3.6.5 Summary on SEM Model .......................................................................................... 459
10.4 CONCLUSION .................................................................................................................... 460
CHAPTER ELEVEN ........................................................................................................................ 461
DISCUSSION OF RESULTS .......................................................................................................... 461
11.1 INTRODUCTION ............................................................................................................... 461
11.2 QUESTIONNAIRE SURVEY RESULTS ......................................................................... 461
11.2.1 DWELLING UNIT FEATURES INFLUENCE ON BENEFICIARY’S RESIDENTIAL
SATISFACTION ................................................................................................................................ 462
11.2.2 BUILDING QUALITY FEATURES INFLUENCE ON BENEFICIARY’S RESIDENTIAL
SATISFACTION ................................................................................................................................ 464
11.2.3 NEIGHBOURHOOD FEATURES INFLUENCE ON BENEFICIARY’S RESIDENTIAL
SATISFACTION ................................................................................................................................ 466
11.2.4 BENEFICIARIES PARTICIPATION INFLUENCE ON RESIDENTIAL SATISFACTION ............. 468
11.2.5 NEEDS AND EXPECTATION INFLUENCE ON RESIDENTIAL SATISFACTION .................... 471
11.2.6 SERVICE PROVIDED BY GOVERNMENT INFLUENCE ON RESIDENTIAL SATISFACTION .. 473
11.2.7 EXTENT THE HYPOTHESISED INTEGRATED MODEL FIT THE IDENTIFIED FACTORS ...... 474
11.3 QUESTIONNAIRE AND DELPHI SURVEY RESULTS ............................................... 476
11.4 CONCLUSION .................................................................................................................... 476
CHAPTER TWELVE ...................................................................................................................... 478
CONCLUSIONS ............................................................................................................................... 478
12.1 INTRODUCTION ............................................................................................................... 478
12.1.1 RESEARCH OBJECTIVE RO1 .......................................................................................... 478
12.1.2 RESEARCH OBJECTIVE RO2 .......................................................................................... 479
12.1.3 RESEARCH OBJECTIVE RO3 & RO4 .............................................................................. 479
12.1.4 RESEARCH OBJECTIVE RO5 .......................................................................................... 480
12.1.5 RESEARCH OBJECTIVE RO6 .......................................................................................... 481
12.2 CONTRIBUTION AND VALUE OF THE RESEARCH ................................................ 481
12.2.1 THEORETICAL CONTRIBUTION AND VALUE .................................................................. 481
12.2.2 METHODOLOGICAL CONTRIBUTION AND VALUE ......................................................... 482
12.2.3 PRACTICAL CONTRIBUTION AND VALUE ...................................................................... 483
12.3 RECOMMENDATIONS ..................................................................................................... 485
12.3.1 METHODOLOGICAL ....................................................................................................... 485
12.3.2 THEORETICAL ................................................................................................................ 485
12.3.3 POLICY IMPLICATION AND PRACTICAL RECOMMENDATION ........................................ 486
xviii
12.4 LIMITATIONS .................................................................................................................... 487
12.5 RECOMMENDATIONS FOR FURTHER RESEARCH ................................................ 488
12.6 CONCLUSION .................................................................................................................... 489
REFERENCES .................................................................................................................................. 492
APPENDIX A .................................................................................................................................... 539
APPENDIX B .................................................................................................................................... 540
APPENDIX C .................................................................................................................................... 541
APPENDIX D .................................................................................................................................... 543
APPENDIX E .................................................................................................................................... 557
APPENDIX F .................................................................................................................................... 576
APPENDIX G .................................................................................................................................... 577
APPENDIX H .................................................................................................................................... 586
xix
LIST OF TABLES
TABLE 2.1: INDIVIDUAL CONCEPTUALIZATION OF RESIDENTIAL SATISFACTION MODELS ................. 57
TABLE 3.1: LADDER OF CITIZEN EMPOWERMENT ................................................................................. 94
TABLE 3.2: A LADDER OF PARTICIPATION ............................................................................................ 95
TABLE 4.1: THE EVOLUTION OF HOUSING POLICY ............................................................................. 130
TABLE 5.1: PERFORMANCE OF PUBLIC HOUSING IN NIGERIA (1960- 2010) ....................................... 160
TABLE 5.2: MAJOR HOUSING POLICY STEPS IN NIGERIA (1928- 2010) .............................................. 163
TABLE 6.1: IRDP HOUSING SUBSIDY QUANTUM AMOUNTS FOR THE 2009/2010 FINANCIAL YEAR ... 209
TABLE 6.2: EPHP HOUSING SUBSIDY QUANTUM AMOUNTS FOR THE 2009/2010 FINANCIAL YEAR ... 212
TABLE 6.3: BREAKING NEW GROUND ELEMENTS AND OBJECTIVES ................................................... 223
TABLE 6.4: ESTIMATED HOUSING DELIVERY FROM 2008 TO 2014 (DHS) ......................................... 228
TABLE 7.1: RESEARCH PROCEDURE .................................................................................................... 256
TABLE 7.2: DELPHI QUESTION FORMULATION .................................................................................... 269
TABLE 7.3: RESIDENTIAL LOCATION OF EXPERTS ............................................................................... 275
TABLE 7.4: QUALIFICATION OF EXPERT’S PANELIST .......................................................................... 276
TABLE 7.5: EXPERT’S PANELIST FIELD OF SPECIALIZATION ............................................................... 276
TABLE 7.6: EXPERT’S PANELIST YEARS OF EXPERIENCE .................................................................... 277
TABLE 7.7: EXPERT’S PANELLIST PUBLICATION HISTORY .................................................................. 277
TABLE 7.8: INFLUENCE OR LIKELIHOOD SCALE .................................................................................. 283
TABLE 7.9: IMPACT SCALE .................................................................................................................. 283
TABLE 7.10: CONCEPTUAL MODEL INDICATOR VARIABLES ............................................................... 299
TABLE 7.11: STUDY POPULATION AND SAMPLE ................................................................................. 303
TABLE 7.12: CUT-OFF CRITERIA OF FIT STATISTICS ............................................................................ 318
TABLE 8.1: DWELLING UNIT ATTRIBUTES .......................................................................................... 330
TABLE 8.2: NEIGHBOURHOOD AND ENVIRONMENTAL CHARACTERISTICS ........................................ 331
TABLE 8.3: HOUSEHOLD CHARACTERISTICS ...................................................................................... 332
TABLE 8.4: SOCIAL FEATURE ASPECTS .............................................................................................. 333
TABLE 8.5: BUILDING QUALITY ASPECTS .......................................................................................... 334
TABLE 8.6: COMMUNITY SERVICES PROVIDED BY GOVERNMENT ..................................................... 335
TABLE 8.7: FACTORS THAT MAKES SUBSIDISED HOUSING UNSUSTAINABLE IN SOUTH AFRICA ........ 339
TABLE 8.8: PREFERRED NATIONAL HOUSING DELIVERY PROGRAMME THAT WILL BETTER SERVE THE
LOW-INCOME GROUPS ................................................................................................................ 344
TABLE 8.9: PREFERRED NATIONAL HOUSING DELIVERY MODEL ...................................................... 344
TABLE 8.10: CURRENT NATIONAL, PROVINCIAL AND LOCAL GOVERNMENT HOUSING DEVELOPMENT
ISSUES ......................................................................................................................................... 351
xx
TABLE 8.11: FORECASTED NATIONAL, PROVINCIAL AND LOCAL GOVERNMENT HOUSING
DEVELOPMENT ISSUES ................................................................................................................ 352
TABLE 8.12: PARAMOUNT NEED OF THE POOR AND LOW-INCOME ASIDE HOUSING ........................... 357
TABLE 9.1: CONCEPTUAL MODEL LATENT CONSTRUCTS .................................................................. 381
TABLE 9.2: FACTORS OF RESIDENTIAL SATISFACTION ....................................................................... 388
TABLE 10.1: RESPONDENTS’ DEMOGRAPHIC AND SOCIO-ECONOMIC CHARACTERISTICS ................. 393
TABLE 10.2: AVAILABLE DWELLING UNIT FEATURES ....................................................................... 396
TABLE 10.3: AVAILABLE DWELLING UNIT SERVICES FEATURES ....................................................... 396
TABLE 10.4: AVAILABLE PRIVATE / PUBLIC NEIGHBOURHOOD FEATURES ....................................... 397
TABLE 10.5: AVAILABLE GOVERNMENT NEIGHBOURHOOD FEATURES ............................................. 397
TABLE 10.6: UNIVARIATE AND MARDIA’S NORMALIZED MULTIVARIATE ESTIMATES ..................... 401
TABLE 10.7: POSTULATED DWELLING UNIT FEATURES MODEL ........................................................ 404
TABLE 10.8: RESIDUAL COVARIANCE MATRIX FOR DWELLING UNIT MODEL (UNSTANDARDIZED) 405
TABLE 10.9: RESIDUAL COVARIANCE MATRIX FOR DWELLING UNIT MODEL (STANDARDIZED) ..... 405
TABLE 10.10: ROBUST FIT INDEXES FOR DWELLING UNIT FEATURES CONSTRUCT .......................... 406
TABLE 10.11: FACTOR LOADING AND Z-STATISTICS OF DWELLING UNIT FEATURES MEASUREMENT
MODEL ....................................................................................................................................... 407
TABLE 10.12: RELIABILITY AND CONSTRUCT VALIDITY OF DWELLING UNIT FEATURE MODEL ...... 408
TABLE 10.13: POSTULATED NEEDS AND EXPECTATION MODEL ........................................................ 410
TABLE 10.14: RESIDUAL COVARIANCE MATRIX FOR NEEDS AND EXPECTATION MODEL
(UNSTANDARDIZED) ................................................................................................................... 411
TABLE 10.15: RESIDUAL COVARIANCE MATRIX FOR NEEDS AND EXPECTATION MODEL
(STANDARDIZED) ....................................................................................................................... 411
TABLE 10.16: ROBUST FIT INDEXES FOR NEEDS AND EXPECTATIONS CONSTRUCT .......................... 412
TABLE 10.17: FACTOR LOADING AND Z-STATISTICS OF NEEDS AND EXPECTATION MEASUREMENT
MODEL ....................................................................................................................................... 413
TABLE 10.18: RELIABILITY AND CONSTRUCT VALIDITY OF NEEDS AND EXPECTATION MODEL ...... 414
TABLE 10.19: POSTULATED BENEFICIARY PARTICIPATION MODEL ................................................... 416
TABLE 10.20: RESIDUAL COVARIANCE MATRIX FOR BENEFICIARY PARTICIPATION
(UNSTANDARDIZED) ................................................................................................................... 416
TABLE 10.21: RESIDUAL COVARIANCE MATRIX FOR BENEFICIARY PARTICIPATION (STANDARDIZED)
.................................................................................................................................................... 417
TABLE 10.22: ROBUST FIT INDEXES FOR BENEFICIARY PARTICIPATION CONSTRUCT ......................... 418
TABLE 10.23: FACTOR LOADING AND Z-STATISTICS OF BENEFICIARY PARTICIPATION MEASUREMENT
MODEL ....................................................................................................................................... 419
TABLE 10.24: RELIABILITY AND CONSTRUCT VALIDITY OF BENEFICIARY PARTICIPATION MODEL . 420
TABLE 10.25: POSTULATED BUILDING QUALITY FEATURES MODEL ................................................. 421
xxi
TABLE 10.26: RESIDUAL COVARIANCE MATRIX FOR BUILDING QUALITY FEATURE MODEL
(UNSTANDARDIZED) ................................................................................................................... 422
TABLE 10.27: RESIDUAL COVARIANCE MATRIX FOR BUILDING QUALITY FEATURE MODEL
(STANDARDIZED) ....................................................................................................................... 422
TABLE 10.28: ROBUST FIT INDEXES FOR BUILDING QUALITY FEATURES CONSTRUCT ..................... 423
TABLE 10.29: FACTOR LOADING AND Z-STATISTICS OF BUILDING QUALITY FEATURES
MEASUREMENT MODEL ............................................................................................................. 424
TABLE 10.30: RELIABILITY AND CONSTRUCT VALIDITY OF BUILDING QUALITY FEATURE MODEL . 426
TABLE 10.31: POSTULATED NEIGHBOURHOOD FEATURES MODEL .................................................... 427
TABLE 10.32: RESIDUAL COVARIANCE MATRIX FOR NEIGHBOURHOOD FEATURE MODEL
(UNSTANDARDIZED) ................................................................................................................... 428
TABLE 10.33: RESIDUAL COVARIANCE MATRIX FOR NEIGHBOURHOOD FEATURE MODEL
(STANDARDIZED) ....................................................................................................................... 428
TABLE 10.34: ROBUST FIT INDEXES FOR NEIGHBOURHOOD FEATURE CONSTRUCT .......................... 429
TABLE 10.35: FACTOR LOADING AND Z-STATISTICS OF NEIGHBOURHOOD FEATURES MEASUREMENT
MODEL ....................................................................................................................................... 430
TABLE 10.36: RELIABILITY AND CONSTRUCT VALIDITY OF NEIGHBOURHOOD FEATURE MODEL .... 432
TABLE 10.37: POSTULATED SERVICES PROVIDED BY GOVERNMENT MODEL ................................... 433
TABLE 10.38: RESIDUAL COVARIANCE MATRIX FOR SERVICES PROVIDED BY GOVERNMENT MODEL
(UNSTANDARDIZED) ................................................................................................................... 434
TABLE 10.39: RESIDUAL COVARIANCE MATRIX FOR SERVICES PROVIDED BY GOVERNMENT MODEL
(STANDARDIZED) ....................................................................................................................... 434
TABLE 10.40: ROBUST FIT INDEXES FOR SERVICES PROVIDED BY GOVERNMENT CONSTRUCT ........ 435
TABLE 10.41: FACTOR LOADING AND Z-STATISTICS OF SERVICES PROVIDED BY GOVERNMENT
MEASUREMENT MODEL ............................................................................................................. 436
TABLE 10.42: RELIABILITY AND CONSTRUCT VALIDITY OF SERVICES PROVIDED BY GOVERNMENT
MODEL ....................................................................................................................................... 437
TABLE 10.43: POSTULATED RESIDENTIAL SATISFACTION MANIFEST MODEL ................................... 439
TABLE 10.44: RESIDUAL COVARIANCE MATRIX FOR RESIDENTIAL SATISFACTION MODEL
(UNSTANDARDIZED) ................................................................................................................... 439
TABLE 10.45: RESIDUAL COVARIANCE MATRIX FOR RESIDENTIAL SATISFACTION MODEL
(STANDARDIZED) ....................................................................................................................... 440
TABLE 10.46: ROBUST FIT INDEXES FOR RESIDENTIAL SATISFACTION CONSTRUCT ........................... 441
TABLE 10.47: FACTOR LOADING AND Z-STATISTICS OF RESIDENTIAL SATISFACTION MEASUREMENT
MODEL ....................................................................................................................................... 442
TABLE 10.48: RELIABILITY AND CONSTRUCT VALIDITY OF RESIDENTIAL SATISFACTION MODEL ... 443
TABLE 10.49: RELIABILITY AND CONSTRUCT VALIDITY OF THE LATENT VARIABLES ...................... 445
xxii
TABLE 10.50: ROBUST FIT INDEXES FOR STRUCTURAL MODEL 2.0 ................................................... 452
TABLE 10.51: MODEL 2.0 FACTOR LOADINGS AND Z-STATISTICS ..................................................... 456
TABLE 10.52: MODEL 2.0 FACTOR LOADINGS, Z-STATISTICS, VARIANCE ACCOUNTED FOR &
RELIABILITY AND CONSTRUCT VALIDITY ................................................................................... 458
xxiii
LIST OF FIGURES
FIGURE 2.1: ACSI MODEL FOR GOVERNMENT AGENCIES .................................................................... 38
FIGURE 3.1: HOUSING NEEDS ORDER ................................................................................................... 70
FIGURE 5.1: SUSTAINABLE MASS HOUSING DELIVERY FRAMEWORK ............................................... 170
FIGURE 7.1: VISUAL MODEL OF MIXED METHODS DESIGN ............................................................... 248
FIGURE 7.2: STEPS IN THE RESEARCH DESIGN PROCESS ...................................................................... 253
FIGURE 7.3: RESEARCH DESIGN OUTLINE .......................................................................................... 258
FIGURE 7.4: DIAGRAM OF THE DELPHI PROCESS ................................................................................ 272
FIGURE 7.5: EXPERT’S PANEL CONTRIBUTION TO THE ABOVE MENTIONED PUBLICATIONS .............. 278
FIGURE 7.6: OUTLINE OF DELPHI PROCESS ......................................................................................... 284
FIGURE 8.1: INFLUENCE OF CORE ATTRIBUTES ON RESIDENTIAL SATISFACTION IN SOUTH AFRICA
LOW-INCOME HOUSING OCCUPANTS .......................................................................................... 329
FIGURE 8.2: ECONOMIC FEATURES ..................................................................................................... 333
FIGURE 8.3: PERSONALITY VARIABLES .............................................................................................. 335
FIGURE 8.4: AESTHETICS VARIABLES ................................................................................................. 336
FIGURE 8.5: LOCATION VARIABLES .................................................................................................... 336
FIGURE 8.6: HEALTH (PERSONAL AND ENVIRONMENTAL) FEATURES VARIABLES ............................. 337
FIGURE 8.7: HOUSING POLICY INSTRUMENTS .................................................................................... 342
FIGURE 8.8: DOMINANCE OF PUBLIC HOUSING (SUBSIDY SCHEME) .................................................. 342
FIGURE 8.9: LACK OF ATTENTION NO OTHER POLICY INSTRUMENTS.................................................. 343
FIGURE 8.10: SOUTH AFRICA HOUSING DELIVERY SYSTEM .............................................................. 346
FIGURE 8.11: WAITING TIME ON HOUSING DATABASE ...................................................................... 348
FIGURE 8.12: PREDICTION OF CURRENT GOVERNMENT PUBLIC HOUSING SUBSIDY MODEL ............ 349
FIGURE 8.13: STATE SUBSIDISED PUBLIC HOUSING BEING THE MAJOR DELIVERY MODEL .............. 350
FIGURE 8.14: STATE SUBSIDISED PUBLIC HOUSING BEING THE MAJOR DELIVERY MODEL ................. 353
FIGURE 8.15: EFFECT OF BENEFICIARY PARTICIPATION IN RESIDENTIAL SATISFACTION ................... 354
FIGURE 8.16: BENEFICIARY PRIOR EXPOSURE TO HOUSING ................................................................ 355
FIGURE 8.17: LOW-INCOME BENEFICIARY HOUSING ORDER NEEDS ................................................... 356
FIGURE 8.18: HOUSING AS PARAMOUNT NEED OF THE POOR AND LOW-INCOME GROUPS IN SOUTH
AFRICA ....................................................................................................................................... 356
FIGURE 9.1: AN INTEGRATED CONCEPTUAL MODEL OF RESIDENTIAL SATISFACTION ...................... 390
FIGURE 10.1: SURVEY LOCATIONS ...................................................................................................... 393
FIGURE 10.2: MEASUREMENT MODEL OF DWELLING UNIT FEATURES ................................................ 403
FIGURE 10.3: MEASUREMENT MODEL OF NEEDS AND EXPECTATION ................................................ 410
FIGURE 10.4: MEASUREMENT MODEL OF BENEFICIARY PARTICIPATION .......................................... 415
FIGURE 10.5: MEASUREMENT MODEL OF BUILDING QUALITY FEATURES ........................................ 421
xxiv
FIGURE 10.6: MEASUREMENT MODEL OF NEIGHBOURHOOD FEATURES ........................................... 427
FIGURE 10.7: MEASUREMENT MODEL OF SERVICES PROVIDED BY GOVERNMENT ............................. 433
FIGURE 10.8: MEASUREMENT MODEL OF RESIDENTIAL SATISFACTION MANIFEST SONSTRUCT ...... 438
FIGURE 10.9: HYPOTHESISED MODEL OF RESIDENTIAL SATISFACTION ............................................. 447
FIGURE 10.10: MODEL 2.0 - INTEGRATED HOLISTIC RESIDENTIAL SATISFACTION MODEL............... 449
FIGURE 10.11: MODEL 2.0 - INTEGRATED HOLISTIC RESIDENTIAL SATISFACTION MODEL
COVARIANCES ASSOCIATION ..................................................................................................... 450
xxv
LIST OF MAPS
Map 1: Map of Nigeria……………………………………………………………..……………...…155
Map 2: Map of Ghana..………………………………………………………………………………178
xxvi
LIST OF APPENDICES
APPENDIX A Invitation Letter to Participate in a Delphi Study
APPENDIX B Request for Expert’s Curriculum Vitae
APPENDIX C Delphi Method and Application to this Study Background
Information
APPENDIX D Delphi Instructions for Round One and Questionnaire
APPENDIX E Delphi Instructions for Round Two and an example of
completed Questionnaire with Group Median
APPENDIX F Instructions to Experts on Delphi Study Round Three
APPENDIX G Research Introduction Letter and Questionnaire
APPENDIX H Model 2.0 Residual Covariance Matrixes (S-Sigma)
1
CHAPTER ONE
INTRODUCTION
1.1 BACKGROUND
The thesis studied the residential satisfaction of public housing beneficiaries’ in developing
nations, using South Africa as a case study. This first chapter is a collection of basic background
information for the research. It starts with research about the contextual background and further
brief supporting aspects in literature.
1.1.1 Residential Satisfaction
Residential satisfaction is defined by Galster (1987) as the perceived gap between a
respondent’s needs and aspiration and the reality of the current residential context. McCray
and Day (1977) also refer to housing satisfaction as the degree of contentment experienced by
an individual or a family member with regard to the current housing situation. Satsangi and
Kearns (1992) and Lu (1999) assert that housing satisfaction is a complex attitude. Hence,
Onibokun (1974) posits that residential satisfaction encompasses satisfaction with the dwelling
unit and the entire neighbourhood. Likewise, Ogu (2002) informs that the concept of residential
satisfaction is often employed to evaluate residents’ perceptions of and feelings for their
housing units and the environment. Lastly, the concept of housing satisfaction has also been
used as a key predictor of an individual’s perceptions of general ‘quality of life’ (Campbell et
al., 1976; cited in Djebarni & Al-Abed, 2000). Scholars such as Andrews and Whitney (1976),
Morris, Crull & Winter (1976), Kleinhans (2007) and Diaz-Serrano (2006) also affirms that
residents’ perception of their environment defines the quality of their lives and determine the
propensity to move.
Research on residential satisfaction has been a major topic in various disciplines such as
sociology, psychology, planning, and geography (Baillie & Peart, 1992; Bruin & Cook, 1997;
Canter & Rees, 1982; Cutter, 1982; Galster & Hesser, 1981; Marans & Rodgers, 1975; Nathan,
1995; Weidemann & Anderson, 1985). In addition, Yiping (2005) states that most of the social
psychology scholars that have dominated satisfaction research vary from consumer
satisfaction, job satisfaction, to patient satisfaction. An understanding of people’s satisfactory
evaluation toward a product or a service is believed to bring improvements, which could thus
2
be found and allocated to the right place and direction, which will improve the effectiveness of
the production or service provision. As such, residential satisfaction research deals with the
occupants’ satisfaction of the, and aims to inform policy and planning intervention.
Furthermore, Lu (1999) posits that the motivation for the high interest and popularity of
residential satisfaction is twofold. First, residential satisfaction is recognized as an important
component of an individuals’ general quality of life; arguing that for most people, housing is
the largest consumption item in their lifetime. Secondly, a home is the place where one most
often finds refuge, rest and satisfaction (Adams, 1984). This means that the degree to which an
individual’s needs and aspirations are met by their housing condition is a concern for
researchers but most importantly for housing developers, planners and specifically for the
Departments of Human Settlement (DHS). This is because the DHS are vested with the
responsibility of policy formulation and implementation of housing delivery programmes for
the low-income groups, as well as the society at large.
Similarly, Nathan (1995) found that in housing programmes targeting the low-income groups,
measures of residential satisfaction will provide additional insights regarding individuals’
experience with housing, and can be used to evaluate the success of the programmes and set
the tone for future developments. Secondly, individuals’ subjective evaluations of their
housing, determine the way they respond to the residential environment and form the basis of
demand for public action (Dahmann, 1985). Speare’s (1974), Mohit, Ibrahim and Rashid
(2010) and Lawhon (2009) findings support this idea, informing that in ‘behavioral
conceptualization’ of migration, low levels of residential satisfaction precede housing and
mobility behaviour. That is, if individuals feel dissatisfied with their current housing situation,
they may well consider relocating and actually moving to a different unit or location. However,
Rory, Maarten and Peteke (2010) in their work on longitudinal analysis of moving desires,
expectations and actual moving behaviour, revealed that housing dissatisfaction, and especially
dissatisfaction with the neighbourhood strongly increases the propensity to desire a move, but
not to expect a move, and the propensity to desire and expect a move. It is further emphasized
in contrast to other previous research that dissatisfaction has a much smaller effect on expecting
an undesired move. Hence, housing dissatisfaction, is closely associated with moving desires,
but not with moving expectations, except when expectations are simultaneously stated with a
desire.
3
Despite a sizeable amount of literature that has developed in this field, an understanding of
how individuals form their residential satisfaction is still inadequate. An obvious sign of this
inadequacy is the existence of inconsistent, sometimes even conflicting, research results about
the factors that shape the residents’ level of satisfaction with their housing and neighbourhood.
This may be as a result of the differences in samples, most samples were not representative of
the population being investigated and the way the key variables were defined, but it may also
be because of how the data was analyzed, hence the current study is determined to overcome
these problems in order to achieve a better understanding of constructs that determines low-
income housing beneficiary satisfaction.
The study of residential satisfaction in developed nations was fostered by two phenomena.
According to Campbell et al. (1976), the first was the postwar housing boom of the 1950s,
early 1960s and the new residential environment through growth of suburban development.
Next is the plight of central city residents under the active programme of slum clearance and
central city rebuilding. Likewise, urban development in developed countries over the years is
similar to the low-income and the large scale inner city redevelopment currently taking place
in cities like Johannesburg, Durban, and Cape Town amongst others in South Africa and in
other developing nations. This has also fostered the increase in the study of how residential
satisfaction is created in developing countries. This thesis is focused on exploring whether
residents are satisfied with the redeveloped and newly built environment low-income housing,
in order to inform policy-making process. There has been much discussion about residents’
satisfaction in South Africa, but the majority has centered on measuring residents satisfaction
in the informal settlement areas and privately owned low-income estates. Only a few studies
have been done with regard to subsidised low-income schemes, with the major focus of the
research being done on the post-occupancy evaluation of the residences, where residential
satisfaction was treated as a sub-objective in these studies (Aigbavboa, 2010; Darkwa, 2006;
Ria & Bontle, 2004). Also, the work of Westaway (2006) on the longitudinal investigation of
satisfaction in a Soweto informal settlement, focused on the effect of satisfaction with personal
and environmental quality of life. The study is aimed to ascertain group and time effects on
satisfaction with personal and environmental domains of quality of life, and to determine
personal and environmental predictors of life and neighbourhood satisfaction in the informal
settlement. From the longitudinal study, it was found that the group from the squatter camp had
the lowest levels of satisfaction with their personal and environmental quality of life. The group
4
was found to be the most disadvantaged in this regard, when compared with the relocated, the
awaiting relocation and the site tenure allocated groups.
Furthermore, Robin, Brian and Kingstone (2007) also measured the quality of life in three
informal settlements in South Africa. In their work, they focused on the factors that are most
important in improving the quality of life of residents in informal housing, as well as the main
obstacles to a better quality of life. Likewise, Moller and Saris (2001) in their work on the
relationship between subjective well-being and domain satisfaction in South Africa explored
the effect of domain satisfaction of finances, housing and social contrast in relation with the
developed countries. However, Ria and Bontle (2004) in their work on the post-occupancy
evaluation of the Hope City Housing Complex (a private low-income housing estate), found
that the residents of the housing estates were satisfied with the facilities in the dwelling units,
the complex and the management components of the estate. Despite the residents being
satisfied with their overall housing situations, it was observed that females were less satisfied
than male. Hence, since no study relating to the absolute study of residential satisfaction in
state subsidized low-income housing in South Africa is available, this study intends to fill the
gap in this area.
Housing occupants responses’ that determines residential satisfaction towards a given housing
situation has been a subject of dispute over time amongst many researchers. Rosenberg and
Hovland (1960; cited in Yiping, 2005) inform that social psychologists generally categorize
people’s responses to any social or physical object into three kinds: the affective, the cognitive
and the conative or behavioral. Affect refers to a person’s feeling towards and evaluation of
some object, person, issue, or event. The cognitive denotes his or her knowledge, opinions,
beliefs, and thoughts about the object. Lastly, conation refers to his or her behavioral intentions
and actions with respect to or in the presence of the object. These three categories provide a
useful framework in understanding and testing of the theoretical development underpinning
residential satisfaction research (Francescato et al. 1987; Weidemann & Anderson, 1985). Kim
(1997) informs that these categories also provide an understanding of the relationships between
objective conditions, subjective experiences and the level of satisfaction with the people’s
living environment. As a result, residential satisfaction is not only used as an indicator in
evaluating housing policies, but also as a predictor of housing quality, propensity to mobility
and the quality of life and well-being of the residents. High residential satisfaction levels have
been considered an indication of the success of specific policies, programmes or designs.
5
Hence, an understanding of the factors that facilitate a satisfied or dissatisfied response can
play a critical part in making successful housing policy decisions.
Furthermore, there is no consensus about what type of evaluation residential satisfaction is.
Whilst some authors conceive residential satisfaction as a purely cognitive evaluation (Canter
& Rees, 1982; Mandler, 1984; Oseland 1990), others have held that it is affect (Weidemann &
Anderson, 1985). However, authors like Francescato et al. (1989) do not think that evaluation
such as satisfaction can be neatly separated into cognition or affect. In addition, satisfaction
studies have been approached from two main perspectives over time namely; satisfaction as a
measure of the degree to which the environment facilitates or inhabits the goal of the user,
called the purposive approach (Canter & Rees, 1982; Oseland, 1990), and those which conceive
satisfaction as a measure of the gap between consumer actual and aspired needs called the
Aspiration-Gap Approach (Galster, 1987). The implication of the purposive approach is that
researchers emphasize goals or associated activities in relation to the attributes of the physical
environment. This approach is entrenched in a cognitive view. However, it is useful because it
assists researchers in understanding the degree to which different aspects and roles of users
contribute to their satisfaction. In addition, people are not only goal-oriented but they have
affective relationships with the environment (or any psychological object), usually involved
with comparisons. This is the comparison between what the beneficiaries have and what they
would like to have or have previously experienced. This is the proposition on which the
Aspiration-Gap methodology is based and the more common conceptual frameworks of
residential satisfaction (Galster, 1987; Morris & Winter, 1975; Weidemann and Anderson,
1985) have all conceived residential satisfaction from this perspective. Also, Morris and Winter
(1975, 1978) and Morris and Jakubczak, (1988) introduced the notion of ‘housing deficit’ and
conceptualize residential satisfaction as a dynamic process. In their housing adjustment theory
of residential mobility, they theorize that individuals judge their housing condition according
to normatively defined norms, which are dictated by societal standards or rules for life
conditions, and family/personal norms, which amounts to household’s own standards for
housing. This means that when the housing norms are met, the household is likely to express a
high level of satisfaction with the housing and the surrounding neighbourhood. On the other
hand, an incongruity between the actual housing situation and housing norms results in a
housing deficit, which gives rise to residential dissatisfaction, leading to some form of housing
adjustments that may be either in situ, such as revising their housing needs and aspirations in
order to reconcile the incongruity, to improve their housing conditions through remodeling, or
6
else they may move to another place and bring their housing into conformity with their
aspirations or needs (Morris & Winter, 1978).
In addition, a more vigorous view of residential satisfaction was developed by Francescato et
al. (1989) when they conceptualized residential satisfaction as an attitude and a multifaceted
construct, which has cognitive, affective and conative dimensions. In their work on evaluating
the built environment from the user’s point of view, they assert that this definition of residential
satisfaction is more comprehensive and that it accounts for the low predictive strength of the
construct in previous studies. Residential satisfaction has also been conceptualized as a multi-
dimensional construct (Bonaiuto et al., 1999), focusing on different specific aspects of a place,
such as the spatial features, human features, and functional features. Various attributes of
housing to which users respond to, in relation to satisfaction are categorized according to a
number of dimensions. Canter and Rees (1982:185) referred to these attributes as the “referent
of interaction”, whilst Francescato (2002) referred to them as the territory of the location.
Generally, these attributes have been categorized in the literature as social/psychological,
management/organizational and physical attributes. Social attributes include privacy,
relationship with neighbours, safety and security, social densities, freedom of choice, social
relations and personalization (Francescato et al., 1979; Rent & Rent, 1978; Spencer & Barneji,
1985). The management attributes usually examined are rules and regulations, maintenance,
management staff and policies, participation and rents (Paris & Kangari, 2006). However, the
physical attributes have received less attention in literature. They usually include the lack or
presence of certain facilities, spatial density, location and size of the bedroom (Galster, 1987;
Kahana et al., 2003; Peck & Stewart, 1985; Turkoglu, 1997). Other physical attributes used in
the literature include the appearance of the building and the floor level (Kaya & Erkip, 2001).
This study will help to bridge the gap that has been created in this area, as the physical attributes
of the subsidised low-income houses being used as a case study will be explored in greater
detail on how they influence housing satisfaction.
On the other hand, physical attributes are not so simple to measure in a way that data may be
obtained about them with confidence (Francescato, 2002). This is why very few physical
characteristics have been examined in most studies on residential satisfaction. An important
physical characteristic, which is not often used in evaluating satisfaction, is the morphological
configuration. Morphological configuration is referred to in architecture as typologies. These
are the spatial and organizational forms of the building, based on certain physical
7
characteristics. This is an important aspect of the design of buildings. However, it is usually
the ‘type of house’, which has usually been examined in satisfaction studies. The ‘type of
house’ refers to terraces, apartments, single family houses or duplexes. The differences
between these house types are more or less functional, rather than morphological. This
classification is not useful in all contexts of housing and especially not in the context of mass
low-income housing, where functional differences do not exist. However, differences in the
morphological characteristics of buildings need to be captured for the purpose of evaluation.
Contrary to some studies (Day, 2000; Francescato et al., 1979), which found that, the type of
site layout (site morphology) and the type of housing (low rise/high rise and detached/attached)
were not predictors of satisfaction. There is enough evidence however to suggest that the
morphological configuration of the residence would significantly affect the level of satisfaction
(Davis & Roizen, 1970; Gifford, 1997:204; Hourihan, 1984). For example, Baum and Valins
(1977) and Baum and Davis (1980) have shown that the length of the corridor of dormitories
has a significant influence on the perception of crowding. In addition, whenever residential
satisfaction has been examined, it usually focused on one (but rarely more than one) of the
levels of the environment, which also refers to as the scale of the environment (Aragones et al.,
2002) or levels of environmental interaction (Canter & Ress, 1982) with very little
differentiation between the levels. In other words, the focus has been on satisfaction with a
level (or scale) of the physical environment, such as the dwelling unit, the neighbourhood, the
community or country of residence. The current thesis differentiates morphological
configuration of the subsidised low-income housing from the physical attributes of the dwelling
units.
Likewise, various demographic characteristics such as sex, age, length of residence, socio-
economic status, race and ethnicity, which influence housing satisfaction, have also been
studied, which researchers have found that they also influence the satisfaction level of housing
units. However, as noted earlier, not much evidence is available on residential satisfaction for
state subsidized low-income housing schemes in South Africa. It is not certain whether the
characteristics, which predict satisfaction for residents in private medium-income and private
low-income housing, informal settlements, and rented apartments would also predict
satisfaction for the beneficiaries of state subsidized low-income housing units in a developing
country context, using South Africa as a case study.
8
1.1.2 Housing Adequacy Issues
Housing production, access and affordability and maintaining existing dwellings in habitable
conditions has been the emphasis of government policies and programmes designed to help the
most poor and those that cannot access housing in South Africa. The United Nations Centre
for Human Settlements (2003) informs that in spite of the national and international efforts
aimed at developing appropriate shelter policies and strategies, no effective remedy has been
found to cure housing ills, as little consideration is most times placed on the issue of the
satisfaction of the beneficiaries and occupants of the housing projects. There are at least two
concerns about housing: one is quantitative – too few housing units for those needing them;
that is, the number of houses provided do not meet the demands for the low-income group. The
other concern is qualitative/- the housing units being provided are unsuitable for the
beneficiaries (the housing type not being satisfactory to the beneficiaries housing needs, even
though there is an improvement of comfort compared to most publicly funded beneficiaries’
former housing situations). Quantitative problems come and go cyclically, depending on the
economy and on the extent of population changes. However, qualitative problems always seem
to stay with us. But their nature however, changes from decade to decade. Qualitative concerns
are very important as it influences the quality of life and affects the psychosocial aspects of the
inhabitants.
Housing has been a major concern for all people in the world, as it has always been considered
as a basic human need. According to Yong (2008) housing fulfills physical needs by providing
security and shelter from the weather and climate and fulfills psychological needs by providing
a sense of personal space and privacy. Housing in South Africa emphasizes the provision of
adequate, affordable and quality houses for all, with a particular emphasis on the low-income
groups, as determined in the South Africa Constitution of 1996 and the Housing White Paper
Policy Framework of 1994. Yong (2008) further emphasized that industrialization and
urbanization have been some of the influencing factors contributing to the acute housing
demand amongst the lower and middle income groups in cities and other larger urban
agglomerations of many developing countries. However, the general demand for housing in
the urban areas far outstrips supply, which in most times is due to the scarcity of suitable
residential land and competing land uses in the urban areas of developing countries. Therefore,
in spite of the South African government commitment and effort in providing adequate,
affordable and quality houses with emphasis on the development of low-income houses for the
9
poor, the houses are still not enough. Also, there are complaints about the housing products
being delivered in the generally housing sector to fully meet the housing need of the low-
income group.
Low-income housing provision has been a major focus of the government in post-apartheid
urban South Africa, as the government attempts to address the historical race-based
inequalities, poor municipal service provision and contemporary rapid urbanization. The South
Africa Housing White Paper of 1994 (Republic of South Africa, 1994), which has undergone
several modifications, prioritized the needs of the poor, encouraged community participation
and the involvement of the private sector, and committed to deliver one million houses in the
first five years (Jenkins, 1999) after the democratic elections. The delivery of one million
houses has since been surpassed. But, the housing needs of the poor and their participation in
the process has not been fully incorporated into the developmental process, resulting in
complaints with the delivered housing products. Since 1994, the low-income housing
programme has mostly involved building serviced townships on urban peripheries, which in
itself presents a myriad of environmental, social and political concerns. Despite this problem,
by the end of 2010, government had built and handed out more than three million houses,
giving shelter to more than 13,5 million people, free of charge, according to the Department of
Human Settlement (2010). Many problems with the process have become clear as the
progression has unfolded. These problems according to Jenkins (1999), include:
1. new houses and infrastructure are of poor quality, and are rapidly deteriorating and
require maintenance;
2. new houses and Human Settlement Development continue placing the poor and low-
income blacks in ‘ghettos’ on urban peripheries, far from jobs and services;
3. people dislike the model of housing used, and would prefer larger houses (the main
model was first changed in 1998 when the Department of Housing, now the Department
of Human Settlement increased the minimum size of new houses to 30m2, and was
further increased in 2004, during the launching of the Breaking New Ground Policy, to
40m2);
4. the dominant model of free-hold tenure inadequately deals with the dynamics of
poverty, and several categories of the poor, such as temporary workers and many
women, which would be better served by rental accommodation as opposed to giving
them houses;
5. because of these problems, people often sell or rent out their low-income houses bought
10
through the subsidy, and move back to squatter or other informal settlements closer to
economic activities; and
6. environmental concerns regarding the new developments include increases in vehicular
traffic caused by urban sprawl and land use changes.
Aigbavboa’s (2010) findings on the post occupancy experience of government housing subsidy
beneficiaries verified the above problems. The study found that most of the occupants in the
South Africa Government Reconstruction and Development Programme (RDP) Housing
Scheme are dissatisfied with the characteristics in the dwelling unit, such as the size of the unit
and inadequacy of rooms in some units. Also, Husna and Nurijan (1987) from their first study
of residential satisfaction of public low-cost flat dwellers in Kuala Lumpur, Malaysia found
that most beneficiaries of the public low-cost flats were also dissatisfied with the characteristics
in the dwelling unit, because there were no dining spaces; bathroom and toilet were
incorporated together in the units. These findings are thus not met making the outcome of the
policies guiding public-funded housing schemes unfulfilled. This further revealed that good
and quality housing is a reflection of the well-being of the community, which refers to the
residents’ acceptance of the houses and what housing provides for the people. Determinants of
good housing can be accessed through the investigation of the satisfaction levels perceived by
the housing residents, through the objective and subjective measures of the right domains that
can determine satisfaction in a given context, like the present studies focusing on subsidized
public housing in South Africa. Further, Marans and Rodger (1975) assert that the concept of
housing satisfaction has been used as an ad-hoc evaluative measure for judging the success of
housing developments constructed by the public sector, which enhances the image of the
housing provider, the public, and contributes towards good housing environment.
However, the success of publicly funded housing schemes does not only depend on just the
provision of housing units, but also on other factors that affect the needs and requirements of
the beneficiaries. The failure of many laudable housing projects in developing countries may
be attributed to the lack of knowledge on the determinants of housing satisfaction. This is
because beneficiaries’ satisfaction reflects the degree to which individuals’ housing needs are
fulfilled, which acts as a guide to policy makers to monitor the implementation of low-income
housing policies and in effect, the quality of life of the people. An understanding of how
individuals form their housing satisfaction is important because the subjective evaluations
determines housing adjustment and mobility behaviour and are the basis of demand for public
11
action. The knowledge is also used to design more effective housing programmes and to avoid
problems that may result because of the perceptions of the planners and policy makers that do
not always coincide with those of the residents.
1.2 THE RESEARCH PROBLEM STATEMENT
It has been shown that residential satisfaction is a very important issue. It is important because
it deals with the housing occupants’ satisfaction, and aims to inform policy and planning
intervention. Similarly, it has also been revealed that measures of residential satisfaction will
provide additional insights regarding individuals’ experience with housing, and can be used to
evaluate the success of the programmes. Also, occupants’ objective and subjective evaluations
of their housing units determines the way in which they respond to the residential environment
and form the basis of demand for public action.
It has also been shown that despite the numerous empirical studies that have been conducted
on residential satisfaction, there is still confusion on the attributes that determine residential
satisfaction. This is because very few researchers have organized these variables into a model,
so as to be able to study and analyze, as a guide, the relationships produced amongst them.
Also, in spite of the sizeable amount of literature that is available in this field, an understanding
of how individuals form their residential satisfaction in public provided subsidised low-income
housing is still inadequate. It is also clear from the background to this study that other studies
that have been conducted on residential satisfaction in South Africa have centered on
measuring residents satisfaction in the informal settlement areas and privately owned medium
and low-income estates with only a few making reference to subsidised low-income housing
as a sub-objective in the studies. Furthermore, the method that was used in these studies may
not always be completely successful. An obvious sign of this inadequacy is the existence of
inconsistent, sometimes even conflicting, research results about the factors that shape
individuals’ satisfaction with their housing units and neighbourhood. This may be as result of
the differences in samples; as the sample for most studies might not be representative of the
population under study and the way the key variables were defined. It may also be because of
how residential satisfaction was treated in the global context of the studies or how the data was
analyzed. Hence, the current study is determined to overcome these problems in order to
achieve a better understanding of the constructs that determine public funded low-income
beneficiaries’ housing satisfaction. Moreover, residential satisfaction of low-income housing
12
in the developing countries, using South Africa as a case study, is still a major concern and the
Human Settlement Department yearns for this to be a thing of the past through programmes
and interventions that are being developed daily and those already implemented.
Therefore the problem that has been addressed in this study may be stated as follows:
Given that the previous models of residential satisfaction established in the developed countries
cannot be relied upon in developing countries, and the findings of what determines residential
satisfaction in public funded low-income housing in developing countries are rarely known
from the previously conducted research, the lack of research into the overall impact and
influence of the direct and holistic active involvement of residential satisfaction constructs, and
the absence of a residential satisfaction model in subsidised low-income housing, the
achievement of occupants’ residential satisfaction is unlikely.
The above problem will be addressed in this study, as outlined in the next sections.
1.3 AIM OF THE STUDY
The research aim of this study is to examine the organize the relationship between publicly
funded beneficiaries dwelling units, neighbourhood and environment facilities, building
quality; which are the essential variables that have been measured in the majority of previous
studies, to develop a holistic integrated residential satisfaction model. The study also takes into
consideration the beneficiaries’ needs and expectations and meaningful consultation with the
beneficiaries (participation), which are all classified as the exogenous variables and their role
in predicting overall beneficiaries’ satisfaction (endogenous variable). This was achieved by
constructing a cohesive model to measure beneficiaries’ satisfaction. The model that was
constructed aided in determining and measuring housing satisfaction in low-income housing
areas among low-income groups. Because these are the groups who usually cannot move away
if they are dissatisfied with the areas or housing units they live in, because of where most of
them are coming from. Also, their economic ability for alternative housing is limited, as most
probably depend on the government for provision of housing. The integrated model when
correctly applied will help to avoid mistakes that have been previously made and will inform
new publicly funded housing schemes, to be developed. The proposed model is context specific
as it relates to the South Africa housing situation.
13
1.4 RESEARCH MOTIVATION
Since no study of residential (beneficiaries’) satisfaction in publicly funded housing in South
Africa is available, the study hopes to fill the gap in this area. Furthermore, the motivation
behind this research is to determine the level of the beneficiaries’ satisfaction in publicly
funded housing schemes in South Africa. The relationship of the exogenous variables, which
are grouped into six components, namely, dwelling unit, neighbourhood and environmental
features, services provided by government, building quality, beneficiaries participation, needs
and expectations (beneficiaries’ participation, needs and expectation are the new attributes
peculiar to the present model to be developed as it has not been previously considered in the
foregoing models of residential satisfaction in the literature); and there resultant influence on
the endogenous variables (residential satisfaction) will be examined. Since there is some
disagreement between researchers as to the relative importance (influence) of the variables,
apart from the new inclusions; the research will assess these housing satisfaction variables and
measure all of them to determine their level of influence (impact) in predicting residential
satisfaction. The principle components of the factors affecting the beneficiaries’ satisfaction
will also be examined.
1.5 SIGNIFICANCE OF THE STUDY
A gap exists in literature on the determinants of residential satisfaction in subsidised low-
income housing in developing countries. This study therefore contributed to existing
knowledge by establishing the factors that determine residential satisfaction in subsidised low-
income housing by establishing the impact of the factors, classified as exogenous variables in
the thesis. In addition, instead of using the existing models as conceptualized for the developed
countries and using instruments such as SERVQUAL amongst others, factors of residential
satisfaction, which have not been considered in previous studies holistically, were evaluated as
outcome variables. The study further used an innovative mixed methodology of Delphi and
Structural Equation Modeling to analyse and model subsidised low-income residential
satisfaction.
Hence, the study adds new knowledge on the factors that determine residential satisfaction in
subsidised low-income housing. The innovative method and the outcome variable measures
used in the study also contribute to the existing body of knowledge on residential (housing)
satisfaction determinants. Similar to the empirical study, a critical literature review on
14
residential satisfaction and housing research theory expanded existing knowledge by providing
synthesized literature that will be useful for improving low-income housing in South Africa
and in other developing countries.
Since the governments of developing countries’ and most specifically the South African
government have been actively providing subsided public housings in different types of
development projects in various locations in South Africa. Therefore, it is important to assess
whether or not these government development projects have met the needs and expectations of
the users. Particularly on the low-income housing development programmes; that came into
being after the draft and implementation of the new Housing Policy Framework in South Africa
in 1994. Also, it is paramount for the government to know, which attributes bring about
residential satisfaction in the constructed low-income housing. This study thus measured the
success of the government body (Department of Human Settlement - DHS) that has been
entrusted with the responsibility of delivering affordable quality housing for the low income
group. The study further provides a guide of vital factors to consider in low-income housing
development. For most individuals, housing is one of the largest investment items of their
lifetime and, as a result satisfaction with their housing situation, is an important component of
their quality of life and well-being.
1.6 THE STUDY
1.6.1 Research Questions
Based on the research problem statement and aim of the study, a few research sub-questions
emerged as stated below:
RQ1 To determine whether the extent of beneficiary’s residential satisfaction is
influenced by the dwelling unit features in the subsidised low-income houses?
RQ2 To determine whether the extent of beneficiary’s residential satisfaction is
influenced by building quality features?
RQ3 To determine how much the level of satisfaction depends upon the
neighbourhood features of the housing subsidy scheme?
RQ4 To determine how much the level of satisfaction is influenced by beneficiaries’
meaningful participation in the entire housing process?
15
RQ5 To determine how much does meeting of the beneficiaries’ needs and
expectations influences their level of the housing unit satisfaction?
RQ6 To investigate the extent to which the level of residential satisfaction is
influenced by the services provided by the government?
RQ7 To what extent the hypothesized integrated residential satisfaction model fits
into the identified factors?
1.6.2 Research Objectives
In order to provide answers to the research questions and achieve the aims of the research, the
following objectives were set:
RO1 To establish the factors that determines residential satisfaction in low-income
housing;
RO2 To investigate and incorporate the current theories and literature that has been
published on residential satisfaction and to identify the gaps that need
consideration;
RO3 To determine the main and sub-attributes that brings about residential
satisfaction and to examine if the attributes that determine satisfaction in other
cultural contexts, is the same in South Africa;
RO4 To evaluate the critical factors and issues that affects the delivery of low-income
housing in South Africa;
RO5 To develop a holistically integrated residential satisfaction model for subsidised
low-income housing; and
RO6 To determine the validity of the conceptualized integrated residential
satisfaction model for subsidised low-income housing.
The achievement of the above research objectives is explained in detail in chapter 8. However,
it is essential that the reader is made aware at this stage as to how the above objectives have
been achieved.
Research objectives RO1 and RO2 where achieved through conducting a literature review on
the subject in question. It is essentially a theoretical understanding of the debate on residential
satisfaction and general housing theories. The objectives RO3 and RO4 were achieved by
16
conducting a Delphi Study. Research objective RO5 was achieved by drawing on the
conclusions from the extensive literature review and the findings from the qualitative Delphi
Study. The Delphi Method is explained in detail in Section 7.3.4 and the reasons why this
method was chosen above other methods is also delimited. The final objective RO6 was
achieved through conducting a field questionnaire survey and modeling of the results, using a
structural equation model’s software known as EQations Software (EQS).
1.6.3 Research Methodology
Research methodology is an arrangement of techniques and guidelines to facilitate the
collection and analysis of data. It provides the starting point for choosing an approach made up
of theories, ideas, concepts and definitions of the topic. Therefore, it is the basis of a critical
activity, which consists of making choices about the nature and character of the social world.
Over time, one of the most important outcomes of the research and review is the
identification of methodological traditions, which, in turn, help to identify data-collection
techniques that can be considered for use in research. This study draws concurrently from
several different methods of investigation. In this thesis, both quantitative and qualitative
research design were used. This is usually referred to as mixed method research design. The
pragmatic (Mixed Method) approach was used in order to answer the research questions and
meet the research objectives thus developing an integrated residential satisfaction model that
applies to the study area, as well as other developing countries. Below is a basic discussion of
the research design; however, the chapter on research methodology (Chapter 7) elaborates on
the individual methods.
Quantitative Research
Quantitative research places emphasis on measurement when collecting and analysing data.
Quantitative research is defined, not just by its use of statistical measures but also that it
generally follows a natural science model of the research process measurement to establish
objective knowledge (that is, knowledge that exists independently of the views and values of
the people involved). Generally, it makes use of deduction, that is, the research is carried out
in relation to an informed proposition, or speculation, that is expressed in a way which can be
tested from theory. It focuses on the possible relationship between two or more variables. Just
as the case of the present study, which is positioned to observe the relationship between low-
income beneficiaries housing dwelling units, neighbourhood features, services provided by
17
government, building quality, beneficiary’s participation, needs and expectations in predicting
their residential satisfaction. The quantitative method of data collection for the present study is
done by users Survey Method with the application of a structured questionnaire, which had
been piloted with the housing occupants. The analysis was done through the use of Structural
Equation Modeling with EQS, which was used in the development of the integrated residential
satisfaction model.
Qualitative Research
Qualitative research emphasises meanings (words) rather than frequencies and distributions
(numbers) when collecting and analysing data. Some researchers argue that qualitative research
is also concerned with issues of measurement, but with measurement that is of a different order
to numerical measurement. Thus, qualitative ‘measurement’ is often binary in that it is
interested in the presence or absence of phenomena, or it works implicitly with simple scales.
For instance, how much conversation or laughter or aggression or mutual touching in a
particular interaction? Predominantly qualitative research seeks to understand and interpret the
meaning of situations or events from the perspectives of the people involved and as understood
by them. It is generally inductive (the process of inferring a generalised conclusion from
particular instances) rather than deductive in its approach. That is, it generates theory from
interpretation of the evidence, though against a theoretical background. The qualitative method
implored in this study is the structured and semi-structured (using an interview guide)
interview. This was made possible through the use of the Delphi Technique. The findings from
this section of the study help to refine the survey tool (structured questionnaire) for the study
and to validate the findings. The Delphi findings were further used to resolve conflicting issues
surrounding residential satisfaction and other housing study issues in the study setting (South
Africa) through the consensus (agreement and disagreement) that was reached when the Delphi
studies was conducted. The Delphi technique is discussed in great detail in the Methodology
Chapter (Chapter 7).
Mixed Method Research
According to Tashakkori and Teddlie (2003) Mixed Method research is a research design with
philosophical assumptions, as well as different methods of inquiry. As a methodology, it
involves philosophical assumptions that guide the direction of the collection and analysis of
data and the mixture of qualitative and quantitative approaches in different phases in the
research process. As a method, it focuses on collecting, analysing, and mixing both quantitative
18
and qualitative data in a single study or a series of studies. Its central premise is that the use of
quantitative and qualitative approaches in combination provides a better understanding of the
research problems than only one approach alone. The quantitative data in a typical mixed
method includes closed-ended questions such as those found about attitude, behaviour, or
performance instruments. The collection of this kind of data involves using a closed-ended
checklist, on which the researcher ‘checks’ the behaviour seen. Sometimes quantitative
information is found in documents, such as census records or attendance records. The analysis
of the quantitative data consists of statistically analysing scores collected on instruments and
checklists to answer research questions or to test hypotheses or to answer the research questions
(Creswell, 2003; 2010).
In contrast, typical qualitative data in a mixed methodology study consists of open-ended
information that the researcher gathers through interviews with participants. The general open-
ended questions asked during interviews allow the participants to supply answers in their own
words. These can be thematically analysed and converted into qualitative data, which can also
be transcribed in quantitative data, such as the case of Delphi Studies where frequencies of
measures of central tendencies are used to draw consensus. Also, qualitative data may be
collected by observing participants or sites of research, gathering documents from a private or
public source, or collecting audiovisual materials, such as videotapes or artifacts. The analysis
of the qualitative data (words or text or images) typically follows the path of aggregating the
words or images into categories of information and presenting the diversity of ideas gathered
during data collection. The open- versus closed-ended nature of the data differentiates between
the two types better than the sources of the data.
The combination of both approaches (qualitative and quantitative) can offset the weaknesses
of either approach when used itself. For instance, mixed method research provides more
comprehensive evidence for studying a research problem than either quantitative or qualitative
research alone. Also, researchers are given permission to use all of the tools of data collection
available, rather than being restricted to the types of data collection typically associated with
qualitative or quantitative research (Creswell, 2003). Further, Mixed Method research helps
answer questions that cannot be answered by qualitative or quantitative approaches alone. It is
also very practical because the researcher is free to use all methods possible to address a
research problem. It can be used to increase the generalisability of the research result, which in
this present day and age is a major consideration. It can also provide stronger evidence for a
19
conclusion through convergence and verification of findings. It can furthermore add insight
and understanding that might be missed, when only a single method is applied. It also provides
complete knowledge necessary to inform theory and practice and can answer a broader and
more complete range of research questions because the researcher is not confined to a single
method or approach (Cameron, 2011; Lisle, 2011). In addition because individuals tend to
solve problems using both numbers and words, that is, a combination of inductive and
deductive thinking, this method is more reliable and valid (Johnson, Onwuegbuzie, & Turner,
2007). Therefore, it should become natural, to employ Mixed Method research as the preferred
mode of understanding the world.
For instance, when the discussion about the housing satisfaction of the South Africa public
housing schemes beneficiary is debated both numbers and words comes to mind. This is
because the debate is natural, psychological and persuasive which neither words nor numbers
can adequately represent. Hence, words, pictures and narratives can be used to add meaning to
numbers, and numbers can be used to add precision to words, pictures and the narrative
(Creswell, Tashakkori, Jensen, & Shapley 2003; Mayring, 2007). Therefore, the Delphi
Technique was combined with the Survey Method in this research study which provided the
basis for the validation of the conceptual framework for the development of an integrated
residential satisfaction model, in developing countries, using South Africa’s three metropolitan
municipalities and one district municipality, as a case study.
Data needed and means of obtaining it (empirical measures)
Literature on residential satisfaction, general housing and other aspects as related to the study,
was reviewed to provide a background to the study. Various sources were reviewed including
books, articles in accredited journals, published and unpublished works such as dissertations
and web-based publications on the specified filled of study.
Two methods were used to collect the empirical data. These methods were the Delphi and Field
Method Questionnaire Survey Method, respectively. A detailed description of both these
methods is presented in Chapter 7 of the thesis.
With the Delphi Method, the data that needed to be collected was the prediction of the
likelihood of residential satisfaction factors and other housing issues surrounding low-income
housing in South Africa. The method was used for the second stage of the study to identify the
20
main attributes that bring about residential satisfaction and to examine if the attributes that
determine satisfaction in other cultural contexts, as identified from the literature, is the same
within South Africa (developing countries). Also, the Delphi Technique was used to explore
the extent to which to which these main attribute sub-factors impact / influence residential
satisfaction in South African low-income housing. This data was obtained through the use of a
Structured Questionnaire Interview. Experts were asked to complete the questionnaires and
consensus was reached on the rated likelihoods and impact of various factors. The process
involved a three round iterative process, with the main aim of getting experts to reach consensus
on the questions raised in the questionnaires. Experts were also encouraged to give reasons for
their dissenting views.
In the questionnaire survey, the data that needed to be collected was the evidence of factors
that determine residential satisfaction, namely: dwelling unit features, neighbourhood features,
building quality, beneficiaries’ participation, needs and expectations. In addition, data
regarding the outcome of overall residential satisfaction from the occupants were also needed.
This data was obtained with the use of structured questionnaires. These were completed by the
low-income housing occupants, who are residents of the particular subsidised low-income
housing location.
Data sources
In the Delphi Study, data regarding the rating of the likelihood and impact of the factors that
determine residential satisfaction and other issues regarding low-income housing in South
Africa was obtained from the expert panels. Likewise, data from the questionnaire survey was
obtained from residents of subsidised low-income housing locations. The houses which have
been completed and allocated by the respective municipality to the occupants’ were only used
for the survey.
Data analysis
Data obtained from the Delphi Study was analysed in Microsoft Excel, a spread-sheet software
programme. The output from the analysis was a set of descriptive statistics such as means,
standard deviations and their respective derivatives.
As for data obtained through the Field Questionnaire Survey, Structural Equation Modeling
(SEM) was utilised using the EQS 6.2 software package. Outputs from the analysis were
21
univariate and multivariate descriptive statistics, as well as measures of goodness of fit of the
hypothesis model. Other outputs included measures of statistical significance of parameter
estimates. However, the demographic characteristics of the respondents were analysed using
descriptive statistics.
1.6.4 Results
An effort directed at measuring residential satisfaction requires an understanding of the major
influences and determinants on the residential facility. Results of the study related to the
relationship between the exogenous variables (dwelling unit features, neighbourhood and
environment features, building quality, beneficiaries’ participation, needs and expectations)
and the endogenous variable (residential satisfaction); will be presented in graphs and tables of
values describing the extent of the attributes and the sub-attributes on overall residential
satisfaction.
Utilising data from both the Delphi and the Questionnaire Survey, a holistic integrated model
for residential satisfaction in developing countries was conceptualized and validated. This
model will be presented in the final output of the study.
Delphi Specific Objectives
The objective for conducting the Delphi Survey, for this study, was to determine the following:
DSO1 To identify the attributes (main and sub) that determine residential satisfaction
and to examine if the attributes that determine satisfaction in other cultural
contexts, is the same in South Africa;
DS02 To determine the factors that makes subsidised public housing unsustainable in
South Africa;
DS03 To identify the combination of housing policy instruments that will better serve
South African subsidised low-income housing groups;
DS04 To identify the critical factors affecting the delivery of low-income housing and
their effects on beneficiaries’ residential satisfaction;
DS05 To predict the life span of the present South African public housing subsidy
delivery model;
22
DS06 To evaluate the management issues affecting the national, provincial and local
government housing agencies in the delivery of housing in South Africa;
DS07 To determine the influence of beneficiary participation on their overall housing
satisfaction; and
DS08 To determine the effect of meeting beneficiary’s housing needs and
expectations on their overall housing satisfaction.
The main outputs from the Delphi Study will be the identification of the factors of residential
satisfaction and to respond to other pressing housing issues in South Africa, with significant
influence and a conceptual model defining residential satisfaction in South Africa. The
conceptual model will be validated by results from the Field Questionnaire Survey.
Specific Objectives of the Field Questionnaire Survey
The specific objectives for conducting the questionnaire survey and thereby satisfying the
general research objective RO6 of validating the conceptual model were to:
QS1 Identify the factors that had a higher (direct) influence on low-income housing
occupants’ residential satisfaction;
QS2 Establish the influence of the identified factors on occupants’ residential
satisfaction;
QS3 Determine the influence of the overall residential satisfaction on subsidised low-
income occupants’ behaviour; and
QS4 Determine the goodness-of-fit of the hypothesised integrated residential
satisfaction model to the sample data.
However, the general hypothesis tested in the study is based on the fact that overall residential
satisfaction is directly related to the influence of the exogenous variables’ in predicting /
determining overall beneficiaries’ satisfaction in publicly funded housing schemes in
developing countries using South Africa as a case study.
1.6.5 Delimitation of the study
The scope of this thesis studied beneficiaries’ satisfaction in publicly funded housing schemes
in South Africa. The study focused on the influence of the exogenous variables (dwelling unit
23
features, neighbourhood and environment features, building quality, beneficiaries’
participation, needs and expectation) on the endogenous variable (residential satisfaction). A
causal relationship between the exogenous variables and the endogenous variable was
established through Structural Equation Modelling (SEM) using the software EQS version
programme 6.2. SEM was used in order to aid testing for the factorial validity and scores from
the measuring scale adopted and therefore, to predict a reliable model. The result of the study
will be presented as statistical measures in literature, tables, charts and graphs.
The study was limited to the beneficiaries of the public housing scheme or better known as the
Reconstruction and Development Programme (RDP) of housing in South Africa. Only RDP
houses with the traditional design of one room (30-40m2) and two rooms (40-45m2) units were
considered in the study. The research thesis empirically studied whether beneficiaries living in
these houses are satisfied with their living conditions, in order to inform the housing policy-
making process, and contribute to the theory building of beneficiaries (residential) satisfaction.
Households that have benefited from government housing subsidy schemes were engaged in
conducting the research. Only the low-income group and the previously disadvantaged were
considered in the research. Considering the households that have benefitted from the housing
subsidy scheme; housing units that were subsidized but with a little level of participation
(procured through the Peoples’ Housing Process or the enhanced People’s Housing Process)
from the beneficiaries was not considered in the study. Only housing units that were
constructed through the capital subsidy schemes (Project Linked Subsidy now known as
Integrated Residential Development Programme) where beneficiaries are shortlisted and given
a completed housing product, were considered for the study.
The study was restricted to three recognized metropolitan municipalities (MM) and one district
municipality (DM) in the Gauteng Province of South Africa. These were: Ekurhuleni (East
Rand) MM, City of Johannesburg (Johannesburg) MM, and Tshwane (Pretoria) MM and
Mogale City (Krugersdorp) DM, where the government has extensively been involved in the
provision of housing for the low-income groups. The same criteria were used in selecting the
survey samples for the study areas. The study analysed the level of beneficiaries’ satisfaction
of public housing in South Africa. This was done by examining the level of satisfaction and
the factors affecting it.
24
Finally, the study discussed the implication of its findings on housing policy and made
recommendation to improve the existing strategies of low-income housing development in
South Africa, and in other developing countries.
1.7 ETHICAL STATEMENT
Ethical issues were a key consideration in undertaking this study. The principal of voluntary
participation was upheld. This required that people were not coerced into participating in the
research. Further, participants were only involved in the research where an informed consent
had been established. Great effort was made to help protect the privacy of research participants
by ensuring confidentiality in not making available identifying information to anyone who was
not directly involved in the study. Confidentiality was further enhanced by keeping participants
anonymous throughout the study.
1.8 STRUCTURE OF THE THESIS
The compilation of the entire thesis will be organised as follows:
Chapter 1 – Introduction
This chapter presents information on the background to the study, and also the main research
problem. The chapter also presents a general description of the study stating the aim, and
objectives of the study. In addition, a description of the methods that were used to conduct this
study, including the results that were obtained and ethical considerations are presented in this
chapter.
Chapter 2 - Theoretical and Conceptual Perspectives of Residential Satisfaction Research
This chapter presents an account of theoretical and conceptual perspectives of literature on
residential satisfaction that was needed to inform this study. The chapter presented a survey of
related literature from books, journal articles, conference papers, and internet searches from
notable databases. The chapter is presented relative to the guiding questions of the study.
Chapter 3 - Gaps in Residential Satisfaction Research
This chapter addresses the gaps observed in residential satisfaction research, which was not
evaluated as an all-inclusive construct in the previous models. Though they are mentioned in
the discussion of the previous models but not evaluated as single constructs. These gaps form
25
the additional new constructs in the current study’s conceptual framework. The gaps identified
are: beneficiaries’ needs and expectations, and their participation in the housing process. The
identified gaps are discussed and also how to achieve them in low-income housing
development process is explained.
Chapter 4 - Housing Research Theory
This chapter of the thesis presents a general overview of housing research, housing theoretical
framework and overview of the most influential perspectives on housing. This is followed by
a discussion on the methodological approaches to housing studies. An evolution of housing
policy framework is also presented, with the various forms that housing policy has been
attended to over time. Also, the objectives and purpose of housing policy are further examined.
Lastly, the chapter closes with an outline of housing policy instruments, which enables the
intentions of housing policy to be actualized.
Chapter 5 - Housing in Developing Countries – An African Experience
Provision of adequate housing is one of the major challenges faced by most African countries.
Hence, in this chapter, housing policies and other housing issues in the African countries of
Ghana and Nigeria are discussed. The roles played by the different bodies such as government,
non-governmental organisations in the provision of housing were also considered. This is
because in most West Africa countries, like Nigeria and Ghana, the ownership of affordable
good quality housing has been a problem justifying serious private sector intervention. Unlike
South Africa, these African countries and other developing countries have reluctantly refused
to include the right to housing in their national constitution, undermining the multiplier role of
housing in national development. The chapter set out a background on housing in developing
countries. However, the two African countries (Ghana and Nigeria) policies’ and schemes’
were explored.
Chapter 6 - Housing in South Africa
This chapter of the thesis provides an outline of housing legislation and jurisprudence, policy
and implementation in South Africa since 1994 because South Africa is the focal point of the
study. This section outlines the housing legislative and policy framework in South Africa;
examining the Constitution with specific reference to the Bill of Right and the Right to
Housing; National Housing Code; and the National Housing Programmes categorized therein
with a specific focus on state subsidised housing (Housing Subsidy Scheme - HSS). This
26
chapter also presents an overview of the developments in Housing Policy since 1994, including
a summary of the negotiations at the National Housing Forum held between 1992 and 1994.
The section further examines the supreme policy framework contained in the 1994 White Paper
on Housing, and the problems associated with the Reconstruction and Development
Programme (RDP) houses (subsidised houses) built after 1994. This is necessary because the
central focus of the thesis is on the RDP houses built after 1994. The chapter also discusses the
2004 BNG Policy Amendment. Other specific policies, which are examined in this section
include the People’s Housing Process (PHP) now enhanced ePHP and the Inclusionary
Housing Policy (IHP), to ascertain if the programmes would have guaranteed a better housing
satisfaction compared to the model used for the initially built houses and those still being
constructed. This section also included an outline of the National Housing Code published in
2000 and recently updated in 2009, amongst others. Lastly, a summary of the lessons learnt to
date from the literature is presented and a comparison is drawn between South African and the
two Africa countries (Nigeria and Ghana) previously reviewed (Chapter 5).
Chapter 7 - Research Methodology
This chapter contains a detailed description of the methodology, methods and the tools used to
collect data for this study. It also describes the participants of this study, as well as a detailed
description of the results, analysis of the results and how results are presented in the findings
sections. The above is done for every method of data collection used in this study, namely: the
Delphi and Field Questionnaire Survey methods. The research design is also described in this
chapter and how the collected data was treated. Finally, a description of the population, the
sampling design and the interpretation of results was also presented in this chapter.
Chapter 8 – Results from the Delphi Study
This chapter discusses the findings from the Delphi Method. These findings are presented
relative to each Delphi Objective. Also, a discussion and interpretation of the Delphi results is
presented at the end of this chapter.
Chapter 9 – The Conceptual Integrated Residential Satisfaction Model
Chapter 9 is a discussion of findings from the review of literature and the Delphi Study. This
discussion forms the basis of the conceptual model’s theory. The hypothesised integrated
holistic residential satisfaction model is then presented in this chapter. This chapter describes
the integrated holistic model and the variables of the model in detail except beneficiaries’
27
participation, needs and expectations, which has already been discussed in chapter three of the
thesis.
Chapter 10 –Survey Results
This chapter explains the findings from the questionnaire survey. Findings from this survey are
discussed relative to each questionnaire survey objective.
Chapter 11 – Discussion of Results
This chapter covers discussion, analysis and interpretation of the results obtained from the
questionnaire. Discussions on the goodness-of-fit of the postulated integrated holistic
residential satisfaction model are presented in this chapter.
Chapter 12 – Conclusion and Recommendations
This chapter concludes the study and contains recommendations based on the conclusions
drawn from the study. Recommendations for future research are also presented in this chapter.
As part of the recommendations, a description of the finalized integrated holistic model is done.
The study was also subject to independent review by the promoter (Professor Wellington
Thwala) to help protect all participants and the researcher against potential any legal
implications of neglecting to address important ethical issues and also to uphold integrity,
honesty and quality assurance.
1.9 CONCLUSION
Chapter 1 introduced the subject of the research study; it gave insight into the structure, the
background and significance thereof. It relays information on how the research report is
presented. The next chapter looks into the theoretical and conceptual perspectives of residential
satisfaction research and other related literatures in line with the objectives of the study.
28
CHAPTER TWO
THEORETICAL AND CONCEPTUAL PERSPECTIVES OF
RESIDENTIAL SATISFACTION RESEARCH
2.1 INTRODUCTION
This chapter is focused on the review of theoretical and conceptual perspectives on residential
satisfaction research. Also discussed are previous residential satisfaction models and theories,
residential satisfaction research (conceptual issues on residential satisfaction), such as the
appropriate ways of measuring residential satisfaction, followed by the discussion on the
determinants of residential satisfaction. The problems raised to date in residential satisfaction
research were also reviewed.
2.2 SATISFACTION THEORY AND DEFINITION
The foundation for satisfaction lies in “mankind’s ability to learn” from previous experiences
(Peyton, Pitts, & Kamery, 2003:41). Likewise, user’s preferences are constantly being updated
by way of the learning process. Learning theory posits that “… a given response is strengthened
either positively or negatively to the extent that it is followed by a reward. Reward, in turn,
leads to an evaluation that the purchase or achievement was satisfactory… and hence it can
exert an effect on product beliefs and attitude. The probability of engaging in a similar buying
act or continuance in a housing scheme will be increased if there are positive consequences in
the act of purchase”, use of the unit and vice versa (Engel, Kollat & Blackwell, 1968:532).
The word satisfaction first appeared in English in the thirteenth century. The word itself is
derived from the Latin word satis (meaning enough) and the Latin ending -faction (from the
Latin facere - to do/ make). Early usage of the word satisfaction focused on satisfaction being
some sort of release from wrong doing. Later citing’s of the word emphasises satisfaction as a
freedom from uncertainty (The Oxford Library of Words and Phrases, 1993:1258). Modern
usage of the word has tended to be much broader, and satisfaction is clearly related to other
words such as satisfactory (adequate), satisfy (make pleased or contented) and satiation
(enough). The study on satisfaction has grown rapidly over time with more than 500 studies
carried out on the concept in the 1970’s alone as informed by Hunt (1977). However, despite
29
the overwhelming quantity of literature surrounding the concept in the present time, Anderson
and Fornell (1994: 244) noted that “certain key issues have either gone unresolved” or have
recently been brought into question, with regards to a comprehensive understanding of the
attributes that determine satisfaction in a typical user environment.
Satisfaction is a concept that has appeared in many fields, such as evaluation and employee
satisfaction, patient satisfaction, and visitor satisfaction of sites, amongst others. However, this
concept has been fundamental to the marketing concept for over three decades; as the most
extensive use of satisfaction has been in literature concerned with customer satisfaction. Wilton
and Nicosia (1986) inform that several models of satisfaction have emerged over time in this
field (customer relations) and in others. Kim (1997) argues that the models developed to date
all view satisfaction as a consumer attitude in relation to the consumer’s belief and evaluation
about merchandise and buying behaviour. This broad use of behaviour demonstrates the
appealing validity of the concept and its utility in explaining the success of a range of
phenomena (Parker & Mathews, 2001).
Hence, Day (1980:593) asserts that “while everyone knows what satisfaction means, it clearly
does not mean the same thing to everyone”. Initial conceptualization of user’s satisfaction
views it as a “single variable which involves a single evaluative reaction from the users”
(Peyton et al., 2003:42), which may or may not be related to pre-evaluation concepts. Further
conceptualization of satisfaction, notes that “… satisfaction is a kind of stepping away from an
experience and evaluating it… One could have a pleasurable experience that caused
dissatisfaction because even though it was pleasurable, it was not as pleasurable as it was
supposed to be; so satisfaction is not an emotion, it is the evaluation of the emotion” (Hunt,
1977:39).
The most generally acknowledged conceptualization of the user satisfaction concept, is the
Expectancy Disconfirmation Theory (McQuitty, Finn and Wiley, 2000). Expectancy
Disconfirmation Theory was developed by Oliver (1980a), who proposed that a user’s
satisfaction level is a result of the difference between expected and perceived product
performance, and expectations as predictions of future performance. The inclusion of
expectations suggests that products satisfying high expectations are predicted to generate
greater user satisfaction than products that meet low expectations. Other researchers employ
perceived performance as an additional predictor of satisfaction (Churchill & Surprenant 1982;
30
Tse & Wilton 1988). Satisfaction (positive disconfirmation) is known to occur when a product
or service is better than expected. On the other hand, a performance worse than the expected
results is labeled as dissatisfaction (negative disconfirmation). In this theory, expectations
originate from beliefs about the level of performance that a product/service will provide, which
is the predictive meaning of the expectations concept. In comparison, Kotler (2000) defined
satisfaction as a person’s feeling of pleasure or disappointment resulting from comparing a
product’s perceived performance (or outcome) in relation to his or her expectations. Hoyer and
MacInnis (2001) promote that satisfaction can be associated with feelings of acceptance,
happiness, relief, excitement, and delight. Similarly, Hansemark and Albinsson (2004),
established that satisfaction is an overall customer attitude towards a service provider, or an
emotional reaction to the difference between what users anticipate and what they receive,
regarding the fulfillment of some need, goal or desire. While, in the case of a state-subsidized
housing scheme for the poor and the low-income groups’, where the government is responsible
for provision of the houses, it would mean the overall beneficiaries’ emotional reaction towards
the government. This is in comparison to the difference between what they had anticipated and
what they later received.
Churchill and Surprenant (1982:491) inform that “the vast majority of satisfaction studies have
used some variation of the disconfirmation model”. Oliver (1993) found that a variety of
scholar’s definitions of satisfaction are consistent with the Expectancy Disconfirmation Model;
Tse and Wilton (1998:204) reports that “it is generally agreed that satisfaction can be defined
as … the evaluation of the perceived discrepancy between prior expectation … and the actual
performance of the product”; thus concurring with the assertion by Iacobucci and Oston (1995)
that satisfaction is a function of the discrepancy between a users’ (beneficiaries’) prior
expectation and the perception regarding the product, for instance, an allocated subsidised
housing unit. From the above, it is clear that the study of satisfaction both in academic spheres
and in the real world is to understand determinants and process of users (beneficiaries)
evaluation. Hence, this shows the relevance of this study in determining the factors, which
influence residential satisfaction in the South African low-income housing subsidy scheme.
This is with a view to ensure that the beneficiaries of the state subsidised housing schemes are
satisfied with the product and that it will enhance their well-being.
Nonetheless, Parker and Mathews (2001) also defined satisfaction as a process of evaluation
between what was received and what was expected, which is the most widely adopted
31
description of satisfaction in most current literature. Satisfaction can also be viewed as an
outcome of a consumption activity or experience; which is also referred to as a process.
Currently, there are two principal clarifications of satisfaction within the literature as defined
by Parker and Mathews’ (2001): satisfaction as a process and satisfaction as an outcome.
However, these are complementary classifications as; often one depends on the other. Parker
and Mathews (2001:38) promote that when satisfaction is viewed as a process, the definition
concentrates on the “antecedents to satisfaction” rather than satisfaction itself. When viewed
as an outcome, it is perceived as a consumption activity or experience, which is moderated by
different variables. The current study conforms to the later conceptualization as the satisfaction
of residents’ in public housing in South Africa is said to be moderated by different variables,
which exacts influences on the occupants.
2.2.1 Approaches to the Study of Satisfaction
The theory of satisfaction has its origins in the Discrepancy Theory (Porter, 1961) and other
scholars have over the years, used some form of comparison to model satisfaction (Parker &
Matthews, 2001). A number of theoretical approaches have been developed to explain the
relationship between satisfaction or positive disconfirmation and dissatisfaction or negative
disconfirmation. According to Oliver (1980a), these approaches can be seen as variants of the
consistency theories and focus primarily on the nature of the user’s post-usage evaluation
process. Consistency Theory conceptualizes that when expectations and the actual product
performance do not match, the consumer will feel some degree of dissatisfaction (Peyton et al.,
2003:42). In order to discharge this dissatisfaction the user will make adjustments either in
expectations or in the perceptions of the product’s actual performance. This theory informed
Morris and Winter’s (1978) Mobility Theory of residential satisfaction.
Over the years, a number of authors have used some form of comparison to model satisfaction.
Some theoretical approaches which have been advanced amongst others, include: Assimilation
Theory, Contrast Theory, Assimilation-Contrast Theory, and Negativity Theory.
2.2.1.1 Assimilation Theory
Kurt Lewin (1890-1947) originally formulated the Theory of Cognitive Dissonance
(dissonance theory), which was further developed and refined by Festinger (1957), which
formed the basis of the Assimilation Theory (Smith, 2001). The Dissonance Theory posits that
32
the users of a particular product make some kind of cognitive comparison between expectations
about the product and the perceived product performance. Hence, if there is a discrepancy
between expectations and perceived product performance then dissonance or negative
disconfirmation arises. This view of the user post-usage evaluation was introduced into the
satisfaction literature in the form of Assimilation Theory by Anderson (1973). In his work on
consumer dissatisfaction, the effect of disconfirmed expectancy on perceived product
performance was extrapolated. According to Anderson (1973), consumers seek to avoid
dissatisfaction by adjusting perceptions about a given product to bring it more in line with
expectations. Consumers can also reduce the dissatisfaction resulting from a discrepancy
between expectations and product performance either by altering expectations so that they
coincide with perceived product performance or by raising the level of satisfaction by
minimizing the relative importance of the dissatisfaction experienced (Olson & Dover, 1979).
However, the Assimilation Theory has a number of weaknesses. First, the approach postulates
that there is a relationship between expectation and satisfaction but does not specify how
disconfirmation of an expectation leads to either satisfaction or dissatisfaction. Next, the theory
also postulates that users are inspired enough to adjust either their expectations or their
perceptions about the performance of the product (Forman, 1986). According to Peyton et al.
(2003:42) “if the user adjusts either expectations or perceptions about product performance
then dissatisfaction would not be an outcome of the post-usage evaluation process”. A number
of scholars such as Olson and Dover (1979) and Andrson (1973), have found that controlling
actual product performance can lead to a positive relationship between expectation and
satisfaction. Therefore, it would seem that dissatisfaction could never occur except if the
evaluative processes were to begin with negative user’s expectations (Bitner, 1987), which is
not always the case.
2.2.1.2 Contrast Theory
The Contrast Theory was first introduced by Hovland, Harvey and Sherif (1957; cited in Peyton
et al., 2003:43). However, Cardozo (1965) claims that the theory presents an alternative view
of the user post-usage evaluation process in contrast to the Assimilation Theory that
hypothesized that post-usage evaluation leads to outcomes in opposite predictions for the
effects of expectations on satisfaction. The Contrast Theory posits that consumers would
exaggerate any contrast between expectation and product evaluation. Dawes, Singer and
Lemons (1972) further define Contrast Theory as the propensity to magnify the discrepancy
33
between one’s own attitudes and the attitudes represented by opinion statements validated by
persons with opposing views. Whereas Assimilation Theory suggests that users will seek to
minimize the discrepancy between expectation and performance; Contrast Theory argues that
a surprise effect arises leading to the discrepancy being exaggerated (Peyton et al., 2003). This
theory was further developed into the Assimilation-Contrast theory by Anderson (1973).
2.2.1.3 Assimilation-Contrast Theory
The Assimilation-Contrast Model has been proposed as another way of explaining the
relationship amongst the variables in the Disconfirmation Model (Peyton et al., 2003). This
model is a combination of both the Assimilation and the Contrast Theories. This model
postulates that satisfaction is a function of the magnitude of the discrepancy between expected
and perceived performance. Generally, users of any product have “zones or latitudes of
acceptance or rejection with respect to their perceptions” (Peyton et al., 2003:43). As with
Assimilation Theory, the user will tend to adjust differences in perceptions about product
performance to bring it in line with prior expectations, but only if the discrepancy is relatively
small. Peyton et al. (2003:44) noted that when there is a “large discrepancy between
expectations and perceived performance, contrast effects occur and the consumer tends to
magnify the perceived difference”. However, it should be noted that most discrepancies with a
given product such as the case of subsidised low-income housing in South Africa are not just
magnified or exaggerated, but they are simply the true evaluation of the products (houses),
based on what the users’ have experienced. However, some evaluation can be an emotional
expression of the user’s judgement with regards to the functionality of the product. On the other
hand, whether assimilation or contrast occurs depends upon the perceived disparity between
expectations and actual product performance, which is both objective and subjective, based on
the user’s judgment.
The Assimilation-Contrast Theory argues that Cardozo’s (1965) attempt in the Assimilation
Theory at reconciling the two earlier theories was methodologically weak as informed by
Anderson (1973). Anderson asserts that users possess a noticeable difference in the
disconfirmation threshold. Further, Assimilation-Contrast Theory attempts to illustrate that
both the Assimilation and the Contrast Theory Models have applicability in the study of user
satisfaction. The approach makes it possible to “… hypothesize variables other than the
magnitude of the discrepancy that might also influence whether the assimilation effect or the
34
contrast effect would be observed… when product performance is difficult to judge,
expectations may dominate and assimilation effects will be observed… contrast effects would
result in high involvement circumstances. The strength of the expectations may also affect
whether assimilation or contrast effects are observed” (Bitner, 1987:13).
Attempts by researchers to empirically test the Assimilation-Contrast Model have met with
varied results. For instance, Anderson (1973) and Olson and Dover (1979) found some
evidence to support the assimilation theory approach. In debating both of these studies, Oliver
(1980b) argues that Anderson (1973) and Olson and Dover’s (1979) findings cannot be
accepted because they only measured expectations and supposed that there were perceptual
differences between disconfirmation and satisfaction. This criticism is of some significance
because most researchers do not actually measure satisfaction or dissatisfaction; instead,
researchers usually assumed that it is the perception of disconfirmation that leads to satisfaction
or dissatisfaction (Forman, 1986). In contrast, this present study measures satisfaction and
dissatisfaction through the exogenous variables as set out in the research aim, to determine
their influence in predicting residential satisfaction as being the dependent variable. The current
study also measures for residential satisfaction as an independent construct, to further
understand the outcome variables of being satisfied or dissatisfied. In contradiction to the
findings supporting the Assimilation Theory, Cadotte, Woodruff and Jenkins (1983) reported
negative correlation between expectation and disconfirmation. They therefore resolved that
satisfaction is truly an additive function of the two concepts. Further, Peyton et al. (2003),
asserts that uncertainty was created by the results of studies from the works of Oliver (1977a,
1977b, 1979), which found no relationship between expectation and disconfirmation.
Moreover, Olshavsky and Miller (1972) in their study on consumer expectations, product
performance and perceived product quality supported the Assimilation-Contrast Theory.
However, it has been found that if the discrepancy was too large to be assimilated, then the
contrast effect occured.
2.2.1.4 Negative Theory
Similar to the three previous models, Negativity Theory also has its foundations in the
disconfirmation process. This theory was introduced into the consumer satisfaction literature
by Anderson (1973). Negativity Theory postulates that when expectations are strongly held,
users will respond negatively to any disconfirmation. Therefore, dissatisfaction will occur if
35
perceived performance is less than expectations or if perceived performance surpasses
expectations (Anderson, 1973; Carlsmith & Aronson, 1963).
2.2.2 Further Approaches to the Study of Satisfaction
The most well-known descendent of the Discrepancy Theory is the Expectancy
Disconfirmation Paradigm (Oliver, 1981), which states that, if performance exceeds
expectations, users will be positively disconfirmed or satisfied. On the other hand, if
performance fails to meet expectations, users will be negatively disconfirmed or dissatisfied.
Positive disconfirmation leads to increased satisfaction, with negative disconfirmation having
the opposite effect, whilst zero disconfirmation occurs when performance matches
expectations (no effect on satisfaction). Kotler, Siew, Swee and Chin (1996), state that this is
because user’s expectations are formed on the basis of past experience, statements made by
friends and associates. Oliver (1989) and Mastura, Noor, Osman and Ramayah (2007),
proposed that expectations could be exceeded in two different ways:
1. The level of performance is within a normal range (product was better than
expected); and
2. The level of performance is surprisingly positive (one would not expect that the
product would have performed so well) and delights.
The Disconfirmation Paradigm, which is the most dominant theoretical paradigm used in many
satisfaction research studies, is also what the present study’s paradigm is based on. The
paradigm has its roots in social and applied psychology (Oliver 1977:23). Therefore, the
Disconfirmation Paradigm presents its satisfaction judgments in three ways: satisfaction,
higher satisfaction, and dissatisfaction. But for the present study, satisfaction judgment will be
related to satisfaction and dissatisfaction.
Harris (1998) informs that when performance is greater than the users’ expected level of
performance of the service, higher user satisfaction will result because the service performs
better than expected. Users’ dissatisfaction occurs when the performance is less than the users’
expected level of service, as the service performs poorer than the users’ expected level. A
confirmation of expectations, or zero disconfirmation, is considered a state of satisfaction. A
negative disconfirmation indicates that their expectations were not met and yields a state of
dissatisfaction. The Expectancy Disconfirmation Model not only explains satisfaction with
36
product performance, but also service satisfaction, as is the case of the government being
responsible for the provision of low-income houses for the poor. There has been strong support
for the Disconfirmation Paradigm as a measurement of satisfaction. However, Churchill and
Surprenant (1982) found some inconsistencies in their developed model whereby neither
disconfirmation nor expectations have any effect on user satisfaction with durable products.
Satisfaction, according to Churchill and Surprenant (1982) is determined exclusively by the
performance of the durable good. This again puts the burden of a genuine evaluation result of
a typical low-income building or any other building product in the hands of the users’; because
they are the ones that can determine if the building is durable in terms of how the different
aspects of the building helps to meet their needs.
In their review, Poisz and Van Grumbkow (1988) view satisfaction as a discrepancy between
the observed and the desired. This was found to have been consistent with the Value-Percept
Disparity Theory in the work of Westbrook and Reilly (1983), which was developed in
response to the problem that users could be satisfied by aspects for which expectations never
existed (Yi, 1990). The Value-Percept Theory views satisfaction as an emotional reaction
caused by a subjective evaluative process, which is the comparison of the ‘object’ to one’s
values rather than an expectation (Parker & Matthews, 2001). What users want is a no disparity
level between their values: needs, wants and desires, and the object of their evaluations. Recent
developments in this study include the concept of desire congruency (Spreng et al., 1996).
Besides discrepancy theories, Equity Theory has also been applied to model satisfaction
(Parker & Mathews, 2001). Equity Theory holds that individuals compare their input/output
ratios with those of others (Yi, 1990) and that the consumer will be satisfied if the net gain is
perceived to be fair. Also, Parker and Mathews (2001) inform that renewed attention has been
focused on the nature of satisfaction to establish concrete attributes that determine the tenet of
the present study in relation to the public subsidised housing scheme in South Africa.
Due to the wide variance in the nature and meaning of satisfaction, many associations and firms
are using different reference points as a standard to compare their own user’s satisfaction
figures. To resolve this, a number of organizationally harmonized national customer
satisfaction indices have been developed (Hackl & Westlund, 2000). For example, the
American Consumer Satisfaction Index (ACSI) and the European Customer Satisfaction Index
(ECSI) represent the two major customer satisfaction indices for the United States and the
European countries, respectively; from which the South African Satisfaction Index is based
37
(SAS Index). The American Customer Satisfaction Index (ACSI) in Fornell et al. (1996)
defines satisfaction as a weighted average of three survey ratings: perceived quality, perceived
value, and customer expectations. This index has been used to measure satisfaction in the
manufacturing/nondurables, manufacturing/durables, transportation, communications and
utilities, retails, finance and insurance, services, public administration industries, and even in
government establishments. Although the ACSI index has an accepted satisfaction evaluation
methodology, it has not been found suitable for the construction industry and it is the lowest
with the rating of government and public agencies (Jyh-Bin and Sheng-Chi, 2006). This is
because the evaluation result for customer satisfaction is highest for competitive products,
lower for competitive services and retailers, and lowest for government and public agencies.
Figure 2.1 presents the model used by ACSI to measure satisfaction with government agencies.
In the ACSI model, users’ expectations influence the evaluation of quality and predicts how
well the product or service will perform. Perceived quality in the model is the extent to which
a product or service meets the user’s expectation and this normally has the greatest impact on
user satisfaction. Lastly, satisfaction has an inverse relationship to customer complaints, which
is measured as the percentage of respondents who reported a problem with the measured
product or service within a specified time frame.
Another theoretical description of satisfaction is the GAP Analysis Model developed by
Parasuraman et al. (1985). This is also referred to as the ‘disconfirmation paradigm’ in
customer satisfaction literature. The main theme of the Gap Analysis Theory is the fact that
gaps between user expectations and user experiences lead to user dissatisfaction. Consequently,
measuring gaps is the first step in enhancing user or service satisfaction, which results, in a
better understanding of users’ or beneficiaries’ perceptions; which is important to an
establishment of policy performances. Aziam (2009) informs that the Gap Analysis Model is
used as a tool to narrow the gap between perceptions and reality, thus enhancing user’s
satisfaction. Parasuraman et al. (1985:42) posits that [housing satisfaction] “quality is a
comparison between expectations and performance” and reiterated service quality as “the
discrepancy between users’ expectations and perceptions” (Parasuraman et al., 1985:111).
They further developed a [service] quality model based on the Gap Analysis Theory, which
informed that the measurement of the product quality gap is attained in the same manner as
service quality gap (Brown & Plenert, 2006). From the above discussion on the
conceptualization of satisfaction, the present study is tend towards the Gap Theory in ideology.
38
However, the study conceptualized its own attributes in order to measure residential
satisfaction.
Figure 2.1: ACSI Model for Government Agencies
Source: The American Consumer Satisfaction Index (ACSI). The ACSI Model for Most
Government Agencies, http://www.theacsi.org/government/govt-model.html (ASCI, 2006)
2.3 ASSESSING RESIDENTIAL SATISFACTION
This section will review residential satisfaction research and bring up the theoretical framework
that has been generalised in housing and residential satisfaction studies to date.
2.3.1 What is a Theory?
Theories and concepts are tools used for human thinking. According to Lundequist (1999), a
theory is a system in which a number of concepts and propositions have been methodically
ordered. Nachmias and Nachmias (1997), state that scientific ideas are constructs representing
certain aspects of the empirical world and they are concerned with the how and why of the
empirical phenomena, not with what should be.
Generally, scientific theories help to explain and understand a phenomenon. According to
Lundequist (1999) and Nguluma (2003), the world is generally seen through concepts.
Lundequist (1999), further states that the world is interpreted by the way we see things and that
as changes occur in society, concepts also change. Further, concepts also change as relations
39
to one another change, and as our way of seeing the world changes. Nachmias and Nachmias
(1997) define concepts as constructs representing certain phenomenon from which a meaning
or way of seeing the world can be assessed, informing that scientists begin the process of
research by developing concepts as shorthand for describing the empirical world. This process
is fundamental for communication of new knowledge. Without sets of agreed upon concepts it
is difficult for scientists to communicate their findings. It has become a common notion that
without the understanding of theory, and conceptual frameworks, advances are unlikely in any
field of study. It is against this background that previous theoretical framework will be
reviewed relating to housing and residential satisfaction studies in order to inform the model
constructs for the present study.
2.3.2 Residential Satisfaction Research
Residential satisfaction describes an ‘end-state’, where an individual or household is satisfied
with the residential status they have attained. This section of the study reviews previous studies
of residential satisfaction, bringing together a synthesized collection of housing and non-
housing components that are related to the formation of housing satisfaction. This is because
residential satisfaction is not one constant experience or state; it is an outcome, perceived by
an individual or household, that their current housing status meets their needs and aspirations.
Because residential satisfaction is based upon perception, the determinant factors essential to
attain will certainly be different in each case. Influencing this perception are factors such as
expectation (as already discussed above), history, demographic characteristics, and the
employment situation amongst others. In addition, variables such as culture (Guney, 1997);
age/older adults (Taylor, 1995a); individuals with severe handicaps (Leder & Sayre, 1989);
low-income single-parent families (Bruin & Cook, 1977); assisted living (Kalymum, 1989);
life satisfaction (Amerigo, 1990), financial status (Bruin & Cook, 1977); homeownership
(Montero, 1991); neighbourhood and environs, house and neighbours (Amerigo & Aragones,
1997; Kim, 1997a); perceived atmosphere, apartment evaluation, maintenance and friends
nearby (Weidemann & Anderson, 1982) have all been found significant to the study of resident
satisfaction. Also, Potter et al. (2001) reports that gender variances may appear with factors
such as safety in the study of residential satisfaction.
The study of ‘satisfaction’ dates back to the 1940s and is currently used in many disciplines
such as housing, consumer satisfaction, marketing, landscape architecture and medical fields
40
(Potter et al., 2001) and is being dominated by social psychology scholars. Residential
satisfaction has been studied in a wide variety of housing settings, ranging from mini-suites or
small self-contained apartments (Sidjak, 1995), gate-guarded neighbourhoods (Carvalho,
1995), low income housing (Montero, 1991), owner-occupied homes (Oseland & Raw, 1996),
college residence halls (Amole, 2009; Davis & Roizen, 1970), high-rises buildings (Guney,
1997), and multifamily housing (Weidemann & Anderson, 1982).
The formation of residential satisfaction is not simply based upon freedom from dissatisfaction;
it is more complex (Lu, 1999). Residential dissatisfaction is a differently composed construct;
the causes of dissatisfaction are more likely to be a universal source of discontent for everyone
(Hourihan, 1984), whilst the sources of satisfaction are much more diverse. This is the case
with the beneficiaries of the South Africa low-income subsidy schemes, where there is a
general dissatisfaction with the initially provided 30 square meters housing unit situated in a
250 square meters of land; which have since been increased to a minimum of 40 square meters.
One of the main reasons of dissatisfaction was because the area estimation of the constructed
housing units never took into consideration the housing life cycle of the beneficiaries and other
dynamics that are relevant to the beneficiaries’ optimal usage of the housing units. This is
because a building’s success depends not only on how effective the building provides for the
setting of activities of daily living, but may also depend on the perceptions of its residents. To
understand building occupants’ satisfaction evaluation towards a product or a service, it is
believed that improvements could thus be found and allocated to the right places and in the
right direction (Yiping, 2005), which will thus enhance the efficacy of the production or service
provision.
Residential satisfaction research deals with the housing products’ in terms of consumer
satisfaction, and is aimed to inform housing policy, planning intervention and to enable future
housing development to perform better than previously done. Over the past decades, much
research has been conducted on residential satisfaction (Campbell et al.; 1976; Francescato et
al., 1987; Weidemann & Anderson, 1985). These studies evaluate housing environments and
the residents’. The residents’ satisfaction studies of their environment have had a tendency to
focus on: research techniques, methods, and specifics design/planning frameworks for a
specific site but not on a more general basis. Also, most housing satisfaction studies have used
a direct theoretical approach in relating to a person’s beliefs, perceptions, or affect on his or
her satisfaction with the housing environment. From this perspective, any belief, idea, or fact
41
is thus a potential predictor of residential satisfaction. However, this present study is geared
towards an integrated holistic view of the study of residential satisfaction in subsidised low-
income housing. Also, residential satisfaction has been discussed in numerous empirical
scholarships which study characteristics of the users (either cognitive or behavioural) or
characteristics of the environment, both physical and social (Amerigo & Aragones, 1990). But,
Wiesenfeld (1994) states that, very few scholars have organized these variables into a model
so as to study and analyse, as a guide, the relationships produced amongst them.
There are two general approaches to empirical research studies about residential satisfaction
according to the classification used by Weidemann and Anderson (1985). One approach is to
view residential satisfaction as a criterion of evaluation of residential quality. Amerigo and
Aragones (1990:47) argue methodologically that the studies that falls into this category are
characterized by their “treatment of satisfaction as a criterion variable and, therefore, as a
dependent variable”. The theoretical background guiding this type of research is displayed by
the work of Marans and Rodgers (1975); Galster and Hesser (1981) and Cutter (1982). A
second approach is to view residential satisfaction as a predictor of residential mobility. In this
case, residential satisfaction is considered as a predictor of behaviour and, therefore as an
independent variable. The theoretical model developed by Speare (1974) is a good example of
empirical research belonging to this second approach. The present research thesis emphasises
the former approach, where satisfaction is treated as a dependent variable.
Nevertheless, the concept of satisfaction has been used in at least four different ways: as a key
predictor of an individual’s perception of general quality of life (Campbell et al., 1976); as an
indicator of incipient residential mobility, and hence has altered housing demands and effected
neighbourhood change (Spear, 1974; Varady, 1983); also as an ad hoc evaluative measure for
judging the success of housing developments constructed by the private sector (Lansing et al.,
1970; Zehner, 1977), and by the public sector (Marans & Rodgers, 1975) and also to assess
residents’ perceptions of inadequacies in their current housing environment so as to direct
forthcoming private or public efforts to improve the status quo (Anderson et al., 1983; Craik
& Zube, 1975; Michelson, 1977; Sanoff & Sawhney, 1971).
Nevertheless, a broader view on residential satisfaction is provided by Campbell, Converse,
and Rodgers (1976). They perceived housing satisfaction as one of the predictor’s of life
experience, where satisfaction with that variable might contribute to a person’s quality of life.
42
Developing from this theory were the more specific studies of a resident’s housing satisfaction
and the development of theoretical models explaining the sources of satisfaction (Marans &
Rodger, 1975). Together with these studies of housing satisfaction as a predictor of a person’s
quality of life, satisfaction was seen as a criterion for evaluation of the housing environment.
The initial work by Francescato et al. (1979); Kim (1997b) investigated the concept of users’
satisfaction and focused primarily on a specific housing type - subsidised multi-family housing
units. Although it was recognised that there are many different perspectives for housing
evaluation, they viewed the occupant as the primary user and stated that the evaluation from
this perspective has been neglected; which is what the present thesis advances in the context of
the study area.
Weidemann and Anderson (1985) noted that past models developed by numerous researchers
reflect the use of both approaches and include affect, cognition, and behaviour. The two
approaches as stated above were first combined by Weidemann and Anderson (1985) based on
Fishbein and Ajzen’s (1975) model of reasoned actions that considers how attitudes reflect
beliefs and evaluations of residences (Amerigo, 1992; Kim & Anderson, 1997). Others have
also supported integrated approaches (Michelson, 1977). Francescato et al. (1989) proposed an
all-inclusive model of relationships among the environment, satisfaction, and behaviour.
Amerigo and Aragones (1990) attempting to understand how the residential environment, the
house, the neighbourhood, and neighbours related developed the systematic model of
residential satisfaction. However, the integrated model as proposed by Weidemann and
Anderson (1985) deals with the complicated nature of housing quality. The integrated model
embodies three basic components of housing quality evaluation. These are: objective attributes
of the physical environment; residents’ perception and beliefs regarding their housing quality;
and satisfaction with the housing environment. The integrated model according to Weidemann
and Anderson (1985) can serve as a framework for research on relationships that have not been
empirically tested. An integrated model can also be used to organize existing literature that
many feel is disjointed and unorganized. The current study investigates the organized
relationship between publicly funded beneficiaries dwelling units, neighbourhood features,
building quality and services provided by the government, which are the essential variable that
have all been measured in a majority of previous studies, together with the assessments of the
beneficiaries’ needs and expectations and meaningful participation in the housing development
process. These are all classified as the exogenous variables and their role in predicting overall
beneficiaries’ satisfaction classified as the endogenous variable.
43
Nonetheless, the theories of residential satisfaction are based on the notion that residential
satisfaction measures the difference between households’ actual and desired housing and
neighbourhood situations (Galster & Hesser, 1981). It is believed that households make their
judgment about residential conditions based on their needs and aspirations. Hence satisfaction
with their housing conditions indicates the absence of complaints as their needs meet their
aspirations. On the other hand, they are likely to feel dissatisfied if their housing and
neighbourhood do not meet their residential needs and aspirations. This notion informs the
inclusion of beneficiary’s participation, needs and expectation on the represented model to be
developed so that a holistic view of beneficiaries’ residential determinants can be studied in
the evaluation of their housing satisfaction.
There is considerable evidence in literature which shows that residential satisfaction is
influenced by a broad array of objective and subjective perceived conditions. Previous studies
on residential satisfaction have analysed many variables such as housing, neighbourhood and
users’ characteristics that affect housing satisfaction (Galster, 1987; Lu, 1999). A buildings’
physical and social features are also strongly related to housing satisfaction. Satisfaction with
neighbourhood features has also been noted to be an important factor of housing satisfaction.
These include neighbourhood facilities, such as schools, clinics, shops, community halls
amongst others. However, numerous studies as carried out in developing countries have mainly
analysed three main components of the identified residential satisfaction attributes such as the
dwelling units, facilities and services, and the neighbourhood. For instance, a study by Husna
and Nurizan (1987) found that the residents of low-cost flats in Malaysia were generally
satisfied with their housing conditions and environment. Amongst the predictive variables that
contribute to the overall housing satisfaction, neighbourhood satisfaction contributed the most.
In another study by Savasdisara et al. (1989) the residents in a private low-income housing unit
were generally satisfied with the dwelling units and the neighbourhood. In contrast, a study by
Ukoha and Beamish (1997) found that the residents in a public housing unit in Abuja, Nigeria
were dissatisfied with their overall housing situation but satisfied with the neighbourhood
facilities. Overall, the concept of residential satisfaction does not lie on the individual’s
dwelling characteristics alone, but it is a composite of the overall physical and social
components that makeup the housing system (Francescato et al., 1987). Hence, this thesis is
focused on observing the effect of other variables that have not yet been studied collectively
on how they influence residential satisfaction.
44
Residential satisfaction, as a measure, has been criticized by some scholars as being subjective
(Campbell et al., 1976). Others have acknowledged the criticism but informed that all measures
have limitations and satisfaction should not be dismissed as a measure because it is a useful
concept (Potter et al., 2001). Another criticism of satisfaction is that operational definitions
vary greatly because they are defined as cognitive, emotional, and/or conative (Anderson &
Weidemann, 1997). All evolving models tend to support the belief that satisfaction can and
does include all areas. Francescato, Weidemann and Anderson (1986) defined satisfaction as
an attitude and stated that satisfaction and responses to questions directed at measuring
satisfaction could be considered affective, cognitive, and conative. Therefore, as a construct,
resident satisfaction must be both conceived and interpreted as multifaceted, including
affective, cognitive and conative responses. Researchers have continued to explore this broad-
based approach (Anderson & Weidemann, 1997). However, social psychologists have
suggested that there are three general categories of responses to any socio-physical object: the
affective, the cognitive and the conative or behavioural (Rosenberg & Hovland, 1960), which
is in line with the conceptualization made by Francescato et al. (1986). According to Rapoport
(1977), this suggestion or environmental perception is also referred to as a process of knowing,
feeling and doing. Thus, these categories create a useful framework in understanding the
theoretical development of satisfaction research (Weidemann & Anderson, 1985; Francescato
et al., 1987). As a result, residential satisfaction is not only used as an indicator of housing
policy, but also as a predictor of housing mobility as many researchers has shown.
As previously discussed, satisfaction is a subjective response to an objective environment. As
such, measures of satisfaction have been met with criticism. The criticisms, as outlined by
Francescato et al. (1987:48), include:
1. a report that residential satisfaction tends to be uniformly high and therefore, cannot
be assumed to indicate the ‘true’ state of affairs;
2. that subjective measures of satisfaction do not correlate with objective measures of
context and behaviour, therefore they cannot be considered valid measures of the
objective reality;
3. satisfaction with an object varies, for the same individual or social group, with time
and with personal and social norms and expectations, thus it is too fickly an
indicator on which to base action;
45
4. satisfaction tending to be higher where there is a lower beneficiary awareness of
‘better’ alternatives, which was argued that it tends to reflect unenlightened
assessments on which policy and decisions should not be based; and
5. focusing on satisfaction rather than attacking ‘real’ problems may result in sub-
optimal environments.
Francescato et al. (1987) further provided answers to these criticisms and established that while
the criticisms point to limitations that should be taken into account when interpreting results,
they also seem to warrant using the construct of satisfaction (Potter et al., 2001). Additionally,
Campbell et al. (1976) concluded that exaggerated skepticism of subjective responses is not
warranted based on extensive consideration of (among others) the following:
1. the reliability and validity of measures; the comparison between objective and
subjective indicators of well-being;
2. the levels of reality of domains being assessed, and the analytical intentions (Anderson
& Weidemann, 1997).
Hence, it is important to be aware of these limitations. However, it is clear that they do not
prevent satisfaction from being a useful concept. This is because there are limitations to all
research investigations; for example, there are always limitations to the operationalization of
subjective concept. However, Kim (1997) states that the criticisms in residential satisfaction
research point out the need for research that addresses these criticisms, and illustrates the
impact on theoretical models, and then proposes a research direction with clear theoretical
foundation. This is one of the prime reasons why the present research was conceived.
To this end, the next section will review the common problems that are mostly raised in
residential satisfaction research and offer possible solutions which will guide the present
research. The present research will develop and extend the concept of housing satisfaction, by
which the study of housing satisfaction will be treated as a criterion of evaluation; meaning
that residential satisfaction will be treated as a dependent variable. This will thus help to
identify the most appropriate variables to be considered in the provision of housing for the low-
income group by the government in developing countries using South Africa as a case study.
46
2.3.3 Problems Raised in Residential Satisfaction Study
When attempts are made to empirically prove models of residential satisfaction and the
relationship between the individual and his/her environment, scholars have faced several kinds
of problems. The problems that are mostly faced by researchers are usually grouped into three
dimensions (Amerigo & Aragones, 1997:52). The first relates to the definition of the terms of
the interaction being studied, which is the residential environment. Most times, what is
understood by residential environment, is not clearly defined; or likely a clear assertion of how
has it been empirically defined is necessary. The second dimension according to Amerigo and
Aragones (1997:52) raises the problem of the interaction between the individual and his/her
residential environment. Since the interaction is a dynamic two-way constantly changing
process. Lastly, the above-mentioned problems, together with the social attractiveness inherent
in the term ‘satisfaction’, result in the difficulty of formulating this variable. In order to meet
the aim and objectives for this study, the above problems will be described, highlighting
important theoretical and methodological queries which have hindered research in this field,
and possible solutions offered, which will again guide the present research.
2.3.3.1 Definition of residential environment
According to Amerigo and Aragones (1997:47) numerous studies of residential satisfaction
have mostly been applied to the house and to its surrounding neighbourhood. Both have been
researched from two points of view namely:
1. physical, conforming to equipment and services; and
2. social, referring to the social linkages established both in communal areas of the
building and in the neighbourhood.
However, there seems to be a problem in trying to outline the physical boundaries of the house
and that of the neighbourhood. For example, when referring to a house, we should take into
account not only its private space, but also the semi-public spaces immediately surrounding it.
According to Rapoport (1977), the appropriate definition of these areas is very significant at
certain socio-economic levels due to the perceptions they involve. Rapoport further argues that
spatial perceptions may vary substantially as a function of variables, like social and cultural
status. The concept of neighbourhood on the other hand is even more confusing according to
Lee (1968) and Amerigo and Aragones (1997). Very few scholars have clearly define this
concept exactly to which physical area it involves, while most other scholars uses terms like
47
community, district, neighbourhood amongst others, without defining them specifically.
However, Marans and Rodger (1975) are one of the few exceptions that put forward clearly
differentiated levels within the residential environment. In their work on the understanding of
community satisfaction, they defined the environment as the intermediate zone between the
macro-neighbourhood and micro-neighbourhood, including a more-or-less large area near the
occupant’s house, and where relationships are formed with other people living in it. This means
that the individuals residing in a space and the physical objects they use are closely bound into
one unit, thus forming an outline (Amerigo & Aragones, 1997:53). According to this concept,
neighbourhood cannot be specifically defined; rather the concept should be referred to as a
personal category, which is what the residents’ consider it to be. Another significant
characteristic when trying to define the neighbourhood is given by the sense of belonging to it,
or identification with it. Hence, the present thesis presupposes that the definition of
neighbourhood does not refer to the geographical area which limits it, but rather to the
occupant’s perception and to their sense of belonging, as supported by Amerigo and Aragones
(1997:53). In this way, the neighbourhood does not have a fixed surface, but varies from one
occupant to another.
2.3.3.2 Interaction between the Resident’s and their Residential
Environment
Amerigo and Aragones (1997:54) theorize that the study of residential satisfaction is most
interesting when it is applied to residents of low-income housing, who cannot move away if
they are dissatisfied with their present residential environment. When the degree of residential
satisfaction is low in this type of residence, in most cases mental (cognitive) restructuring
occurs in the resident, which keeps them in equilibrium with the residential environment. This
interaction between the individual and the residential environment implies that all the
residential environment intervening elements are considered as part of a process, which greatly
complicates empirical treatment of conceptual models. This practice sets in motion internal
mechanisms which determine the evaluations, which will make occupants experience a higher
or lower degree of residential satisfaction. Marans and Rodgers (1975) described the internal
mechanism as a standard of comparison when assessing the residential environment. This
notion denotes issues such as expectations, levels of aspirations, degree of equity
(participation), reference norms, needs and values. However, Marans and Rodgers (1975) limit
their study to enumerating the elements that form the standard of comparison, making no
48
reference to the meanings of such elements, or to the actual role played by each of them in the
evaluation of residential environment.
Likewise, Amerigo (1990) on the description of the theoretical model of residential satisfaction
refers to a similar element, which is essential when transforming the objective characteristics
of the residential environment into subjective ones, since it is the latter which will primarily
define the degree of residential satisfaction experienced by the individual. This element
according to Amerigo and Aragones (1997:55) is called standard of residential quality, which
is determined, amongst other things, by the occupant’s norm of reference. For example, having
an indoor toilet would be the maximum aspiration in housing quality for certain cultures or
certain socio-economic levels; whilst for others, it will have a much lower place on the scale.
Hence, it is assumed that each occupant has a specific standard of residential quality, by which
comparisons are made with the actual environment, so that as the gap between both decreases,
satisfaction with the actual residential environment increases. This cognitive view is formed
on the basis of the background of the occupant or household whose residential satisfaction is
being studied. Thus, this element implicitly includes individual, social and cultural influences
on the interaction with the residential environment.
Amerigo (1990) argues that the empirical demonstration of these internal processes in the
assessment of the residential environment has not always been effectively clarified, given the
complexity of interrelations posed by the processes themselves. Although, in Galster and
Hesser’s (1981) research on the model of residential satisfaction, and the works of Lindberg et
al. (1987, 1988) on residential preferences, both have attempted to resolve the complexity of
the interrelations. The work of Lindberg et al. (1988) on the model of housing preferences
which was based on the Fishbein and Ajzen’s (1975) theory of ‘reasoned action’, empirically
support the premise that the evaluations people make of a series of housing characteristics are
defined by their structure of essential values and beliefs about the effects that certain types of
behaviour will have on achieving these values (Amerigo & Aragones, 1997).
Also, with the traditional approaches used for residential satisfaction data analysis, such as
multiple regression analysis; it is extremely difficult to prove empirically the intervention of
mechanisms, such as those already mentioned. It has been found that it is not possible to derive
cause-effect relations from the technique of regression models, which bring out a serious
weakness for previous models of residential satisfaction, which try to determine the causes that
49
generate a certain degree of satisfaction and the effects they imply. Nevertheless, more and
more reliable statistical techniques for multi-variant analysis are being developed that allows
the extrapolation of causal relations between different variables. This considers the existence
of the independent variables, which increase the reliability and validity of the relations
established. For example, Galster and Hesser (1981) in their model of residential satisfaction
used Path Analysis to analyses the relation produced amongst the variables. Also, Lu (1999) in
his work on the determinants of residential satisfaction used the Ordered Logit Model, which
is seen as a more appropriate method than the widely-used regression models due to the ordinal
nature of the dependent variables. In order to overcome the above problems and thus, achieving
the desired cause-effect, this study employed the use of structural equation modeling to
determine the relationship between the exogenous variables and how they predict residential
satisfaction. This helped to overcome the weaknesses that the use of regression models posed
in residential satisfaction studies. Though, the techniques of analysis of structural equations
was first applied to an attitudinal model by Bentler and Speckart (1979).
2.3.4 Methodological Issues in the study of Residential Satisfaction
Despite sizeable literature and models that have been developed in the field of residential
satisfaction, it has not proven easy to quantify residential satisfaction empirically. According
to Amerigo and Aragones (1997:54), there are two problems associated with this: first, “social
desirability generated by direct questions of the type ‘to what extent are you satisfied with. . .
?’and secondly, the difficulty of determining ‘objective’ levels of residential satisfaction”.
With regards the first problem, there are many studies which obtain high levels of satisfaction,
not only with the residential environment, but with life in general, and with other domains of
life, when these are measured through items or scales, which ask the residents directly about
their degree of satisfaction (Argyle, 1987; Campbell et al., 1976). Amerigo and Aragones
(1997:54) offer a solution with the use of indirect scales since they do not ask directly about
the degree of satisfaction, but about the residents. They further argue that in addition to the
weaknesses of social desirability already mentioned, it may be noted that the word satisfaction
has general connotations of meaning referring to a global state of the occupant, more than to a
specific aspect of the residential environment, such as the neighbourhood, the house, or the
neighbours. This they posit may obviously influence the judgments of satisfaction referring to
these three components of the residential environment.
50
In contrast, the indirect scale has the weakness of its validity in comparison with the direct
scale, insofar as it is not certain that what is being measured is actually satisfaction and not
some other concept. The present research is oriented towards this measurement method and
will refer to specific aspect of the occupant’s residential environment and not to a global state
of satisfaction. This is because they offer more possibilities when aiming at a more valid
measurement of residential satisfaction which is what this study wants to attain.
The second problem, referring to the measurement of residential satisfaction, poses more
problems and questions the traditional definitions of the concept of satisfaction. Amerigo and
Aragones (1997:55) argues that if it is defined as the gap existing between achievements and
aspirations (Bardo & Hughey, 1984; Canter & Rees, 1982; Marans & Rodgers, 1975), how can
the situation of individual be explain who, despite having a wide gap between residential
achievements and aspirations, profess to be satisfied with their residential environment when
asked about it. They further clarify that the obvious answer is either that the question is
influenced by “social disability, as referred to above, or that these types of individuals have
reduced the dissonance that their objective residential conditions generate”. In either case, this
presumes that the outward expression of the degree of residential satisfaction experienced by
this type of individual varies obviously from the actual residential situation they are in, and is
determined by the gap that separates residential achievements from aspirations.
Consequently, the question of how to measure ‘objectively’ the degree of residential
satisfaction has been a problem for researchers. However, the question of an accurate
‘objective’ measurement of residential satisfaction was first empirically proven by Amerigo
(1990), who asserted that a sample of housewives living in council housing in Spain obtained
a higher degree of satisfaction when expressed directly, than when it was obtained through the
definition of satisfaction mentioned above. Specifically, the degree of satisfaction for the study
was obtained by calculating the distance between the perceived residential environment and
the ideal residential environment. Rationally, it was found that if this distance is small, the
person is more satisfied than if it is greater. Besides, when a profile was established of those
who claimed to be very satisfied, it was noted that those who lived in the areas which was
shortly going to be rebuilt, and their residential conditions were, consequently, very poor at
that moment (Amerigo, 1990). The high levels of residential satisfaction were achieved in cases
where the objective residential conditions were predicted. While the opposite leads to suspect
51
that the measurement strategies used up to the moment are inadequate. However, to overcome
these weaknesses, the use of indirect items and multi-term scales will be used in the current
study to help overcome this problem. To this end, the next section will discuss and review the
previous residential satisfaction models and conceptual frameworks, which will inform the
development of the holistic integrated conceptual framework for the present study.
2.3.5 Residential Satisfaction Conceptual Models
Previous models of residential satisfaction have been dedicated to three main components that
constitute the environmental interrelationship between man and his environment, that is, the
residents (the human part of the system) and the socio-physical environment and satisfaction
(the regulator of this complex relationship). In explaining the residents’ satisfaction, behaviour,
choice and mobility, Michelson (1977) came up with an integrated model.
2.3.5.1 Michelson’s Integrated Model
Michelson (1977:30) proposes an integrated model that explains residents’ mobility and
choice, user needs, and environment and behaviour. The major constructs of Michelson’s
model were: aspirations, primary demands, expectations, the physical settings, perception and
culture, and residents’ behaviour. The theoretical framework of the model focuses on residents’
satisfaction as a major determinant in residents’ mobility from their homes. In his model,
Michelson (1977) starts by supposing that residents have aspirations and primary demands in
interacting with their housing environment. Through residents’ expectations, residents’
aspirations and primary demands influences the physical and social characteristics of the
physical settings.
After residents’ experience the physical setting, an assessment occurs through perception and
culture, and spatial, social and psychological factors. Successively, this evaluation shapes
residents’ foreseen and unforeseen behaviour according to Michelson (1977:31). The resultant
behavioural pattern is a consequence of supportive or restrictive characteristics of the new
physical setting (Ahmad, 1994). Michelson’s model likewise assumes changes in the users’
primary needs, as a result of actual contact with the physical setting. These changes in the
users’ primary needs affect the evaluation of the housing environment. The users’ evaluation
may yield negative or positive perception followed by an action in the physical setting. Actions
related to negative perception, such as dissatisfaction, could be moving out of the
52
neighbourhood, altering the use of space or adopting physical means to change or modify the
design of the space. In addition, the model describes that the residents’ action may not depend
on the residents’ evaluation of their housing setting, but rather on the ability to achieve their
aspirations. In other words, a resident may not move because they cannot afford something
better or it is not available at all. Michelson (1977) further informs that the ability to achieve
aspirations may lead to negative evaluation of present homes in favour of new or better ones.
However, it can also be blocked by lack of affordance or absence of a better environment.
2.3.5.2 Onibokun ‘Habitability’ Model
Onibokun (1974) postulated that assessing habitability means evaluating the satisfaction of a
tenant living in a specific housing unit. This housing unit, according to Onibokun, would
normally be part of a housing project located within a particular community under some type
of institutional management. Onibokun (1974) emphasized that the housing habitability
systems usually involve four interacting subsystems, which include: the tenants subsystem, the
dwelling subsystem, the environment subsystem and the management subsystem. In the
Onibokun model, it was hypothesised that the adequacy of a housing unit, as determined by
the internal space, the structural quality, the household services, and the amenities and the
quality of the internal environment will impact the extent to which the resident is satisfied with
the unit. It was argued that the housing unit by itself is not the only variable or the only
determinant of housing need satisfaction. The unit subsystem according to the model is only a
part of the whole system, which constitutes housing habitability.
The habitability model thus emphasized that the variables that will affect the satisfaction level
with a housing unit are: tenant, external environment, management and dwelling variables. In
particular the model singled out the inhabitant as the recipient of all the feedbacks from the
subsystems and is therefore the central focus of the conceptual model of habitability on which
a study on housing habitability should be based (Falah et al., 1995). However, this concept
remains limited with respect to the real and complex situation of housing satisfaction.
2.3.5.3 Marans-Rodger Model
Another conceptual model of residential satisfaction is that developed by Marans and Rodger
(1975). The model conceptualized that an individual’s satisfaction with housing depends on
their perception of the various neighbourhood characteristics and their assessment thereof. The
53
neighbourhood attributes include several aspects of the physical environment and the quality
of local or community services. The Marans-Rodgers Model conceptualized that both the
perpetual evaluative process and the overall satisfaction level are related to the person’s own
characteristics, such as social class, housing status, amongst others. These socio-demographic
variables involve a smaller portion of residential satisfaction that does the assessment of
neighbourhood features. However, when personal characteristics were combined with
valuation variables as predictors of residential satisfaction, it was found that the former were
largely taken into account through the latter, and did not have much independent influence on
the level of satisfaction.
Nevertheless, in spite of adding new factors that will have an impact on satisfaction
(neighbourhood and community), Marans and Rodgers (1975) assessed personal characteristics
through the assessments of housing and neighbourhood attributes. The assessed variables were
found to be insufficient to fully assess personal characteristics. This limitation in the model
was what led to the development of the Path Analysis Model, which emphasizes the impact of
other significant variables which are neighbourhood and community variables. The Marans-
Rodgers (1975) model measured satisfaction with the community, the macro-neighbourhood,
and the micro-neighbourhood, and found that satisfaction with community related more to
social factors, while satisfaction with neighbourhood related more to physical factors.
2.3.5.4 Path Analysis Model
The Path Analysis Model as proposed by Hourihan (1984) hypothesizes that personal
characteristics are inter-related. The principle concern of the Path Analysis Model is the
relationship between the personal characteristics of residents and their levels of satisfaction.
The model specifies that residential satisfaction begins with residents’ personal characteristics.
These comprise the measure of social class, local social attachments, residential experience,
life cycle stages and housing type. For example housing type is dependent on social class, and
social attachment in turn, may well be related to housing type. Only social class and length of
residence were treated as being totally predetermined variables, and these would then influence
all other personal indices.
Neighbourhood attributes have also been found to have a direct contribution to housing
satisfaction. Attributes, such as safety, design of dwelling unit, stability and friendliness, were
54
found to form a fairly comprehensive profile of each resident’s perception and assessment of
the neighbourhood. The model confirms the importance of personal and neighbourhood
variables.
2.3.5.5 Housing Adjustment Model
The Normative (Housing Adjustment Model) model was first proposed by Morris and Winter
(1978). They introduced the notion of housing deficit to hypothesize residential
(dis)satisfaction. In their Housing Adjustment Model of residential mobility, they theorize that
individual’s judge their housing conditions according to normatively defined norms, which are
dictated by societal standards or rules for life conditions, and family/personal norms, which
amount to households’ own standard for housing. Thus, families evaluate their own residential
situation and that of others using definite culturally derivative benchmarks as norms. However,
a family whose housing does not meet these standards, experiences one or more deficits. The
housing adjustment theory contends that if a household’s current housing meets the norms, the
household is likely to express a high level of satisfaction with housing and neighbourhood.
Thus, an incongruity between the actual housing situation and the cultural and/or familiar
housing norms, results in a housing deficit, which in turn gives rise to residential
dissatisfaction. Households with a housing deficit who are dissatisfied are likely to consider
some form of housing adjustment to meet the known norm. The occupants’ needs are defined
in terms of both cultural and housing norms. Cultural norms are the standards by which the
behaviour or conditions experienced by members of a culture are evaluated as ‘good’ or ‘bad’.
Housing norms are standards related to the dwelling and its environment. They vary from
zoning regulations that specify, amongst other things, the minimum distance a house must be
set back from the street to very informal rules about having a quiet place to live in (Falah et al.,
1995:460).
The Normative Model emphasizes, in addition to the many variables common to the previous
models, the significance of culture which is a very important factor in satisfaction research and
in all research involving developing countries. But the factor (culture) was not considered as a
separate factor in the current study, since every subjective decision made by the occupants’ is
influenced by their cultural background. Although, the Normative Model postulates that a
standard (good or bad) should be set according to the cultural environment of each country.
Standards should be set in relation to local housing needs, which take into account cultural and
55
ethnic factors, rather than using some universal standards set in different countries. In a
developing nation, like South Africa, and in most developing countries where housing
standards for the low-income are fully regulated, it is essential to measure, as objectively as
possible, the physical quality of housing and its environment without predetermined ideas of
housing standards derived from the developed nations of the world. However, a number of
empirical studies have demonstrated that housing deficit is a useful notion in explaining
residential satisfaction and mobility behaviour (Bruin & Cook, 1997; Cook et al., 1994).
2.3.5.6 Francescato Model
In an attempt to understand the man-environment relationship, which was a question of an
understanding of the ‘users needs’ by the Design and Planning Professions during the 1960s;
Francescato et al. (1974) began to examine residents’ satisfaction with the housing
environment. In that study, issues identified were thought to be important to residents of
housing and developed self-report measures for these issues and for satisfaction. The
empirically derived causal model revealed a range of issues to be direct and indirect predictors
of housing satisfaction. The empirically derived causal model has been a prototype of later
housing satisfaction studies and has been used as framework by several housing evaluation
researchers. The Francescato model (1974) shows the multi-faceted character of the housing
environment. The model identifies important predictors of residents’ satisfaction with their
housing environment, such as safety, physical convenience, and social interaction. Some of the
predictors related to the physical environment, some to the social environment and some to the
housing environment. This brought about the notion that residential satisfaction can be
conceived in three levels, which are: the residential (physical) environment, the social
environment and housing (individual) environment or characteristics. The Francescato (1974)
Model assumes a direct functional relationship between satisfaction and each of the above
components, that is, residents’ characteristics and socio-physical components. Although the
model includes all the basic components for measuring satisfaction, residents’ behaviour and
values were not included in the model (Ahmad, 1994). The model also ignores the various
levels of the physical environment, such as the home, neighbourhood and city. Another
criticism of this model is that it focuses on satisfaction as an outcome of one side of the
equation, while the human behavioural aspects were completely ignored.
56
2.3.5.7 Weidemann and Anderson Model
Weidemann and Anderson (1985) seeking a more advanced understanding of resident housing
satisfaction, developed a conceptual framework for housing satisfaction by drawing on other
theories and models. This led to their development of a conceptual framework for the
Residential Satisfaction Model. The model was based on the concept found in Fishbein and
Ajzen’s (1975) general Theory of Reasoned Action. The conceptual framework makes explicit
several of the theoretical orientations and the assumptions that underline housing satisfaction
approach.
The model explicitly recognizes the causative role of the physical and social environment by
indicating these as categories of ‘objective attributes’ of the particular environment. The
objective environmental attributes have an influence upon a person’s satisfaction, which is
defined as an attitude and affect on the physical and social environment – through the person’s
perceptions and beliefs about those environmental attributes. In addition, this model recognizes
that the person’s affective attitude toward the environment influences the person’s intentions
to behave with respect towards the environment. Subsequently, the occupant’s intention to
behave has an influence upon behaviour related to the environment (Weidemann & Anderson,
1985). They therefore, propose an interpretation of satisfaction in purely affective terms,
informing that housing satisfaction is the subjective response to the dwelling, the positive or
negative feeling that the occupants have towards the place they live in. Hence, it is a global
representation of the affective response of people to the social-physical environment in which
they live.
The theoretical model explicitly includes personal and social information, which many social
researchers’ have neglected. However, the characteristics of the individual resident in relation
to personal and social attributes should be considered as potential predictors of housing
satisfaction. Therefore, with respect to housing satisfaction, inputs of resident’s characteristics
with personal and social attributes are important, and housing satisfaction cannot be properly
interpreted without them, which have been considered in the present study.
2.3.5.8 Marans and Sprecklemeyer ‘Inclusive’ Model
Marans and Sprecklemeyer (1981) further suggest a theoretical framework which attempts to
clarify the relationship between objective conditions, subjective experience and residential
57
satisfaction. The basic three components of the Marans and Sprecklemeyer Conceptual Model
are: the physical environment, the perception and attitude of the residents towards their housing
environment, and residents’ satisfaction. Marans and Sprecklemeyer (1981) in their study,
assumed a linear relationship between objective attributes of the physical environment, and
residents’ satisfaction. The model posits that satisfaction is a function of the physical
environment through one’s perception and beliefs of the physical environment. Hence, housing
satisfaction is a result of an integrated relationship between environment and the human
perception of beliefs. Additionally, the model assumes that human behaviour is a result of the
satisfactory or dissatisfactory outcome of the relationships produce amongst the variables.
Further to the above presentation of the previous models of residential satisfaction, Table 2.1
gives a summary of the previous models conceptualizations.
Table 2.1: Individual Conceptualization of Residential Satisfaction Models
Models Conceptualization
Francescato et al. (1979)
Francescato et al.’s Model of
Housing Satisfaction
The Francescato Model shows the multi-faceted character
of the housing environment. The model identifies important
predictors of residents’ satisfaction with their housing
environment, such as safety, physical convenience, and
social interaction. Some of the predictors relate to the
physical environment, some to the social environment and
some to the housing environment. This brought about the
notion that residential satisfaction can be conceived in three
levels, which are: the residential (physical) environment,
the social environment and housing (individual)
environment or characteristics.
Hourihan (1984)
Path Analysis Model
Personal characteristics are inter-related. The relationship
between the personal characteristics of residents and their
levels of satisfaction. The model specifies that residential
satisfaction begins with residents’ personal characteristics.
These comprise the measure of social class, local social
attachments, residential experience, life cycle stages and
housing type.
Marans and Rodger (1975)
Marans-Rodger Model
An individual’s satisfaction with housing depends on their
perception of the various neighbourhood characteristics
and their assessment of them.
Marans & Sprecklemeyer
(1981)
Inclusive Model (Basic
Conceptual Model)
This model posits that satisfaction is a function of the
physical environment through one’s perception and beliefs
they have of the physical environment. Hence, housing
satisfaction is as a result of an integrated relationship
between environment and the human perception of beliefs.
58
Michelson (1977)
Michelson’s Integrated
Model
Residents’ satisfaction as a major determinant in resident’s
mobility from their homes. Residents’ expectations,
residents’ aspirations and primary demands influences the
physical and social characteristics of the physical settings
Morris & Winter (1978)
Housing Adjustment Model
The notion of “housing deficit”. They theorize that
individual’s judge their housing conditions according to
normatively defined norms, which are dictated by societal
standards or rules for life condition, and family/personal
norms, which amount to households’ own standard for
housing.
Onibokun (1974)
Habitability Model
The adequacy of a housing unit is determine by the internal
space, the structural quality, the household services, the
amenities and quality of the internal environment impact
and the extent to which the resident’s is satisfied with the
unit.
Weidemann & Anderson
(1985)
Integrated Conceptual Model
They therefore proposed an interpretation of satisfaction in
purely affective terms, informing that housing satisfaction
is the “emotional response to the dwelling, the positive or
negative feeling that the occupants have for where they live.
Hence, it is a global representation of the affective response
of people to the socio-physical environment in which they
live.
The model explicitly recognizes the causative role of
physical and social environment by indicating these as
categories of ‘objective attributes’ of the particular
environment. The Model recognizes that the person’s
affective attitude toward the environment influences the
person’s intentions to behave with respect to the
environment.
Source: Author’s Literature review
2.3.6 Measuring Residential Satisfaction
From the conceptualized models discussed above, the work of Marans and Rodger (1975) and
Marans and Sprecklemeyer (1981) are the most comprehensive conceptual models of
residential satisfaction and many other studies have been based them. Also, all previous models
suggest a general sense of causality by moving left to right in their thematic diagrams with the
exception of the Marans and Sprecklemeyer’s (1981) Conceptual Model. Likewise, all models
suggested the indirect associations between objective attributes and satisfaction through other
components in the model, which was not in their model. Marans and Sprecklemeyer’s (1981)
model was the only one which stressed the direct effects of objective environmental attributes
on overall neighbourhood satisfaction. The direct association between these two variables is
considered in the current study. Both Marans and Sprecklemeyer’s (1981) and Marans and
59
Rodger’s (1975) models posited that residential satisfaction together with satisfaction with
other domains can influence the quality of life as an individual experiences it. Despite the
robustness of both models, they excluded beneficiary participation, needs and expectation as
significant factors which have been considered in the current model to develop a more robust
holistically integrated model of residential satisfaction.
Marans and Rodger’s (1975) model hypothesized that the resident’s satisfaction depends on
their perception of three domains (the dwelling unit, the neighbourhood and community
facilities). This is also a major hypothesis of the current study. In disparity, the Path Analysis
and the Onibokun ‘Habitability’ models specify that housing satisfaction is based on the
occupants’ personal characteristics. This thus informed that the central focus of their models is
the inhabitant. This is a limited assessment of the nature of housing satisfaction which in
practice has several other dimensions. In addition, the Marans-Rodgers (1975) Model further
postulates that overall satisfaction should be related to the person’s characteristics and other
variables. However, Hourihan (1984) inform that socio-demographic variables account for a
small proportion of housing satisfaction level, than do the assessments of the three domains of
measurement. The normative and Marans-Rodgers models emphasized that as well as looking
out from the inside, residential satisfaction should be evaluated by looking in from the outside
by measuring the eccentricities of actual environments from norms or standards (Falah et. al.,
1995, Michelson, 1977).
Studying these previous satisfaction models, leads to the conclusion that certain essential
variable that have an impact on residential satisfaction are contained within the following four
main domains:
1. Personal characteristics (socio-economic variables);
2. Dwelling unit;
3. Neighbourhood; and
4. Community services.
Because of the disagreement between researchers as to the relative importance of these four
different variables, the current research assesses the four variables and will also take into
consideration the impact of the two new variables (beneficiaries’ participation, needs and
expectations) that will be added. These domains include beneficiary’s participation after they
were shortlisted to being given houses, as well as the impact of considering the needs and the
60
expectations of the beneficiaries’. Therefore, because of the reasons stated above, this research
model will combine the essential variables of the residential environment together with that of
the newly added constructs in the assessment of low-income housing satisfaction.
2.3.7 Measuring Residential Quality and Adequacy (Satisfaction)
It is worth noting that previous theories of residential satisfaction all centre around the concept
that residential satisfaction measures the difference between households’ actual and desired (or
aspired-to) housing and neighbourhood characteristics (Galster, 1987; Galster & Hesser, 1981).
A significant issue in all models of residential satisfaction is how the housing attributes are
measured. There are commonly two types of measurements available, namely objective and
subjective measures of housing attributes found in literature (Francescato, 2002; Wiedemann
& Anderson, 1985). All the models discussed above, assess the level of housing satisfaction
subjectively, which is only one indicator of beneficiaries’ satisfaction as defined already.
Consequently, an objective measure of housing adequacy is essential to complete the present
research model. Objective measures refer to the actual measurements, such as the presence, the
lack of, or quantities of attributes, whilst subjective measures refer to perceptions, emotions,
attitudes and intentions towards the housing attributes. The objective measures of the attributes
of housing satisfaction have been shown to be weaker predictors than the subjective measures
(Francescato et al., 1989; Wiedemann & Anderson, 1985). The main objective measuring
technique, which has been used in assessing housing satisfaction quality, was first adopted by
Morris et al. (1972) who classified three areas of housing quality:
1. Structural quality, which refers primarily to durability of the shell;
2. Service quality, which is concerned with the kinds of equipment, facilities and
conveniences, which the dwelling provides; and
3. The state of maintenance and caretaking.
The measure of quality used by Morris et al. (1972) consisted of 26 items that measured these
three identical areas (Djebarni & Al-Abed, 1998). The presence or absence of a particular
characteristic was used as the prime basis for the assignment of a score to the various items.
The resulting item score was summed up to provide a measure of housing quality. Morris et al.
(1972) further emphasized that the procedure was based on traditional scaling techniques.
However, the work of Duncan (1971) titled Measuring Housing Quality, presented a number
of different models for measuring housing quality in the United Kingdom and in America.
61
Though, these vary significantly in their origins and scope because of the cultural differences.
To incorporate a more flexible and practical approach, Duncan (1971) in his study developed
the Housing and Environment Defects Index (HEDI), which has been used in the Scottish
Development Department, and is similar to the work of Morris et al. (1972). Duncan (1971)
categorized three dimensions of housing quality which are:
1. Basis of the dwelling interior schedule;
2. Basis of the dwelling exterior schedule; and
3. Basis of the environment schedule.
The method used a weighting system to distinguish between defects of greater and lesser
significance so that the aggregate result can be meaningful and capable of comparison. The
rationale behind Duncan’s (1971) method is that when some elements or condition in the
housing environment deviates from its set point or standard, a deficit results. This is what the
general principle of housing satisfaction is based on in all the models reviewed previously. This
technique was found to be flexible for measuring quality under different housing standards.
Because of the setting, the specifics indices developed by Duncan (1971) are equally applicable
to developing countries. Duncan’s (1971) method of measuring the quality of housing schemes
was used in the current research with the necessary adaptations to fit the South African
conditions in measuring the quality of the dwelling units of the low-income housing schemes.
However, the weight scoring system was not employed, as the presence and absence of vital
housing variables were only assessed in the present study.
Also, it has become common, in measuring residential satisfaction, to use an index of highly
correlated items rather than a single-item variable of how satisfied you are with your housing
unit; which is insufficient to illustrate satisfaction as a multifaceted constructs. Francescato et
al. (1986) addressed this issue in two ways. First, they proposed a list or index of four questions
reflecting overall satisfaction with the housing unit:
1. How satisfied are you with living here?
2. How long do you want to live in this housing development?
3. If you move again, would you like to live in another place like this?
4. Would you recommend this place to one of your friends if they were looking for a
place to live?
62
Secondly, recognizing that interpretation of responses to questions could have affective,
cognitive, and conative components, they proposed that this index of satisfaction would best
be described if many questions reflecting this multiplicity were used in the evaluation. The
reason for the above proposals was conceptual, which has been adopted in the present thesis.
This system helps to increase the reliability of the criterion since an index is essentially better
than a single item (Carvalho et al., 1997; Wiedemann & Anderson, 1985). Based on the above,
the present study measured the constructs in the conceptual framework with the aid of the
domains objective and subjective characteristics.
2.3.8 Determinants of Residential Satisfaction
Most empirical studies of residential satisfaction have been based on one of the conceptual
frameworks discussed above. The emphasis of the studies have been to establish the effects of
various housing, neighbourhood, and household features on residential satisfaction (Bonnes,
Bonaiuto & Ercolani., 1991; Lu, 1999). Various studies have analysed the correlates of
observed residential satisfaction for specific population groups, as already described above.
Also, a host of variables representing housing and neighbourhood characteristics, individuals’
socio-demographic attributes, as well as their perceptions of housing and neighbourhood
conditions have also been analysed in previous studies. They vary from the home size to
personal collections. According to Lu (1999), differences in model specification and data type
could also prevent a direct comparison of the empirical results. Two features that have arisen
from empirical studies on residential satisfaction are: firstly, Weidemann and Anderson (1982)
promotes that the significance of the perceptual variables have been emphasized because of the
belief that what is important in determining individuals’ residential satisfaction is their
perception rather than the actual structure of the residential surroundings. In other words,
objective measures of housing and neighbourhood attributes alone do not provide an adequate
explanation of satisfaction. Secondly, Crull et al. (1991) states that dwelling satisfaction and
neighbourhood satisfaction are measured differently and are often analysed separately. But, it
must be noted that the two types of satisfaction are also interrelated. This is because the
assessment of one’s housing unit, for example, is likely to include its immediate surroundings,
even one’s relationship with neighbours (Lu, 1999).
Overall, empirical studies have identified a number of important determinants of residential
satisfaction, such as income, tenure, life cycle stages, house size, and housing quality. For
63
instance, being older, having a higher income, having a smaller family have been related to
more housing satisfaction (Campbell et al., 1976; Galster & Hesser, 1981; Morris & Winter,
1978). Homeownership, particularly owners of single family homes, are almost always more
satisfied with their homes and neighbourhood than are renters (Rohe & Basolo, 1997). Also,
available space in the house has a significant positive effect on residents’ dwelling satisfaction.
Neighbourhood satisfaction has also been found to be an important predictor of dwelling
satisfaction; but there are also inconsistent and conflicting results in the literature on several
variables. Onibokun (1974) argued that the residential satisfaction (habitability) of a house is
influenced not only by the engineering elements, but also by social, behavioural, cultural, and
other elements in the entire socio-environmental system. The house, as informed by Onibokun
(1974), is only one link in a chain of factors that determine beneficiaries’ relative satisfaction
with their accommodation.
According to the literature, housing characteristics, neighbourhood characteristics, and
household characteristics have also, been viewed as the essential determinants of residential
satisfaction (Amerigo & Aragones 1997; Galster & Hesser 1981; Lu 1999). Housing
characteristics include the age of houses (He, 2009), interior and proximal exterior
environments (Phillips, Siu, & Yeh 2005), and other aspects of housing, such as, building
quality and disrepair (Amerigo & Aragones, 1990; Paris & Kangari 2006). Sirgy and Cornwell
(2002) also identified neighbourhood, social, economic, and physical features as the major
determinants of residential satisfaction. The social features most often regarded as important
include interaction with neighbours, attachment to the communities, perceptions of privacy,
safety at home, and others (Bruin & Cook 1997; Weidemanh & Anderson, 1982).
Neighbourhood socio-economic status and home values, and community cost of living are
factors used to measure the economic features of the neighbourhood (He, 2009; Lu 1999).
Physical features are other infrastructural and equipment settings, and these regard the quality
of the environment of the community, such as lighting of streets as informed by Dahmann
(1983), crowding and noise level, hypothesized by Gomez-Jacinto and Hombrados-Mendieta
(2002) and Bonnes et al. (1991), and green area or open space by Turner (2005) .
Largely, the concept of housing does not lie in the individual’s dwelling alone. It is a composite
of the overall physical and social components that makeup the housing system (Francescato et
al., 1987). Furthermore, housing satisfaction is influenced by the numerous components in the
system and the background characteristics of the occupants. Other factors that have been found
64
related to housing satisfaction, include: marital status (Tan & Hamzah, 1979), number of
children and family size (Miller & Crader, 1979; cited in Theodori, 2001), socio-economic
status, education, employment and welfare (Varady et al., 2001), housing physical
characteristics of the house (Yeh, 1972), satisfaction with the housing physical condition of the
house and management services (Varady & Carrozza, 2000), social participation and
interaction (Mohd Zulfa, 2000) and past living conditions, as well as residential mobility and
future intention to move (Morshidi et al., 1999).
However, there is little agreement on the effect of these factors on residential satisfaction. Lu
(1999:265) argues that the inconsistent in empirical findings may be attributed to the fact that
such key variables such as residential satisfaction are often defined differently in different
studies as well as the inappropriate statistical techniques that was employed in measuring the
determinants. The most widely used method in the previous studies has been Multiple
Regression Analysis, with levels of residential satisfaction being the dependent variable.
Because residential satisfaction is usually measured at the ordinal level even though it actually
has an interval scale, regression models have been found not to be appropriate. In the current
study, Structural Equation Modeling (SEM) utilizing EQS software was used in measuring the
variables, which helped to overcome the limitations of other measurement techniques that have
so far been used. This is because SEM is a statistical methodology that takes a confirmatory
approach rather than an exploratory approach, to the analysis of a structural theory, which
represents causal processes that generate observations on multiple variables (Bentler, 1988).
2.4 CONCLUSION
From the literature reviewed, one of the most outstanding findings in this chapter revealed that
the work of Marans & Rodger (1975) and Marans & Sprecklemeyer (1981) are the most
comprehensive conceptual models of residential satisfaction and as such many studies on
residential satisfaction have been based on them. Also revealed by the review of literature is
the important fact that previous theories of residential satisfaction all centre upon the concept
of residential satisfaction measuring the difference between households’ actual and desired (or
aspired-to) housing and neighbourhood characteristics. This is because a significant issue in all
models of residential satisfaction is the actual measurement of housing attributes. The literature
further reveals that there are two common types of measurement available, namely objective
and subjective measures. All previous models have all tended to assess the level of housing
65
satisfaction subjectively, which is only one indicator of beneficiaries’ satisfaction as defined
already. Hence, the current study hopes to fill this gap where residential satisfaction will be
measured both objectively and subjectively because objective measure of housing adequacy is
essential to complete the present research model.
Further findings from the reviewed literature reveal that a lot of factors determine housing
satisfaction ranging from income, tenure, life cycle stages, house size, housing quality, being
older, higher income bracket, having a smaller family, age of houses, interior and proximal
exterior environment, building quality, disrepair, place attachment, perception of privacy,
safety at home, home values and homeownership amongst others. However, homeownership
is said to be a major determinant of satisfaction within the owners of single family homes, than
renters, but not considered as a variable in the present study. Also, available space in the house
was found to be a factor that influences residents’ dwelling satisfaction. Neighbourhood
satisfaction has been found to be a major predictor of housing satisfaction; but at times, there
are inconsistencies and conflicting results in the literature on several other variables that also
affect the occupants’ satisfaction. Likewise, housing satisfaction is said to be determined by
the social, economic, behavioural, cultural, physical features and other elements in the entire
socio-environmental system. This is because the house is only one link in a chain of factors
that determine beneficiaries’ relative satisfaction with their accommodation. Lastly, literature
reveals that the measures of satisfaction have been met with criticism. Though, it is important
to be aware of these limitations. However, it is clear that they do not preclude satisfaction from
being a useful concept, as there are limitations to all research investigations. The criticisms
point out the urgency for research that addresses them with a clear theoretical foundation. The
next chapter addresses the gaps that have been observed from the review of past residential
satisfaction models, which the current study is positioned to address.
66
CHAPTER THREE
GAPS IN RESIDENTIAL SATISFACTION RESEARCH
3.1 INTRODUCTION
This chapter addresses the gaps observed in residential satisfaction research, which have not
been evaluated as all-inclusive constructs in the previous models. Though, they were
mentioned in the discussion of the previous models, only as distinct variables. These gaps form
the additional new constructs for the current study’s conceptual framework. The identified
gaps are: beneficiaries’ needs and expectations, and participation in the housing development
process. These identified gaps are discussed in relation to how to achieve them in low-income
housing development.
3.2 GAPS IN RESIDENTIAL SATISFACTION CONCEPTUAL
FRAMEWORK
Using the conceptual frameworks of Marans and Rodger (1975) and Weidemann and Anderson
(1985) as examples, it becomes clear that most of the research findings relating to residential
satisfaction are done in the developed western countries. Studies on residential satisfaction
completed in developing countries mostly relate to public housing provided by private
developers. However, these studies have not adequately provided an overview of the concept
compared to those done in developed countries. Hence, the result of the studies conducted in
the developed countries application and relevance will not be consistent with those of the
developing countries, as a result of the peculiarity of the developing countries housing
situations. This section of the study identifies the gaps in the residential satisfaction conceptual
framework. Since the conceptual framework provides the perspectives from which problems
are highlighted, it is most likely that there are some gaps in the western conceptual framework
that have failed to capture the factors affecting residential satisfaction in South Africa and other
developing countries; and in public housing studies in general. This section attempts to address
the two gaps that have been identified, namely: lack of understanding of beneficiary’s needs
and expectations, and the lack of practical involvement of beneficiaries’ meaningful
participation prior to the housing development process and the eventual allocation of the units
to the beneficiaries. It is of importance to inform the reader that the houses being evaluated and
67
referred to in this study are subsidised government allocated housing units. These houses are
built by the government and allocated to the poor and low income groups within a certain
category, as recommended by the housing department allocation guidelines. The consideration
of these identified gaps is based on the notion that resident satisfaction cannot be achieved
without good internal (needs and expectations) and external (participation in the housing
process) assessment of the occupants. This is because residential satisfaction is not a simple,
single-track factor assessment; but a combination of numerous variables.
3.3 GAP ONE: UNDERSTANDING BENEFICIARY’S NEEDS AND
EXPECTATIONS
Before a conclusion can be reached on how residential satisfaction of subsidised low-income
housing beneficiaries in South Africa is formed, it is important to explore why there is a
contrast in residential satisfaction research findings and whether the existing theoretical
framework proposed by western researchers has some gaps that have not fully accommodate
the developing nation’s context.
As already established from the literature, residential satisfaction is an individual evaluation of
living conditions. Collectively in social research, individual’s evaluation has the power to
appraise the housing policy’s performance, and possibly predicting housing mobility. In order
to examine the difference in residential satisfaction research as a combination of residents’
individual housing evaluation, it is essential to explore it both independently and collectively
(Yiping, 2005). As a result, this calls for research on residential satisfaction across individual
and social perceptions, which studies this at both psychological and sociological level. So far,
most residential satisfaction studies have focused on certain social goals, such as assessing
housing policy as a measure and also forecasting housing mobility as a predictor. According to
Francescato et al. (1987), the supposition behind housing policy research is that higher
satisfaction levels are a good indication of the success of specific policies, programmes, or
designs like the South Africa housing subsidy scheme amongst others. Also, Speare (1974)
noted that for housing mobility research, mobility results from the increase in dissatisfaction
beyond a person’s threshold of tolerance level. It is assumed that satisfied residents choose to
stay, rather than move out without a thorough examination of other factors that could have
contributed to the none-mobility decision. In contrast to the previous uses of residential
satisfaction, the focus should be primarily to determine which components of the housing
68
system most strongly and consistently predict residential satisfaction, so that it can be used to
direct efforts in those directions (outcomes) and to those aspects in which an intervention is
likely to yield the most beneficial effect (Yiping, 2005) for the occupants of the houses, which
is the focus of the current research. The idea of aiming to intervene in some components of the
housing system have tilted most research focus towards the physical and social setting of the
housing system, away from the beneficiaries needs and expectations.
Apart from the above, some residential satisfaction conceptual frameworks have included
needs and expectations, as part of the factors researched. Bruin and Cook (1997), Greenberg
(1999) and Parkes et al. (2002) proved in their research that needs and expectation data is a
better determinant of residential satisfaction, as opposed to socio-demographic data. Some
proven and useful personal needs and expectation variables, according to Greenberg (1999),
include mistrust of authority, negative emotions and pessimism amongst others researched,
during the housing development process. Moreover, because there is a lack of psychological
understanding in most residential satisfaction research, the focus has been put on sociological
understanding alone (Yiping, 2005). However, Sundstrom et al. (1996) state that psychological
understanding in residential satisfaction research has increasingly incorporate situational and
contextual variables as evidence has shown, but still, fundamental exploration from a
psychological perspective is lagging behind in most conceptual frameworks. As a result, the
effort to understand the individual psychological differences in the residential satisfaction
evaluation process is necessary.
In addressing the gap of understanding personal need, Maslow’s (1970) needs theory will be
drawn upon, whilst for the gap of resident’s expectation, the expectancy disconfirmation
theoretical framework postulated by Oliver (1981), as discussed in the previous chapter, will
be drawn upon as well. As already discussed, research in residential satisfaction is a valid way
of assessing the overall performance of the housing system as a criterion (Francescato et al.
1987), and it is useful to explore the meaning of satisfaction from the residents’ perspective in
order to have a holistic view of the housing system. Studying satisfaction requires real
understanding of the individual needs and expectation; and according to the Marine-Webster
Dictionary, satisfaction means fulfillment of a need or want. Thus, an understanding of
occupants’ housing needs and expectation will be useful to investigate the gaps in previous
residential satisfaction research.
69
The work of Maslow (1970) on ‘Motivation and Personality’ conceptualized the well-known
Needs Hierarchy Theory. Maslow theorizes that basic human needs are organized into an order
of relatively importance. Maslow also believes that human needs arrange themselves in order
of ‘pre-potency’. The needs order theory further postulates that: the appearance of one need
usually rests on the prior satisfaction of another more pre-potent need. Also, no need or drive
can be treated as if it were isolated or discrete; thus every drive is related to the state of
satisfaction or dissatisfaction of other drive. It is believed that once a need is satisfied, it ceases
to motivate behaviour. But man being an “ever-wanting” creature; as soon as one need is
satisfied, another appears in its place. Thus ‘human needs’ form a hierarchy according to
Maslow’s (1970) theory, which includes the physiological needs, safety needs, need to belong
and love needs, esteem needs, and need for self-actualization. The Maslow’s Needs Theory
postulates that each need must be satisfied in turn, starting with the first, which deals with the
most obvious needs for survival. It is only when the lower order needs of physical and
emotional well-being are satisfied that a concern can be placed on the higher order needs of
influence and personal development. However, if the things that satisfy the lower order needs
are swept away, people are no longer concerned about the attainment and satisfaction of the
higher order needs.
The Maslow’s Needs Order Theory has been extensively used in many fields, but, mainly in
the business and management domain to provide help on how to stimulate employees’
motivation. However, the needs order clearly set up the levels of order which individuals have
tried to pursue, as each level’s need satisfaction will motivate the desire for the next level. This
theory has also been tested to be useful in job satisfaction and customer satisfaction; and as
such it has the potential to help in addressing the gap in residential satisfaction research. Turner
(1972) using Maslow’s (1970) Need Theory suggested that material and existential needs and
priorities exist in housing choices, in which the existential functions includes identity, security,
and opportunity. Turner (1972) found that in any given context, housing priorities across
different income groups show difference in their vital need (Turner, 1972). However,
Greenberg (1999) attempted to combine Maslow’s (1970) Needs Order; but failed to explore
the full coverage of the theory and only indicated that controlling crime and physical problems
are the basic criterion for a satisfied neighbourhood. Though Greenberg (1999) was right in his
conceptualization, his attempt only touched upon the holistic understanding of human complex
needs in relation to their environment.
70
It must be recognised that a housing need exists, in which studies are mostly focused on
different levels of the need. The work of Taylor (1995b) claims that the fear of crime, or the
feeling of safety, is the dominant predictor of satisfaction. Also, Greenberg (1999) found that
crime and severe physical problems are the necessity need of any given neighbourhood.
Furthermore, to achieve a satisfied living environment, there exists the needs order in the
resident’s housing consumption (Yiping, 2005); but little attention has been paid on researching
this order. However the work of McCray and Day (1977) on housing satisfaction based on the
Maslow’s (1970) Theory of Needs, evaluated individual needs towards housing. McCray and
Day (1977) found that when housing needs are fulfilled, the individual will indirectly be
satisfied with his/her houses. Despite the general acceptance that housing is the primary
component for quality life, McCray and Day (1977) emphasize that housing construction rarely
refers to the needs and types of families who are going to inhabit the houses, whereas these
criteria are critical in the establishment of human habitats. But the emphasis in most housing
construction that involves the low-income is on the delivery at scale (quantity), which has
compromised the principles enacted in most housing edicts that are committed to the provision
of low income houses.
Figure 3.1: Housing Needs Order
Source: Yiping, 2005.
Personalised
Housi
ng C
hoic
e
The need for comfortable housing at acceptable
location, enough living space Physiological
The need for safe environment, free of eviction risk
and crime around Safety Needs
The need for being part of a group, social
contact with neighbours Social Needs
The need for prestigious location,
and being recognised Esteem Needs
The need to design,
develop and build houses Self-actualization
71
In another work of Maslow (1998) on ‘Towards a Psychology of Being’, Maslow posits that
housing order is constructed to understand factors that lead to a satisfied or dissatisfied
neighbourhood, which is a major ideology of the current research. Residential satisfaction is
determined by the fulfillment of individual housing needs, which is fundamentally determined
under the condition of what level of housing need is pursued. This is because unless one level
of needs is satisfied, they remain in the consciousness and become the prime determinant of
behaviour towards the neighbourhood. Also, the effects of satisfying higher order needs will
be neutralized if the lower-order needs are not fulfilled or when partially fulfilled. In other
words, the residents who do not have sufficient living space in their apartment will care little
about how the historic design of their building could affect their satisfaction. Figure 3.1 shows
the housing needs based on the Needs Hierarch Theory.
Physiological needs refer to the quality of the shelter provided by the housing unit and the
location the house is situated in. The location of the housing unit should be within acceptable
distance from the resident’s work place. Using the available transportation type, the daily
commute distance should be acceptable by a majority of the residents, and the house has to be
a permanent structure, with the necessary ventilation, acceptable climatic condition according
to the prevailing weather conditions in the particular place, access to water, and electricity
supply. All these services can be varied in different societies and geographical location, and
should be among the prerequisite norm in the housing selection process. Also, the size of the
housing units should be big enough to accommodate all family members, providing necessary
separation of rooms for different generation and different gender members.
When the minimum needs (physiological needs) for housing are satisfied, residents will start
pursuing the next level, which is the security need. According to Turner (1972), the security
need includes physical, emotional, and financial security. Residents would then want their
living condition to be secured with the appropriate structure and supply. For instance, the low-
income groups prefer a safe, orderly, predictable, lawful, and organized dwelling environment,
where they need to feel emotionally safe and secure. It means a safe environment free of crime
and risk of eviction. They want their properties to remain free from damage or theft. Yiping
(2005) stated that when crimes happen to a resident’s neighbours, they could generate a feeling
of insecure. The broader aspects of the attempts to seek safety and stability in the world are
seen in the common preference for familiar rather than unfamiliar things, or for the known
72
rather than the unknown (Yiping, 2005). When residents already have satisfied their minimum
living needs, and the fulfillment of their safety needs, they start considering the social needs.
The desire for social needs comes after the security needs. When the first two order needs are
fairly well satisfied, there will emerge the love and affection and belonging needs. Within the
living environment, these order needs will be typically exhibited towards family members and
neighbours. Residents expect affectionate relations with family members and look for friends
amongst their neighbours. It is the need to become part of the small society, a feeling of
belonging to a group and associating with it. This will enable the residents’ to participate in
activities and have contact with their neighbours. Especially for new comers who just move
into a neighbourhood, and who’s lower order needs are all satisfied, this level order need will
start developing. Moreover, the aspect of beneficiaries’ housing participation will only happen
when physiological and safety needs have been met. This need is followed by the esteem needs.
The esteem need is the desire for a stable, firmly based, usually high evaluation and regard
from others. It is the need for being recognized and respected by people around you. In the
living environment, it relates to the desire for reputation, prestige, authority and appreciation
from others. The form of esteem needs includes the desire for living in a neighbourhood with
a good reputation, and being respected by neighbours. Satisfaction of the esteem need leads to
the feeling of self-confidence, being useful and necessary within the society.
Self-actualization is a term that has been used in various psychological theories, often in
slightly different ways. The term was originally introduced by Goldstein (1934) in his work on
organismic theory for the motive to realize one’s full potential. In his view, it is the organism’s
master motive, the only real motive to actualize itself as fully as possible in the basic drive of
self-actualization. Similarly, Rogers (1951) further shed light on the concept through his work
‘On Becoming a Person’, informing the curative force in psychotherapy, that is man’s tendency
to actualize himself. This is the ability to become his inherent potential and to express and
activate all the capacities of the organism. However, the concept was brought fully to
prominence in Maslow’s Hierarchy of Needs Theory as the final level of psychological
development that can be achieved when all other basic and mental needs are fulfilled and the
‘actualization’ of the full personal potential takes place. When all of the above-mentioned
needs are satisfied; then, and only then, are the needs for self-actualization activated.
73
Maslow (1980) describes self-actualization as a person’s need to be and do that, which the
person was born to do. Self-actualization needs refer to man’s need for self-fulfillment, such
as to the tendency of man to actualize what is inherent in his/her potential. Within the housing
needs order, self-actualization refers to the desire of achieving a living environment true to the
residents own nature. Possible forms of self-actualization might include the personalization of
the interior design of the dwelling unit, and the customized designing or construction of his
own house. Nevertheless, a clear emergence of these needs usually rests upon some prior
satisfaction of the lower order needs.
The housing needs order is only a framework for general residents. It thus has different
meanings to different groups of people like children or older people; or even the low-income
groups that are the subject of this research. For instance, a lower-income resident in a
subsidised housing unit will not have the need for self-actualization, when the lower order
needs of good quality housing close to their place of work and with enough rooms to cater for
his/her needs and that of the family have not yet be met. However, the change and progression
of peoples’ needs as the society develops are clearly evident. For example, the South African
experience on evaluating quality of life has shown that the indicators of quality of life have
shifted from goals, which are basically a concern for food, shelter, needs for equity and racial
respect, to the need for participation, challenge and personal development.
In addressing the gap of resident’s expectations, the Expectancy Disconfirmation Theoretical
Framework postulated by Oliver (1981) was drawn upon, as already stated above. Research on
satisfaction using disconfirmation of expectations suggests that satisfaction is the result of a
comparison of that which was expected and that which was received (Woodruff, Cadotte, &
Jenkins, 1983). A fundamental premise of disconfirmation of expectation is that expectation is
related to satisfaction. Erevelles and Leavitt (1992) describe that post-purchase evaluation of a
product can be explained, at least in part, by a comparison of the pre-purchase performance.
Also, Spreng et al. (1996) extended the disconfirmation of expectations theory to include
desires by proposing a new model which integrates desires and expectations. Tse and Wilson
(1998:204) suggest that in addition to the influences from expected performance and subjective
disconfirmation, “perceived performance exerts direct influence on satisfaction”. Hence, the
Expectancy Disconfirmation Model claims that user’s satisfaction is a response to the
congruency between an individual's expectations and the actual performance of a product
(Oliver, 1981). Applied to the public housing subsidy scheme, satisfaction is viewed as a
74
function of the interrelationship between what beneficiaries expect from the government and
their perceptions of the housing unit they have received, that is, the quality of the houses
received and the satisfaction derived from the housing services provided.
According to Reisig and Chandek (2001), the expectancy disconfirmation model can be
conceptualized in a four-stage process. Firstly, the user formulates expectations regarding a
product. Expectations contrast across users (Tse & Wilson, 1988). For example, based on
occupant knowledge of the product, an individual may estimate what the performance will be
(Oliver, 1980). On the other hand, expectations may also be more normative in nature, and
thought of as what the user believes performance ought to be (Tse & Wilson, 1988; Woodruff
et al., 1983). Secondly, the individual makes certain attributions regarding the performance of
that product; and thirdly, compares the perception of the product’s performance against the
initial expectations. The last stage in the Expectancy Disconfirmation Process is the user’s
determination of how well the product measures up to initial expectations. Here, expectations
provide a standard from which to compare perceptions of product performance. Consequently,
the individual may judge product performance to be better than, worse than, or equal to what
he/she expected. The extent to which perceptions of performance match expectations dictates
the type of disconfirmation the occupant experiences, and has a direct effect on satisfaction
(Oliver, 1980). For example, an individual might experience positive disconfirmation, wherein
the expectations are exceeded (increases likelihood of satisfaction). Negative disconfirmation
is another probability, and arises when the user’s expectations are not met by the product or
service performance (decreases likelihood of satisfaction). Lastly, zero disconfirmation can
also occur when performance of the product matches expectations (no effect on satisfaction).
While disconfirmation is assumed to have a major effect on user satisfaction, research shows
that disconfirmation is not the only direct outcome (Reisig & Chandek, 2001). Reisig and
Chandek (2001) further claim that expectations have also been found to directly affect
satisfaction. For instance, individuals with lower expectations often report higher levels of
satisfaction. Similarly, the second component of disconfirmation, which is performance, has
also been interrelated to outcome satisfaction. Additionally, Oliver (1981) maintains that as
performance increases, so too do levels of user satisfaction. Expectations and performance,
therefore, are believed to have both direct and indirect effects on user’s satisfaction (Reisig &
Chandek, 2001).
75
The Expectancy Disconfirmation Theory not only explains satisfaction with product
expectation performance, but also service satisfaction (Churchill & Suprenant, 1982). For
example, Caughey et al. (1998) found that expectations have a significant effect on overall
satisfaction of occupants, as satisfaction normally occurs based on a comparison of that which
is expected, with that which is received. Also, prior exposure to what is to be received has the
tendency to influence occupant’s satisfaction towards a property. While a negative prior
experience can generate a lower expectation, which will result in lower satisfaction. So, an
understanding of how resident’s expectations are formed is significant in ascertaining how
beneficiaries’ satisfaction is ultimately formed. However, the dominant hypothesis guiding
recent research on satisfaction has been disconfirmation of expectations. Some researchers
have been challenging and expanding the disconfirmation theory and suggesting that many
other determinants also affect satisfaction apart from expectation (Ereveller & Leavitt, 1992;
Woodruff et al., 1983), which is a proposition that the current study advocates. Based on this
assertion, expectation together with needs is considered as one of the exogenous variables
included in the current research conceptual model to ascertain how occupants’ satisfaction in
subsidized low-income housing is formed in a developing country context, using South Africa
as a case study.
3.3.1 Satisfying Housing Needs and Expectations
As part of the conceptual framework of residential satisfaction research, the gratifications of
housing needs and expectations should have noteworthy prominence. For people with different
housing needs and expectations, the same housing condition could bring different satisfaction
levels because their needs and expectation are different. From the reviewed literature above,
residential satisfaction is basically formed under the condition of what the level of housing
needs are currently being pursued and the priori expectation as held by the beneficiaries. Unless
the level one needs are sufficiently satisfied, they will remain in the occupants’ consciousness
and will thus become the prime determinants of housing behaviour. The living condition that
is currently pursued forms the housing expectation of the individual, which influences the
overall residential satisfaction.
From the literature on housing research, from various perspectives, there have been many
studies separately addressing different levels of needs of individuals and social groups, or its
significance in informing policies on how best to deal with a need of a particular social group.
76
For instance, Marcus (1995) studies the self-actualization level and believes that housing is like
a mirror, which has a powerful effect on our sojourn toward a state of wholeness. Furthermore,
research on social needs in the housing environment has increased to such an extent that social
capital is the focus (Putnam, 1995). Social capital refers to social trust, norms and networks
that people can draw upon to resolve common problems, such as a housing problem. All over
the world, and in South Africa, there is growing agreement that social capital constitutes a
significant new dimension of community development and establishment, as occupants are
directly involved. This means that their needs and expectations would have been taken care of
through their active participation in the housing development process (Lang & Hornburg,
1998), which is the second gap addressed in this chapter. Furthermore, the security needs of
housing extend to another large area of research. For example, Newman (1972) addresses the
relationship between the built environment and security using his theory of defensible space.
Related to the security issue, there have been proposals and projects on urban renewals (Smith,
1996); debate on the gated community (Landman, 2004; Wilson-Doenges, 2000), and on social
issues of residential segregation (Hamnett, 2001). Housing needs, as a shelter, are mostly a
concerned of those who struggle for these needs, such as the homeless, or those previously
disadvantaged from owning property, as a result of the previous South African government
rule- defined as those constituting the low-income groups.
All this social research on housing can be grouped within a system relating to a different other
of needs. Individually, every household is inspired to pursue a higher level of needs in the
housing needs order, when the lower needs have been satisfied. Collectively, it brings out social
issues regarding the processes of different level of housing need satisfaction. Discrepancies in
housing priorities are so big that housing provision sectors have to provide a wide variety of
dwelling types with all forms of tenure to meet the demand. This is because residents are only
satisfied when their current housing needs and expectations are satisfied. However, it must be
noted that satisfaction will not stay unchanged, because soon, there will be other higher level
needs and expectation that will have to be satisfied.
More so, households who are dissatisfied are likely to consider some form of adjustment. They
may attempt to make adjustment to reduce dissatisfaction by revising their needs and
expectations to reconcile the incongruity, or by improving their housing conditions through
remodeling (Lu, 1999). According to Morris and Winter (1976), they may also move to another
place to bring their housing into conformity with their needs and expectations. Both mobility
77
and adjustments are subject to the constraints posed by financial resources at one’s disposal
and by the information given regarding alternative adaptation opportunities (Lu, 1999). Thus
moving behaviour is only one type of adjustment residents perform during the time of
dissatisfaction with housing needs; but in the case of the low-income groups, it might not be
possible, as most cannot access housing on their own and the subsidised houses received might
be their only life time opportunity of access to housing.
3.4 GAP TWO: UNDERSTANDING PARTICIPATION OF
BENEFICIARY
As already discussed, research in residential satisfaction is a valid way to assess the overall
performance of the housing system (Francescato et al. 1987). Hence, it is useful to explore the
meaning of satisfaction from the residents’ perspective in order to have a holistic view of the
housing system. Studying satisfaction requires the real understanding of beneficiaries’
meaningful participation, as housing issues affect an entire community or group of people who
in the present context are the low income and disadvantaged groups of the South African
society.
The beneficiary’s participation offers an opportunity to engage those who are affected by
housing issues in a dialogue; defining problems and creating solutions. The inclusion of
community stakeholders in the housing process helps ensure appropriate housing strategies and
policies are developed through more efficiently evaluation, development and implementation
to guarantee the satisfaction of the beneficiaries’. Inadequate beneficiaries’ participation in the
process can lead to community conflict or as a worst case scenario, anti-development initiatives
and ultimately housing dissatisfaction, which impacts on the quality of life of the final
beneficiaries. Successful beneficiaries’ participation is important because a mixed cross section
of the population that has a housing need can be involved in defining the housing problem and
in crafting community sensitive solutions. However, there is disagreement among planners and
professionals about the contribution of beneficiaries’ participation in improving the lives of the
people, particularly the poor and disadvantaged (Rifkin & Kangere, 2002:37). Some
completely dismiss its value altogether, while others believe that it is the “magic bullet”,
(Rifkin & Kangere, 2002:39), that will ensure improvements especially in the context of
poverty alleviation, and community ownership. The emphasis of this section is to provide an
overview of the policy framework, which informs participatory development in South Africa
78
together with the legislative and policy framework for participatory housing. Also the reasons
for engaging and encouraging beneficiaries’ participation, importance of beneficiaries’
participation and the advantages and disadvantages of beneficiaries’ participation in
determining housing satisfaction are discussed in this section.
Too frequently, development initiatives have been designed by those who have no real
knowledgeable understanding of the real needs of a specific community. Hence, most times,
the produced ‘housing plan’ is based on the different stakeholders’ perceived needs of the low-
income groups instead of the beneficiaries’ true needs (Davy, 2006:1). Kotze and Kellerman
(1997) attribute this to the fact that the idea that development consists of a transfer of skills or
information creates a role for the expert as the only person capable of facilitating the transfer
of these skills from them to the community or society. In order to create developmental efforts
that echo the real needs and expectations of specific groups, inclusive of development that will
satisfy the people, a paradigm shift is needed in the current conceptualization of residential
satisfaction research. This is a shift from the so-called blue-print approach to development
toward a more process and people-centered development that should produce beneficiaries’
participation. According to Oakley (1991) the role of beneficiaries’ participation in South
Africa cannot be undermined or may not override economic, personal or technological
aspirations in the South Africa public sector as the country’s past governance situation should
compel the government to correct injustices by actively involving the affected in policy
development.
In addressing the gap of understanding beneficiaries’ participation as identified, the Meyer &
Theron (2000:1) Building Blocks of Development will be used as criteria against which to
assess beneficiaries’ participative role in predicting housing satisfaction. The building blocks
include: beneficiaries’ participation, social learning, empowerment, sustainable development,
capacity building and self-reliance. Though, Meyer & Theron (2000:2) inform that all the
above-listed building blocks should be present for development to be considered a success,
which has also been recognised in the present thesis, however, beneficiaries’ participation and
empowerment will be given specific emphasis in this section with a brief mention of the other
building blocks during the dialogue. This is because beneficiaries’ participation has since been
envisaged as a diminution of the state’s involvement and a strengthening of the role of the civil
society, as a means to empower ordinary citizens, and the poor in particular, and to promote
more sustainable and satisfying form of development (Tapscott & Thompson, 2010:1-2).
79
Likewise the objectives of beneficiary participation as an active process are: empowering the
residents, building beneficiary capacity, increasing project effectiveness, improving project
efficiency, and sharing of project costs. These frameworks identify four levels of intensity of
participation: information sharing, consulting, decision making, and initiating action (Abbott,
1996). According to Thwala (2009:39) the framework has been largely accepted by
development agencies worldwide. However, a criticism of the model is that it is project-based
and does not include the full spectrum of community participation approaches. Also, the
seminal work of Arnstein’s (1969) Ladder of Participation will be drawn upon, inclusive of the
Burns et al. (1994) modified version of Arnstein’s Ladder of Participation, which was
conceptualized as the ladder of citizen empowerment. The Wilcox (1999) theory of the Ladder
of Participation will also be drawn upon concurrently.
3.4.1 Origin of Beneficiary Participation
Beneficiary participation in the public sector organisation has undergone a significant change.
Prior to this, people were more tolerant of poor service deliveries; more patient in long queues
and enduring inefficient public administration than they are now (Olivier, 2003:2). Nowadays,
people are expecting quality delivery of public services and are beginning to hold elected
representatives increasingly accountable, when their expectations are not met. Hence, the
origin of beneficiary participation can probably be traced to three root sources, which are:
participation as good development project practice (Abbot, 1996); participation as good
governance (Kooima, 1993) and participation as political empowerment (Bond, 2001; Freire,
2000). These concepts are discussed below in greater detail.
3.4.1.1 Participation as Good Development Project Practice
According to Rahnema (1992), participation was first used in the early 1950’s by social
activists and project field workers, as a necessary facet of development. Also, the World Bank,
and other international agencies, as well as the Development Bank of Southern Africa (DBSA)
have since taken the notion of participation as a requirement for successful project
implementation in the society. Hence, it has become a common practice to include some or
other form of public participation in the implementation of infrastructure projects within
development initiatives. A large amount of development work case studies tend to focus on
project specific participation and it is arguably the most well-known participation framework
of reference (Olivier, 2003:4).
80
3.4.1.2 Participation as Good Governance
The United Nations Development Programme (1998) defines governance as the exercise of
economic, political, and administrative authority to manage a country’s affairs at all levels and
the means by which states promote social cohesion and integration, to ensure the well-being of
their population. This entails all methods used to distribute power and manage public resources,
and the organizations that shape government and the execution of policy. Governance also
encompasses the mechanisms, processes and institutions, through which citizens and groups
articulate their interests, exercise their legal rights, meet their obligations and resolve their
differences. According to this definition, good governance therefore depends on public
participation to ensure that political, social and economic priorities are based on a broad
collective agreement and that the poorest and most vulnerable populations can directly
influence political decision making, particularly with respect to the allocation of development
resources. Good governance is also effective and equitable, and promotes the rule of law and
the transparency of institutions, officials, and transactions (UNDP, 1998).
Governance can be used in several contexts, such as corporate governance, international
governance, national governance and local governance. Participation within the framework of
good governance has its origins from within western democracies since the 1980’s and 90’s
Olivier, 2003). This was because falling voter turn-out and a general sense of disillusionment
with practices of the period, resulted in a rethink in the way civil society should be engaged.
The causes of oppressive, unresponsive and inefficient bureaucracies (Bennington, 1997) in
addition to a sense of powerlessness and marginalized local political structures within the state
brought about the idea of good governance to better serve the citizen.
Participation by the citizens of a state is a key cornerstone of good governance. Participation
could be either direct or through legitimate intermediate institutions or representatives.
However, it should be noted that representative democracy does not necessarily mean that the
concerns of the most vulnerable in society would be taken into consideration in decision
making, but it does create a platform for participation with the vulnerable in the society.
Participation needs to be informed and organized. This means freedom of association and
expression on the one hand and an organized civil society on the other hand. Good governance
has eight major characteristics. It is participatory, consensus oriented, accountable, transparent,
81
responsive, effective and efficient, equitable and inclusive and follows the rule of law. It
guarantees that corruption is curtailed, the views of minorities are taken into account and that
the voices of the most vulnerable in society are heard in decision-making process. It is also
responsive to the present and future needs of society.
Lastly, participation as good governance refers to a high quality of processes by which
decisions affecting public affairs are reached and implemented. This process ensures that all,
including the poor and other disadvantaged groups, are included and have the means to
influence the direction of development in particular as far as it affects their lives. Also, to make
contributions to development and have these recognised and to share in the benefits of
development and to improve their lives and livelihood. Participation as good governance helps
to ensure that all people have adequate access to basic services.
3.4.1.3 Participation as Political Empowerment
The empowerment approach to participation is located within the radical paradigm of
alternative development and is manifested in the mobilization of popular political power. This
originated from the economic development theory and theories of development. This approach
positions participation within a broader political struggle that links the condition of under-
development with access to political power (Freire, 2000). Originally, this tradition found
expression in popular resistance movements within South America, Asia and South Africa
(Bond, 2001).
These three approaches to participation sometimes intermingle and sometimes are confusing
in practical engagement between the government and communities. From the above, it should
be noted that there is no single universally applicable or perfect model of participation. It is
important to recognize different circumstances require a different style of participation from
authorities. However, the responsibility is to understand the context within which communities
are engaged, so as to design the most appropriate participative mechanism and process.
Participation in the current study is positioned on the good governance approach because good
governance depends on public participation to ensure that civil, societal and cost-effective
priorities are based on a broad collective agreement and that the poorest and most vulnerable
populations can directly influence political decision making, particularly with respect to the
allocation of development resources. Also, since the current study is based on South Africa,
82
where public participation is considered to be one of the key tenets of democratic governance,
the concept of participation as used, in this study, inclines towards this approach.
3.4.2 Beneficiary Participation Defined
Participation is a rich concept that varies with its application and definition. The way
participation is defined, depends on the context in which it occurs. For some scholars, it is a
matter of principle; for others, a matter of practice; for even more it is an end in itself as
described above. However, Rahnema (1992:116) informs that participation is a stereotyped
word, like children use Lego pieces. Like Lego pieces, the words fit arbitrarily together and
support the most fanciful constructions. They have no content, but do serve a function. As these
words are separate from any context, they are ideal for manipulative purposes. ‘Participation’
belongs to this category of word.
Most times, the term participation is modified with adjectives, resulting in terms such as
community participation, citizen participation, people’s participation, public participation,
popular participation or even beneficiary participation as used in the current study. However,
the Macmillan English Dictionary (2002:1032) defines participation as “to have a share in” or
“to take part in,” thereby emphasizing the rights of individuals and the choices that they make
in order to participate. Whilst, Arnstein (1969:216) claims that the idea of citizen participation
is a “little like eating spinach: no one is against it in principle because it is good for you”. But
there has been little analysis of the content of citizen participation, its definition, and its
relationship to social imperatives such as social structure, social interaction, and the social
context where it takes place. Bearing this in mind, the present study hopes to advance the
concept further by incorporating it into the definition and object of beneficiary satisfaction with
their housing unit. However, it can also be a method to co-opt dissent, a mechanism for
ensuring the receptivity, sensitivity, and even accountability of social services to the users, as
is the case of satisfaction with publicly provided houses by the South African government to
the poor and low-income groups.
Mathbor (2008) defined citizen participation as a process by which citizens’ act in response to
public concerns, voice their opinions about decisions that affect them, and take responsibility
for changes to the community. Likewise, citizens’ participation may also be referred to as a
response to the traditional sense of powerlessness felt by the general public when it comes to
83
influencing government decisions. This is because citizens often feel that housing development
issues are beyond their control because the decisions are made outside their community by
unknown bureaucrats and technocrats. Hence, Westergaard (1986:14) defined participation as
“collective efforts to increase and exercise control over resources and institutions on the part
of groups and movements of those hitherto excluded from control”. This definition points
towards a mechanism for ensuring community participation. Williams (2006) further informs
that beneficiary participation is the direct involvement of the citizenry in the affairs of planning,
governance and overall development programmes at local or grass roots level. Likewise,
Davidson et al. (2007:101) inform that it involves how and why members of a community are
brought into these affairs. Likewise, Meyer and Theron (200:1) inform that participation is a
social learning process linking the building blocks of development.
A vivid definition of participation programmes would indicate the involvement of a significant
number of persons in situations or actions that enhance their well-being, for example, their
housing, income, security, or self-esteem (Chowdhury, 1996). Chowdhury further states that
the ideal conditions contributing towards meaningful participation can be discussed from three
aspects which are:
1. What kind of participation is under consideration?
2. Who participates in it?
3. How does participation occur?
Mathbor (2008) also points out the importance of the following issues in order to assess the
extent of community participation:
1. Who participates?
2. What do people participate in?
3. Why do people participate? There are:
a) Cultural explanations (values, norms, and roles, etc.);
b) Cognitive explanations (verbal skills and knowledge about the organizations); and
c) Structural explanations (alternatives, resources available, and the nature of benefit
sought)
4. Implications (how the benefit contributes to the ends or principles they value).
The significance of beneficiary participation is said to draw from three main factors. Primarily,
it is alleged to allow for cost reduction through the utilization of local labour and expertise
84
(Davidson et al., 2007:102). Secondly, it potentially leads to the implementation of appropriate
responses through the involvement of locals in collective decision-making, through the
assessment of their needs and expectations, (Davidson et al., 2007:102) thus guaranteeing
housing satisfaction. Thirdly, it helps in directing scarce resources towards the more needy,
identified by fellow locals (Davidson et al., 2007:102; Mayavo, 2002). Beneficiary
participation is perceived as an undertaking that results in the empowerment of the local
population. However, it also has numerous non-benevolent political significances. It is referred
to as a curious element in the democratic decision-making process (Mcdowell, 1986). While
the roots of beneficiary participation can be traced to ancient Greece and colonial New
England, its significance reflects a contemporary recognition that societies are simply too
remote to be truly “of, by and for the people” without their involvement in the development
that affects them (Mcdowell, 1986).
Nevertheless, in principle, beneficiary participation requires the involvement of local actors in
the conceptualization, implementation, monitoring and evaluation of projects. In practice it
sometimes tends to be confined to specific activities (Mafukidze & Hoosen, 2009:7). As such,
beneficiary participation can further be referred to as local involvement within a continuum of
possibilities where locals may participate only as providers of labour, in decision-making or at
all levels (Davidson et al., 2006; Mafukidze & Hoosen, 2009:7). The level of local involvement
is most times conditional since there are no rules that prescribe the levels of involvement
(Lizarralde & Massyn, 2008). In some development, beneficiary participation could be
confined to the discussion of a proposed idea of building low-income houses. For instance, the
Reconstruction and Development Programme (RDP) capital subsidy low income housing
under consideration in this present study considers a minimal involvement from the local
population as most participatory process is simply aimed at bringing them together to endorse
an idea rather than to achieve empowerment, contribution of ideas and capacity building
(Hemson, 2007; Khan and Haupt, 2006).
Also, Jennings (2000:1) infers that participation refers to involvement by local populations in
the creation, content and conduct of a programme or policy designed to change their lives.
Beneficiary participation requires recognition and use of beneficiaries’ capacities and avoids
the imposition of priorities from the outside. It increases the odds that a programme will be on
target and its results will more likely be sustainable and satisfactory to meet the needs and
expectation of the beneficiaries. Ultimately, participatory development is driven by a belief in
85
the importance of entrusting citizens with the responsibility of shaping their own future.
Likewise, the World Bank Resource Book (1996:9) defined participation as a process through
which stakeholders influence and share control over development initiatives and the decisions
and resources which affect them. In this perspective, the benefits of participatory development
are perceived to be self-evident.
Furthermore, Stoker (1997) argues that participation is defined as members of the public taking
part in any of the processes of formulation, passage and implementation of public policies. This
is seen as a wide-ranging definition, which extends the emphasis of beneficiary participation
beyond the development of policy, to decision-making (outlining their needs and expectations,
and what is most important to them in the proposed housing units) and implementation. Meyer
& Theorn (2000:1) attempting to conceptualize participation as a people’s rights, defined
beneficiary participation as an active process by which the beneficiary influences the direction
and execution of a development project with a view to enhancing their well-being, in terms of
income, personal growth, self-reliance and other values they cherish, thereby guaranteeing their
housing satisfaction and eventually good quality of life with respect to housing development.
This definition helps to understand that the process does not deviate from the objective of
authentic and empowering beneficiary participation. Also, beneficiary participation and its
processes is being emphasized as a fundamental part of peoples’ rights to choose how they are
governed and how they, together with the governments, carry out the work of development
(Long, 2001).
To this end, Wates (2005) defined beneficiary participation as the act of being involved in
something, for instance housing development, amongst others. Habraken (2005) further posits
that participation has two definitions with opposite meanings. Habraken (2005) argues that
participation can either represent assigning certain decisive roles to the users, where they share
the decision-making responsibility with the professionals, in guiding them to design
satisfactory buildings. While the other type of participation, is where there is no shift of
responsibilities between the users and professionals but instead only the opinion of the user is
considered while making decisions. Beneficiary participation also means some form of
involvement of people, with similar needs and goals, in decisions affecting their lives. Abrams
(1971) also defined it as a theory that the local community should be given an active role in
programme and improvements directly affecting them. However, it should be noted that it is
rational to give control of affairs and decisions to people most affected by them. Moreover,
86
since no government or authority has the means to solve all the public problems adequately, it
is necessary to involve people, mostly the low-income groups / poor in matters that affect them,
because they might not have the opportunity to express their needs and expectations with
regards to the functionality of the housing unit that will be most suitable for them. However,
delegating powers to people to make decision concerning them when their financial
contribution is meaningless with a limited level of knowledge is not an easy task and involves
great inquiry into the change in the attitudes of the authorities and professionals (Davy, 2006).
Furthermore, Hamdi (1991) informs that beneficiary participation is a ‘powerful idea’, which
refers to the process by which professionals, families, community groups, government
officials, and others get together to work something out, preferably in a formal or informal
partnership.
Hauptmann (2001) emphasized the importance of beneficiary participation, arguing that
involvement gives people a better understanding of their own interests and the interests of
others, and, in some cases, brings them to see what would be best for the entire group. However,
this depends on the level at which beneficiaries are involved. Moote et al (1997) informs that
beneficiary participation facilitates decision implementation by resolving conflict issues during
the planning process, rather than delaying implementation of completed plans, whilst decisions
are reviewed through appeals and adjudication in some cases, while in other instances, it must
be accepted the way it is.
Beneficiary participation in housing delivery, and as used in this thesis, agrees with the
aforementioned definitions and can be summarized as a localized collective learning process.
This is where all stakeholders acquire and share information and learn to accept responsibility
for decisions, whilst working towards achieving the shared objective of improving housing
delivery. The definition acknowledges and tolerates the interest of different knowledge, pursuit
of cooperation and deliberate minimization of clashes along interest, knowledge and power
lines, which reinforces beneficiary participation. Thus, enabling the eventual satisfaction of the
beneficiaries with the subsidised houses being constructed and allocated to them by the
government.
87
3.4.3 The Legislative and Policy Framework for Participation in
South Africa
Since 1994, the South African government has put in place policy and legislative frameworks
that seek to promote participatory governance. The notion of beneficiary participation is
embedded in the South African Constitution. Recognizing the adverse impact of Apartheid on
the settlement of the majority of South African citizens, the incoming democratic government
in 1994, from the outset, placed emphasis on the provision of housing, as a basic human right.
The 1994 Housing White Paper asserted that the government was under a duty to take steps
and create conditions which will lead to an effective right to housing for all (Tapscott &
Thompson, 2010:4). It is alleged in South Africa that a person has a right to live in dignity, in
habitable conditions, and that government will vigorously promote an effective right to housing
for all, within the resources and other limitations applicable to it (Republic of South Africa
Constitution, 1996, Section 4.4.2). The principles of citizen participation was clearly
articulated in the Housing White Paper and further advanced in the Development Facilitation
Act of 1995, of which the policy goals were later given legal effect by the 1996 Constitution
(Tapscott & Thompson, 2010:4). Enshrined in the Constitution’s Bill of Rights (Section 26)
is the declaration that: “Everyone has the right to have access to adequate housing. The State
must take reasonable legislative and other measures within its available resources, to achieve
the progressive realisation of this right”. Following this edict, was a new National Housing Act
promulgated in 1997, committing the state, inter alia, to prioritize the needs of the poor in the
design and delivery of housing development programmes.
In harmony with the RDPs emphasis on beneficiary consultation, the 1994 Housing White
Paper committed the government “to a development process driven from within the
communities” (Section 4.4.4.), which would promote “the participation of affected
communities in the planning and implementation of new developments” (Section 4.5.1). This
viewpoint was also advanced in the 1997 National Housing Act which emphasizes, in Section
2(1) that national, provincial and local spheres of government must: “give priority to the needs
of the poor in respect of housing development; consult meaningfully with individual and
communities affected by housing development; ensure that housing development … is
administered in a transparent and equitable manner, and upholds the practice of good
governance”
88
The SA Government’s commitment to consultation, public participation transparency, and the
adherence to agreed norms and standards is further evident in the 2008 Social Housing Act
(Act No. 16 of 2008), which, in Section 2.1, states the need to: “consult with interested
individuals, communities and financial institutions in all phases of social housing development.
Facilitate the involvement of residents and key stakeholders through consultation, information
sharing, education, training and skills transfer, thereby empowering residents;...” (Department
of Housing, 2008).
Moreover, the South Africa constitutional requirements for beneficiary participation is found
in its mandate for local government, but more specifically in Chapter 10, Section 195, which
states that: “public administration must be development-oriented; people’s need must be
responded to, and the public must be encouraged to participate in policy making and good
human resource management and career development practices must be cultivated to maximize
human potential”.
Also, on a national level, the South Africa government introduced, what is commonly known
as the Batho Pele Principles, which are found in the White Paper on Transforming Public
Service Delivery (1997) and embodies the evolution of public participation in South Africa.
Batho Pele means ‘people first’. Through this principle, the government established the
importance of the South African public (citizens) and their valued input through participatory
means, and called “for a shift away from inward looking, bureaucratic systems, processes and
attitudes, and a search for new ways of working which put the needs of the public first, better
and sustainable development, which is faster, and more responsive to the citizen’s needs and
expectations” (White Paper on Transforming Public Service Delivery, 1997:2). The Batho Pele
concept is based on eight interrelated principles (Batho Pele Policy Review, 2003:14):
consultation in terms of quality of service received;
service standards should be indicated to the citizens to ascertain if it meets their needs
and expectations;
access to the services the citizenry are entitled to;
courtesy: a right to which each citizen is entitled to and as such beneficiaries’ should
be treated with consideration in the development that affect them;
information should be concise, accurate and about the service beneficiaries’ are entitled
to receive;
89
openness and transparency, so as to bring about greater accountability;
redress should occur if the pre-determined standards of service delivery are not met;
and
value for money, as the delivery of services should be done efficiently and effectively
to bring about satisfaction to the beneficiaries.
The ‘people first’ principle, which is a participatory bottom-up approach is derived from the
recognition that the total dependence on professionals (top-down approach) to implement
development initiatives is grossly inadequate and contributes to greater underdevelopment
(Oakley, 1991) as the needs and expectations of the citizens are in most cases not met by such
development. Because this creates a new level of underdevelopment, as the people end up in a
disgruntled situation that they have to live with until they are able to meet these needs by
themselves. Kotze and Kellerman (1997) state that the role and status of the technocrat and
technocratic top-down approaches contribute not only to the devaluation of the citizens’
indigenous knowledge and experience, but also to the side-tracking of the role of people’s
psychological and cognitive feeling in development. The top-down approach to development
has resulted in the deepening of the poverty cycle, greater underdevelopment, dissatisfaction
with housing development and other service delivery processes as development officials do not
implement participatory processes with the beneficiaries. The Batho Pele Principle advocates
for a bottom-up approach whereby the beneficiaries will have the opportunity to play an active
role in the decision-making processes which affect them. According to Oakley (1991), the
realization of the inadequacies of total dependence on a professionally dominant manner of
intervention has resulted in a search for alternative ways to bring about development, which
has led to the bottom-up approach to development which puts the people first and putting the
last first.
Nevertheless, whilst various legislative agenda and policy papers provide an enabling
framework for the delivery of public housing, none specified precisely how this is to be reached
by the different levels of government. With regards to the Constitution, as indicated, the
delivery of housing is a synchronized responsibility of all three levels of government (Tapscott
& Thompson, 2010:7). Following this model, the Housing Act ascribes responsibility to the
national government to determine provincial policy in respect of housing development.
Subsequently, provincial governments must accept responsibility for promoting the adoption
90
of provincial legislation to ensure effective housing delivery; and take all necessary steps to
support and strengthen the capacity of municipalities to effectively exercise their powers and
to perform their duties in respect of housing development (Tapscott & Thompson, 2010:8). In
the final instance, municipalities are vested with the responsibility of ensuring that housing is
delivered within the policy framework, as formulated by the national government and endorsed
by the provincial government. The placement and fragmentation of policy between different
levels of government has proven to be a major challenge to the democratic state since its
inception and this applies no less to the delivery of housing (Tapscott, 2000; Tapscott &
Thompson, 2010:8).
However, the National Housing Code (2009), prepared by the then Department of Housing
provides a framework of the procedures to be followed in implementing the National Housing
Act. The National Housing Code in Sections 2.4.1, 2.4.3 and 2.4.5, re-emphasizes the need for
participation, fairness and accountability in the development and allocation of public housing,
informing that: “the human settlement process will be participatory and decentralized allowing
effective response to priorities and opportunities at the local level and enabling all role players
to contribute their skills, labour, creativity, financial and other resources to the housing process.
Therefore, Government’s Human Settlement Policy must promote fairness and equity among
all South Africans and achieve equal and equitable access to housing opportunities, goods and
services. Transparency is seen as a key to guard against inequitable systems, in which some
segments of the population benefit more than others. Coupled with transparency, systems that
monitor progress and ensure accountability are equally important” (Department of Housing,
2009).
3.4.4 Beneficiary Participatory Process in South Africa
In the delivering of public housing projects and in giving effect to the rules of national policy
in South Africa, a municipality is mandated to work with community representatives through
what are called beneficiary committees (Tapscott & Thompson, 2010). Beneficiary committees
are understood to be elected by communities, and it is evident that they are established in
different ways by different municipalities and in some cases in different ways by the same
municipal authority, depending on the community dynamic, or nature of the project. As such
beneficiary committees cannot be expected to serve the communities they are purported to
represent in similar ways. For instant, in the Department of Human Settlement capital subsidy
91
projects, in which beneficiaries are selected from a general waiting list, this differs substantially
from a committee established in an, in situ, upgrade area (where shacks are replaced by houses
on site) where a sense of community is likely to be stronger. The effectiveness and legitimacy
of the beneficiary committees as seen by the representatives themselves is markedly different
between the different types of projects, in spite of the signed agreement between beneficiary
representatives, the municipality and the housing developer (Tapscott & Thompson, 2010).
In giving influence to the idea of people-centered development, the South Africa Housing Code
stresses the need for a structured agreement (also referred to as a ‘social compact’ or ‘contract’)
between a municipality and the community in the delivery of housing projects. This agreement
ensures that community members assume ownership of their own development and project.
The involvement of the beneficiaries from the onset is of vital importance. Hence, beneficiary
participation is undertaken within the context of a structured agreement between the
municipality and the community (National Housing Code, 2009).
When preparing a housing project for provincial government approval and funding, the
Housing Code stipulates that a municipality must submit a copy of a social compact, which
reflects the agreement of beneficiary groups and other stakeholders in the community on a
number of key issues relating to the project. Amongst the issues under consideration are the
following: the housing needs of the relevant community; the extent to which the housing
project will meet the housing needs of an identified target market with particular reference to
the appropriateness of the location; the number and type of residences to be constructed; the
full cost to the beneficiary if any; and the level of services to be provided (National Housing
Code, 2009). Unfortunately, this has not been held to occur as intended, as participation is often
and only interpreted to mean acquiescence and voluntary contributions of labour and resources
by the low-income beneficiaries who have no real influence on a projects’ goals and design or
in establishing the rules within which it must operate as intended (Hassen, 2003). If the social
compact agreement were created, there would be a state of total satisfaction by the beneficiaries
of all low-income houses in South Africa, as their needs and expectations would have been
considered during the conception stage of the project. But this is not the case as beneficiaries
are only made to endorse the projects being created without any concern for their needs and
expectations. Moreover, in most housing developments, no social compact groups are formed.
This ultimately leads to a state of total dissatisfaction with the houses being received, as only
the beneficiaries’ needs perceived by the experts are considered.
92
Although, the determination of norms and standards, in terms of the Housing Act, is the
responsibility of provincial governments, the South Africa Housing Code does not provide
explicit details on how the social compacts, should be drawn up, which is a major short coming
in giving the embedded concept of participation expression. In other words, the framework for
determining who should represent communities in drawing up social compacts, the content of
the agreement and the expected roles of those involved, is left to provincial governments to
decide or to delegate to local governments. In the Gauteng Province for example, the domain
for the empirical aspect of this study, the provincial government has left responsibility to the
municipalities to draw up social compacts. As a result, the interpretation and implementation
of participatory policies is left to the differential capacities of local housing officials and their
understandings of participatory development, which is highly inadequate and inconsistent.
Beneficiary participation is generally more successful when the community (‘beneficiaries’)
takes on much of the responsibility, as compared to situations in which the government attempt
to assess beneficiaries’ preferences for housing through surveys or meetings. In order for
beneficiary participation to work, projects must include special components that address it
directly. Beneficiaries should be recruited to help in all phases of designing, implementing,
maintaining, supervising, and evaluating a new housing construction, but only if the time,
effort, and money are spent to do it correctly (Thwala, 2009). Despite these constraints, when
the process is started early enough, this aspect will enhance the production of a housing product
that would have be specifically designed to meet the needs of the community in all aspects.
Also, special consideration must be given to the development of local committees and
governance structures to adequately oversee local participation. These local committees and
governing structures when developed will direct and execute development (housing) projects,
rather than merely receiving a share of project benefits.
3.4.5 Levels of Beneficiary Participation
Theories of citizen participation have received considerable academic attention particularly
since the early 1900’s, but have been a source of debate since at least the early 1960s. However,
the influential theoretical work on the subject of community participation was done by Arnstein
(1969). The precise importance of Arnstein’s work comes from the obvious recognition that
there are different levels of participation, from manipulation or therapy of citizens; through to
93
consultation, and to what we now view as genuine participation, that is the levels of partnership
and citizen control. The fundamental point in Arnstein’s model (is) that “participation without
redistribution of power is an empty and frustrating process for the powerless. It allows the
power holders to claim that all sides were considered, but makes it possible for only some of
those sides to benefit, thus maintaining the status quo” (Arnstein, 1969:217).
Arnstein (1969:217-219) describes the type of ‘non-participation’ represented by the lower two
rungs on the ladder as attempts to ‘educate’ participants. Levels 3 and 4 allow participants to
hear and have a voice, but they have no power to ensure that their voice has influence. At level
5 participants can advise, but the right to decide is retained by the agency. Arnstein (1969)
alludes that true participation begins where ‘partnerships’ enable negotiation and shared
decision-making responsibility. Arnstein considers that partnership working is most effective
when participants have an organized and resourced base from which to work, and to which
they are accountable. At levels 7 and 8 participants form the majority in decision-making
arenas, or hold managerial power.
Despite this conceptualization, the framework had some limitations. The limitations of
Arnstein’s (1969) framework are twofold. First, each of the steps represents a very broad level,
within which there are likely to be a wide range of experiences. For instant, at the level of
‘informing’ there could be significant differences in the type and quality of the information
being conveyed. Convincingly therefore, stages of participation are likely to reflect a
multifaceted level, than a simple series of steps. Secondly, the use of a ladder also implies that
more control is always better than less control. Nevertheless, increased control may not always
be wanted by the community and increased control without the necessary support may result
in failure; thus making the Arnstein (1969) theory a complex one. Also, there is a failure to
acknowledge the different spheres of decision-making in which the level of participation
occurs. However, the Arnstein’s (1969) Theory of Participation has since been modified and
new terminologies added. In particular, there has been a shift towards understanding
participation in terms of the empowerment of individuals and communities. This came from
the growing importance of the idea of the citizen as consumer, where choice amongst
alternatives is seen as a means of access to power. Within this framework, people are expected
to be responsible for themselves and should, therefore, be active in public service decision-
making, as this is where their needs and expectations could be tabled. In this framework, Burns
94
et al (1994) improved Arnstein’s Ladder of Participation and postulated a Ladder of Citizen
Empowerment as shown in Table 3.1.
Table 3.1: Ladder of citizen empowerment
CITIZEN CONTROL
12. Independent control
11. Entrusted control
CITIZEN PARTICIPATION
10. Delegated control
9. Partnership
8. Limited decentralized decision-making
7. Effective advisory boards
6. Genuine consultation
5. High quality information
CITIZEN NON-PARTICIPATION
4. Customer care
3. Poor information
2. Cynical consultation
1. Civic hype
Source: Burns et al., 1994.
The Burns’ theory of ‘Ladder of Citizen’s Empowerment’ is more elaborate than Arnstein’s
(1969) conceptualization because of the qualitative breakdown of the different levels. For
instance, a difference is drawn between ‘cynical’ and ‘genuine’ consultation, and between
‘entrusted’ and ‘independent’ citizen control. The phenomena of ‘civic hype’, which was
gradually recognised during the 1990s (Harvey, 1982), is integrated on the bottom rung of the
ladder. This essentially treats beneficiary participation as a marketing exercise, in which the
desired end result is ‘sold’ to the community.
Adopting the Ladder of Citizen Empowerment as a conceptual framework, Wilcox (1999)
developed a Ladder of Participation, in which five interconnected levels of beneficiary
participation were identified (Table 3.2). Wilcox’s work arose from the UK regeneration
context and reflects a philosophical progression in thought around participation. This model of
participation is what the current study advocates for in order for, the beneficiaries of South
Africa housing subsidy schemes to participate in the low-income housing developing projects
and ultimately have satisfaction with the houses that are to be provided.
95
Table 3.2: A ladder of participation
Information
Consultation
Deciding together
Acting together
Supporting individual
Community initiative
Source: Wilcox, 1999.
The adopted framework indicates that different ‘levels’ of participation are acceptable in
differing contexts and settings. The progression recognizes that even though power is not
always transferred in participative processes; the processes still have value. This was in contrast
to the common interpretation of Arnstein’s (1969) framework, which brings the thought that it
is only acceptable to be striving towards citizen control. Exclusive to some contexts, this shift
in philosophy has been further developed to describe levels of involvement as a continuum.
The Wilcox (1999) Framework provides useful insights into the scope of experiences
associated with beneficiary participation, which by their nature represent simplifications of a
more complex reality.
Wilcox’s (1999) model defines that you need to inform the beneficiary of what is planned so
that they are informed about what is happening. This principle is a major tenet in the South
African Batho Pele’s Principle and other legislative policy guiding participation in South
Africa, which is strongly aligned with the Africa Charter; which states that there must be
meaningful consultation with the beneficiaries of any development programme. The Africa
Charter further states that the government must be clear about the opinions they are asking
from the citizens. In the consultation stage, a number of options are provided and a careful
analysis of the resultant feedback is completed. From this point, a decision is reached taking
into account the results of consultation alongside other factors. In order for consultation to be
meaningful, it must be initiated at an early stage so that the people can have a holistic view and
idea of what is to come. In the level of deciding together (Table 3), beneficiaries are encouraged
to provide some additional ideas and options and to decide with the government or their
representative the best way forward. According to Wilcox (1999), deciding together is a
difficult standpoint, because it can mean giving people the power to choose without fully
sharing the responsibility for carrying out decision. Also, deciding together can mean accepting
other people’s ideas, then, choosing from options you have developed together, which is what
96
participation should be like in the low-income communities as most participants might not have
the required expertise to make a significant contribution. However, the basic rules of
consultation must apply including the need to generate other options together, choose between
them, and agree on a way forward. This is because people need more confidence to get
involved, as their levels of understanding and knowledge plays a part in their contribution to
the process. The level of deciding together usually takes more time as the inter-play between
the people and the government representatives need to be on par. Furthermore, from deciding
together, the level of acting together comes into scene. Acting together involves short-term
collaboration or forming more permanent partnerships that will last, as the people has been
involved all along in the process. Acting together involves both deciding together and then
acting together. Acting together is having a common language, a shared vision of what is
needed and the means it needs to be carried out. At this point, partners need to trust on another
as well as agree on what they want to do. At this stage, each partner needs to feel they have an
appropriate stake in the partnership, a fair say in what happens, and a chance of achieving what
they want (Wilcox, 1999). The level of acting together is a critical point in any partnership as
this stage brings about the benefit that any participatory process has to offer. Acting together
is not likely to be appropriate when one party holds all the power and resources and use this to
impose its own the solutions. Acting together is a point of power sharing, not the taking over
of power by the people. It must be known that people want to have a say in making decisions,
but not a long term stake in carrying out solution. The last stage on the modified Wilcox (1999)
Ladder talks about ‘supporting local initiatives’. This level is the most empowering level,
provided the people want to do things for themselves. Carrying through this level may involve
people setting up new forms of organisations to handle funds and carry out projects or
programmes. This process has to be owned by, and must move at the pace of those who are
going to run the initiative. However, the government or funders may set the deadline (Wilcox,
1999). This level of participation may be appropriate where there is commitment to empower
individuals or groups within the community; where people are interested in starting and running
an initiative. Wilcox further informs that this level will not be ideal when the following applies:
community initiatives are seen as ‘a good thing’ in the abstract and pushed on people from the
top down; there is no commitment to training and support; there are no resources to maintain
initiatives in the long-term and where time is of the essence. When the above processes are
given thorough consideration and the beneficiaries are involved in all the stages as outlined by
Wilcox (1999), housing satisfaction will be guaranteed. This is because involvement or
participation brings about interest which ultimately leads to satisfaction. It usually takes more
97
time for a fully participatory process to accomplish its goals, but the end result in the form of
housing satisfaction and improvement in the quality of life of the people will go a long way.
3.4.6 Beneficiary Empowerment
In order for beneficiaries to participate meaningfully in projects initiated with the goal of
improving their quality of life, it is imperative that they are empowered (Thwala, 2009).
Rowlands (1997:14) claims that “empowerment is concerned with the processes by which
people become aware of their own interests and how those related to the interests of others, in
other both to participate from a position of greater strength in decision-making and actually
influence such decisions”. Participation leading to empowerment is the common direction one
thinks of when discussing these concepts. The debate also centres on the question whether or
not a certain amount of empowerment is necessary in order to be able to participate at all.
The principle of empowerment states that people participate because it is their democratic right
to do so (Wignaraja, 1991) and participation also means having power (Tacconi & Tisdell,
1993). According to the concept, participation is the natural result of empowerment.
Empowerment is not a means to an end but it is the objective of development (Ogunfiditimi &
Thwala, 2007; Thwala, 2009). Thwala (2009) further informs that in addition to having the
power to make decision, it demands the knowledge and understanding essential to making
correct decisions because communities cannot make wise decisions if they do not have the
required information.
However, support organisations are required to be sources, as well as channels of information
to the communities so that they will be able to make informed decisions to developments that
relate to them. There are numerous developmental organisations, agencies and government
departments that regard local people as a good source of information (Thwala, 2009). But,
some other agencies and governmental departments do limit the people’s participation to an
advisory role. In cases where this happens, there is no participation, but a case of tokenism.
According to El Sherbini (1986), power must accompany participation; while Arnstein (1969)
stated that participation without power is an empty and frustrating process for the powerless.
Swanepoel (1997) states that under the banner of participation, people can be used as cheap
labour. This is because decision-making and planning are considered as being outside the orbit
of the powerless because participation is seen as interfering with the effective provision of
98
basic needs (Spalding, 1990; Thwala, 2009). Yet, empowerment involves more than having the
power to make decisions. Empowerment requires the knowledge and understanding needed to
make the correct decisions, of which the development practitioner has a special task in this
respects (Thwala, 2009). Beneficiaries cannot be expected to make wise decisions if they do
not have the necessary information to make these decisions. Kilian (1998) posits that
empowerment can be misused; it can become a ‘radical cloak hung around conservative ideas’.
Thwala (2009) further informs that empowerment does not mean giving people facilities that
were previously denied them or were not available to them, or giving them skills that they lack;
but that, in its purest form, empowerment is the acquisition of power and the ability to give it
effect. Such power is not an amorphous or indefinable entity, but manifests itself in groups of
people working together (Kent, 1981).
3.4.7 Benefits of Beneficiary Participation
The benefits of participation are usually seen differently because of the various interests
involved. According to Reed (2008) the many purported benefits of beneficiary participation
have to an extent been incorporated into national and international policy. Reed (2008) further
emphasis that, at the same time, cynicism has been growing amongst practitioners, stakeholders
and the general public, who feel let down when these benefits are not realized. The benefits of
beneficiary participation can be generally categorized under the normative and pragmatic
arguments for stakeholder engagement in developmental decision-making.
3.4.7.1 Normative Benefits
Normative benefits focus on the benefits for a democratic society, citizenship and equity (Reed,
2008). For instance, it is contested that beneficiary participation reduces the likelihood that
those on the margin of the decision-making milieu or society are disregarded. In this way, more
significant stakeholders can be included in decisions that affect them and active citizenship can
be promoted, with benefits for the wider society (Martin & Sherington, 1997). Beneficiary
participation is said to increase public trust in decisions and civil society, if participatory
processes are perceived to be transparent and conflicting claims and views are considered
(Richards et al., 2004). Beneficiary participation, it is claimed, can empower beneficiary
through the co-generation of knowledge with researchers and increasing participants’ capacity
to use this knowledge (Greenwood et al., 1991; Reed, 2008). Thus empowering them to make
informed decisions. It is also claimed that beneficiary participation may increase the likelihood
99
that developmental decisions are alleged to be holistic and fair, accounting for a diversity of
norms, values and needs and identifying the complexity of human-environmental interactions
(Reed, 2008; Richards et al., 2004). It may also promote social learning (Blackstock et al.,
2007). This is where the beneficiary and the wider society in which they live, learn from each
other through the development of new networks, building on existing relationships and
transforming adversarial relationships as individuals learn about each other’s’ dependability
and learn to appreciate the usefulness of each other’s views (Forester, 1999). Newig and Fritsch
(2009) argue that social learning may be one of a number of mechanisms that can deliver more
pragmatic benefits from participation, with groups of people developing more creative
solutions through reflective deliberation. The highlighted normative benefits are essential for
the achievement of beneficiary satisfaction in the South African housing subsidy scheme. This
is because most participants representing the communities would have developed a capacity
that will enable them to actively participate and make informed decision that will benefit the
entire community.
3.4.7.2 Pragmatic Benefits
Pragmatic benefits centre on the quality and durability of developmental decisions that are
made through engagement with the beneficiary. Reed (2008) argues that beneficiary
participation enables interventions and development to be better adapted to local socio-cultural
and environmental conditions. This enhances the rate of developed adoption and diffusion
amongst beneficiary groups, and their capacity to meet local needs and priorities (Martin &
Sherington, 1997; Reed, 2007). Beneficiary participation may make research more robust by
providing higher quality information input (Hansen, 1994). By putting local interests and
concerns into account at an early stage, it may be possible to inform project design with a
variety of ideas and perspectives, and in this way increase the likelihood that local’s needs,
expectations and priorities are successfully met (Dougill et al., 2006). Thus giving them
satisfactorily developed projects that concern them. It is also argued that beneficiary
participatory processes should lead to better quality decisions, as they can be based on more
complete information, anticipating and improving unexpected negative outcomes before they
occur (Fischer, 2000; Newig & Fritsch, 2009). By establishing common ground and trust
between participants and learning to appreciate the legitimacy of each other’s’ viewpoints;
participatory processes have the capacity to transform adversarial relationships and find new
ways for participants to work together (Stringer et al., 2006). This will lead to a sense of
100
ownership over the process and outcomes, when the participation is shared by a wide-ranging
combination of beneficiary, long-term support and active implementation of decisions will be
enhanced (Reed, 2009; Richards et al., 2004). Depending on the nature of the initiative, this
may significantly reduce implementation costs.
3.4.8 Beneficiary Participation in Housing Development
When implementing a participatory process, beneficiary participation should be considered
right from the onset, from concept development and planning, through to implementation, to
monitoring and evaluation of outcomes. Engagement with the beneficiary should commence
as early as possible in the decision-making process, which has been frequently cited as essential
if participatory processes are to lead to higher quality and durable decisions (Chess & Purcell,
1999; Reed, 2008; Reed et al., 2006). Normally, beneficiaries only get involved in decision-
making at the implementation phase of the project cycle, and not in earlier project identification
and preparation phases. Gradually, they may also be involved in monitoring and evaluating the
outcomes of the decision-making process (Estrella & Gaventa, 2000), to see how the process
is undertaken for subsequent empowerment to participate meaningfully in other development
that will concern them. However, unless flexibility can be built into the project design, this can
mean that beneficiaries are invited to get involved in a project that is at variance with their own
needs and priorities (Reed, 2008). This can make it a problem to motivate beneficiaries to
engage with the decision-making process, and those who are engaged may be placed in a
responsive position, where they are asked to respond to proposals that they perceive have
already been concluded (Chess & Purcell, 1999; Reed, 2008).
3.5 CONCLUSION
This chapter addresses the gaps observed in previous residential satisfaction research
frameworks which were not evaluated as all-inclusive constructs in the previous models. The
identified gaps form the new constructs in the current study’s conceptual framework. The
identified gaps are: needs and expectations and the beneficiary’s meaningful participation in
the housing process. In addressing the needs and expectation gap, the study draw on the seminal
works by Maslow and established that as part of the conceptual framework of residential
satisfaction research, the gratifications of housing needs and expectations should have the
noteworthy prominence. This is because people have different housing needs and expectations
that cannot be satisfied with the same housing conditions; and this could bring about different
101
satisfaction levels because their needs and expectations are different. From the reviewed
literature, it was found that residential satisfaction is basically formed under the condition of
the current level of housing needs that is being pursued. Unless the level one need is sufficiently
satisfied, they will remain in the occupants’ consciousness and will thus become the prime
determinants of housing behaviour. Hence the living condition that is currently pursued forms
the housing expectation of the individual, which is related to the overall residential satisfaction.
Also, in addressing the gap of the beneficiary’s meaningful participation, several works were
drawn upon but the seminal work of Arnstein’s Ladder of Participation, Burns et al. modified
version of the Arnstein’s Ladder of Participation conceptualized as the Ladder of Citizen
Empowerment and the Wilcox’s Theory of the Ladder of Participation was used in addressing
the gap. However, the work of Wilcox bears more prominence. The Wilcox Framework
indicates that different ‘levels’ of participation are acceptable in differing context and settings.
The framework recognizes that even though power is not always transferred in the participative
processes, but the processes still have value. This view was in contrast to the common
interpretation of Arnstein’s Framework, which brings to the fore thought that it is only
acceptable to strive towards citizen’s control. Exclusive to some contexts, this shift in
philosophy has been further developed to describe levels of involvement on a continuum. The
Wilcox’s Framework provides useful insights into the scope of experiences associated with
beneficiary participation, which by their nature represent simplifications of a more complex
reality. The Wilcox Model further posits that you need to inform the beneficiary about what is
planned so that they are informed about what is happening. Because of the information stated
above, this model of participation is what the current study advocates for, in order for the
beneficiaries of the South African housing subsidy schemes to participate and ultimately have
satisfaction with the houses being provided.
102
CHAPTER FOUR
HOUSING RESEARCH THEORY
4.1 INTRODUCTION
This chapter of the thesis presents a framing overview of housing research, housing theoretical
frameworks, overview of the most influential perspectives on housing, followed by a
discussion on the methodological approaches to housing studies. An evolution in housing
policy framework is also presented, with the various forms that housing policy has been
attending to over time. Also, the objectives of housing policy and the purpose of housing policy
are examined. Lastly, the chapter closes with an outline of housing policy instruments, which
enables the intentions of housing policy to be actualized.
4.2 HOUSING THEORETICAL FRAMEWORK
An understanding of housing theory is imperative because it can be claimed that discussions
of policy formulation and evaluation will be deprived of direction and reason, unless there is
an understanding of a clear theoretical basis. However, one must have an understanding of the
concept of housing practice and the nature of housing to generate a useful housing theory.
Literature on housing has developed since the mid-1960s, and many housing or social scholars
write from a position of commitment to a specific theoretical framework, be it neo-classicalism,
institutionalism, or neo-Marxism, amongst others. However, Pugh (1986) raised questions as
to whether the concept of housing in its essential nature can precisely fit into any single
theoretical framework.
Sullivan and Gibb (2006) state that housing is a difficult thing to hypothesize about. This is
because it is an inherently multifaceted commodity, with ‘spatial fixity’ a defining
characteristic, and asset, investment and consumption dimensions to account for. They insisted
that the economics of housing remains a challenge to those who seek equity and efficiency
improvements for society to this day. Lux (2003) also argues that housing is not a simple
category that can be viewed from a single perspective. On one hand, housing is one of the basic
human needs and the right to adequate housing has been classified as a basic social human right
in most developed countries around the world, with most of the developing countries currently
enshrining it into their constitutions. On the other hand, housing constitutes a special type of
103
private property, traded on the market. Although, trade-offs between the social and economic
aspects of housing may have to be made sometimes, which thus necessitate the searching for a
consensus that assures both the effective functioning of the housing system for all social groups
within a society (Lux, 2003). Furthermore, housing is also a field of inquiry, which Robinson
(1979) termed; thinking in ‘theory time’ is a poor substitute for recognizing the role of real,
non-abstract, historical time. Moreover, conceptualizing from the spatial aspects of housing
leads to quote Maclennan (1982) informing that it is a curious form of ‘pointless economics’.
Good theory according to Maclennan essentially involves jettisoning many of the assumptions
usually made in neoclassical economics, rendering the work more difficult and ‘messy’.
Also, housing is a multidisciplinary field, therefore housing research draws on a number of
disciplines and professions, including economics, geography, political science, planning, and
architecture, among others (Van Vliet, 2003). Van Vliet (2003) further emphasized that
respective disciplines and professions tend to concentrate on a particular set of questions. For
example, economists are inclined to concern themselves with issues of housing finance,
dynamics of supply and demand, and house prices. Geographers on the other hand often
explore spatial aspects of housing, together with the patterns of residential segregation and
urban form, local and regional housing markets, urban regeneration and gentrification, and
residential mobility. Likewise, architects usually focus on aspects of design, building materials,
and construction techniques. This emphasis of attention is neither mutually exclusive nor
exhaustive. They show an inclination within disciplines to employ certain types of questions
according to hypotheses that characterize those disciplines more generally.
However, it should be noted that most times, disciplinary labels apply to paradigms rather than
individuals. This is because the researchers may have been formally appointed in a certain
discipline but their scholarly work draws on conceptualizations or approaches originating from
or dominating in another discipline. A defining characteristic of a discipline is that it concerns
itself with a diverse array of topics from a demarcated paradigm that helps specify the questions
asked and that guides research design, choice of data collection methods, and explanatory
theory (Van Vliet, 2003). In contrast, a field of study comprises a fairly clearly delineated
subject matter, such as housing, but draws on the contributions from a wide range of
disciplines.
104
4.2.1 Theoretical Perspective of Housing Study
The theory of housing has its origin in the Paleolithic period when homo-sapiens began to use
natural materials like stone, wood, leaves, animal skin and other similar items to create shelter
from elements of weather (Ifesanya, 2003). It has been defined by different scholars in different
ways depending on the emphasis and focus of analysis. However, the basic definition has been
housing as shelter and for the provision of human needs (Sharipah, 2007). Ifesanya (2003)
informs that the initial form of housing was the post and beam construction of the Stone Age,
when the principal and perhaps, the only motivating factor for housing was fortification from
external aggression and from climatic elements like sun, rain, heat, cold and other extreme
weather conditions. This nonetheless cannot be referred to as housing, but ordinary shelter. The
United Nations (1978) however defined housing not simply as a shelter, but also as a means of
creating communities, providing great emphasis on the functions which housing has to
perform, thus making housing a multi-dimensional concept.
As described above, housing in today’s expression has become a multi-dimensional bundle of
services, encompassing the need for privacy, aesthetic value, conformity to statutory standards,
fiscal economy and other related issues of importance in contemporary society (Ifesanya,
2003.). According to Randava (1979), housing must not be misunderstood or narrowed to
describe a single unit of dwelling. The house is only a constituent part of housing, and its
functionality and quality is determined by the surroundings, that is, the environment. Also, it
is the process of providing a large number of residential buildings on a permanent basis with
adequate physical infrastructure and social services, planned decent, safe and sanitary
neighbourhoods to meet the basic and special needs of the people.
This fundamental perception of housing has since given way to an all-inclusive definition.
McLead (2002) states that adequate housing offers a refuge for emotional and physical rest,
and the stability found therein empowers families in their pursuit of a better quality of life. The
importance of a decent place to live cannot be overstated, for with it comes stability and
promise, family unity, hope and a foundation from which individuals reach their full potential.
Further, Schirnding and Dodd (2002) suggest that adequate housing enhances healthy living.
Learning and academic accomplishment are also enhanced by adequate housing, according to
Hodson and Pelullo-Willis (2002); while emotional stability and psychological balance are
constructed by having a decent and comfortable place to live (McLead, 2002).
105
Housing is one of the most important elements in our lives and community. It is both a shelter
and a link to the neighbourhood and larger community. Housing also refers to both the physical
product and the process of its attainment. Housing is mostly perceived according to its
performance, and its usefulness varies with the level of comfort and hygiene it provides. The
significance of people in housing is recognised not when housing complies with municipality
or city by-laws, but when people come to live in it and it is acceptable to the community, which
is the primary reason this study is being undertaken. Housing also means privacy and is an
expression of ways of life, aspirations and social relationships. Thus, Dwijendra (2007) informs
that housing is the provision of comfortable shelter with available infrastructure, services and
facilities that address the need of the occupants. Furthermore, Li (2002) also defined housing
as:
a heterogeneous, durable and essential consumer good;
an indirect indicator of status and income differences between consumers;
a map of social relations within the city;
an important facet of residential structure;
a source of bargaining and conflict between various power groupings; and
a source of profit to different institutions and agents involved in the production,
consumption and exchange of housing.
Hence, these diverse classifications make the study of housing a complex issue opened to
various interpretations. Tan (2001) thus classified these various perspectives as follows:
political perspective;
social perspective;
developmental perspective;
institutional perspective;
radical perspective;
comparative approach;
historical approach;
experiential perspective; and
neo-classical perspective.
While Tan’s (2001) classification of housing gave a valid view, it would have been an all-
106
inclusive view if the economic perspective was included as a separate entity, thus developing
a framework that would have disentangled the multi-dimensional concept attached to housing.
However, Li (2002) says that the diversity of approaches to the study of housing is partly a
manifestation of the multifaceted nature of the topic. It should be noted that other renowned
theoretical positions such as that of Mancur Olson’s (1965) perception on collective choice and
special group interest are also capable of explaining some important aspects and individual
behaviour within the housing systems. The next section provides an overview of the three most
influential approaches to housing: the neo-classical, the institutional, and the neo-Marxist (the
radical perspective) perspectives, with the introduction of two new separate views, the
economic and social perspectives.
4.2.1.1 Neo-Classical Perspective
The Neo-Classical Perspective on housing draws its theoretical guidance from neo-classical
economics. Neo-classical economics describes a distinct and relatively homogenous school of
thought in economic theory that became prominent in the late nineteenth century and that now
dominates mainstream economics. The term was first introduced by Thorstein Veblen (1900)
to describe developments in the discipline (of which Veblen did not entirely approve)
associated with the work of figures such as William Jevons, Carl Menger, and Leon Walras
(Brennan & Moehler, 2010). Brennan and Moehler (2010) inform that neo-classical economics
relies on subjective preferences for determining prices in order to escape from the so-called
objective value theory of classical economics, according to which the value of goods could be
established by a reference to some basic commodity or the labour input required to produce a
good. However, it is largely concerned with the analysis of utility maximization on the part of
individual consumers in an atomistic housing market (Li, 2002). Put at its simplest, neo-
classical economics views society as collection of individuals whose nature is assumed to be
given. The realisation of individual preferences shapes the form of the economy and the nature
of the society.
In its study of the economy, neo-classical economics makes four suppositions. Firstly, the
creation of goods and services reveals the preferences of consumers. Next, it is assumed that
all households and organisations have perfect information. Thirdly, from this basis of perfect
information, households get the most out of utility and organisations and maximize profits.
Lastly, the creation of goods and services is assumed to be flexible in that the factors of
107
production can easily be interchanged. It should be borne in mind that the theoretical roots of
these suppositions are in ‘methodological individualism’. That is the methodological position
that aims to explain all economic phenomena in terms of the characteristics and the behaviour
of individuals. Because everything ultimately reduces to what individuals do, methodological
individualism states that any theory of how the economy runs should be built up from an
understanding of how the individuals within it behave. This is the classic idea the current study
hopes to underpin, in that the issues of housing should be understood from the occupants’
(users’) point of view, which this study refers to as ‘methodological occupancy’. The Neo-
Classical Perspective’s commitment to methodological individualism means that neoclassical
economics puts clear boundaries around what it is attempting to explain (since theories cannot
explain everything).
The neo-classical perspective is considered the orthodox approach to cities and housing
because of the suppositions in ‘methodological individualism’. Since housing has ‘externality’
value, one can locate housing in this theoretical framework by relating it to the economic theory
of externality and public goods. However, the central tenets of the neo-classical approach
include equilibrium, individual utility maximization, and the absence of severe information
problems (Arrow & Hahn, 1971; Hodgson, 1988, 1998, 1997). Li (2002) posits that the
assumptions and overall importance of the equilibrium of conditions continue to provide the
footings of many studies of urban structure and housing. In this dimension, the adjustment
model of residential location in the determination of micro house prices are worthy of attention.
The adjustment model of residential location suggests an association between the consumption
of housing space and travel costs. The general hypothesis is that households trade-off travel
costs (which increase away from the city centre), against housing costs (which are shown to
decrease from the city centre) in an attempt to maximize utility subject to an overall budget
constraint (Alonso, 1964; Kain, 1962; Muth, 1969; Mills, 1972). The relationships between
transport costs, housing costs and income and other trade-offs are joined together in this
framework to predict the relationship of individual households and those of different income
groups within the city (Li, 2002). Mills’ (1972) model of urban structure effectively predicts
the decline of land values and population density from the city centre, both of which are found
in most large cities of North-West Europe and North America (Li, 2002). Thus, individuals
will always trade-off one commodity for another, but only in situations where there is some
exogenous change (McMaster & Watkins, 1999). Thus in the neoclassical theory, the
individual’s behaviour is explained by concentrating on the changes in the constraints to which
108
he or she is exposed; preferences are assumed to be constant.
Basically, neo-classical economists have a habit of using their theoretical framework to explore
principles of effectiveness, incentive and maximization of utility or minimization of costs. But
when it comes to housing they are drawn into social and public policy where they face
considerable problems. The most general criticism about this approach is that the models fail
to consider the structuring of household’s housing decisions. This is because households do
not make choices in a vacuum. The preferences household’s express and the constraints that
they experience are influenced by the nature of the wider social structure and by the more
immediate effects of the specific character of certain systems of housing production and
allocation. Also, another criticism of the neo-classical approach is that optimization is not
practicable (McMaster & Watkins, 1999). This criticism is based on the grounds that the
optimization process itself is costly and requires cognitive abilities that are scarce (Arrow,
1962).
Arrow (1962) further argues that there is a fundamental paradox in the determination of the
demand for information; its value for the purchaser (user or beneficiary) is not known until
after he has the information, but then he has in effect acquired it without cost; given incomplete
appropriability, the potential buyer will base his decision to purchase information on less than
optimal grounds. Arrow’s (1962) critical criticism recognizes that information is costly to
obtain, and that its value is uncertain ex-ante, so the utility maximizing individual faces a
considerable conundrum. Today, market-based economic studies in housing policy are
conscious of the effects of wider social factors and tend to have varied relative merits in the
context of balancing economic and social values.
4.2.1.2 Institutional Perspective
Institution is a term that is often employed in varied field of studies, but not frequently defined.
Hodgson (1998) quotes Walton Hamilton’s 1932 definition of an institution as, “a way of
thought or action of some prevalence and permanence, which is embedded in the habits of a
group or the custom of a people”. Hodgson (1998) further informs that this is a broad definition
that includes organisations, such as universities and firms, but also ‘integrated and systematic
social entities’, such as money, language, law and religion. Hence, institutions represent a co-
ordination of belief that is durable, although not unchanging.
109
It has long been accepted, even amongst neo-classical economists, that institutions matter.
However, the challenge initially was finding a way to integrate institutions into housing
analysis. At present, there has been a serious debate concerning the position and viability of
existing institutional arrangements, including nation states. Giddens (2002) argues that many
institutions have become ‘shell-like’, and have become inadequate to the tasks they are called
upon to perform. However, a common set of institutions can be found in most societies,
including public and private enterprises, public utilities, financial establishments, educational
institutions, trade unions and government/quasi-governmental agencies. The relative strength
of these institutions can vary, as also the manner in which they interact. North (1989) outlined
that an institution consists, primarily, of informal constraints, formal rules, and the enforcement
characteristics of both. These rules are separate from the players, or organisations such as
schools, firms, trade associations, and government agencies that are also called institutions.
North’s (1989) clarification provides a strong background to discourage the loose but common
sense use of institutions only as organisations but not as rules.
Institutional economists believe that evaluation of policy is socio-political and that public
policies are necessarily expressed through institutional arrangements (Gruchy, 1972; Myrdal,
1978). Contrasting with the neo-classical approach that accentuates preferences and value of
individual; the institutional perspective tends to focus on groups and organisations. This is
because the institution arrangement believes that members of a group interact with each other
on a regular basis; which is the premise that public participation is based upon. Li (2002) states
that this regularity encourages the formation of shared values and commonly accepted rules
and norms (institutions). Likewise, a group’s capacity to pursue its members’ common interests
depends on its collective power, which in turn is the function of the amount of resources at the
group’s disposal – wealth, position in the government and society, and the size of the group.
Thus, Guy and Henneberry (2000) assert that the institutional approach to housing offers an
alternative to the positivist theories, which reify, idealize and isolate economic structures and
individual behaviour.
One way of theorizing the institutional perspective into housing analysis is to define a housing
system (Murie et al., 1976) or structures of building provision as suggested by Ball (1996). The
latter identifies historically contingent networks of relations (structures) associated with the
provision of particular types of buildings. These networks Ball (1996) claimed are historically
contingent; implying the pattern of networks is shaped by history and is an empirical question.
110
When viewed from this perspective, the analysis of institutional power and behaviour of
interest groups is a major strand of research to be explored. However, Bassett and Short (1980)
using Maxist analysis provided an analytical framework of the power relations and interactions
amongst the institutions involved in the UK’s public housing system, which was categorized
as an institutional approach to housing.
Bengtsson (2001) promotes that housing policies in most Western countries are best perceived
as the state providing corrective measures to the housing market. This means that institution
contracts serve as the main mechanism for distributing housing, and state intervention takes
the form of correctives defining the economic and institutional settings of those institutional
contracts. The emphasis on institutional power and behaviour leads to the development of
strategies to contain or defuse conflict. Institutionalists realize that public policy economics is
necessarily within a context of political power, conflicts of interests, and the wider social and
historical impacts, which affects the way institutions are developed (Li, 2002). Housing would
reside in this context of political economy. Moreover, institutional scholars see the scope of
political economy as widening and becoming increasingly relevant. They also recognised that
various devices can be used to bring cohesion, including markets and the State’s social,
political, and economic roles.
Unlike the neo-classicalists, the institutionalists do not feel uncomfortable working with the
‘political’ and they often draw upon useful neo-classical methods and empirical findings.
Unlike the neo-Marxists, the institutionalists do not look for a ‘final’ solution where conflicts
disappear after the demise of capitalism. Rather, they perceive a continuously evolving society
with well-designed institutional arrangements offering only temporary and tentative solutions;
knowing that as society changes, so it will become necessary to revise institutional
arrangements. Sometimes, neo-classical economists criticize the institutional approach for its
free-ranging characteristics and remoteness from deductive and prognostic techniques and
thinking (Li, 2002) making the neo-Marxists to conclude that the institutional perspective
scholars are dealing with symptoms and offering mere prescriptions of justice without attention
to class conflicts and recurring internal contradictions in capitalism. Institutional scholars are
still striving to develop and strengthen their paradigm based on the criticism laid upon them.
In his work on Transaction Costs, Institutions, and Economic History, North (1984) attempted
to integrate the useful features of neo-classical and neo-Marxist approaches, in which he
111
hypothesised three basic assumptions: individualism; specifying and enforcing the rules that
underlie contracts is costly; and ideology modifies maximizing behaviour. North (1990) later
put forth an explanation of institutional and organizational change that is endogenous. The
main points in the work are:
Agents: entrepreneur, the decision makers in organizations;
Sources: opportunities perceived by entrepreneurs;
Process: overwhelmingly incremental; and
Direction: determined by path dependence.
The explanation of institutional change provides profound implications for citizens and
policymakers to enable them to evaluate the gains and losses of alternative policies in a more
accurate way (North, 1993). Streeck and Thelen (2005) assert that it is exactly in this interplay
that institutional dynamics are created informing that political institutions are not only
periodically contested; they are the object of on-going skirmishes as actors try to achieve
advantage by interpreting or redirecting institutions in pursuit of their own goals, or by
subverting or circumventing rules that clash with their interests. Summing up, housing policy
in an institutional perspective is about the interaction between rules, rule makers and rule
takers. Here the interaction concerns a specific attempt to change rules that are guiding ‘quality,
quantity, price and ownership and control of housing’ with regard to adequate housing.
4.2.1.3 Neo-Marxist Perspective
The neo-Marxist perspective draws its foundation from the thought of Karl Marx (1818-1883),
the founder of the Marxist tradition of political economy. The perspective is centered on the
analysis of the contradictions of capitalism with a view to replace it with socialism. However,
the radical road to socialism, such as those taken by Lenin and Mao Zedong in China, has been
stopped and is prohibited. Capitalism as an economic and political system has proven to be
more durable and flexible, than Marx maintained. In the modern social system, for example,
the advent of communism does not appear imminent and it can be concluded that the Marxist
theory is far from dead. However, the old Marxist theories have since been transformed and
reworked.
According to Li (2002), the ‘guiding thread’ of Marxist analysis is the economic structure, or
the approach of production consisting of the factors (techniques) of production and the relations
112
(ownership) of production. As Marx (1976) put it, the guiding principle of his studies could be
summarized as follows. In social production of their existence, men inevitably enter into
definite relations, which are independent on their will, namely relations of production
appropriate to a given stage in the development of their material forces of production. The
totality of these relations of production constitute the economic structure of society, the real
foundation, out of which arises a legal and political superstructure and to which correspond
definite forms of social consciousness correspond. The mode of production of material life
conditions the general process of social, political and intellectual life. At a certain stage of
development, the material productive forces of society come into conflict with the existing
relations of production, from forms of development of the productive forces these relations
turn into their fetters (Marx, 1976). Then begins an era of social evolution. The changes in the
economic foundation lead sooner or later to the transformation of the whole immense
superstructure.
Since housing is one of the three basic necessities a person requires for a nominal standard of
living, Karl Marx and other Marxist scholars debate that the state should embrace the social
obligation of housing provision to the disadvantaged sections of the society (Chereni, 2010).
Hence, the state is seen as the main driver of the processes of shelter provision. Until the early
90s, this thinking has been the motivation behind housing policies in Asian socialist countries
and other African countries which shared the same ideologies.
Volume III of Marx’s (1976) work on Capital paid considerable attention to land rent, but with
much debate. His categories of rent go beyond merely the Ricardian residual value arising from
differences in the fertility of the soil, which he called Differential Rent I. He introduced
Differential Rent II arising from differences in outputs by varying the capital intensity on the
same plot of land. There is some dispute over the usefulness of this rent category according to
Ball (1977). Marx did not write a lot on housing and what he wrote is not closely integrated
into his broader theories of economic structure and social change. Nevertheless, Engels (1970)
provided an early analysis of what he called ‘the housing question’. Engels (1970) work
opposed the suggestion that homeownership is the answer to the housing problem, insisting
that it will ‘chain’ the worker to the spot from mortgage debts and prevent them from looking
for employment elsewhere. Thus, insisting that the housing problem was inseparable from the
capitalist mode of production; it could only be solved by the elimination of that mode of
production, the abolition of the big city and the ending of the separation of town and country.
113
The Neo-Marxists perspective views the delivery of housing as the role of the state and not of
the private sector. Their core argument is that, should this role be vested in the private sector
then this will result in commodification of housing. This is because of the basic tenets of the
neo-Marxist scholars that the poor are regarded as a necessity for the purpose of capital
accumulation. The neo-Marxist believes that the Commodification of housing will therefore
lead to inequalities and exploitation by capitalists towards the poor. The neo-Marxist classic
has the advantage of viewing housing in broader political economy. Smith (1999) explains that
the Neo-Marxist thought focus on the role of the state in the process of capital accumulation,
including its role in housing. This is viewed in a twofold manner namely: to support capital
accumulation and to maintain social stability (legitimation).
Smith (1999) further states that the neo-Marxist concept is based on the fact that the poor are
vulnerable to exploitation in the capitalist mode of housing delivery. That is, if housing is
provided with the primary aim of serving as a means of poverty alleviation, then the state must
be the exclusive provider of housing. This, they believe, will lessen the manipulation of the
poor by capitalists as the state will be able to own land and therefore, the poor will be in a
position to be accommodated within well-located land and in close proximity to services and
opportunities (Manikela, 2008). The neo-Marxists’ belief is that the problems experienced in
housing provision are solely due to capitalism. They see capitalism as dividing societies in
terms of economic classes and the poor remain the most vulnerable and exploited (Manikela,
2008). Neo-Marxists also accentuate that the housing problems cannot be answered whilst the
capitalist system is still in existence, it must be removed. Omenya (2002) argues that Marxists
are the popular critics of what they term the ‘Capitalist System of Urban Planning’. They claim
that the urban planning system exercised by most capitalist countries has only served the
bourgeois interests. They view the urban planning system as a means to segregate spatial land
uses into economic class antagonism. According to the Marxists, urban development is just a
reflection of mode of production in order to gain more economic surplus.
Nevertheless, the over-concentration on economic relationships, considered by neo-Marxist to
be the most important basic relationship in society has led to a number of criticisms. The main
weakness of the neo-Marxist perspective is that Marxian categories are difficult to test
empirically and the theories tend to be general rather than specific to housing. In particular,
Popper (1974) argues that Marxism is not a theory that can be tested and possibly falsified,
114
mainly because it sees the replacement of capitalism by communism as ‘historically inevitable’.
4.2.1.4 Economic Perspective
Purely from an economics perspective, housing constitutes private property because it is not
accessible in a non-competitive manner and does not have the same features as other public
goods (Truett & Truett, 1987). However, housing economics notes that the following important
differences between housing and standard market commodities exist:
1. Housing is a very heterogeneous, multifaceted and multidimensional good. This is
because individual houses, and apartments or flats, differ in floor space, design, age,
quality, standard, furnishing, tenure, size and number of additional spaces or buildings
location, quality of the environment, and accessibility amongst others. Likewise, it is
also very challenging to measure the unit of output and the demand for housing in
general because the rate/rent paid for a small flat can be the same as for a large family
house, even under conditions of optimal distribution and market equilibrium.
According to Lux (2003), the above expressions have necessitated housing economists
to introduce a theoretical construct called housing service. Housing service theory states
that in a state of equilibrium, the price per housing service unit will be the same in all
types of dwelling units. Therefore, households or individuals thus demand housing
service rather than housing on the market.
2. Housing is a durable good and as such it becomes subject to both consumption and
investment. Fallis (1985) claims that there are two housing markets. In one, the
consumer good, housing service, is exchanged and the price per unit of housing services
is determined. In the other, the investment good, housing stock, is exchanged and the
price per unit of the housing stock is determined. However, Lux (2003) argues that
housing consumption and investment motives may conflict. The consumer wishes to
maximize utility but the investor chooses from all the options a housing unit with a
maximum net present value of expected future returns.
3. Also, housing is a spatially fixed good and cannot be moved from one location to
another. To buy a dwelling means not only a particular dwelling but also to buy the
socio-economics status of the neighbourhood and the level of accessibility to the place
of employment.
Housing constitutes a significant share of household expenditure as well as total wealth.
115
According to Chetty and Szeidl (2004) the mean expenditure share for shelter (i.e. housing) is
about 20%, household income, supplies and furniture is about 6%, transport (including gas and
maintenance) is 16%, food and apparel each is 15%, utilities, fuels, and public services is 7%,
health care is 6%, the rest are for education, entertainment, and miscellaneous items.
Greenwood and Hercowitz (1991) found that the value of the residential capital stock is larger
than that for business capital, and usually, the annual market value of residential investment is
larger than that for business capital investment. Clearly, housing is not just ‘another’
consumption good. Significant fluctuations in housing development prices would imply
significant fluctuations in wealth, and thus effects potentially significant household wealth
(Skinner, 1989). For instance, Morris and Heathcote (2003) find that the market value of the
United States residential property stock is approximately equal to the annual average Gross
Domestic Product (GDP).
Also, housing is usually associated with a high transaction cost of potential moving, that is, the
finding and furnishing of a new dwelling and moving involves considerable expenses, not only
monetary expenditure, but also time and emotions invested that do not relate directly to the
acquisition of a new dwelling. The housing market adapt to changes in household income very
slowly, compared to potential adaptations, if the transaction cost equals zero. Maclennan
(1982) informs that such cost may range between five and ten percent of the total price of a
house, particularly where movement entails both selling and purchase costs. However, Monk,
Tang and Whiteheadat (2010) suggest that at the city level, it has been argued that sufficient
housing supply, underpinned by new housing investments, both from the government or
developers, helps to support a vibrant urban system and contributes to urban competitiveness.
The economic perspective of housing claims that housing accrues benefits, such as employment
creation, from investing in housing. Hence, the opportunity cost of investing in housing as
compared to investing in employment creation directly would need to be estimated by
measuring the value of the benefits of each investment – a research project in itself. In principle,
a house is a house, and housing as an investment is no different from any other kind of
investment. In reality, housing wealth can be converted into additional income, and as a result
alleviate income poverty among home-owners, especially in later life. A study by Dewilde and
Raeymaeckers (2008) found that being a home-owner effectively shielded older people from
different forms of poverty as home-owners had a significantly lower risk of being income poor,
of being deprived and of being cumulatively deprived. However, the poverty-reducing effect
116
of being a home-owner diminished significantly as the home-ownership rate increased, because
as more households own their own home, there are more low-income homeowners. It is also
argued that people use the equity in their house as part of their pension, or at least reduce
savings because of their investment in a house (Muellbauer & Murphy, 2008). Hence, Monk
et al., (2010) inform that, not investing in housing will tend to amplify market inequality and
social exclusion. Also, the impact of housing on a family include the opportunity that living in
better housing could lead to a better economic position of the household. For instance, a move
to an area of expanding employment could enable family members to get a better paid job or
to at least get a job (Monk et al., 2010). This is particularly likely for social tenants, who
previously may have lacked a permanent address, causing them difficulty in accessing basic
services that others take for granted, such as overdraft facilities or even bank accounts. Thus,
the macro-economy and the housing market are indeed interrelated and co-determined.
Also, Glossop (2008) claims that in delivering healthy and attractive communities, housing can
contribute to the development of a knowledge-based economy and plays a vital role in
attracting and retaining the most talented and skilled members of the workforce that will be the
catalysts of economic growth in the future. A study by Bramley and Morgan (2003) in Central
Scotland also confirmed the pivotal role of new house-building in supporting city
competitiveness. It is argued that new housing increases the competitiveness of cities in three
main ways:
by ensuring an adequate and responsive supply of housing;
by providing a high quality living environment; and
by promoting urban vitality
Bramley and Morgan (2003) establish that new housing is principally important for mobile
workers, especially those with higher skills, partly because of its relatively easy purchase
process.
However, with all the over-arching benefits that come with the economic perspective of
housing, standard macroeconomics textbooks either treat housing as one of many consumption
goods, or neglect it all together. ‘Mainstream macroeconomics,’ simply put, ignores the
housing market. Conventional housing economics and urban economics research for its part
virtually ignores interactions with the macro economy. At best, some of the theoretical and
empirical analyses for urban and housing economics, include macroeconomic variables (such
117
as the inflation, the economic growth, GDP, the unemployment rate, etc.) as exogenous ‘control
variables.’
4.2.1.5 Social Perspective – The right “to Adequate Housing”
As previously stated at the beginning of the theoretical framework, housing is professed to be
a basic social need of human beings and its standard greatly influences the standard of welfare
of the whole society. Housing insecurity can have far reaching consequences for the labour
market, as well as for the political stability in a particular country. Lux (2003) informs that in
view of the increased acceptance of the theory of the welfare state after World War II the right
to adequate housing has become one of the fundamental social rights in all economically
developed and developing countries and the responsibility for housing has progressively
transferred from the occupants and family to the government of states and public finances.
The right to housing is a social right, which constitutes the third element of human rights apart
from political and civil rights; which is the tenet of the social perspective. The key norm of the
right to housing is equal and non-discriminatory access to housing with respect to race, creed,
and sex. In a country like South Africa, housing earns the particular ‘attention’ of the State and
is included directly in the Constitution. South Africa has defined the right to housing in great
detail in its Constitution. The South Africa Constitution contains justifiable socio-economic
rights and enshrines everyone’s right to have access to adequate housing. In the Bill of Rights
in Chapter 2 of the Constitution, section 26 outlines: “26 (1) everyone has the right to have
access to adequate housing. (2) The state must take reasonable legislative and other measures,
within its available resources, to achieve the progressive realisation of this right. (3) No one
may be evicted from their home, or have their home demolished, without an order of court
made after considering all the relevant circumstances. No legislation may permit arbitrary
evictions”.
The right to housing as one of the human rights is also expressed in a number of international
documents, the oldest being the Universal Declaration on Human Rights, which the United
Nations General Assembly adopted in December 1948. Article 25(1) of this document states
that: “Everyone has the right to a standard of living adequate for the health and well-being of
himself and of his family, including food, clothing, housing and medical care and necessary
social services…” In Principle 4 of the Declaration on the Rights of the Child, adopted in
118
November 1959, it is stated: “the child shall have the right to adequate nutrition, housing,
recreation and medical services...” Likewise, Part II, Article 10 of the Declaration on Social
Progress and Development, adopted in December 1969, states that: “the basic freedoms can be
attained also by provision for all, particularly persons in low income groups and large families,
of adequate housing and community services.” Furthermore, Article 11 of the International
Covenant on Economic, Social and Cultural Rights adopted in 1966 states: “the State Parties
to the present Covenant recognizes the right of everyone to an adequate standard of living for
himself and his family, including adequate food, clothing and housing, and to the continuous
improvement of living conditions”. To substantiate the recognition of the obligations under
these International documents, the UN Council for Human Rights has worked out numerous
recommendations, explanations and commentaries, supporting the adoption of the documents
(not emphasized in the current study).
The social perspective of housing is mostly accepted as the right to ‘adequate housing’ which
is understood as ensuring affordable housing for the disadvantaged and endangered social
groups such as the seniors citizens, children, physically handicapped individuals, victims of
natural and other disasters. This right is a general awareness and acceptance of a housing price
level in the society at large that will ensure the fulfillment of basic needs in the field of housing
(Lux, 2003), as well as the likelihood of obtaining social support in cases when the family
cannot ensure this fulfillment by its own means and the availability of housing. The continuous
efforts to guarantee the greatest possible degree of general and financial affordability of
housing is especially important in relation to groups of the population with little social power,
that is, those who cannot by themselves ensure adequate housing on the free housing market
(Lux, 2003).
4.2.2 Methodologies in Housing Studies
With respect to the development of theories and methodical approaches to housing studies,
Kemeny (1992) states that regardless of the growing interest in theoretical housing issues, there
remains a strong propensity for housing scholars to incline themselves in their own empirical
and policy issues, with almost complete disinterest in ‘abstract’ questions. This contemporary
approach to housing studies was further illustrated by Marston (2002). Marston (2002) claims
that where theory is used in contemporary housing research it tends to be mid-range versions
of political economy theory, for instance the concept of a ‘housing system’, competing
119
definitions of housing need, or comparative and state-centered social policy. These studies have
an important place, particularly comparative studies that confront us with our own assumptions
about what's ‘natural’. However, very few empirical studies are explicit about the
epistemological foundations or the theoretical frameworks that inevitably inform them.
According to Saugeres (1999), one of the principal reasons that this trend continues to be the
case is that most housing scholars have tended to ignore the definitions of housing policy
makers. For instance, Wang (2004) says that housing scholars bend to this mind-set because
‘policy problems’ are taken to be objective facts, rather than contested realities. Within this
mind-set, much of the focus of this type of housing research is a theoretical empirical work
focused on addressing policy problems defined by governments and their instrumentalities
(Wang, 2004). Nevertheless, this mind-set does not mean that researchers leave an empty space
for the theoretical basis of their research; in actual fact, the paradigm indicates that there are
hidden theoretical evidences for these housing studies rarely to be challenged.
There are two major methods in planning the theoretical background in housing study, as
suggested by Harloe (1995). One stresses the similarities and the other emphasizes the
differences. An example of a study on the different methods is the famous Esping-Anderson’s
(1990) Welfare State Typology. In the book titled: “Three Worlds of Welfare Capitalism”, the
capitalist world is characterised into three different types of welfare systems: the liberal, the
corporate and the social-democratic regimes. In the liberal welfare-state regime, the state
constructs a safety net directed at the lowest-income population and it is carefully separated
from the free market. This is a system currently practiced in South Africa to house the lower
income groups. The other system, the corporatist, does not consider the conceding of social
rights harmful to the market mechanism. The corporatist system believes that rights are
attached to class and status, as also claimed by the Neo-Marxist scholars, but the philosophy
background is against the Marxian School of Thought. The Neo-Marxist believes that the
market mechanism controlled by the capitalist should not be the controlling factor for housing
provision. Whilst the soco-democratic system, rather than tolerate a dualism between state and
market, between working class and middle class, it pursues a welfare state that would promote
an equality of the highest standards not an equality of minimal needs as was pursued elsewhere
(Esping-Anderson, 1990). As pointed out by Brandsen (2001), the advantage of the welfare
state approach is ‘a safeguard against notions that housing systems will inevitably converge to
a single type’, which is far from the reality the global world has conceded to.
120
Likewise, a typical representative of the focus on the similarities is the neo-classical scholars
that concentrate on the dynamics of the demand and the supply of housing in a market
mechanism, which most time brings about trade-offs. The supposition is the universality of the
market mechanism and the identical behaviour of its participants. This has been criticized in
various ways. As asserted by Brandsen (2001), the critiques are mainly stressing three different
points in neo-classical economics theory which are: the neglect of historical and geographical
variations, the neglect of relationships other than economic ones, and the neglect of the active
roles of the participants, which has been seen to be a more sustainable way of delivery
development that affects the citizens of a particular state. This is a major tenant of the current
study. The debate between these two groups forms the basis of the methodological background
in housing studies. Hence, the next section explores the two most dominant methodologies that
have been discussed in varied ways, which are the positivist and social constructionist
methodologies that have been adopted over time in housing studies.
4.2.2.1 Positivist Methodology
Before social scientist and housing scholars shifted to the social constructionist approach in the
1990s, housing studies were dominated by positivist thinking. Positivism Methodology brought
the trend of paying little attention to theorization. According to Jacobs and Manzi (2000), the
Positivist Methodology in the UK for instance was decisively influenced by Fabianism.
Fabianism refers to a tendency in English thinking based on the technique of empirical research
that emerged in the late nineteenth century, which is essentially non-revolutionary, pragmatic,
and rational, with a belief in government intervention and the perfectibility of the welfare state
(Marshall, 1998). This method emphasizes the scientific qualities of housing research by
verifiable quantitative methods, which it thinks will convince the decision-makers.
Consequently, this methodology has moulded the motivation and expectation attached to
housing studies over time. Jacobs and Manzi (2000) argue that for many housing scholars, their
studies are carried out principally to improve policy practice; the expectation being that new
research can apprise policy makers in their efforts to resolve social problems. Additionally,
Jacobs and Manzi (2000) explain that while the absence of an explicit theory remains a defining
characteristic of mainstream housing research, it primarily relies upon a positivist
epistemology.
121
Positivist methodology was initiated from separate movements in nineteenth-century social
science and early twentieth-century philosophy (Kincaid, 1998). Fundamental positivist ideas
were that philosophy should be scientific; that metaphysical speculations are meaningless; that
there is a universal and a priori scientific method; that the main function of philosophy is to
analyse that method; that this basic scientific method is the same in both the natural and social
sciences; that the various sciences should be reducible to physics; and that the theoretical parts
of good science must be translatable into statements about observations (Kincaid, 1998). In
housing research and the philosophy of housing studies, positivism has supported the emphasis
on quantitative data and precisely formulated secondary theories.
According to Kincaid (1998), this assertion was criticized on several grounds because: the
theory/observation peculiarity is difficult to draw in any sharp way, efforts to translate
theoretical terms into observational ones often presupposed theoretical terms in the course of
describing the observational data. Also, theoretical terms can be applied in indefinite ways to
observations, and even if the theory/observation distinction could be drawn, every scientific
tests involves background theoretical assumptions, thus showing that observational evidence
has no absolute epistemic ontological status. The above criticisms have led to doubts about the
positivist idea of a unified science. Because if theoretical terms cannot and need not be reduced
to observational ones, then it seems implausible that the special sciences are reduced to physics
or that they must then be a good science. Kincaid (1998) further emphasizes that these
criticisms certainly undercut positivist doctrines in the social sciences. They have also led many
to conclude, somewhat implausibly, that any standards of good social science are merely
matters of rhetorical persuasion and social convention.
According to Wang (2004), the duty of the housing researcher within this paradigm is one of
discovering objective facts, presenting them in a descriptive format with the expectation that
policy makers will take notice and act accordingly. This ideology is what the current research
is hinged on. However, research within this pragmatic tradition achieves a level of complexity
in its analysis of social phenomena, but the primary purposes are to establish facts and to
recommend effective action once problems are acknowledged. Most housing studies have been
quantitative and problem-solving oriented over time. In turn, the positivist methodology
reinforces the ignorance of theorization in the housing study, or put another way, gives housing
study the characteristic of being conservative to the policy of the regime it is serving. Not
surprisingly, the conceptual categories used in housing research are seldom examined within
122
this paradigm; instead they depend on the collection of material evidence to support policy
recommendations.
The resulting research product is often methodologically conservative (Wang, 2004.). Besides,
it is difficult to pursue new lines of research or, for that matter, to develop different
conceptualizations of the policy process. Thus, the Positivist Methodology has had an impact
on the approach used in housing research over time. Housing issue debates tend to be steered
within an agenda dominated by two competing ideologies: either policy should be formulated
to strengthen market mechanisms, or the role of the state should be extended (Jacobs & Manzi,
2000; Wang, 2004). This still remains the situation today. Because government funds are
critical to housing research, housing studies generally are mostly conducted through the
positivist methodology with a heavy reliance on quantitative survey and secondary data
analysis, which was also adopted in the current research. Despite the criticism that has been
levied on this methodology, most research on housing is based on a positivist approach either
because of its roots in policy sponsored work or in economic or psychological analysis.
4.2.2.2 Social Constructionist Methodology
Since the late 1990s, increasing attention has been given to the inadequacy of the positivist
approach. This necessitated the implementation of the Social Constructionist Methodologies in
housing research. According to Jacobs and Manzi (2000) the re-emergence of research drawing
upon social constructionist epistemologies marks an attempt to enlarge the scope of housing
studies. Hence, more and more scholars have begun to make use of the methodological insights
offered by explanations rooted in constructionism.
A constructionist epistemology proposes that an individual’s experience is an active process of
interpretation rather than a passive material apprehension of an external physical world (Jacobs
& Manzi, 2000). Social constructionism is an amalgamation of different strands of work that
have different emphasis, although they sometimes share the same fundamental assumptions. It
is not a clear unified tradition of thought with an agreed research modus operandi (Clapham,
2009). In contrast, it is a dispersed field that draws on different traditions and in which there
are many differences of approaches (Clapham, 2009). Social constructionism as an
epistemology has its origin in a number of theoretical developments (not discussed in the
current thesis).
123
Jacobs and Manzi (2000) gave an in-depth account of the philosophical basis and development
of Social Constructionist Approaches in housing studies, arguing that a major claim advanced
by those adopting a social constructionist epistemology is that actors do not merely provide
descriptions of events, but are themselves constituents of wider policy discourses and conflicts.
Viewing society and social policy as malleable and subject to power struggles, constructionists
do not accept social facts as permanently ‘accomplished’. This emphasis on contestation is
important in offsetting any tendency by actors to objectify social phenomena or reify
abstractions into material realities. Social Constructionist therefore offers an altogether
different conception of reality from the one advanced by positivism, as well as a basis from
which to understand the contexts and the processes of housing. An important goal of
constructionist research is therefore to examine how certain issues become defined as
‘problem’ and to identify the collective strategies developed to confront these issues. Within a
housing context, Kemeny (2004) argued that what becomes a ‘problem’ is, to a considerable
extent, contingent on how interest groups compete with one another to gain acceptance of a
particular definition, whilst rejecting others. In this respect problems are constructed, as policy
makers attempt to establish their policy agendas in response to changing economic and social
conditions and in accordance with their own needs.
Clapham (2009) further discusses that though there are many aspects in Social Constructionism
housing research, which has largely been confined to four areas. The most popular being what
is termed the ‘social construction of social problems’. The emphasis on the ‘social construction
of social problems’ is trying to understand different definitions of social problems such as
homelessness by using investigative policy histories. Jacobs, Kemeny and Manzi (2004)
emphasized that the problems are perceived not to have ‘objective’ foundations, but are
constructed on shifting sands of public rhetoric, coalition building, interest group lobbying and
political expediency. The strength of the method is in its probing of existing ways of thinking
in policy (Clapham, 2009). One can infer that it opens the eyes of policy-makers to query
existing assumptions of policy.
The second area of social constructionist housing research is focused on interaction, but is far
less advanced than the social problems construct. For example, the work on the social
construction of housing management (Clapham, Franklin & Saugeres, 2000) has taken an
explicitly ‘interactionist’ stance. Also, what is being socially constructed through interaction is
124
a relationship and an occupational role, which is a variant of the social problems tradition.
Clapham (2009) argues that it is not a social problem that is being purported but a small part
of social reality as constructed through interaction between individuals in different positions in
social space. In this way, the analysis is uniquely social constructionist in a way that other
research as epitomised in discourse analysis may not be (Clapham, 2009). The symbolic
interactionist component of social constructionism places prominence on discovery of the life-
worlds of individuals and groups and describing the world as they see and experience it
(Clapham, 2009).
The third area of social constructionist housing research methodology is in international
comparative research. The highlight of this is on how and why social problems are defined
differently in different countries. This approach has led to a questioning of theories of national
housing policy that have laid emphasis on the convergence of policy and instead suggested a
notion of continuing separation (Kemeny & Lowe, 1998). There is the general notion of
culture, that countries should retain an individual housing structure that is in agreement with
the country’s political structure and in particular the type of welfare policy pursued by the
country. This approach builds on the philosophies of Esping-Andersen (1990) who identified
what he saw as the three worlds of welfare. Most housing research has sought to place housing
within this framework and to map the relationship between housing and other elements of the
welfare state (Clapham, 2009). Haworth, Manzi and Kemeny (2004) assert that a social
constructionist approach to international comparative housing research can abate the
‘ethnocentrism inherent in much research of this kind’.
The fourth type of social constructionist research has been in developing an all-inclusive view
of the housing field. Despite this area being part of the methodology, most social
constructionist housing research has been focused on specific research topics rather than trying
to describe a social constructionist view of the housing field and its relationship to other fields.
There are many general criticisms of Social Constructionism reflecting the many other
sociological approaches. A common criticism of Social Constructionism is its relativism. King
(2004) argues that there is a fundamental contradiction at the heart of the approach, informing
that if all discourse is socially constructed and there is no such thing as objective truth, then
Social Constructionism itself is only one discourse among many and cannot claim superiority
over any other approach. Another criticism is that humans only achieve their humanity through
125
social interaction. King (2004) further argues that social constructionism sees individual
subjects as empty vessels filled up through discourse. However, it should be noted that in
research associated with the meaning of home; there has been an understanding of the
importance of emotions in influencing the relationship between people and their physical
environment. Though people may have a strong emotional relationship with home, their
relationship to the physical fabric of the house is mediated through their physical
characteristics. Hence, without an understanding of the influence of embodiment on meaning,
social constructionism can offer only partial explanations. Clearly, the importance of the Social
Constructionist Methodology is that it moves forward to challenge the ‘given’ context of
previous housing studies, and tries to bring forward alternative understandings to the question.
Therefore, a housing study is no longer providing answers to given questions, but to challenge
and redefine the given question.
4.3 WHAT IS HOUSING POLICY
Anderson (2005) defined policy as a guiding principle used to set direction in an organisation.
It can be a course of action to guide and influence decisions. Policies are used as a guide to
decision making under a given set of circumstances within the framework of objectives, goals
and management philosophies as determined by management (Anderson, 2005). Likewise,
policy may also refer to the process of making important organizational decisions, including
the identification of different alternatives such as programmes or spending priorities, and
choosing among them on the basis of the impact they will have. Similarly, Jiboye (2011) states
that a policy can be understood as political, management, financial, and administrative
mechanisms arranged to reach explicit goals. Furthermore, Agbola and Alabi (2000) defined it
as a plan of action, a statement of aim and ideas. Anderson (2005) also asserts that there are
two types of policies. The first being rules: frequently used as employee policies, formulated
by the management of an organisation guiding the conduct of all employees. The second
is mini-mission statements frequently associated with procedures, which is the category the
housing policy falls into; which is a think rules versus missions. Policy is not usually used to
denote what is actually done; this is normally referred to as either procedure or protocol.
However a policy will contain the ‘what’ and the ‘why’, while, procedures or protocols contain
the ‘what’, the ‘how’, the ‘where’, and the ‘when’. Policies are generally adopted by the board
of an entity of governance within an organisation or a national state or an arm of the state.
Policies are thus well reasoned, carefully articulated and presented documents (Olatubara,
126
2002).
Thus, a housing policy is a guideline provided by government through the negotiation of
various bodies in a country as put together by the government, which is intended at meeting
the housing need and demand of the people through a set of suitable approaches including
fiscal, institutional, legal and regulatory frameworks (Agbola, 1998). Housing policies
provides a guide which defines action and sets goals and in most cases specify strategies for
achieving the goal (Jiboye, 2011). It further institutes guidelines and limits for discretionary
actions by individuals liable for implementing the overall plans of action (Olatubara, 2002).
Other scholars define it as a system of courses of action, regulatory measures, laws,
and funding priorities concerning a given topic promulgated by a governmental entity or its
representatives (Kilpatrick, n.d.). According to Duruzoechi (1999) some housing policy
decisions (written or implied) prompt the overall past work of government, whilst others are
goal statements or prescriptions of elemental rules for the conduct of personal or organizational
affairs. Housing policy is fundamentally necessary in any country as a guide or control on the
various actors in the housing sector.
Furthermore, housing policy as a governmental action is generally the principled guide to
housing action taken or to be taken by the administrative or executive branches of a state with
regard to a class of issues in a manner consistent with law and institutional customs. In general,
the foundation of any housing policy is the pertinent national and subnational constitutional law
and implementing legislation. Housing policy is sometimes embodied in constitutions,
legislative acts, and judicial decisions (Schuster, 2008). The main goals of any housing policy
according to Duruzoechi (1999) and Jiboye (2011) is to achieve the best possible use of existing
housing resources in order to ensure adequate housing for the people, guide the location of new
housing, and be responsive to the housing needs of ‘special people’ such as the low-income
groups.
Housing policies are typically legislated through official written documents. Housing policy
documents often come with the endorsement or signature of the executive powers within the
country or organisation to legitimize the policy and show that it is considered enforced.
Housing policies often have standard formats that are particular to the country or organisation
issuing the policy. While such formats differ in form, housing policy documents usually contain
certain standard components including:
127
1. A purpose statement, outlining why the policy is being issued, and what the desired
effect or outcome of the policy should be;
2. An applicability and scope statement, describing who the policy affects and which
actions are impacted by the policy. The applicability and scope may expressly exclude
certain people, organisations, or actions from the policy requirements. Applicability
and scope is used to focus the policy on only the desired targets, and avoid unintended
consequences where possible;
3. An effective date which indicates when the policy comes into force, however,
retroactive policies are rare, but can be found;
4. A responsibilities section, indicating which parties and organisations are responsible
for carrying out individual policy statements. Some policies may necessitate the
formation of new institutions or functions or actions to effectively execute the goals;
5. Policy statements showing the specific regulations, requirements, or modifications to
organizational behaviour that the policy is creating. Policy statements are extremely
diverse depending on the nation state or organisation and intent, and may take almost
any form;
6. Background, indicating any reasons, history, and intent that led to the creation of the
policy, which may be listed as motivating factors. This information is often quite
valuable when policies must be evaluated or used in a confusing situation; and
7. Definitions, providing clear and unambiguous definitions for terms and concepts found
in the policy document.
All housing policies usually have a cycle. A housing policy cycle is a system used for the
evaluation of the development of a policy item. The policy cycle is also referred to as a ‘stagist
approach’. A typical housing policy cycle includes the following stages: Agenda Setting
(Problem identification); Policy Formulation; Adoption of the policy; Implementation; and
Evaluation of the policy.
However, an eight step policy cycle approach as developed by Althaus, Bridgman and Davis
(2007) includes the following: issue identification; policy analysis; policy instrument
development; consultation (which permeates the entire process); coordination; decision;
implementation; evaluation of the policy. Policy cycles are usually considered as adopting a
classical approach. Hence some postmodern academics challenge cyclical models as
unresponsive and unrealistic, preferring systemic and more complex models (Young &
128
Enrique, 2009). Young and Enrique (2009) consider a broader range of actors involved in the
policy space that includes civil society organisations, the media, intellectuals, think
tanks or policy research institutes, corporations, lobbyists, amongst others. Housing policy
addresses the intent of the state or any organisation, whether government, business,
professional, or voluntary. It is intended to affect the ‘real’ world, by guiding the decisions that
are made. Whether they are formally written or not, most organisations and national states have
identified and formulated policies in almost all relevant aspects of the economy such as the
housing sector.
4.4 THE EVOLUTION OF HOUSING POLICY FRAMEWORK
Whilst there is no universally established definition of housing policy as observed from the
previous section, however, there are two established views of what housing policies need to
be. Malpass and Murie (1999) highlighted these two viewpoints as the static view and the
dynamic views. The first is associated with how things are done as a matter of routine,
characterised by general rules and conventions governing practice. It reflects recognised
positions on housing topics. This is very significant in many African countries, where there are
no specific policy documents, although the practice has changed in some countries, yet there
are entrenched practices in housing. The vigorous view of policy tends to be more prevalent
where there is overt action to resolve a housing problem. Malpass and Murie (1999) further
informs that this implies specific actions, relating to a problem defined in a specific way, in
pursuit of some objectives. This view of policy involves change, towards some end and also
needs a policy process. The policy process includes: problem formulation, planning, execution
and evaluation (Malpass & Murie, 1999). However, most of the housing policy formulated in
African countries either end up not being implemented or when they are implemented, they are
never evaluated. This present study is on the evaluation of the South African housing subsidy
delivery system and to profile solution towards the variables that are considered essentials in
the creation of sustainable and habitable human settlements.
Globally, housing policies have been used as an attempt to try and address housing problems,
especially with respect to the low-income earners, with the view of helping them access better
housing (Harris & Giles, 2003; Mukiibi, 2011). The evolution of housing policy in developing
countries has been studied and identified in different ways. According to Harris and Giles
(2003) the following phases have been identified by scholars as a definite time lag in the
129
evolution of housing policy: the period of public housing provision (1945-1960s); sites-and-
services (1972-1980s) and market enabling (1980s – present). However, Harris and Giles
(2003) further claim that this grouping is based on the policy recommendations that
international agencies recommended and focusses less on policies that nations pursued, as can
be seen from the current adoption of the Cities Without Slum Agenda and the Millennium
Development Goals (MDG) by developing countries. In each period, scholars have supposed
that the policies adopted by national governments in the developing world matched the
recommendations of international agencies. Harris and Giles (2003) also claim that the
assumption by the various researchers was based on ‘meager evidence’ and is challenged by
the earlier statements made by informed observers. For instance, it was observed that about 40
years ago the UN’s Bureau of Social Affairs observed that several countries in Latin America
had been providing public housing regardless of the UN’s preference for self-help (Harris &
Giles, 2003).
The established world organisations such as the UN and the World Bank, amongst others,
began to influence the housing policy agenda of nations in the developing world after World
War II. The most significant amongst these organisations were national agencies of the United
States and United Kingdom, inclusive of the United Nations.
Housing Policy Framework for planning about development has changed in important ways
over the last three decades. According to UN-Habitat (2006), the initial planning concepts of
the top-down strategies have gradually given way to the market and people-based solutions,
process approaches, and strong emphasis is repeatedly being placed on building capacities and
institutions through the policy. Thus, current housing policies of nations of the world have been
influenced by these dynamics. Table 4.1 shows a summary of housing policy developments
since the 1960s. The early focus of housing policy framework was on physical planning and
public housing, which quickly gave way, first to ‘self-help’ housing projects, which was used
to service the middle income households. This proved to be an unsustainable option over time
to address the needs of the poor due to the high subsidies that were involved. This later evolved
into the ‘enabling approach’ which concentrated on maximizing the contributions of all the
stakeholders in housing production within a supportive legal and regulatory framework
(Erguden, 2001; UN-Habitat, 2006).
However, in reality, the evolution of housing policy is never so neat nor linear as this, and there
130
are always instances of which agencies or governments seek to return to ways of doing things,
which have long-been discredited. This is principally the case where there is political pressure
to show quick results through, for instance, large-scale evictions of squatters or construction of
public or subsidised private housing (UN-Habitat, 2006). For instance, the Botshabelo Accord
of 1994 in South Africa, which sought support for the construction of one million housing units
each year through the commercial private sector is a good example. However, due to lack of
resources and neglect in policy to utilize rental housing alternatives which could mobilize
private capital, a mere ten per cent were actually built and none went to the poor (Bolnick,
1996) in the first two years; but the initial projection has since been exceeded.
Table 4.1: The Evolution of Housing Policy
Phase and
Approximate
Dates
Focus of Attention Major Instruments Used Key Documents
Modernization
and urban
growth: 1960s-
early 1970s
Physical planning and
production of shelter
by public agencies
Blueprint planning: direct
construction (apartment
blocks, core houses);
eradication of informal
settlements
Redistribution
with
Growth/Basic
Needs: mid
1970s-mid
1980s
State support to self-
help ownership on a
project-by-project
basis
Recognition of informal
sector; squatter upgrading
and sites-and-services;
subsidies to land and
housing;
Vancouver Declaration (Habitat I. 1976);
Shelter, Poverty and Basic Needs (World
Bank, 1980); World Bank evaluations of
sites-and-services (1981-83); UNICEF
Urban Basic Services
The Enabling
Approach/
Urban
Management
late 1980s-early
1990s
Securing an enabling
framework for action
by people, the private
sector and markets
Public/private partnership;
community participation;
land assembly and housing
finance; capacity-building
Global Shelter Strategy to the Year 2000
(1988); Urban Policy and Economic
Development (World Bank 1991); Cities,
Poverty and people (UNDP, 1991);
Agenda 21 (1992); Enabling Housing
Markets to Work (World Bank, 1993)
Sustainable
Urban
Development
mid 1990s
onwards
Holistic planning to
balance efficiency,
equity and
sustainability
As above, with more
emphasis on
environmental
management and poverty-
alleviation
Sustainable Human Settlements
Development: Implementing Agenda 21
(UNCHS, 1994)
HABITAT II:
1996
"Adequate shelter for
all" and "Sustainable
human settlements
development"
Culmination and
integration of all previous
policy improvements
The Habitat Agenda;
Global Report on Human Settlements
(UNCHS, 1996)
Istanbul+5 Review of the Renew commitments and Declaration on cities and other human Process and the
Implementation of seek/device more effective settlements in the new millennium(GA, Review and the Habitat
Agenda strategies 2001),Cities in a globalizing world, The Appraisal of GA state of the world’s cities (UNCHS,
2001) 2001
Source: UNCHS, 1995, pp15 and expansion to include Istanbul+5
In spite of the evolution, there is a common agreement today on the enabling approach in the
formulation of housing policy. However, UN-Habitat (2006) argues that changes continue to
131
show up between scholars who place more faith in markets to deliver both efficiency and equity
goals, and those who emphasize sustainable human development as an agenda within which
markets must be carefully managed. On the other hand, capacity-building for a developed urban
management, institutional reform especially in the public sector, and ‘local ownership’ over
policy decisions have significance in both approaches (Erguden, 2001). Both systems have also
identified the key roles of NGOs and other civil society groups in the housing process both as
care givers and in other roles, such as community participation and support; and both place
gender equity and other issues-of difference at the centre of policy choices (UN-Habitat, 2006).
4.5 FORMS OF HOUSING POLICY
Forms of housing provision can be defined by the processes through which such provision is
achieved. A useful systematic tool for identifying and examining these processes is the concept
of structures of provision which is based on the identification of social relations and
interactions of agents involved in all aspects of housing provision, such as, production,
exchange and consumption (Ball & Harloe, 1992; Healey & Barret, 1990). However, there are
a limited number of ways in which the governments can act to improve housing conditions,
especially for the low income and disadvantaged households universally. However, at times
governments can build (provide) housing inexpensively, usually for rent. They can help
households build their own homes, through the subsidization of materials and other help where
most needed; or they can try to make the housing market more efficient by delivering affordable
homes.
In any major city, there are many forms of housing stock through which people seek shelter.
For example, the forms in operation in South Africa, include private sector-produced housing,
catering to a range of income groups; public sector housing of various forms, frequently
housing more than one family; shacks in areas, which are both ‘legal’ and ‘illegal’; garages;
backyard shacks in formal housing areas; shacks in the relatively undeveloped landscape;
single quarter migrant hostels etc. Nevertheless, three main forms of housing policy exist
universally. They are: Public housing; Aided self-help; and Enabling the building industry
(popularly refer to as market enabling- a World Bank sponsored approach). In the next section,
these forms of housing provision mechanisms will be discussed briefly.
132
4.5.1 Public Housing
Public housing is a form of housing tenure in which the property is financed and owned by a
government authority, which may be central or local and operated by public housing
authorities, for low-income families and for elderly and handicapped individuals. For example,
in China, the government provides public housing through various sources, such as new
housing, abandoned properties, and old flats which are rented at a very low price called ‘low-
rent house’ or ‘low-rent housing’. Also, in Hong Kong, the government offers public housing
through flats, which are rented at a lower price than the markets, and through the Home
Ownership Scheme, which are sold at a lower price to the public, with a special emphasis on
the low-income earners. The houses are built and managed by the Hong Kong Housing
Authority and the Hong Kong Housing Society (Hong Kong Census and Statistics Department,
2006).
Public housing was established to provide decent, safe and affordable rental housing for
eligible low-income families, which would otherwise occupy housing units of very reduced
quality. This is supported by the notion that housing carries several positive externalities, for
example that housing affects children depending on the quality of the housing unit. It comes in
all sizes and types, from scattered single family houses to high-rise apartments for low-income
and elderly families. According to Leigh and Mitchell (1980), the public housing aim is not
only the provision of housing for low-income families but also for stimulating the economy
through the construction and finance sectors. In addition, Harris and Giles (2003) argues that
publicly sponsored construction, offered governments a means of nurturing a local building
industry, providing on-the-job training in the handling of modern construction materials and
methods. They further contended that public projects are effective vehicles for consolidating
political support, which is obvious in the South African low-income housing setting. The
medium makes it possible for governments to reward followers, whether by the judicious
allocation of building contracts and jobs, or through the allocation of constructed housing units.
Public Housing is named differently in different countries. In the United Kingdom it is often
referred to by the British public as ‘council housing’ and ‘council estate’, based on the historical
role of district and borough councils in running public housing. In Sweden, it is called the
Million Programme a term for an ambitious housing programme executed in Sweden between
1965 and 1974 with the aim of building one million new dwellings in ten years. In Canada,
133
public housing is usually a block of purpose-built subsidised housing operated by government
agencies, which are often referred to as projects or community housing. While in South Africa,
it was initially referred to as the Reconstruction and Development Programme Housing (RDP),
a major policy document developed by the pro-apartheid government in 1994 to bridge the
inequity that trailed during the apartheid government. It is now known as Breaking New
Ground (BNG) housing, named after the revision of the 1994 version of the Housing Act. The
provision of the housing is based on different programmes as directed by the National Housing
Policy. The present study is based on the residential satisfaction of the occupants that inhabit
these public houses. There are several methods of developing public housing, of which the
peculiarities of the system are mostly country specific and to a large extent determines if the
occupants’ will be satisfied with the houses received. For instance, Leigh and Mitchell (1980)
inform that in the US, four systems are mostly in use, which are: conventional method; the
turnkey method; a nonconventional means of acquisition of existing units; and Leasing. Whilst
in South Africa, there are different systems, which will be elaborately discussed in Chapter 6
of this thesis.
The United States was the first country to have an established public housing sector, with the
introduction in 1937 through the Housing Act also known as Wagner-Steagall Act. Since public
housing is limited, government all over the world usually set eligibility standard based on the
following: annual gross income; whether you qualify as elderly, a person with a disability, or
as a family; citizenship or eligible immigration status of the particular country. When the
eligibility statuses are met, the administration also check applicant’s background to make sure
that they will be good tenants in the rental option and in the non-rental option, the applicant
must have been on the waiting list, of which the names of all eligible persons are kept.
In general, occupants can stay in public housing as long as they comply with the lease in the
rental option. While with the other option of outright ownership without any contributions as
operated in South Africa, occupants are given ownership once the houses are allocated to them.
However, in some other countries, like the US, at the re-examination of a family's income, if
found that the income is sufficient to obtain housing on the private market, the Housing
Authority may determine whether the family should stay in public housing or not. However,
no occupant is required to move unless there is affordable housing available on the private
market (Leigh & Mitchell, 1980).
134
However, this system of housing provision is not sustainable to the government of nations
because of the numerous limitations that were encountered. A major problem with public
housing delivery is its cost to the state. Also, there is the problem of the lack of skills and
resources to handle the production rate. As a result of this, international agencies and various
countries that had adopted the form of housing provision, soon concluded that public housing
could not solve the housing problem (Harris & Giles, 2003). Hence, Atkinson (1960) concludes
that public housing is not a solution to the general housing problem. However, Harris and Giles
(203) assert that an International Labour Organisation (ILO) study claimed that public housing
was fine ‘in principle’ but that high costs often make it a ‘low priority’ to government of certain
nations.
4.5.2 Aided Self-help
The term aided self-help was coined in about 1948 by Jacob Crane as a result of the limitations
that surrounded the public housing provision. Aided self-help housing generally revolves
around the idea that governments might help families to build their own homes (Harris, 1998).
After 1945, aided self-help housing was propagated by agencies of the United States, in the
first place by Jacob Crane and later by the United Nations and the British colonial office.
However, Crane (1949) used the expression ‘aided self-help’ for the first time in 1945, and
linked it at first with rural projects and equal to ‘minimum urbanization’. Crane (1949) claims
that the government can build very few excellent houses which will actually accomplish almost
nothing as measured against the problem; or undertaking an ‘aided self-help’ programme. Early
advocates of self-help apart from Jacob Crane are Abrams, Mangin and Turner who mostly had
positive views on the families’ power to build the house through self-help. Self-help provides
household with a medium of upward mobility (Bredenoord, 2010). Initial experiences with
aided self-help housing took place mostly in Western countries and were based on the notion
that governments should assist families to build their own dwellings. Throughout the
developing world, millions have been building their own homes, where they have title to land
without government aid. Generally, the people’s power to construct their homes with or without
assistance happens almost everywhere in the world especially in the developing countries.
Aided self-help housing was practiced in South Africa, before World War I. A site-and-service
scheme with plots and services was first developed at Pimville in existing Soweto near
Johannesburg. The ‘sites and services approach’ was later revitalized after World War II in
135
South Africa. Massive land invasions and squatting during the 1940s obliged the city of
Johannesburg to convert many into controlled site and service camps (Harris, 1998). However,
after the abolishment of the Apartheid rule, self-help was later replaced with the People
Housing Process (PHP), which has now been enhanced to address limitations that were
observed in the first twelve years of implementation. Aided self-help housing is usually
implemented to help alleviate housing problems at minimum costs. According to Bredenoord
(2010) aided self-help housing was above all liberal, having a social component, but
predominantly perceived in contrast to public housing.
Bredenoord (2010) informs that the ‘aid’ or ‘assistance’ is usually referred to as: land for
construction, urban services, knowledge-development and the option to construct a house step-
by-step which is also called the ‘incremental’ housing process. However, the scarcity of land
and the increasing land prices make sites-and-services projects relatively hard to develop. For
instance: where can suitable and sufficient land b found to be divided into parcels and to be
sold or allocated for self-help housing? Another concern is how to improve individual houses
and even whole neighbourhoods without constraints? Moreover, the growing demand for new
urban land is causing financial and organizational difficulties for local governments, which
necessitated some countries to formulate urban land laws. A major feature of self-help is the
freedom a shelter offers; if it was constructed without using loans, the responsibility to pay-off
every month is restricted or even zero, which is the best for poor and disadvantaged families.
Self-help is usually linked to informality, since the possibility of realising a mass public
housing system is difficult with all the limitations. Also, because the customary housing sector
cannot provide enough dwellings for the low-income groups, most have to resort to the
informal housing market, where the self-help principle is a significant feature (Bredenoord,
2010). However, self-help housing is present almost everywhere, in formal as well as informal
developments, except in countries that have shown a strong economic development and where
large scale or industrial development is adopted. In real life self-help housing is present: it
happens on a larger scale and is unavoidable for many in developing countries where there is
little or no support from the government, despite the criticism and constraints.
The advantages that comes with the aided self-help housing are: it costs much less than public
housing; it stimulates owner-occupancy, which many believed would boost social stability,
give people pride in their homes and a stake in the society, whilst encouraging savings and
136
investment (Harris & Giles, 2003). furthermore, it is a way of assuring residential satisfaction
since occupants are involved in the construction and will accepted any defect since it came
from them, compared to the allocation of public housing that has already been constructed.
During the implementation stage of the aided self-help housing form, owner-builders use
methods and materials with which they are familiar, unless obliged to do otherwise. Owner-
builders produce homes that most experts’ regards as temporary and that some judged as
substandard. Critics of this housing forms charged that self-help produced instant slums.
Hence, Ward (1982) argues that self-help housing failed to become a significant housing
solution in most countries of the South because of this limitation. Also, Burgess (1982) claims
that there are six constraints of self-help housing and Marcuse (1992) even claimed they were
ten. With time, it became clear that the vision and the power of self-help cannot be over-
valuated (Bredenoord, 2010). Other opponents further argued that the process of learning with
self-help would be ineffective and that most families can only master the process, when they
have almost completed the house (UN-Habitat, 2005). However, families having some
experience with initial self-help housing, have a better chance of managing construction related
work. It is obvious that the disabled, single-parent households and elderly people cannot
participate, except getting help from others. A good condition for the success of self-help is if
people with the same attitude take the initiative and receive added help from the government
or support organisations. Another problem of self-help housing is that the initial target group
is not always fully serviced, since aided self-help housing increases and becomes more
exclusive.
UN-Habitat (2005) asserts that self-help housing is the most affordable and intelligent way of
providing sustainable shelter. Because it is based on minimum standards and incorporates a
substantive amount of ‘sweat’ equity and can be cost-reducing. Also, it can be useful because
individuals and communities engaged in it, acquire valuable skills, and it can be practical
because it answers to people’s needs and levels of affordability.
4.5.3 Market Enabling Strategy
Housing provision can also be promoted by helping the building industry to become more
efficient at providing decent and affordable housing. Harris and Giles (2003) state that this
strategy may involve the provision of better building materials and methods; the training of
tradesmen and entrepreneurs; assistance to small savings and mortgage lending institutions;
137
and efforts to guarantee secure tenure. The form of housing provision through the private
market mechanism has been on the agenda of housing studies since the mid-1970s (World
Bank, 1975). Throughout the 1980s, through the recommendations of scholars and housing
practitioners, there was an almost universal acceptance of reducing the role of the government
in direct provisory roles in the economy and increased reliance on the private sector (World
Bank, 1988). In different capacities, most countries irrespective of ideology, political structures
or levels of development have pursued this policy (Keivani & Werna, 2001; World Bank,
1988). Subsequently, the extension of the role of the private market in the provision of housing
has become increasingly the focus of attention.
This strategy was particularly promoted by the World Bank and its associated writers during
the 1980s who developed an enabling strategy for public sector support of private market
activity in housing provision particularly in developing countries (World Bank, 1988). The
World Bank (1984) paper on ‘Housing and Financial Institutions in Developing Countries’
policy document set the tone for this strategy universally. The document lay emphasis on the
need for developing countries governments to encourage financial innovation in the
provision of housing finance to households on a financially viable basis, as far up as the
upper level of the informal sector and to develop specific programmes for the lower
income households, who could still contemplate some form of house ownership with the
major focus centered on urban infrastructure and services (World Bank 1984). These
concerns were taken over by the Vienna Recommendations in 1986, and from there key words
like ‘enabling environments’, ‘co-operation’, ‘efficacy’ and ‘financial sustainability’ were
developed. The responsibility of governments was to shift from being a housing provider to
policy ‘enabler’ in order to incorporate shelter as an integral part of the national macro-
economic plans. By the late 1980s, this form of housing delivery was widely promoted as the
‘enabling strategy’ (World Bank, 1993). Harris and Giles (2003) state that the strategy was new
to the housing delivery sector with the notion that a bundle of measures might be defined as a
coherent strategy. Taken individually, however, many of the interventions themselves were
not, and neither was the underlying assumption that governments should help the market work
its ‘magic’ (Harris & Giles, 2003).
The ‘Market Enabling Strategy’ approach meant that housing provision policies were to
become self-sustainable, and therefore all participants involved in the process should
interact, co-ordinate, re-organize and revise their roles towards the new approach (Duran,
138
1995). The role of communities in the decision making process was emphasized so that they
could determine their own set of priorities and the pace and extent of the development process.
Likewise, at the national level, the government’s role is to lean towards policy-making and
institutional support, which is aimed at concentrating on the highest priority needs, while the
private sector is to play a leading role.
Although the need for recognizing the importance of supporting and encouraging the
construction sector has been recognised by many developing countries, few questions have
been asked about its implications upon the poor, who create the biggest demand for
houses in many developing countries. This policy could leave the poor completely dependent
on the success or failure of the free market and the private sector. This is because the approach
encourages governments to ‘enable’ the housing market to work but they also propose the
gradual abolition of subsidies (Duran, 1995). The ability of a free market economy to provide
cheap housing with reasonable profit margins remains to be proven. For instance, in South
Africa, subsidies are the only means available for the lower income sector to acquire housing
as a result of the socio-economic problem facing the lower-income groups. This approach of
the enabling strategy, however, has been subject to much argument and criticized for its over-
concentration on the private markets and rejection of alternative/complementary modes of
housing provision from serious policy consideration (Keivani & Werna, 2001). While private
markets can and should be supported they cannot form the attention of the enabling strategy in
most developing countries. As an alternative, a comprehensive approach to enabling strategies
which combines adjustments to overall supply and demand conditions with the identification
and inclusion of different modes and agents of housing provision in a holistic integrated policy
will better serve the dire need of housing provision in the under-developed and developing
countries of the world and most especially to the low-income group and the disadvantaged.
4.6 OBJECTIVES OF HOUSING POLICY
It is impossible to put emphasis on the importance of creating a housing policy that will provide
a framework for affordable and decent living environments for all people. As discussed
previously, housing impacts on almost all dimensions of an individual and the family’s life.
Dewar (1996) claims that from a societal approach, widespread conditions of inadequate
housing contribute considerably to social instability, as witnessed by the ‘tragic violence in
South African black townships, and to general the attitudes of despair and de-socialization’.
139
A major feature that makes the issue of housing multifaceted is that ‘housing’ does not simply
relate to the provision of shelter. This is because; in obtaining shelter the individual gains access
to a number of different products. Dewar (1996) asserts housing has the potential to contribute
to an improved quality of life and this should direct the formulation of a housing policy. Most
countries universally set their housing policy objectives based on the current housing situation
prevalent in the economy at that time. Thus, there is no country with the same wordings of
housing policy objectives, but each has unique features that most times follow in the same line
of thinking as proposed by the international housing agencies responsible for monitoring of
shelter provision. However, the main products associated with a housing policy decision, and
the objectives which should be associated with them in an idea situation are access to land,
access to a good habitable socio-economic location, access to adequate services, access to
adequate shelter and access to an adequate external, social and physical environment. All of
the factors have some relevance in any housing policy formulation and do not represent a
sequential list of priorities, which must be satisfied on a one-to-one basis, but all must be
present to some degree or other for any housing policy to succeed, which is the reason for its
formulation. If all are present, and all of the objectives satisfied, the situation would be optimal.
Dewar (1996), state that it is obvious that the decisions facing individuals in the arena of
housing are intricate and priorities will vary widely with individual circumstances. It follows
from this that no single ideal approach or package exists and that centralized, external agencies
cannot decide on priorities. The clear implication is that more individuals can determine their
own priorities, the better the situation and the greater the range of choices available to people,
the more they are assisted. Thus, it should be recognised that housing policy objectives require
a framework which locates the housing issue within its broader urban context - a framework of
broader issues, which should inform the direction of housing policy in the specific context of
any country.
4.7 THE PURPOSE OF HOUSING POLICY
The effectiveness of a housing policy is about the ability of the policy to achieve its purposes.
The effectiveness of a housing policy, therefore, cannot be judged in isolation from what it is
expected to do. It would be completely wrong to design a new housing policy or to propose
changes to an existing policy or to implement a housing policy without a clear view about the
purpose of the housing policy (new or revised). This should logically be examined within the
140
context of the overall purpose of housing policy and the place of the entities overall goal within
the policy. Hence, the primary purpose of any housing policy should be to provide the whole
population adequate and secure housing for all by unblocking and unleashing all potential
energies and resources, from a wide variety of sources, which can play a role in improving
living conditions. Possible but not exhaustive lists of housing policy purposes according to
Oxley (2009) are:
1. help low income groups access decent housing;
2. help low income households have adequate post housing expenditure incomes;
3. improve the quality of housing consumed by low income groups;
4. increase housing choices for households with unmet housing needs;
5. increase the supply of housing in the society;
6. improve the quality of urban neighbourhoods;
7. improve the functioning of urban labour markets;
8. promote community cohesion;
9. improve the functioning of the macro-economy; and
10. promote environmental sustainability.
In most nations of the world, housing policies are becoming increasingly integrated with a wide
range of social and economic objectives, which mean that housing policy is reaching further
down the list above than was the case in past decades. A housing policy that has broad goals is
much more multifaceted than one that solely emphases on housing low income groups (Oxley,
2009). Therefore, any finance and other economic systems that will enhance the actualization
of the policy must be compatible with the goals of policy in the given country. When housing
policy focusses on the purpose of helping low income and disadvantaged groups to access
decent housing, the setting of the appropriate standards for decent housing is a key element in
the success of the policy. Hence, meeting housing needs for decent housing in that particular
context will mean that governments will have to have policies that will bridge the gap between
what is needed and what is demanded. If the standards of decent housing are set too high and
what is needed is too great, housing policies will be extremely expensive. It has been argued
that setting standards at inappropriate levels has been one of the failures of policies in
developing countries (Habitat, 1994, UN-Habitat, 2009).
141
4.8 HOUSING POLICY INSTRUMENTS
Housing policy instruments may be defined as the set of techniques by which governmental
authorities wield their power in attempting to ensure support and effect or prevent (Vedung,
1998) the provision of affordable housing to the citizens. Over time, ranges of housing policy
instruments have been developed to respond to policy imperatives of housing delivery.
However, most times, it is not anticipated that any instrument will be discontinued or
terminated, but when the instrument can no longer meet the need it was designed for, they are
usually discontinued, modified or terminated. Rationally, most existing instruments that are
not meeting the policy goals are sometimes supplemented by additional instruments to provide
flexible solutions to demand-side needs. No one singular policy instrument has been found to
fulfill the housing policy goals of a nation’s housing policy. Most housing arrangements
include a mix of conditional subject and conditional object subsidies (Oxley, 2009). When
housing problems are regarded as demand side affordability problems, the greater is the
preference to use conditional subject subsidies (Oxley, 2009). Likewise, the more the emphasis
is on supply side housing shortage problems, the greater the emphasis is likely to be on
conditional object subsidies. Although housing finance systems can support housing suppliers
directly by means of conditional object subsidies, they can also support suppliers indirectly by
conditional subject subsidies that underpin the rental revenue stream (Oxley, 2009).
A fundamental mix of housing policy instruments aimed at securing greater provision of
housing stock ranges from: rent regulation; allocation and rental policies in current social
housing; support for the construction of new social flats provided by municipalities or non-
profit housing associations; housing allowances; tax relief and interest subsidies for ownership
housing; and housing subsidies for special social groups in the society etc. According to Lux
(2003), social housing represent the subsidies aimed at decreasing the cost of housing, whilst
housing allowance represents the subsidies aimed at increasing income of households from the
pillars of public housing policies in most countries housing policy instrument. Oxley (2007)
informs that housing allowances are sometimes regarded as demand-side subsidies and social
housing as a set of arrangements that involve supply-side subsidies. Oxley (2007) further
argues that, given that they go to individuals, the demand-side subsidies are also termed subject
subsidies and the supply-side subsidies, given that they support buildings, are termed object
subsidies. However, there are, in practice, very few examples of either pure subject or pure
object subsidies. Pure subject subsidies represent income supplements with no housing-related
142
conditions attached, where households would be able to spend the additional resources on
whatever they wished. Pure object subsidies are used to build new dwellings without any
conditions about who occupied the dwellings and how they were priced (Oxley, 2007).
There are many variants of conditional subject subsidies and conditional object subsidies found
throughout the world. It is the conditions that make them unique in any set of environments
and is the conditions that make them successful or not successful. According to Galster (1997),
it is ultimately the conditions, how they are used and what effects they have, that is at the
essence of the choice of policy instruments. For instance, one housing allowance system with
one set of conditions is very different from another with another set of conditions. The
conditions attached to housing allowances usually include considerations of the size of the
household, household income, and housing costs (Galster, 1997; Oxley, 2007). There may also
be conditions associated with the size and quality of the housing occupied, which is mostly
determined by the allocated amount of finances for each respective building. The relationship
between the amount received and these items can vary within a country. There are thus
occasionally locational elements to the conditions. The conditions also specify who gets the
resources; the household or the housing supplier. The application of the conditions turn what
is superficially termed a demand-side subsidy into a measure with important supply-side
features. The conditions usually have substantial influences on the way that housing allowances
affect housing quality and quantity.
Most times, new housing policy instruments tend to focus attention on sectors which have been
previously neglected. These instruments are inclined to place greater emphasis on flexibility
and responsiveness to local circumstances particularly the physical context within which
housing is to be delivered. This usually is a natural consequence of an increased focus on the
development of sustainable human settlements as opposed to the delivery of
‘commoditized housing units’ (Breaking New Ground, 2004). For example, in South Africa,
there is consequently a greater emphasis on the process of housing delivery emphasizing
planning and engagement, the quality of the housing product both in terms of location but also
in terms of final housing form and the long-term sustainability of the housing environment
leading to a focus on institutional capacity, which were observed as the problems with housing
delivery in the first decade of applying different housing policy delivery instruments when the
new housing policy was formulated in 1994. Nevertheless, there is no explicitly ‘best’ housing
policy instrument in all circumstances. Every policy instrument has its own comparative
143
advantages to a degree as determined by the particular country housing market context and by
the goals of the particular housing policy.
4.9 CONCLUSION
This chapter has outlined a number of issues upon which housing research studies should rest.
Housing by its nature is multifaceted as observed from the literature. It consumes natural
resources and produces impact on the natural environment. It constitutes a major economic
activity and impacts on the general economy. It is an important component of social
development and quality of life. It is often used by government to achieve political and
economic end. It is also a cultural attribute, manifesting the aesthetic value and the way of life
of man in his particular setting. Thus, a holistic perspective of housing studies is therefore
needed if we wish to chart the future of housing development. Although housing research has
tended not to engage with the theoretical debates about methods and focus. Rather research has
either been based on the general shared assumption or the analysis has been based on one
particular approach. The most important point to stress in conclusion is that the central
objective of housing policy must be the stimulation of environments which give dignity to
people’s lives. It is not simply the provision of shelter. Against this criterion, the record of
housing policy and implementation in South Africa in recent decades has been really poor.
Thousands of millions of Rands have been spent on housing development but the environments
which have resulted are almost unfailingly sterile, monotonous, hostile and inconvenient. In
order for the money spent not to be wasted, there is a need to evaluate the residential satisfaction
of the housing occupants so that errors made can be corrected for future development. A narrow
focus on the individual housing unit and the provision of shelter, which is the prevalent
disposition, gives rise to a particular mind-set and approach which ensures the generation of
poorly-performing, sterile environments. Significant improvement demands a paradigm shift-
a shift which places not the individual unit but collective spaces, institutions and facilities at
the centre of ‘housing research’. The next chapter will focus on the review of literature in
developing countries, with special emphasis on two West African States, namely Nigeria and
Ghana.
144
CHAPTER FIVE
HOUSING IN DEVELOPING COUNTRIES – AN AFRICAN
EXPERIENCE
5.1 INTRODUCTION
Provision of adequate housing is one of the major challenges faced by most African countries.
In this chapter, housing policies and other housing issues in the African countries of Ghana and
Nigeria will be discussed. The roles played by the different bodies such as government, non-
governmental organisations in the provision of housing will be looked at. This is because in
most West Africa countries like Nigeria and Ghana, the ownership of affordable good quality
housing has been a problem. This justifies serious public and private sector intervention
(Cudjoe, 2010). Unlike South Africa, most African countries have refused to include the right
to housing in their national constitution, knowing all too well the role of housing in national
development. This chapter sets out a background review on housing in developing countries
and a summary of the lessons learnt to date from the literature presented.
5.2 HOUSING IN DEVELOPING COUNTRIES
The need to provide adequate, suitable and equitable housing has remained a major priority of
every government in the developing world. Hitherto, since the problem of housing is complex
and pressing, no developing nation has been able to provide adequate housing of an acceptable
standard for all its citizens. Besides, it is worth noting that the minimum housing standard
differs from one country to another, depending on geographic and economic conditions. The
United Nations, recognizing the seriousness of the housing problem, declared 1987 as the
International Year of Shelter for the homeless. This was in recognition of the fact that there are
no encouraging signs that the housing problem for the world’s population in the developing
world would be solved soon.
Housing is usually regarded as the most valuable asset for all people in the developed world
and most especially in the developing countries. However, some scholars have advanced the
argument that it is incorrect to assume that housing is a priority for everyone (Alder, 2002;
Chambers, 1995) not even in the developing countries. Housing has tremendous social and
145
economic impact on the total living environment of the world. It has direct and immediate
influence on health, education, economy, environment, political and social life of any society
(Sinha, 1978). However, a significant number of people in the developing countries tend to
prioritizes housing as their most urgent need. It must however be acknowledged that income
generation or livelihood is usually the highest priority for people in developing countries
(Skinner 1989), but often that is linked to housing, the house being also a workplace in many
developing countries. Thus, homeownership assumes high priority and offers home-based
subsistence livelihood activities (Ahmed, 2011). But the shortage of housing in developing
countries most especially for people of low-income groups has been termed a problem of
‘colossal magnitude’ according to Alaghbari, Salim and Ali (2010). The extent of the housing
needs of the populace in these countries rises phenomenally by the day on account of rapid
population growth and urbanization occurring in these countries, and the lack of a
commensurate increase in housing stock (Olotuah, 2006). In many developing countries, urban
housing crisis is growing relentlessly even though a number of new policies, programmes and
strategies are being engaged by the public and private sectors in addressing this problem (Ibem,
Anosike & Azuh, 2011). The governments of the developing countries have recognized that
the bulk of those in dire need of housing are in the low income categories and that some require
special housing programmes to be able to live in decent housing, outside of the programmes
currently in place.
In developing countries, the low-and moderate-income majority build their own homes
incrementally over a period of five to 15 years, largely without the support of formal-sector,
private and public institutions (Ferguson, 2001). Ferguson (2001) further states that, the term
‘housing’ in the developing countries is used as a verb because households must actively
perform most of the tasks to gain access to land and construct adequate shelter during a longer
time period, regardless of the many interventions that have sprung up in most developing
countries. Whereas, the term ‘housing’ has become a noun in the high-income industrialized
countries, because it is a product delivered mainly by a sophisticated network of private firms
and public institutions. Most developing countries lack social safety nets of all kinds as reported
in most literature. The low-and moderate-income households constitute a greater majority in
these countries, which live in a world of few windfalls and many lay-offs from jobs and income
changes, disease, and other events (Ferguson, 2001). In developing countries, a home of one’s
own represents a precious refuge. Indeed, literature on housing in developing countries often
shows that housing ranks above education and health services as a priority. Thus, households
146
in developing countries value homeownership more than households in advanced industrialized
countries. In fact, in most developing nations when a family does not have a house they can
call their own, that family is regarded as the poorest of the poor.
However, acquiring access to a home and to the components that comprise housing and housing
policy- land and property rights, building materials, basic services, regulations, subsidies, and
credit- are extraordinarily difficult for most households in developing countries (Angel, 2000).
Poor land records and dysfunctional legal systems and regulatory bodies typically cloud
ownership rights for a large number of households (Ferguson, 2001). De Soto (2000) claims
that investment in housing in developing countries has been severely hindered by regulation,
and that investment in informal housing has been further slowed down by confusing property
rights, and that titling land with ambiguous property rights will sharply stimulate investment
in low- and middle-income housing. De Soto (2000) was quite right with the first point, as
showed by the inelasticity of the formal housing sector supply and the high price of titled land
in most cities in developing countries. With regards to the second point, he may or may not
have been right, but on the third point he was largely wrong. This is because titling by itself
appears to do little to solve land market problems. According to Arnott (2008), land titling
process is costly and time consuming; titling land that is illegally occupied raises legal and
compensation problems; titling may conflict with traditional property rights; and titling a
property is not enough to obtain a mortgage.
De Soto’s (2000) idea of making capitalism work for the poor through formalizing their
property rights in houses, land and small businesses is very logical, but many of his policy
recommendations may be inappropriate for the poorest and most vulnerable, and could have
negative impacts on their security and well-being. Because titling does not necessarily increase
tenure security or certainty, in many cases, it does the opposite. Also, formalization of property
rights does not promote lending to the poor rather than turning their property into ‘capital’,
formalization could increase the rate of homelessness. Formalization through registered title
deeds creates unaffordable costs for the poor in developing countries. However, developing
countries’ informal property systems support a vibrant rental market, so formalization could
undermine this, producing unintended negative consequences for the poor and disadvantage.
One of the major features of the developing countries is the increasing inequality between the
urban and rural areas. Igboeli (1992) argues that this feature has its roots in the neo-classical
147
economic theories, which acknowledged that development can be accelerated by concentrating
investments in the cities and that rural poverty will be ameliorated by the trickle down of
benefits from the urban industrial growth. However, Ikejiofor (1998) asserts that with the so-
called growth-centered strategy, the developing countries have continued to witness
imbalances in the living conditions between the urban and rural dwellers. Thus, development
theories over the years have been searching for alternative strategies that would not only
accelerate growth but also spread the benefits of development to all areas. Therefore, much of
the housing crisis in the developing countries is predominantly an urban one. The rural people
living at reasonably low densities are often able to meet their own housing needs, though at a
less sophisticated level than city dwellers (Akpomuvie, 2010). According to Adeniyi (1985)
and Akinjo (1984) it is in the cities where the legal acquisition of land is far beyond the means
of most people where the traditional forms of housing are often not acceptable and where the
population is raising that the problem of housing is more acute. Furthermore, housing in
developing countries is characterised by some general features. The first is the relatively high
house prices to income ratio. Malpezzi (1990) informs that in countries with less elastic supply
for whatever reason, asset prices usually a bit up. Malpezzi (1990) further commented that the
poor housing finance performance is responsible for the inelastic housing supply and the
resulting high housing price to income ratio. Most developing countries’ housing stocks are
dominated by rental tenant units. About 40 percent of the world’s urban dwellers are said to be
renters; with two-third of the developing country cities, housing stock being in rental (Farzana,
2004). Another feature of housing in developing countries is the domination of the informal
housing supply because of the deficiencies in formal housing. Formal housing is that which
have legal approval of the planning agency prior to their development and has been developed
within the framework of government housing policy rules, regulations and controls and meet
the minimum required standard of environmental quality and infrastructure. While informal
housing is illegal and is composed of unauthorized colonies and squatter settlements, these
have mostly emerged because of non-availability or unaffordability of housing in the formal
housing market (Farzana, 2004).
According to Ferguson (2001), the building materials industries in the developing countries
frequently suffer from cartelization and inefficient production methods that result in high
prices. The Provision of water, roads, drainage, and electricity occurs at a low level
equilibrium. Government usually provides poor, incomplete services, whilst households refuse
to pay a substantial share of these infrastructure costs. This vicious circle greatly limits the
148
delivery of the infrastructure departments and companies, to extend their services to new areas
and improve their services in existing areas. Most national agencies in some developing
countries set high building and subdivision standards, attempting to regulate local land use, and
imposing transfer and other taxes on real estate and mortgage transactions as an easy means of
raising revenue, which has been a problem of housing development in this region (Ferguson,
2001). Ferguson further stressed that the greatest setback, however, often occurs with accessing
credit. Widespread access to long-term, competitively priced mortgages has revolutionized
housing in high-income industrialized countries over the past 60 years as opposed to that of
developing countries. An overwhelming share of household in these countries has now
acquired homes with market-rate mortgages from private sector financial institutions. The
reality of mortgage finance is the opposite in developing countries. Only a small minority of
households-typically less than 20 percent of the population obtain a mortgage to finance their
homes. This system is gradually changing as compared to the last twenty years in most
developing countries. Regardless, many of the mortgages receive substantial subsidies from
government in one form or another. Once these subsidies are taken into account, the private-
sector mortgage market in which market rate intermediation occurs often is extremely small
(often less than five percent of new household formation) or is missing entirely (Bruce, 1998).
Various factors lie behind the low levels of mortgage finance in developing countries. Much
of the blame is attributed to inflation. High and explosive inflation destroyed the existing
mortgage finance systems of many countries in the 1980s and 1990s, like in Nigeria, Ghana
and Kenya. Likewise, the re-payment of international debt has contributed to the stabilizing of
inflation and interest rates in some of these countries, but a considerable improvement has not
been shown in the housing sector of most developing countries. This macro-financial stability
provides the opportunity for reconstruction of mortgage finance systems, as is currently being
done in countries, such as Nigeria and Ghana.
In spite of the remarkable increase in literature on the demand for housing attributes in
developed countries over the last decade, there exists little or no scholarly studies estimating
the parameters of the demand for housing attributes in developing countries. Arimah (1992)
states that very little is known about the working of housing markets in this region and that the
state of affairs of housing in developing countries could be attributed to the difficulty associated
in obtaining data on the operations of the housing market in developing countries. The apparent
belief that housing markets in developing countries are inhibited by socio-cultural and political
institutions raises doubts as to the applicability of microeconomic models to such markets; and
149
the fact that data from such markets is unreliable. Nonetheless, it is still possible to undertake
an analysis of the demand for housing in a developing country. This is because public policy
in terms of government intervention, which is the often-cited reason for housing market
imperfections, in the form of public housing programmes and rent control, is quite ineffective
in preventing substantial market transactions (Koenigsberger, 1986). Also, most policymakers
and executors in developing countries have a genuine lack of understanding about the
operations of the housing market, as well as the fact that housing programmes in developing
countries aimed at improving the huge housing problems are usually formulated based on ad
hoc notions of ‘needs’ and ‘standards’ without due consideration of the actual demand for
housing (Arimah, 1992). This has led to the general failure of housing programmes in many
developing countries.
Around the world and most especially in the developing countries, no nation can claim to have
solved the housing problem of their people as shown by various authors’ report that highlighted
various countries related housing issues. The following examples illustrate the housing
shortages that prevail:
In Ethiopia, the Ministry of Works and Housing (2008) stated that studies conducted in
the last five years found that a housing shortage of between 900,000-1,000,000 in urban
centres and only 30% of the existing urban housing stock is in good or fair condition;
For the Metropolitan Region of Sao Paulo (MRSP), the urban housing deficit is
approximately 611,936 units (UN-Habitat, 2010);
The housing shortage in Nigeria is estimated to affect between 14 and 16 million people
(UN-Habitat, 2008a). Mabogunje (cited by Kabir and Bustani, 2009) indicated that
R600 billion (N12trillion; Nigerian naira) will be required to finance the housing
deficit;
For Pakistan, in 2008, the yearly estimated housing demand was 570,000 units. Actual
supply was 300,000 units, leaving a shortfall of 270,000 units every year. The
consequence of this situation is that almost half of the total urban population now lives
in squatters or informal settlements (ICA, 2009a);
In the year 2007, the housing deficit in India was estimated to be 24.7 million houses
in urban areas and 15.95 million houses in rural areas, totaling 40.65 million units
overall (ICA, 2009b);
150
Bell (1996) emphasized that the burden of cumulated housing shortage in Algeria is
still high and it is expected to reach nearly two million dwellings by 2025 but was
estimated at 763,176; 994,357 in 2010 with a population of 33.8 million in 2007;
In Mexico, Centro de Investigacion Documentacion de la Casa (CIDOC) and Sociedad
Hipotecaria Federal (SHF) (2006) established that 1.8 million new housing units and
2.7 million housing improvements are needed in a country with a population of 103.3
million people;
In Kenya, Government of Kenya (GoK) showed that the country has a deficit of units
of 127,700 in urban and 303,600 in rural areas;
In Uganda, the Uganda Bureau of Statistics estimates that Uganda has a housing deficit
of 550,000 units. About 160,000 of this backlog is in urban areas. Kampala alone has a
housing deficit of 100,000 units.
In South Africa, in spite of the delivery of 3.0 million houses since 1994 until 2011, the
backlog is still at 2.1 million (Sexwale, 2010; Zuma, 2010)
The above mentioned situation calls for concern. This concern was reflected by Tibaijuka
(2005) who stated that the need for housing production in developing country’s cities is
estimated at around 35 million per year. Breaking this figure down, Tibaijuka added that some
20 million units are required to meet demographic growth and new household formation, while
the remaining 15 million units are to meet the requirements of the homeless and people living
in inadequate housing. Summing this up, some 95,000 new urban housing units are needed to
be constructed each day to ensure acceptable housing conditions. Oruwari (2006) emphasized
that globally the housing conditions of the poor are deteriorating with the developing world
accounting for the worst rate of deterioration. Approximately 998 million people were living
in slums in 2007; the projection for 2010 was 1.12 billion people (UN-Habitat, 2007). These
figures indicate that the housing challenge in the developing countries of the world is
enormous.
5.3 NIGERIA
This section looks into housing in Nigeria. The policies and agencies supporting housing
delivering in Nigeria, such as the government, private sector and others are presented. An
evaluation of past policies and government intervention programmes is also presented. The
philosophical basis for housing provision in Nigeria is discussed. Also explored in this section
151
are the challenges facing provision and the needs, demand and supply of housing in Nigeria.
5.3.1 Background
Nigeria, a former British colony, became an independent nation on October 1, 1960, and a
Republic in 1963. According to the country comparison by the Central Intelligence Agency
World Factbook (2011), Nigeria’s population is currently estimated at 155,215,573 million;
with conservative estimates concluding that more than 20% of the world’s black population
lives in Nigeria. The rate of urban population growth is thought to be 5.5% annually, roughly
twice the national population growth rate of 2.9%. More than seven cities have populations that
exceed one million, and over 5,000 towns and cities of various sizes have populations of
between 20,000 and 500,000. Greater Lagos, the former national capital, has grown from 1.4
million in 1963 to 3.5 million in 1975. It is currently over 7 million strong, and is projected to
be 24 million people by 2020 (Nwaka, 2005). Furthermore, Nigeria has the greatest diversity
of cultures, ways of life, cities and terrain. With the current annual population growth rate of
2.9 percent, which is nearly the same as the annual GDP growth rate (3.5 percent), the Nigerian
population is set to double in the next twenty five years (AFRODAD, 2010). By 2015 (the year
benchmarked for the attainment of the MDGs) Nigeria’s population is estimated to be about
178 million.
Up until, 1989 the country’s capital was Lagos, with a population of about 2,500,000; but the
government has since moved the capital to Abuja. The country’s governmental structure was
increased to twelve states in 1967, then to nineteen states in 1976, with Abuja as the new federal
capital. Nigeria is located in western Africa on the Gulf of Guinea with a land and water total
area of 923,768 km² making it the world’s 32nd-largest country (after Tanzania). Nigeria has
a varied landscape. From the Obudu Hills in the southeast through to the beaches in the south,
the rainforest, the Lagos estuary and savannah in the middle and southwest of the country and
the Sahel to the encroaching Sahara in the extreme north. The highest point in Nigeria is Chapel
Wadi at 2,419 m (7,936 ft). According to the United Nations, Nigeria has been undergoing
explosive population growth and has one of the highest growth and fertility rates in the world.
By the UN’s projections, Nigeria will be one of the countries in the world that will account for
most of the world’s total population increase by 2050. According to current data, one out of
every four African is a Nigerian. Presently, Nigeria is the most populous country in Africa,
the seventh most populous country in the world, of which the majority of the population
152
is black. The economy of Nigeria is one of the fastest growing in the world, with
the International Monetary Fund (IMF) projecting a growth of 8% growth in the Nigerian
economy in 2011 (Odueme, 2011).
Map 1: Map of Nigeria
Source: CIA Factbook, 2010.
Nigeria has become progressively more urbanized in the last five decades; the proportion of
the population living in urban areas rose from 15 per cent in 1950, to 23.4 per cent in 1975 and
to 43.3 per cent in 2000. Projections indicate that more than 60 per cent of Nigerians will live
in urban centres by 2025, and a substantial percentage of these are likely to live in slums if
urgent action is not taken. Over time, one of Nigeria’s greatest challenges is providing adequate
housing for its growing population. The country is estimated to have a deficit of about 11 - 16
million housing units, requiring the construction of about two million units annually to meet
the shortfall.
5.3.2 Housing in Nigeria
Housing in Nigeria is a highly contentious and politicized issue that is of great concern to
administrators, scholars and the Nigerian general public. The influx of people into the urban
areas, the natural population increase and insufficient responses by the government have
contributed to the deterioration of the housing situation in Nigeria over time. The extent of this
situation is such that the economic development and the welfare of the citizens are adversely
153
affected (Federal Republic of Nigeria, 1991; Ibem, 2010). The housing problem in Nigeria is
more critical in the cities, where a huge housing supply deficits exist, which is, coupled with
the dilapidated housing conditions, high cost of housing construction, as well as proliferation
of urban slums and squatter settlements the norm. As such, a large majority of urban residents,
particularly the low-income earners who constitute about 50% of Nigeria’s 155 million people,
are forced to live in conditions that constitute an affront to human dignity (Aribigbola, 2008;
Ibem, 2010). In acknowledgment of the fact that neither the public nor the private sector are
able to address this problem individually, current efforts in addressing the housing situation in
Nigeria are mostly based on collaborative efforts (National Economic Empowerment and
Development Strategy, 2004; Mabogunje, 2003); also, Public-Private Partnerships (PPPs) are
amongst the most collective forms of such collaborative efforts (Ibem, 2010).
Thus, the provision of affordable housing for its citizens has remained the principal focus of
every successive government in Nigeria. This is because of the pivotal roles played by housing
in national development and growth on the one hand and its being a necessity in the life of the
people, on the other. Since Nigeria’s independence, affordable housing has been the major
policy concern of relevant Housing and Mortgage Institutions. However, critics are of the view
that despite the policy efforts, only the needs of the middle and high-income classes are met,
which defiles the generally acknowledged fact that the right to housing is one of the most
important basic human rights recognised in many international Human Rights Treaties (Bret,
2002), because housing is supposed to addresses basic human needs of all citizens.
In Nigeria, a home of one’s own represents a precious refuge. Ademiluyi (2010) informs that
in the traditional African setting and in most developing nations, housing is one of the greatly
cherished material properties. This is because of the other functions that a house performs in
the traditional society, which includes the protection of family cohesion and values, taking care
of the aged through the extended family system, and the protection of the ancestral values,
amongst others. Indeed, the literature on housing in Nigeria often shows that housing ranks
above education and health services as a priority. Thus, households in Nigeria value
homeownership more than households in advanced industrialized countries. In fact, in most
developing nations like Nigeria, when a family does not have a house they can call their own,
that family is regarded as the poorest of the poor. Thus, the importance of providing adequate
housing (housing that meets the needs and expectations of the people, ranging from the
supplied quantity to the quality, thus assuring the quality of life of the people) in Nigeria, cannot
154
be over-emphasized. However, acquiring access to a house and to the components that
comprise housing such as land and property rights, building materials, basic services,
regulations, subsidies, and credit are extraordinarily difficult for most households in Nigeria.
This is because poor land records, tenure right with a dysfunctional legal systems and
regulatory bodies typically cloud ownership rights for a large number of households.
However, efforts of the governments aimed at providing shelter especially for the low-income
earners have not yielded expected positive results compared to the level of effort. This is
because it has been apprehended as the product of a politically motivated official intervention
in the housing crisis, usually at urban level, than a genuine concern to shelter the poor and
needy. For instance, housing schemes designed for the citizens especially the low-income
groups has been forcefully taken over by the high and medium income earners. According to
Shyllon (1999), most chairmen (persons) of the allocation committees of these housing estates
are politicians, who allocated the houses to their party members who neither had the need for
them nor could be classified as low-income earners. In the same vein, Omoniyi (1994) notes
that the housing corporations that are supposed to provide shelter for the public (low-income
and disadvantaged group) only catered for the high-income groups who can afford or meet up
with their stringent terms and pre-qualification conditions. Furthermore, frantic efforts by the
low-income earners for owner-occupier houses are thwarted by their inability to mobilize
adequate funds to execute the project (Jolaoso et al., 2008). As a result, the Nigerian housing
question is primarily that of a crisis situation, manifesting and expressing itself in qualitative
and quantitative forms. As a matter of fact, past Nigeria governments cannot be said to have
ignored housing in the process of national development. The emphasis and implementation
may be weak, but past governments have shown some commitment to housing matters as
reported by various scholars (Adeyemo & Dekolo, 2000; Aigbavboa & Thwala, 2009, 2011;
Federal Government of Nigeria, 2004).
In spite of the efforts of various stakeholders that have been involved in the provision of
housing, there are still numerous challenges affecting these efforts as highlighted: the limited
use of co-operative housing approaches as one of the challenges facing delivery of adequate
housing for the majority of people needing housing (Onukwugha, 2000). Jinadu (2004) states
that the high cost of building materials; inadequate housing statistics for proper planning;
institutional challenges and low housing investment are the factors responsible for inadequate
housing in Nigeria. Furthermore, Ademiluyiand Raji (2008) asserts that difficulties in land
155
acquisition; inability to access long term mortgage finance and high cost of building material
are the factors responsible for inadequate housing delivery. While Jimoh and Olayiwola (2008)
are of the opinion that a disproportionate number of professional builders (less than 2000)
relative to the population of Nigeria (more than 155 million) contributes indirectly to the
inadequate delivery of housing. Aigbavboa and Thwala (2009) established the following
factors as being responsible for the inadequate housing delivery experienced in Nigeria:
legislation, contracts enforcement, inadequate infrastructure, unstable macroeconomic
environment and lacklustre implementation of the National Housing Policy. In addition to the
above challenges, high interest rates charged by banks for mortgage loans and concentration of
Primary Mortgage Institutions (PMIs) in urban centres, such as Lagos and Abuja are also
responsible for the slow pace of affordable housing delivery for the low income group.
The present housing difficulty in Nigeria does not necessarily arise from poverty, but because
of the absence of an effective administrative arm to mobilize and organize the country’s natural,
human, and industrial resources, amongst others for housing and urban development (Jolaoso
et al., 2008). However, the problem of poor co-ordination and ineffectiveness of some public
housing agencies in Nigeria is in most cases responsible for the failure of certain laudable
housing policies and programmes. Some of the agencies with adequate knowledge have refused
to perform their duty of seeing to it that the beneficiaries of the housing schemes fulfill their
obligations and hence, this has constituted a threat to the successful execution of the housing
scheme for the low income groups (Shu’aibu, 2007).
5.3.3 Philosophical Basis for Housing Development in Nigeria
The fundamental philosophy underpinning housing development vision in Nigeria is the
existing Nigerian Constitution of 1999. Chapter II of the Nigerian Constitution popularly
referred to as the “Fundamental Objectives and Directive Principles of State Policy”, states that
the security and welfare of the people shall be the primary purpose of government. The
constitution further expresses that the government is entrusted with the responsibility to harness
the resources of the nation to serve the common good of all and to promote national prosperity
based on an efficient, dynamic and self-reliant economy (Eboh, 2010). Also, it states that
government should manage the economy so as to secure the maximum welfare, freedom and
happiness of every citizen. Furthermore, it specifically requires that the State shall ensure that
suitable and adequate shelter is provided for the citizens (Federal Republic of Nigeria, 1999).
In effect, this national philosophy (which was also enunciated in previous constitutions) has
156
been the generic basis for housing development and planning since Nigeria gained her
independence. Despite this background, the Nigerian Constitution does not employ the
expression ‘adequate housing’, but it provides in section 16(1) (d), under the Chapter dealing
with the Fundamental Objectives and Directive Principles of State Policy, that the State shall
ensure that “suitable and adequate shelter” is provided to all citizens. Section 13 of the
Constitution further states that, “it shall be the duty and responsibility of all organs of
government, and of all authorities and persons, exercising legislative, executive or judicial
powers, to conform to, observe and apply the Fundamental Objectives and Directive
Principles”. However, Section 6(6) (c) diminishes the impact of section 13 by expressly
stipulating that it does not establish enforceable rights. This contrasts with the constitutional
provisions on fundamental rights, which though enforceable do not include access to adequate
housing or shelter. Thus, the Nigeria Constitution does not explicitly contain justifiable socio-
economic rights and does not directly enshrines everyone’s right to have access to suitable and
adequate housing, but indirectly informs of the support for everyone to have accesses to
‘suitable and adequate shelter’. This indirect support has been a major hindrance in the effort
to adequately house the poor and the low-income groups.
5.3.4 The History of Housing Policy in Nigeria
Before the arrival of the colonial rule at the beginning of the 20th Century, a communal system
of housing delivery was practiced in most Nigerian communities. Houses were built through
communal efforts by peer groups (Kabir & Bustani, 2009). Members of all age and different
groups and other social organisations would turn out as a group on an appointed day to assist
the builder in whatever task of the project. In return, the builder would provide meals, whilst
the project lasted and vice versa. This system continued up to 1928, and still remains in some
communities to date despite the disruption of the people's ‘communitarian’ values by
‘westernization’ and urbanization. According to Adisa, Agunbiade and Akanmu (2008),
housing issues in colonial Nigeria were based on the politics of ‘separate areas’. During the
early years of colonial administration, the Nigerian government was involved in the
construction of official residences for expatriates and senior indigenous staff (in the public
services, such as the police and the railways personnel) in specially zoned Government
Reserved Areas (GRAs) through the Public Works Department (PWD) (Omole, 2001). Also,
the then three Regional Governments, through the colonial housing policy in place and catered
exclusively for the housing needs of the top echelons in the civil services. When Nigeria
157
became an independent state in 1960, the colonial politics of discriminatory housing changed
somewhat. After the attainment of independence and the immediate period after achievement
of sovereignty, public housing was limited, elitist and largely in the form of middle class
housing estates for government officials (Federal Republic of Nigeria, 1981). The exit of the
colonial masters afforded Nigerians the opportunity to move into residential areas hitherto
reserved for the European workers. Despite the removal of the discriminatory housing format
operated by the colonial masters with the advent of the Nigerian independence, housing in
Nigeria has remained a contentious issue. Regardless of the laws and all attempts made to date,
Nigeria still faces a severe housing shortage with more of the population living in sub-standard
housing. However, evidence from literature clearly shows that public housing provision in
Nigeria, has not recorded any impressive results in matching housing production to housing
demand, as there are huge housing supply deficits.
Furthermore, the outbreak of the Bubonic Plague in Lagos, the then capital city of Nigeria in
the 1920s, led to the first slum clearance and settlement upgrading policy programme by the
Lagos Executive Development Board (LEDB) (Aribigbola, 2008). Agbola and Jinadu (1997)
claim that between 1973 and 1995 about 36 other cases of slum clearance were reported in
urban areas in Nigeria, including the widely publicized demolition of Maroko, Lagos, in 1990.
This was regarded as a violation of the housing rights of the occupants. Although the aim of
the slum clearance was to upgrade blighted areas in the cities (Nwaka, 2005), scholars like
Agbola and Jinadu (1997) and Umeh (2004) contended that the approach failed to provide
decent and affordable housing to Nigerians. This is due to the non-availability of land in
locations that were acceptable to displaced persons, as well as the lack of adequate funds to
resettle them; however, the vacant land was sold at high prices to the rich in the society.
In 1958, the defunct regional governments, that is, Western, Eastern, Northern and Mid-
Western Regions, established Housing Corporations to construct and manage housing estates,
as well as to grant ‘soft’ loans to individuals wishing to build their own houses. Unfortunately,
the established agencies were unable to extend their services to the low-income group due to
the lack of commitment to low-cost housing (Federal Republic of Nigeria, 1991; Ibem, 2010).
The first National Development Plan (1962-1968), brought about the establishment of the
Federal Housing Authority (FHA), Federal Ministry of Environment Housing and Urban
Development (FMEHUD), Ministry of Works and Housing, State Housing Corporations and
Federal Mortgage Bank of Nigeria (FMGN), which spilt into the 1970s. At this stage, Nigeria
158
was set to witness a massive government involvement in housing construction, which was
supposed to have extend to the low-income group. According to Ibem (2010), the first
government-assisted self-help housing programme took off in the then newly created states of
Bauchi, Benue, Gongola, Imo, Niger, Ogun and Ondo, as well as in Lagos State, in the mid-
1970s. With assistance from the World Bank, the scheme succeeded in providing serviced
plots, soft loans and technical assistance for few low-income groups toward owning houses in
the eight states’ capitals, but it could not be extended to other state because of logistic and
funding constraints on the part of the government. Conversely, the plan made provision for the
erection of at least 59,000 dwelling units, 15,000 for Lagos and 4000 for each of the then 11
states. However, the success rate was only 12% (Okpala, 1986). Its spread was wide and even
but was skewed in favour of low-income workers, which constituted 60% of the workforce.
The middle-income group constituted 30%, while 10% was allocated to the upper-income
group. The Second National Development Plan of 1970-1974, with the launch of the National
Low-Cost Housing Scheme in 1975 committed a total of 53.35 million Naira ($380 000) to the
building of new houses and upgrading of the old ones.
In the Third National Development Plan (1975–1980), both the Federal and the State
Governments attempted, for the first time, direct construction of housing units to be let out at
subsidized rates. Similarly, the Federal Mortgage Bank (FMB), a latter day convert of the
Nigeria Building Society (Adisa et al., 2008) was established primarily to grant loans to
mortgage institutions in order to facilitate housing delivery. About 2.6 billion Naira was
earmarked for the construction of 202,000 housing units across the country. With the proposal
of 202,000 units during this planned period, Lagos State, the then national capital, was
allocated 46,000 units and the remaining 156,000 were to be built in the different parts of the
country. Ogu and Ogbuozobe (2001) argue that the dismal performance of the public’s direct
involvement in housing is shown by the fact that only 19% of the intended dwelling units
(8500) were built in Lagos, whilst only 13% (20,000 units) of the proposed units were
constructed in the rest of the country by the end of the planned period. In all, about 24% of the
202,000 housing units were constructed at the expiration of that development plan, while many
of the uncompleted housing units were abandoned. As a result, the dream of homeownership
by many Nigerians could not be realized through that scheme (Onibokun, 1985). The targets
were not met and it necessitated the incorporation of another housing plan in the Fourth
National Development Plan.
159
The Fourth Development Plan period (1981-85) witnessed the launching and implementation
of the third national housing programme by a civilian administration. During this period, only
20% of the planned 1600,000 dwelling units were constructed. According to Aina (1990) and
Ogu and Ogbuozobe (2001) the housing programme was frustrated by fraud and politicization
of these projects, lack of supervisory technical staff at building sites and lack of service
infrastructure. Likewise, the plan earmarked 600 million dollars for housing development, but
from 1984, the Military Administration abandoned direct intervention in the housing market
and embarked on demonstration projects. Table 5.1 below shows a summary of the planned
and constructed number of housing units in the different public housing programmes, initiated
between 1962 and 1999. An analysis of Table 5.1 reveals that an aggregate of 618,498 housing
units were intended for construction in the various public housing schemes across the country.
Nevertheless, about 85,812 housing units represented around 14% of the intended housing units
that were actually constructed. This achievement level clearly shows that many of the public
housing programmes initiated by government within that period failed to meet the targeted
number of housing units (Ibem et al., 2011). The cumulative effect of this failure is that an
estimated 75% of Nigeria’s 60 million urban population live in slums, and not less than 700,000
housing units are required annually to improve on this appalling housing situation across the
country (Federal Republic of Nigeria, 1991; Olotuah, 2010). This denotes that suitable
measures need to be urgently put in place to combat the challenge of low productivity in public-
sector housing in the country.
Furthermore, the failure of the National Low-Cost Housing Scheme led to the initiation of the
National Site-and-Services Programme. The National Site-and-Services programme was
proposed to make serviced plots available to housing developers without many inconveniences
(Onibokun, 1985). Under the guidance of the Ministry of Works and Housing, Federal Housing
Authority (FHA) and other related housing provision agencies, the programme recorded initial
success in seven states, namely, Lagos, Kano, Imo, Kwara, Ondo, Rivers and Imo, as well as
in FCT. The majority of beneficiaries for this scheme were the middle- and high-income groups
who could afford the high cost of transfer fees and who met the requirements for allocation of
the serviced plots. Regardless of the obvious poor performance of the public housing schemes,
as amongst others, the number of produced housing units fell far below the intended housing
units. Hence, this led to the launch of a new public housing scheme in 1994 by the Military
Administration. Thirteen months after the inauguration of the housing scheme, a review
committee set up by the Federal Ministry of Works and Housing (FMWH) admitted that the
160
new housing scheme have failed to meet its objective (Agbo, 1996). The number of units
produced was too expensive for the low-income groups to afford, even when the cost of
infrastructure and land were waived.
Table 5.1: Performance of Public Housing in Nigeria (1960- 2010)
PERIOD PROGRAMME TARGET ACHIEVEMENT LEVEL First National
Development
Plan (1962-1968)
Planned construction of 61,000
housing units.
Only 500 units less than 1% of the
planned units were constructed. The
political chaos and the resulting
civil war (1966-1970) contributed
to the marginal progress recorded
during this period.
Second National
Development
Plan (1971-74)
Establishment of National
Council of Housing (1972) to
advise the government on
housing matters and Federal
Housing Authority (FHA) in
1973 to co-ordinate public
housing provisions - Plan direct
construction of 59,000 ‘low-cost’
housing units across the
Federation.
7,080 housing units representing
12% of planned houses were
actually built.
Third National
Development
Plan (1975-
1980)
- Creation of Federal Ministry of
Housing, Urban Development
and Environment and
conversion of Nigerian
Building Society to Federal
Mortgage Bank of Nigeria
- (FMBN).
- Promulgation of the Land Use
Decree (1978).
- Planned construction of
202,000 low-cost housing units
nationwide.
30,000 housing units representing
less than 15% of planned houses
were completed
4th National
Development
Plan (1981-1985)
- National Housing Programme
launched for the first time in
1980. Earmarked N1.9billion
for the construction of 160,000
housing units, for low-income
groups
- The second phase of the
housing
programme set out to construct
20,000 housing units across the
Country.
A total of 47,234 housing units
representing about 23.6% of
planned housing units were
constructed in the first phase. The
second phase was cut short by
the military coup of 1983
Military
Government
- National Housing programme
planned 121,000 houses with
- 5,500 housing units (less than 5%)
of planned houses were actually
161
(1986-1999) National Site-and-Services
Housing Programme between
1993 and 1995-1988
- National Housing Policy
launched to provide Nigerians
access to quality housing and
basic infrastructure.
- 1991, National Housing Policy
was launched with the goal of
granting all Nigerians access to
decent housing by 2000 in
response to the slogan “
Housing for All by the year
2000” of the United Nations.
constructed.
- Provision of rural infrastructure
through the Directorate of Food,
Roads and Rural Infrastructure
(DFFRI)
Civilian
Governments
(1999-2010)
- The New National Housing and
Urban Development Policy
(NHUDP) launched in 2002
with the goal of ensuring that
“all Nigerians own or have
access to decent housing
through private sector-led
initiatives”.
- Planned construction of about
10,271 housing units through
the Public-Private Partnership
(PPP) arrangements in different
PPP housing schemes across the
country.
- Planned construction of 500
housing units in the Presidential
Mandate Housing Scheme in all
36 State capitals and Abuja.
- Government planned a pilot
projects involving the
construction of 40,000 housing
units per annum nationwide.
- 2000 serviced plot through PPP
site and service in Ikorodu, Lagos.
- 4,440 housing units completed in
Abuja, Port Harcourt, Akure and
Abeokuta, through PPP.
- The Presidential Mandate
Housing Scheme did not take off
in many States. In Ogun State
about 100 housing units
representing 20% of the planned
units were constructed.
- Records of the achievement level
of the pilot projects are not yet
available.
Source: Ibem, 2010
Yet, in the last few decades, the Nigerian housing agencies have delivered an insufficient
number of low-quality and expensive housing units for few middle- and high-income earners
(Awotona, 1990; Ogu, 1999; Ogu & Ogbuozobe, 2001) thus, creating a huge backlog for the
low-income group in the country (Ibem, 2011; Onibokun, 1990). The problem of poor funding,
bureaucracy, the politicization of housing programmes and the absence of proper organisation
and transparency in the management of housing programmes all account for the minimal
successes recorded by the housing schemes (Ibem, 2011; Onibokun, 1985). However, the UN-
162
HABITAT II Conference held at Istanbul, Turkey, in June 1996 heralded a renewed advocacy
of an initiative in housing delivery in Nigeria. At this time, people started subscribing to the
National Housing Fund (NHF), which was established by Decree 3 of 1992. Conversely, the
subscribers were disappointed by the upward review in the agreed amounts. Since then, many
Nigerians have been suffering from problems of housing, they are forced to pay higher rent in
spite of the existence of the State Rent Control and Recovery of Residential Premises Edict of
1997.
5.3.5 Housing Policy in Nigeria
In an attempt to meet the housing challenges of the Nigerian populace, a number of
programmes and policies have been articulated and introduced. The first explicitly formulated
National Policy on Housing was launched in 1991 by the Federal Government of Nigeria with
a set goal of providing housing accommodation for all Nigerians by the year 2000. This was in
response to the Agenda 21 of Global Shelter Strategy, aimed at achieving sustainable human
settlement development. This was in response to the exploding population and urban growth
rate and also an acceptance, for the first time, that government alone was incapable of
addressing the alarming gap in the housing needs of all Nigerians (Aigbavboa & Thwala,
2011). The failure of the set goals for the first housing policy necessitated the adoption of a
new policy in 2002 aimed at providing necessary solutions to the previously intractable housing
crisis in Nigeria (Aribigbola, 2008).
The Federal Government of Nigeria revised the National Urban Development and National
Housing policies in 2002, in line with the new democratic dispensation, which required the
promotion of sustainable urban development and social order in the country, centered on
citizen’s participation in decision making and programme implementation, monitoring and
evaluation (Aribigbola, 2008). The resultant 2002 National Housing Policy was directed
toward all Nigerians owning or having access to decent, safe and sanitary housing
accommodation at affordable cost with secure tenure through private sector initiative with
government encouragement and involvement. The 2002 policy introduced some new measures
and innovations that were considered suitable to making housing accessible to all Nigerians in
line with global thinking and action.
The adopted 2002 housing policy, which is a review of the 1991 policy has again been reviewed
163
in 2004 with the ultimate goal to ensure that all Nigerians own or have access to decent, safe
and healthy housing accommodation at an affordable rate (National Housing Policy, 2006).
The most significant differences between the new policy and the previous ones is that housing
was now seen in the context of overall national development in contrast to when housing was
regarded as a social service and a natural fall-out of the national economic development.
Secondly, the policy identified the fact that different people both within and between income
groups tend to have different demands for housing, moving away from the one-size-fits all
syndrome. This is evident from the ultimate goal of the Housing Policy which is, to ensure that
all Nigerians own or have access to decent housing accommodation at affordable cost. Thirdly,
the current focus on removing all barriers to the supply of housing and provision of incentives
to all parties involved in the housing delivery system. The previous policies from the past until
now have always been formulated with good intentions, but the formulators of the policies do
not spell out the direction, neither do they take into consideration the amount of involvement
required from the would-be beneficiaries. A typical short coming of the previous policies
usually carried the slogan of ‘housing for all Nigerians’. This statement ordinarily assumed that
all families in Nigeria would be provided with adequate housing regardless of who needed one
and with no contribution or participation being expected from the beneficiaries (Ajanlekoko,
2001).
In Nigeria, the major policy steps taken, so far, towards solving the housing crisis in the country
can be summarized, as shown below in Table 5.2. Others are the formulation of the National
Housing Policy (NHP) in 1984, the establishment of the Infrastructural Development Fund
(IDF) in 1985, and the Urban Development Bank (UDB) in 1992 (Federal Republic of Nigeria,
1997). In addition to the above, all the introduced National Development Plans (NDPs) from
1962-1985 and the National Rolling Plans (NRPs) from 1990 to date plainly recognize the
importance of providing affordable housing in the country as a tool for stimulating the national
economy (Ademiluyi, 2010; Gbolagade, 2005), but little has been done to bring about an
actualization of the policy intentions.
Table 5.2: Major Housing Policy Steps in Nigeria (1928- 2010)
Policies to date Reason for policy formation Formation of the Lagos Executive
Development Board (LEDB) - 1928
The Board was authorized to carry out slum
clearance, land reclamation, and the
development of residential and industrial
164
estates.
The setting up of Nigerian Building Society
(NBS) - 1956
Provision of housing loans to both civil
servants and the Nigerian public.
Establishment of housing corporations –
1964
The regional housing corporations were
mandated to develop estates and at the same
time provide mortgage for the people to
build houses and pay back over many years.
Establishment of the Federal Housing
Authority, established under Decree No. 40
of 1973 and amended by CAP 136 LFN of
1990 – 1973
Its role was to make proposals to the
government for housing and ancillary
infrastructural services and implementing
those approved by government.
The creation of the National Site and
Services Scheme - 1986
Formed to provide land for government
housing with essential infrastructural
facilities for housing developments in well-
planned environments. Also, to provide well
laid-out serviced plots in each of the 36 state
capitals of the federation, including FCT
Abuja.
The formation of the National Prototype
Housing Program by the Federal Ministry of
Works and Housing
To complement the objectives of the
National Site and Services Scheme. The
project was embarked upon to demonstrate
the feasibility of constructing functional,
effective, and affordable housing units
through imaginative designs, judicious
specification of materials, and efficient
management of construction.
The setting up of the State Housing
Corporation
To provide housing to the populace at
affordable prices.
The creation of the Federal Mortgage Bank
of Nigeria - 1977
To finance housing loans to prospective
housing developers at minimal interest rates.
The setting up of the National Housing
Program (NHP) in 1991 and the National
Housing Fund (NHF) scheme by Decree No
3 of 1992
To provide self-loans to potential housing
developers and also monitor developments
in the housing sector.
The deconsolidation of the Federal Mortgage
Bank of Nigeria (FMBN) through the
establishment of the Federal Mortgage
Finance Limited (FMFL)
To take over retail mortgage portfolios
previously handled by the bank and also to
facilitate effective management of the
National Housing Fund (NHF) Scheme.
The setting up of a Housing Policy Council
(HPC)
To monitor development in the housing
sector and also to set up the machinery for
the review of the 1978 Land Use Decree
(LUD) in order to make more land available
for large scale land developers.
The creation of the ministry of Housing and
Urban Development in June 2003
Charged with the responsibility of ensuring
adequate and sustainable housing delivery
and maintenance of a conducive living
environment that meets the needs and
aspirations of Nigerians.
The review of the mandate given to the The authority also plans to facilitate the
165
Federal Housing Authority to include
provisions of the National Social Housing as
part of the strategy towards meeting the
Millennium Development Goal.
provision of two million housing units
before 2015, which is not a realistic goal
with the way housing issues are still being
treated in the country.
Source: Compiled by Researcher from various sources. Adisa et al., (2008); Ajanlekoko
(2002); Aribigbola, (2008); Ogu and Ogbuozobe, (2001); Olotuah (2010); Omole (2001)
5.3.6 Challenges Facing the Provision of Housing in Nigeria
Housing provision in Nigeria faces major challenges coupled with the limitations imposed by
the income levels of her citizenry. In the face of a daunting deficit put at about 16 million
houses and the level of achievement over the past 36 years of the government programmes and
policies, the Nigerian government is faced with difficult task of overcoming this backlog and
delivering homes to the low, medium and high income groups, respectively. There are several
challenges facing the provision of housing in Nigeria, with particular emphasis on the low-
income groups. This is unlike the case of South Africa, where government provide houses free
of charge to the low-income groups or disadvantaged poor, through the South Africa National
Housing Subsidy schemes; to ensure the beginning of economic independence and freedom.
Shu’aibu (2007) states that the problems associated with the attainment of affordable housing
in Nigeria, includes compromises during implementation, lack of political sensitivity, and
corruption, amongst others. Any compromises made during implementation that sought to alter
basic policy goals are normally detrimental to the successful execution of any housing policy.
Though policy implementation is a tedious process that requires a great deal of analysis before
starting it, the reality of basic housing can only be realized if there is proper implementation of
policies. Furthermore, lack of adequate data relating to the magnitude of the problem, which is
as a result of the absence of a national data bank for housing; inconsistency in policies and
programmes, including regular changes of policies with changes of government without proper
assessment of the existing ones; lack of efficient and sustainable credit delivery to the housing
sector; and income levels of those needing houses. This is relatively low in comparison with
house market prices, resulting in an affordability problem; high cost of building materials; the
rapid annual growth rate of the Nigerian population, which was estimated at 3.3% on the basis
of annual birth rate of 49.3 per 1,000. All these coupled with the rapid population
growth/urbanization with the problem of an increasing poverty level among the citizenry,
which has risen from 65% in 1996 to about 70% in 2009 (United Nations Development
Programme and World Bank estimates confirms UN-habitat (2009) statement that 70% of the
166
urban dwellers in Nigeria live in slums); and lack of effective coordination amongst Housing
Agencies. While all the three tiers of the government are involved in one way or the other in
housing matters, their activities are hardly coordinated.
5.3.7 Programmes Supporting Housing Creation in Nigeria
Nigeria’s efforts towards sustainable housing development since 1986 are marked by policy
formulation and the establishment of agencies for implementation. One of such ‘efforts’ is the
establishment of the Family Support Programme (FSP) initiated by the then First Lady of the
Federal Republic of Nigeria, Her Excellency, Late Mrs. Maryam Babaginda. The FSP
recognizes that one of the most important needs for the survival of any family and healthy
living is the provision of decent and affordable housing, as most low income families in cities
do not own houses because they cannot afford them. The FSP also recognizes that women are
handicapped in their access to land and property in the country. This seriously affects their role
since they need a secure place to live to carry out their subsistence farming and generate
income. Widows and single women are worse off as they are denied rights to inherit landed
property. In this regard, the FSP sets out, amongst others, the above objectives to ensure
adequate housing for the less privileged in the society.
Furthermore, the Federal Government through the Federal Housing Authority (FHA) started
the National Housing Programme in 1994. The objective was to produce 121,000 housing units
for low, medium, and high income earners. Currently, records show that about 5% of the target
has been achieved, as at 2010. Further efforts on direct construction of houses continue to be
made through the National Prototype Housing Programme aimed at demonstrating the
feasibility of constructing functional, cost effective and affordable housing units. So far 600
housing units in various stages of completion are being constructed in Lagos, Kaduna, Port-
Harcourt, and other areas as at the end of 2010. Also, the National Housing Fund was
established in 1992 to solve the problem of finance for housing development. All workers (both
public and private sector) earning N3, 000 and above per annum contribute 2.5% of their
income to the fund. The funds are disbursed as mortgage loans through primary mortgage
institutions to the subscribers of the fund (UN-Habitat, 2008). Currently, the Nigerian Urban
Renewal Programme is directed at improving existing slums in the core of cities. The
programme has been implemented in 18 cities across the nation. A total of about N20 million
had been spent on the programme since 1992. The Federal Government has also provided
167
through the National Sites and Services Programme over 15,000 plots at subsidized rates to the
public; and over N250 million have been committed to the programme in the last six years as
reported by the UN-Habitat (2007). Moreover, the on-going Federal Housing Schemes will,
according to plans, make possible the provision of adequate housing for all government
workers in Nigeria so that at retirement they and their families will have a place to live.
Further to the achievement of the goal of improving urban management, the country is now
participating in the Sustainable Cities Programme (SCP) under the Urban Management
Programme (UMP) of the United Nations Centre for Human Settlement (UNCHS) / World
Bank/United Nations Development Programme (UNDP). Through this programme, the
Sustainable Ibadan Project (SIP) is being implemented. Also, through this initiative, Local
Governments, NGOs, Community Based Organizations, and private individuals are
encouraged to participate and contribute to urban improvement and management. Presently,
the process of replicating the sustainable city programme in other cities has already begun.
Two other cities, Kano and Enugu have commenced their projects. The sustainable Kano
Project has already prepared the Kano environment profile study forming the basis for
consultative actions on the management of Metropolitan Kano. The SCP emphasizes a two-
way relationship between development and environment, which promotes better awareness and
understanding of the priority issues to be addressed in urban environment and development and
a consideration of modern urban and environmental management methods, and the most
effective and lasting impact (UN-Habitat, 2008).
In addition, the determination of the government toward the success of the objectives of
programmes under the Infrastructure Development Fund Programme, the Urban Basic Services
Programme is being undertaken in the country to promote the integrated provision of
environmental infrastructure, water, sanitation, drainage, and solid waste management. The
project involves the identification of core areas in some Nigerian cities and the packaging of
improvement programmes targeted at women and children. Also, the Nigerian Government is
currently working on developing future programmes aimed at improving the human settlement
development and management sectors. These sectors include: poverty alleviation programmes
in collaboration with the World Bank and UNDP; a programme support document for
Governance in collaboration with UNDP; a National strategy for the replication of the
Sustainable City Programme in other Nigerian cities; and replication of the Urban Basic
Services Programme in collaboration with the United Nations Children’s Fund. The
168
government intends to concentrate efforts in the near future on the following areas: capacity
building for improved management; institutional and policy reforms; social reorientation;
increased participation of NGOs and the private sector; and promotion of appropriate
technologies (UN- Habitat, 2008). These future plans are aimed at achieving a state of
environmentally sound human settlements, free of slum conditions, in which every
disadvantaged and poor Nigerian will have access to adequate and affordable shelter, and
efficient infrastructure and services, which will foster sustainable economic growth, and an
improved standard of living and well-being.
5.3.8 Housing in Nigeria – Need, Demand and Supply
Housing needs are considerable in Nigeria, the deficit is currently estimated at over 14 to 16
million units (UN-Habitat, 2008); if put in monetary terms, it will amount to four times the
annual national budget of Nigeria. The rate of urbanization in Nigeria has witnessed
tremendous increase in the last two decades. Census in the early fifties showed that there were
about 56 cities in the country and about 10.6% of the total population lived in these cities. This
rose dramatically to 19.1% in 1963 and 24.5% in 1985 and according to the country comparison
report by the Central Intelligence Agency (2011), Nigeria’s population is currently estimated
at 155, 215, 573 million with the urban population constituting approximately 60%. The rapid
growth rate of urban population in Nigeria since the early seventies is mainly due to migration
promoted by the concentration of the gain from the oil sector in the urban areas (Ajanlekoko,
2001). While the United Nations estimates that Nigeria’s population would reach 289 million
by 2050, the United States Census Bureau projects that the population of Nigeria will reach
264 million by 2050. The rapid growth in population will create demand for shelter and
efficient supply and distribution of basic utilities and services for the city dwellers. The effect
of the explosion from the population growth will manifest in overcrowding in houses. Nigeria
being one of the fastest urbanizing countries in the African continent faces a huge challenge of
adequately providing affordable housing to its citizenry. As more and more Nigerians make
towns and cities their homes, the resulting social, economic, environmental and housing need
should be urgently addressed (Raji, 2008). A study of the housing situation in Nigeria put
existing housing stock at 23 per 1000 inhabitants whilst housing deficit is put at 16 million
houses, and about 12 trillion Naira will be required to finance the housing deficit.
169
As at the beginning of 1999, housing development had been so neglected by successive
governments, who for years did not regard housing as a priority and who on many occasions
made no annual budgetary provision for housing; this brought about a ‘no-housing’ situation
in Nigeria. According to the Nigeria National Housing Policy (2006) about 60% of Nigerians
are said to be homeless. If this is represented by the current population of the country as
represented by the CIA report, which put the Nigeria population at 155 215 573 million people,
it means that about 93 million Nigerians are ‘homeless’. In the context of six members per
household as recommended by the United Nations, it thus confirms that Nigeria has a housing
backlog of 15.6 million units. However, as of March 2009, the Federal House of
Representatives Committee on Environment and Habitat informs that Nigeria is in dire need of
a minimum of 23 million houses if it must meet the minimal housing demands of its citizens
based on the United Nations standard. The House of Representatives Committee also estimated
that Nigeria currently has about 6.3 million houses and to maintain the United Nations standard
of six persons per house for its population, Nigeria actually need about 23.33 million houses in
addition to the 6.33 million that already exist today. There is no doubt that the housing need of
Nigerians is as of yet an unmet judging by the high demand and the limited supply of residential
accommodations in Nigeria.
While virtually all governments in Nigeria since independence have highlighted housing as a
major priority, Nigeria is yet to develop a vibrant housing market, and housing continues to be
provided through traditional methods of buying land and building over some years, which
could mean an individual’s entire life. The problem of housing in Nigeria has been a concern
for both the government and individuals. In acknowledging these challenges, both the public
and private sector developers make an effort through various activities to bridge the gap
between housing supply and demand, but the cost of building materials, deficiency in housing
finance arrangement, stringent loan conditions from mortgage banks, government policies
amongst other problems have significantly affected the rate of housing delivery in Nigeria
(Raji, 2008). Currently, the main goal and target of the Nigeria Housing Authority is to deliver
100,000 housing units in the next four year period from June, 2009 to June, 2013 being the
confirmed tenure of the present management. Despite the relevance of the initiative, housing
development should not be centered on office holders, but a matter of national interest and goal
to adequate house those needing housing (Federal Housing Authority, 2011). The proposed
housing units according to the Nigerian Housing Authority will be achieved through the
following Sustainable Mass Housing Delivery Framework and as typified in Figure 5.1:
170
1. Direct Development - This will be done by utilizing internally generated resources and
institutional financing of projects to be executed by the Authority. A total of 12,000
housing units are to be built through this delivery model;
2. Public Private Partnership – The key objective of the present model is to foster
partnerships, which are profitable and cost effective, leveraging on the strength of the
private sector partners in project financing, financial prudence and effective
management. About 14,000 housing units are to be built through Public Private
Partnership;
Figure 5.1: Sustainable Mass Housing Delivery Framework
Source: Federal Housing Authority, Nigeria (2012)
3. Public-Public Partnership – This is a newly introduced housing procurement system in
Nigeria, which is a partnership between the FHA with other tiers of government also
involved; military and para-military organizations to facilitate housing delivery that
meet the needs of their staff. A total of 4,000 housing units are targeted for delivery
through Public-Public Partnership;
4. Cooperative Housing - The FHA is set to partner with housing cooperatives to facilitate
housing delivery for the needs of their members. About 8,000 housing units are targeted
for delivery through this model;
5. Site-and-Services will be made available within estates to be developed. Additionally,
Site-and-Services will be provided on parcels of land to be acquired to create serviced
plots that will be sold to citizens who desire to build their own houses. About 7,000
serviced plots will be provided under the Site-and-Services scheme;
6. New Towns Development – The FHA also wants to promote the development of new
171
towns across the nation in key cities starting with Abuja, FCT. The new towns will
reduce pressure on the existing cities by providing additional accommodation at
affordable rates. For example, the Bwari Area Council has been identified for
development of a new town in FCT, Abuja, Nigeria. A total of 25,000 housing units are
targeted under the FCT, Abuja New Town;
7. Rental Housing - Additional to the development of housing units for outright sale to the
public, the FHA has proposed to provide rental housing for low income dwellers in the
cities. About 3,000 housing units will be provided for rental in the period under
consideration;
8. Estate Regeneration - Regeneration of old estates will be pursued in partnership with
private investors. The estates being proposed for regeneration include: Maitama,
Asokoro and Lugbe Estates in Abuja, and Festac Town Estate in Lagos.
The Nigerian government needs to make available supportive mechanisms to provide her
citizens with decent and affordable housing. Because housing constitutes an essential need to
complement other social assets of human beings if they are to lead a productive life. It is of
higher importance than food, education, and medical care because of its intrinsic connection to
security, which is paramount to human survival. Without adequate security, preservation of life
is difficult and without shelter, man is exposed to the vagaries of weather and predatory animals
(Akintokunbo, 2008). Despite these initiatives, the implementation of a subsidised low-cost
housing scheme in Nigeria is still in its developmental stage whereas in South Africa it has
been through the developmental threshold and hence properly implemented. In most
developing nations, 90% of the houses are built by private individuals in an informal market,
which is highly unreachable for the low-income to access.
5.3.9 Lessons Learnt from Nigerian Housing Studies
The following are the lessons learnt from the review of literature on the housing situation in
Nigeria:
The provision of affordable housing for its inhabitants has remained the principal focus
of every successive government in Nigeria. This is because of the pivotal roles played
by housing in national development and growth on the one hand and it being a necessity
in the life of the people, on the other;
Housing in Nigeria is a highly contentious and politicized issue that is of great concern
172
to stakeholders, such as administrators, scholars and the Nigerian general public;
Likewise, the difficulties in land acquisition; inability to access long term mortgage
finance and high cost of building materials are some of the factors responsible for
inadequate housing delivery in Nigeria;
Housing problems in Nigeria are more critical in the cities, where a huge housing supply
deficits, coupled with the dilapidated housing conditions, high cost of housing
construction, as well as proliferation of urban slums and squatter settlements is the
norm;
Households in Nigeria value homeownership more than households in advanced
industrialized countries, as the literature reveals that housing ranks above education and
health services as a priority;
The Federal Government of Nigeria should not engage in direct housing construction,
because such effort yields no results. Studies have shown that individuals build better
and cheaper houses and at a faster rate than the government agencies;
The government should place more emphasis on the use of local materials for building
construction so as to reduce building costs;
The government must promote alternative strategies for housing construction, such as
provision of services and sites with basic infrastructure before making them available
for sale to individuals who need them;
The present housing difficulty in Nigeria arises not necessarily out of poverty, but
because of the absence of an effective administrative arm to mobilize and organize the
country’s natural, human, and industrial resources, amongst others for housing and
urban development;
Also revealed from this section, is that the fundamental philosophy underpinning
housing development vision in Nigeria is the existing Nigerian Constitution of 1999;
The Constitution of Nigeria does not contains justifiable socio-economic rights that
enshrines everyone’s right to have access to adequate housing;
Current governmental intervention in housing is at a level of providing an enabling
environment for the various stakeholders to help ameliorate the housing backlog.
5.4 GHANA
This section of the thesis reviews housing in Ghana. This section provides a historical overview
of housing provision in Ghana. It attempts to systematically and brings to focus the challenges
173
of housing delivery by reviewing past and present housing schemes, as was the case in Nigeria.
The policies and agencies supporting housing delivery in Ghana are also explored, such as the
government, private sector and others. Lastly, a summary of the lessons learnt to date from the
literature is also presented.
5.4.1 Background
The modern day Ghana was created from the British Gold Coast Colony, established in 1874,
and the UK-administered Trusteeship Territory of Togoland, incorporated in 1956 following a
referendum. Anxiety for independence grew strongly after the Second World War. From the
early 1950s, self-government was introduced with elections in 1951, 1954 and 1956. In 1957,
Ghana became the first sub-Saharan country in colonial Africa to gain its independence. Ghana
before Independence on March 6, 1957 was called the Gold Coast. The earliest Europeans to
arrive here were the Portuguese in the 15th Century. Upon their arrival, they found so much
gold between the River Ankobra and the Volta, they subsequently named it ‘da Mina’, meaning
The Mine (Republic of Ghana, 2011).
The Republic of Ghana is a country located in West Africa Coast. Ghana is a country located
on the Gulf of Guinea, only a few degrees north of the Equator, therefore giving it a warm
climate. The country covers an area of 238,500 square kilometers (92,085 square meters). It is
surrounded by Togo to the east, Côte d'Ivoire to the west, Burkina Faso to the north and the Gulf
of Guinea (Atlantic Ocean) to the south. Map 2 shows the Republic of Ghana with the
neighbouring countries and internal sub-division.
Map 2: Map of Ghana
Source: CIA Factbook, 2012
174
The Ghanaian population is estimated to be about 24,233,431 million, up from about 17.4
million in 1995 and 6.7 million in 1960 (Ghana Statistical Service, 2011). However, according
to the CIA, the World Factbook report of July, 2011, the population is estimated at 24,791,073,
which represent a growth rate of 1.02 percent. According to the Ghanaian Statistical Services,
the present population count is an average of 48.7 percent male and 51.3 percent female. The
Ghanaian population has increased by 28 percent compared to the last census of 2000. The
population annual growth rate is estimated at 2.4 percent. The United Nations projected that by
the year 2025; about 60 percent of the Ghanaian population will reside in urban areas (United
Nations, 1995). The urban population is currently estimated to be 51 percent of the total
population, with an urbanization growth rate of 3.4 percent estimated for the period 2010-2015.
The main city is Accra with a population of 2.269 million followed by Kumasi with 1.773
million. The rapid rate of urbanization in Ghana is sustained not only by the high fertility rate,
but also by the continuing rural-urban migration. In the period ranging from 1948-1960, urban
population increased in Ghana by almost one million. Today, migration is not quite as rapid,
but it still continues. In the 1980s, the average annual growth of the population for the country
was about 3.2 percent, but for the urban areas, it was about 4.2 percent (Berry, 1995); of which
much of the difference is attributed to rural-urban migration.
5.4.2 Housing in Ghana
The need to provide adequate, suitable and equitable housing has remained a major priority of
every Ghanaian government. Even though housing is a basic necessity of life, more than half
of the Ghanaian population lives in adequate houses (Government of Ghana, 2005) where they
have no access to adequate sanitary facilities, water or warmth to meet their daily physical
needs. Adequate housing is seen as one of the effective means to alleviate rural and urban
poverty, which has further external and internal effects. The lack of adequate housing has
lowered the life expectancy of the homeless according to the European Federation of National
Organizations Working with the Homeless (FEANTA, 2007), as expose they are to serious
health risks, which gravely affect their contribution to society.
In Ghana, the housing situation is said to be inadequate, but, improving. Many households,
particularly those in the cities and other urban areas, continue to live in overcrowded and
unsanitary conditions (UN-HABITAT, 2008). Their houses lack the basic amenities, such as
toilets, kitchen, bathroom, and refuse facilities. UN-Habitat (2008) has observed that the
175
shortage of houses, especially in the urban areas, has given rise to very high occupancy rates,
exorbitant rent, unstable tenancies and poor living conditions. These factors, combined with
issues of land litigation, high cost of urban residential land, multiple sale of urban land, high
cost of building materials, and shortage of infrastructure and services, underline the difficulties
of housing delivery in Ghana.
Since Ghana’s independence, provision of housing has remained central to the development
agenda of the various governments. Various policies, programmes and institutions have sought
to address issues, such as land ownership, land title regulation, and the provision of affordable
housing units to the working and non-working population. However, a number of these housing
approaches were negatively affected by the lack of funds, a poor macro-economic environment
and lack of private sector participation (Bank of Ghana, 2007). Thus, compared with other
advanced countries, Ghana’s housing industry remains rudimentary. According to the Ghana
National Shelter Draft Policy Document, housing tenancy in the country ranges from home
ownership to weekly tenancy. More than half (57.4%) of the national housing stock, is owner-
occupied. Rental units represent 22.1% of the stock while rent-free and ‘perching’ constitute
19.5% and 1.0% respectively (Government of Ghana, 2005). According to the Draft Policy
Document, quite a number of the houses in which people live in are rent-free. These are usually
called ‘ebusua fie’ (family home) usually for the extended family, long after the original owners
have died. A large number of extended family members and even some ‘strangers’ live rent-
free in these homes (UN-Habitat, 2010).
In recent times however, and within the context of the improved macroeconomic environment,
characterized by low inflation rates, low interest rates and relatively stable exchange rates,
activities in Ghana’s housing sector is gaining momentum (Bank of Ghana, 2007). Although
housing demand and supply gaps that are driven by a rapidly growing middle-class, as well as
increased urbanization remains, and as asserted by the Bank of Ghana, the rising mortgage debt
outstanding to GDP ratio provides evidence that the sector has recorded moderate growth over
the past seven years in particular, albeit from a low base. The gradual improvement in housing
supply notwithstanding, the sector still faces a number of challenges, such as land acquisition,
prolonged land title and registration processes, high costs of rental units and house prices that
require policy intervention.
176
5.4.3 Philosophical Basis for Housing Development in Ghana
Unlike housing development in Nigeria, which has its underpinning in its Constitution, the
philosophical basis for housing development in Ghana is not completely supported in their
constitution. The foundation is found in separate Ghana Development Plans as deemed
necessary by the previous and even the current government as a follow-up to their manifesto’s
before being elected into government. For instance, the present government involvement in
housing provision is as a result of the manifesto’s declaration, which states that: every
Ghanaian must have a home, though not necessarily own a house, as a meaningful expression
of the right to shelter. The Government Manifesto Agenda further states that: housing,
especially for the low and lower-middle income earners, will be one of the government’s top
priorities.
However, Chapter Five of the Ghanaian Constitution usually referred to as the Fundamental
Human Right and Freedom which contains the central passage on the human rights sets out a
range of rights to which every person in Ghana is entitled. Amongst these rights is that, every
person has the right to own property either alone or in association with others. That no person
shall be subjected to interference within the privacy of his home, property, correspondence or
communication except in accordance with the law and as may be necessary in a free and
democratic society for public safety or the economic well-being of the country, for the
protection of the rights or freedoms of others. Unlike the Nigerian Constitution that expresses
that the government is entrusted with the responsibility to harness the resources of the nation
to serve the common good of all and promote national prosperity based on an efficient, dynamic
and self-reliant economy, the Ghanaian Constitution is completely silent on this. However,
amongst the rights enshrined in Chapter Five of the Ghanaian Constitution, the right to housing
or adequate housing as a responsibility of the government is also not included. However, the
fact that the constitution is limited in its conception of human rights should not, in principle,
be a challenge to the protection of rights since the framers of the constitution purposefully
made the provisions on rights expandable. For instance the Ghanaian constitution states in
Article 33(5) that: “The rights, duties, declaration and guarantees relating to the fundamental
human rights and freedoms specifically mentioned in this chapter shall not be regards[sic] as
excluding others not specifically mentioned which are considered to be inherent in a democracy
and intended to secure the freedom and dignity of man”.
177
The constitution accordingly makes room to include into the legal framework those rights that
are articulated and protected outside the national space, and those that might exist in the future.
Despite Ghana being a signatory to numerous international and regional treaties that enforces
the right to adequate housing like the Africa Charter, there is no visible manifestation of those
rights by the government in the provision of housing to the citizens. In effect, the governments
of Ghana present and past only make housing a priority in their political development agenda;
none have tried to include it as a right in the constitution. The national philosophy for housing
development in Ghana is the political manifestos of the ruling parties, which are not sustainable
in the long run since it is not a legal framework. Thus, the Ghanaian Constitution does not
contain justifiable socio-economic rights that directly enshrine everyone’s right to have access
to suitable and adequate housing.
5.4.4 History and Development of Housing Policy in Ghana
There has been continuous government support in the housing sector in Ghana since the
Colonial Era. According to Tipple and Korboe (1998), in the first-half of the 20th Century, the
colony of the Gold Coast including Ashanti and the Northern Protectorate, had a fairly standard
attitude towards housing construction for the time up to independence in 1957. It started with
a motivation to house the British Civil Servants in some ‘splendour’ and separate from the local
people. This was to prevent the spread of the tropical diseases such as malaria, yellow fever,
and other debilitating diseases, which had no cure at that time. After from this, nothing was
done to improve the housing conditions of the local population. However, the public housing
schemes embarked upon by consecutive governments through the State Housing Company
Limited have only be successful in providing housing for the few who are relatively rich
(Wiredu, 2000). Indeed, the huge public investments in housing throughout the years have
produced less than the expected results, which is obvious in the severe shortage of housing in
Ghana up to date.
Public housing delivery has undergone socio-political transformation in Ghana’s history. The
pre-independence era witnessed the direct involvement of government in public housing. The
emphasis on developing the housing industry gained prominence in Ghana from the late 1950s
to the early 1960s as it attained independence from colonial rule (Bank of Ghana, 2007).
According to Nelson and Ayeh (2009) and Agyemang (2001) all housing schemes initiated by
various governments from pre to post independence era were unsuccessful due to a host of
178
factors and as such, the history of housing delivery in Ghana is ‘de-facto, a tale of failed
economic policies’.
As previously noted, the housing interventions during the pre-independence era took the form
of provision of staff houses for the senior public officer of the colonial governments in many
parts of the country, especially in regional capitals, towns and mining areas through direct
funding by the colonial government (Agyemang, 2001). The available housing schemes
initiated by major companies in the then Gold Coast only provided accommodation for their
expatriates, senior officers and junior workers and were not extended to the rest of the
population. It was only until the 1920s and after the incidence of the 1939 earthquake in Ghana
(Agyemang, 2001) that housing issues became more paramount. However, the main
approaches for housing improvement waited until after the Second World War; the new
emphasis on colonial development, and the need to show some gratitude to the heroic soldiers
returning from the battle fronts to their colonial homelands.
For instance, the Gold Coast Government’s first recorded and direct involvement in housing,
was in the 1920s when the Dispossessed Person‘s Housing Scheme was introduced to provide
housing for the natives dispossessed as a result of various government development
programmes. Under the scheme, which began in 1923, affected persons were given building
materials loan to commence the construction of their own houses. By 1933, 118 loans had been
approved and disbursed. The scheme was discontinued in 1933 because the Gold Coast
Government perceived it to be very expensive. This was during the tenure of Governor G.
Guggisberg‘s reign- (9th October 1919-24th April 1927) (Agyemang, 2001). From that time
onwards, little attention was paid to housing the natives until the 1939 earthquake. But, the
22nd June, 1939 earthquake in Accra, the now capital city of Ghana, called for the direct
intervention of the then government in the provision of affordable housing for the affected
population. The government provided funding to build 1000 two bedroom unit houses for the
affect people in various locations (Kwofie, Adinyira, & Botchway, 2011). By 1955, 1250 units
only were completed, which still exist up to date, but now are occupied by civil and public
servants and the armed forces. The housing provided through the earthquake intervention was
subsidized rental, and the tenants were given the opportunity to acquire them through hire-
purchase.
179
Further, the Alan Burns government (29th June 1942 - 2nd August 1947) also introduced a four
year Development Plan in 1943 of which housing was considered a top priority. The plan
sought to implement the construction of inexpensive but well-built houses with as much local
material content as possible on a limited budget (Agyemang, 2001). In 1946, two new housing
schemes (Schemes A and B) under the government plan and policies were published. Scheme
A of the new housing scheme was under the direction of Department of Social Welfare. Under
this scheme, three, two and one bedroom apartments were to be constructed and rented to all
people at economic cost. Only employees were required to pay non-economic rents. The
Scheme B was termed Town and Council Housing to be concentrated in Accra, Kumasi and
Sekondi-Takoradi. Under this scheme, a person could apply for financial assistance to build
within the municipality on his own design or to adopt a pro-forma building plan from the
Department of Social Welfare of which the plan and the contractor to be responsible for the
housing construction must be approved by the town council.
In the post-independence era, several interventions have been undertaken in the Ghanaian
housing industry in an attempt for the new government to address the housing situation of those
that were deprived of housing during the pre-independence period. However, all attempts made
were considered unsuccessful by experts and stakeholders. Under the reign of Dr. Kwame
Nkrumah (6th March 1957 - 24th February, 1966), the first president of Ghana, three
development plans were formulated all aimed at the provision of adequate housing for the
citizenry. The first development plan was the five year plan from 1951-1956. This plan
established the Tema Development Corporation (TDC) and the State Housing Corporation
(SHC) (Agyemang, 2001; Bank of Ghana, 2007). The main objective of the TDC was to
provide affordable housing for the low income workers of the newly created Tema Province.
The activities of the TDC led to the creation of the Communities of Tema such, as the
communities one to eight; thus contributing over 2255 housing units. The Schockbeton
Housing Scheme was also established, which was targeted at the provision of 168 houses in
Accra, Kumasi, and Sekondi-Takoradi. This scheme was specifically under the consultancy of
a Dutch firm, which introduced pre-cast materials perceived to be cheaper but later became
more expensive than estimated and hence this led to the abandonment of the whole scheme.
Likewise, the Roof Loan Scheme, which sought to grant loans and assistance to public sector
workers under the recommendation of the United Nations, also made contribution to the
provision of housing units in Ghana. However, due to inefficiencies with the system only 2517
180
units out of the originally proposed 6700 housing unit were constructed (Nelson and Ayeh,
2009). The established SHC was to provide housing for the workers in the civil and public
service class and also provide long-term housing finance. The activities of the SHC were
expanded to cover all the nine regions in the country. Their activities were monitored under the
Ministry of Works and Housing with direct funding of their projects from the central
governments and in 1995, it was converted into a limited liability company and operated as a
commercial enterprise. Their schemes operated with the flexibility of workers owning their
home through years of gradual monthly payment from salary deductions (Agyemang, 2001).
The Kwame Nkrumah Second Development Plan was instituted to continue provision of
housing from 1959-1964. This was to support the UN Commission‘s recommendations and
initiated programmes for the government to provide housing units for the citizenry. The plan
sought to continue and expand the ‘Roof Loan Scheme’, which was focused on assistance from
employers to employees through the provision of housing loans and self-help housing sites and
services (Nelson & Ayeh, 2009). The shortfall of this plan was that there was no needs-
assessment and as a result, there was no indication of projected targets and outputs in the
development plan (Agyemang, 2001). Despite the Development Plans version to provide
housing for all Ghanaians, Nkrumah‘s vision on housing was to house those in urban areas
where shortage was at its peak due to uncontrolled urbanization. However, he was not able to
see to the end of this plan before he was replaced by a different government in 1966. Generally,
the First Republican (Nkrumah) government used housing provision as a way of intervening in
population distribution through housing in towns where employment opportunities were
planned.
After the collapse of the Nkrumah’s government, the National Liberation Council (NLC) under
the leadership of Joseph Ankrah (24th February, 1966 - 3rd April 1969) took over control and
immediately adopted a two year Development Plan, which was run through the existing
systems put into place by the formal government. The NLC’s plan through the TDC and the
SHC was to produce 2,000 housing units annually. However, only a total of 1000 units were
realized. Out of the produced houses, 2.7 percent were one room apartments. According to the
housing location, 63.6 percent were constructed in Accra, 9 percent in Kumasi, 7.5 percent in
Sekondi-Takoradi and 11.3 percent in the Cape Coast. The main objective of this scheme was
to ensure that housing was generated by the productive sectors of the economy through rational
and balanced approach (Nelson & Ayeh, 2009). The Development Plan was also targeted at
181
the clearance and slowing down of the growth of slums in the urban areas.
The Kofi Busia’s (1st October 1969 - 13th January 1972) administration also showed
commitment to alleviating the housing crisis confronting the nation especially in the major
Ghanaian cities by introducing a one year Development Plan. The initial one year Development
Plan (1970-1971) of the second republic, later proceeded to a seven years Development Plan.
The main objective of this plan was aimed at a house occupancy rate of ten persons per house
as against a housing occupancy rate of six persons per household as recommended by the
United Nations. The plan was also aimed at the construction of an estimated 26,000 housing
units per year. However, the plan failed to specify the housing units with their associated cost.
At the end, the scheme added just 25% (764 - SHC, 1012 – TDC = 1776) of the targeted 8,000
units, which was mostly attributed to a lack of funds (Nelson & Ayeh, 2009).
When the National Redemption Council under I. K. Acheampong (13th January 1972 - 9th
October 1975) took over government in 1972, they quickly established the National Low Cost
Housing Committee under the auspices of the Ministry of Works and Housing. This plan
initially received a capital injection of 10 million Cedis to construct low cost housing for low-
income households in the urban areas across the ten Ghanaian regions. The plan had an annual
projected delivery of 2,300 units (Nelson & Ayeh, 2009). By June, 1975, the scheme had only
realized 5,466 units at a cost of 47,602,678 Cedis. The scheme was however abandoned in
1976 because of its failure to serve the targeted population due to its associated high cost
(Nelson & Ayeh, 2009). For instance, the original estimate indicated a cost of between 2,000 -
4 ,000 Cedis depending on the size; but upon completion of 5,466 units, an average per unit
cost stood at 10,000 Cedis ($9,803.92). Furthermore, 6,000 units cost a total sum of 62.6
million Cedis, thus increasing the average cost to over 12,000 Cedis. The government
acknowledging its limitation with funding sought to encourage the private sector to
complement her effort (Agyemang, 2001).
The Hilla Liman (24th September 1979 - 31st December 1981) Government also recognized
the enormity of the housing problems and thus contributed to the building of 1990 rental units
through the SHC and 228 by the TDC (Benjamin, 2007; Nelson & Ayeh, 2009). However,
because of the poor economic performance in Ghana with the rising energy crisis, rising costs
of oil, excessively high rise in imported building materials, decline in external funding, the
Ghanaian construction industry was brought to a halt which impeded the construction of houses
182
as planned by the Liman Administration. It was against this background that the Liman
government sought to invest in the development of the use of local materials leading through
to the establishment of the Tile and Brick Factory (Benjamin, 2007).
The Jerry J. Rawlings Era (31st December, 1981 - 7th January, 1993, 7th January, 1993 - 7th
January, 2001) saw the implementation of many schemes in an attempt to solve the housing
problems that other past administrators could not solve. Amongst the introduced schemes were
the National Shelter Strategy, Ghana Vision 2020 and the Structural Adjustment Programme
(SAP) and Economic Recovery Programme.
The National Shelter Strategy initiated in 1986, led to the formation of the National Housing
Policy Committee by the Ministry of Works and Housing (MOWH) to examine the housing
situation in Ghana. The intention was to establish an appropriate government policy and action
plan that seeks to provide adequate and decent housing units in order to improve the quality of
life of the people in urban and rural areas (Bank of Ghana, 2007). The focus of this committee
was on constraints in housing delivery, especially in the area of housing finance, land, physical
planning, infrastructure, building materials, design and construction and coordination delivery
efforts. The report of the committee culminated in a National Housing Policy and Action Plan
covering the period 1987 through to 1990. Prior to the formation of the Action Plan, the
MOWH had identified the need for a comprehensive National Shelter Strategy (NSS) and an
enhancement of the Ministry’s planning capacity to implement housing policies. The strategy
sought to: implement a revised national housing sector policy and action plan for short, medium
and long-term strategies for Ghana, with emphasis on rural communities in order to assist them
to improve their existing shelter or improve access to the means for providing their own shelter;
develop non-conventional housing delivery systems which encourages community
participation at all levels and with emphasis on local authorities playing a substantial role in
the management and development of housing. In summary, the NSS was to create an enabling
environment and framework to enhance housing provision rather than the full participation of
the government to the delivery of housing
The Ghana Vision 2020 Scheme had the First Medium-Term Development plan from 1997-
2000, target at the provision of affordable low-income housing units, which is within the reach
of the poor in order to improve their living conditions (Bank of Ghana, 2007). The vision
introduced a new facility under the Social Provident Scheme, which permitted contributors to
183
withdraw part of their contributions to purchase a house. Unfortunately, due to lack of funds,
low private sector participation and political will, none of the housing strategies under this plan
were implemented or have been implemented to date as 2020 is still a long way away.
No considerable additions had been made to public housing from 1985 to 2000. Nevertheless,
the John A. Kuffour government (7th January, 2001 - 7th January, 2009) also pursued the
reduction of the crisis situation of the housing sector through the initiation of about 20,000
affordable housing units in 2001. In 2007, building of about 4,500 units from bed sitter, single
and two bedroom apartment had started at various locations. This represented the new
government‘s effort to ease the housing problems in the country. The main target group of this
scheme was the civil and public servants. Unfortunately, before the expiration of the Kuffour
Government, not a single unit was completed and most has been taken over by squatters. The
scheme was discontinued by the new government in 2009.
The government of John Evans Atta Mills (7th January, 2009 – December 2012) drove for the
provision of housing in Ghana, which is fully enshrined in the Coordinated Programme of
Economic and Social Development Policies, 2010 – 2016; an agenda for shared growth and
accelerated development for a better Ghana. The Coordinated Programme of the Economic and
Social Development Policies is a strategic blueprint for directing national priorities in the
medium-term and for providing a framework for channeling national aspirations towards
accelerated industrial development. On careful scrutiny, the agenda is an expanded manifesto
of the National Democratic Congress presented in 2004 entitled ‘A Better Ghana’. The agenda
is operationalized to guide the execution of development policy and related activities at all
levels of the governance structure (GoG, 2010). The present government recognizing the ever
increasing population growth, rural-urban migration and the re-classification of settlements
from rural to urban and other population growth dynamics, which have contributed to the rapid
urbanization of Ghanaian towns and cities refuses to make a commitment to the exact number
of housing units to be constructed. Instead, a different approach was adopted to provide an
enabling environment for all stakeholders to participate in the housing delivery process. The
government informed that a special effort would be made, together with local and international
private sector partners, to launch a new national housing initiative to begin the rationalization
of the housing market in order to provide affordable housing for Ghanaians.
The involvement of the quasi-government institutions in the Ghanaian housing delivery cannot
184
be overlooked. Towards the end of the decade (1980-1990), the Social Security and National
Insurance Trust (SSNIT) expanded on its programmes to build housing for its staff across the
country. Though this was originally targeted at its staff, it was later expanded in 1988 in a
significant investment in housing at a ‘social’ and not market prices, providing a lower option
for the general public to also benefit from the scheme. However, at the end of construction,
SSNIT could not attain its objective for the poor and low-income earners/groups. The project
however benefited the middle and upper income classes (Benjamin, 2007). Also, another
SSNIT’s notable schemes were the 1637 units at Sakumono, where its success led to it being
repeated across the country in Anaji- Takoradi, Koforidua, Wa, Adenta-Accra, Kumasi
amongst others (Amoa-Mensah, 1999). Conversely, in 1999, when Ghana’s financial crisis hit
its peak, SSNIT was unable to continue operating its social rental units at a loss in so much that
its reduced rents were higher than what most Ghanaians could afford. Being saddled with huge
operational and maintenance cost, the trust began the process of divesting most of its real estate
assets. As of today, they have sold out almost more than 92 percent of its housing units.
Likewise, the State Housing Company (SHC), after its recapitalization in 1995, adapted a new
approach to housing delivery. Under this scheme, prospective home owners were given the
opportunity to finance their own home after first making a down payment of about 20-25% of
the cost of the building. This has come with little success, as many of the units are overrun in
time and cost. This scheme was carried out in all the regions where SHC operates. Also, the
Ghanaian Real Estate Development Association (GREDA) was formed to help improve the
dismal housing deficit, especially through the adoption of best practices in construction and its
management. Notwithstanding the expansive role of the GREDA in recent times, housing
supply has not increased any further (Ahadzie, 2008; Bank of Ghana, 2007). Since its formation
in 1988, the association has delivered a total of 10,954 housing units (Mahama, 2004). The
delivery from the formal Real Estate Developers (GREDA) annually at its peak, averaged 2,500
units constituting less than 10% of the total annual delivery. This is a drop in the ocean,
compared to the then annual housing requirement of 199,000 units. The actual delivery was
purported to be between 25,000 to 30,000 units.
5.4.5 Housing Policy in Ghana
From the above analysis of the history and development of housing policy in Ghana, it can be
inferred that the main policy directions in Ghana have been in the direct supply of quite a small
185
numbers of dwellings and a number of measures to influence demand. These have included
provision of low-income housing for the civil servants and others, rent subsidies and
subsequent purchase of government-built dwellings; subsidised interest rates for borrowers
from the few institutions concerned with housing finance, and a very successful rent control
regime starting in 1943 and only relaxed in 1987 (Tipple & Korboe, 1998). The National
Shelter Strategy and the Ghana Vision 2020 represent the current direction of urban housing
policy in Ghana with the United Nations assistance ran by the Ministry of Works and Housing.
Its six main objectives are:
improving the quality of shelter;
improving the environment of human settlements;
making shelter programmes more accessible to the poor;
promoting private sector involvement through an enabling policy environment;
encouraging rental housing; and
promoting orderly growth with infrastructure in place.
Until the early 1980s, housing was regarded as a social service in Ghana, which was to be
enjoyed by both the rich and poor. In the light of high incidence of poverty, income inequality
and acute housing shortages in Ghana, the consideration of housing as a social policy became
a matter of critical necessity. Housing provision in Ghana reflected this beliefs until it was
realized that this approach had left the government-approved bodies seriously indebted and
liable, and even unable to house the poor. A general shift in the national ideology in Ghana in
the early 1980s when the economy was liberalized which resulted in a re-orientation of the
approach to housing delivery. Housing is now seen as an economic commodity, which is
produced and sold for profit. The corresponding institutional response was the metamorphosing
of the State Housing Corporation into a limited liability company, which is expected to now
show maximum returns on investments in housing. However, public housing has and is still
not meeting the housing demand in Ghana. Besides, the supplied houses are widely not suitable
for the users. Grave housing inequalities are therefore visible in Ghanaian urban areas in the
form of slums and squatters settlement.
5.4.6 Challenges Facing the Provision of Housing in Ghana
The forgoing discussions reveal that the housing industry in Ghana is inundated / plagued with
an array of challenges. Currently, Ghana is facing an acute housing problem with a housing
186
deficit being in excess of 400,000 units. The most vulnerable groups are the urban and rural
poor whose houses are mostly constructed with sub-standard materials with little or no basic
services and infrastructure, including proper drainage and waste disposal systems. The key
challenges mitigating against effective housing delivery in Ghana, include the following:
poverty, land cost and its accessibility; finance, high cost of mortgage; infrastructure
development; development approval procedures; availability and cost of building materials;
institutional co-ordination; and a lack of adequate governance for shelter provision. In view of
this, the ultimate goal of the country’s shelter policy is to provide adequate, decent and
affordable housing that is accessible and sustainable with infrastructural facilities to satisfy the
needs of Ghanaians. Other factors also responsible for the inadequate provision of housing in
Ghana include: absence of clearly defined national housing policy; managerial inefficiencies;
high cost of building materials; lack of access to sustainable capital/finance; and Lack of
control and regulatory policy framework for rent (Bank of Ghana, 2007).
5.4.7 Housing in Ghana – Needs, Demand and Supply
Provision of housing in Ghana has witnessed fragmented and un-sustained effort from
individuals, private developers and the government. This situation has contributed to the huge
housing deficit encounter currently experienced in Ghana. The shortage of housing continues
to be one of the most critical socio-economic challenges facing the country at the present
moment. The rapid increase in population has resulted in a large housing deficit, especially in
the urban areas. This has manifested in overcrowding and the development of slums. According
to the Government of Ghana (2010), there is insufficient housing stock to meet the ever-
increasing demand for housing in the urban areas. Recent estimates indicate that there is a
housing backlog of more than one million units of houses nationwide. It is further estimated
that to replace this shortfall, annual national housing delivery should be approximately 120,000
housing units. However, the supply capacity nationally is 42,000 units per annum. Thus 60 per
cent of the national requirements will remain unsatisfied each year. Demand for housing is
higher in the major cities of the country, such as Accra, Kumasi, Tema, amongst others.
The main sources of housing demand in Ghana include: locally resident Ghanaians, non-
resident Ghanaians, expatriates living in Ghana, corporate organizations, and foreigners. An
increasing number of people from the West Africa sub-region also contribute greatly to the
demand for housing in Ghana. The Ghana housing supply source has a long history as
187
enumerated above. The government through the state-owned housing corporation has played a
significant role in housing delivery albeit, not enough to cater for the entire low-income earners
and poor groups. However, under the Government Liberalization Policy, since 1992 and the
current agenda for shared growth and accelerated development for a better Ghana, the state has
significantly reduced its role in the housing sector and encouraged the participation of the
private sector to provide housing. The Social Security and National Insurance Trust (SNIT), a
para-statal organization, has over the years developed a number of high income and middle
income housing either directly or through joint ventures.
5.4.8 Lessons Learnt from Ghana Housing Studies
The lessons learnt to date from the Ghanaian housing studies are not far from those achieved
by the Nigerian Government, which are:
Since Ghana’s independence, provision of housing has remained central to all
development agenda of the various governments as evident from the review of the
housing policy from the pre to post-colonial eras;
Also, various governmental administrations and institutions have sought to address
issues such as land ownership, land title regulation, and the provision of affordable
housing units to the working and non-working population;
From the above analysis of the history and development of housing policy in Ghana,
the literature revealed that the main housing policy directions in Ghana have been in
the direct supply of quite a small number of dwellings and a number of measures to
influence demand;
Also, the National Shelter Strategy and the Ghana vision 2020 is the country’s current
direction of urban housing policy in Ghana with the United Nations’ assistance, which
was formulated by the Ministry of Works and Housing. However subsequent
governments have re-modified the strategies and the vision to suit their own various
manifestos and purpose;
There was a shift from the formal national ideology in Ghanaian housing when the
economy was liberalized, resulting in the re-orientation of the approach to housing
delivery. Housing is now currently seen as an economic commodity in Ghana, which is
produced and sold for profit;
The Constitution of Ghana does not contain justifiable socio-economic rights that
enshrines everyone’s right to have access to adequate housing;
188
Further revealed from this section is that fundamental philosophy underpinning housing
development vision in Ghana are the various political manifestos of the various
governments that have ruled Ghana;
Current government intervention in housing is at a level of providing enabling
environment for the various stakeholders to help ameliorate the housing backlog.
5.5 HOUSING POLICY ISSUES: NIGERIA AND GHANA
Since housing has been recognised as a basic human need, policy programmes that promote a
housing sector capable of supplying adequate housing to the population are fundamental goals
of government’s social development strategy. Well-designed policies supporting the
production and consumption of housing services have significant impact on development. Not
only do they promote the expansion of the construction industry and hence, a push to growth
in GDP, they also help stabilize an economy that is responsive to price bubbles. Above all, they
increase the welfare of the population, particularly the urban and rural poor, by improving
living conditions, quality of life and expanding their physical assets. In view of these, a number
of policy considerations are imperative for Nigeria and Ghana to make giant strides in the area
of housing development and provision to their citizens:
1. There is the need to establish a well-defined system of housing finance to fund the
construction of new structures and trading of existing properties. This can be achieved
with the creation of a vibrant secondary mortgage market. This is based on the notion
that an efficient mortgage industry thrives on the increased secondary housing market
activities.
2. Establishment of a strong legal and regulatory framework, which comprises consistent
and holistic set of laws in areas such as property rights, collateral, foreclosures amongst
others, in the housing sector is highly required.
3. Also, alternative strategies to enhance mortgage financing that will benefit low to
middle-income earners and increase their access to affordable housing units could also
be pursued in conjunction with the private sector as part of a broader strategy to narrow
the countries housing deficit (Nigeria – 16 million housing deficit whilst Ghana has a
million housing deficit).
4. Finally, creation of land banks with major infrastructure facilities like roads, electricity
and water, as well as the establishment of land courts and decentralized land
189
administration systems may be considered as part of efforts toward the downscaling of
housing prices in the country. This calls for the amendment of the Land Use Decree of
1978 in Nigeria.
5.6 CONCLUSION
In this chapter, housing in developing countries was discussed with emphasis being given to
Nigeria and Ghana. According to the literature, the government of Ghana has a detailed track
record of a holistic national support for the provision of housing to its populace. However, with
the recent general development plan for the Ghanaian economy, the government has distanced
itself from the old practice of housing provision to supporting housing delivery by other
stakeholders. Likewise, the Nigerian government has been very influential in the provision of
housing to its citizens, but the challenge is that the low-income and the disadvantaged groups
for which the housing programmes, designed do not benefit from the programme as the
processes are politicized. The Nigerian Government has embarked on the provision of housing
for its citizens with a good Housing Policy Framework to fight the huge housing backlog.
However, the problem with management and implementation of housing policy by the agencies
given the responsibility has made these laudable policies ineffective. Also, it was found that
the present housing difficulty in Nigeria arises not necessarily from poverty, but because of the
absence of effective administrative machinery to mobilize and organize the country’s natural,
human, and industrial resources, amongst others; for housing and urban development. To date,
there is no government housing subsidy programme in place to help provide houses for the low
income groups. Nevertheless, a fundamental finding from the reviewed literature in both
countries revealed that their Constitutions does not explicitly contain justifiable socio-
economic rights that enshrine everyone’s right to have access to adequate housing despite being
signatories to international treaties, which advances the right to adequate housing. The next
chapter will focus on the review of literature of housing development in South Africa.
190
CHAPTER SIX
HOUSING IN SOUTH AFRICA
6.1 INTRODUCTION
This chapter of the thesis provides an outline of housing legislation and jurisprudence, policy
and implementation in South Africa since 1994. It outlines the housing legislative and policy
framework in South Africa; examining the Constitution with specific reference to the Bill of
Rights and the Right to Housing; the National Housing Code; and the National Housing
Programmes categorized therein with a specific focus on State subsidised housing (Housing
Subsidy Scheme). This chapter is also focused on an overview of the developments in housing
policy since 1994, including a summary of the negotiations at the National Housing Forum
held between 1992 and 1994. The section further examines the supreme policy framework
contained in the 1994 White Paper on Housing, and the problems associated with the
Reconstruction and Development Programme (RDP) houses (subsidised houses) built after
1994. This is necessary because the focal point of the thesis is on the RDP houses built after
1994. The chapter also discusses the 2004 Breaking New Ground Policy Amendment, amongst
others. Lastly, a summary of the lessons learnt to date from the literature is presented and a
comparison is drawn between South African and the two Africa countries (Nigeria and Ghana)
reviewed in the previous chapter.
6.2 HOUSING POLICY TRENDS IN SOUTH AFRICA
The housing environment in South Africa (SA) is complex, in large part due to the deliberate
policy and legislative framework of socio-economic and spatial exclusion and marginalization
created during the Apartheid Era. Also, the complexity of the housing process in SA is due to
many failures and a full understanding of the problems by the Apartheid Government and the
inability of the Post-Apartheid State Government to satisfactorily redress these problems since
1994. However, it must be genuinely acknowledged that the Post-Apartheid State Governance
has been actively involved in trying to create a level playing ground field for the previously
disadvantage and also trying to repair the disadvantaged condition created by almost 42 years
of the Apartheid Government. Simply put, it is easier to destroy than to create - so much so that
the Post-Apartheid Government has been faced with a situation that is not irreparable and
manageable, but a situation that needs patience and a little firmness to address. Hence, as with
191
other socio-economic rights, the legislative and policy framework created by the national
government around housing is progressive in addressing the situation on the ground. However,
implementation to date has been skewed and unable to address the land, housing and basic
services needs of millions of poor South Africans, who still lack adequate housing and access
to water, sanitation and electricity (Tissington, 2011). Whilst the urban and rural spatial divide
still remains pronounced in respect of access to socio-economic goods and services, the
phenomenon of the inadequately housed urban poor is increasing. Redressing the inherited
inequalities of the Apartheid State has established a complex and challenging context for
meeting basic needs in contemporary South Africa. Given the physical and political segregation
of Apartheid, meeting the demand for housing has been a central development challenge since
1994 (Pottie, 2004). However, apartheid alone cannot be held responsible for the housing
conditions in South Africa but equally no account of housing policy and conditions can be
credible if it does not take into account the history of South Africa and the colonial legacy of
the African continent (Goodlad, 1996).
Much of the debate in respect of the SA housing policy has been centered on the politico-
economic background of the policy itself. The policy has been described by some as an
extension of World Bank neo-liberalism, while others have labeled it as economically
conservative (Pottie, 2003). However the South Africa housing policy is rather the result of a
mixed bag of international influences and local creativity - mostly due to the policy of spatial
segregation in the Apartheid State, which contributed to a policy which is defined in terms of
‘scan globally, reinvent locally’ principle according to Gilbert (2004). Clearly seen, a number
of World Bank policy elements comprise an integral part of the South Africa policy, for
example, the emphasis on incremental housing, economic conservatism, the once-off subsidy
element and the instrumental role of formal ownership (Marais, 2007). However, it should also
be acknowledged that two important differences are discernible in the South African policy.
Firstly, the South African policy suggests that only housing structures should be subsidised.
Whilst the World Bank policy suggests that only site and services should be subsidised.
Secondly, as the South African policy has developed, an increasing emphasis has been placed
on housing size – a factor which does not comprise part of the World Bank policy (Marais,
2007).
Although it can be contended that the current South African housing policy was founded on
the RDPs basic needs approach, which emphasises providing the poor with basic shelter, public
192
services and addressing the social unequal nature of the Apartheid State. The South African
housing agenda has objectives that go beyond this. For instance, an important aim in the policy
is that housing should contribute to the national economy. Also, in line with the World Bank’s
view of the 1990’s, the South African housing policy also aims to enable the housing markets
to work. Hence, the housing programmes are expected to contribute to the development of
urban citizenship with the creation of new communities of home owners helping to develop a
democratic and integrated society.
However, certain key dimensions of the current policy framework cannot be fully grasped in
isolation from an understanding of the legacy bequeathed by South Africa’s peculiar history
(Wilkinson, 1998). In a very direct sense, the problems of the past have profoundly shaped the
situation to which the present policy seeks to respond. The next section of this chapter presents
a historic development of the evolution of housing policy in South Africa.
6.2.1 The Evolution of Housing Policy in South Africa
According to Wilkinson (1998), housing policy is an element of the trend of all modern states
in the twentieth century, in order to intervene extensively in the societies over which they
exercise power. The moment South Africa first emerged as a recognisable policy arena in the
early 1920s, housing policy was greatly involved in the state’s efforts to establish and maintain
a particular social order, sometimes referred to as ‘racial capitalism’ (Saul & Gelb, 1981;
Wilkinson, 1998). The housing policy in South Africa has generally been a contentious issue,
since 1910. De Loor (1992) refers to housing as either or both emotional and a very personal
issue in South Africa.
When the Union government was established in 1910, they developed several strategies in form
of ‘ACT’ (decrees) to control the movement of blacks, especially in areas referred to as white
urban areas. Since 1910, to the end of the segregation rule, various approaches were used to
advance the inhumane idea. Most significantly from the literature was the drafting of the
Segregation Policy, which was advanced at national and provincial level of government.
Foremost in the Acts, was the Natives’ Land Act 27 of 1913. This Act was concerned with land
issues, and since land and housing issues are inextricably linked, this also effected the provision
of housing for the blacks and other groups (Phago, 2010). The enactment of the Native Land
Act 27 of 1913, cemented housing policy issues in the apartheid era, which created the divide
in housing issues up to date. This meant that houses could only be built where the land had
193
been made available through proper government approval. As such, the Native Land Act 27 of
1913 had direct implications on the housing situation in the country since it specified the
territorial separation of the black and white races (Morris, 1981; Reader’s Digest, 1989; South
African Native National Congress, 1916).
According to De Loor (1992) and Morris (1981), following the establishment of the Native
Land Act 27 of 1913, the then Central Housing Board introduced the first housing policy
documents in 1920, called the Housing Act 35 of 1920. The purpose of the Board through the
drafting of the Housing Act was to have a control of the development of houses in local
authorities, with a special control on the mechanisms of financing, which had a sinister motive
to deprive the blacks and other groups of any assistance to receive housing. De Loor (1992)
and Morris (1981) further inform that during the first two decades of the Board’s existence,
expenditure was allocated to alleviating the housing plight of poor whites only without any
extension to the blacks and other groups. Nonetheless, a broader evaluation of the Housing Act
by Rodgers (1980) indicates that the Housing Act only strengthened the policy of separate
development. Besides, although these policies were introduced with good intentions such as
developing communities based on their ethnic locations, it was later more evident that housing
became an instrument for the implementation of the policy of separate development.
Following the Land Act of 1913, the Housing Act 35 of 1920, the Native Act of 1923 were
also enacted. This Act lasted for more than 60 years, until its desertion in 1986 on attempts to
enforce ‘influx control’ on African urbanization. The key provisions of this legislation
remained at the core of efforts to be achieved, during the 1930s, ‘total segregation’ and, after
the National Party Government came to power in 1948, ‘Grand Apartheid’. The ‘Stallard
Principle’ (1923 Natives Act) itself held that the right of municipal ‘enfranchisement’ should
be denied to African urban residents only if they are given right to permanent residence in those
areas (Wilkinson, 1998). This policy directly withdrew the rights of the blacks to freehold
tenure of urban land. Consequently, ‘the native’ was to be permitted to enter the ‘white’ cities
and towns only ‘to minister to the needs of the white man and should depart therefrom (to
return to the ‘Reserves’) when he ceases so to minister’ (Transvaal Provincial Administration,
1922). Thus, blacks were considered as ‘temporary citizens’ in all areas outside of their
homelands. As a result of this, South Africa continued to develop housing backlogs, which still
continues to date. Inherent in all the enacted decrees was the policy of separate development
propagated by the Apartheid Government to deceive the world into accepting Apartheid
194
Policies as another developmental approach with no racial degradation pertinent to South
Africans (Phago, 2010).
On the other hand, a considerable number of Africans were already long established in the
major urban centres, with some already having acquired freehold on properties. The
implementation of the Stallardist doctrine required that a comprehensive system of social
control be established (Wilkinson, 1988). Hence, the first element of this system eventually
became a massive accretion, of truly Kafkaesque complexity, of ‘pass laws’ and labour bureaus
which, apart from their primary task of regulating the supply of African labour to the various
sectors of the economy, could be used to control the movement of Africans to the ‘white’ urban
areas (Hindson, 1987). The second element of the system was the institutionalisation of the
form of residential segregation known successively as the ‘location’ or ‘township’. According
to Wilkinson (1998), the fundamental purpose underlying the prolonged and often ferociously
contested efforts to segregate the African urban population into separate residential areas, was
to regulate the degree of permanence with which the African population could establish itself
there. Thus, it is in relation to this strategy of ‘containing’ African urbanization (Fair &
Schmidt, 1974) through a cruel but highly developed racial oppression that the evolution of
South African housing policy must be understood (Wilkinson, 1998).
Prior to 1920, the only efforts to regulate or improve the generally very poor housing conditions
of Africans living in the urban areas in South Africa were irregular ventures by the larger local
authorities to clear so-called ‘plague spots’ and a few half-hearted efforts to establish municipal
‘native locations’, invariably far removed from the rest of the city or town. In contrast to this,
the provision of barracks and compounds to house single and domestic workers, usually
migrants, was already well established. However, for Africans and, in general, the poorer
sections of the population as a whole, were left to cater largely for themselves, with many
ending up in squalid, overcrowded and very unhealthy slum tenements or ‘yards’ (Wilkinson,
1998). However, Morris (1981) states that during this period, blacks in the rural areas were
accustomed to building their own traditional dwellings. According to Phago (2010), this
finding from the work of Morris (1981) has attracted considerable interest from the post-1994
government as this form of development (building of traditional dwellings) has contributed to
the fact that proper and quality housing in South Africa is an integral part of government policy
to the provision of housing.
195
Furthermore, the formation of the very first Central Housing Board (CHB) in terms of the 1920
Housing Act was a direct consequence of public concern about the impact of the devastating
influenza epidemic of 1918. The formation seems also to have reflected a growing unease,
which subsequently underpinned the codification of the ‘Stallard Principle’ in the 1923 Urban
Areas legislation about the accelerated influx of Africans into the urban and industrial heartland
during the First World War (Rich, 1978). The CHB provided the somewhat limited means to
enable black access to adequate housing, while the Urban Areas Act had the objective, of the
programme of residential segregation of the African population, which gradually unfolded
during the 1920s and 1930s in the larger centres. For the most part, the initiation of ‘slum
clearance’ schemes and the building of municipal ‘locations’ were hindered during the period
between the World Wars by the continuing unwillingness and the limited capacity of the local
authorities to bear the costs involved in fulfilling their statutory obligations (Wilkinson, 1998).
The central government, on the other hand, prevented any extension of its financial
responsibilities for executing residential segregation in the urban areas, which devolved
essentially to making subsidies available for the provision of very basic ‘sub-economic’ houses
by the local authorities (Wilkinson, 1998).
In the beginning, the newly installed Nationalist Government directed its attention primarily to
choke off what had become an only partially controlled flow of Africans to the cities. By the
end of 1950, Hendrik Verwoerd the key ideologue of Apartheid was brought into the Cabinet
as Minister of Native Affairs and, with the assistance of Dr W. M. Eiselen as his Departmental
Secretary; he moved with a single-minded determination to resolve the ‘Bantu housing
problem’ to truly accord with the commitment of the National Party’s 1948 election manifesto
which was the ‘ideal of total Apartheid’. According to Hart’s (1990), the policy of separate
development was further reinforced in spite of media reports condemning it and despite the
international stance towards South Africa. One aspect of the proposed resolution of the ‘Bantu
housing problem’ involved an effort to eliminate the apparently uncontrollable areas ‘held in
Native ownership’. According to Wilkinson (1998), the instant target was undoubtedly the
various so-called ‘black spots’ on the perimeters of some of the larger centres, which generally
had their origins in irregular sales of land to Africans during a slump in the property market
after the South African War at the turn of the century. However, the fundamental attack was
on the right of Africans to freehold land tenure in ‘white’ areas. In retrospect, all efforts to
address the other aspects, such as the ‘provision of adequate housing in properly planned Native
196
townships’, can now be seen to have produced a material legacy which came to lie at the centre
of subsequent attempts to restructure the South African housing policy (Wilkinson, 1998).
According to De Loor’s (1992) the period between 1978 and 1986 was evidently a resistance
period through the perpetuation of violence in townships and mass industrial actions.
Accordingly, through such pressures, the then National Party Government introduced reform
measures to address the economic and political crises, such as poverty and political unrest
during that period. One of the most important victories for blacks was the acceptance by the
then Apartheid Government of the permanence of black settlements in non-homeland cities
and towns, as well as the introduction of new institutions to accommodate regional labour
markets. Furthermore, Hart (1990) argues that the release of political prisoners and the
subsequent commencement of the democratic negotiations during 1990 brought about a new
direction in the housing policy, especially through the appointment of the De Loor Commission
of enquiry in 1991. The Commission was tasked to investigate the status quo regarding housing
matters and to advise on the new housing policy and strategy. The new housing policy and
strategy, generally viewed as the housing vision, was intended to encompass principles such as
adequate shelter for all, security of tenure, equitable access to potable water, sanitary facilities
and refuse removal as well as access to energy sources, including electricity (De Loor, 1992).
6.2.2 Housing Statutory and Policy Framework in South Africa
It is imperative to note at this very point, that the details of the current housing policy in South
Africa remains somewhat fluid and that the overall policy framework has yet to be cast in its
final statutory form. According to Wilkinson (1998), this on-going fluidity is undoubtedly due,
for the most part, to the significant political changes that the country has undergone in the last
few years, since the installation of South Africans’ first fully democratic government in mid-
1994 and the initial adoption of the ‘Reconstruction and Development Programme’ (RDP)
presented by the African National Congress (ANC) in 1994. This is an overarching framework
for the formulation and implementation of policy in a wide range of social and economic policy
arenas, including housing.
The South African current housing policy is rooted in the 1994 Housing White Paper. The
fundamental policy and development principles introduced by the Housing White Paper
remains relevant and guide all developments in respect of housing policy and implementation.
The fundamental policy framework of the South Africa housing policy established in the White
197
Paper, which was published in December 1994, contains the fundamental principles of the
government’s housing policy to achieve the housing vision. The housing vision is underpinned
by principles of sustainability, viability, integration, equality, re-construction, holistic
development and good governance. According to the Department of Human Settlement, the
South African housing policy and strategy must contribute to a non-racial, non-sexist,
democratically integrated society. The goal is to improve the quality of living of all South
Africans with an emphasis on the poor and those who cannot independently satisfy their basic
housing needs. Furthermore, the government’s human settlement development mandate
emanates from the South Arica Constitution, of 1996. Based on the provisions of the
constitution, it is the government’s duty to work progressively towards ensuring that all South
Africans have access to secure tenure, housing, basic services, materials, facilities and
infrastructure on a progressive basis. The government is therefore required to apply legislative,
administrative, financial, educational and social measures to fulfill its housing obligations. The
following is a list of some of the primary and secondary legislation relating to various
regulatory, financial, technical, environmental, institutional and developmental aspects of
housing in South Africa enacted within the past seventeen years:
Primary legislations
Constitution of the Republic of South Africa, 1996 (Constitution);
Prevention of Illegal Eviction from and Unlawful Occupation of Land Act 19 of 1998
(PIE Act);
Housing Act 107 of 1997 (Housing Act);
Housing Consumers Protection Measures Act 95 of 1998;
Housing Amendment Act 28 of 1999;
Rental Housing Act 50 of 1999 (Rental Housing Act);
Housing Second Amendment Act 60 of 1999;
Local Government: Municipal Systems Act 32 of 2000 (Municipal Systems Act);
Housing Amendment Act 4 of 2001;
Rental Housing Amendment Act 43 of 2007;
National Norms and Standards for the Construction of Stand Alone Residential;
Dwellings Financed through National Housing Programmes (2007) (National Norms
And Standards);
Social Housing Act 16 of 2008 (Social Housing Act); and
198
Housing Development Agency Act 23 of 2008.
Secondary legislations
Expropriation Act 63 of 1975;
National Building Regulations and Building;
Standards Act 103 of 1977 (NBRA);
Sectional Titles Act 95 of 1986 (amended by Acts 24 and 29 of 2003);
Environment Conservation Act 73 of 1989 (amended by Act 79 of 1992);
Land Titles Adjustment Act 111 of 1993 (LTA);
Development Facilitation Act 67 of 1995 (DFA);
Land Reform (Labour Tenants) Act 3 of 1996;
Interim Protection of Informal Land Rights Act 31 of 1996;
Extension of Security of Tenure Act 62 of 1997 (ESTA);
Water Services Act 108 of 1997;
National Environmental Management Act 107 of 1998 (NEMA);
Public Finance Management Act 1 of 1999 (PFMA);
Home Loan and Mortgage Disclosure Act 63 of 2000;
Division of Revenue Act 7 of 2003 (DORA);
Municipal Finance Management Act 56 of 2003 (MFMA);
Intergovernmental Relations Framework Act 13 of 2005 (IRFA); and
Co-operatives Act 14 of 2005.
In the following sections, highlights of the important information contained in the Constitution
of the Republic of South Africa, 1996 (Constitution) and key housing-related policy and
legislation, including the Housing Act 107 of 1997 (amended by Acts 28 and 60 of 1999; Act
4 of 2001) (Housing Act) will be discussed.
6.2.2.1 Constitution of the Republic of South Africa (1996)
The fundamental philosophy underpinning housing development goals in South Africa is the
existing South Africa Constitution of 1996. The Constitution of the Republic of South Africa
of 1996 is the supreme law of the country. It is the basis of all activity in the Republic of South
Africa. This means that any law or conduct that is inconsistent with the Constitution is invalid,
and that the obligations that it imposes must be fulfilled. Everything done must conform to
199
what is contained within the Constitution. In the context of Post-Apartheid South Africa, it is
significant that on the basis of the Constitution, National Housing Policy as contained within
the Housing White Paper and the Housing Codes applies equally to the total geographic area
of the Republic of South Africa. At a most basic level, the constitution defines the fundamental
values, such as equality, human dignity, and freedom of movement and residence, to which the
housing policy subscribes. These notions are also contained broadly in the Bill of Rights, in
Chapter 2 of the Constitution. Two components of the constitution especially relevant to
housing are: the specific right to have access to adequate housing, as enshrined in section 26
of the Constitution; and the powers of national, provincial and local governments with respect
to housing activities are framed by the concept of ‘concurrent competence’ and developmental
local government.
The Constitution also contains justifiable socio-economic rights and enshrines everyone’s right
to access to adequate housing. For instance, in the Bill of Rights in Chapter 2 of the
Constitution, section 26(1-3) outlines: “Everyone has the right to have access to adequate
housing. The state must take reasonable legislative and other measures, within its available
resources, to achieve the progressive realisation of this right…” The Bill of Rights also includes
a number of other rights, which relate either directly or indirectly to the enjoyment of the right
to housing. These include, among others:
Everyone has inherent dignity and the right to have their dignity respected and
protected (section 10);
No one may be deprived of property except in terms of the law of general
application, and no law may permit arbitrary deprivation of property (section
25(1)); and
Everyone has the right to sufficient water (section 27(b)) (The Water Services Act
108 of 1997);
Every child has the right to basic shelter (section 28(c)).
While Section 26(1) of the Constitution enshrines that everyone has the right to access to
adequate housing and that it is the government’s responsibility to take reasonable legislative
and other measures, within its available resources, to achieve the progressive realisation of this
right. When defining the concept of adequate housing, the wording of the housing right
200
provision corresponds with the International Covenant on Economic, Social and Cultural
Rights (1966). However, the concept of ‘adequate housing’ is not easy to define as it depends
on the exact context and circumstances of households and individuals, together with their needs
and priorities (Tissington, 2011). Similarly, the right to housing is also bound to the rights of
other socio-economic goods and amenities including rights to land, water, sanitation,
electricity, livelihood, transport, clinics and hospitals, schools, universities and other cultural
and recreational amenities, such as parks, libraries, public spaces, swimming pools, sports
fields, etc. In addition, Tissington (2011) further claims that achieving the right to housing is
also fundamentally bound to a number of other cross-cutting rights, including rights to public
participation, equality, human dignity, just administrative action, freedom of expression, access
to information and access to justice etc. This inter-relatedness and ‘interdependency of rights’
is acknowledged by the international human rights law as the principle of interdependency of
rights, which means that socio-economic rights and civil and political rights are interrelated,
and that the enjoyment of one right (or group of rights) requires enjoyment of others (which
may or may not be in the same group).
In terms of Section 26 of the South Africa Constitution, the government should endeavour to
ensure that all people living in South Africa are able to satisfy all the requirements with regard
to access to adequate housing as engraved in the constitution. In responding to the
constitutional right to ‘access to adequate housing’ for all South Africans, government is under
an obligation to not only pass enabling legislation, but also to apply other measures of an
administrative, a financial, educational or a social nature to fulfill its housing obligations. This
commitment according to Tissington (2011) would characterize the move to transformation,
equality and socio-economic well-being for all citizens (and non-citizens). Since the foundation
to housing provision is enshrined in the new South Africa Constitution, individuals unable to
access housing through the ‘normal’ residential market should be further assisted with new
schemes that will make that right a realization. Thus, state prioritization should be people-
centered, so as to ensure their access to livelihoods as contained in the constitution.
6.2.2.2 The Housing Act (1997)
Apart from the South Africa Constitution that contains justifiable socio-economic rights and
enshrines everyone’s right to have access to adequate housing; the South Africa Housing Act
of 1997 is the primary piece of housing legislation in South Africa. It legally entrenched
201
housing policy principles outlined in the 1994 White Paper on Housing. The South Africa
housing vision is confirmed in the Housing Act, 1997 (No. 107 of 1997). Within the Housing
Act, ‘housing development’ is defined as: 1(vi) “… the establishment and maintenance of
habitable, stable and sustainable public and private residential environments to ensure viable
households and communities in areas allowing convenient access to economic opportunities,
and to health, educational and social amenities in which all citizens and permanent residents of
the Republic will, on a progressive basis, have access to:
(a) permanent residential structures with secure tenure, ensuring internal and external
privacy and providing adequate protection against the elements; and
(b) potable water, adequate sanitary facilities and domestic energy supply.”
The housing goals are further reiterated in both the Urban and Rural Development Frameworks.
In each of these documents, the environment within which a house is situated is recognised as
being equally as important as the house itself in satisfying the needs and requirements of the
occupants (Housing Code, 1997). The ultimate goal of the Housing Act is that the housing
process must make a positive contribution to a non-racial, non-sexist, democratic and
integrated society. The goal of the Housing Vision for both urban and rural areas is to improve
the quality of life of all South Africans with a special emphasis on the poor and those who have
been previously disadvantaged.
The Housing Act further provides for a sustainable housing development process, laying down
general principles for housing development in all spheres of government. It defines the
functions of national, provincial and local governments in respect of housing development; and
it lays the basis for financing national housing programmes. In Section 2(1) the Act states that
“all spheres of government must give priority to the needs of the poor in respect of housing
development, and consult meaningfully with individuals and communities affected by housing
development”. This forms the basis for housing participation in South Africa, which is the
major subject discussed in the current thesis. They must ensure that housing development
provides, as wide a choice of housing and tenure options as is reasonably possible; is
economically, fiscally, socially and financially affordable and sustainable; is based on
integrated development planning; is administered in a transparent, accountable and equitable
manner; and upholds the practice of good governance in all sectors. In addition, Section 2(1)(e)
of the Act states that all spheres of government must promote inter alia the following: a process
of racial, social, economic and physical integration in urban and rural areas; measures to
202
prohibit unfair discrimination on the ground of gender and other forms of unfair discrimination
by all actors in the housing development process; higher density in respect of housing
development to ensure the economical utilization of land and services; the meeting of special
housing needs including the needs of the disabled; the provision of community and recreational
facilities in residential areas; the housing needs of marginalized women and other groups
disadvantaged by unfair discrimination. However, amendment was made to the principal Act
in 1999 and 2001, respectively. Section 4 of the Housing Act requires the minister to publish a
Code which includes the national housing policy and procedural guidelines for the
implementation of the policy.
6.2.2.3 National Housing Code (2000, revised in 2009)
The National Housing Code (NHC), which was first published in 2000 in accordance with the
Housing Act, set out the underlying policy principles, guidelines and norms and standards,
which apply to the National Housing Programmes. The NHC 2009 sets the same underlying
policy principles, guidelines and norms and standards, which apply to government’s various
housing assistance programmes introduced since 1994. Some of the initially created
programmes have been updated or removed, and new programmes included, after the adoption
of Breaking New Ground in 2004. The NHC is binding on provincial and local spheres of
government. The Housing Code includes the national housing vision, housing goal, basic points
of departure and the fundamental principles of the Housing Policy. The NHC set the tone for
the understanding of existing policies and the development of new ones. It is the basis for all
housing activities in South Africa. The National Housing Code 2000 has been substantially
revised. The revised NHC of 2009 is aimed at simplifying the implementation of housing
projects by being less prescriptive while providing clear guidelines, as against the initial
provision of the NHC of 2000.
The NHC set out the national housing policy of South Africa, collectively with practical
guidelines for its effective implementation through the inclusion of the National Housing
Programmes. The Code’s vision for housing in South Africa echoes the definition of ‘housing
development’, as outlined in the Housing Act. The initial 2000 Code states that the
government’s housing goal is subject to fiscal affordability, to increase housing delivery on a
sustainable basis to a peak level of 350 000 units per annum until the housing backlog is
overcome (NHC, 2000). In 2004, Breaking New Ground Policy (BNG) made provision for a
203
new National Housing Code to be published, with the intention to align and cohere with the
policy so that its goals and aims can be implemented. The new Housing Code is meant to
accommodate any changes made since 2000 and to convert the National Housing Programmes
into flexible provisions and guidelines (Tissington, 2011). Further to the BNG provision, the
revised National Housing Code was adopted and published in February 2009. A few old
programmes have been removed from the new Code, such as the Project Linked Subsidy
Programme, Relocation Assistance Programme, Blocked Projects Programme and
Rectification of RDP Stock 1994-2002 Programme. However, it is important to note that in
respect to all programmes, which are not contained in the Housing Code 2009, the rules of the
National Housing Code 2000 still apply whenever they are initiated. The National Housing
Code is proposed to be revised on an annual basis in order to ensure that it keeps up-to-date
with legislative or policy changes. However, this has not been done since the last revision of
2009. The revised NHC is all-embracing and addresses a variety of housing programmes
mentioned in BNG.
6.2.2.4 National Housing Programmes
The government’s primary housing objective since the dawn of the new administrative
government is to undertake housing development, which Section 1 of the Housing Act, No.
107 of 1997 (‘the Housing Act’) defines as being the establishment and maintenance of
habitable, stable and sustainable public and private residential environments to ensure viable
households and communities in areas allowing convenient access to economic opportunities,
and to health, educational and social amenities in which all citizens and permanent residents of
the Republic will, on a progressive basis, have access to: “Permanent residential structures with
secure tenure, ensuring internal and external privacy, and providing adequate protection against
the elements, and potable water, adequate sanitary facilities and domestic energy supply”.
The past and existing national housing programmes have been based on this objective and the
principles embodied therein. However, in response to the South Africa Constitutional
imperative, government has in terms of the Housing Act, 1997 (Act No 107 of 1997) introduced
a variety of programmes which provide the poor households access to adequate housing. Thus,
the policy principles set out in the White Paper on Housing aims to provide poor households
with houses, as well as basic services such as potable water and sanitation on an equitable basis.
The limited resources available from the ‘fiscus’ however necessitate the provision of housing,
204
security and comfort to all over time. The national housing programmes are categorized into
different ‘Intervention Categories’ as follows:
Financial Programmes
Individual Housing Subsidies;
Enhanced Extended Discount Benefit Scheme (EEDBS);
Social and Economic Facilities;
Accreditation of Municipalities;
Operational Capital Budget (OPS/CAP);
Housing Chapters of IDPs; and
Rectification of Pre-1994 Housing Stock.
Incremental Housing Programmes
Integrated Residential Development Programme (IRDP);
Enhanced People’s Housing Process (ePHP);
Informal Settlements Upgrading Programme (UISP);
Consolidation Subsidies; and
Emergency Housing Assistance.
Social and Rental Housing Programmes
Institutional Subsidies;
Social Housing Programme (SHP); and
Community Residential Units (CRU).
Rural Housing Programmes
Rural Subsidy: Informal Land Rights; and
Farm Residents Housing Assistance Programme.
From the above list, the Department of Human Settlement (DHS) has identified three
programmes as core programmes for future housing delivery, which are the IRDP, UISP and
Social/Rental Housing Programme. The following section provides a detailed review of the
IRDP, the Social housing Programmes and the enhanced People’s Housing Process (ePHP).
Even though ePHP is not listed as a core programme for housing delivery, but relates more to
205
the present study because of its central theme of citizen participation. Also, it should be noted
that there is a set of generic qualifying criteria, which must be fulfilled by those applying for
state housing subsidies under the National Housing Subsidy Scheme (NHSS) for the national
housing programmes. However, there are also specific rules that apply to each subsidy
programme and in some cases there are specific eligibility criteria that apply over and above
the generic criteria (DHS, 2009). The specific criteria are discussed below.
6.2.2.4.1 Qualifying Criteria for Housing Subsidy Programmes
The set of generic qualifying criteria as outlined in the revised National Housing Code of 2009
for consideration in the national housing programmes are summarized as follows:
1. Citizenship: All applicants must be citizens of the Republic of South Africa, or be
in the possession of a Permanent Resident Permit;
2. Competent to contract: Applicant must be legally competent to contract (i.e. over
18 years of age, and of sound mind);
3. Not yet benefited from government funding: The applicant or their spouse may not
have received previous housing benefits from the government. In the event of a
divorce involving a person who previously derived benefits, the terms of the divorce
order will determine such person’s eligibility for further benefits;
4. First time property owner: The applicant or their spouse may not have owned and/or
currently own a residential property. Except for the following cases, such as for
disabled persons; persons who own a vacant stand that was obtained through the
Land Restitution Programme; have acquired a residential property for the first time
without government assistance and the house/dwelling on the property, if any, does
not comply with the National Norms and Standards in respect of permanent
residential structures;
5. Married or financial dependents’: The applicant must be married or constantly be
living together with a spouse. A single person with proven financial dependents’
(such as parents or parents-in-law, grandparents or grandparents-in-law, children,
grand-children, adopted children, foster children) may also apply;
6. Monthly household income: The applicant’s gross monthly household income must
not exceed R3 500. Adequate proof of income must be submitted;
7. Beneficiaries of the Land Restitution Programme: Beneficiaries of the Land
Restitution Programme, should they satisfy the other qualification criteria, may
apply for housing subsidies;
206
8. Persons classified as military veterans as confirmed by the South African National
Defence Force (SANDF): Military veterans who are single without financial
dependents’ may also apply for housing subsidies;
9. Persons classified as aged: Aged persons who are single without financial
dependents’ may also apply for housing subsidies. Aged persons are classified as
male and female persons who have attained the minimum age applicable to
government’s old age social grant scheme; and
10. Persons classified as disabled: Persons, who are classified as disabled, whether
single, married or co-habiting or single with financial dependents’, may apply for
housing subsidies. If a person who has already received state funding for housing
and/or who already owns or owned a house, is or becomes disabled, or if his or her
dependant(s) is/are or become disabled, such a person may receive an additional
variation on the subsidy amount to finance special additions to provide independent
living conditions.
6.2.2.4.2 Integrated Residential Development Programme (IRDP)
According to the DHS (2010), one of the key lessons learnt in the review of the outcomes of
housing programmes since 1994 is that, owing to an array of reasons, the initially developed
low income settlements continued to be located on the urban periphery without the provision
and consideration for social and economic amenities, as in the Apartheid Era. Hence, this
necessitated the introduction of the IRDP, which replaces the previous Project Linked Subsidy
Programme, in order to facilitate the development of an all-inclusive human settlement in well-
located areas that provide convenient access to urban amenities, including places of
employment. As stated on the IRDP document (DHS, 2009), IRDP provides a tool to plan and
develop integrated settlements that include all the necessary land uses and housing types and
price categories to become a truly integrated community.
The programme is also aimed at creating social cohesion. The IRDP provides for the
acquisition of land, servicing of stands for a variety of land uses including commercial,
recreational, schools and clinics, as well as residential stands for low, middle and high income
groups. The land use and income group mix is based on local planning and needs assessment,
which is a basic tenet of the present thesis. This is because if the concept of needs assessment
is prioritized, the end product will eventually satisfy the need of the occupants as they will be
aware of what they will receive when the houses are allocated to them. The IRDP provides for
207
phased area-wide planning and development of integrated housing projects in situations where:
1. a project is undertaken in an area where unoccupied vacant land is developed; or
2. a project is undertaken in an existing township where an undeveloped parcel of land
is utilised for development purposes.
However, IRDP moves away from the requirements found in the previous Project Linked
Subsidy Programme version which requires the identification of subsidised housing
beneficiaries up front and provides for both subsidised, as well as finance-linked (Misselhorn,
2008); housing that only caters for households earning between R3 500 and R7 000 a month.
IRDP further provides for social and rental housing, commercial, institutional and other land
uses to be developed.
In the plans of the IRDP, a municipality assumes the role of the developer (where they lack
financial, technical and managerial capacity, a provincial department can take on this role),
undertaking all planning and project activities. As developers the municipality appoint
construction industry professionals (who design and establish the township, design and monitor
the installation of services, and design the houses) and contractors (who construct the services
and housing) to assist with the housing development process. Municipalities apply for funding
from the provincial Department of DHS who approves project applications, reserves and
distributes funds, as well as assesses and adjudicates various aspects of the project process. The
plans for projects undertaken within the scope of the IRDP are only based on approved housing
chapters of the Integrated Development Plan (IDP). An IDP is a single, inclusive strategic plan
for the development of a municipality. It incorporates, integrates and organizes plans and take
into account proposals for the development of the municipality (Tissington, 2011). It aligns
resources and capacity of the municipality with the implementation of the plan, complies with
the requirements of the Municipal Systems Act and is compatible with national and provincial
development plans and planning requirements binding on the municipality in terms of
legislation. Hence beneficiaries are able to obtain non-residential stands in the development.
According to Tissington (2011), most provinces and municipalities used the Turkey
Contracting Strategy in the past, which transfers all the development duties to a private sector
contractor, including the administration of beneficiaries. This model encountered numerous
problems and gaps in policy and process of beneficiary registration and allocation. In the
208
present IRDP model, the developer must ensure that identified beneficiaries complete and sign
the application form for the grant of an IRDP Individual Subsidy, which is submitted to the
principal municipal department. The identification of beneficiaries to receive housing
construction subsidies are undertaken before the design and housing construction planning
phase commences, which enable beneficiaries needs assessment to be conducted shifting away
from the one-size-fits-all system of the past. This ultimately will guarantee housing satisfaction
as the variation between what they receive will be in-line to what they expect.
Phase 1 of IRDP involves planning, land acquisition, township establishment and the provision
of serviced residential and other land use issues to ensure a sustainable integrated community.
Phase 2 consists of the house construction phase for qualifying housing subsidy beneficiaries
and the sale of stands to non-qualifying beneficiaries and for commercial interest (DHS, 2009).
During Phase 1’s execution, non-residential stands are allocated, for instance institutional
stands, such as police stations, clinics, public sector use amongst other; business and
commercial stands; serviced stands for use by not-for-profit community service providers, such
as churches, crèche/pre-school/nursery schools, old age homes, etc.; and public uses, such as
parks, recreation areas, informal trading areas, taxi ranks, etc. Conversely, there are rules as to
how these stands are allocated. However, the allocation of the stands and the submission of
application forms for housing subsidies and applications to buy the stands are undertaken and
finalized before the approval of the housing construction project phase; which enables the
houses to be constructed for those needing them, as against the previous patterns where houses
are constructed for the sake of construction. Once the provincial department has received and
approved a subsidy application within three months, it records the name and identity number
of the applicant (and their spouse plus dependents’, if applicable) on the National Housing
Subsidy Data Base to prevent the duplication of an applicant.
At present the subsidy quantum for the IRDP is R55 706 for the construction of the top structure
only. Those earning between R1 501 to R3 500 per month must contribute R2 479 (Table 6.1).
The cost for the provision of internal municipal engineering services is financed from
alternative sources and the use of the housing subsidy allocation for the financing of internal
services, are only approved as an option of last resort. If the latter is the case, a subsidy amount
of R22 162 is available for the preparation of a serviced stand (DHS, 2010). This programme
currently benefit persons, who satisfy the generic housing qualifying criteria, as outlined in the
revised National Housing Code of 2009.
209
Table 6.1: IRDP Housing subsidy quantum amounts for the 2009/2010 financial year
The 2009/2010 housing subsidy guideline amounts
Income category Top structure funding Own contribution Product price
Integrated Residential Development Programme:
R0 to R1 500 R55 706,00 None R55 706,00
R1 501 to R3 500 R53 227,00 R2 479,00 R55 706,00
Indigent: Aged,
disabled and health
stricken
R0 to R3 500
R55 706,00 None R55 706,00
Source: National Department of Human Settlement, 2010 (Subsidy Quantum - Incremental
Interventions)
6.2.2.4.3 Enhanced People’s Housing Process (ePHP)
When the South African state strengthened its role in low-cost housing delivery, a parallel
process was created to increase beneficiary participation in the housing process by getting them
involved in savings and construction through the People’s Housing Process programme. The
People’s Housing Process (PHP) was initially adopted by the Minister of Housing in 1998 to
assist communities in supervising and driving the housing delivery process by building their
homes themselves. However, the notion of community participation, which was enacted on the
White Paper on Housing, reflected in the requirement for a social compact between developers
(government) and communities. According to Charlton & Kihato (2006) regardless of this
provision, community participation had not been clearly defined and its interpretation varied
widely across projects. Charlton & Kihato (2006) further informed that the PHP scheme was
developed partly in response to lobbying by grassroots organisations, such as the South African
Homeless People’s Federation (SAHPF), for greater beneficiary participation and also pressure
from international organisations, such as the UN, which had a track record of beneficiary
participation resulting in more responsive and effective low-cost housing delivery. The PHP
scheme was intended to work with registered NGOs in the housing sector to support
communities in planning and implementing the construction of their own housing settlements
through ‘sweat equity’ to be offset against the NHSS savings requirement. According to
Tissington (2011), this was envisioned to enable poor households overcome the affordability
210
barrier and gain access to a house without the long wait to access housing finance.
The process was holistically supported by a number of South African developmental NGOs
who later formed part of a PHP Reference Group that lobbied for changes to the programme
on the basis that it would achieve more response and effective delivery. The major criticism of
this programme was that it shifted part of the cost of housing onto the poor and that there was
a fundamental dissonance between the collective nature of community-based processes and the
individualized ones (Charlton & Kihato, 2006). Another notable criticism of the PHP was made
by Huchzermeyer (2004) who states that participation in the programme was only limited to
housing construction, with little influence by beneficiaries over key issues, like location of
housing projects and layout around the existing patterns of land occupation. Huchzermeyer
(2004) claims that, organized communities had not been able to identify and manage
infrastructure projects. Yet another criticism is that through the programme, the state abdicates
its responsibility and shifted the burden of delivery to the poor, thus COHRE (2008) claims
that the PHP process has often failed at the local level because of that singular reason.
As a result of the above issues and criticism, the Enhanced People’s Housing Process (ePHP)
was adopted to replace the old PHP programme in July 2008, which was rolled out in April
2009. The new policy programme was the result of lengthy and difficult negotiations between
the PHP Reference Group and the default National Department of Housing, dating back to
2004. The PHP Reference Group had for some time challenged the narrow definition of the
PHP as ‘self-build’ housing involving contributions of ‘sweat equity’, as opposed to the use of
contractors. They were of the notion that the PHP should fundamentally concern a collective,
community-based process of decision-making that would seek to address housing in the context
of other social needs and community priorities (Himlin, 2008). In addressing the social and
community priorities as advocated by the PHP reference group, beneficiaries would ultimately
be actively involved in decision making of housing development issues and at the end, would
be satisfied with their housing situation.
The ePHP adopts a broader definition than the PHP, allowing for greater flexibility and choice
while maintaining the central principles of people-centered development in any participatory
process. In the new programme, the DHS (2009) acknowledged that there are a number of
different approaches to community development that needed to be accommodated with
community involvement in the decision-making processes, community empowerment and the
211
leveraging of additional resources being the determining factors for making it a project. The
enlargement of the scope of the PHP, with a focus on the outcomes of the housing process as
a whole rather than just how the housing product is delivered, informed the development of the
ePHP (Tissington, 2011). The ePHP allows for a process in which beneficiaries actively
participate in decision-making over the housing process and housing product so as to: empower
beneficiaries, create partnerships, mobilize and retain social capital, build housing citizenship,
encourage beneficiaries who are aware of their rights and responsibilities, promote local
economic development, foster stable communities, build houses that are better suited to the
needs of individual households, involve women and youth more directly, and finally, create
sustainable and inclusive human settlements, which are more responsive to the needs of the
community, thus, guaranteeing housing satisfaction.
The ePHP is fundamentally a community-driven process that takes place over a period of time
and is not orientated towards delivery at scale over limited time frames (DHS, 2010). It
encompasses the participation of organized community groupings in the housing process,
through amongst other things the provision of ‘sweat equity’, which replaces their requirement
for a financial contribution. In the ePHP, an approved Community Resource Organisation
(CRO) such as an NG or faith-based organisation is appointed by the provincial government to
provide technical and administrative assistance to the community organisation. Both of these
groups have specific roles and responsibilities, as set out in the ePHP.
According to the DHS (2010), there are four different funding streams for the ePHP, which
include capital funding, capacity-building funding, community contribution/equity funding and
bridging finance. In terms of capital funding, the standard housing subsidy amount for the top
structure applies. The subsidy amount available for the ePHP programme is currently R55 706
(Table 6.2), which covers the top structure but excludes the cost of internal municipal
engineering services (R22,162.00), which is financed from alternative sources. Funding for
infrastructure is provided through the applicable grants if available or as a last resort, accessed
from the province (DHS, 2009). The municipality is responsible for all land packaging and
town planning/township establishment funding (including the undertaking of Environmental
Impact Analysis and Rezoning) and could provide land purchase funding or donate land to
communities. The municipality is also responsible for funding additional facilities and
amenities. In the ePHP, capacity-building funding relates to six aspects of the housing process,
which are: pre-project consumer education funding; project specific capacity-building and
212
facilitation funding; funding for building the physical structure to be used as the Housing
Support Centre (HSC); funding for facilitation and capacity-building for the sector; and
funding for unblocking blocked projects. The ePHP outlines a number of community
contributions/equity that should also be incorporated into an ePHP project, both pre- and during
the project.
According to the DHS (2010), the programme requires skillful technical expertise to assist,
train and guide the house building processes. The realization of quality housing products
remains a fundamental objective of the programme. Also, community contribution is broadly
defined in the process and is not limited to a labour contribution (sweat equity). The programme
may apply in a variety of development circumstances, such as in informal settlement upgrading
projects, rural housing developments and ‘Greenfield’ developments. The benefits of the
programme are only available on a project basis and community members must establish
appropriate community groupings to facilitate representation and decision-making. The main
role players are the Community Based Organisation (CBO) that represents the beneficiaries,
the Community Resource Organisation (CRO) that provide technical and administrative
assistance to the CBO, the municipality and the provincial department responsible for human
settlements. This programme benefits persons who satisfy the generic housing qualifying
criteria as outlined in the revised National Housing Code of 2009.
Table 6.2: ePHP Housing subsidy quantum amounts for the 2009/2010 financial year
The 2009/2010 housing subsidy guideline amounts
Income category Top structure funding Own contribution Product price
Enhanced People’s Housing Programme
R0 to R3 500 R55 706,00 None R55 706,00
Source: National Department of Human Settlement, 2010 (Subsidy Quantum - Incremental
Interventions)
6.2.2.4.4 Social Housing Programme (SHP)
Security of tenure remains one of the fundamental ideologies of the South African housing
policy because of the disadvantaged nature of the past policies against a majority of the citizens.
Where most of the created policy programmes provide freehold tenure to households, there has
213
been an increasing need for affordable rental units, which provide security of tenure to
households who prefer the mobility provided by rental accommodation (DHS, 2010).
The Social Housing Policy for South Africa was approved in June 2005 and the Implementation
Guidelines published in November 2006. According to the DHS (2009), the policy has been
included in the new National Housing Code of 2009, as the Social Housing Programme (SHP).
The Social Housing Act provides the legal framework for the implementation of the Social
Housing Policy. According to the DHS (2010), social housing is defined as follows: a rental or
co-operative housing option for low income persons at a level of scale and built form, which
requires institutionalized management and which is provided by accredited social housing
institutions or in accredited social housing projects in designated restructuring zones. Social
Housing Programme applies only to ‘restructuring zones’, which are identified by
municipalities as areas of economic opportunity and where urban renewal/restructuring
impacts can best be achieved. Designated restructuring zones are geographic areas identified
by local authorities and supported by provincial government for targeted and focused
investment. Restructuring zones were approved (as of 2006) in the following areas: Ekurhuleni
Metropolitan Municipality, City of Cape Town, City of Johannesburg, eThekwini Metropolitan
Municipality (Durban), Nelson Mandela Bay Municipality (Port Elizabeth), City of Tshwane
(Pretoria), Buffalo City Municipality (East London), Mangaung Local Municipality
(Bloemfontein), Msunduzi Local Municipality (Pietermaritzburg), Polokwane, Potchefstroom,
Kimberley and Nelspruit. In the context of the present thesis, the three metropolitan cities with
designated restructuring zones form the basis of the empirical data collection for the study. The
programme is also aimed at developing affordable rental in areas where bulk infrastructure
(sanitation, water, transport) may be under-utilized, therefore improving urban efficiency.
There is a significant capital contribution from government for the development of social
housing in these defined localities as part of a broader goal of social restructuring in South
Africa. Social housing in restructuring zones, according to the guiding principles, take the
form of medium density multi-unit complexes requiring institutionalized management. This
includes townhouses, row housing, multi-storey units, walk-ups amongst others, but they all
come in detached housing units.
The SHP adhere to the general principles laid down in the Housing Act, 1997 (Act 107 of 1997)
Part 1 Section 2, as well as in relevant sections of subsequent legislation such as the Rental
Act, 1999 (Act 50 of 1999). In addition, the policy is read in conjunction with the White Paper
214
on Housing (1994), the Urban Development Framework (1997) and with the National Housing
Code. Thus, it is important to note that the Department of Housing’s ‘Comprehensive Plan for
the Development of Sustainable Human Settlements’ (2004) provides direction in matters of
the principle guiding SHP. The fundamental principles underpinning SHP include (DHS,
2010):
1. promoting urban restructuring through the social, physical, and economic integration
of housing development into existing areas, likely to be urban or inner-city areas;
promote the establishment of well-managed, quality rental housing options for the poor;
2. responding to local housing demand; delivering housing for a range of income groups
(including, inter alia, middle income, emerging middle class, working class and the
poor) in such a way as to allow social integration and financial cross subsidization;
3. supporting the economic development of low income communities in various ways.
This is done by ensuring that projects are located close to job opportunities, markets
and transport, and by stimulating job opportunities to emerging entrepreneurs in the
housing services and construction industries;
4. fostering the creation of quality living environments for low-income persons;
5. promoting a safe, harmonious, and socially responsible environment both internal to
the project and in the immediate urban environment;
6. promoting the creation of sustainable and viable projects;
7. encouraging the involvement of the private sector where possible;
8. facilitating the involvement of residents in the project and/or key stakeholders in the
broader environment through defined meaningful consultation, information sharing,
education, training and skills transfer;
9. ensuring secure tenure for the residents of projects, on the basis of the general
provisions for the relationship between residents and landlords as defined in the
Housing Act, 1997 and the Rental Act, 50 of 1999 - Chapter 3, Section 4 (1) to (5).
10. supporting mutual acceptance of roles and responsibilities of tenants and social
landlords, on the basis of the general provisions for the relationship between residents
and landlords as defined in the Rental Act, 50 of 1999 - Chapter 3, Sections 4 and 5, in
the Co-operatives Act, 1981 (Act 91 of 1981), as well as in the envisaged Social
Housing Act;
11. ensuring transparency, accountability and efficiency in the administration and
management of social housing stock; and
12. promoting the use of public funds in such a manner that stimulates and/or facilitates
215
private sector investment and participation in the social housing sector.
A fundamental constraint to the ability of social housing to assist poor individuals and
households to access affordable and well-located rental housing is that SHIs have tended to
look ‘up-market’ (above the eligibility cut-off point) in order to survive and very little has been
done to increase the range of options available to those in the lower bands of the subsidy range,
also known as the ‘deep-down market’. The Social Housing Policy acknowledges that there is
a perception that social housing is for a small, relatively privileged elite and does little to
contribute to the housing challenge in South Africa (DHS, 2010; Tissington, 2011). While the
SHP aims to tackle these challenges, it admits that the primary policy objective of the SHP is
restructuring, not mass delivery (Charlton & Kihato, 2006). Despite the fact that South Africa
has made great progress in the past seventeen years since the election of its first democratic
government, a number of structural constraints in accomplishing vital changes remain a cause
for concern. Though political constraints have largely been removed, obstacles arising from
the economic structure and spatial patterning of South African society have proven stubborn
and persistent (DHS, 2010). For instance, some Post-Apartheid programmes have even
unconsciously reinforced Apartheid inequities. Thus, there is therefore a need, to ensure that
the links between processes of social restructuring and housing policies and instruments are
brought into closer orientation. Hence, social housing is a useful key instrument in this regard,
and can contribute strongly toward the achievement of urban restructuring and urban renewal
through urban integration and impacting positively on urban economies (DHS, 2010).
Specifically, the involvement of social housing to such restructuring objectives comprises of
three dimensions, which are spatial, economic, and social. While the SHP aims to maximize
‘deep-down reach’ and target those earning R1 500 a month (and less if possible) it aims for
mixed-income projects and requires participants to demonstrate a regular income, which is able
to sustain the monthly rental, and the payment of a deposit equal to rental of three months
(Charlton & Kihato, 2006).
Currently, the SHP has moved away from the earlier individual subsidy-based approach to a
project-based approach. It is predicted that appropriate targeting will be addressed in the project
approval process and that it will be a pre-condition for the awarding of a project grant or
subsidy. Respective projects will specify a range of housing products targeted at income groups
appropriate to the area and context, based on tested demand and in line with the broader
restructuring aims of the Social Housing Policy (DHS, 2010). The DHS states that the
216
difference between rental revenues and the cost of providing the units will be subsidised by
way of a grant from government. This grant will be calculated with reference to the project as
a whole rather than with reference to particular unit types. However, the units meant for the
very poor will attract proportionately more subsidy than units meant for those low-income
groups with more substantial incomes (NDoH, 2004).
The mechanism and funding for the SHP is through a dedicated capital fund at national level,
which the DHS (or SHRA) disburses to accredited local authorities and provinces (who apply
for social housing development in restructuring zones). There are two ways of accessing this
grant. The first is through a standard/fixed component of the social housing grant by way of
the capital restructuring component, which is allocated by the national government. In order to
qualify for the capital grant on every unit, a social housing project must have at least 30 percent
of units contributing to ‘deep down-market reach’ and maximum rentals not higher than R2
500 (implying an income of R7 500 per month, within the top of the income band) according
to Charlton & Kihato (2006) and DHS (2010). A major constraint in the provision of social
housing for the very poor is the on-going management and operations and management
associated costs. The unwillingness of private SHIs to share the risks associated with very low-
income rental housing provision is a major concern and should necessitate a policy rethink by
the DHS with regard to this aspect.
6.3 HOUSING POLICY PROGRESS IN SOUTH AFRICA (1994 – 2010)
In the past seventeen years, there have been several shifts in SA housing policy, which
corroborates the socio-economic significance and political imperative of housing provision in
the country. According to Charlton and Kihato (2006) the housing policy shifts that have
occurred since 1994 were most often responses to flaws in policy execution, or were compelled
by other agendas, such as political pressure or internal departmental politics. Charlton and
Kihato (2006) further emphasized that, housing policy shifts in South Africa are not, explicitly
rooted in a rigorous interrogation of the needs of the poor, such as the impact of housing
programmes on livelihoods and economic activity of the poor beneficiaries. They argued that
much of this had to do with the movement of personalities and senior housing officials out of
the policy and research division of the housing department, and a lack of continuity and
institutional memory to carry the policy development forward strongly and decisively.
217
While the SA housing policy may have been moderately reformed, urban policy, Integrated
Development Plan (IDP) processes and land availability, which are crucial elements for
successful housing provision, have lagged behind. The IDP is a single, inclusive strategic plan
for the development of a municipality. It links, integrates and coordinates plans and takes into
account proposals for the development of the municipality. It aligns resources and capacity of
the municipality with the implementation of the plan, complies with the requirements of the
Municipal Systems Act and is compatible with national and provincial development plans and
planning requirements binding the municipality in terms of legislation (City of Johannesburg,
2011).
Also, the contentious issue of well-located land for housing was never adequately addressed
and almost all housing scholars and practitioners have affirmed that this has to do with the
reluctance of the urban elite to grapple with an issue in which they themselves may hold a
significant stake (Charlton & Kihato, 2006). This section of the chapter provides an overview
of the housing policy development since 1994. It briefly examines the National Housing Forum
(NHF), a process which preceded the development of a National Housing Policy and moves to
an analysis of the guiding national housing policy; the White Paper on Housing; as well as the
Breaking New Ground Policy.
6.3.1 National Housing Forum (1992-1994)
It is useful to briefly recount the debates that dominated the National Housing Forum (NHF),
which was a multi-party non-governmental negotiating forum which met between 1992 and
1994 to discuss the Post-Apartheid housing situation. The formulation of South Africa’s
Housing Policy commenced prior to the democratic elections, with the formation of the
National Housing Forum. The multi-party non-governmental negotiating body comprising 19
members from business, the community, government, development organisations and political
parties outside the government at the time. At these negotiations the foundation for the new
government’s housing policy were developed and agreed. This culminated in the achievement
of the broad housing sector convention also referred to as the Housing Accord that concluded
with the White Paper on Housing in 1994. The Government of National Unity in 1994 made
use of these negotiations and investigations when it formulated South Africa’s National
Housing Policy.
The objective of the NHF was to formulate a consensus around a new non-racial housing
218
policy, which centered around two main debates. First amongst these was whether housing
should be provided by the state or the market; and, secondly, whether the housing standard
should be a completed four-room house or a ‘progressive’ (incremental) house (Tomlinson,
1999). The NHF set the tone for the first democratic national housing policy in 1994. Despite
a holistic number of stakeholders that participated in the NHF, criticisms still emerged about
the dominance of the private sector and big business at the negotiations, and the implications
of this on the final housing policy that emerged. Furthermore, the impact of international
experience affected the outcome, as well as the need for pragmatism around operational
implementation and the heightened sense of urgency of the need to demonstrate delivery
(Charlton & Kihato, 2006).
The NHF debate was centered on who would provide the housing and how it will be provided.
According to Tomlinson (1999) who was an active member of the NHF, the constituencies on
the ‘left’ and the private construction sector argued, for different reasons, that the government
should provide mass rental housing. The ‘left’, also referred to as the Mass Democratic
Movement (consisting of the ANC, COSATU and the civic movement) argued that this would
immediately entail a high standard of provision. The private sector held a similar standard with
the caveat that the private sector should be employed as contractors and not developers so as
to limit their financial risk. However, the opposition to this view was centered on a concern
that the proven financial and organizational burdens of this approach would be too onerous for
a fledgling government (Tomlinson, 1999). Besides, critics argued that local authorities were
keen to rid themselves of the responsibility of managing rental housing because of difficulties
in collecting rent, maintaining stock and applying qualifying criteria to tenants (Tomlinson,
1999). Those in favour of a mass state rental programme were challenged to explain how the
state would finance and manage it; conversely they were unable to persuasively do so and
hence, a more practical approach was pursued relating to the Mass Democratic Movement idea
which at the current time is not sustainable.
In October 1994, the newly-elected government hosted a National Housing Summit in
Botshabelo, where it was able to secure formal support from a broad range of key stakeholders
for the new housing policy and strategy in what is known as the Botshabelo Housing Accord.
The National Housing Accord was signed by a range of stakeholders representing the homeless,
the government, communities and civil society, the financial sector, emerging contractors, the
established construction industry, building material suppliers, employers, developers and the
219
international community. This accord set down the beginning of the common vision that forms
the core of South Africa’s National Housing Policy. Primarily, it comprised an agreement that
all of these stakeholders would work together to achieve the vision summarized in the accord.
As such, the National Housing Accord was soon followed by the Housing White Paper, which
was promulgated in December 1994.
In terms of the debate around the role of the state versus the market in driving housing delivery,
it was decided that the government would provide the context for housing provision and
facilitate delivery, while the private sector would apply for subsidies on behalf of communities,
identify and service land, and construct structures where possible. This approach was heavily
criticized by many who believed that it would not address endemic flaws in the South African
housing market and would simply perpetuate them; thus this was rightly predicted as the
policies initially formulated had been revised to adequately cater for the needs of the larger
disadvantaged members of the society. However, according to Huchzermeyer (2004) the
approach was a consensual one and related to the ‘pacted’ nature of the South African
transition, in which the private sector had a powerful leverage over both the National Party and
incoming ANC governments. The debate about what type of housing would be delivered
concerned the cost of addressing the housing backlog and different estimates of budgets, time-
frames and standards. It was agreed that a once-off capital subsidy scheme would be adopted
to benefit households with an income of less than R3 500 per month and hence the government
launched the NHSS. The subsidy was linked to individual ownership (as opposed to rental),
and households effectively ‘bought’ a housing option with their subsidy (Tomlinson, 1999).
This was to enable housing delivering opportunities and options to as many previously
disadvantaged and deprived South Africans as quickly as possible.
Eventually, the White Paper on Housing - emerged from the NHF process and the Housing
Accord which was influenced by the broad principles and targets of the ANC’s RDP in 1994.
However, the RDP cabinet was disbanded and replaced with the Growth, Employment and
Redistribution; a Macro-Economic Policy Framework (GEAR), which favoured a market-
oriented approach and relied heavily on the private sector for housing delivery. This latter
strategy focused on meeting basic needs and was heavily concerned with delivery (Charlton &
Kihato, 2006).
220
6.3.2 White Paper on Housing (1994)
The ANC government adopted the White Paper on Housing after the historic 1994 democratic
elections, with the aim of creating viable, integrated settlements, where households could
access opportunities, infrastructure and services, within which all South Africa’s people will
have access on a progressive basis to (NDoH - White Paper, 1994): a permanent residential
structure with secure tenure, ensuring privacy and providing adequate protection against the
elements; and potable water, adequate sanitary facilities, including waste disposal and domestic
electricity supply.
The White Paper marked the end of the NHF process. From its inception in 1992, the NHF
played a seminal role in creating the conditions necessary for a national accord in housing,
most visibly evident at the National Housing Summit in Botshabelo on 27 October 1994. Out
of the accord, South Africa earned the task of harnessing the skills, resources and energy that
the nation has in abundance, and directing it to the task at hand - which was to adequately house
all South Africans. The White Paper also marked the beginning of a process. This is because
for the first time in its history, South Africa now had a policy framework that would cater for
all of its citizens. The approach adopted by the White Paper was the search for the creation of
an enabling environment, and not for the publication of a new set of rules. It aimed to contribute
to the certainty required by the market, as well as give the provincial and local governments
their capacity to fulfill their Constitutional obligations. Throughout the document, a partnership
between the various tiers of government, the private sector and the communities was envisaged.
This is seen as a fundamental prerequisite for the sustained delivery of housing at a level
unprecedented in the history of the country as the community can be truly for the citizens, if
they participate in developmental decision that affects them. It also required all parties not only
to argue for their rights, but also to accept their respective responsibilities.
The objective of the policy was to increase the national budget allocation of housing to five
percent and to increase housing delivery on a sustainable basis to a peak level of 338 000 units
each year and to reach the government’s target of one million houses in five years, of which
this target has since been met. The White Paper outlined seven key strategies it would pursue
in order to achieve its objectives, which are:
1. stabilizing the housing environment in order to ensure maximal benefit of state housing
expenditure and mobilizing private sector investment;
221
2. facilitating the establishment or directly establishing a range of institutional, technical
and logistical housing support mechanisms to enable communities to, on a continuous
basis, improve their housing circumstances (supporting the PHP);
3. mobilizing private savings (whether by individuals or collectively) and housing credit
at scale, on a sustainable basis and simultaneously ensuring adequate protection for
consumers;
4. providing subsidy assistance to disadvantaged individuals to assist them in gaining
access to housing (through the NHSS and National Housing Programmes);
5. rationalizing institutional capacities in the housing sector within a sustainable long-term
institutional framework;
6. facilitating the speedy release and servicing of land (utilizing the Development
Facilitation Act and the Housing Development Act); and
7. coordinating and integrating public sector investment and intervention on a
multifunctional basis.
On adoption of the White Paper in 1994, the government’s intention was to deliver a ‘starter
house’ (sometimes consisting of building materials, where the subsidy only covered land and
servicing costs), which beneficiaries would add to and consolidate over time (Charlton &
Kihato, 2006). This incremental approach of achieving the right to housing was related to a key
assumption in the policy that beneficiaries would be able to access loan finance, which would
be spent on improving the house, which never materialized. Charlton and Kihato (2006) further
inform that by the late 1990s, the nature of the houses being delivered shifted from the open-
ended concept of a ‘starter house’ to a minimum 30 square meters housing unit with a defined
specification, which has since been adjusted, but without the consent or any participatory
involvement of the beneficiaries.
However, in 1999 the National Norms and Standards (NNS) for the Construction of Stand
Alone Residential Dwellings were introduced by the Minister of Housing in terms of Section
3(2) (a) of the Housing Act. The NNS placed an increasing focus on the size and quality of the
top structure or house and stipulated minimum standards. As stipulated on the NNS, each house
must have a minimum gross floor area of 40 square meters; two bedrooms; separate bathroom
with a toilet, a shower and hand basin; combined living area and kitchen with wash basin; and
ready board electrical installation, if electricity is available in the project area (DHS, 2009).
However, this norm was never followed in most developments, which has brought about
222
complaints and other social issues with the houses developed to date. Thus beneficiaries’
quality of life is not better enhanced with the provided housing, as the quality of these buildings
are unappealing and as such are even a burden with further problems of having to carry out
repairs and upgrading of the housing units.
In addition, an undesired effect of this new policy, according to Charlton and Kihato (2006),
was that service standards relating to sanitation, water and roads were often dropped in order
to deliver houses in greater numbers and of greater size. Thus, substandard latrines, communal
standpipes and gravel roads were accepted as adequate, reinforcing the trend towards
development on peripheral land as housing projects were built in areas where lower service
levels were more acceptable. This policy adjustment, driven by a political need to deliver
acceptable houses, was not rooted in a deeper understanding of the consequences of the service
levels/location/top-structure trade-off on beneficiaries. Rather, it was an irritable move related
to the historic rejection of the notion of ‘incrementalism’ – the gradual consolidation of a starter
house over time by the end-user – and may again, in fact, have further contributed to the spatial
marginalization of the poor. Nevertheless, the NHSS was used to finance the construction of
over 1.5 million housing units across South Africa between 1994 and 2003. In March 2007, the
NDoH announced that a total of 3 043 900 subsidies had been approved and 2 355 913 houses
built since 1994. While this achievement has been greatly applauded, the government often
notes that the backlog is increasing due to rapid urbanization, amongst other factors. The DHS
currently estimates the housing backlog at around 2.2 million units (DHS, 2012).
6.3.3 Breaking New Ground (2004)
From its inception, the Housing Policy and Strategy of 1994 focused on stabilizing the
environment to transform the extremely fragmented, complex and racially-based financial and
institutional framework inherited from the previous government, whilst simultaneously
establishing new systems to ensure delivery to address the housing backlog (BNG, 2004). The
policy has come with significant achievements, which have been recognized both nationally
and internationally with a significant socio-economic, demographic and policy shifts occurring
within the first ten years of the New South Africa Government. Whilst the government believes
that the fundamentals of the 1994 policy remained relevant and sound, a new plan was required
to redirect and enhance existing mechanisms to move towards more responsive and effective
delivery. This brought about the formulation of a new human settlements plan also referred to
as the Comprehensive Plan for the Development of Sustainable Human Settlements, also
223
known as Breaking New Ground (BNG). BNG reinforces the vision of the then Department of
Housing, to promote the achievement of a non-racial, integrated society through the
development of sustainable human settlements and quality housing.
As of 2002 to 2003, the NDoH carried out a comprehensive appraisal of its housing programme
after identifying a number of ‘unintended consequences’ of the existing programme. These
unexpected problems included peripheral location of residential development; poor quality
products and settlements; the lack of community participation; the limited secondary low
income housing market; corruption and maladministration; a slowdown in delivery; underspent
budgets; limited or decreasing public sector participation; the increasing housing backlog; and
the continued growth of informal settlements (NDoH, 2008). The review process aimed at
providing a new policy direction and to establish a research agenda to inform and support
policy decision-making within the housing programme, particularly to counter the dispersal of
knowledge and intellectual capacity that had occurred over the previous decade (Tissington,
2011).
The BNG comprises nine elements or programmatic interventions, and seven objectives as
shown in Table 6.3. The elements and objectives of BNG are a mix of substance and procedure
and outcomes, outputs and inputs. Nevertheless, BNG can be understood to set out the ends to
be achieved, the means to achieve them and the instruments to be used in the process of
achieving the defined policy goals (Rust, 2006a). With the review, BNG shifted away from a
focus on quantity of houses delivered to quality, with better size and workmanship of housing
product, settlement design, alternative technology, amongst others and a choice on tenure type
and better locations. BNG implementation has increased the rate of delivery of well-located
housing of satisfactory quality through a range of innovative and demand-driven housing
programmes and projects. It also sought to place increased emphasis on the process of housing
delivery, which is the planning, engagement and the long-term sustainability of the housing
environment (NDoH, 2008). Its key objective was to eradicate all informal settlements by 2014
(National Treasury, 2009).
Table 6.3: Breaking New Ground elements and objectives
BNG elements BNG objectives
1 Supporting the entire residential Accelerate the delivery of housing as a key
224
property market strategy for poverty alleviation
2 Moving from housing to sustainable
human settlements
Utilise the provision of housing as a major job
creation strategy
3 Using existing and new housing
instruments
Ensure that property can be accessed by all as
an asset for wealth creation and empowerment
4 Adjusting institutional arrangements
within government
Leverage growth in the economy
5 Building institutions and capacity Combat crime, promote social cohesion and
improve quality of life for the poor
6 Defining financial arrangements Support the functioning of the entire single
residential property market to reduce duality
within the sector, by breaking the barriers
between the first economy residential property
boom and the second economic slump
7 Creating jobs and housing Utilize housing as an instrument for the
development of sustainable human
settlements, in support of spatial restructuring.
8 Building information,
communication and awareness
9 Establishing systems for monitoring
and evaluation
Source: Adopted from Rust, 2006a.
The BNG policy recognised the change in the dynamics of the housing demand, the increasing
average annual population growth, the drop in average household size, significant regional
differences, increasing urbanization, skewed growth of the residential property market, growth
in unemployment and a growing housing backlog despite substantial delivery in the first decade
of the initiated housing policy. Also, it recognised that the lack of affordable, well-located land
for low-cost housing had led to development on the periphery of existing urban areas, achieving
limited integration. Thus, NDoH (2008) informs that the dominant production of single houses
on single plots in distant locations with initially weak socio-economic infrastructure is
inflexible to local dynamics and changes in demand. Hence, the new human settlements plan
moves away from the current commoditized focus of housing delivery towards more
responsive mechanisms, which addresses the multi-dimensional needs of sustainable human
settlements.
In addition, the BNG recognised that the initially constructed subsidised houses had not
become the valuable assets envisioned in earlier policy. Besides, beneficiaries’ inability to pay
for municipal services and taxes revealed that municipalities viewed such housing projects as
liabilities, and were not particularly responsive to the national department’s more progressive
225
intentions around housing. The BNG document frames housing delivery more clearly as a
catalyst for achieving a set of broad socio-economic goals. Hence, the key intention of BNG
was to move away from a supply-centered model to a demand-centered model driven by the
needs of the beneficiaries, which the present thesis has considered a tangible way to guarantee
a lasting satisfaction with the produced houses by the beneficiaries.
Furthermore, the BNG introduces an expanded role for municipalities. That is, in shifting away
from a supply-driven framework towards a more demand-driven process, it places an increased
emphasis on the role of the State in determining the location and nature of housing, as part of
a plan to link the demand for, and supply of housing. In so doing the problems of placing
housing in urban boundary will be done away with - however, the BNG has not succeeded in
doing this since the endorsement. Thus, this approach has only enabled municipalities to
assume overall responsibility for housing programmes in their areas of jurisdiction, through a
greater devolution of responsibility and resources to them. However, the BNG defines four
primary ends (BNG, 2004; Rust, 2006a), which are the basis of its acceptance:
1. Sustainable human settlements: well-managed entities in which economic growth and
social development are in balance with the carrying capacity of the natural systems on
which they depend for their existence and result in sustainable development, wealth
creation, poverty alleviation and equity.
2. Integration: The shift from housing units, to sustainable human settlements in BNG
largely captures the integration end. Spatial restructuring is also critical and sustainable
human settlements are seen to support spatial restructuring. BNG utilizes housing as an
instrument for the development of sustainable human settlements, in support of spatial
restructuring. There is also an institutional dimension as integration is both intra-
governmental (within a sphere of government) and inter-governmental, requiring
integrated planning and coordinated investment.
3. Housing assets: ensuring property can be accessed by all as an asset for wealth creation
and empowerment and supporting the functioning of the entire residential property
market to reduce duality.
4. Upgraded informal settlements: progressive eradication of informal settlements and
urban inclusion.
The means to achieve the above mentioned primary responsibilities, as defined on the policy
226
includes:
1. Municipal accreditation: to reduce transaction costs and unnecessary administration,
funds will flow directly from national government to accredited municipalities.
2. Effective inter-governmental relations: coordination and alignment is essential to
ensure the effective and efficient flow of resources. This will be achieved through the
enhanced planning framework; bilateral co-operation between the Department of
Housing, the Social Cluster Partner Departments and the other spheres of government,
particularly municipalities; the DoH and the metros working together to achieve the
required alignment.
3. Delivery defined by demand: demand responsiveness cuts across many BNG proposals
and the notion of demand definition underpins the rationale for an expanded role for
municipalities.
4. Effectively functioning housing markets: the BNG develops a strategy around
supporting the entire residential property market, which includes: assisting lower-
middle income groups (expanding the scope upwards); a more flexible approach to
accommodate demand responsiveness and shift from product uniformity; enhancing the
role of the private sector; and creating linkages between the primary and secondary
residential property market.
Already, the cities of Johannesburg, Ekurhuleni, eThekweni, Tshwane and other metropolitan
cities have all redraft their housing strategies in line with BNG. While their policies are in line
with the budgetary allocations and inherent conditionality’s defined at national level for BNG;
the local implementations are thus, giving expression to the intentions of BNG.
6.4 HOUSING DELIVERY AND BACKLOGS
In 2009, the DHS admitted that the data it relies on to estimate the housing backlog in South
Africa is most likely unpredictable, and that a state-subsidised house that is delivered may
remain inadequate because of lack of access to basic services. Consequently, in terms of
eliminating the housing backlog and delivering adequate housing to the low-income groups
especially, the department is not really clear where it stands (Tissington, 2011). Hence, the
DHS has further indicated that in relation to its statistics collection and verification, they are
hoping that the 2011 Census will be able to give the department a better sense of the accuracy
227
of the data they will need to measure backlogs and access delivery. However, there are various
reasons for unreliability of data such as poor provincial and municipal record-keeping in many
parts of the country, as well as incomplete data on house construction amongst others. The
South Africa National Treasury informs that detailed records of spending on subsidy
instruments per municipality are not readily available on a national basis; and that these data
weaknesses are a problem for the sector and it reflects an on-going coordination problems being
experienced by the municipalities. Thus, this severely undermines any detailed analysis,
oversight and accountability. As such, this section of the thesis provides review on housing
delivery since 1994, as well as on the current and growing housing backlogs.
6.4.1 Housing Delivery since 1994
The National Department of Housing states that from 1994 to 2004, about R29.5 billion was
spent on state-assisted housing investment generated 1.6 million housing opportunities and
provided 500 000 families with the opportunity to secure titles of old public housing stock
(NDoH, 2004). Also, NDoH (2007) informs that the government had made good progress in
eradicating backlogs and providing adequate housing. It further reported that over 3 million
subsidies had been approved, benefiting over 13.5 million poor people with housing.
Cumulatively, the government stated it had spent R40 billion on housing developments since
the inception of the housing programme, contributing to 2.4 million houses being constructed
and those still in the process as of 2007. However, while approximately 3.3 million subsidies
were approved as of 2009, actual delivery of subsidised housing units had been much slower
(National Treasury, 2009). However, between 2001/2002 and 2007/2008 delivery was said to
have declined in most provinces, while the allocation from the Integrated Housing and Human
Settlement Development (IHHSD) grant increased (National Treasury, 2009). The National
Treasury further informed that between 1995 and mid-2008 the IHHSD grant disbursed
approximately R49 billion, which provided a total of 2.6 million housing opportunities at a
gross average cost of R18 850 per unit, and an average annual delivery rate of 200 000 units a
year. According to the then Department of Housing, although accurate data is not readily
available, the bulk of this spending occurs through project-linked subsidies (now IRDP), where
developers implemented housing projects at scale and a qualifying household obtains
ownership of a complete residential unit.
In 2010, the Minister of Human Settlements stated that since 1994 more than 3.0-million
housing units have been made available for nearly 13.5 million people. Hence, housing scholars
228
argued that there are three major problems with reaching an aggregate figure for state-
subsidised housing. First, they informed that records in the deeds office do not indicate whether
the house was constructed with or without a state subsidy, whilst data on the approval of
housing subsidies is incomplete and difficult to match with actual house construction.
Secondly, it seems that a substantial proportion of state-subsidised RDP and BNG houses,
perhaps as high as 50 percent have not been registered with the deeds office. Lastly, it seems
that state subsidies have been used in some cases to finance transfers of ownership from the
state to occupants, for example leasehold being converted to freehold, which means a house
was not actually constructed. Rust (2006a) informs there are probably less than one million
registered RDP and BNG housing units, and that the figure of 2.8 million can only be reached
if there are as many unregistered RDP and BNG housing units and about the same number of
properties where the ownership has been transferred from state to occupant. As such, if this
was the case there would still only be about two million new houses.
Table 6.4 shows the preliminary units delivered in 2009/10 and an estimated delivery till 2014.
It is envisaged that if delivery occurs at this pace - on average 230 000 units per year (which is
unlikely) - it would still only mean that by 2014 approximately 1.1 million housing units will
be delivered. This is over one million units short of the current and growing backlog of 2.1
million households, which is thought to be a conservative estimate (Tissington, 2011).
Table 6.4: Estimated Housing Delivery from 2008 to 2014 (DHS)
Estimated Delivery
Preliminary units
delivered in
2009/10
2010/11 2011/12 2012/13 2013/14
Eastern Cape 28 644 23 400 23 400 24 463 26 058
Free State 18 829 21 462 21 462 22 438 23 058
Gauteng 39 922 48 553 48 553 50 760 54 071
Kwazulu-
Natal
27 376 26 626 26 626 27 837 29 652
Limpopo 23 079 22 613 22 613 23 641 25 182
Mpumalanga 8 291 8 181 8 181 8 553 9 111
Northern
Cape
6 257 6 512 6 512 6 808 7 253
229
North West 35 141 30 954 30 954 32 361 43 472
Western
Cape
32 371 31 698 31 698 33 139 35 300
Total 219 899 220 000 220 000 230 000 245 000
Source: National Department of Human Settlement, 2010.
For instance, in 2009, South Africa’s national budget for housing and community amenities
was R73.2 billion (8.7 percent of the national budget) of which about R13 billion was
distributed to provinces to administer housing subsidies to the low-income group. Likewise, in
2010, more than R93 billion was allocated to housing and community amenities, representing
a nominal increase of more than 14 percent on the previous year. From the approved allocation,
more than R20 billion was set aside for housing development purposes only. Hence, the
Minister of Finance stated in his National Budget Speech that the human settlements grant is
one of the faster growing items on the budget. Conversely, as has been specified above,
although billions of rands are being allocated to housing development, and countless number
of subsidies are being approved for poor and low-income households, in contradiction, the
delivery of housing units is not occurring at scale or at pace with the monetary allocation.
6.5 HOUSING DELIVERY IN SOUTH AFRICA
The delivery of housing in Post-Apartheid South Africa is characterised by three streams of
provision and allocation that are running parallel and often overlapping to some extent. These
three streams are public sector built; private sector provided and self-provided housing.
However, the public sector built is only delivered through the housing subsidy schemes, thus
the name state subsidised housing. In this section of the thesis, the concept of state subsidies
housing in South Africa will be discussed, with a particular emphasis on the delivery
mechanism and the problem associated with the housing delivered to date.
6.5.1 State Subsidised Housing in South Africa
State subsidised housing in South Africa is a form of housing delivery system in which the
property and associated infrastructure is financed by the government and transferred to either
a group of qualified low-income families, elderly and handicapped individuals with little or no
contribution coming from them. However, it has never proved easy to help the poor through
housing subsidies, particularly in developing countries (Gilbert, 2004) such as South Africa
with numerous social-economic issues and racial divides. Today, very few governments are
230
prepared to offer housing subsidies to the poor unless they are delivered as up-front or as,
targeted capital subsidies. Also, the lack of resources has forced most governments into making
difficult decisions about the size and the number of subsidies to be offered.
In order to address the housing shortage and the urban and rural housing backlog in the Post-
Apartheid South Africa State, the government instituted a number of programmes and
mechanism to assist lower income households. Foremost amongst these include the housing
subsidy system, as well as other innovative mechanisms to encourage the increase of affordable
housing to the poor (Landman & Napier, 2010). Also dependent on these decisions, has come
a series of implementation problems relating to the quality of construction, the location of the
new housing solutions, the use of credit and how to allocate subsidies between so many
beneficiaries. While there have certainly been positive experiences from the South Africa
housing delivery mechanism, there is also a very long list of failures particularly with the
process of delivery and the product that was and is still being delivered. However, housing
delivery for the low income groups in South Africa is reliant on the Housing Subsidy process.
The subsidy scheme facilitates the provision of a range of housing types. Prominent amongst
this is the RDP housing (named after the Reconstruction and Development Programme initiated
to promote delivery in 1994) which was developed by government and allocated to
beneficiaries with a household income of less than R3, 500. Beneficiaries of the housing
subsidy scheme receive a once off grant. The Housing Subsidy grant is a grant by government
to qualifying beneficiaries for housing purposes. The qualifying criteria have been discussed
above. The grant is not paid in cash to beneficiaries; it is either paid to a developer, or in new
housing developments for the provision of a house. The grant is used to construct a house (top
structure as approved) that complies with the minimum technical and environmental norms and
standards, land for the house, and basic services (water and sanitation), which is then
transferred to the qualifying beneficiary after completion.
Despite the impressive delivery of houses, the housing backlog has increased from 1.5-million
in 1994 and has continued to grow and is currently estimated at between 2.1-million and 2.5-
million. This, according to the DHS translates into approximately 12.5 million people still
needing houses in the country. South Africa has spent billions of rands to provide housing to
the poor, but the government’s 2009 General Household Survey showed that the portion of
households still living in shacks has remained steady at about 13 percent since 2002. Also,
despite the housing budget which was increased in 2011 by 38 percent of 22.5 billion rands
231
($3.3 billion, 2.3 billion euros) the authorities still regularly battle protests in shantytowns by
destitute black residents angered by rampant joblessness and poor amenities like water, toilets
and electricity. However, some housing scholars argues that there may be many more people
still needing housing, as the number of those living in informal settlement are not adequately
captured and the number most times do not include foreign nationals. However, the housing
subsidy scheme is in line with the government housing strategy as contained in the National
Housing Policy Framework, which is to provide subsidy assistance to the low-income groups,
thus enabling them to become home owners and to improve their quality of life.
According to Gardner (2004), subsidy housing for homeownership comprise 15.2% of the total
stock in South Africa, with social housing (rental) constitutes 0.5% and co-operatives 1.6%
(based on the 2001 Census). Though delivery through social housing has significantly increased
since then, but still constitutes a small percentage of the overall housing supply in the country.
Other forms of state provided housing, include public housing at a local level. According to
Gardner (2004), ex-council housing (now owned) comprised 7.6% of the total stock in South
Africa, with current stated owned housing available to rent, also constituting 7.6%. However,
Statistics SA (2009) informs that around 18.9 percent of South African households live in
‘RDP’ or government-subsidised homes according to its General Household Survey report and
another 13 percent are waiting for a state subsidised dwelling. A similar percentage of
households (13.5%) had at least one household member on a demand database/waiting list for
state-subsidised housing. Based on these statistics, a study of the current nature is imperative
in order for government to provide houses that will meet the need of the low-income who are
allocated these houses.
Government-subsidised homes or ‘RDP’ housing, usually includes a stand-alone house of
about 30-40 m2 on a 250 m2 plot. However, lately, developers and designers have started to
experiment with alternative housing types where RDP units are semi-detached and located on
smaller sites to accommodate densification. These are evident in projects, such as the
Alexandra Urban Renewal Programme and the Pennyville Development in the City of
Johannesburg (Landman & Napier, 2010). For instance in Pennyville, the developers have also
experimented with semi-detached RDP units on two levels such as the semi-detached
simplexes, while in another project in the Nelson Mandela Metropolitan Municipality in the
Eastern Cape Province, namely Sakhasonke, semi-detached duplexes have also been built.
According to Landman and Napier (2010), these newly experimented houses are significantly
232
larger (46 square meters) than the previous models of 40 square meters. This shows some
flexibility being adopted regarding the implementation of subsidy housing for ownership in
South Africa. Currently, there is very little supply of housing for households earning between
R3, 500 and R8, 000. These households do not qualify for a housing subsidy as recommended
by the Housing Policy Framework, yet are unable to afford housing in the market. This gap in
the housing market is known as Gap Housing (Rust, 2006b). In attempting to address part of
this gap, housing institutions are also making use of institutional subsidies to co-fund the
development of social housing units for the more affordable housing market sector; thus
addressing the gap market. This current initiative is supposed to sufficiently address the Gap
Housing as government has committed resources to it. For example, the Brickfields and Carr
Gardens projects developed in the inner city of Johannesburg and the Amalinda Housing
Project in the Buffalo City Municipality are plot projects of this initiative. Cross (2008) informs
that the majority of government-delivered dwelling units are concentrated in the urban sector,
whilst the rural sector is not being given a considerable attention.
Despite the efforts of the NHSS to deliver housing to all, there have been problems with both
the quantity and quality of housing delivered since 1994. Prime amongst these, as noted by the
NDoH more recently, indicates that housing delivery has had a limited impact on poverty
alleviation and houses have not become the financial, social and economic assets as envisioned
in the early 1990s and as stated in the Housing White Paper. This was supported in a study by
Aigbavboa (2010) study on the housing subsidy post-occupancy evaluation, which found that
a majority of the beneficiaries do not consider their houses an asset for wealth creation. This is
because most of the previously built houses via the project-linked subsidies for large-scale
housing developments (now IRDP) were often located on the periphery of existing townships;
land previously acquired or zoned for township development under apartheid (Charlton &
Kihato, 2006). This system maintains the marginalization of the poor and does not contribute
to the compaction, integration and restructuring of the Apartheid City (Charlton & Kihato,
2006). This trend has thus reinforced the spatial segregation of cities, which isolates the poor
from livelihood opportunities and social services, as well as the tendency towards urban sprawl.
This problem has often been exacerbated by the fact that there has also been little co-ordination
between government departments to ensure that public transport, schools, clinics, libraries and
police stations are provided for the new community. However, with the modification and
enhancement of the NHP and with particular emphasis on the SHP design to bring about
restructuring of the urban divide, this has greatly been attended to through the IRDP scheme.
233
Because of the numerous problems associated with the RDP houses over the years, Tissington
(2011) asserts that they have become residential dormitories, thus many beneficiaries choose
to trade their houses and move back to informal settlements or other informal housing to be
closer to work. Also, Urban LandMark (2010) states that, since 2005, approximately 11 percent
of all RDP houses were unofficially traded by owners who were barred from selling their
houses due to the mandatory lock-in period of eight years set for all state subsidised housing
in the country. Wessels (2010) informs that over half of the unofficially traded transactions
were between R5 750 and R17 000. Hence, Rust (2006) states that eight years is a long time,
as this duration makes it incredibly difficult for a household RDP to be used as a financial asset
and get lending against it. Also, Tissington, Rust, Mcgaffin, Napier, Charlton (2010) found that
it is difficult for households to use their RDP houses as financial assets and get lending against
them as the location of the constructed houses often fails to match beneficiaries’ needs, thus
financial institutions are reluctant to lend using the houses as collateral security. As a result,
the low-income groups accept the RDP houses but rent them out to generate income, while
they choosing to live in an informal settlement or backyard shack in a township to be closer to
jobs and livelihood opportunities. Also, Sexwale (2010) mentioned that only about 34.0% of
the beneficiaries that were originally allocated state-subsidised are occupying the housing
units.
Another problem that concerns the RDP houses according to Wessels (2010), is the delays in
the transferal of title deeds and the protracted length of township establishment have meant
many beneficiaries do not have their title deeds or proof of ownership. Also of late, the Minister
of Human Settlements announced that the government would be using R1.3 billion, or tem
percent of the department’s budget, to demolish and rectify badly constructed RDP houses
(Prinsloo, 2010). The Minister of DHS since coming to office in 2009 has focused greatly on
the issue of the quality of RDP houses previously built and those being constructed, as well as
corruption in housing projects through his national audit task team (Naidu & Isaacson, 2009).
There have been several problems with this model including corruption in the allocation of
subsidised housing units as well as in construction tenders, the latter resulting in construction
short cuts being taken and poor quality houses being built.
Statistics South Africa, through its 2009 General Household Survey (GHS) across the country
states that about 16.1% of households living in RDP or state-subsidised dwellings felt that the
234
walls of their dwellings were weak/very weak, whilst 14.9 percent felt that their roof was
weak/very weak. Furthermore, the GHS report informs that there was a considerable variation
between provinces in perceptions about housing quality. For instance, in the Western Cape and
Eastern Cape, nearly a third of all households informed that they have a problem with their
walls and roofs. While in the Northern Cape, 17.0% of households had problems with their
walls and 18.0% had problems with their roofs. Also in KwaZulu-Natal, 14.9% of households
had problems with their walls. Hence the DHS claims that approximately R359 million would
be needed to demolish and rectify approximately 20 000 shoddy RDP houses in the Eastern
Cape. However, towards the end of 2009, the DHS stated that nearly 3 000 RDP houses
identified in the Eastern Cape and KwaZulu-Natal provinces will be demolished because of
inferior workmanship. Mbanjwa (2009) researched that these houses had been built in the 18
months prior to their identification as intolerably poor structures.
Furthermore, in delivering state subsidised RDP housing, the government has overlooked a key
principle of the RDP policy, which is to promote integrated development. However, this is
currently being addressed through the IRDP scheme, which is aimed at creating social cohesion
in the developed settlements.
6.6 LESSONS LEARNT FROM SOUTH AFRICA HOUSING STUDIES
The following are the lessons learnt on the study of the South Africa housing situation:
the housing environment in South Africa is complex, in large part due to the deliberate
policy and legislative framework of socio-economic and spatial exclusion and
marginalization created during the Apartheid Era;
the Post-Apartheid State Governance has been actively involved in trying to create a
level play field for the previously disadvantaged and also trying to repair the
disadvantaged condition created by the almost 42 years of the Apartheid Government;
the implementation of the South African housing policy to date has been skewed and
unable to address the land, housing and basic services needs of millions of poor South
Africans who still lack adequate housing and access to water, sanitation and electricity,
also, the SA housing policy is the result of a mixed bag of international influences and
local creativity - mostly due to the policy of spatial segregation in the Apartheid State,
which contributed to a policy which is defined in terms of ‘scan globally, reinvent
locally’ principle;
235
the South African housing agenda objectives go beyond simply providing housing
structures, but that housing construction should contribute to the national economy,
demonstrating delivery, contributing to the economic performance and assisting with
poverty alleviation;
that the enactment of the Native Land Act 27 of 1913, cemented housing policy issues
in the Apartheid Era, which created the divide in housing issues, which exist to date,
thus, ascertaining that the housing backlog and problem of the present date back to the
1913 Native Land Act;
the first effort to adequately assist local authorities in their task of meeting the housing
needs in their areas of jurisdiction was through the adoption of the first South African
Housing Code in 1964;
the fundamental principles of the South Africa housing policy and developmental
framework is established in the Housing White Paper, which was published in
December 1994;
the fundamental policy and development principles introduced by the Housing White
Paper remains relevant and guide all developments in respect of housing policy and
implementation in South Africa;
the fundamental philosophy underpinning housing development goals in South Africa
is the existing South African Constitution of 1996;
the Constitution contains justifiable socio-economic rights and enshrines everyone’s
right to have access to adequate housing;
the South Africa Housing Act of 1997 is the primary piece of housing legislation in
South Africa. It legally entrenched housing policy principles outlined in the 1994 White
Paper on Housing;
the Housing Act provides for a sustainable housing development process, laying down
general principles for housing development in all spheres of government; it defines the
functions of national, provincial and local governments in respect of housing
development; and it lays the basis for financing national housing programmes;
the National Housing Code (NHC), which was first published in 2000 in accordance
with the Housing Act, set out the underlying policy principles, guidelines and norms
and standards which apply to the National Housing Programmes;
housing policy shifts in South Africa are not, however, explicitly rooted in a rigorous
interrogation of the needs of the poor, such as the impact of housing programmes on
236
livelihoods and economic activity of the poor beneficiaries;
there is a lack of continuity and institutional memory to carry the policy development
forward strongly and decisively, as housing department personnel’s are changed from
time to time; and
housing delivery for the low income groups in South Africa is reliant on the Housing
Subsidy process. The subsidy scheme facilitates the provision of a range of housing
types. Prominent amongst this is the RDP housing which is developed by government
and allocated to beneficiaries with a household income of less than R3, 500.
6.7 CONCLUSION
This chapter has provided an outline of housing legislative and policy framework in South
Africa; examining the Constitution with specific reference to the Bill of Right and the Right to
Housing; National Housing Code; and the National Housing Programmes categorized therein
with a specific focus on State subsidized housing (housing subsidy Scheme). Also, an evolution
of housing policy in South Africa was discussed. Further, an overview of the developments in
housing policy since 1994 was illustrated, including a summary of the negotiations at the
National Housing Forum held between 1992 and 1994. The section further examined the
supreme policy framework contained in the 1994 White Paper on Housing, and the problems
associated with the Reconstruction and Development Programme (RDP) houses built after
1994.
In many ways, despite the ‘miracle’ of the negotiated settlement, which established a fully
democratic order without the liberation of a full-scale revolution, the new South African state
remains a hostage of its past, as evident from the history of its housing policy down to the
development and amendments made to the policies developed after 1994. Precisely, the legacy
of Apartheid in the form of profound social polarization, extreme economic inequalities and
spatially divided cities massively complicates the task of building a new society in which, as
informed in the Constitutional Right that all South Africans should have access to adequate
housing. Though, Apartheid alone cannot be held responsible for the housing conditions in
South Africa but equally no account of housing policy and conditions can be credible if it does
not take into account the history of South Africa.
From the reviewed literature, housing policy in South Africa appears to remain wedded to the
237
sort of supply side approach which, for the most part, can deliver only a fairly standardised
product or range of products. The degree to which the present housing policy and the supporting
programmes will overcome certain of the major problems inherited from the past even if the
obstacles to more rapid delivery at scale can be overcome, remains questionable. In particular,
the current policy framework has clear limits in terms of addressing the fundamental social and
spatial divisions, which characterize the country’s cities. Also, from the literature, the failure
to clearly and steadily integrate housing policy into a coherent strategy of urban restructuring
in South Africa was the former housing framework’s, deficiency, which has now been
addressed through the social housing programme framework which applies only to
‘restructuring zones’, which are identified by municipalities as areas of economic opportunity
and where urban renewal/restructuring impacts can best be achieved.
Despite the numerous challenges that have come with the delivery of housing in the new South
African State, over three million subsidies had been approved, benefiting over 13.5 million
poor people with housing. Cumulatively, the government stated it had spent R40 billion on
housing developments since the inception of the housing programme, contributing to 3.0
million houses being constructed. However, while approximately 3.3 million subsidies were
approved as of 2009, actual delivery of subsidised housing units had been much slower. The
houses delivered have housed 18.9% South African households according to the General
Household Survey report and another 13.0% are waiting for a state subsidised dwelling. The
most important point to stress in conclusion is that the vital objective of housing policy
framework must be to stimulate the environments which give dignity to people’s lives: it is
not simply the provision of shelter. Against this criterion, the record of housing policy and
implementation in South Africa in recent decades has been really poor. Thousands of millions
of rands have been spent on housing but the environments which have resulted are almost
unfailingly sterile, monotonous, hostile and inconvenient. In order for the money spent not to
be wasted, there is a need to evaluate the residential satisfaction of the housing occupants so
that errors made can be corrected for future development. A narrow focus on the individual
housing unit and the provision of shelter, which is the prevalent disposition, gives rise to a
particular mind-set and approach which ensures the generation of poorly-performing, sterile
environments. Significant improvement demands a paradigm shift, a shift which places not the
individual unit but collective spaces, institutions and facilities at the centre of housing research,
which is beneficiary driven. The next chapter will focus on a discussion of the methodology
adopted in order to realize the research objectives.
238
CHAPTER SEVEN
RESEARCH METHODOLOGY
7.1 INTRODUCTION
This chapter discusses the various steps undertaken by the researcher to explore the objectives
of this research. As discussed in the introduction, the aim of this research is to investigate and
organize the relationship between publicly-funded beneficiaries’ dwelling units,
neighbourhood features, services provided by government, building quality, beneficiaries’
participation, needs and expectations, into a model. Beneficiaries’ participation, needs and
expectations, are new constructs that are peculiar to the present model to be validated as they
have not been previously considered in the existing models of residential satisfaction; whilst
other variables have been measured in a majority of the previous studies. This chapter provides
details about the methodological research framework for the current study. The chapter consists
of the following sections: research design and methodology, the quantitative study and the
qualitative study. The research design and methodology section focuse on the research
procedures, including the choice of research methods and the selection of participants. This
investigation combined quantitative and qualitative methods (Mixed Method): the Delphi
Study and a Structured Questionnaire Survey. The use of a Mixed Method approach is rooted
in both philosophical and practical reasons, which are explained in detail to justify the Mixed
Method approach for this thesis.
7.2 QUANTITATIVE VERSUS QUALITATIVE RESEARCH
METHODOLOGY
This particular section of chapter 7 explores the methodological options available to undertake
research. When deciding on the appropriate research methodology, researchers are usually
influenced by the research aims, as well as the type of data that they have to collect. Hence
researchers have to choose between different options of methodology which fall into two broad
categories namely, quantitative and qualitative. Some researchers may decide to use either one
of the methodologies or a combination of the methodologies to carry out the research, provided
that they are appropriate in answering the (researchers) research questions.
239
However, despite the influence of the research aim, these methodologies are sometimes
influenced by the research paradigms (Jean, 1992). A paradigm is a “set of beliefs that
researchers use to make sense of the world or a segment of the world” (Crotty, 1998:35). In
other words a paradigm provides insight into the way, in which researchers look at and perceive
the world (Kuhn, 1996). Paradigms guide the conceptual framework that researchers use in
seeking to understand and make sense of reality (Maguire, 1987). Paradigms set boundaries for
researchers in terms of the manner in which they can execute the research process, with regards
to research methods, strategies for inquiry, as well as the purpose and use of knowledge (Crotty,
1998). Thus, paradigms influence what researchers regard as accepted knowledge and ways of
doing research (Crotty, 1998) and shapes researchers’ “…perceptions and practices within
their research disciplines” (Maguire, 1987:11).
Similarly, the choice of method is typically influenced by major philosophical considerations
(ontological and epistemological) underlying the research process (discussed in the next
section). Both quantitative and qualitative research methodologies are based on the
epistemological assumptions regarding the nature of knowledge and the methods of abstracting
that knowledge, as well as ontological assumptions, which relate to the nature of reality or the
phenomena being investigated (Jean, 1992). These philosophical considerations which
influenced the choice of the research approach for this thesis are discussed briefly in the next
sections.
7.3 PHILOSOPHICAL CONSIDERATIONS IN RESEARCH
METHODOLOGY
As stated above in Section 7.1, the choice of research methodology is usually influenced by a
set of assumptions underlying each research methodology (Crotty, 1998). According to Crotty
(1998), the choice of a method has to be supported by the statement of assumptions that have
been brought into the research process and are reflected in the methodology. These assumptions
though varied, tend to fall into the philosophical areas of ontology and epistemology. A brief
discussion of these considerations follows in Section 7.3.1 and 7.3.2.
7.3.1 Ontological Consideration
Ontological assumptions revolve around the question of ‘what is’ with the nature of reality
(Crotty, 1998). In other words, it is an effort to elucidate what reality is and why things happen
240
the way they do. In an attempt to explain reality, Jean (1992) advocates two opposite
assumptions of reality, which are objectivity and subjectivity. Jean (1992) viewed the
objectivist stance as, reality existing out there, intact and tangible, but it is independent of
individuals’ appreciation and cognition (Crotty, 1998; Jean, 1992). Thus, regardless of whether
or not individuals perceive and attach meaning to this reality, it remains unchanged (Burrell &
Morgan, 1994). Hence an individual is thus; “...born into and lives within the social world that
has its own reality, which cannot be created by that individual” (Burrell & Morgan, 1994:4).
Thus, in order to create a better understanding of reality, objectivist’s researchers propose the
need to study the causal relationships among the elements constituting reality (Burrell &
Morgan, 1994; Jean, 1992) which is advanced in the current research.
Additionally, objectivist view of reality is closely related to a theoretical position called
positivism (Crotty, 1998). Positivism holds the objectivist assumption that reality is
independent of human cognition (Guba, 1990). Positivists postulate that the world exists as a
system of observable variables waiting to be discovered (Maguire, 1987). Similarly, positivists
believe that the use of scientific methods of inquiry can assist in discovering the true meaning
of reality (Crotty, 1998; Guba, 1990; Maguire, 1987). In this regards, scientific methods are
those research methods that lack human involvement in arriving at the meaning of reality. The
aim is to avoid the researchers’ bias in the research process and produce scientifically verified
knowledge (Guba, 1990). The results of such investigation generate rules and theories that help
to explain and sometimes provide a guide for understanding social behaviour (Maguire, 1987).
This is exactly what the currently research aims to achieve, in that through the model to be
developed, an understanding will be created of the social construct that brings about residential
satisfaction of occupants living in low-income housing in South Africa.
Though, objectivism has been criticized for its inflexible assumption of an independent reality
outside human cognition (Guba, 1990). According to Maguire (1987), the supposition by
objectivists that reality exists outside human conception is inconsistent because reality is
humanly and socially created. Besides human beings are not passive spectators but rather they
participate actively in the construction of meaning. Thus, Maguire (1987:19) argues that
“objectivity is illusion because it suggests that it is possible to separate the subject of
knowledge, the knower, from the object, the known”.
241
Furthermore, the opposite view to objectivism is subjectivism or constructivism. It assumes
that the world consists of labels, names and concepts that are used to create the meaning of
reality (Burrell and Morgan, 1994). According to the subjectivist interpretation, reality is not
discovered but it is constructed by human beings as they engage with the world they live in
(Crotty, 1998). In that way understanding and interpretation of reality occurs when human
beings interact with their environment and others and assign meaning to the world around them
(Crotty, 1998). Thus, in research, meaning is “…an expression of the manner in which the
researcher as a human being has arbitrarily imposed a personal frame of reference on the
world…” (Jean, 1992:89). The next section is an extension of the discussion on philosophical
suppositions that influence researchers’ choice of methodology. Having discussed the question
of ‘what is’ reality, the next section looks into ‘how’ reality or knowledge is created.
7.3.2 Epistemology
Epistemology is associated with an explanation on the nature of knowledge in terms of how
knowledge is created (Hill, 1995). In research, epistemology provides the grounds for deciding
on the kind of knowledge that is considered appropriate, adequate and legitimate for the
research at hand (Crotty, 1998). Furthermore, Hill (1995) submits that research methodology
is applied epistemology, and therefore, methodology has to be supported by an epistemology.
Therefore, researchers are expected to point out, explain and justify the epistemology that
informs their choice of research methodology.
Consequently, the choice of epistemology is widely influenced by the ontological
considerations within a particular discipline (Quattrone, 2000). Though, both dimensions of
ontology (objective and subjective), plays an important role in the epistemology and ultimately,
the methodology chosen to conduct the research. Therefore, the next Sections 7.3.3 and 7.3.4
will elucidate how the ontological dimensions (objective and subjective), as well as
epistemological considerations, affect the choice of research methodology.
7.3.3 Quantitative Methodology
According to Jean (1992) the objectivist view of an integral and independent reality encourages
researchers to adopt the epistemology of positivism. Likewise, Kent (1999) informs that the
objectivist researcher strives to observe measure, analyse and predict relationships between
components that comprise reality. Kent (1999:11) further states that certain principles guide a
positivist’s search for reality, which includes:
242
1. Only phenomena that can be observed can be used to validate knowledge;
2. Scientific knowledge is arrived at through the accumulation of verified facts derived
from systematic observation or record-keeping;
3. Scientific theories are used to describe patterns of relationships between these facts to
establish causal connections between them; and
4. The process is neutral and judgment free. Observations are uncontaminated by the
scientist’ own prediction. Thus ethical issues can be included only if they are included
as part of the research.
Thus, a positivist epistemology would result in the use of a scientifically guided research
methodology where the aim is to explain and predict causal relations between elements that
constitute reality (Jean, 1992; Quattrone, 2000), which was considered in the present research.
The success of positivist research depends on the collection of data that can be quantified and
analysed using mathematical formulas (Maguire, 1987) as also adopted for the current thesis.
Likewise, positivist’s researchers advocate the use of quantitative methodology to explore and
explain relationships between variables. The presentation of research findings under this
methodology usually follows an approach that emphasises explicit, exact, scientific and formal
procedures (Sarantakos, 2005). For instance, positivist researchers have to use statistical
rhetoric such as reliability, unidimensionality, validity, correlation, cause and effect
relationships, to mention a few, which are in line with the scientific presentation of results
(Kent, 1999). Hence, the whole research process is considered to be highly neutral and
judgment free with limited room for personal bias (Sarantakos, 2005). As Sarantakos (2005:33)
further accentuates, “... the task of the researcher is to discover the scientific laws that explain
human behaviour using quantitative methods, similar to those of natural sciences”. According
to Kent (1999:11) “a researcher using quantitative methodology has to follow a number of steps
in conducting their research which usually include, generating the research problem, coming
up with expectations based on reality, generating hypothesis, defining variables, sampling, data
collection, analysis of data, report of findings and relating findings to the theory”.
Critics of quantitative methodology argue that placing emphasis on quantitative research
methods often leaves out important social phenomena that cannot be quantified (Maguire,
1987). In some cases “...complex social phenomena are reduced to meaningless quantitative
243
results in a bid to follow the norms of the methodology” (Maguire, 1987:22). May (2001)
suggests that detachment of the researcher from the research process neutralizes their
(researchers) influence on the researched, thereby depersonalizing and alienating them from
the world they are supposed to study. “This reduces researchers to research tools that do not
have a mind, while respondents become research objects and are treated as such” (Sarantakos,
2005:35). In reality it is not possible to totally detach the researcher from the research process
since their perceptions, expectations, experiences and interpretations ultimately become part of
the research process (May, 2001). Therefore the researcher’s subjectivity is considered an
integral part of the research process. May (2001) further suggests that the relationship between
the researcher and the research should not be exclusive, but should be “...a continuous ebb and
flow of information...” Thus advocates of subjectivity suggest that it is a better option for
undertaking research as opposed to objective quantitative methods (Brieschke, 1992).
In summary, quantitative methodology is appropriate in certain instances, for example in
scientific research where emphasis is on explicit, exact, and formal procedures (Sarantakos,
2005). Beside this, it has been heavily criticized as already discussed above. However, these
limitations can be reduced if it (quantitative methodology) is supported by qualitative
methodology (Tashakkori & Teddlie, 1998). Section 7.3.5 discusses the advantages of
combining qualitative and quantitative methodology in more detail. In the following section
qualitative methodology is discussed. The discussion focuses on both the advantages and
limitations of qualitative methodology.
7.3.4 Qualitative Methodology
According to Jean (1992), the subjectivist’s view of reality advocates for appreciation of human
involvement in the creation and shaping of knowledge. The subjectivist epistemology thus
suggests that meaning or reality is not discovered but is rather imposed on the object by the
subject, and in a research situation, imposed by the researcher (Crotty, 1998). In other words,
with the subjectivist epistemology, the object being studied contributes less to the meaning or
reality. Thus, researchers’ input in the research process is recognised under subjectivism. The
research methodology recommended by subjectivists is qualitative methodology. According to
Jean (1992:92) qualitative research is “…a form of social interaction in which the researcher
converses with, and learns about the phenomenon being studied”. In other words, the researcher
is part of the research process and is actively involved in creating the meaning of reality (Crotty,
1998).
244
Qualitative research is suggested as more applicable to the study of people and their
environment (social sciences) than natural sciences (Bryman, 2001). The intention is that the
object of research for natural sciences (chemicals, metals, atoms and others) cannot make sense
of their environment and are easy to manipulate while people can, and are, able to attribute
meaning to their environment. Consequently advocates of qualitative research advanced the
use of qualitative methodology when studying people as it enables the researcher to see through
the eyes of the researched (Bryman, 2001). Besides, the social world needs to be studied from
people’ viewpoints rather than to treat them as objects that cannot attach meaning to their
environment.
Constructing meaning through engagement with people involves interpretation. Thus, the
process by which information is extracted through interpretation is sometimes called
interpretivism (Sarantakos, 2005). Under interpretivism, researchers seek information relating
to people’s views, opinions, perceptions and interpretations of the social world (Crotty, 1998),
which was also partly utilized in the current research. Subjectivism, constructivism, and
interpretivism form part of a broader list of research methods commonly employed in
qualitative research. Quintessentially, qualitative research is a broad area with diverse research
methods. In that way, it (qualitative research) will not be extensively discussed here, since this
thesis is not solely qualitative.
Despite the positive contribution of qualitative methodology to the research process, it has
some setbacks. For instance, qualitative methodology has been criticized for lacking in efficacy
due to its inability to study with a degree of accuracy, the relationships between variables
(Sarantakos, 2005). Also, in qualitative research, the researcher is the main player, in the sense
that he or she decides on what to concentrate on. Moreover, what is observed and heard may
not necessarily be the same as what another researcher will observe (Bryman, 2001). It is
difficult to replicate and generalize the findings of qualitative research with ease because they
are more likely to be controlled given that only a small number of cases are studied compared
to large sample sizes common in quantitative research (Bryman, 2001). Consequently, the
number of cases may not be representative of the majority of the population being studied.
However, advocates of qualitative research argue that generalizations are made on the
assumption that the findings and inferences made during the research are supported by sound
theoretical reasoning (Mitchel, 1983). According to Ruyter and Scholl (1998),
245
representativeness in the case of qualitative research is not concerned with the size of the
sample representing the research population but rather representativeness in accordance with
the subject of investigation, which is highly subjective and a narrow-minded view of events
and what is being observed.
Another possible disadvantage of solely using qualitative research is that it is difficult to subject
findings of qualitative research to rigorous quality verification requirements such as reliability
and validity (Creswell, 1994). Validity requires measurement of the object of enquiry and that
is not possible in qualitative research because its purpose is not to measure but to generate ideas
(Stenbacka, 2001). Thus, it would be difficult to prove the validity of qualitative research
findings through measurement. On the other hand, reliability is concerned with producing the
same result with consistency. This is not possible under qualitative research because of the
involvement, influence, subjectiveness and the possibility of bias of the researcher in
qualitative research. Qualitative researchers have, however, argued that quality verification
using validity and reliability checks is not necessarily applicable to qualitative research because
it owes its origin to scientific rhetoric and positivist paradigms common in quantitative research
(Creswell, 1994; Stenbacka, 2001). Stenbacka (2001:555) further defending the paradigm
suggests that “...new concepts relevant to qualitative research been used instead of quality
concepts borrowed from quantitative research”.
Nevertheless, both qualitative and quantitative research methodologies can be used in different
situations depending on the aims and objectives of the study. For instance, Ellram (1996:98)
claims that most research is centered on four primary objectives. These are; “...exploration,
explanation, description and prediction” (Ellram, 1996:98). Research where the objectives are
either exploration and/ or explanation would normally require qualitative research methods.
This is because qualitative research has the ability to provide insight and explanation into a
phenomenon that was relatively unknown (Ruyter & Scholl, 1998). It provides answers to
questions, such as ‘how’ or ‘why’ which are common in exploration and explanation of
phenomena (Ellram, 1996). On the other hand, research that is descriptive and or predictive
would, in most cases, require quantitative research methods that utilize statistical techniques to
predict and describe relationships between variables (Ellram, 1996). This therefore implies that
the choice between the two areas of methodology should not be driven by like or dislike of
either method, but by the aims and objectives of the study, as well as the nature of the study.
246
In some cases the two methods may be used jointly to cover for the weaknesses inherent in
each method (Amaratunga, Baldry, Sarshar, & Newton, 2002; Tashakkori & Teddlie, 1998).
The process of combining quantitative and qualitative research methods is usually called
triangulation, which Tashakkori and Teddlie (2003) called Mixed Method or pragmatism. The
discussion on the combination of qualitative and quantitative methods; follows hereafter.
7.3.5 Combined Quantitative and Qualitative Methods
The combination of qualitative and quantitative methods has been supported theoretically by
many scholars such as Uysal and Crompton (1985); Creswell (1994); Tashakkori and Teddlie
(1998); Bryman (2001); and Amaratunga et al. (2002. There has been a suggestion that
quantitative and qualitative research methods are not dichotomous but rather can complement
one another to produce improved research findings (Tashakkori & Teddlie, 1998). Those
advocating the use of combined methods reject the forced choice between positivism and
constructivism as none of the methods work best in isolation (Tashakkori & Teddlie, 1998).
The use of combined methods, often called Mixed Method (MM), has been found to alleviate
the weaknesses linked with using either of the methods on their own (Tashakkori & Teddlie,
1998; Amaratunga et al., 2002; Bryman, 2001; Mangan, Lalwani, & Gardner, 2004). For
example, Bryman (2001:450) suggests that “...in some instances neither qualitative nor
quantitative research methods may be adequate on their own, thus researchers cannot rely on
just one method and have to use both to support the research process”. Quantitative and
qualitative methods supplement each other by providing richness and details that are otherwise
unavailable if each method were pursued separately (Jack & Raturi, 2006). Combining the
methods provides a multidimensional insight into the research problem, and thus assists in
getting a broader understanding as well as a truer analysis of the situation at hand (Mangan et
al., 2004), which is also one of the strong points of consideration for the current research. The
use of combined methods compensates for the weakness embedded in each of the research
method by “... counter balancing the strengths of another” (Amaratunga et al., 2002:23). Jack
and Raturi (2006) inform that triangulation provides confirmation of the research findings by
improving the ability of researchers to draw conclusions from their studies thereby resulting in
more robust and generalizable research findings.
The next section discusses the research design followed in this research. As identified in the
next section, this study adopted a Mixed Method methodology in order to counter balance the
strengths and weakness embedded in each of the research methods when used separately as
247
already discussed above. Further details on the justification and how quantitative and
qualitative methods were used to collect data in this thesis are provided.
7.3.6 Mixed Method Approach
In this thesis, a Mixed Method Approach was adopted based on the philosophical and practical
reasons, as discussed above. While the quantitative survey provides us with a snapshot of
phenomena, qualitative method -\interview from the Delphi Panelist (Delphi Study), provided
contextual information and human subjective information to interpret and inform the
quantitative results. Creswell, Clark, Gutmann and Hanson (2003) identified six commonly
used designs in Mixed Methods research. However, the present study uses two of those which
are: Sequential Explanatory and Concurrent Triangulation Design. A visual model of Mixed
Methods, as discussed by Creswell et al. (2003), and Teddlie and Tashakkori (2009), is also
used in the present study (Figure 7.1), to summarize and clarify the procedure. The quantitative
survey is the main driver of this study, complemented by the qualitative study. The use of both
methods provides richer understanding of phenomena and an explanatory account of
triangulation and illuminates significant survey findings in what Teddlie and Tashakkori
(2009) infer as Crossover Track Analysis. Although the quantitative and qualitative studies are
independent, however, both sets of data and analyses are used in analysis. In the following
chapters, the survey (quantitative) examining the relations and associations between the key
variables, and likewise the Delphi Study (qualitative) will be presented.
As briefly indicated, the current study used Mixed Methods Research, which involved both
quantitative and qualitative approaches. According to Tashakkori and Teddlie (2003) mixed
methods research is a research design with philosophical assumptions, as well as methods of
inquiry. As a methodology, it involves philosophical assumptions that guide the direction of
the collection and analysis of data and the mixture of qualitative and quantitative approaches
in many phases of the research process. As a method, it focuses on collecting, analysing, and
mixing both quantitative and qualitative data in a single study or series of studies. Its central
premise is that the use of quantitative and qualitative approaches in combination provides a
better understanding of research problems than either approach alone. The quantitative data in
a typical Mixed Method includes closed-ended information, such as those found on attitude,
behaviour, or performance instruments, which was also adopted in this thesis.
248
Figure 7.1: Visual Model of Mixed Methods Design
Source: Tashakkori and Teddlie (2003)
The collection of this kind of data involves using a closed-ended checklist, in which the
researcher checks the behaviour seen. Sometimes quantitative information is found in
documents such as census records or attendance records. The analysis of the qualitative data
consists of statistically analysing scores collected on instruments and checklists to answer
research questions or to test hypotheses or to answer the research questions (Creswell, 2003).
In contrast, qualitative data consists of open-ended information that the researcher gathers
through interviews with participants. The general open-ended or closed-ended structured
questions asked during interviews allow the participants to supply answers in their own words.
These can be thematically analysis and converted into qualitative data, which can also be
transcribed in quantitative data, for instance when the Delphi Technique is used, frequencies
of measures of central tendencies are used to draw consensus. Also, qualitative data may be
collected by observing participants or sites of research, gathering documents from a private or
public source, or collecting audio-visual materials such as videotapes or artefacts. The analysis
of the qualitative data (words or text or images) typically follows the path of aggregating the
words or images into categories of information and presenting the diversity of ideas gathered
during data collection. The open-versus closed-ended nature of the data differentiates between
the two types better than the sources of the data.
Mixed Methods Research is commonly used as a strategic research approach that is able “(a)
to demonstrate a particular variable will have a predicted relationship with another variable and
QUAN Data
Collection
QUAN Data
Collection QUAL Data
Analysis
QUAL Data
Collection
Comparison and integration of
QUAN and QUAL results
249
(b) to answer exploratory questions about how that predicted (or some other related)
relationship actually happens” (Tashakkori & Teddlie 2003:15). Hence, the current study used
both qualitative and quantitative methods to identify a variety of factors associated with
residential satisfaction and to indicate the statistical significance of these factors in determining
residential satisfaction. It also explored the relationship between the identified factors and
residential satisfaction process to be tested (predicted).
In addition, the current study sought to gain a better understanding of the impact of the
identified independent variables on the overall beneficiaries’ residential satisfaction in
subsidised housing schemes in South Africa. Thus, the study was interested in exploring the
way in which the factors identified by the qualitative study/- using a Delphi Technique and
through literature study predict residential satisfaction. For example, it investigated the
perceived importance, impact and influence of beneficiaries’ dwelling units amongst others
from different households and individuals and the way it influences their satisfaction with the
houses allocated to them. Hence, using only a quantitative research approach would not have
allowed more detailed information to be obtained. Therefore, a qualitative method was
employed to explore and gain a more comprehensive understanding of the way in which the
selected factors influence residential satisfaction from the perspective of selected housing
experts in the low-income housing industry and ultimately verifying through a quantitative
occupant survey, the effects of the factors on their overall satisfaction with the houses.
Mixed Methods Research was considered to be an appropriate research framework for the
current study because it helped to demonstrate a useful predictive process and provided a better
understanding of the factors which bring about residential satisfaction in subsidised low-
income housing in South Africa. The use of Mixed Method helped to confirm the findings of
both the quantitative and qualitative approaches (Flick, 2009). Hence, the research findings
could thus be used to establish a better understanding of the factors that needed to be considered
in developing sustainable human settlements low-income housing for the poor and
disadvantaged groups in South Africa.
The qualitative method adopted in this present study is the structured (using an interview
schedule) and semi-structured (using an interview guide) interview. This was made possible
through the use of the Delphi Technique (which is elaborately discussed in Section 7.4.3). The
findings from this section of the study helped to refine the survey tool (structured
250
questionnaire) for the study and to validate the findings. The Delphi findings were further used
to resolve conflicting issues surrounding residential satisfaction and other housing studies
issues surrounding South Africa low-income housing and in the developing countries at large
through the consensus that was reached during the Delphi study. While the quantitative method
of data collection for the study was through the Survey Method with the use of a structured
questionnaire. The analysis done through the use of structural equation modeling with EQS
Version 6.2, which was used in the development and validation of the beneficiary satisfaction
model.
7.3.7 Justification of the Mixed Method Approach
Both quantitative and qualitative methods have their strengths and weaknesses. For example,
quantitative methods have been criticized for being “sanitized and lacking in contextual
realism” (Tashakkori & Teddlie 2003:516). Qualitative methods are suitable for addressing
questions of how and why things occur, whereas quantitative methods are more appropriate for
answering what and how questions (Yin, 1994). In studying the variables that predict
residential satisfaction in low-income subsidised housing, the use of only one approach was
limiting as other housing issues relating to the creation and sustainability of low-income
housing needed to be explored from the various stakeholders (identified in the Qualitative –
Delphi study as experts) as well as the beneficiaries (occupants) of the housing units. For this
reason a Mixed Method approach that integrated qualitative and quantitative methods was
required.
One of the merits of a Mixed Method Approach in the current study is that the techniques of
the qualitative and quantitative domains, which are interwoven helped to “maximize the
knowledge yield of the research endeavour” (Tashakkori & Teddle, 2003:518). Also, another
advantage of the methods is that it allowed the researcher to discover and justify the model
components within one study. For instance, qualitative research involves people in order to
provide the realism and detail needed for the generation of hypotheses and building of theory
(Tashakkori & Teddle, 2003). Additionally, qualitative techniques permitted the gathering of
data that is rich in detail, which the researcher was interested in. By using techniques such as
the questionnaire survey interview for data gathering from the occupants of the housing units
and Delphi Technique; the language and context of the low-income housing stakeholders and
the people being studied were captured. It should be noted that although qualitative data was
gathered in both the first and second stage, the analysis approach was more aligned with the
251
positivist paradigm as it sought to identify patterns and repetition within each key research
issue and also explored the level of impact, influence and agreement through the use of scales.
Further, the adoption of a Mixed Method Approach helped to answer questions that would have
not been answered by qualitative or quantitative approaches alone. The method was very
practical because the researcher was free to use all methods possible to address the research
problem. The method was also used to increase the generalisability of the research result, which
was a major consideration in the present study. It also provided stronger evidence for a
conclusion through convergence and verification of findings. It also added insights and
understanding that would have been missed if only a single method was used. It also provided
complete knowledge necessary to inform theory, practice and was able to answer a broader and
more complete range of research questions because the researcher was not confined to a single
method or approach. Also because individuals tend to solve problems using both numbers and
words, that is, a combination of inductive and deductive thinking, therefore, it became natural,
to employ Mixed Methods research as the preferred mode of understanding the thesis
statement. For instance when a discussion about the satisfaction of the South Africa public
housing schemes beneficiary is debated (which is the research focus), both numbers and words
comes to mind. This is because the debate is natural, psychological and persuasive than either
words or numbers can adequately represent. This is because words, pictures and narratives can
be used to add meaning to numbers, and numbers can be used to add precision to words,
pictures and narrative. Hence, the Delphi Technique was combined with the Survey Method in
the current research which provided the basis for the validation of the conceptual framework
for the development of a holistic beneficiary’s satisfaction model in developing countries using
South Africa’s three metropolitan municipalities and one district municipality as a case study.
7.4 RESEARCH DESIGN
The research design has been defined as the framework for conducting research, which helps
researchers to ensure that the study will be carried out successfully (Churchill, 2001). Usually,
the research design is used to justify decisions and choices relating to the research procedure
(Sekaran, 2000). Following the decision on the appropriate methodology to use in this study
based on the ontological and epistemological assumptions, this section of the thesis is to decide
on the research design. Generally, the choice of research design is influenced largely by the
methodology (whether quantitative or qualitative) as well as the philosophical assumptions
252
guiding the research process (ontology and epistemology). For instance objectivist ontology
will influence a researcher to adopt a more positivist epistemology, which accentuates the use
of quantitative methods in the research process (Sarantakos, 2005) viz-a-viz the constructivist
ontology, which culminates in qualitative methodology. When objectivist ontology is adopted,
the research design will be more fixed and in line with the requirements of objectivism, which
advocate a scientific way of abstracting data. Ultimately the instruments to be used in collecting
data will also be determined by the research design, and in the case of quantitative design, the
Survey Method will be used to collect the data.
Accordingly, the course of deciding on a specific research design can be hypothesised in the
form of a connection starting from the philosophical underpinnings (epistemology and
ontology). Thus, the philosophical underpinnings provide a guide to the methodology followed
in a research process. Following the decision on the methodology the researcher has to decide
on the research design guided by the research questions and aims. Ultimately the research
design will influence the researcher on the choice of instruments to use in the execution of the
research process (Sarantakos, 2005). Figure 7.2 illustrates these connections.
Additionally, the precise justifications for a research design should show that the five aspects
– research purpose, theoretical framework, research questions, research methods, and sampling
strategy – are appropriately inter-connected according to Robson (2002). The current study
follows these aspects of research design. Hence, the choice of research methods for the current
study was influenced by the research aim, sub-questions and objectives.
253
Figure 7.2: Steps in the research design process
Source: Adapted from Sarantakos (2005:29)
Hence, there were three considerations made for selecting the research methods to answer the
predetermined set of goals for the research. Firstly, the selected research methods needed to
identify the variety of factors associated with residential satisfaction. Secondly, the selected
research methods had to be able to predict the relationship between each of these identified
factors and on how they can predict residential satisfaction. Finally, the selected research
methods needed to allow in-depth information to be collected and analysed in order to show
how beneficiaries perceive the identified factors as important (influential) in determining
residential satisfaction. Accordingly, the current study adopted the Mixed Methods research
(quantitative and qualitative combined) approach as already stated, discussed and justified
above. This methodology was adopted in order to answer the research questions and meet the
research objectives thus developing a residential satisfaction model that applies to the study
area.
Therefore, in order to meet the stated research objectives (Chapter 1, Section 1.3.2), the
following strategies were adopted:
1. For the first general objective, which was to establish the factors that determine
residential satisfaction in low-income housing and literature review was conducted
ONTOLOGY
EPISTEMOLOGY
METHODOLOGY
DESIGNS
INSTRUMENTS
254
about the factors that determine residential satisfaction. Published articles, housing post
occupancy reports, development reports and status reports were reviewed. Both
international and local Southern African reports and literature were reviewed. The
expected outcome from this objective was information and a global picture of the
determinants of residential satisfaction. This information and the general picture are
useful for the reader to have an understanding of how residential satisfaction is formed
and the extent of its relevance to the housing occupants.
2. The second general objective of the research was to establish the current theories and
literature that has been determined on residential satisfaction and to identify the gaps
that needed consideration. The established constructs were included as part of the
theory for the development of the holistic residential satisfaction model. The process
leading to this entailed a rigorous and exhaustive review of literature. This review was
conducted from a wide source of publications including journals, conference
proceedings, books and monographs etc. A review of literature on the subject of general
housing and with specific emphasis on the theories of residential satisfaction was done.
The expected output from this second general objective was information on the current
theories on residential satisfaction and especially, to determine the gaps, which other
scholars have not yet addressed; common themes and the type of methodologies that
have been used in the research and how terms had been defined. This information was
necessary as it was the core literature to inform the current research project.
3. The Delphi Method was used to achieve the third and fourth objectives which were to
determine the main and sub-attribute(s) that bring about residential satisfaction; to
examine if the attribute that determine satisfaction in other cultural contexts is the same
within South Africa. Also, to evaluate the critical factors and issues that affects the
delivery of low-income housing in South Africa. The Delphi Method was the best
method to use in this instance as the objectives entailed soliciting expert opinions on
the factors that determine satisfaction in low-income housing to rate their influence on
the occupants’ eventual disposition/- be it either satisfied or dissatisfied. These types of
questions can only be addressed by methods such as the Delphi and/ or focus groups
apart from experimental procedures which were not feasible for this study. Apart from
the Delphi study, focus groups could have been used except that there was the challenge
of bringing housing researchers and experts into one room and make them deliberate
255
for a minimum duration of eight hours per day for at least two days. Not only was the
focus group method not feasible, it was also considered costly beyond what was
budgeted for and it would have defeated the purpose of conducting a rigorous process
to achieve the objective. In the Delphi method, bias is eliminated by members of the
expert panel remaining completely anonymous to each other and therefore, there is no
undue influence from other peers. This is not the case for a focus group. A detailed
explanation of the Delphi Technique is explained in Section 7.4.3 in order to give an
idea to the reader of how the Delphi Method is conducted and what should be expected
from the Delphi Technique. The expected output is an estimation of the extent to which
residential satisfaction is influenced by the established factors; and consensus on the
critical factors and issues that affect the delivery of subsidised low-income housing in
South Africa. From these factors and interrelationships a conceptual model was
developed for residential satisfaction in subsidised low-income housing.
4. Research objective five was achieved by drawing on the conclusions from the extensive
literature review and the results and findings from the qualitative Delphi Study.
5. An empirical questionnaire survey was conducted and analysed using structural
equation modeling in order to achieve the sixth objectives of the research. With regards
to the sixth objective which was to test and validate the conceptual model developed
from the RO5, data obtained from the questionnaire sought to establish
interrelationships between the factors that determine residential satisfaction and to
establish the relationship produced amongst them and which constructs have a greater
influence on the determination of residential satisfaction in subsidised low-income
housing. The method was considered to be suitable for the type of information that was
being collected as the aim was basically to establish the core determinants of residential
satisfaction in subsidised low-income housing. A detailed explanation of the survey
concerning population, sampling procedure and analysis of results is presented in
Section 7.4.4. The expected output for the sixth objective was information to validate
the conceptualized holistic model and based on this; to finalize the best fit model for
residential satisfaction in public subsidised low-income housing in developing
countries, using South Africa as the case study.
256
The table below (Table 7.1) summarises the research methods that have been employed to
achieve the objectives of the research. The sections that follow explain in detail the methods
that have been used to achieve the objectives of the research.
Table 7.1: Research procedure
Stage Research
objective
Data collection
Method
Data analysis
method
Output
1.0
Literature
review
RO1: Establish the
factors that determine
residential satisfaction
in low-income housing.
Literature
review
Information and a
global picture of the
determinants of
residential
satisfaction.
RO2: Establish the
current theories and
literature that has been
advanced on residential
satisfaction and to
identify the gaps that
need consideration.
Literature
review
Information on the
current theories on
residential
satisfaction and
especially to
determine the gaps
which other
scholars have not
addressed.
Outline of
constructs (factors)
associated with
residential
satisfaction which
have not been
considered in the
previously
developed models.
2.0
Delphi
Technique
RO3: Determine the
main and sub-attributes
that brings about
residential satisfaction
and to examine if the
attribute that determine
satisfaction in other
cultural context is the
same with South Africa.
Delphi
Technique
Descriptive
statistics
Consensus on the
influence and
impact level of the
various attributes on
residential
satisfaction.
RO4: Evaluate the
critical factors and
issues that affect the
delivery of low-income
housing in South Africa.
Delphi
Technique
Descriptive
statistics
Consensus on the
critical factors and
issues that affect the
delivery of
subsidised low-
income housing.
257
RO5: Develop a
holistic residential
satisfaction model for
subsidised low-income.
Desk study
Questionnaire
Survey
Delphi
Technique
Theory Holistic integrated
residential
satisfaction model.
3.0
Questionnaire
Survey
RO6: Determine the
validity of the
conceptualized holistic
residential satisfaction
model for subsidised
low-income housing
Questionnaire
Survey
Structural
Equation
Modeling
(SEM)
EQS
Information to
validate conceptual
model;
Validated best-fit
model.
7.4.1 Methods
This section gives an overview and detailed description of the methods that were used in order
to meet the objectives of the research. The methods described in this section include the
literature review, the Delphi Method and the Questionnaire Survey. The section of the analysis
(how data were treated) of the data obtained using each of the methods mentioned are also
described in detail. Figure 7.3 below, is an outline of how the study was conducted and will aid
the reader in understanding the detailed description of the methods described in the next
sections.
7.4.2 Literature Review
The literature review is one of the most important aspects of developing a study and also as a
way to research what has already been written on the subject, methodologies that have been
used to investigate similar concepts or phenomena and to establish the trends on the solutions
that are being advanced to solve the many problems that face mankind (Heppner & Heppner,
2004:52). Hence, Boote and Beile (2005) inform that literature is the foundation of research.
It was therefore necessary to conduct a literature review for the current study in order to
establish the:
1. determinants of residential satisfaction from the researched work of others;
2. theories and literature on residential satisfaction and the concept of housing; and
3. residential satisfaction gaps which can improve the satisfaction of low-income
subsidised housing occupants.
258
Stage 3 Stage 2 Stage 1
Need for study
Literature review
Develop conceptual
model based on the
determinants of
residential satisfaction
Questionnaire
Delphi
Evaluate factors
influence on occupants’
residential satisfaction
Analyse & model
results to validate
conceptual model
Best fit model &
recommendations
The literature review on the above aspects was essential as it sets the broad context of the study
to the reader, highlights what has already been done before on the subject under consideration,
relates the present research to the on-going debate on the subject and hence, provides a
framework for comparing the results of the present research with other studies on the subject.
Figure 7.3: Research Design Outline
Source: Manu, Ankrah, Proverbs and Suresh (2010:29) and Musonda (2012:91)
In order to assure integrity and sophistication of the study, an effort was made to ensure that
the literature analysis was thorough and comprehensive. Studies reviewed were cohesive and
also considered methods adopted or used in other studies. A detailed analysis of the methods
used in other studies similarly ensured that the review did not only report the claims made in
the existing literature as this is one of the remedies against that trap (Boote & Beile, 2005). The
materials used for the literature review were books, reviews of articles on the subject both
published and unpublished, such as dissertations, and thesis. In addition, names of leading
authors and contributors on the subject where drawn from the references of the consulted
articles and explored to establish their publication history, as well as conducting focused
searches within research databases. Articles from the above mentioned sources were read and
re-read in order to establish the progression of research in the area, specifically on the topic
under study. Hence, this conformed to the methods of conducting a literature review as
described by Boote and Beile (2005), which involves:
1. finding a broad range of high-quality, specific articles, books, dissertations and reviews,
directly related to the study;
259
2. reading and re-reading to establish progressions and trends;
3. summarizing the studies read;
4. identifying methodologies adopted in the studies;
5. relating the current study to those reviewed; and,
6. writing the literature.
The eventual output from the literature review was a clear standpoint on the topic and an
indication of where the study fits in, in relation to other studies on the subject, as well as
providing a framework for comparing the results of the study with others. Considerable time
was spent on the review of existing literature because as Boote and Beile (2005) states, ‘good’
research is good because it advances our collective understanding.
Findings from the literature review were that, there are various factors which determine
residential satisfaction, but in varied contexts, as there are no universal factors, which guarantee
occupants satisfaction. It was also found that there are other factors, which should be
considered, key constructs in the ‘bundle of factors’, which brings about residential
satisfaction, but have not been considered in the previous models developed. After establishing
the above from literature, theories were developed about the influence of the missing factors
and their interrelationships with other factors, which have been advanced in literature to
determine satisfaction with low-income housing. These therefore needed to be tested on
whether they would influence residential satisfaction of the occupants of subsidised low-
income housing and if so, to what extent. In order to achieve this, the Delphi Method described
below, was used.
7.4.3 The Delphi Method
The Delphi Method was used for the second stage of the study to identify the main attributes
that brings about residential satisfaction and to examine if the attributes that determine
satisfaction in other cultural contexts as identified from the literature, is the same within South
Africa. Also, the Delphi Technique was used to explore the extent of these main-sub attributes
- factors impact / influence on residential satisfaction in South African low-income housing.
The Delphi Technique was originally developed in the 1950s, as a tool for forecasting and
problem solving of complex topics at the Rand Corporation by Helmer and Dalkey (Buckley,
1995). It was named after the ancient Greek temple where the oracle could be found. According
260
to Greek mythology, the oracle at Delphi was consulted to forecast the future so that correct
and timely decisions could be made before embarking upon a major course of action, such as
waging war. The method adopted by the research team at the Rand Corporation was that
subject-matter experts could be solicited for their opinion about the likelihood of future events
or scenarios.
The Delphi Technique is a qualitative methodology seeking to produce a consensus of a group
of experts on an issue of concern (Miller, 1993) through a survey consisting of rounds. The
Delphi Method is based on structural surveys and makes use of intuitively available
information of the participants, who are mainly experts in their various fields. The method
provides qualitative as well as quantitative results, and has beneath its explorative, predictive
even normative elements (Cuhls, 2003). There is agreement that Delphi is an expert survey in
two or more ‘rounds’ in which the second and later rounds of the survey (the results) of the
previous round are given as feedback. Thus, the experts answer from the second round based
on the influence of the other experts opinions. Thus, the Delphi method is a relatively strongly
controlled group communication process, in which matters, on which naturally unsure and
incomplete knowledge is available, are judged upon by experts (Häder & Häder, 1995). The
technique requires knowledgeable and expert contributors individually responding to questions
and submitting the results to a central coordinator (researcher). The coordinator (researcher)
processes the responses, looking for central and extreme tendencies, and their validations
(Grisham, 2008). The results are then fed back to the input provided by the coordinator
(researcher). The experts are then asked to resubmit their opinions, aided by the input provided
by the coordinator (researcher). This process continues until the coordinator sees that a
consensus has been formed on the questions asked.
The method was intended to remove the bias that is possible when diverse groups of experts
meet, which is common with other methods of decision making. With the Delphi Method, the
experts do not know who the other experts are. Hence, the Standard-Delphi-Method is a survey,
which is directed by a coordinator as already stated and comprises several rounds with a group
of experts, who are kept anonymous and for whose subjective-intuitive prognoses a consensus
is aimed at (Cuhls, 2003). After each survey round, standard feedback about the statistical
group judgment calculated from the median, the percentages and the interquartile range of
single projections is given and if possible, the arguments and counter arguments of the extreme
answers are fed back. In the Delphi process, nobody ‘loses face’ because the study is done
261
anonymously using a questionnaire. Rowe et al. (1991) and Häder and Häder (1995) states that
it is commonly assumed that the method makes better use of group interaction whereby the
questionnaire is the medium of interaction. The Delphi Method is especially useful for long-
range forecasting, as expert opinions are the only source of information available.
The Delphi Technique is part of a group of decision-making (policymaking) techniques that
includes the nominal group technique (NGT) and interacting group method (IGM). The Delphi
Technique differs in various ways from NGT and IGM respectively, but primarily due to the
fact that Delphi is individual based, anonymous and independent. The element of group
interaction is eliminated from the process and feedback to questionnaires is in written format
(Loo, 2002). Over time, the Delphi Method has gained popularity across many scientific
disciplines, as a method of inquiry. Czinkota and Ronkainen (1992) indicate that the Delphi
Method has gained considerable approval across disciplines. They further claim that it has been
used as a study instrument in the fields of library and information science (Buckley, 1995), in
the medical disciplines (Linstone & Turoff, 1975), in multi country studies of communications
in Europe, and by actuaries to predict economic conditions (SOA, 1999). Czinkota and
Ronkainen (1992) further report that those experienced with the Delphi Technique report that
the method produces valuable results which are accepted and supported by the majority of the
expert community. Similarly, in the business field, the technique has been highly rated by some
as a systematic thinking tool, but has been challenged in its ability to serve as an identifier of
strategic issues (Schoemaker, 1993).
Since the current thesis might be the initiation of further study and is aimed to attract a wide
spectrum of inputs from various geographically dispersed experts in South Africa; the Delphi
Technique is well suited as a research approach and method for the current study, more so as
the techniques has not been used in a similar study in South Africa or in any other developing
country. The Delphi Method was preferred to common survey methods as the current study
was addressing the ‘what can -if’ kind of questions, as opposed to the ‘what is’ kind of
questions. Delphi is more suited for these kinds of questions to explore concepts that are
difficult to measure except through experimental methods. Unfortunately, an experimental
survey was not feasible and appropriate for the current study. The Delphi Method was also
considered to be a robust method of rigorous query of experts. Unlike ordinary survey research,
the Delphi’s strength also lies in the rounds used, which provide an opportunity for initial
feedback, collation of feedback, and distribution of collated feedback to participants for further
262
review (Stitt-Gohdes and Crews, 2004:62). This unique process requiring group
communication is central to the strength of the Delphi (Stitt-Gohde and Crews, 2004:62). Also,
Loo (2002) informs that the Delphi process should be used when investigating policy-making
or policy-evaluation strategies that will set the future direction for the public or private sector,
respectively. The current thesis is aimed at setting the future direction for residential
satisfaction in low-income public housing in South Africa, hence the method was also
considered useful.
Delphi as a research method has had its fair share of criticism, support and debate on
epistemology (Mullen, 2003). Foremost amongst the criticism is Delphi’s alleged failure to
follow accepted scientific procedures, in particular the lack of psychometric validity (Sackman,
1974). In response to Sackman’s criticism, Coates (1975) states that if it is believed that the
Delphi Technique is of value not in the search for public knowledge, but in the search for public
wisdom; not in the search for individual data but in the search for deliberative judgment, one
can only conclude that Sackman missed the point. However, it should be noted that the
approach deals with areas that do not lend themselves to traditional scientific approaches;
hence, Helmer (1977) argues that the forecasting tendency, one of the major applications of the
Delphi, is inevitably conducted in a domain of what might be called ‘soft data’ and ‘soft law’.
Helmer (1977) further determines that standard operations research techniques should be
augmented by judgmental information and that the Delphi Method cannot be legitimately
criticized for using mere opinion and for violating the rules of random sampling in the ‘polling
of experts’. Such criticism Helmer (1977) argued rests on a gross misunderstanding of what
the Delphi Method is; it should be pointed out that a Delphi inquiry is not an ‘opinion poll’. As
all the above definitions illustrated, in no instance is reaching a majority opinion the ultimate
goal in a typical Delphi study; it is rather the reaching of agreement (consensus). According to
Buckley (1995), Delphi is a tool for discovering agreement and identifying differences rather
than forcing consensus. Buckley (1995) further informs that: in principle, agreement alone is
not a sufficient condition for arguing the acceptance of the Delphi Method. But as with the
majority of research methods, the method of use and application has an enormous influence on
the eventual success of the inquiry. Hence, where no agreement is achieved, the Delphi still
helps to clarify the issue being investigated. Linestone and Turoff (2002) assert that one of the
common reasons for failure in a Delphi Study is ignoring and not exploring disagreement.
Therefore, the current research is not only about reaching or forcing a consensus, but will
recognize disagreement and explore the reason for such.
263
In addition to the above criticism of the Delphi Technique, different authors also state different
weaknesses of the Delphi Technique including:
i. that it has not been shown consistently that the results from the Delphi Method are any
better than those achieved through other structured judgmental techniques (Rowe,
Wright & Bolger, 1991);
ii. that the Delphi Study is at the mercy of the world view and biases of the coordinating
or monitor team (researcher), who chooses the experts, interprets the returned
information and structures the questions. There is an enormous debate whether the
experts should be chosen from within or outside of the organisation initiating the study
and whether they should be experienced in the subject area of the study in question
(Masini, 1993);
iii. that another limitation according to Linstone (1978) is in the way the process and
questionnaire is structured, which Linstone (1978) believes can lead to a bias (like IQ
tests), which assume a certain cultural background. Hence, the experts may give
responses they think the monitoring group wants to hear, or they may not respond at
all. Consequently, the cultural background of respondents will have an impact upon the
results;
iv. likewise, Simmonds (1977) debates that one of the key flaws in the Delphi Technique
is that certain questions are not asked as they do not seem important when the study
begins. Nonetheless, once the study begins, new questions cannot be added, which in
turn can weaken the study considerably;
v. Lang (1995) states that the process of choosing the panelists is often not considered
seriously enough. Yet, it is the caliber of the panelists that determines the quality of the
outcomes of the study (Lang, 1995);
vi. that in the process of achieving consensus, extreme points of view run the risk of being
suppressed, when in fact they may provide important new information or insights
(Lang, 1995);
vii. that the flexibility of the technique means it can be adapted to a whole range of
situations, which in turn can make it vulnerable to misrepresentation and sloppy
execution (Amara 1975); and
viii. Garrod (2008) found that the Delphi Technique can be extremely sensitive to: the level
of panelists’ expertise; the composition of the panel; clarity of the questions; the way
264
the research or coordinator reports reasons for outliers and the administration of the
questionnaire.
Despite the limitations noted above from different scholars, Brill et al. (2006) describes the
Delphi as a particularly good research method for developing consensus amongst a group of
entities having expertise on a particular topic where information required is subjective and
where participants are separated by physical distance (Linstone & Turoff, 1975). Brill et al.
(2006) further states that the Delphi Method has been validated in the literature as a reliable
empirical method for reaching consensus in a number of areas. Amongst these areas are:
distance education (Thach & Murphy, 1995), journalism (Smith, 1997), visual literacy (Brill et
al., 2000), electronic commerce (Addison, 2003), health care (Whitman, 1990) and others.
Beside these areas, the method has also been used in many other disciplines, such as in
information technology (IT) research to identify and rank key issues for management attention
(Delbecq et al., 1975); scientific study of GIS (Hatzichristos & Giaoutzi, 2005), quality
management (Saizarbitoria, 2006), terrorism (Parente et al., 2005), banking (Beales, 2005),
social sciences (Landeta, 2006), privatization of utilities (Critcher & Gladstone, 1998),
education (Yousuf, 2007), etc. Based on the extensive usage of the method over time, the
Delphi Method in research is an accepted practice. However, as discussed above, it is not
appropriate for all research activities.
7.4.3.1 Epistemological Approach towards the Delphi Design
The variance amongst the various group techniques and the definition of the Delphi Method as
complied by various scholars and cognizance of the various criticisms forms the
epistemological foundation for defining the approach towards a typical Delphi Study design.
This is done by assuring that all expert feedback is anonymous.
According to Scheele (2002), the concreteness of the framework of the Delphi Design is vital
in researching the overall objective of the study. The basic premises of the Delphi research
design towards the development of a residential satisfaction model for the low-income groups
in developing country; is entrenched in some form of general agreement and consensus
regarding the core ingredients and components of the subsequent framework. Given the current
status of low-income housing in South Africa and the absence of a general agreed upon
residential satisfaction model framework for low-income housing, the search for consensus and
265
a point of departure in attributes that determine residential satisfaction in other low-income
housing issues is therefore justified. Hence, the objective of the Delphi Design for this study is
to obtain the most reliable consensus of opinion of a group of experts in the field being studied.
According to Lang (1995), the Delphi Technique is mostly used to solicit the opinions of
experts to determine the timing and possible occurrence of future events. It is a method that is
best used where there is little past data available to extrapolate from, and where social,
economic, ethical and moral considerations are pre-eminent. Considering the outcome of the
literature review of the current research (there is no structured research so far carried out or the
presence of a developed model to identify and predict the determining factors which brings
about residential satisfaction in subsidised low-income housing in South Africa) and definition,
function and nature of the Delphi Technique, it is justified that the Delphi Technique is one of
the best methods to explore the subject of this research and to further explore the research aim
and objectives.
7.4.3.2 When to use the Delphi Technique
The Delphi Method is mainly used when long-term issues have to be assessed such as the
subject of the current research. This is because it is a procedure used to identify statements
(topics) that are relevant for the future; it reduces the tacit and complex knowledge to a single
statement and makes it possible to be judged (Cuhls, 2003). Hence the use in combination with
other methodologies, like survey design in modeling residential satisfaction can be interesting.
On the other hand, in more complex issues, when the themes cannot be reduced that much or
when thinking and discussions in alternatives are the major target, the Delphi is not the method
of choice. It is also suitable if there is the (political) attempt to involve many persons in
processes (Eto, 2003). Hence, Linstone and Turoff (2002) argue that one or more of the
following properties could lead to the need for the use of the Delphi technique:
i. When the problem of inquiry does not lend itself to precise analytical Techniques but
can benefit from subjective judgments on a collective basis (Buckley, 1994);
ii. The research need to contribute to the examination of a broad or complex problem with
no history of adequate communication and may represent diverse backgrounds with
respect to experience or expertise, which is a major premise of the current research;
iii. More individuals are needed than can effectively interact in a face-to-face exchange;
iv. Time and cost to make frequent group meetings is limited;
266
v. The efficiency of face-to-face meetings can be increased by a supplemental group
communication process;
vi. Disagreements among individuals are so severe or politically unpalatable that the
communication process must be refereed and/or anonymity assured; and
vii. The heterogeneity of the participants must be preserved to assure validity of the results,
such as the avoidance of domination by quantity or by strength of personality called the
‘bandwagon effect’.
According to Cuhls (2003), the Delphi Method as a foresight tool seems to possess certain
degrees of invariance to survive the changing challenges of the past 50 years. Hence, the
process could serve different understandings of predicting or premonition and is probably
understood by the users as being relevant for covering technical perspectives, organizational
perspectives, but also personal perspectives (Cuhls, 2003). Cuhls (2003) further emphases that
what the users of the Delphi Technique especially like are the sets of data about the future that
are collected. Writing down future topics seems to have an immense psychological effect
because it transfers implicit to tacit knowledge to the more visible, explicit, and therefore
transferable knowledge (Cuhls, 2003).
7.4.3.3 Components of the Delphi Technique
The main components of the Delphi Technique according to Loo (2002), consists of five major
characteristics, which are adopted in the current study:
i. The study should consists of a panel of carefully selected experts representing a broad
spectrum of opinion on the topic or issue being examined;
ii. The participants are usually anonymous;
iii. The coordinator (researcher) constructs a structured questionnaires and feedback
reports for the panel over the course of the Delphi process;
iv. It is an iterative process often involving three to four iterations called ‘rounds’ of
questionnaires and feedback reports; and
v. There is an output, typically in form of a research report containing the Delphi results,
the forecasts, policy and program options (with their strengths and weaknesses),
recommendations to senior management and possibly an action plan for developing and
implementing the policies programs.
267
Likewise, Hasson et al. (2000) recommended that the following research guidelines for using
the Delphi technique. They suggest that the following subject matter be addressed in designing
a Delphi approach:
i. Research problem identification: Turoff (1970) outlined four objectives that call for the
use of the Delphi Technique. One of those objectives was to relate informed judgments
on a topic that spans a wide range of disciplines. Reid (1988) contended that the
decision to use the Delphi Technique must centre upon the appropriateness of the
available alternatives. Reid (1988) claims that the use of experts in a field of study is a
perfectly suited technique if:
the technique has not been utilized in the past, based upon the research
performed, such as the current study that has not employed the Delphi
Technique as a tool of investigation in low-income housing in South Africa; and
it offers the opportunity to check the validity of the cross-disciplinary (social,
psychological, ethical, managerial, cultural, anthropological, etc.) nature of the
issue.
ii. Understanding the process: The Delphi Technique is a multistage process designed to
combine opinions into group consensus (McKenna, 1994). The process being:
Pilot testing of a small group;
Initial questionnaire – qualitative comments solicited (not in all cases);
Initial feedback – quantitative after statistical analysis of the initial opinions;
Subsequent questionnaire – qualitative comments solicited again; and
Subsequent feedback – quantitative after statistical analysis. This provides an
opportunity for participants to change their opinions.
iii. Selection of experts: It is important to select panel members who are impartial, and are
interested in the topic. Some studies have over 60 experts, some as few as fifteen.
Selection of people knowledgeable in the field, and their commitment to multiple
rounds of questions on the same topic are essential. In the section that provides more
details regarding the practical design and execution of the Delphi Study for this thesis,
an elaborate detail on how the experts were chosen for the study, will be presented.
iv. Informing / invitation to experts: It is imperative to explain what is required of them,
how much time it will require, what they will be required to provide, what the objective
of the study is, and what will be done with the information.
268
v. Data analysis: This is the process where opinions of the experts are solicited. According
to Green et al. (1999) two or three rounds are preferred. Green et al. (1999) suggest that
an 80 percent consensus should be the goal. Likewise, Crisp et al. (1997) suggests that
percentages should not be used, but rather the process should stop when stability of the
data occurs. Contrary to Crisp et al. (1997) preposition, percentage estimation was
found suitable to this study as one of the means to achieve consensus, hence, a 60
percent consensus goal was set for the three round Delphi Studies. Also, analytical
software can be utilised to analyse the responses, and provide feedback to the experts
on the central tendencies (median and interquartile range) and on the levels of
dispersion (standard deviation). Hence, Lincola and Guba (1985) state that the criteria
for qualitative studies, such as the Delphi Technique should be credibility
(truthfulness), fittingness (applicability), audit ability (consistency), and confirmatory
ability.
vi. Presentation and interpretation. There are a number of methods for presenting the data
from a typical Delphi Study, with two methods being graphical and statistical. These
two methods have been used in the current research.
Therefore, given the nature of the current research, it is further believed that the Delphi
Technique is well-suited to obtain credible inputs from experts in industry, academics,
government and NGOs to serve as key input in the development of a residential satisfaction
model for low-income housing. The next section provides an overview of how the Delphi
Technique is used in this thesis.
7.4.3.4 Designing, Constructing and Executing the Delphi Study
Given the rationale behind the Delphi Technique and the main features explained above, the
design, construction and execution of the Delphi Study for the current research followed a
sequential process as suggested by Loo (2002). According to Loo (2002), four vital planning
and execution activities should be followed, which are:
i. Problem definition;
ii. Panel selection;
iii. Determining the panel size; and
iv. Conduction the Delphi iterations.
269
Supporting Loo’s (2002) approach, Delbecq et al. (1975) suggest a basic Delphi Methodology
that includes distinct stages such as, Delphi Question Development (objective), expert panel
selection, sample size, first questionnaire, first questionnaire analysis and follow-up
questionnaires. This methodology forms the basis of the current Delphi research study and is
explained in the subsequent sections. Table 7.2 gives a summary of the Delphi design,
construction and execution.
Phase 1 – Delphi Question Development
The formulation of the Delphi question is vital to the whole process. It is paramount that the
panel of experts understands the broad context within which the questionnaire is designed,
especially with the current research where the concept of what determines housing satisfaction
has different connotations; hence the concept had to be broadly clarified. For the Delphi Study
to achieve the objectives, key questions were asked. The basis of constructing the questions for
this current study was based on the guidelines given in Table 7.2, with corresponding wording
and phrasing given for this study.
Table 7.2: Delphi question formulation
Key Delphi questions?
Phrasing for this study
Why are you interested in this study? This study was initiated because of the belief
that not all beneficiaries who received
government low-income houses are satisfied
with what was allocated to them. Therefore,
this assumption is concrete because there is
lack of understanding of the diverse features
that determine housing satisfaction.
What do you need to know that you do
not know now?
Despite the knowledge about the features that
bring about residential satisfaction; they have
not been placed together and put into a model
to inform policy and predict housing
satisfaction in the low-income groups into
South Africa. At the end of this study, it
should be obvious what the attributes are that
determine residential satisfaction in low-
income government provided housing.
How will the results from the Delphi
Study influence residential
satisfaction?
The result of the Delphi Study will enable the
development of a conceptual framework for
the residential satisfaction model to be
developed. Hence, the attributes which would
collectively predict and assure housing
270
satisfaction with South Africa’s low-income
housing will be established.
Phase 2 – Delphi Expert Panel Selection
A critical part of conducting a Delphi interview technique is selecting the right experts (also
known as panellists, participants or respondents) and their role is crucial to the success of the
research (Hasson et al., 2000). Experts to be selected must be sufficiently interested and
involved in the subject being examined to ensure a high commitment response rate. According
to Hasson et al. (2000), controversial debate occurs when a professional becomes an ‘expert’.
The claim that one group represent’s valid expert opinion has been criticized as scientifically
untenable and overstated (Hasson et al., 2000).
For the purpose of this research McKenna’s (1994) defines ‘expert’ as being a panel of
informed individuals (otherwise called experts hereafter is used). McKenna’s (1994) definition
was further supported by Goodman (1987:730) stating that the Delphi technique “tends not to
advocate a random sample of panellist … instead the use of experts or at least of informed
advocates is recommended”. Likewise, Helmer (1977:18-19) argues that since a “Delphi
inquiry is not an opinion poll, relying on drawing a random sample from the population of
experts is not the best approach, rather, once a set of experts has been selected (regardless of
how – but following a predetermined qualifying criteria), it provides a communicative device
for them that uses the conductor of the exercise as a filter in order to preserve anonymity of
responses’, which is the core of the Delphi Technique. Therefore, Linstone & Turoff (2002)
states that the most significant danger in selecting the panel of experts lies in the path of ‘least
resistance’ through the selection of a group of cosy friends and / or like-minded individuals,
which thus negates the strength of the process.
Since panellists form the cornerstone of the Delphi Technique, clear inclusion criteria should
be applied and outlined as a means of evaluating the results and establishing the study’s
potential relevance to other settings and populations (Igbal & Pipon-Young, 2009). The
selection of panellists for the study was based on criterion sampling. Panellists were selected
for a purpose to apply their knowledge to a concept raised in the study based on the criteria that
was developed from the research questions under investigation. A Delphi Study does not
depend on a statistical sample that attempts to be representative of any population. It is a group
decision mechanism requiring qualified experts who have deep understanding of the issues
271
(Okoli & Pawlowski, 2004). Therefore, one of the most critical requirements is the selection
of qualified experts as it is the most important step in the entire Delphi Process because it
directly relates to the quality of the results generated (Hsu & Sandford, 2007). In agreement to
this, Stitt-Gohdes and Crews (2004:61) argue that careful selection of the panel of experts is
the keystone to a successful Delphi Study.
According to Dalkey and Helmer (1963), there are detailed criteria for the selection of panel
experts; recommending that in a typical Delphi Study, experts should meet the following two
recommendations, which were also postulated by Rodgers and Lopez (2002). The first
recommended criterion is that the experts should exhibit a high degree of knowledge of
experience in the subject matter. Another criterion is that they should be representatives of the
profession so that their suggestions may be adaptable or transferable to the population.
Similarly, Adler and Ziglio (1996) stated that the Delphi participants in any study should meet
four ‘expertise’ requirements, which are: knowledge and experience with the issues under
investigation; capacity and willingness to participate; sufficient time to participate in the Delphi
studies; and effective communication skills.
In choosing panellists for this study, each expert was required to meet at least five (5) of the
following minimum criteria:
1. Residency: Have lived or is living within one of the South Africa Metropolitan or
District municipalities cities; for at least a year.
2. Knowledge: Has knowledge of the low-income housing situation in South Africa,
specifically of the RDP housing; knowledgeable in the housing field; knowledgeable in
the field of housing through reading or occasional reading of housing related materials.
3. Academic Qualification: Has an earned degree (National Diploma/B-Degree/M-
degree/PhD) related to any field. Post-doctoral, training, certification
employment/experience focusing on sustainable development issues.
4. Experience: Has a history of / currently is performing consultation services for a South
African organ of State, individuals, businesses, agencies, companies, and/or
organizations, related to low-income or other sustainable development or human
settlement context. The experts must exhibit a high degree of knowledge of experience
in the subject matter and an extensive theoretical knowledge thereof.
272
Figure 1.4: Diagram of the Delphi Process
5. Employment: Currently serves (or has previously served) in a professional or voluntary
capacity (e.g. at place of employment - institution, business, agency, department,
company) as supervisor or manager of an establishment that is involved with housing
or sustainable human development related to issues in South Africa.
6. Influence and Recognition. Has served / currently is serving as a peer-reviewer for one
or more manuscripts received from a journal editor prior to its publication in the
primary literature, with focus on manuscript(s) on housing or sustainable development.
Step 1:
Literature review of housing studies
Step 2:
Housing studies expert selection
Step 3:
Development of 1st Round Questionnaire
Step 4:
Questionnaire Quality Check
Step 5:
Completion of 1st Round
Questionnaire
Step 6:
Analysis of 1st Round Questionnaire
Step 7:
Development of 2nd Round Questionnaire
Step 8:
Completion of 2nd Round
Questionnaire
Step 9:
Analysis of 2nd Round Questionnaire
Step 10:
Has Consensus been
reached?
Step 11:
Produce Report
No Rei
tera
te S
tep
s 7-9
un
til
con
sen
sus
is
reac
hed
273
7. Authorship: Is an author/co-author of peer-reviewed publications in the field of housing
with emphasis on South Africa; has prepared and presented papers at conferences,
workshops or professional meetings focusing on housing, sustainable development and
human settlement.
8. Research: Has submitted one or more proposals to or has received research funding
(grant/contract) from national, provincial, local government, regional, and/or private
sources that support housing development and studies for the low-income group or
other human settlement-related issues.
9. Teaching. Has organized, prepared, and successfully presented one or more housing or
human settlement or sustainable development training workshops focusing on the group
for which expertise is sought. The workshop or course must have been inclusive of the
low-income group. Or, has served as an individual or as a collaborative instructor in the
teaching of one or more college or university courses focusing on the sustainable
development or related field.
10. Membership: Be a member of a professional body. Should be the representative of a
professional body so that their opinions may be adaptable or transferable to the
population.
11. Willingness: Panel members must be willing to fully participate in the entire Delphi
study.
The adoption of five of these criteria was considered more stringent than the recommended
number of at least two criteria by Rogers and Lopez (2002) and Dalkey and Helmer (1963).
The five minimum criteria were framed after the four recommendations made by Adler and
Ziglio (1996), with the inclusion of experts’ residency status, which was considered to be
compulsory for all selected experts. This was considered significant because experts were
required to have a wide-ranging understanding of the low-income housing context in their
residential metropolitan and district municipality cities since the setting for the study is based
on three metropolitan municipalities and one district municipality. Also, a minimum number
of five criteria were set because the method may be undermined if panellists are recruited who
lack specialist knowledge, qualifications and proven track records in the field (Keeney et al.,
2001) amongst others. Although of course expertise comes in many guises and may include
those who are ‘experts by experience’ (Hardy et al., 2004).
274
Panel members were identified from four sources. The first source was from the South African
institutions of higher learning faculties, departments, research institutes like the Council for
Scientific Institute for Research (CSIR) and Non-Governmental Organisations (NGOs) like
FinMark Trust, Affordable Housing Finance in Africa, a division of FinMark Trust amongst
others, who engage in housing and other sustainable human (e) settlement-related issues. This
is because most consultants recruited for research and other consultation by the South African
Human Settlement Department are mostly from institutions of higher learning and research
bodies. The second source was the Department of Human Settlement. This is because they are
the ones who are vested with the responsibility for the initiation and development of subsidised
low-income housing in the country. Hence, their involvement in the Delphi Process was a key
consideration. The third source was from various conference proceedings, such as the annually
held Built Environment Research Conference hosted by the Association of Construction
Schools of Southern African, the Construction Industry Development Board Biannual Post
Graduate Research Conference, amongst others. Individuals who had frequently appeared as
authors or key speakers related to housing and human settlement issues in these proceedings
were identified as potential experts on the study. The fourth and last source was the references
of individuals who had committed their lives working in the area of sustainable human
settlement and housing related issues in Southern Africa.
With regard to the recruitment process itself, panellists were recruited via e-mail, with a brief
overview of the study objective included therein. Thereafter, those that consented to the
preliminary invitation were sent a detailed description of the Delphi Study (See Appendix C);
and were requested to send their curriculum vitae in order to confirm their areas of expertise
and to ascertain whether they meet the qualifying criteria. Hence, all experts selected for the
current study met the five criteria’s set for the study. After the verification exercise, selected
experts were then sent the First Round Questionnaire Survey (See Appendix D), which was
presented in form of both closed and open-ended question. Panellists were judged whether they
qualified to be experts and included in the study based on their curriculum vitae that they were
requested to submit in response to the initial invitation. From all the sources mentioned above,
55 invitations were sent out. Out of 55 invitations, 17 responded to the invitation, 17 completed
the first round and 15 were retained throughout the study as one panellist could not meet with
the demands of the study, whilst the other passed away during the course of the study, but had
sent through his opinions for the first round. Therefore, the Delphi Study involved invited
panellists and it retained 15 active members. This number of panellists was considered
275
adequate based on literature recommendations from scholars which have employed the
technique previously. For instance, Delbecq et al. (1975) suggest that 10 to 15 panellists could
be sufficient if the background of the panellists is homogenous, which was achieved in the
current study. Also, a critical review of literature by Rowe and Wright (1999) indicates that the
size of a Delphi panel ranges from three to 80 in peer reviewed studies. Okoli and Pawlowski
(2004) Skulmoski, Krahn and Hartman (2007) also indicated a panel size of about 10 to 18
members. Whilst Hallowell and Gambatese (2010), suggests that since most studies
incorporate between eight and 16 panellists, a minimum of eight is suggested, which was
surpassed in the current study. Further, Hallowell and Gambatese (2010) argue that the size of
a panel should be dictated by the study characteristics, number of available experts, the desired
geographical representation and capacity of the facilitator. Based on the above and the fact
that the Delphi Method does not depend on statistical power, but rather on group dynamics for
arriving at consensus amongst experts, the panel of 15 experts was considered adequate.
The Delphi Method is a very rigorous and time consuming process and this could have be the
reason while most of the potential experts that had consented to participate, fall out at the
introductory stage when they learnt of their obligations. A drop of only two members from the
first round with 15 eventually completing the study was also verification of the quality of the
study and its engaging nature in the present South African housing space. All (100%) panel
members were from South Africa. Two (2) are currently residing at the Nelson Mandela Bay
Metropolitan Municipality; seven (7) reside in the City of Johannesburg, four (4) in Ekurhuleni,
one (1) in Tshwane and another (1) in the City of Cape Town (Table 7.3). Also 80% of experts
were male, while 20% were female. Although none of the panel members reside in Mogale
City which is a district municipality, however, it was still considered a survey area for the
study. This is because Mogale City (West Rand) is one of the prime areas of the greater
Johannesburg which also includes the Ekurhuleni Metropolitan Municipality.
Table 7.3: Residential location of experts
South Africa
Metropolitan Municipality
Number of experts
City of Johannesburg
Ekurhuleni
Nelson Mandela Bay
Tshwane
City of Cape Town
7
4
2
1
1
Total 15
276
The highest qualifications held by the experts are tabulated in the Table 7.4 below. Three of
the experts had a Doctor of Philosophy (PhD) degree, ten experts had a Master of Science
(MSc) Degree and two had a Bachelor of Science or an equivalent degree. All experts were
from various fields, ranging from urban and spatial planning, housing studies, urban and social
policy amongst others (Table 7.4), but from their curriculum vitae analysis, they are all
involved with low-income housing issues.
Table 7.4: Qualification of expert’s panelist
Highest qualification Number of experts
Doctor of Philosophy (PhD)
Master of Science degree (MSc)
Bachelors of Science or equivalent degree
3
10
2
Total 15
In terms of their current occupation, eight (8) of the selected experts were employed by
universities, one (1) works as a housing practitioner with the City of Johannesburg, four (4)
were employed by the government, one (1) by an NGO and another one (1) by a research
institution. All expert panellists held very senior positions in their organizations and were
involved in low-income housing issues at different levels.
Table 7.5: Expert’s panelist field of specialization
Field of specialization Number of experts
Urban & Spatial Planning
Housing
Urban & Social Policy
Development studies
Informal Settlement
Sociology and Urban Planning
Construction Management
Building Construction
Community Development
Strategy and Organisation dynamics
Education and Training
1
1
4
1
1
1
2
1
1
1
1
Total 15
Three (3) had worked as consultants for the government in the development of low-income
housing policy, five (5) had undertaken various kinds of research to improve the quality and
277
spatial planning of the low-income housing for the Department of Human Settlement, one (1)
currently serves as a board member in the Ekurhuleni Department of Human Settlement whilst
another (1) is a reference group member of the Integrated Development Planning and
Modelling project (IPDM) and a Steering Committee member for the Gauteng Economic
Development Agency. The panel consisted of two (2) professors, two (2) chief executive
offices, three (3) executive directors in the Department of Human Settlement in three different
municipalities and one (1) worked as an educational trainer, focusing on housing (Table 7.5).
Table 7.6: Expert’s panelist years of experience
Years of experience Number of experts
1- 5
6-10
11-20
21-30
Over 31 years
1
2
4
6
2
Mean 18.93
Median 21
Mode
Range
15
28
Cumulative total years of experience 284
From Table 7.6, one expert panellist had 1-5 years of experience, two had 6-10 years of
experience, four had 11-20 years of experience, and six had 21-30 years of experience and two
had above 31 years of experience. All experts were professionally registered at the highest
level with various professional regulating bodies, such as the South African Council for
Construction Project Managers, International Sociological Association, etc. Two were certified
Geographic Information Systems Professionals, ten were registered with the South African
Planning Institute, two the with American Planning Association, one with the International
Development Network, two others with the Royal Town Planning Institute, five with
Anthropology Southern Africa, one with South African Foundation for Public Management
and Development, three with the Institute for Housing of South Africa, and one with the South
Africa Architectural Profession.
Table 7.7: Expert’s panellist publication history
Panel publications No. of publications
Books and monographs 22
Chapters in books 66
278
Peer reviewed Journals 295
Peer reviewed Conference proceedings 365
Funded research 66
Other publications 341
Editorial board membership 32
Referee for journals 32
Referee for Conference proceedings 40
Official reports and Policy papers 14
In terms of publications, all expert panellists had published in peer reviewed journals,
conferences and book chapters. Amongst them, they had published 22 books and monographs,
66 book chapters, 295 peer reviewed academic journals, 365 recent conference papers and 341
other publications comprising of articles in professional journals, 14 official reports, and 8
policy papers (Table 7.7). Between them, they had led and managed 66 funded research
projects. Eight of the expert panellists’s serve on editorial boards of 32 peer reviewed journals
and five on 40 peer reviewed conference proceedings. The Figure 7.5 below shows the
contributions of the panellists to the above mentioned publications. Five of the panellist’s had
published books and monographs, six had published chapters in books, fifteen had published
articles in peer reviewed journals and peer reviewed conference proceedings, eleven of them
had led and managed funded research projects and fourteen of the experts had published an
article in other publications such as professional journals and official reports.
Figure 7.5: Expert’s Panel Contribution to the above mentioned publications
5
6
15
15
11
11
5
5
8
3
0 2 4 6 8 10 12 14 16
Books and monographs
Chapters in books
Peer reviewed journals
Peer reviewed conference proceedings
Funded research
Other publications
Editorial board membership
Referee for journals
Referee for conference proceedings
Official reports and policy papers
Number of contributors
Pu
bli
cati
on
279
Five of the experts served on editorial boards for journals, eight had served as a referee for
conference proceedings and three had been appointed as referee or reviewer for journal
publications, while three had also produced official documents for the Department of Human
Settlement.
Phase 3 – Determining the Panel Size
Since the nature of the Delphi Technique calls for a qualitative rather a quantitative approach,
the use of experts for input indicates that the number of participants should be expected to be
much lower than normal quantitative surveys. Determining the minimum number of experts to
participate in a typical Delphi Survey has been a subject of debate over time. Various scholars
have recommended different sample sizes. For instance, Dalkey and Helmer (1963) used a
panel of seven experts in their original Delphi experiment in 1953 (Helmer, 1983). Linstone
(1978:296) finds that “a suitable minimum panel size is seven”. Linstone (1978) justified this
by saying that the research runs the risk of accuracy deteriorating rapidly as numbers increase.
Hence, Linstone’s (1978) observation was supported by Cavalli-Sforza and Ortolano
(1984:325) who postulated that a “typical Delphi panel has about eight to twelve members”,
while Phillips (2000:193) also stating that the optimum number of participation should be
between seven and twelve members both citing the same reason as Linstone (1978). Miller
(1993) assumes that beyond the first thirty responses, additional responses do not generate
much new information. Similarly, Dunn (1994) suggest ten to thirty participants, apprising that
as the complexity of the policy issue increases, the sample size needs to be larger to include
the entire range of participants both for and against the policy issue area. Dunn (1994) further
emphasises that the type of participants that should be selected should include both formal and
informal stakeholders who have a vested interest in the policy issue. These participants
according to Dunn (1994) should have varying degrees of influence, hold a variety of positions,
and be affiliated with different groups; which was the premise that the present Delphi Study
was also based upon.
According to Andranovich (1995), if the group of experts is fairly homogeneous (sharing
similar opinions) then ten to fifteen panellists will be enough and if there are diverse interests
present amongst the experts, then the size of the group will need to be increased to ensure
balance (Zami & Lee, 2009). However, for most community-oriented Delphi Studies, thirty
panellist is about as large a group as the panel should be since the Delphi Technique is a labour
intensive procedure; the greater the number of panellists, the greater the information load (Zami
280
& Lee, 2009). Skulmoski, Hartman and Krahn (2007) highlighted a number of factors, which
should be considered in order to determine the sample size for a Delphi Technique:
1. Heterogeneous or homogeneous sample: where the group is homogeneous, then a
smaller sample of between ten to fifteen people should yield sufficient results.
Nevertheless, if an unrelated group is involved, for instance in an international study,
then a larger sample will likely be required and several hundred people might
participate (Delbeq et al., 1975). However, the researcher needs to exercise caution
because heterogeneous groups can greatly increase the complexity and difficulty of
collecting data, reaching consensus, conducting analysis, and verifying results.
2. Decision quality/Delphi manageability trade off: there is a reduction in group error (or
an increase in decision quality) as sample size increases (Linstone & Turoff, 2002).
However, above a certain threshold, managing the Delphi Process and analysing the
data becomes cumbersome in return for marginal benefits.
3. Internal or external verification: the larger the group, the more credibly the results can
be said to be verified. However, a smaller sample might be used, with result verification
conducted through follow-up research. The current research adopted a smaller sample
premise and will be verified through a follow up questionnaire survey.
Nonetheless, with the Delphi Study, the selection of an initial respondent panel is variable.
However, from the reviewed literature, it was concluded that a typical sample size varies
between seven to fifty panellist, as there is no agreement on the desired ‘typical’ number of
panellist to be adopted in a Delphi Studies; rather that the method is modified to suit the
circumstances and the research question. In the current research, it is not possible to involve
large numbers of participants in the Delphi Survey because of time constraints and the
conflicting schedule of the experts. Hence, a sample size of 15 expert panellists was adopted
based on the following premise, in conjunction with the qualifying criteria’s as established in
phase two of the Delphi Study:
1. Experts should be fairly and practically split between academics and practitioners. The
two categories may provide input for various perspectives and balance the theoretical
and practical considerations.
2. Panellist in both categories should have extensive experience relating to general
housing studies, low-income housing, or in other sustainable development or human
settlement development contexts.
281
Regardless of the above criteria, the current study also adopted Rowe’s, et al. (1991)
recommendations that the selected participants should represent a wide variety of backgrounds
to guarantee a wide base of knowledge and experience. Also, the number of panellists depends
on the topic area, as well as the time and resources at the researchers’ disposal. The adopted
experts’ number of 15 experts seems appropriate, given the amount of data and subsequent
analyses that each panellist generates.
Phase 4 – Conducting the Delphi iterations
Data collection through Delphi
Sequences of questionnaire rounds are used to obtain iterative responses to issues in a Delphi
Study (Masser & Foley, 1987). For instance, Woudenberg (1991) proposes two or ten rounds
as appropriate numbers of rounds supporting that accuracy is expected to increase over rounds,
because of the repetition of judgment and group pressure for conformity. Likewise, Critcher
and Gladstone (1998) suggest between two and five rounds. The Delphi method used in this
study involved three rounds of iterative process, with the view of achieving consensus between
the panel members on the influence and impact of residential satisfaction characteristic of
occupants of low-income housing. Further issues relating to housing in South Africa, which
are associated with low-income housing, were also asked. A Delphi questionnaire, attached in
Appendix D, was sent out electronically to all panel members who were then asked to take the
time and respond to the questions, according to their ability and expertise. The Delphi
Questionnaire was developed based on the findings from the literature review and was
specifically designed to address and achieve the Delphi specific objectives defined for the
study.
The Delphi Study for the current research consists of three rounds. On average, each round
took about a month to complete. A questionnaire was designed for each round based the on
responses to the previous one. However, the Round One Questionnaire was designed, based on
a summary of the comprehensive review of literature highlighting sets of attributes and sub
attributes that are potentially relevant to residential satisfaction decisions by the occupants of
low-income housing (See Appendix D). Additionally, issues relating to provision of low-
income housing, delivery and sustainability beyond the current level were also extracted from
the reviewed literature. These were structurally and constructively put together to frame the
first round of the Delphi survey. Therefore, Round One of the Delphi Study was intended to be
a brainstorming exercise used to produce a list of empirical attributes that determine residential
282
satisfaction and other issues relating to low-income housing in South Africa and other
subsequent issues relating to the objective of the Delphi Study. Closed and Open-ended
questions were used in this round. Thereafter, these were analysed and formed the basis of
Round Two (See Appendix E) and Round Three of the study. Frequencies were obtained to
measure the degree of consensus reached amongst participants regarding the attributes that
determine residential satisfaction in South African low-income housing and for other related
questions. Also, a content analysis methodology was adopted to analyse responses to the open-
ended questions to “minimize redundancy” (Rubin et al., 1998:6).
The purpose of the second round of the study was to allow experts to review and comment on
the attributes that determine residential satisfaction and other issues relating to low-income
housing in South Africa, which were proposed by the expert participants in Round One. Closed
questions were used in this round to investigate participant comments expressing agreement,
disagreement or clarification concerning proposed attributes that determine residential
satisfaction in South Africa. The specific nature of the closed-ended questions stimulated
participants’ reactions. Frequencies were likewise obtained to measure the degree of consensus
reached amongst participants regarding the attributes that determine residential satisfaction and
for other related questions. Also, a content analysis approach was adopted to analyse responses
to the open-ended questions.
The final Round Three (Appendix F), was specifically aimed at:
1. informing the experts of the findings of the analysis of responses to the questionnaire
of Round Two; and
2. requesting their final affirmation / comments on attributes and issues that did not
receive any consensus in Round Two. The Round Three Questionnaire was designed
based on the findings of content analysis and measures of frequency responses to the
questionnaire of Round Two. Closed-ended questions were used and frequencies were
obtained to indicate consensus reached amongst experts regarding attributes that
determine residential satisfaction and other low-income housing issues, as presented in
the study.
Over the three round Delphi Survey, consensus was reached regarding most of the attributes
that determine residential satisfaction and other low-income housing related issues in South
Africa. Based on the findings of the analyses of responses to the Delphi rounds, a list of
283
attributes that determine residential satisfaction was prepared, which informs the conceptual
framework for the broader study, while issues surrounding low-income housing in South Africa
were highlighted, which responded accordingly to the set objective of the Delphi Study. The
Delphi Survey was conducted via electronic mail, and follow-up emails were used to encourage
prompt responses to the questionnaires. Using email provides a free and faster means of
communication.
With regards to the Delphi Questionnaire, experts panellist were requested to rate the likelihood
of an attribute influencing housing satisfaction; the impact of sub factors in predicting
residential satisfaction of the low-income group in South Africa, if they were present. The
probability scale ranged from one to ten representing zero to 100%. Interval ranges were set at
ten (Table 7.8). Furthermore, experts were asked to rate the negative impact that would result
if a particular residential satisfaction attribute were also absent. This was based on a 10 point
ordinal scale ranging from negligible to very high impact. This aspect indicated the importance
of the residential satisfaction as shown on Table 7.9 below.
Table 7.8: Influence or likelihood scale
0-10% 11-20% 21-30% 31-40% 41-50% 51-60% 61-70% 71-80% 81-90% 91-
100%
1 2 3 4 5 6 7 8 9 10
X X
Table 7.9: Impact scale
No impact
Low impact Medium impact High impact Very high impact
1 2 3 4 5 6 7 8 9 10
X
Panel experts were also required to state their level of agreement using a 5-point Likert Scale
with certain statements and to support their choices where necessary, with regards to the South
African housing subsidy; policy issues and the future of public low-income housing in order to
arrive at a consensus.
Responses were received for each round of the Delphi Study. As mentioned above, group
medians were calculated for each response on each element. The group median was deemed
appropriate, as mentioned earlier, as a measure of the central tendency. Thus an indicator of
whether consensus had been reached on the questions for each element was determined. The
284
median was deemed to be more suitable for the type of information that was being collected.
This is because the median eliminates bias and takes into consideration outlier responses. It
makes the consensus notion more reasonable. The mean on the other hand, may not reflect a
reasonable central tendency as it considers only the outlier responses. Group medians from the
Delphi First Round were computed for each element. These were then sent back to the expert
panel members so that responses in the second round could be made taking into account the
group median. Expert panel members were asked in the second round to either maintain their
original response made in Round One, or they could change their initial response to be more in
line with the group median.
Figure 7.6: Outline of Delphi Process
Source: Thangaratinam & Redman (2005:124)
In the Second Round, panel members who had responses to units above or below the group
median were requested to state their reasons for sticking to a response that does not agree with
the group median. In the Third Round, panellists were again requested to consider reasons for
the outliers in making their decisions. Group medians and the absolute deviations were again
computed for the third round. From these calculations and after three rounds of the Delphi
Round 1
Q1
Q
Round 2
Q2
Round 3
Q3
Researcher actions Expert panel actions
Circulates questionnaire to
expert panellists
Compute group medians and
re-circulate questionnaires
Rates probable housing satisfaction
due to factors a influence / impact and
other issues
Reviews individual rating in view of
the group’s median. Gives reason if
required
Re-compute group medians,
standard deviations &
compile comments
Experts given opportunity to
reconsider initial rating
Determine consensus and
terminate process
285
Process, it was determined that consensus had been reached. Reasons for other experts who
stuck to their ratings have however been taken into consideration. After the third round, group
medians and the absolute deviations were again computed for the third round.
Calculations for the third round of the Delphi Process indicated that there was no need to
proceed to the fourth round as there was no further value that could be added to the degree of
consensus attained at that level. Throughout the entire Delphi Process, anonymity of panel
members was maintained to avoid undue influence on other members. The aspect of anonymity
is crucial to the credibility of the Delphi Technique as a rigorous establishment of the complex
‘what would happen if’ kind of questions that ideally should be established from an
experimental study but in-fact extremely difficult and not feasible to do. Figure 7.6 shows an
outline of how the Delphi Study data was collected.
7.4.3.5 Specific Objectives of the Delphi
The literature states that there are various characteristics and factors which determine
residential satisfaction as measured from different housing typologies, and categories, such as
private low-income and owner occupied low-income housing. What was however not clear in
the literature was specifically the level or extent the identified factors contribute to the housing
satisfaction of the occupants of a subsidised low-income house. Attempts have been made
through various studies to determine the influence of these factors in a low-income housing
situation, but none has been specifically related to subsidised low-income housing. However,
it was noted that various factors determine residential satisfaction, albeit, varied in different
cultural and housing backgrounds and typological settings. Also, previous models have not
been adequately organized into a model to form a holistic attribute which determines residential
satisfaction.
Based on the fore-going, a more reliable measure of the determinants of residential satisfaction
was therefore necessary in order to establish not only whether these factors have influence on
housing satisfaction but also the extent to which or level of their influence and to identify which
factors have the greatest influence in the South African context together with the identified gap
of factors from the literature. Based on the context of the thesis, this kind of investigation would
ordinarily call for an experimental kind of research. However, the experimental method of
research was not feasible and practical considering the time frame, ethical issues and the
286
willingness of would be participants. Hence, the Delphi Method was considered the most
suitable method to achieve the general objective of determining the influence and impact of the
identified factors on low-income occupants’ residential satisfaction.
The broader research aim was to develop a resident satisfaction model on publicly subsidised
housing in developing countries: a case study of South Africa. The Delphi Method was
therefore chosen at the first stage to formulate the conceptual model. This had to be validated
later from responses obtained from the questionnaire survey analysed using the structural
equation modeling software EQS Version 6.2.
At the Delphi Stage, factors that were identified from literature that defined and determined
residential satisfaction were formulated into questions, which the experts rated as being
influential or had an impact on residents’ satisfaction. Likewise, the output from the Delphi
Process was a set of attributes which determine residents’ satisfaction that should be
implemented and given consideration in order to achieve better resident satisfaction for all
levels of government provided housing in South Africa. This is because a number of studies
have identified and shown that different attributes determine residential satisfaction. For
instance, housing characteristics, neighbourhood characteristics, and household characteristics
have been observed as the essential determinants of residential satisfaction (Amerigo &
Aragones, 1997; Galster & Hesser, 1981; Lu, 1999). Further observations included the age of
houses (He, 2009), interior and proximal exterior environments (Phillips, Siu, & Yeh, 2005),
and other aspects of housing, such as, building quality and disrepair (Amerigo & Aragones,
1990; Paris & Kangari 2005). Sirgy and Cornwell (2002) similarly identified neighbourhood,
social, economic, and physical features as the major determinants of residential satisfaction.
Further, Marans and Sprecklemeyer (1981) assert that residential satisfaction is a result of an
integrated relationship between environment and the human perception of beliefs. In addition,
Marans and Rodgers’s (1975) model of residential satisfaction determines that overall
residential satisfaction levels are related to an occupant’s own characteristics, such as social
class, housing status etc. The current study extends the above mentioned studies by looking at
residential satisfaction holistically with the addition of new constructs (occupants’ participation
and the assessment of their needs and expectation) to develop a model that will predict
residents’ satisfaction in low-income housing and to what extent.
In formulating questions for the Delphi Study, the following questions where thus raised
287
considering that there are various attributes that determine residential satisfaction; because
meaningful improvement in low-income housing satisfaction in South Africa could be
achieved only by determining the attributes that brings about housing satisfaction. The
objectives of conducting the Delphi Survey for this study were to determine the following:
DSO1 To identify the attributes (main and sub) that determine residential satisfaction
and to examine if the attribute that determines satisfaction in other cultural
contexts is the same in South Africa;
DS02 To determine the factors that makes subsidised public housing unsustainable in
South Africa;
DS03 To identify the combination of housing policy instruments that will better serve
South African low-income housing groups;
DS04 To identify the critical factors affecting the delivery of low-income housing and
their effects on beneficiaries’ residential satisfaction;
DS05 To predict the life span of the present South African public housing subsidy
delivery model;
DS06 To evaluate the management issues affecting the national, provincial and local
government housing agencies in the delivery of housing in South Africa;
DS07 To determine the influence of beneficiary participation on the overall housing
satisfaction and;
DS08 To determine the effect of meeting beneficiary’s housing needs and
expectations on their overall housing satisfaction.
The philosophy behind the above objectives is to do away with the tendency of a non-coherent
dialogue on low-income housing in South Africa. Therefore, achieving the above objectives
resulted in the following outcomes:
1. Determining the key factors and constructs that are of critical significance (influence)
to determining residential satisfaction in subsidised low-income housing;
2. A holistic conceptual model on residential satisfaction in South African low-income
housing.
288
7.4.3.6 Computation of Data from Delphi Study
Computation of data from the Delphi Study was conducted using Microsoft Office Excel, a
spreadsheet software programme. The first stage involved analysis to determine consensus on
responses to the predetermined criteria. This involved determining the group median responses
for each question. After the third round of the Delphi, absolute deviations ( iD ) of the group
medians ( Xm ) of each rating for the relevant questions as pre-determined were calculated
using Equation 1.0 below.
Equation 1. 0
Where:
tendencycentral of Measure
rating Panellist
deviation Absolute
Xm
x
D
i
i
A computation of each and every question element was completed for the likelihood and impact
of the attributes in predicting residents’ satisfaction and improvement of the low-income
housing context in South Africa. Also, the influence or impact of the absence or presence of a
particular residential satisfaction element on the overall residential satisfaction of the other
elements is presented, likewise on issues relating to low-income housing in South Africa.
Additionally, for every round of responses from the experts, besides the group median value
computation, their respective interquartile deviation (IQD) were also computed as a measure
of the central tendencies to determine consensus.
The median value was adopted as a measure of central tendency because of its effect to
minimize the effects of potentially biased individuals. While the IQD scores were used to
summarize the variability in the data. The IQD helped to identify which measure were most
appropriate to influence residents’ satisfaction. Also, through the use of the IQD, a clearer
picture of the overall dataset was provided as the IQD removes / ignores outlying values. The
inter-quartile range is a measure that indicates the extent to which the central 50% of values
within the dataset are dispersed. However, it is based upon, and related to, the median. Though,
Xmi
xi
D
289
for studies of this nature, quite often the median is adopted as a measure of central tendency,
as opposed to the mean and IQD, although it is sometimes used. To compute the variation of
the median from the responses for each question in each round, the absolute deviation given in
Equation 1.0, was done. This is the absolute difference between a response within a data set
and a given point. In this case, the point from which the deviation is measured is a measure of
central tendency, which is the median. Results from the Delphi analysis will be presented as
numbers and percentages in tables, columns and bar charts in the results Section 8.3. The
outcomes show the predictions of the influence of residential satisfaction factors and other
housing issues surrounding low-income housing in South Africa.
7.4.3.7 Determination of Consensus from the Delphi Process
It was imperative that consensus should be reached on all questions. Consensus was determined
by measuring the central tendency of the various responses on all questions. As mentioned
earlier, the group median and the IQD were computed for all responses. In order to achieve
consensus, the deviation of all responses about the group median was determined not to be
more than one (1) unit, likewise for the IQD. This is considered to be suitable as the scale that
was used for both probability (influence) and impact was 1 to 10. The deviation of all responses
was calculated using the absolute median (Equation 1.0), while the IQD was calculated based
on the recommended statistical process of the absolute value of the difference between the 75th
and 25th percentiles. A percentile (or centile) is the value of a variable below, which a
certain percent of observations fall. The percentile is often used in the reporting of scores
from norm-referenced tests, as in the present situation. The 25th percentile is also known as
the first quartile (Q1), the 50th percentile as the median or second quartile (Q2), and the 75th
percentile as the third quartile (Q3). Hence, the deviation between the 75th and 25th percentiles
give an absolute value referred to as the interquartile range or deviation. The interquartile range
deviation is a robust statistic, having a breakdown point of 25%, and is thus often preferred to
the total range, with smaller values indicating higher degrees of agreement (consensus).
Smaller values in the inter-quartile range would then indicate higher degrees of consensus.
However, consensus is difficulty to measure in Delphi Studies. The foregoing has been
established from literature, that actually there is no consensus on how to determine consensus
regarding a set of opinions. Holey et al. (2007:2) suggested that consensus is the same as
agreement and that agreement can be determined by the following:
290
1. The aggregate of judgments;
2. A move to a subjective level of central tendency; or
3. Alternatively by confirming stability in responses with the consistency of answers
between successive rounds of the study.
Other researchers have used frequency distribution to assess agreement and the criterion of at
least 51% responding to any given response category being used to determine consensus
(McKenna, 1994). Other studies, such as one conducted by Rayens and Hahn (2000), have used
means and standard deviations with a decrease in standard deviations between rounds
indicating an increase in agreement. Likewise, inter-quartile deviation (IQD) has also been
used to determine consensus (Rayens & Hahn, 2000), which has also been adopted for the
present study. In their study, Rayens and Hahn (2000) included another criterion to determine
consensus in addition to the IQD in order to achieve stability. The criterion to achieve
consensus was that the IQD should equal one (1) unit for which more than 60% of respondents
should have answered either generally positive or generally negative. Items which had an IQD
≠1 for which the percentage of generally positive or generally negative responses was between
40% and 60% were determined to indicate a lack of consensus or agreement. Also, Raskin
(1994) identified an IQD of 1.00 or less as an indicator of consensus. Spinelli (1983) considered
a change of more than 1.00 IQD point in each successive stage as the criterion for convergence
of opinion. Clearly, there is no consensus in the literature about how to use or interpret IQD as
a method of data analysis for the Delphi process. The potential range of IQD values depends
on the number of response choices, with larger IQDs expected as the number of response
choices increases. Thus, the use of a particular IQD as a cut-off for consensus requires
consideration of the number of response choices. Furthermore, Holey et al. (2007) used the
following criteria to determine consensus:
1. Percentage response;
2. Percentages for each level of agreement for each question to compensate for varying
response rates;
3. Computation of median, standard deviation and their associated group rankings;
4. Computation of the means, standard deviation and their associated group rankings using
the importance ratings; and
5. Computation of the Weighted Kappa (k) values to compare the chance eliminated
agreement between rounds.
291
According to Holey et al. (2007), consensus is reached when the following is present:
1. An increase in percentage agreements;
2. Convergence of importance rakings;
3. Increase in Kappa values;
4. A decrease in comments as rounds progressed;
5. A smaller range of responses; and
6. Smaller values of standard deviations.
The studies above suggest that there is little agreement on how to measure consensus in a
Delphi Study. It is however agreeable that for consensus to have been achieved, there has to be
a convergence of ideas and reasoning towards a subjective central tendency measure. Hence,
in the current study, consensus was determined to have been reached if the following was
achieved:
1. More than 60% of responses are generally positive or negative with certain questions;
2. The average of the absolute deviation was not more than one unit. The absolute
deviation is calculated from Equation 1.0., and
3. The IQD was less than 1.00. Meaning that items with IQD = 0.00 were considered to
have reflected high consensus.
Therefore the scales of consensus adapted for this research are:
1. Strong consensus - median 9-10, mean 8-10, interquartile deviation (IQD) ≤1 and
≥80% (8-10);
2. Good consensus - median 7-8.99, mean 6-7.99, IQD≥1.1≤2 and ≥60%≤79% (6-7.99);
and
3. Weak consensus - median ≤ 6.99, mean ≤5.99 and IQD≥2.1≤3 and ≤ 59% (5.99).
7.4.3.8 Reliability and Validity of the Delphi Method
According to Els and Delarey (2006:52), reliability is the extent to which a procedure produces
similar results under constant conditions at all times. However in a Delphi Study, this kind of
statistical reliability is not possible because another panel may reach a different conclusion
depending on their knowledge of the subject area and interest. To ensure reliability therefore,
care was taken that credibility showed in truthfulness, fittingness exhibited in applicability,
audit ability shown in response consistency and conformability was exhibited in the responses
from all participants. Credibility was also assured during the selection of the panel. All
292
panellists had distinguished themselves based on the set criteria for the selection of expert
panellist and the depth of their knowledge and experience as presented in Section 7.4.3.8.
Validity was boosted by the removal of preconception or influence from other members by
keeping all members completely anonymous from each other and hence, eliminating the
‘bandwagon’ effect, which is one of the strengths of the Delphi Method. Furthermore, the
number of iterations that were implemented in the Delphi Study also enhanced the internal
validity. Thus, expert panellists were given a chance to change their opinion or maintain it with
a written explanation or argument for dissenting views. Feedback to the researcher and constant
email communication between the researcher and the panellists individually was another way
of ensuring internal validity of the study.
The external validity of a study deals with the extent the results from the study can be
generalised to a larger population. This is usually determined by how participants are selected
to be part of the study. This process was however not necessary as the validation process of the
conceptual model has been done using the Questionnaire Survey. Nevertheless, the selection
of participants for the Delphi Study guaranteed external validity as scientific criteria as pre-
determined (Section 7.4.3.3) based on previous scholarly works were adopted. The panel
comprised of members from varied sectors, all with in-depth knowledge on sustainable human
development. All reside in South Africa and in one of the metropolitan / district municipality,
and all members were highly experienced, with a sound publication history. The study therefore
fulfilled requirements for external validity in line with standard research ethics.
7.4.4 Questionnaire Survey
Phase Three of the research involved collecting data from the field through the use of
questionnaires in order to meet the general objectives RO5 and RO6 of the overall research
respectively. Phase Three formed the pinnacle of the research. The Delphi Study findings were
that:
Residential satisfaction is a multidimensional construct, which consists of the occupants’
satisfaction with the dwelling unit features, neighbourhood features, building quality, services
provided by government, beneficiary participation and needs, and expectations. These factors
have been collectively considered for the development of a holistic integrated residential
satisfaction model in this study. Four of the factors have been previously considered in the
293
development of residential satisfaction model in other cultural contexts, but none of the existing
models to date have included both beneficiary participation and needs and expectations as
factors to develop a model to assist housing authorities in the construction of houses that will
be satisfactory to the poor and low-income group.
Therefore, in order to validate findings from the Delphi Study, the specific objectives of the
questionnaire survey were to:
QS1 identify the factors that had a higher influence on low-income housing
occupants’ residential satisfaction;
QS2 establish the influence of the identified factors on occupants’ residential
satisfaction;
QS3 determine the influence of the overall residents’ satisfaction on subsidised low-
income occupants’ behaviour; and
QS4 determine the goodness-of-fit of the hypothesised integrated holistic residents’
satisfaction model to the sample data.
Given that the previous models of residential satisfaction established in the developed countries
cannot be relied on in developing countries, and the findings of what determines residential
satisfaction in developing countries are rarely known from the previously conducted research,
the lack of research into the overall impact and influence of the direct and holistic active
involvement of residential satisfaction constructs, and the absence of a residential satisfaction
model in subsidised low-income housing, the achievement of occupants’ residential
satisfaction is unlikely.
The integrated conceptual model shown in Figure 9.1 (Model 1.0; page 428) was made up of
the following interrelationships:
1. Dwelling unit features have an impact on residential satisfaction and greatly influence
its determination;
2. Neighbourhood features have an impact on residential satisfaction and greatly influence
its determination;
3. Building quality has an impact on residential satisfaction and greatly influence its
determination;
294
4. Services provided by the government have an impact of residential satisfaction and
greatly influence its determination;
5. Beneficiary participation has an impact on residential satisfaction and greatly influence
its determination;
6. Beneficiary needs and expectations have an impact of residential satisfaction and
greatly influence its determination; and
7. The integrated holistic residential satisfaction model describes the determinant
(constructs) which determines residents’ housing satisfaction.
As a result of the objectives of the study, it was obvious that collecting facts would involve the
actual beneficiaries; hence a field survey was considered the most suitable method of collecting
the required data. Therefore, a complete biographical section was required that related to the
housing beneficiary’s, the presence or absence of physical features in the housing units such as
bedrooms(s), living rooms etc. were also required. Likewise, the presence or absence of
services and facilities in the housing units and in the neighbourhood were also required.
Similarly, data relating to the beneficiaries’ extent of satisfaction with the dwelling unit
features, neighbourhood features, building quality, services provided by the government,
beneficiary participation, beneficiary needs and the overall satisfaction of the beneficiaries was
required. These types of information could not be obtained through other means of data
collection except a field survey. This is because other means of data collection would not give
an adequate representation of the above stated relationships. Also, the decision to choose a
survey method was based on a number of factors, which included the sampling technique to be
adopted, the type of population, the question form, the question content, the response rate, the
costs, and the duration of data collection. The most appropriate survey method for this research
was a personally administered one.
This method was chosen for the following reasons:
1. A list of low-income housing suburbs could easily be obtained from the municipalities
and from the Affordable Land & Housing Data Centre suburb profile;
2. The questions could be answered by circling the proper response format and with an
interviewer present; respondents could seek clarity on any question as to meet
consistent question objectives (Aaker et al., 2009; Sekaran, 2000);
3. The respondents are more motivated to respond, as they are not obliged to admit their
confusion or ignorance to the interviewer (Burns & Bush 2002; Sekaran, 2000);
295
4. A higher response rate of almost 100% can be assured since the questionnaires are not
left with the respondents, but collected immediately once they are completed (Malhotra
1999; Sekaran, 2000); and
5. Higher anonymity of respondents because respondents were not required to disclose
their identities (Burns & Bush 2002; Sekaran, 2000).
However, this method of survey can be very time consuming, if a wide geographic region is
involved. However for this survey, the respondents were all in the Gauteng Province, hence
data was collected from low-income housing suburbs, which have been identified.
Apart from the above reason for adopting a personally administered Questionnaire Survey
Method, the following reasons justify the use of the Questionnaire Survey Method:
1. The philosophy underpinning the research is based on the positivist theory as discussed
above, which uses quantitative methods and collection of data by use of questionnaires;
2. Validation of the conceptual model developed at Phase One (literature review) and Two
(Delphi) entailed using an alternative method to the previous two to validate the
findings. This therefore eliminated the use of methods similar to the Delphi and its
derivatives and hence, called for collection of data on the subsidised housing occupants
through a questionnaire survey;
3. The field survey was considered to be more descriptive in that there are many
subsidised low-income housing locations in South Africa, but are restricted in some
sense to the Gauteng Province as it is the economic hub of the nation;
4. Likewise, the interpretation and presentation of the data can easily be done and
understood by various readers when adopting a positivist philosophy of research as it
follows a logical explanation of the method;
5. A large number of research questions can be asked in a questionnaire to target many
respondents within a stipulated time frame;
6. A questionnaire requires minimal investment and is relatively easy to obtain
generalizations from (Bell, 1996);
7. Specific information about views, attitudes and perceptions of a group of respondents,
which are difficult to measure using observational technique, can be easily elicited via
a questionnaire (McIntyre, 1999; Yuen, 2007);
8. Another reason for choosing a questionnaire is because the data collected through a
questionnaire can be analysed easily; and
296
9. Data entry and analysis for the questionnaire can be easily done using computer
software packages, such as the SPSS and EQS (Bell, 1996; Hishamuddin, 2007; Yuen,
2007).
7.4.4.1 Questionnaire Survey Instrument
The questionnaire is defined as a method used to gather information related to the opinions of
a large group of people (Pinsonneault & Kraemer, 1993). A standardised questionnaire exposes
each respondent to same set of questions (Brace, 2008). This study applies a formal
standardised questionnaire in order to achieve the research objectives.
Hence, a well-structured questionnaire was used to collect data during the field survey. The
questionnaire was based on the literature review conducted in the first stage of the research, as
well as the findings from the Delphi Study in stage two. The questionnaire consists of three
sections. The first section was designed to collect general and comprehensive information
about the subsidised low-income housing beneficiary. This information included biographical,
socio-economic and other information, as deemed necessary to meet the research objectives.
Section Two of the questionnaire, included questions on the presence or absence of features in
the dwelling unit, services in the dwelling unit, facilities in the neighbourhood, and the services
as provided by the government.
Further, Section Three of the questionnaire included questions on the beneficiaries’ levels of
housing satisfaction, as related to the dwelling unit features, neighbourhood features, building
quality, services provided by the government, beneficiary participation, beneficiary needs and
expectations, and the overall satisfaction levels of the beneficiaries. This section was meant to
collect information on the extent of the beneficiary’s level of satisfaction or dissatisfaction for
each sub-attribute variable as provided. Further information on what defines residential
satisfaction to the beneficiaries was asked in the same section. In summation, the questionnaire
was designed to assess the influence of the identified constructs on beneficiary’s residential
satisfaction.
The first section had twenty one questions relating to the respondents biographical, socio-
economic information. Section Two had five questions with the respondents rating 26 items on
a present and not present scale, while Section Three had seven questions with the respondents
297
rating 85 items on a Five Point Likert scale regarding the extent to which they are satisfied or
dissatisfied; agreed or disagreed on factors that determine residential satisfaction. The length
of the questionnaire was nine pages long, including the cover letter (See Appendix G). This
was in line with the recommended length of between 5 to 12 pages.
To avoid bias resulting from questionnaire design, the questions were constructed in such a
way that they were direct, short, comprehensible, avoided ambiguity, not vague,
generalizations, not leading, not double barrelled, presumptuous questions, simple and familiar
to the respondents. Instructions of the questionnaire were kept simple with no technical or
specialized words being used, likewise less bias from the interviewer (fieldworker). However,
it was recognised that this type of questionnaire has a few weaknesses in that
1. There is an absence of probing beyond the answer given;
2. Lack of control over who answers the questionnaire; and
3. They can be characterized by a low response rate because of cost.
The above weaknesses were addressed by refining the questions and keeping them relatively
simple but care was taken not to deviate from the objectives of the instrument, keeping the
overall questionnaire within the recommended limits and ensuring that only the right person(s)
completed the questionnaire by constantly communicating with the fieldworkers by the lead
researcher. The absence of further probing is characteristic of this type of questionnaire. This
aspect was not a major concern as the data to be collected was meant to validate the integrated
conceptual model initially developed in the previous phases of the study.
Two types of response formats were chosen: dichotomous, interval and ordinal close-ended
and labelled scales were used. In order to obtain information pertaining to respondents’
demographics, a dichotomous, interval and ordinal close-ended question format was used. This
scale was also used for the measurement of the presence or absence of features in the dwelling
unit and neighbourhood environment. In addition, to obtain the respondents extent of
satisfaction towards the model identified constructs, a labelled scale response format was used.
Apart from the simplicity of administering and coding in further statistical analysis (Burns &
Bush, 2000) labelled scale response format is appropriate for residential satisfaction research,
as it allows the respondent to respond to perception and attitudinal questions to varying degrees
that describe the dimensions being studied (Aaker et al., 2009).
298
For this study, labelled Likert Scales were appropriate to measure the responses. This scale
was adopted based on the following reasons:
1. It yields higher reliability coefficients with fewer items than the scales developed using
other methods (Hayes, 1998);
2. This scale is widely used in residential satisfaction research and has been extensively
tested in both marketing and social science studies (Garland, 1991);
3. It offers a high likelihood of responses that accurately reflect respondent opinion under
study (Burns & Bush, 2002; Zikmund, 2000); and
4. It helps to increase the spread of variance of responses, which in turn provide stronger
measures of association (Aaker et al., 2009; Wong, 1999).
In relation to the number of scale points, there is no clear rule indicating an ideal number.
However, many researchers acknowledge that opinions can be best captured with five to seven
point scales (Aaker et al., 2009; Malhotra, 1999; Sekaran, 2000). In effect, researchers indicate
that a five-point scale is just as good as any other (Malhotra, 1999; Parasuraman, 1991;
Sekaran, 2000). That is, an increase in scale does not improve the reliability of the ratings
(Elmore & Beggs, 1975) and may cause confusion to the respondents (Aaker et al.,
2009; Hair et al., 2003). Thus, a five-point Likert Scale was used in this research.
7.4.4.2 Variables
The research instrument was designed to measure the exogenous variables namely: dwelling
unit features (DUF); neighbourhood features (NDF); building quality features (BQF); services
provided by the government (SPG), beneficiary’s participation (BNP) and needs and
expectations (NAE). These variables were hypothesised to be characterised by indicator
variable, which collectively constituted the questionnaire items apart from the socio-economic
characteristics, as also measured by the questionnaire. Table 7.10 gives a comprehensive
summation of the latent and indicator variables.
299
Table 7.10: Conceptual model indicator variables
Latent Variable Constructs Measurement Variables
Dwelling Unit Features (DUF) Location of bedroom
Number of bedrooms
Size of the bedroom
Location of living room
Location of dining room
Location of kitchen
Size of the kitchen
Size of bathroom(s)
Size of wardrobe/closet
Size of children’s play space
Size of children’s study space
Amount of privacy within the house
Amount of brightness / sunshine in the house
Quality of ventilation in the house
Quality of floor level in the house
Overall appearance of the house
Overall size of the house
Neighbourhood Features (NDF) Location of the dwelling unit in the
neighbourhood
Quality of relationship with neighbours
Quality of landscape in the neighbourhood
Quality of walkways
Ease of access to main roads
Amount of privacy from other neighbours
Quality of street lighting at night
Amount of security in the neighbourhood
Physical condition and appearance of the
neighbourhood
Cleanliness of the neighbourhood
Proximity of house to workplace
Proximity of house to shopping areas
Proximity of house to the nursery school
Proximity of house to the high school
Proximity of house to hospitals/clinics
proximity of house to places of worship
Proximity of house to police services
Proximity of house to parking facilities
Proximity of house to the disabled facilities
Proximity of house to the community hall
Proximity of house to playground / recreational
facilities
Proximity of house to public transportation and
services
Building Quality Features (BQF) External construction quality
Internal construction quality
Water pressure
Wall quality
300
Floor quality
Window quality
Door quality
Internal painting quality
External painting quality
Plumbing quality
The finished quality of the sanitary system
Electrical wiring quality
Electrical fittings quality
Numbers of electrical sockets
Level of socket
Overall unit quality
Services Provided by Government The drainage system
(SPG) Garbage and waste collection
The fire protection services
Electricity supply
Water supply
Telephone service
Safety
How well resident complaints are handled
Housing Department rules and regulations
Enforcement of rules by the Department of
Human Settlement (Housing)
Overall services provided by the government
Beneficiary Participation (BNP) Owners should be consulted about the housing
location
Owners should be consulted about the house
design
Owners should be consulted about the house
construction
Owners should be consulted about the internal
finishes of the house
Owners should be consulted about the external
finishes of the house
Needs And Expectation (NAE) Owners should be told beforehand the type of
house they will receive
Owners should be asked the type of house they
need
Owners expect good quality houses
Our houses should meet our family need
Residential Satisfaction (RS) I am satisfied living here
I am taking proper care of my house
I am taking proper care of my neighbourhood
I am constantly maintaining my house
I am not intending to move to another place in the
future
I like to live in another place like this
I will recommend to my friend to obtain a house
in the same way that I did
301
7.4.4.3 Population
A study population refers to the entire group of items in which the researcher has an interest
(Cooper & Schindler, 2006; Neuman, 2006). The population for this study comprised of
subsidised low-income housing in the Gauteng Province of South Africa. Three metropolitan
Municipality cities and a District Municipality were chosen as the study population because
the overlapping boundaries between urban and the rural areas have made it difficult to define
settlements in South Africa. For instance, the unpublished draft of the South Africa Urban
Developed Framework (UDF) uses three indicators to define various settlement typologies,
such as population size tied up with function, strength of the economy and the institutional
situation in the municipalities in which the space is located (States of the Cities Report, 2006).
The indicators as developed by the UDF were used to define the scope of the present research
population. Therefore, the scale and settlement characteristics of a metropolitan municipality
city in the context of this study, refers to cities with a population over one million, with an
established formal core of industrial, commercial and suburban development, formal
townships, hostels, and backyards. Also, there is the presence of informal settlements with a
significant low-income housing development on the periphery. The metropolitan municipality
cities have a huge economic base, plus a core economic potential, with high concentrations and
absolute numbers of urban poor. Institutionally, the metropolitan municipality government has
a consolidated fragmented municipal history and the urban benefits have not yet seamlessly
applied to all residents of the metropolitan city. Likewise, the chosen district municipality has
similar characteristics, but with a population above 500 000 inhabitants, with a growing
economic base, but not fully developed like the metropolitan municipality. Consequently,
based on this background, the metropolitan municipality cities that were selected are:
Ekurhuleni (East Rand – Gauteng), City of Johannesburg (Johannesburg - Gauteng), and
Tshwane (Pretoria - Gauteng), while the district or category B municipality is Mogale City
(Krugersdorp).
The municipalities were selected as the study population because they are a typical
representation of the South Africa cities urban space. Their significance and relevance to the
low-income housing development is an example of what the progress and development of low-
income housing typology should be in the country. Hence, they were adopted as the study
population. The unit of measurement was the beneficiary’s (occupants’) of the subsidised low-
income housing that have been built, allocated and are being inhabited. The selected low-
302
income housing locations were chosen based on the history of the areas to the South Africa
housing space. The questionnaires were administered to the heads of the households. In cases
where the head of the household was not available, their spouses were chosen. However, when
both were not around or a spouse is not present, another housing unit is then chosen. One
household head or spouse per house was engaged in the questionnaire administration.
By measuring the opinions of the beneficiary’s regarding their satisfaction with the housing
units, it was observed that the survey was related to the beneficiaries’ psychology and not the
dwelling units. This therefore revealed that houses become a home based on what the occupants
made of it. Hence, Ha (2008) states that the failure of many housing projects may be traceable
to the lack of knowledge or opinion on the determinants of housing satisfaction from the
occupants of the houses. This importance is based on the fact that many problems in the existing
low-income housing environment are the result of neglecting the beneficiaries’ views before,
and mostly, after the houses are built. Further, subsidised low-income housing units in the
selected municipalities were determined based on a comprehensive background study of the
characteristics and situational position of low-income housing in each respective municipality.
In addition, the Affordable Land & Housing Data Centre Suburb Profile was used to refine the
chosen low-income housing area selected.
7.4.4.4 Sample Frame
A sampling frame is the source material or device from which a sample is drawn. It is a list of
all those within a population who can be sampled, and may include individuals, households or
institutions. For the current study, all completed, allocated and inhabited subsidized low-
income housing units in each housing location area were chosen as the sample frame. The
sample frame was the beneficiaries (occupants) of the subsidised low-income housing that are
inhabited. To establish the sample frame, a list of the numbers of subsidised low-income
housing in the respective population was obtained from the municipalities and confirmation
from the Affordable Land & Housing Data Centre Suburb Profile, which has a comprehensive
data, capturing of the number of houses in South Africa. Because of the size of the household
in each frame (Table 7.11), the selection of respondents was based on a probability sampling
technique (discussed in the next section). The respondents selected from these households were
instructed as on the cover letter of the questionnaire (See Appendix G) that their responses
should reflect their personal opinions towards the housing unit and the neighbourhood.
303
Table 7.11: Study Population and Sample
Metropolitan Municipality (MM)
Low-income housing
area / suburb
Population of
Low-income
housing units
Number
sampled
% sampled
City of Johannesburg
Braamfischerville 8,768 100 1%
Pennyville 932 93 10%
City of Tshwane
Mamelodi 589 100 17%
Nellmapius 6,587 100 2%
Ekurhuleni
Tsakane Ext 8 873 87 10%
Reiger Park 1, 845 150 8%
Total (MM) 19,398 624 3.20%
District Municipality (DM) Mogale City
Kagiso
Ext 6 (Chief Mogale)
391 40 10%
Extension 8 869 87 10%
Total (DM) 1,260 127 10%
Total (MM + DM)
20,658
751
3.60%
7.4.4.5 Sampling Method
A sample is a subset of items a researcher selects from a specific population (Neuman, 2006).
Sampling is the process of selecting a sample consisting of units, such as people, organisation
from a population of interest. By studying the sample we may answer the questions posed
regarding some aspects of the population from which they were chosen (Trochim & Donnelly,
2007).
The two general sampling methods are probability and non-probability sampling, which are
usually differentiated by their randomness. A non-probability sample is also known as a non-
random sample where samples are selected in some way not suggested by probability theory,
304
but sampling elements are selected using something other than a mathematically random
process (Neuman, 2006). Whilst probability sampling allows each segment of the population
to be represented in the sample. Probability sampling is also known as random sampling,
hence, Cooper and Schindler (2006) states that a probability sample is one based on the concept
of random selection – a controlled procedure that assures each population element is given a
known non-zero or allows each member of the population to have an equal chance of being
selected. In this case, the samples are chosen from a larger population by a process known as
random selection. The various sampling techniques employed in the selection of a probability
sample are simple random, stratified random, systematic, and cluster sampling.
Simple random sampling allows the sample to be chosen by simple random selection
where every member of the population has an equal chance of being selected. While the
stratified random sampling occurs in populations which consist of different strata or groups.
In order to have equal representation in a stratified sample, the researcher selects samples
equally from each one of the strata or group. Whereas cluster sampling sub-divides an
expansive area into smaller units, for instance a country could be sub-divided into regions and
further into towns. The clusters must be as similar to one another as possible, with each cluster
containing an equally heterogeneous mix of individuals and a subset of the identified clusters
is randomly selected.
Therefore, the current study used the probability sampling method, which allows all segments
of the low-income population as defined above to be represented in the sample, making sure
that a representative sample of low-income housing is selected for this study. Therefore, a
simple random and cluster sampling techniques were used, which allows each member of the
population to have an equal chance of being selected (Kerlinger & Lee, 2000) whilst a cluster
sampling technique divided the population into an expansive area with each cluster containing
an equally heterogeneous mix of individuals. The rationale for selecting this method of
sampling is based on the nature and composition of the low-income housing landscape in South
Africa\- hence cluster random sampling was used to ensure representativeness. The selection
of a representative sample for this study was based on the justification by Smith (2004) who
informed that random sampling must be used for a study of this nature, hence it was adopted.
Hence, from the nine provinces in South Africa, one province was selected; from the eight
established metropolitan municipalities in the country, only three were selected in the same
305
province; likewise, from the forty-four district municipalities in the country, only one was
selected, in the same province as the one in the three metropolitan cities. Similarly, within the
selected metropolitan and district municipalities, two low-income housing suburbs were
selected, making eight in total as in Table 11 above. In addition, the respondents were randomly
selected from the eight low-income housing suburbs. Respondents were selected and
questioned (questionnaire survey) based on the knowledge that they have been residents in the
area for more than a month and the houses have been allocated to them or had been tenants for
more than one month. All households from each location had an equal chance to be drawn and
to occur in the sample. Each category was classified as a cluster. Since the clusters differ in
sizes, a proportional representation of the each cluster was calculated.
7.4.4.6 Sample Size
Sample size is the number of observations or replicates to be included in a statistical sample.
The sample size is an important feature of any empirical study in which the goal is to
make inferences about a population from a sample, just like the current study. The sample size
used in the study was determined based on the expense of data collection, and the need to have
sufficient statistical power to validate the conceptual model. The sample size for the current
study was not based on the entire population of the selected low-income housing; therefore,
the sample size was not equal to the population size.
According to Neuman (2006), the question of how large a sample should be, depends on the
following: the kind of data analysis the researcher plans to use; how accurate the sample has to
be for the researcher’s purposes and the population characteristics. Likewise, Malhotra (1999)
informs that the determination of sample size depends on factors, such as the proposed data
analysis techniques, financial and access to sampling frame. The proposed data analysis
technique for this research is Structural Equation Modeling utilizing EQS software, which is
very sensitive to sample size and less stable when estimated from small samples (Tabachnick
& Fidell, 2001). As a general rule of thumb, at least 300 cases is deemed comfortable, 500 as
very good and 1000 as excellent (Comrey & Lee, 1992; Tabachnick & Fidell, 2001), thus it
was decided to target a sample size of 751 respondents from the eight low-income housing
suburbs as shown in Table 11.
306
Neuman (2006) informs that a large sample size alone does not guarantee a representative
sample. This is because a large sample size without random sampling or with a poor sampling
frame is less representative than a smaller one with random sampling and an excellent sampling
frame. Although, the larger the size of the sample, the more likely its mean and standard
deviation will be representative of the population’s mean and standard deviation. A larger
sample also makes it less likely that the researcher will obtain negative results or fail to
determine the truth. Hence, Leedy and Ormrod, (2005) advice that researchers should
endeavour to maximize the sample size and provide the following guidelines for selecting a
sample size:
1. For small populations with fewer than 100 people or other units, there is little
point in sampling, surveying the entire population;
2. If the population size is around 500, 50% of the population should be sampled;
3. If the population size is around 1 500, 20% should be sampled, and
4. Beyond a certain point (at about 5 000 units or more), the population size is
almost irrelevant and a sample size of 400 should be adequate.
In addition, Neuman (2006) recommends that for small populations, a researcher needs a large
sampling size and for moderately large populations, a smaller sample size of about 10 percent
is needed to be equally accurate. However, Cooper and Emory (1995) and Cooper and
Schindler (2006) disagrees on the 10 percent sample size recommendation for smaller
populations. Cooper and Schindler (2006) inform that a sample size does not have to be large
for it to be representative of the population. They state that the absolute size of a sample is
much more important than its size compared to the population, and how large a sample should
be is a function of the variation in the population parameters under study and the estimating
precision needed by the researcher. They suggested that a sample of 400 may be appropriate
sometimes, while more than 2 000 are required in other circumstances; and in another case,
perhaps only 40 are called for.
Furthermore, Smith (2004) simplify the process of sampling size by recommending that one
may use 20 cases or 5 percent, whichever is greater for the population. Schiller further informs
that the sample size should vary with the type of study, informing that a routine review study
would require 5 percent or 30 observations; a query review study would require 10 percent or
40 cases, whichever is greater, and an intensive review study would require a sample size of
15 percent or 60 cases and a sentinel event would require a 100 percent of the observations.
307
Hence, because of the kind of data analysis method (Structural Equation Modeling) to be used
in this study and the avoidance of negative results, which will jeopardize the model goodness-
of-fit; thus failing to establish the truth with regards to the constructs which predict residential
satisfaction, a large sample size of 751 was considered. This is because the role of sample size
is crucial in SEM analysis (Lucko & Rojas, 2010). Therefore the sample size requirement in
this thesis was a function of the model framework development consideration. For instance,
Harris and Schaubroeck (1990) proposed a sample size of 200 at least to guarantee robust
Structural Equation Modeling. Kline (2010) suggested that a very complicated path model
needs a sample size of 200 or more, while Bagozzi and Yi (2012) proposed that the sample size
should be above 200. Also, based on Smith’s (2004) research classification, the study is both a
query and an intensive review; hence the selection of the sample size demanded a 10 percent
or 40 cases and 15 percent or 60 cases whichever is greater. The study sample size also agrees
with Neuman’s (2006) recommendation of 10 percent sampling size. Therefore, the total
sample size of the respondents from the eight low-income housing suburbs was 751, which
aligned perfectly with the analysis of covariance structure estimation requirement.
7.4.4.7 Sample Selection
This step required a detailed specification of all the steps discussed above (Malhotra, 1999).
In this study, a total of 751 households were chosen in all localities for the research, which was
equivalent to the sample size. This was achieved as follows: each locality was divided into
different clusters using the streets, with each cluster containing 10 houses or more. A
systematic random sampling was then applied through the selection of every 5th house in each
cluster; for easy of identification of the 5th houses, house numbers were used to calculate the
number of the next 5th house. This process was essential in obtaining true representativeness
of the entire sample.
7.4.4.8 Site Visits
Prior to data collection, site visits were conducted by the researcher and the fieldworkers. The
site visits assisted in verifying the low-income housing locations that were selected and it also
gave the researcher and the fieldworkers’ first-hand experience of the types of dwellings and
knowledge of the areas for each potential survey, which was very useful in the data collection.
Before the data collection, copies of maps indicating the low-income areas were obtained from
308
the respective municipality. These maps showed the entire landscape and the streets, as well as
a birds-eye view of the houses in which the survey was conducted. These maps were further
confirmed during the site visit to ascertain the streets and households as revealed on the maps.
7.4.4.9 Fieldworkers
A fieldworker is defined as an objective collector of data. He or she may or may not have
formal qualifications but is perceived to have access to a particular community. Maart and Soal
(1996:1) determined that a fieldworker primarily mediates or facilitates learning of individuals
and groups to create an environment in which people can participate. Hence, Tamblyn and
Shelton (1996) defined in the comprehensive market research manual the data collection skills
that fieldworkers should possess, and thus, recommended that fieldworkers be selected with
great care and be trained for a stipulated minimum of four hours before undertaking
quantitative data collection. They concluded that it is essential for fieldworkers to have a good
understanding of the area and the respondents and to be trained in the skills necessary for
relating to people, analysing situations and designing strategies.
For this study, fieldworkers were recruited from the final year students of the Construction
Management and Quantity Surveying Department at the University of Johannesburg. They
were selected based on the researcher’s working knowledge of their ability and competence in
sustainable human development issues and their resident status in the survey areas. After their
recruitment, the researcher personally trained them on the use (administering) of the
questionnaire. An intensive four hours training workshop was conducted, followed by a day’s
technical training (different days for respective fieldworkers) on the selected sites according to
the areas they were to collect data.
The responsibility of the trained fieldworkers were to identify the specific households to be
surveyed based on the identified cluster and the study sample size; gain the consent of
respondents to be interviewed (surveyed); conduct interviews using the standard questionnaire;
and maintaining standard procedures in conducting the interviews (survey) and recording the
answers. During the collection of data, a total of 10 fieldworkers were used, which was based
on the sample size (751), numbers of days (60) available for the data collection and the
households per interview-day (5). In liaison with the work of the fieldworkers, the researcher
identified the clusters to be surveyed; supervised the first days of data collection in the various
309
areas, performed three to five interviews; ensured that the fieldworkers follow instructions as
stipulated on the survey instrument; control the data quality by checking for errors during the
survey and after each survey, checked that the questionnaires were completed fully and
correctly and by checking that all the respondents answered the questions; and identified
problem areas, which were adequately resolved.
7.4.4.10 Pilot Study
A pilot study is a trial questionnaire survey study designed to test logistics and gather
information prior to the larger study, in order to improve the latter’s quality and efficiency. A
pilot study when properly done reveals deficiencies in the design of a questionnaire and these
can then be addressed before time and resources are expended on large scale studies. A good
research strategy requires careful planning and a pilot study will often be a part of this strategy.
A pilot study is normally small in comparison to the main survey and therefore, can provide
only limited information on the sources and magnitude of variation of response measures. It is
unlikely, that a pilot study alone can provide adequate data on the potential deficiencies in a
study. A systematic review of the literature or even a single publication is a more appropriate
source of information on the sources of inconsistencies.
However, a trial questionnaire survey was conducted for the current research to test logistics
and gather information prior to the larger study. The pilot study was used to modify the survey
instrument; thus it was not analysed for incorporation into the main study. The pilot study was
also used to address a number of logistical issues to ensure that the instructions given to
respondents were comprehensible; to check that fieldworkers could adequately administer the
questionnaire to the respondents; and to identify potential problem areas, such as respondent
ages, ownership of dwellings, housing status and other sensitive questions.
Besides the reasons stated above for conducting a pilot study, Yuen (2007) identified some
procedures which should be adhered to in order to enhance a good pilot study outcome of a
survey questionnaire. All of these procedures were likewise followed strictly when the pilot
study for this research was conducted. These include:
1. Administering the questionnaire to pilot subjects in exactly the same way as it will be
administered in the main study;
2. Asking the subjects for feedback to identify ambiguities and difficult questions;
310
3. Record the time taken to complete the questionnaire and decide whether it is reasonable;
4. Assessing whether each question gives an adequate range of responses;
5. Establishing that replies can be interpreted in terms of the information that is required;
6. Checking that all questions are answered; and
7. Rewording or rescale any questions that are not answered as expected.
A total of 40 subsidised low-income housing occupants from each of the identified location
were chosen as pilot respondents.
7.4.4.11 Data Collection
After determining the sample size for the study, site visit, fieldworker’s selection and training,
together with the pilot testing of the questionnaire; the questionnaire was personally
administered to the low-income residents’ by the researcher and the fieldworkers. This method
was used because it has been suggested that when dealing with a population likely to be of the
low-income group with low interest and motivation, personally administered or face-to-face
structured questionnaire for data collection is the preferable option (Fowler, 1993). Although,
the respondents that could read were given the questionnaire to complete themselves and where
necessary, were given clarity. It took approximately 35 minutes to complete each questionnaire,
although respondents were informed in the cover letter that it would take 30 minutes to
complete.
The process of data collection took two months, from the middle of the month of May 2012 to
middle of the month of July 2012. A majority of the questionnaires were completed by the
fieldworkers and only in exceptional cases did the respondents, who could read and write
request that they were given the questionnaire to fill in their own time without undue pressure
so that they could have time to think about their responses. This was the reason in part why the
data collection process took so long. After the collection of data of 751 questionnaires, each
questionnaire was then marked for identity based on the low-income housing location\-
according to the metropolitan and district municipality where the data was collected and was
then sent for data capturing by the statistics department of the University of Johannesburg.
311
7.4.4.12 Data Analysis from Questionnaire Survey
Data analysis involved steps such as coding the responses, cleaning, screening the data and
selecting the appropriate data analysis strategy. Cleaning and screening of the questionnaire
was done during the data collection. Coding of the questionnaire involved identifying,
classifying and assigning a numeric or character symbol to data, which may be done in two
ways: pre-coded and post-coded (Wong, 1999). This will be elaborated on in the next section.
In this study, most of the responses were pre-coded except for questions 1-3, 8-10, and 17-21
which required post-coding. Taken from the list of responses, a number corresponding to a
particular selection was given. This process was applied to every question that needed this
treatment. Upon completion, the data was then entered onto a Statistical Analysis Software
Package (SPSS) for the next analysis steps.
In choosing the appropriate statistical analysis technique, the following was considered: the
research elements, namely the research problem, objectives, characteristics of data and the
underlying properties of the statistical techniques (Malhotra, 1999). To meet the purposes of
this study, descriptive and inferential analyses and the measures of goodness-of-fit of model
were applied where necessary. The data analysis involved the use of multiple analytical
techniques to facilitate ease of communicating the results, while at the same time improving its
validity. Hence, the researcher chose to use SEM utilizing EQS software. Raw data from the
questionnaire was entered into the Statistical Package for Social Science (SPSS) software and
was later exported to the SEM software EQS Version 6.2 for analysis. The motivation for the
choice of the SEM and particularly the use of the software EQS is explained in the next session.
Inferential analysis refers to the cause-effect relationships between variables which the current
study hopes to establish between the identified model constructs. Inferential statistics use the
results obtained from samples to generalize about a population (Forzano, 2008). Inferential
statistics used for this research were correlations, Confirmatory Factor Analysis (CFA) and
Structural Equation Modeling (SEM). SEM was used for the development and validation of
the residential satisfaction model. Also, the statistical significance of the constructs was
evaluated. The result’s statistical significance was expressed by p-value (Forzano, 2008). When
the p-value is high, there is less possibility of an association between two variables (McClave,
Benson & Sincich, 2008), while a smaller p-value gives a better likelihood of association. The
312
p-value chosen in the present study is 0.05, which implies 95 percent chance that the
populations mean is within a listed range of values (McClave et al., 2008).
SEM is currently the most inclusive statistical procedure in social and scientific research
catering for all operations of the General Linear Modeling (GLM) group of statistics such as
Analysis of Variance (ANOVA), Multivariate Analysis of Variance (MANOVA) and multiple
regression (Kline, 2005:14). Though there are many ways to describe SEM, it is most
commonly thought of as a hybrid between some form of analysis of variance (ANOVA) /
regression and some form of factor analysis. In general, it can be remarked that SEM allows
one to perform some type of multilevel regression/ANOVA on factors. SEM is conceptually
used to answer any research question involving the indirect or direct observation of one or more
independent variables or one or more dependent variables. However, the primary goal of SEM
is to determine and validate a proposed causal process and/or model. In the current study, the
conceptualized holistically integrated residential satisfaction model for public housing
occupants in South Africa is being validated. SEM takes a confirmatory approach to the
analysis of a structural theory bearing on some phenomenon (Byrne, 2010). However, Dion
(2008:365) claims that SEM simultaneously estimates all coefficients in the model and
therefore, it is able to assess the significance and strength of a relationship in the context of the
entire postulated model. Hence, considering that the conceptualized model in this thesis
consisted of unobserved (exogenous) variables, which had to be estimated from the observable
variables, methods of analysis such as ANOVA could not be used as they lack a direct way of
distinguishing between observed measures and the underlying constructs (Kline, 2005:14).
Also, in SEM, as clear distinction is made between true variance and error variance, which
implies that model parameters are estimated by taking measurement error into consideration.
However, before the SEM was performed, CFA was carried out on each exogenous variable to
determine best-fit for the model.
The choice of the software EQS for analysis was enhanced by the benefit of utilizing the
Satorra-Bentler scaled statistics (S – Bχ2), which provides an adjusted, more robust measure of
fit for non-normal data. This approach is more accurate than the normal chi-square test statistics
(χ2) (Byrne, 2006:22). Likewise Kline (2005:83) and Musonda (2012) inform that EQS offers
several different estimation methods for non-normal data as well, including the Robust
Maximum Likelihood (RML).
313
As already informed EQS Version 6.2, a software package was used for SEM as it is user-
friendly software that provides a graphical user interface, which is easy to understand. EQS
also enables data to be imported directly from SPSS. Other reasons why the researcher adopted
EQS 6.2 and SPSS 20.0 software include: first, the software is available at the University’s
Postgraduate and Statistics Centre; hence, it was easier for the researcher as a postgraduate
student to access the software. Second, EQS is seldom used by previous researchers as revealed
from the literature to enhance conceptual understanding of residential satisfaction in subsidised
low-income housing research as compared to other techniques, such as AMOS and LISREL
(Tong, 2007). Being a user-friendly graphically modeling interface, EQS offers a wider variety
of goodness-of-fit measures (Tong, 2007).
Data screening and preparation
Before a detailed analysis of the postulated model was conducted to determine fit, screening of
the data was essential. Pre-analysis data screening focused on establishing whether there was
any missing data, outliers, the distribution characteristics of the data, and the identification of
the model.
Missing values and outliers can adversely affect SEM results by their presence in the raw data.
Hence, it was therefore necessary to identify at the pre-analysis stage, any missing values and
outliers in order to determine the best way to handle them. Kline (1998) informs that missing
data can be treated in three ways, which are: casewise deletion, pairwise deletion and
imputation. However, other new technologies are now being adopted to handle missing data,
such as maximum likelihood estimation (Arbuckle, 1996). In casewise deletion only cases with
complete records are included, that is, all analyses are conducted with the same cases (Kline,
1998), and hence consistency is maintained. On the other hand, an alternative approach of
pairwise deletion of cases excludes the missing responses for variables involved in a particular
computation. This method uses all possible cases for each calculation. In the pairwise-deletion,
variables containing missing data or alignment gaps are removed from the analysis as the need
arises. This is in contrast to the complete-deletion option in which all such sites are removed
prior to the analysis. Imputation is another method used for analyzing missing responses. This
technique involves pattern matching, which replaces “a missing observation with a score from
another case with a similar profile of scores across other variables” (Kline, 1998:75).
314
Subsequent to outlier tests, an assessment of normality was performed (Churchill & Iacobucci
2004; Tabachnick & Fidell, 2001). The first basic assumption about SEM is that all data have
a multivariate normal distribution. Multivariate normality includes both the distributions of
individual variables and the distributions of combinations of variables (Hooley & Hussey
1994). This assumption is necessary in order to allow significance testing using the t-test and
F statistics (Arbuckle, 1996; Hair et al., 1998; Tabachnick & Fidell, 2001). For example, in
SEM models, estimation and testing are usually based on the validity of multivariate normality
assumption, and lack of normality will adversely affect goodness-of-fit indices and standard
errors (Kassim, 2001). For instance, the common estimation method in SEM is maximum
likelihood, which carries with it the assumption of multivariate normality (Kline, 2005:48).
Also, other problems have been identified with non-normal data. Kline (2005:137) reports that
failure to meet the assumption of multivariate normality could lead to an overestimation of the
chi-square statistics and therefore, to an inflated Type 1 error (rejecting a model, which should
not be rejected). As a result, because the missing values were scattered across items, hence
Maximum Likelihood Estimation (MLE) Method of treating missing values was adopted. This
method was used because under ordinary conditions, ML estimates are (Bentler & Wu, 2002):
consistent (approximately unbiased in large samples, such as the present; N = 751); is
asymptomatically efficient (have the smallest possible variance) and asymptomatically normal
(one can use normal theory to construct confidence intervals and p-values). Hence, the MLEs
for the data set are the values of the parameters, which maximize the probability of the observed
data (the likelihood).
Likewise, the examination of the Skewness, Kurtosis, Mardia’s Coefficient and the
multivariate kurtosis was conducted to establish normality of the data. In addition to identifying
the missing values and outliers, assessment of multivariate normality was equally important
because the choice of the estimation method depends on it (Schreiber et al., 2006:327).
According to Tabachnick and Fidell (2001), skewness refers to the symmetry of a distribution,
that is, a variable whose mean is not in the centre of the distribution is regarded as a skewed
variable. On the other hand, Kurtosis relates to the peakedness of a distribution. A distribution
is said to be normal when the values of skewness and kurtosis are equal to zero (Tabachnick &
Fidell, 2001). However, there are few clear guidelines about how much non-normality is
problematic. Many authors suggest that absolute values of univariate skewness indices greater
than 3.0 seem to describe extremely skewed data sets (Chou & Bentler, 1995; Hu & Bentler,
1992). Regarding kurtosis, there appears to be less consensus and a conservative compromise
315
seems to be that absolute values of the kurtosis index greater than 10.0 may suggest a problem
and values greater than 20.0 may indicate a more serious one (Kassim, 2001; Kline, 1998). In
this study, all variables were tested at a univariate and multivariate level for normality using
EQS. At the univariate level, none of the observed variables in the proposed models, had
skewness greater than 3.0 and none had a Kurtosis index greater than 8.0.
Model Identification
In order to proceed to the model analysis for the study, it was critical to determine whether the
postulated model could be analysed or not. Model complexity is determined through
establishing whether a model is just-identified, under-identified or over-identified. A just-
identified model is one, in which there is a one-to-one correspondence between the data and
the structural parameters. That is, the number of data variances and covariance equals the
number of parameters to be estimated (Byrne, 2006:31). Further, Byrne (2006:31) informs that
despite the capability of the model to yield a unique solution for all parameters, the just-
identified model is not scientifically interesting because it has no degree of freedom and
therefore, can never be rejected. An over-identified model is one, in which the number of
estimable parameters is less than the number of observations (Byrne, 2006:31). Accordingly,
an over-identified model is desirable as it results in a positive degree of freedom that allows
for rejection of the model therefore, rendering it to be of scientific use (Byrne, 2006:31).
Finally, the under-identified model is one, in which the number of parameters to be estimated
exceeds the number of variables and covariances. As a result, there can be an infinite number
of solution, and therefore, defeats the purpose of the analysis (Byrne, 2006:106; Kline,
2005:109). In summation, for a model to be analysed, there has to be at least as many
observations as parameters to be estimated, meaning that the degree of freedom (df) should be
greater than zero (df ≥ 0) (Kline, 2005:100). Therefore, the study model is over-identified
because there were 84 indicators for both the exogenous and the endogenous variable. There
were 3570 data points (meaning, 84 variances and 3486 covariances). The errors were
uncorrelated and each indicator loads on only one factor. In addition, the covariance between
the factors was not zero. Hence, this study’s hypothesised model is said to be identified.
Parameter Estimates and Input Matrix Method
Further to the above, an examination of the degree of freedom of the postulated model in this
thesis revealed that the model was over-identified. The least value for the degree of freedom
was found to be 2; which related to the needs and expectations, beneficiary’s participation,
316
services provided by the government and the residential satisfaction manifest construct.
Likewise, all values of degree of freedom for the model constructs were positive and therefore,
indicative of an over-identification of the measurement models (Kline, 2005:100).
After the screening process was completed, it was established that the data was non-normal
with the lowest Mardia’s Coefficient of 13.1652 (residential satisfaction) and the highest
Mardia’s Coefficient of 56.0118 (beneficiary’s participation). This is shown in Table 10.6 (See
page 439). The non-normality of the data influenced the choice and use of the Robust
Maximum Likelihood (RML) Estimation Method. The RML gives several robust fit indices
(Bartholomew, Loukas, Jowers & Allua, 2006:72; Byrne, 2006:22). Hence, Byrne (2006:22)
suggests that one of the outputs from the RML Estimation Method is the robust chi-square
statistics (χ2) referred to as the Satorra-Bentler Scaled Statistics (S – Bχ2) and robust standard
error, which are corrected for non-normality in large samples, as the case of the present study;
with the sample size being 751. SEM software, EQS Version 6.2 was used in part, due to the
ability of the programme to adjust standard errors for the non-normality of the data (Byrne,
2006:22).
The covariance matrix method was the chosen in-put matrix for the analysis in the current
study. The analysis strategy adopted to analyse the hypothesized model was firstly used to
estimate the measurement part of the model and thereafter, to analyse the measurement and
structural parts of the model respectively. Likewise, the results from the analysis were reported
in the same manner namely, results from the measurement model analysis referred to as the
Confirmatory Factor Analysis (CFA) were presented first and thereafter, the results from the
analysis for the entire structural model referred to as the full latent variable model (FV) were
presented.
Model Analysis / Fit Indices
Analysis of the hypothesised model was the next step after the pre-analysis conditions,
selection of the input matrix of the data and the method of estimation was determined. The
following fit indices identified from Hu and Bentler (1999:5); Boomsma (2000:473); Kline
(2005:134); Streiner (2006:232); and Hooper, Coughlan and Mullen (2008:58) were examined
to determined model fit. These statistics were:
317
1. Chi-square values χ2;
2. Satorra-Bentler Scaled Chi-square (S – Bχ2);
3. Bentler Comparative Fit Index (CFI);
4. Standardised Root Mean Square Residual (SRMR);
5. Goodness of Fit Index (GFI);
6. Root Mean Square Error of Approximation (RMSEA); and
7. Root Mean Square Error of Approximation with its 90% confidence interval (RMSEA
@ 90% CI).
The decision on model fit indices was based on the proposal by Hu and Bentler (1999:28) two-
index presentation strategy of incremental and absolute fit indexes because they seemed to
perform superiorly to a single index presentation strategy. Hu and Bentler (1999:27) suggested
therefore that the maximum likelihood based SRMR and a supplemental fit index such as CFI
or RMSEA, would result in minimum Type I (the probability of rejecting the null hypothesis
when it is true) and a Type II Error (the probability of accepting the null hypothesis when it is
false). The fit indexes χ2, CFI, and (S – Bχ2) belong to the incremental or comparative fit
indexes, which are a group of indices that do not use the chi-square in its raw form but compare
the chi-square value to a baseline model (Hooper et al., 2008:55). While the SRMR and
RMSEA belong to the absolute fit indexes, which are fit indices, which determine how well a
priori model fits the sample data (McDonald & Ho, 2002) and demonstrates, which proposed
model has the most superior fit.
Further, additional fit index (Goodness of Fit Index - GFI) was adopted by the researcher for a
more stringent measure to evaluate the overall model fit. This follows the work of Tong (2007).
According to Tong (2007) and Kassim (2001), the GFI is an important measure of absolute fit.
It refers to the percent of observed co-variances implied by the model (Garson, 2009). Garson
(2009) and Tong (2007) inform that GFI should be equal to or greater than 0.90 for a
parsimonious model (Garson, 2009; Tong, 2007). While researchers such as Joreskog and
Sorbom (1988) and Schumacker and Lomax (2004) suggest that acceptable GFI value should
be closer to 0.95.
These measures (χ2; CFI; S – Bχ2; SRMR; RMSEA; RMSEA @ 90% CI; and GFI) provide the
most fundamental indication of how well the proposed theory fits the data. Unlike incremental
318
fit indices, their calculation does not rely on comparison with a baseline model but is instead a
measure of how well the model fits in comparison to no model at all (Hooper et al., 2008:53;
Jöreskog & Sörbom, 1993). The adopted cut-off values for the above fit indices are as tabulated
in Table 7.12.
Table 7.12: Cut-off criteria of fit statistics
Statistics Cut-off criteria Source Chi-square - χ2
Ratio χ2 to df ≤ 2 or 3 with an
insignificant p value (p > 0.05)
Hooper et al., 2008;
Kline, 2005,
Schreiber et al., 2006
Bentler Comparative Fit
Index (CFI)
Value should be ≥ 0.95 for good
fit
Bartholomew et al., 2006;
Hooper et al., 2008;
Bentler, 1990; Dion, 2008;
Schreiber et al., 2006
Standardised Root Mean
Square Residual (SRMR)
The value should be ≤ 0.08
A value of 0.1 is also acceptable
Kline, 2005;
Hu & Bentler, 1999;
Schreiber et al., 2006
Root Mean Square Error of
Approximation (RMSEA)
Value should be < 0.05 for good
fit
Values < and 0.08 indicate a
reasonable error of approximation
Values of > 0.10 suggests a poor
fit
Hu & Bentler, 1999;
Kline, 2005; Steiger, 2007
Dion, 2008;
Schreiber et al., 2006
Root Mean Square Error of
Approximation with its
90% confidence interval
(RMSEA @ 90% CI)
Values to be < 0.06 to 0.08 with
confidence interval
Byrne, 2006;
Schreiber et al., 2006
Goodness of Fit Index (GFI)
Should be > 0.90 Joreskog & Sorbom
(1988);
Statistical Significance of Parameter Estimates
The statistical significance of parameter estimates was established by examining the ration
output of the parameter estimate divided by its standard error (therefore analogous of Z-values)
and tests that the estimate is statistically different from zero (Byrne, 2006:103; Schreiber et al.,
2006:327). Hence, based on an alpha (α) level of 0.05, the test statistics had to be greater than
1.96 (Z > ± 1.96) before the hypothesis (meaning, the estimate = 0.00) could be rejected (Byrne,
2006:103).
319
Reliability and Validity
In order to establish the score reliability, the internal consistency reliability measure statistics
of Rho coefficient and Cronbach’s (1951) alpha (α) were examined. Byrne (2006:132-133) and
Kline (2005:59) theorize that Cronbach’s alpha measures the degree to which responses are
consistent across all items within a single measure and if this statistics is low, the content of
the items may be so heterogeneous that the total score is not the best possible unit of analysis
for the measure. Hence, the acceptance of Cronbach’s Alpha to measure internal homogeneity
is limited. Byrne (2006:133) argues that the use of the Cronbach’s Alpha Coefficient to judge
latent variable models especially models with multi-dimensional structure is questionable
because it is based on a very restrictive model that requires all factor loading and error variances
to be equal. Therefore, in establishing score reliability for the analysis, the Rho Coefficient was
relied upon more than the Cronbach’s Alpha Coefficient even though it is the most common
method used for assessing the reliability for a measurement scale with multi-point items
(Hayes, 1998). The Rho coefficient provides a good estimate of internal consistency because
the model that was analysed in the current study was a full latent variable mode (Byrne,
2006:133).
Measurement Models - Confirmatory Factor Analysis
This study used Confirmatory Factor Analysis (CFA) to scrutinize the factor structure of the
exogenous and endogenous indicator variables. In contrast to exploratory factor analysis, the
aim of which is simply to identify the factor structure present in a set of variables, the aim of
CFA is to test an hypothesized factor structure or model and to assess its fit to the data. CFA
may be viewed as a sub-model of the more general structural equation modeling (SEM)
approach to analysis. Specifically, it is a measurement model of the relations of indicators
(observed variables) to factors (exogenous variables), as well as the correlations among the
latter. CFA is generally based on a strong theoretical and/or empirical foundation that allows
the analyst to specify an exact factor structure in advance. The CFA approach usually restricts
which variables will load on which factors, as well as which factors will be correlated. In CFA
each observed variable has an errors term, or residual, associated with it that expresses the
proportion of variance in the variable that is not explained by the factors. These error terms
also contain measurement error due to the lack of reliability in the observed variables. The
typical research question in CFA is: Are the covariances (or correlations) among variables
consistent with an hypothesized factor structure? As such, CFA is quite useful for studying the
factorial validity of multi-item construct such as residential satisfaction.
320
Therefore, after establishing the score reliability, the construct validity was conducted to
demonstrate the extent to which the constructs hypothetically relate to one another. This can
also be referred to as the test of measurement invariance (MI), factorial invariance or
measurement equivalence between indicator variables. Measurement invariance is a very
important requisite in Structural Equation Modeling. It attempts to verify that the factors are
measuring the same underlying latent construct within the same condition. MI is aimed to
ensuring that the same attribute must relate to the same set of observations in the same way.
The MI for the present thesis was determined based on examination of the residual covariance
matrix from the CFA output result, which determined the variables to include the full structural
model.
Hence, preliminary Confirmatory Factor Analysis (CFA) was performed to measure the
dimensions of all latent variable indicators to identify which items were appropriate for each
dimension. Indicators variables with an unacceptably high residual covariance matrix (>2.58)
were dropped. Residual covariance matrix values greater than 2.58 are considered large (Byrne,
2006:94; Joreskog & Sorbom, 1988). In order for a variable to be included in a CFA
measurement model analysis, which enables the model to be described as well-fitting, the
distribution of residuals covariance matrix should be symmetrical and centered around zero
(Byrne, 2006:94; Joreskog & Sorbom, 1988). This procedure was adopted as a means to ensure
that the indicator variables were measuring the same latent construct. For instance, when an
investigator wishes to use a given measure or set of measures to make evaluations, the validity
of those comparisons depends on the assumption that the same construct is being measured.
Hence, the assumption of measurement invariance is most times tested in CFA (Meredith,
1993), so as to allow for comparison of indicator variables under the same condition. The
assessment of measurement invariance across latent variables involves the use of multi-sample
CFA as used in this thesis. This procedure has been described and used by Widaman & Reise
(1997) and Reise et al. (1993). Also, Little (1997) investigated invariance of factors of control
expectancy across gender and four cultural groups. Kim et al. (1996) studied invariance of
world views and religious beliefs of older adults over time.
Since this study sought to test the potential relationships among variables, a Confirmatory
Factor Analysis using EQS 6.2 was applied on the indicator variables that passed the first CFA
test of Residual Covariance Matrix Analysis. Further, to achieve construct validity, the
321
measurement should demonstrate convergent validity and discriminant validity. Convergent
validity refers to the items purporting to measure the same construct correlates positively with
one another (Malhotra, 1999; Parasuraman, 1991) as already described above. On the other
hand, the latter requires that an item does not correlate too highly with other items of different
constructs (Hair et al., 2003; Malhotra, 1999). In this study, the correlation matrix and inter-
construct correlation were analyzed for convergent and discriminant validity. In addition, due
to the absence of another external criterion against which comparison could be made of the
measures, discriminant validity was also used to examine construct validity. This is because a
set of variables hypothesised to measure different aspects only shows discriminant validity if
their inter-correlations are not too high (Kline, 2005:60).
7.4.4.13 Ethical Considerations
Finally, it is pertinent to consider the proper conduct of this research. This research
accommodated the responsibilities to protect the interests of the survey respondents. With
regards to the survey respondents, no one was coerced to respond to this survey. The
respondents were asked to participate of their own freewill, that is, they were told of their rights
not to participate or to end their participation if they so wished. In addition, they were briefed
about the purpose of the study and how or why they were chosen. As such they were free from
deception or stress that might arise from their participation in this research. The respondents
were also guaranteed protection through anonymity and all information that may reveal their
identity, which are held in strict confidence.
7.4.4.14 Reliability and Validity of the Questionnaire Survey
The validation of research work is important for studies that will influence the overall welfare
of the public. Studies that may have an influence on the housing, economy, political climate,
or environment are typically validated before the results are used to make influential decisions
(Thorne & Geisen, 2002). Since this study focuses on residential satisfaction in subsidised low-
income housing, which is a much contested space, the validation of the results are extremely
important. Therefore, it is considered good practice for researchers to demonstrate the
reliability of their data collection, analysis and findings.
Validity refers to the degree to which the findings of a research are interpreted in the
correct way. Similarly, validity determines whether the identified inputs within their
322
attributes actually produce the expected output or result (Barrett & Sutrisna, 2009). It is
the extent to which the results of a study can be verified against the stated objectives. Validity
may be content, construct or criterion-related, whilst reliability issues include scoring
agreement, test, equipment forms and internal consistency. Validity may be evaluated
internally or externally. Internal validity refers to whether the identified inputs within their
attributes actually produce the expected output or result; and external validity refers to the
extent to which any research findings can be generalised, or extrapolated beyond the
immediate research sample or setting in which the study took place.
Further, reliability denotes the consistency of results obtained in the research. In other words,
it is the soundness of the method for data collection or the degree to which the findings
of research are independent of any accidental circumstances. In research, reliability is achieved
when the same research process is repeated and reproduces results within the stated confidence
limits. Hence, Eriksson (2002) states that reliability of an investigation is satisfactory if another
researcher conduct the same research and draw the same conclusions. Thus, reliability deals
with the quality of data and this requires the triangulation of the various sources of data, which
provide similar results from different angles. This requires a thorough demonstration of rigour
and clarity of research findings.
Therefore, the questionnaire instrument for this study was designed to reflect the above issues
and therefore, intends to capture all necessary information to accomplish the research. A major
criticism against the use of questionnaires is the fact that they may lack validity.
Likewise, respondents may interpret questions in a different way from what was intended
especially, when ranked responses are asked. Also, respondents may not be totally honest in
their answers (Miller & Brewer, 2003). Hence these limitations were overcome by the
researcher through pilot-testing of the questionnaires on a small sample group and personally
administering them (Refer to Section 7.4.4.10). The findings were intended to demonstrate
rigour and also repeatability within the context of housing satisfaction studies, thus
improving the questionnaire instrument. Another reliability problem such as observer bias was
minimized by the involvement of the main researcher as the quality controller during the
process of data collection only by scrutinizing the work of the field worker. This ensured a
higher level of consistency in the nature of the data collected. Furthermore, the research was
designed to ensure a maximum degree of objectivity within the scope of the study. This was
achieved through the use of a Mixed Method as already discussed.
323
7.4.4.15 Generalisability
This is the extent to which the findings and conclusions of research conducted on a population
sample can be generalised or extended to the entire population. Generalisability is based on the
frequent occurrence of a phenomenon and so when there is sufficient data to support the
validation of a hypothesis, a premise exists to generalize the behaviour of such data in similar
circumstances. However, due to its foundation in probability theory, generalization cannot be
regarded as conclusive (Shakantu, 2004:185). Generalisability is more applicable in
quantitative research involving large samples than qualitative research. The rule is, the larger
the population sample, the more the results tend towards generalization, which was applicable
to the current research. Also, the adoption of Mixed Methodology involving both quantitative
and qualitative data addresses the issue of generalisability of the findings in this research.
7.5 CONCLUSION
This chapter presented the methodology adopted for the conduct of this research. It also
provided the justifications for the philosophical position and methods of data collection. The
research design described in this chapter has linked three important elements of the research
methodology, namely; the underlying philosophical assumptions; the research
methods/approach; and the process followed in the questionnaire administration, as well as an
introduction to the data analysis. Finally, ethical considerations pertaining to the collection of
data were discussed; including issues relating to validity and reliability have also been
discussed. In the following chapter, results of the data analysis from the Delphi Study are
presented.
324
CHAPTER EIGHT
RESULTS FROM THE DELPHI STUDY
8.1 INTRODUCTION
A Delphi Study was conducted to determine and solicit expert’s views on the influence
(probability) and impact of residential satisfaction attributes on low-income housing
beneficiaries, thus, identifying the determinant attributes that bring about residential
satisfaction in the South Africa low-income housing context. The Delphi Technique solicited
from the experts, critical issues affecting the provision of housing for the low-income group in
South Africa. Also, tested through the Delphi study was the expert’s view on housing
development and management issues that affect the national, provincial, and the local
government housing agencies in South Africa. Three rounds of the Delphi process were
conducted before experts could reach consensus on the questions that were posed to them.
This section presents a summary of results from all the rounds from the first to the third round.
Computation for each and every question element is made for the influence (probability) and
impact of the attributes in predicting residential satisfaction and improvement of low-income
housing in South Africa. The influence or impact of the absence or presence of a particular
residential satisfaction element on the overall residential satisfaction of the other elements is
presented; likewise, on issues relating to low-income housing in South Africa. The following
sections provide an analysis of the results. The composition of the expert panel and a general
background to the Delphi Study is first described, followed by the study’s findings. The chapter
concludes with a summative discussion of the findings based on the objectives of the Delphi
Study.
8.2 BACKGROUND TO THE DELPHI SURVEY
It is important that the reader is first reminded of the objectives that were to be met through
conducting a Delphi Study. The objective of conducting the Delphi Survey for this study was
to determine the following:
DSO1 To identify the attributes (main and sub) that determine residential satisfaction
325
and to examine if the attribute that determines satisfaction in other cultural
contexts is the same in South Africa;
DSO2 To determine the factors that make subsidised public housing unsustainable in
South Africa;
DSO3 To identify the combination of housing policy instruments that will better serve
South African low-income housing group;
DSO4 To identify the critical factors affecting the delivery of low-income housing and
their effects on beneficiaries’ residential satisfaction;
DSO5 To predict the life span of the present South African public housing subsidy
delivery model;
DSO6 To evaluate the management issues affecting the national, provincial and local
government housing agencies in the delivery of housing in South Africa;
DSO7 To determine the influence of beneficiary participation on their overall housing
satisfaction; and
DSO8 To determine the effect of meeting beneficiary’s housing needs and
expectations on their overall housing satisfaction.
The philosophy behind the above objectives is to do away with the tendency of a non-coherent
dialogue on low-income housing in South Africa. Thus, the associated outcomes from
achieving the above objectives were:
1. to determination of key factors and constructs that are of critical significance
(influence) to determine residential satisfaction in subsidised low-income housing; and
2. to develop a holistically integrated conceptual model on residential satisfaction in South
African low-income housing.
A panel of 17 experts participated in the first round, while 15 experts were retained who
participated from the second to the third round of the study. In selecting the panel, particular
attention was given to the experts’ residency status, which was considered compulsory for all
selected experts (Refer to Section 7.4.3.4). This was considered significant because experts
were required to have a thorough understanding of the low-income housing context in their
respective residential metropolitan and district municipality cities. Panel members were
selected to achieve balance between practitioners and theorists in the field of sustainable human
development with a core focus on housing for the poor and low-income groups.
326
A questionnaire was designed for each round based on responses to the previous round.
However, the Round One questionnaire was designed, based on a summary of the
comprehensive review of literature highlighting sets of main and sub-attributes that are
potentially relevant to residential satisfaction decisions by the occupants of low-income
housing. Additionally, issues relating to the provision of low-income housing, delivery and
sustainability beyond the current level were also extracted from the reviewed literature. These
were structurally and constructively put together to frame the first round of the Delphi Survey.
Subsequently, Round One of the Delphi Study was intended as a brainstorming exercise to
produce a list of empirical attributes that determine residential satisfaction and issues pertaining
to low-income housing in South Africa in order to achieve the study objectives. Closed- and
Open-ended questions were used in this round; thereafter, these were analysed and thus formed
the basis of Round Two and Three of the study. Frequencies were obtained to measure the
degree of consensus reached amongst participants regarding the attributes that determine
residential satisfaction in South African low-income housing and for other related questions.
Also, content analysis methodology was adopted to analyse responses to the open questions to
“minimize redundancy” (Rubin et al., 1998:6).
The purpose of the second round of the study was to allow experts to review and comment on
the attributes that determine residential satisfaction and other issues relating to low-income
housing in South Africa, which were proposed by experts’ participation in Round One. Closed
questions were used in this round to investigate participant comments expressing agreement,
disagreement or clarification concerning proposed attributes that determine residential
satisfaction in South Africa. The specific nature of the closed-ended questions stimulated
participants’ reactions (refer to Appendix E).
The Delphi Round Three, there was a revision of Delphi Round Two, and panellists were again
asked to respond using the provided rating scale ranging from negligible to very high impact
as applicable to the question. In the third round, statistical information calculated from the
second round was reported to each panel member. The results of each Delphi round were
reviewed and compiled by the researcher. After analysing the responses from the third round,
the characteristics, and features which determine residential satisfaction as agreed upon by the
panel of experts, were organized to create a more complete picture of those attributes that
determine residential satisfaction and likewise to other issues as presented in the study.
327
The results of Round Two indicated that the experts were in general agreement, and that Round
Three successfully refined the discussion to the point that clear points of consensus or lack
thereof, could be determined; therefore, a fourth round was not necessary. The median, mean,
standard deviation, percentages and IQD scores of each question were calculated and in
situations where the score was two points from the median score the experts were requested to
expound on the response. If a consensus had not been formed at the third round, the data from
this third round would have been analysed again by calculations for median, mean, standard
deviation and percentages scores and sent to the experts for consideration in responding to a
fourth round. Also, content analysis approach was adopted to analyse responses to the open-
ended questions.
The researcher predicted the use of only three rounds in order to achieve a consensus and this
prediction was correct. A fourth round was not needed, and the participants were informed of
this, when the third round questionnaires were sent out. The goal of the research technique was
to cycle the questions towards a consensus amongst the experts. During each round of
questionnaires, the experts were given the results of the medium of the previous round. It was
anticipated that by the third round, responses would converge to indicate a consensus from the
experts. In actuality, a consensus is achieved with 100% of the participants in agreement, but
two-thirds in agreement is considered a common consent (Stitt-Gohdes & Crews, 2004). The
goal for this study was that each individual question or statement should have a consensus, but
common consent would be acceptable. Common consent was obtained if 60.0% of the experts
agreed on each statement which was achieved in the study (refer to Section 7.4.3.7). All
statements were examined individually for a consensus. The quantitative results were
statistically analysed after each round of questionnaires to determine if a consensus had been
reached for each question or statement using the provided scale for each question or statement.
If a consensus was reached prior to the final round that question or statement was no longer
required (asked/required) in the next rounds.
After the third round Delphi Survey, consensus was reached regarding most of the attributes
that determine residential satisfaction and on other low-income housing related issues in South
Africa. Based on the findings of the analyses of responses to the Delphi rounds, a list of
attributes that determine residential satisfaction was prepared which informs the conceptual
framework for the broader study, while issues surrounding low-income housing in South Africa
were highlighted, which responded accordingly to the set objective of the Delphi Study. The
328
results of the Delphi Study are therefore presented in relation to the specific Delphi objectives
in the next section.
8.3 FINDINGS FROM THE DELPHI STUDY
DSO1 - To identify the attributes (main and sub) that determine residential satisfaction and
to examine if the attribute that determines satisfaction in other cultural contexts is the same
in South Africa
From the summary of the comprehensive review of literature highlighting sets of attributes and
sub-attributes that are potentially relevant to residential satisfaction decisions by the occupants
of low-income housing were identified. Though the reviewed literature was based on studies
from the developed countries, these were collectively used to examine the attributes that
determine residential satisfaction in subsidised low-income housing in South Africa. The
influence of the attributes on residential satisfaction was obtained as a product of the impact
on the housing occupants. The level of influence of the main attributes as categorized on the
questionnaire was established by assessing the extent to which the listed attributes will
determine the occupant’s satisfaction with their houses. Also, the impact of the sub-attributes
in determining residential satisfaction was assessed if they were present or lacking. The rating
was based on an ordinal scale of one to ten with one being low influence or no impact and ten
being high influence or very high impact. As mentioned above, in the previous chapter, the
levels of influence and impact were then obtained as a product of the consensus achieved as
detailed on Section 7.4.3.7.
From the nineteen (19) identified main attributes that determines residential satisfaction, only
two attributes (dwelling unit and location) were considered by the experts to have a high
influence based on the medium score of 9.0 to determine occupants housing satisfaction. The
IQD score for dwelling unit feature was 1.50 reflecting a low consensus. However, the dwelling
unit was still considered to have a high influence on the occupants in determining if they will
be satisfied or dissatisfied with their residential status. This is because the unit of residential
satisfaction measurement is primary the occupant’s, but the occupant’s objective and subjective
measurement of their situation is based on the dwelling unit. Likewise, ten other main
attributes, as detailed in Figure 8.1, were scored to have an average influence on the
determination of satisfaction based on the good consensus achieved as evident from the median
and IQD scores, respectively. The median scores were not more than 8.99, while the IQD was
329
less than 1.00. However, seven other listed attributes were measured as having low influence
on the determination of satisfaction, as a result of the weak consensus scores (Figure 8.1). The
median scores were less than 7.99, while the attributes IQD were more than 1.0 while others
were within the study IQD cut-off score of 1.0 required to achieve consensus.
Figure 8.1: Influence of core attributes on residential satisfaction in South Africa low-
income housing occupants
Furthermore, when the impact of the above listed characteristic sub-attributes were measured
in the determination of residential satisfaction, if they were present or absent, the following
was found. Of the 21 listed variables for the dwelling unit construct, none were scored to have
a very high impact (VHI: 9.00-10.00) in determining residential satisfaction, while only three
variables had a high impact (HI: 7.00-8.99) and 18 other variables were scored to have between
a low to medium impact (MI: 3.00-6.99). None was found not to have an impact in the
determination of residential satisfaction (Table 8.1). The IQD scores revealed that consensus
was achieved for a majority of the items (19) with a score of between 0.00 and 1.00. But,
consensus was not achieved for two features, which were the number of bedrooms (2.00), and
brightness and sunshine within the house (2.00). The number of bedrooms had a mean score of
7.00 and a standard deviation-SD (σx) of 1.69, while the brightness and sunshine within the
houses had a mean score of 5.40 and an SD (σx) of 1.06; revealing their variableness
9
7
6
8
6
6
7
7
6
8
8
7
6
4
5
9
7
8
7
1.50
0.50
0.75
0.00
2.00
0.00
0.00
0.50
0.00
0.50
0.00
0.00
2.00
0.00
0.50
0.00
0.00
0.00
0.00
0 2 4 6 8 10
Dwelling unit
Housing physical characteristics
Household or personal characteristics
Housing condition or quality of building
Social features
Economic features
Community services
Neighbourhood and environmental facilities
Culture
Tenure
Homeownership
Needs and expectation
Beneficiaries meaningful participation
Personality variables
Aesthetics
Location
Health (personal and environmental)
Safety and Security
Psychological factors
IQD ≤ 1
Median
330
considering the total number of responses.
Table 8.1: Dwelling Unit Attributes
Features M x̅ σx IQD
The number of bedrooms 8 7.00 1.69 2.00
The size and the location of the spaces (rooms) 7 6.67 1.68 0.50
Location of living room 5 4.73 1.10 0.50
Location of dining room 5 4.73 1.10 0.50
Location of kitchen 5 4.93 0.96 0.00
Size of the bedrooms 7 6.80 1.26 0.00
Size of the kitchen 6 6.00 1.20 0.50
Size of the bathrooms 5 5.13 1.06 0.50
Size of the wardrobes or closet 5 4.93 1.49 0.00
Size of the dining room 5 5.20 1.15 0.00
Space for children to play 5 5.33 0.98 0.00
Space for children to study 6 6.00 0.76 0.00
Balcony 3 2.87 0.99 0.00
Privacy within the house 6 5.80 1.61 0.50
Brightness and sunshine 6 5.40 1.06 2.00
Ventilation in the house 6 5.80 1.52 0.00
The floor level 5 4.60 1.24 1.00
Overall appearance of building 5 5.27 1.62 0.00
Interior design 5 5.07 0.88 0.00
Overall size of House 6 6.40 1.35 0.00
Washing room area 5 5.00 1.13 0.00
M = Median; x̅ = Mean; σx = standard deviation; IQD = Interquartile deviation
Likewise, the scores for the listed neighbourhood and environmental characteristics revealed
that from the 26 listed variables, four were scored to have a very high impact (VHI: 9.00-10.00)
in determining residential satisfaction, while, 15 variables had a high impact (HI: 7.00-8.99)
and 7 other variables were scored to have between low to medium impact (LI & MI: 3.00-
6.99). Conversely, none were found not to have an impact in determining residential
satisfaction (Table 8.2). In addition, the IQD scores revealed that consensus was achieved for
a majority of the items (24) with a score of between 0.00 and 1.00. Nonetheless, consensus was
not achieved for two elements; which are the parking facilities and police protection with an
IQD score of 1.50, which was more than the acceptable IQD score for the study. Furthermore,
the parking facilities had a mean rating = 4.67 and an SD = 1.68; whilst the police protection
331
factor had a mean rating = 7.40 and SD = 1.35; showing the level of consistency within the
experts’ rating of the factors.
Of the 18 listed aspects for the household characteristics, only one aspect was scored to have a
very high impact (VHI: 9.00-10.00) in determining residential satisfaction, whilst, 7 aspects
were rated to have a high impact (HI: 7.00-8.99) and 10 other aspects were scored to have a
medium impact (MI: 5.00-6.99). Equally, none were found to have between no to low impact
in the determination of residential satisfaction (Table 8.3). In addition, the IQD scores revealed
that consensus was achieved for 17 aspects with only one aspect (Tenureship of residence) not
achieving consensus.
Table 8.2: Neighbourhood and Environmental Characteristics
Features M x̅ σx IQD
Location of dwelling unit 8 8.20 1.32 1.00
Friends and neighbours 7 6.80 1.32 0.50
Closeness to workplace 9 8.60 1.18 0.50
Closeness to shopping areas 8 7.53 1.19 1.00
Closeness to schools 8 7.67 1.35 0.50
Closeness to hospitals/clinics 8 7.47 1.36 1.00
Closeness to the place of worship 7 6.47 1.41 0.50
Public transportation and services 9 9.07 0.70 0.50
Landscape of the neighbourhood 6 5.87 1.46 0.50
Adequacy of on-street parking (bays) 5 4.93 1.39 1.00
Parking facilities 4 4.67 1.68 1.50
Walkways and access to main roads 7 7.33 0.90 0.50
Privacy from other neighbours 7 6.53 0.99 1.00
Closeness to playground and other recreational
facilities 6 6.27 1.10 0.00
Street and highway noise 6 6.07 1.28 0.00
Smoke or odours 7 6.53 1.88 1.00
Street lighting at night 8 7.73 1.39 0.50
Secure environment 9 8.73 0.80 1.00
Physical condition and appearance of the
neighbourhood 7 6.73 1.16 0.00
General cleanliness of the neighbourhood 7 7.27 0.96 0.50
Proximity to Police services 7 6.80 1.32 1.00
Police protection 8 7.40 1.35 1.50
Incidence of burglary activities 9 8.53 0.99 1.00
332
Elderly centres 5 5.20 1.15 0.00
Community hall 7 6.67 0.90 0.50
Facilities for the disabled 6 6.27 1.53 1.00
Similarly, when the aspects of the occupants social features were assessed, findings revealed
that none of the aspects were perceived by the experts to have a very high impact (VHI: 9.00-
10.00), whilst 9 of the aspects were considered to have a high impact (HI: 7.00-8.99) in
determining residential satisfaction and 5 other aspects were scored to have a medium impact
(MI: 5.00-6.99). Conversely, none were found not to have an impact in the determination of
residential satisfaction (Table 8.4). In addition, the IQD scores revealed that consensus was
achieved for all aspects of the occupants’ social features, in relation to the impact rating.
Table 8.3: Household Characteristics
Aspects M x̅ σx IQD
Gender (sex) 5 4.93 1.00 1.00
Marital status 5 5.36 1.45 0.50
Employment and welfare 8 7.86 1.10 0.50
Number of children 7 7.14 1.17 1.00
Age 6 6.14 0.86 0.50
Occupation 6 6.43 1.40 1.00
Education 6 6.29 0.99 0.50
Household structure 6 6.36 1.39 0.50
Race 5 5.00 1.11 0.50
Ethnicity 6 6.14 0.77 1.00
Tenureship of residence 7 7.21 0.70 1.50
Payments for own house 8 7.71 0.73 0.50
Length of residency 8 8.14 0.66 1.00
Family income 8 8.07 0.73 0.00
The amount of rent 9 8.86 0.53 0.00
Location of last residence 6 5.93 1.44 1.00
Tenureship of last residence 6 6.14 0.66 0.50
Disability 7 6.67 1.32 1.00
Furthermore, when the aspects of the occupants building quality was assessed, findings
disclosed that none of the aspects were perceived by the experts to have a very high impact
(VHI: 9.00-10.00), while 9 of the aspects were considered to have a high impact (HI: 7.00-
8.99) and 11 others were scored to have a medium impact (MI: 5.00-6.99). Equally, none were
333
found not to have an impact in the determination of residential satisfaction (Table 8.5). The
IQD scores revealed that there was a consensus amongst the experts for a majority of the
building quality aspects, while only one aspect (clothes-line facilities) did not achieve
consensus according to the IQD benchmark set for the study; with an IQD score of 1.50.
Table 8.4: Social Feature Aspects
Aspects M x̅ σx IQD
Privacy from neighbours 7 6.73 0.96 0.50
Interactions with neighbours 7 6.93 0.80 0.00
Security around neighbourhood 8 7.87 0.74 0.50
Safety at home 8 8.00 0.93 0.00
Security provision of flat (collapsible, sliding
front gate, window burglary etc.). 8 7.80 1.01 0.00
Density of population 6 5.87 1.41 0.50
Freedom of choice 7 7.13 0.74 0.50
Social relations (social networks) 7 7.07 0.70 0.00
Adequacy of escape route in case of fire 5 5.40 1.35 1.00
Community attachment 7 6.93 0.62 0.00
Anticrime measures (report centres) 7 7.20 0.86 0.50
Special requirement for disabled 5 5.47 1.06 0.00
Accident situation 6 6.13 0.64 0.00
Community relations 6 6.40 1.18 0.00
Also, when the aspects of the occupants economic features were assessed, further findings
revealed that none of the aspects were alleged by the experts to have a very high impact (VHI:
9.00-10.00), one (1) aspect was considered to have a high impact (HI: 7.00-8.99) in determining
residential satisfaction and 2 were scored to have medium impact (MI: 5.00-6.99).
Figure 8.2: Economic Features
Findings also revealed that none were found not to have an impact in determining residential
6.4 6.67 7.14
0.99 1.23 0.951.00 1.00 0.75
0
2
4
6
8
Home value Neighbourhood socio-
economic status
Cost of living (Town-
wide)
Mean (x̅)
SD (σx)
IQD ≤1
334
satisfaction (Figure 8.2). Additionally, the IQD scores revealed that consensus was achieved
for all aspects of the occupants’ economic features.
Table 8.5: Building Quality Aspects
Aspects M x̅ σx IQD
The water pressure 6 6.27 1.22 1.00
The quality of exterior construction 7 7.00 0.65 0.00
The quality of walls 7 7.07 0.88 0.00
The quality of interior construction 7 7.00 0.93 0.00
The quality of the floors 6 6.33 1.11 0.00
The quality of the windows 6 6.40 1.18 0.00
The quality of the doors 6 6.47 1.19 0.50
The quality of the interior painting 6 6.13 0.74 0.00
The quality of the exterior painting 6 6.07 0.88 0.00
The quality of the plumbing works 7 7.20 0.86 0.00
The quality of the sanitary finishing 7 7.13 0.92 0.00
Functioning of the plumbing fixtures 7 7.27 0.88 0.00
Plumbing repairs 7 7.20 0.94 0.00
Electrical wiring quality 7 7.07 0.96 0.00
Rooms and others spaces lighting 6 6.33 1.18 0.50
Electrical fittings quality 5 5.80 1.52 1.00
Numbers of electrical sockets 6 6.07 1.03 0.50
Level of sockets 5 5.47 1.13 0.50
Clothes-line facilities 5 5.60 1.18 1.50
Overall quality of the unit 7 7.13 1.25 1.00
When the community and services provided by the government aspects were assessed, findings
revealed that one aspect was perceived by the experts to have a very high impact (VHI: 9.00-
10.00); 10 other aspects were considered to have a high impact (HI: 7.00-8.99), while, 5 other
elements were scored to have a medium impact (MI: 5.00-6.99). In relation to the other aspects
that have been assessed, thus far, none were found not to have an impact in the determination
of residential satisfaction (Table 8.6). In addition, the IQD scores revealed that consensus was
achieved for all aspects that were assessed.
In addition, when the aspects of the occupant’s personality variables were assessed, the findings
revealed that none of the aspects were considered by the experts to have a very high impact
(VHI). All three (3) assessed aspects were scored to have a high impact (HI: 7.00-8.99) in
335
determining residential satisfaction. None were rated to have between no impacts to medium
impact (Figure 8.3). Also, the IQD scores revealed that consensus was achieved for this
variable among the experts.
Figure 8.3: Personality Variables
Table 8.6: Community Services Provided by Government
Residential satisfaction attributes M x̅ σx IQD
Drainage system 6 6.27 1.22 1.00
Garbage and waste collection system 7 7.00 0.65 0.00
Owned houses 7 7.07 0.88 0.00
Security system 7 7.00 0.93 0.00
Fire protection 6 6.33 1.11 0.00
Maintenance and repair services 6 6.40 1.18 0.00
Convenience of bus and public
transportation 6 6.47 1.19 0.50
Electricity supply 6 6.13 0.74 0.00
Water supply 6 6.07 0.88 0.00
Telephone service 7 7.20 0.86 0.00
Handling of residents’ complaints 7 7.13 0.92 0.00
Management responds to necessary repairs 7 7.27 0.88 0.00
Housing department officials treatment of
beneficiaries 7 7.20 0.94 0.00
Housing department rules and regulations of
the development 7 7.07 0.96 0.00
Enforcement of rules by the Department of
housing 6 6.33 1.18 0.50
Participation by the community 5 5.80 1.52 1.00
Further, when the variables of the building aesthetics were also assessed, findings showed that
6.676.27
6.60
1.88 1.871.5
0.00
1.00
0.000
1
2
3
4
5
6
7
8
Mistrust of authority Negative emotions Pessimism
Mean (x̅)
SD (σx)
IQD ≤1
336
none of the aspects were considered by the experts to have a very high impact (VHI: 9.00-
10.00), nor high impact (HI: 7.00-8.99) in determining occupants housing satisfaction.
However, four of the five listed variables were observed to have a medium impact (MI: 5.00-
6.99), while one variable was scored to have a low impact (LI: 3.00-4.99) in shaping residential
satisfaction decisions. None were rated to have between no to medium impact (Figure 8.4).
Hence, the IQD ratings revealed that consensus was achieved for the variables between the
experts.
Figure 8.4: Aesthetics Variables
The assessment of the location factors revealed that one of the listed variables (nearness to
economic opportunities) was rated by the experts to have a very high impact (VHI: 9.00-10.00);
while three other factors were scored to have a medium impact (MI: 7.00-8.99) in determining
residential satisfaction.
Figure 8.5: Location Variables
None were rated to have a no; low and high impact (Figure 8.5). Nonetheless the IQD scores
5.73
5.2
6
4.87
4.4
1.53
1.15
1.31
0.52
0.83
0.50
0.00
0.50
0.00
1.00
0 2 4 6 8 10
Building forms
Building height
External appearance (compare
with others in the neighbourhood)
Entrance / lobby design
Colour of the building
IQD ≤1
SD (σx)
Mean (x̅)
6.07
8.47
6.13
6.60
9.00
1.33
1.06
1.19
1.3
0.76
0.50
1.00
0.00
1.00
0.00
0 2 4 6 8 10
Size of housing development
Ease of access by public transport
Appropriateness of site forerection of residential building
Nearness to slums
Nearness to economicopportunities
IQD ≤1
SD (σx)
Mean (x)̅
337
revealed that consensus was achieved for all variables amongst the experts.
In addition, when the health (personal and environmental) features were evaluated, findings
established that one factor was considered by the experts to have a very high impact (VHI:
9.00-10.00) in determining residential satisfaction; while four other variables were considered
to have a medium impact (MI: 7.00-8.99). None was rated to have a no; low and high impact
(Figure 8.6). The IQD scores revealed that consensus was achieved for a majority (4) of the
variables, while no consensus was reached for one of the items (cleanliness of public area)
having an IQD score of 1.50, which was more than the IQD cut-off needed to achieve
consensus.
Figure 8.6: Health (personal and environmental) features Variables
Lastly, when the experts were asked to list any missing sub-attributes which they thought had
not been addressed, which could also affect residential satisfaction, the following were the
listed attributes:
1. Size of dwelling land;
2. Suppression of vermin like rats;
3. Location of kitchen sink;
4. Temperature / insulation;
5. Presence of ceiling;
6. Quality of sewerage and garbage removals services;
7. Physical safety (in terms of protected drain covers; fenced off streams, rivers and
railway lines, as well as road safety issues especially with regards to children’s and
6.40
5.47
5.07
8.53
6.40
0.91
1.06
0.88
1.13
1.76
1.00
0.50
0.00
0.50
1.50
0 2 4 6 8 10
Adequacy of daylight distribution in
the units
Adequacy of natural ventilation in the
units
Acoustic quality in the units
Water quality (cleanliness, etc.)
Cleanliness of public areas
IQD ≤1
SD (σx)
Mean (x̅)
338
elderly’s safety); and
8. Beneficiary’s participation in the design of the low-income houses.
The above attributes (7) were thus rated by the experts panelist to have a very high impact
(VHI: 9.00-10.00) except for the suppression of vermin like rats, which was rated to have low
impact (LI: 3.00-4.99). However, the IQD scores revealed that consensus was achieved for all
listed variables, achieving the IQD cut-off set to reach consensus.
DSO2 - To determine the factors that makes subsidised public housing unsustainable in
South Africa
The impact of factors that makes subsidised public housing unsustainable in South Africa was
also evaluated. The issue of unsustainability of subsidised public housing in South Africa has
been recognised in many studies and in policy debate in the country. Literature and housing
commentators inform that the culture of outright public housing ownership without any
contribution from the beneficiary’s and the continual reduction in housing and infrastructure
budget plays a major role in the unsustainability of subsidised public housing in the country.
Despite the above realisation, it is also true that the current economic disparity between the
different race groups in the country and the high level of unemployment plays a major role in
the present public housing delivery system in the country. In view of the above, there is
therefore need for stakeholders to create an enabling environment that will not disadvantage
the poor in the country, yet have a public housing delivery system that will be sustainable for
a long time. In this study, it was intended that the factors which currently makeup the subsidised
housing delivery system, should be assessed. An evaluation was therefore conducted to identify
the factors, which make subsidised public housing unsustainable in South Africa.
The impact of a set of factors as identified from the literature review was rated by the expert
participants in order to achieve the study’s objective. The findings from the Delphi Survey
revealed that consensus was achieved for a majority (15) of the listed factors as the IQD scores
achieved the cut-off score set to reach consensus for the study. The housing backlog factor did
not achieve consensus with an IQD score of 1.50; despite the experts rating the factor as having
a very high impact (VHI: 9.00-10.00) in making subsidised housing unsustainable in the
country. The mean score for this factor was 8.40, while the SD (σx) = 1.12 (Table 8.7).
The impact score rating further revealed that 2 factors were alleged to have a very high impact
339
(VHI: 9.00-10.00), while 8 factors were considered to have a high impact (HI: 7.00-8.99) and
6 other factors having a medium impact (MI: 5.00-6.99).
Table 8.7: Factors that makes subsidised housing unsustainable in South Africa
Factors M x̅ σx IQD
High cost of construction 8 7.60 0.91 1.00
None involvement of the big contractors 6 5.80 1.32 1.00
No solution to general housing problem 7 7.27 1.16 0.50
Undesirable 6 6.27 1.44 0.00
Lack of skills to handle delivery demand 6 6.07 1.83 0.50
Lack of resources to handle the production rate 7 6.40 1.68 1.00
Maintenance cost to the beneficiaries 7 6.93 1.98 0.50
Lack of maintenance plan 6 6.20 1.32 0.50
Illiteracy of the beneficiaries 6 5.60 1.35 0.50
Lack of housing education 6 5.67 1.68 1.00
Growing unemployment 8 7.80 1.08 0.50
Political involvement 9 8.73 1.03 0.00
Housing backlog 9 8.40 1.12 1.50
Size of the national housing budget 7 7.13 1.81 1.00
Bureaucratic capacity 8 8.20 0.94 0.50
Problem of under-spending 7 7.27 0.96 0.50
Furthermore, when the experts were asked to state any factor, which they thought was not
listed, which could further impact the sustainability of low-income housing delivery in the
country; the following 22 factors were itemised:
1. Culture of entitlement;
2. Poor national planning regimes;
3. Lack of strategic management by government;
4. Corruption;
5. Delivery mechanisms;
6. Lack of knowledgeable personnel in the housing departments;
7. Partnership with the beneficiary;
8. Dependency on government;
9. Lack of project management skills;
10. Inadequate beneficiary participation;
11. Nepotism;
12. Undue process;
340
13. Unnecessary delay;
14. Price escalation;
15. Inflation;
16. Land market / land price;
17. NIMBYism (Not-in-my-backyard);
18. Continuity of unfulfilled high expectations;
19. Ignorance of bureaucrats dealing with housing;
20. Beneficiary’s needs assessment;
21. Coordination of housing delivery; and
22. Lack of protection for vulnerable recipients of subsidised low-income housing.
However, 15 of the experts remarked that corruption, partnership with the beneficiary’s,
participation and beneficiary’s needs assessment should be given prime attention in the
delivery of public housing for the low-income groups.
DSO3 - To identify the combination of housing policy instruments that will better serve
South African low-income housing group
The South Africa housing policy was based on international good practice during the early
1990s, but since then there has been a major shift in international trends in housing which have
also been captured in the various revisions that have been implemented in the country’s housing
policy strategy. The South African housing policy instruments contribute to the national
priority of restructuring South African society in order to address structural, economic, social
and spatial dysfunctionalities, thereby contributing to the Government’s vision of an
economically empowered, non-racial, and integrated society living in sustainable human
settlements. Also, it was designed to improve and contribute to the overall functioning of the
housing sector and in particular the low-income segment, insofar as subsidised low-income
housing is able to contribute to widening the range of housing options available to the poor.
Besides, in response to the South African government constitutional imperative, government
has in terms of the Housing Act, 1997 (Act No 107 of 1997) and Housing Codes of 2000
introduced a variety of programmes, which provided poor households access to housing
opportunities. The policy principles set out in the White Paper on Housing aimed to provide
poor households with houses, as well as basic services such as potable water and sanitation on
an equitable basis. But the limited resources available from the government have necessitated
341
the prioritization of the most vulnerable groups, as well as the provision of housing, security
and comfort to all over time. Based on this, the current study objective was positioned to assess
the combination of housing policy instruments that will better serve the low-income housing
groups in the country, as well as their influence on public housing delivery in South Africa.
In assessing the research objective, nine housing policy instruments were identified from
literature. From the experts’ responses, one policy instruments (Public Housing Subsidy
Scheme) was considered by the experts to have a high influence (HI: 9.00) in the delivery of
low-income housing in South Africa. However, there is a high level of variability from the
experts as revealed from the IQD score, which was rated as 2.00 beyond the set consensus
yardstick, reflected a non-consensus. Likewise, five other factors were considered to have an
average influence (AI: 7.00-8.99) on the low-income housing delivery, while three other factors
were alleged to have a low influence (LI: 0.00-6.99) on public housing delivery in South Africa
(Figure 8.7). Additionally, there was consensus amongst the experts on the five factors that
were consider to have an average influence on public housing delivery in the country, while
two factors with low influence also achieved the IQD consensus cut-off, with one not achieving
consensus, scoring 1.50 (Figure 8.7).
Further, when the experts were asked to state the combinations of policy instruments that would
better serve the need of the low-income housing; public housing subsidy scheme (x̅=7.80;
σx=2.11), social (medium-density) housing (x̅=6.67; σx=0.72) and Incremental housing
(x̅=6.47; σx=1.60) were scored as the instruments that would better serve the need of the low-
income housing group in the country. Subsequently, when the experts were asked to state any
policy instrument which they thought was not listed with an impact on public housing delivery
in the country and which could be combined with other existing policies, a majority (9) of the
experts commented that the listed policy instrument is more than sufficient to cater for the
housing needs of the low-income groups provided they are well administered. Whilst the other
six (6) experts inform that Public Private Partnership (PPP) in public housing delivery should
be considered, but, should be administered with care in order not to jeopardize the aim of low-
income public housing through an ill-advised PPP scheme.
342
Figure 8.7: Housing Policy Instruments
In addition to the above questions fed to the experts in assessing the Delphi Study objective
three, they were further asked the following questions. When they were asked if the failure of
the South Africa Housing Policy to adequately respond to the needs of the poor and low-income
groups is hinged on the dominance of the public housing subsidy scheme as the main housing
policy instruments; Figure 8.8 gives a summary of the experts’ responses.
Figure 8.8: Dominance of Public Housing (Subsidy Scheme)
The above findings revealed that 60.0% of the experts agreed that the failure of the South Africa
9.00
7.00
6.00
7.00
7.00
7.00
6.00
7.00
6.00
7.80
6.67
6.00
6.93
7.07
6.33
5.80
6.87
6.47
2.11
0.72
1.60
0.88
0.46
1.72
1.52
1.25
1.60
2.00
0.00
0.50
0.00
0.00
0.50
1.50
0.00
0.50
0.00 2.00 4.00 6.00 8.00 10.00
Public housing (subsidy schemes)
Social (Medium-density) HousingInstrument
Rent regulation
Allocation and rental policies in currentsocial housing
Support for the construction of new socialhousing flats
Housing allowances
Tax relief
Interest subsidies for homeownership
Incremental housing
IQD
SD (σx)
Mean (x)̅
Median
13.0% 13.0%0.0%
60.0%
13.0%
0.00%
20.00%
40.00%
60.00%
80.00%
100.00%
Strongly
disagree
Disagree No Opinion Agree Strongly
agree
Per
cen
tage
Agreement
343
Housing Policy to adequately respond to the needs of the poor and low-income groups is hinged
on the dominance of the public housing subsidy scheme, as the main housing policy
instruments.
When they were further asked if the failure is hinged on the lack of proper
consideration/attention being given to the other policy instruments; findings showed that 50.0%
of the experts were in agreement with this statement (Figure 8.9). Therefore, these findings
further supported the experts’ view that a mix of policy instruments would better serve the need
of the low-income groups than a single-minded focus on one policy instrument.
Figure 8.9: Lack of attention no other policy instruments
Additionally, the experts were made to assess the present South Africa National Housing
Programmes, as contained in the Housing Codes of 2000. This assessment was done in order
to identify and achieve consensus on the programmes that would improve housing delivery that
could be satisfactory to public housing beneficiaries. Findings revealed that the Integrated
Residential Development Programme (IRDP), a programme that has been given national
priority was scored (x̅ = 2.67; σx = 1.54; IQD = 0.50; Rank = 1) by the experts, as the ideal
national programme to serve the housing need of the low-income groups and would assure
housing satisfaction (Table 8.8). Likewise, findings from the assessment revealed that two
other programmes, the Social Housing Programme (SHP) and the Enhanced People’s Housing
Process (ePHP), which are also national priorities like the IRDP, were ranked second and third,
were perceived to better serve the need of the low-income groups. Further findings revealed
that consensus was achieved for the programmes as they all met the IQD cut-off of less than
one or equivalent to one score (IQD ≤1).
10.0%
20.0%
10.0%
50.0%
10.0%
0.00%
10.00%
20.00%
30.00%
40.00%
50.00%
60.00%
Strongly
disagree
Disagree No Opinion Agree Strongly
agree
Per
cen
tag
e
Agreement
344
Table 8.8: Preferred National Housing Delivery Programme that will better serve the
low-income groups
Housing delivery programmes Mean
(x̅)
SD
(σx)
IQD
≤1 Rank
Integrated Residential Development Programme
(IRDP) 2.67 1.54 0.50 1
Social Housing Programme (SHP) 2.93 0.70 0.00 2
Enhanced People’s Housing Process (ePHP) 3.00 0.88 0.00 3
Informal Settlements Upgrading Programme (UISP) 3.13 1.25 0.50 4
Institutional Subsidies 3.47 1.77 0.00 5
Community Residential Units (CRU) 4.20 1.15 0.00 6
Consolidation Subsidies 4.86 0.86 0.00 7
Emergency Housing Assistance 5.57 1.45 0.00 8
Rural Subsidy: Informal Land Rights 5.93 1.90 0.00 9
Farm Residents Housing Assistance Programme 7.14 1.03 0.00 10
Similarly, when the housing delivery models operated by the South Africa Human Settlement
Department were also assessed; in terms of the one that would best respond to the housing need
of the low-income groups, findings revealed that consensus was reached amongst the experts
for the listed models (Table 8.10), but the Social and Rental Housing model was scored and
ranked first (x̅ = 1.67; σx = 0.49; IQD = 1.00; Rank = 1) as the model that would better serve
the needs of the low-income groups (Table 8.9).
Table 8.9: Preferred National Housing Delivery Model
Housing Delivery Models Mean
(x̅)
SD
(σx)
IQD
≤1 Rank
Social and Rental Housing model (Programmes
facilitating access to rental housing opportunities,
supporting urban restructuring and integration, e.g.
Social Housing, Institutional Subsidies etc.)
1.67 0.49 1.00 1
Public Housing Model (through the provision of free
subsidy) 2.07 0.26 0.00 2
Incremental Housing model (Programmes
facilitating access to housing opportunities through a
phased process, such as Informal Settlement
Upgrading, Consolidated Subsidy, and IRDP etc.)
2.09 1.30 1.00 3
Self-help Housing Model 2.27 0.46 0.50 4
Rural Housing Model (Programmes facilitating
access to housing opportunities in rural areas - Rural
Subsidy: Informal Land Rights)
2.64 1.03 1.00 5
345
Whilst public housing (x̅ = 2.07; σx = 0.26; IQD = 0.00; Rank = 2) and incremental housing (x̅
= 2.09; σx = 1.30; IQD = 1.00; Rank = 3) delivery models were ranked second and third and
self-help housing ranked as fourth (x̅ = 2.09; σx = 0.46; IQD = 0.50; Rank = 4), as shown in
Table 8.9. This particular finding suggests displeasure with the public subsidy scheme
dominant model that is currently being used to provide housing to the low-income groups in
South Africa.
Subsequently, when the experts were asked to predict the pivotal context of the South African
housing policy in the next 10 years, findings revealed that the experts foresee a South African
housing policy that will be positioned to respond to the following:
1. Upgrading of informal settlements;
2. Spatial policy and spatial analysis of need;
3. Beneficiaries effective participation;
4. Capacity building of beneficiaries in order for them to build their own housing;
5. Affordability;
6. Homelessness of the poor;
7. Fairness of allocation of housing;
8. Access to adequate services (both urban and rural);
9. Provision of low-income rental stocks;
10. Delivery of sustainable and energy efficient low cost housing;
11. Embracing Public Private Partnership in housing delivery;
12. Construction of Community Residential Units;
13. Better quality control of housing construction tenders with less corruption; and
14. Acceptance of informal settlements as part of the urban areas and working with the
occupants to improve their lives and not their homes.
The above listed aspects predict pivotal contexts that are currently the typical debate of housing
policy shift in South Africa in order for the government to fulfill its constitutional goal of
housing the low-income groups.
From the foregoing findings, when the experts were asked if the SA housing delivery system
could be referred to as a developmental or welfare system, findings revealed that 100.0% of
the experts conceded that it is a welfare housing delivery arrangement (Figure 8.10).
346
Figure 8.10: South Africa Housing Delivery System
DSO4 - To identify the critical factors affecting the delivery of low-income housing and their
effects on beneficiaries’ residential satisfaction
Housing the poor is one of the major challenges that have besieged the government of nations
since the last decade of the twentieth century. The challenges are particularly acute in global
urban areas where populations are projected to grow from less than 300 million in 1950 to
almost 2 billion by the turn of the last century with an increase of more than 50 million every
year throughout the 1990s and an average growth rate of 3.4 percent per annum. Presently, the
major housing problem in South Africa and in other developing countries, is the shortage of
affordable accommodation for the urban poor; the low-income majority.
In South Africa, the challenge is that there has been an average population growth of 2.1 percent
per annum resulting in the population increasing by 10.4 percent or over 4.2 million people
between 1996 and 2001. Over the past 20 years South Africa’s population has grown rapidly:
1990 - 36.1 million, 2010 - 49.1 million and 2011 – 50.59 million (Statistics South Africa,
2011), as a result of high fertility and in-migration rates. However, although the population is
projected to continual growth in absolute numbers over the next 20 years, reaching 52.2 million
by 2030, the growth will be significantly slower than the past two decades. Although the South
African population is projected to continue to increase in size, at least until 2030, the annual
population growth rate has been declining since the early 1990s, and is projected to continue
declining at the rates of -0.412% due to the high HIV and AIDS infection rate. In the absence
of HIV and AIDS, the population growth rate is projected to be significantly higher, but
nevertheless also declining. AIDS thus slows down population growth in South Africa.
The country has also experienced a 30 percent increase in the absolute number of households,
0.0%
100.0%
0.0%0.0%
50.0%
100.0%
150.0%
Developmental Welfare No Opinion
Per
cen
tag
e
Agreement
347
where only a 10 percent increase was expected. Correspondingly, the housing backlog has
increased and current figures indicate that there is a shortfall of over 2.1 million dwellings.
Although the housing institutions and markets in South Africa have developed over the years,
the country still faces a huge backlog of housing needs. Likewise, the strong policy response
has failed to adequately provide housing for the low-income groups who are unable to access
housing by themselves. Therefore, the Delphi Study’s objective was modeled to investigate
from the experts who deal with housing in the country to highlight the factors affecting the
delivery of low-income housing in the country and their effect on the satisfaction of the
beneficiaries of the housing stock.
Results emanating from the study revealed that the following seven factors were considered
critical by the experts, which affect the delivery of low-income housing in South Africa and
eventually on their residential satisfaction. The factors are:
1. Limited budget (dwindling tax base) (VHI);
2. Appropriate policy to handle informal settlement (VHI);
3. Poor planning and coordination from national to local government levels (VHI);
4. Housing delivery mechanisms (HI);
5. Lack of active participation of beneficiaries in the development of housing (HI);
6. Enabling the poor to solve their own housing problem (MI); and
7. The inability of relevant state authorities to consult adequately with affected local
communities to seek joint solutions to the housing crisis (HI).
From the impact ratings of the factors, findings revealed that 3 of the factors have a very high
impact (VHI: 9.00-10.00), while 3 other factors were correspondingly considered to have a
high impact (HI: 7.00-8.99). One factor was rated to have a medium impact (5.00-6.99) whilst
none were rated to have either low or no impact. The IQD scores revealed that consensus was
achieved for all listed variables, achieving the IQD cut-off (IQD ≤1) score set to achieve
consensus. Also, the experts informed that all factors negatively impact on residential
satisfaction of the beneficiaries, as a result of the effect on the final housing product given to
them.
When they were further asked if the waiting time on the housing database has an impact on the
housing delivery and satisfaction of the low-income groups; findings showed that 67.0% of the
experts were in agreement with this statement (Figure 8.11).
348
Figure 8.11: Waiting Time on Housing Database
DSO5 - To predict the life span of the present South African public housing subsidy delivery
model
The first recorded use of the Delphi Research Design was Helmer’s 1952 project that engaged
seven expert panellists from varying scientific and operational disciplines focused on
predicting atomic bomb attacks on American targets within the United States from a simulated
view of a fictional Soviet war strategist. Linstone and Turoff’s (1975) seminal work and
collection of Delphi applications endorsed the use of the Delphi Methodology in prediction of
events. The Delphi Method is especially useful for long-range forecasting, as expert opinions
are the only source of information available; thus the forecasting tendency is one of the major
applications of the Delphi.
The Delphi Methodology embraces the ideology that the best forecast is one that comes from
a collective intelligence formulated through the focused collaboration of experts in the field of
study (Adler & Ziglio, 1996), such as in the current study. This is not to say that quantitative
statistical analysis is of small consequence within the Delphi approach. But the combination of
qualitative inquiry and statistical analysis tools support the Delphi Methodology, making it an
effective and flexible research approach for prediction of events. Hence, the present objective
was designed to solicit, from the expert panelist, when they foresee the government (National
Department of Human Settlement) will end the current model of free housing delivery to the
poor and low-income groups in the country.
Findings from the study revealed that 67.0% of the experts predicted the programme will be
terminated from 10 years and above time, whilst 13.0% simultaneously predicted the period, 8
and 5 years respectively. However, a further 7.0% did predict that the programme will never
0.0% 7.0% 0.0%
67.0%
27.0%
0.0%
20.0%
40.0%
60.0%
80.0%
Strongly
disagree
Disagree No Opinion Agree Strongly
agree
Per
cen
tage
Agreement
349
be stopped while all experts panelist do not see the programme coming to end so soon, as within
the next 6 months to 2 years’ period (Figure 8.12).
Figure 8.12: Prediction of Current Government Public Housing Subsidy Model
Furthermore, the experts that commented that the programme will never be stopped remarked
upon the high incidence of inequality and unemployment, which has resulted in a high, level
of poverty within the previously disadvantaged groups, which was a major reason for this
programme to continue. Hence, they further informed that there is no reason to stop the free
housing delivery model. While another expert informed that the current housing subsidy
scheme might not continue, but some sort of government involvement in free housing for the
poor would continue.
Furthermore, when the experts were asked to predict what they envisage will be the
replacement for the current government housing subsidy scheme when it is eventually stopped
in the near future; the following is a summary of the experts’ predictions:
1. Market and subsidy mix (social housing-type models);
2. Assisted Self-help housing;
3. Upgrading of informal settlement;
4. Providing leasing agreements with informal settlers;
5. Community housing;
6. Community based initiatives;
7. Private buying and financing;
8. Rent-to-buy scheme; and
9. Public Private Partnership in housing delivery.
0.0%
0.0%
13.0%
13.0%
67.0%
7.0%
0.0% 10.0% 20.0% 30.0% 40.0% 50.0% 60.0% 70.0% 80.0%
6 months’ time
2 years’ time
5 years’ time
8 years’ time
10 years and above
It will never be stopped
350
The above predictions show that the housing need of the low-income groups is beyond the
current (Subsidised Housing Delivery Model) housing delivery dominant model in place to
adequately respond to the need of the low-income housing groups.
Subsequently, the experts were further asked if state subsidised (financed) housing will always
be the major housing delivery model to provide housing to the poor and low-income group.
Findings revealed that 50.0% of the experts disagreed with the statement, while, 29.0% agreed
that it will be the major model of housing delivery and another 14.0% strongly agreed with the
statement (Figure 8.13).
Figure 8.13: State Subsidised Public Housing being the Major Delivery Model
DSO6 - To evaluate the management issues affecting the national, provincial and local
government housing agencies in the delivery of housing in South Africa
In additional to the previous Delphi Study objective (DSO-5), DSO-6 investigates the current
management issues affecting the national, provincial and local government housing department
and agencies in the provision of subsidised low-income housing. From the list of factors as
itemized by the experts, a summative content analysis, which involved the counting of the
frequency of word occurrence and comparisons of keywords, helped in the classification of the
management issues. This minimized redundancy amongst the listed issues on a scale of 1-10;
1 being most important and 10 being least important. The summarized management issues are
as summarized in Table 8.10.
Of the 29 listed management issues as shown in Table 8.11, government capacity to facilitate
development and corruption in provincial and local government departments were ranked as
the most critical issues facing the national, provincial and local government in the provision of
7.0%
50.0%
0.0%
29.0%
14.0%
0.0%
20.0%
40.0%
60.0%
Strongly
disagree
Disagree No
opinion
Agree Strongly
agree
Per
cen
tage
Agreement
351
low-income housing based on their higher number of occurrence. Whilst, budget constraints
and poor political will were likewise ranked second as the besieging issues facing government
in the provision of low-income housing. The cost of basic services such as water and electricity
in low-income houses and urbanization were ranked as ninth and tenth management issues
facing the government; being the least problems to low-income housing delivery.
Table 8.10: Current national, provincial and local government housing development
issues
Current management issues hindering housing delivery
Rank
Government capacity to facilitate development 1
Corruption in provincial and local government departments 1
Budget constraints 2
Poor political will 2
Access to well-located land for housing (availability of vacant land) 3
Legislature (planning, environmental and heritage) 3
Housing backlog 3
The current structure of government (provincial government sphere is not
necessary)
4
Informal settlements 5
Government bureaucracy in the housing system 5
Absence of reflection or serious acknowledgement of problems that affect the poor 5
Sustainability 5
Passive beneficiaries (entitlement culture) 5
Lack of technical skills (engineers and artisans, etc.) 5
Unaccountability of government employees 5
Lack of national vision for housing 5
Poor planning and coordination 5
Ideological constraints 5
Lack of participation from communities and CBOs 5
Limited participation of the private sector’s involvement in low cost housing
development
5
Public Private Partnership in housing delivery 6
Lack of project management skills / ineptitude 6
Better regulation and quality control of housing construction tenders 6
Nepotism and cadre deployment 7
Salary and benefit expectations of executives 7
Bloated and duplicated bureaucracies 8
Lack of appropriate skills 8
Cost of basic services (water, electricity etc.) in low-income houses 9
Urbanization 10
Lack of national vision for housing, which was ranked 5th by the experts, does not adequately
352
represent the government’s effort in fulfilling its constitutional mandate. From the provision of
the South Africa primary legislations dealing with the provision of housing such as the 1996
Constitution of the Republic of South Africa; Prevention of Illegal Eviction from and Unlawful
Occupation of Land Act 19 of 1998 (PIE Act); Housing Act 107 of 1997; Housing Amendment
Act 28 of 1999 amongst others and the Secondary legislations, such as the Expropriation Act
63 of 1975; Sectional Titles Act 95 of 1986 (amended by Acts 24 and 29 of 2003); Land Titles
Adjustment Act 111 of 1993 (LTA); Land Reform (Labour Tenants) Act 3 of 1996; Interim
Protection of Informal Land Rights Act 31 of 1996 amongst others, reveals the government’s
comprehensive vision to adequately house the poor and low-income groups.
Consequently, the experts were further asked to list and rank the management issues that would
affect the national, provincial and local government housing departments and agencies in the
next 10 years or in the near future. From the list of issues as garnered from the experts, a
summative content analysis which also involved the counting of the frequency of words
occurrence and comparisons of keywords was also perform as on the previous question; thus
minimizing redundancy amongst the listed issues. This was done on a scale of 1-5; 1 being
most important and 5 being least important. The condensed management issues are as
summarized in Table 8.11.
Table 8.11: Forecasted national, provincial and local government housing development
issues
Future management to hinder housing delivery
Rank
Budget constraints 1
Government capacity to facilitate development 1
Corruption in provincial and local government departments 2
Urbanization 2
Supply of water and electricity 2
Housing backlog 2
Poor political will 2
Informal settlements 2
Lack of technical skills (engineers and artisans, etc.) 3
Challenges from the dissatisfied public, including legal cases 5
The current structure of government (provincial government sphere is not
necessary)
3
Availability of vacant land 3
Legislature (planning, environmental and heritage) 4
Lack of accountability from government employees 4
353
Passive beneficiaries (entitlement culture) 4
Poor planning and coordination 4
Ideological constraints 5
Environmental impact of subsidy housing in South Africa 5
Limited participation of the private sector’s involvement in low cost housing
development
5
Sustainability of low cost housing programmes 5
Public Private Partnership in housing delivery 5
Willingness to work with flexible, multiple housing delivery programmes suited to
local circumstances
5
Beneficiaries’ expectations 5
From the summative factors as listed in Table 8.11, budget constraints and government capacity
to facilitate development were ranked as the most critical issues that the national, provincial
and local governments would face in the next 10 years and or in the near future in the provision
of low-income housing. Whilst, corruption, urbanization, supply of water and electricity,
housing backlog, poor political will and informal settlement were ranked second, as the issues
that will overwhelm the government in the provision of low-income housing in the future. The
present ranking of factors, such as urbanization and supply of water and electricity were in
disparity with the previous question rankings of ninth and tenth on the list of current issues
facing the government in the delivery of low-income housing.
Figure 8.14: Impact of unemployment on low-income housing delivery and eventual
satisfaction of the low-income Groups.
In addition, the experts were asked if the current unemployment situation in the country had an
impact on low-income housing delivery and eventual satisfaction of the low-income groups.
The Delphi Survey findings revealed that 40.0% of the experts agreed with this statement,
27.0% strongly agreed, while 20.0% disagreed, as shown in Figure 8.14.
13.0%20.0%
0.0%
40.0%
27.0%
0.0%10.0%20.0%30.0%40.0%50.0%
Strongly
disagree
Disagree No opinion Agree Strongly
agree
Per
cen
tage
Agreement
354
DSO7 - To determine the influence of beneficiary participation on their overall housing
satisfaction
Beneficiary participation has been an issue in housing development projects since the
emancipation of the new South African state; but its significance has increased since it
principally became part of the official rhetoric in most of the developed country’s policy
documents, post-1994. Though, the criticism of development projects is widespread, and blame
for disappointing results is cast in many directions. One line of criticism, which has become
quite strong in developmental literature, is that developmental projects are too top-down and
need to be more bottom-up. It is generally agreed that housing projects should involve more
participation by the beneficiaries.
Beneficiaries’ participation in the low-income housing delivery process is presumed as the
direct involvement of beneficiaries’ in the construction of their own houses. Such participation
can constitute contribution by beneficiaries and empowering them eventually. But the baseline
is the satisfaction of the beneficiaries with the project. Hence, when the experts were asked if
beneficiary participation in the housing development process could potentially lead to the
implementation of appropriate responses through the assessment of their needs and
expectations and eventual housing satisfaction; 73.0% agreed that it will result in the outcome.
Whilst 13.0% simultaneously responded that they strongly agree and strongly disagree (Figure
8.15).
Figure 8.15: Effect of beneficiary participation in residential satisfaction
DSO8 - To determine the effect of meeting beneficiary’s housing needs and expectation on
their overall housing satisfaction
As part of the conceptual framework of residential research, the gratification of occupants
housing needs and expectations should have noteworthy prominence. For people with different
housing needs and expectations, the same housing condition could bring different satisfaction
13.0%0.0% 0.0%
73.0%
13.0%
0.0%20.0%40.0%60.0%80.0%
Strongly
disagree
Disagree No opinion Agree Strongly
agree
Per
cen
tage
Agreement
355
levels because their needs and expectations are different. Also, residential satisfaction is
basically formed under the condition of what level of housing needs is currently being pursued
by the occupants (Yiping, 2005). Unless, level one need is sufficiently satisfied, they will
remain in the occupant’s consciousness and will thus; become the prime determinants of
housing behaviour. In earnest, the living condition that is currently being pursued forms the
housing expectation of the individual, which is highly related to overall residential satisfaction.
Therefore, when the experts were asked if prior exposure to what is to be received has a
tendency to influence beneficiaries’ satisfaction towards a given housing product; 80.0%
respondents’ said that they agree, while, 13.0% inform they strongly agreed as shown in Figure
8.16.
Figure 8.16: Beneficiary prior exposure to housing
Furthermore, in response to the research objective, experts where thus asked which of the listed
housing needs order should be met in order to satisfy the South Africa subsidised low-income
housing beneficiaries. Findings, as shown in Figure 8.17, revealed that 27.0% of the experts
concurrently agreed that esteem and physiological needs should first be met in order to satisfy
the low-income housing beneficiaries; whilst 18.0% consecutively informed that social and
safety needs should follow the initial housing order needs. Further result analysis revealed that
consensus was achieved for the experts’ classification of South African low-income housing
order needs, as they all met the IQD cut-off (IQD ≤1) score.
0.0% 0.0% 7.0%
80.0%
13.0%
0.0%20.0%40.0%60.0%80.0%
100.0%
Strongly
disagree
Disagree No
Opinion
Agree Strongly
agree
Per
cen
tage
Agreement
356
Figure 8.17: Low-income beneficiary housing order needs
In addition, when the questions of housing being the paramount need of the low-income groups
in South Africa was asked, findings revealed that 40.0% strongly agree that housing is the
paramount need of the poor and low-income groups, whilst, 33.0% agree to the statement.
However, a further 13.0% concurrently disagreed and strongly disagreed that housing is not
the paramount need of the poor and low-income groups in South Africa (Figure 8.18).
Figure 8.18: Housing as paramount need of the poor and low-income groups in South
Africa
Therefore, when the beneficiaries that disagree and strongly disagree with the above statement
were asked to state what the paramount needs of the poor and low-income groups in South
Africa should be, below is a list of needs as listed by these experts. From the listed themes by
the experts, a summative content analysis which involved the counting of frequency of words
occurrence and comparisons of keywords was performed, thus minimizing exclusion of
relevant issues. Likewise, the experts were further asked to rank the most responsive need
amongst the listed paramount needs. This was done on a scale of 1-5; 1 being most important
and 5 being least important. Also, the level of urgency was likewise asked based on the listed
27.0%
27.0%
18.0%
18.0%
9.0%
0.0% 10.0% 20.0% 30.0%
Esteem Needs
Physiological
Social Needs
Safety Needs
Self-actualisation
Percentage
Hou
sin
g n
eed
s ord
er
13.0% 13.0%
0.0%
33.0%40.0%
0.0%
10.0%
20.0%
30.0%
40.0%
50.0%
Strongly
disagree
Disagree No
opinion
Agree Strongly
agree
Per
cen
tage
Agreement
357
needs, on a scale of very urgent, urgent and not urgent. The summarized paramount needs aside
from housing are as shown in Table 8.12.
Table 8.12: Paramount need of the poor and low-income aside housing
Paramount needs
Rank
Level of Urgency
Employment 1 Very urgent
Education 1 Very urgent
Job security 1 Urgent
Food security 2 Urgent
Clothing 3 Not urgent
Transportation 4 Urgent
Health security 5 Very urgent
From the summation of needs, as listed in Table 8.12, employment, education and job security
were ranked as the paramount needs of the poor and low-income groups in South Africa. Food
security was ranked second, followed by Clothing which was ranked third, while transportation
and health security were ranked fourth and fifth, respectively. Further findings from the
categorization of the urgency of the delivery of the paramount needs; revealed that
employment, education and health security were listed as very urgent needs, whilst job security,
food security and transportation were listed as urgent and clothing of the poor and low-income
groups were ranked as not urgent.
8.4 DISCUSSION OF DELPHI RESULTS
Objective DSO1
The first objective of the Delphi Study was to establish the attributes (main and sub) that brings
about residential satisfaction and to examine if the attribute that determine satisfaction in other
cultural contexts is the same within South Africa. Findings emanating from the survey revealed
that the attributes that bring about residential satisfaction in South African low-income housing
are similar to other cultural contexts. However, consensus was not achieved for the dwelling
unit attributes, as these were major attributes that had been found to be a strong determinant of
housing satisfaction in other cultural contexts (Amerigo & Aragones, 1997; Carvalho et al.,
1997; Caughey et al., 1998; Francescato, 2002; Mohit et al., 2010; Salleh et al., 2011). Though,
it was rated to have a very high influence in determining occupants housing satisfaction. Also,
there was disagreement on the attributes of social features and beneficiaries’ meaningful
358
participation in having a significant influence in determining residential satisfaction. Both
factors had an IQD rating of 2.00, which was higher than the IQD cut-off for achieving
consensus as set for the study and a medium score of 6.00. Likewise, the standard deviation in
the level of influence for the three factors was found to be between 1.29 - 1.45. This value is
very high and indicates a significant level of variability considering the number of values in
question. This inconsistency shows that even though these factors are relevant to determine
residential satisfaction, the expert panellists do not view them as such, in South Africa.
In addition, the assessment of the sub-attributes of the 19 major determinants of residential
satisfaction, as assessed, also revealed that the sub-attributes that determine residential
satisfaction in South African low-income housing are similar to other cultural contexts.
However, consensus was not achieved for some sub-attributes, which had been found to
strongly determine satisfaction in other cultural contexts. Prime amongst these were the number
of bedrooms, a variable under the dwelling unit features, which was rated as having a high
impact on satisfaction, but with a high SD value of 1.69 revealing the variability about the
factor by the experts’ rating. Likewise, other factors within the dwelling unit features which
were rated as having medium impact such as the location of the dwelling spaces, location of
living space, location of dining rooms, ventilation in the house, overall appearance of the
building amongst others suggest that they were thus rated because of the lack of consultation
from the occupants as predetermined decisions were already taken with regards to the typology
of the subsidised houses. This is because the design of the houses are already in existence
(approved) and hence, the local government responsible for the development only carry out
their mandate as recommended by the provincial and national government departments to
develop new housing when a suitable land is available.
Furthermore, the assessment of the sub-attributes of neighbourhood and environmental
characteristics, findings replicated the results of a majority of studies on housing satisfaction
in relation to the neighbourhood attributes (Amerigo, 2002; Johnson & Abernathy, 1983;
Marans & Rodgers, 1975; Ukoha & Beamish, 1997), upon which the current study’s sub
attributes were also based. Closeness to workplace, public transportation and services and the
incidence of burglary activities were rated as variables with a very high impact as these impacts
on the quality of life of the occupants. Also, police protection was rated to have a high impact
in relation to the security and safety of the occupants. For instance when a neighbourhood is
not crime free, but there is the presence of police protection in the form of dedicated police
359
postings in the neighbourhood, occupants feel safe and are thus, satisfied with the
neighbourhood. Likewise, when the distance occupant’s commute to their place of work and
the availability of public transport is satisfactory, this will also satisfy them with regards to
their neighbourhood.
The assessment of the sub-attributes of the household features revealed that household
characteristics were classified by the experts as having a high level of impact in determining
residential satisfaction of the occupants, as consensus was achieved for all the listed attributes.
This revealed that residential satisfaction is greatly influenced by the household disposition.
Likewise, similar findings were displayed by the social feature variables, as all variables
achieved consensus and the lowest median score was a medium impact. Despite the features
being scored, as having an average influence from the assessment of the main attributes, the
sub-attribute assessment revealed that low-income housing occupants value privacy from
neighbours, desire interactions with neighbours, want security around their neighbourhood and
want to be safe at home. This finding agreed with previous findings from the studies done by
Mohit et al. (2010), Salleh et al. (2011), Ukoha and Beamish (1997), Landman (2004), Lu
(1999) and Li (2002). Further findings as related to the economic features imply that the sub-
variables play an active role in occupant’s residential satisfaction.
Besides, further findings relating to the building quality variables revealed the importance of
the sub-factors, such as the quality of walls, quality of interior and exterior construction
amongst others, which impact residential satisfaction. The level of sockets and clothes-line
facilities were rated to have a medium impact (MI: 5.00), which concurs with Salleh’s et al.,
(2011) study and Ukoha and Beamish (1997). Other sub-variables that were found to have a
very high impact in the determination of residential satisfaction were: community services
provided by the government; personality variables; location variables and health (personal and
environmental) feature variables. With regard to the influence of the location variables on
satisfaction, expert ratings revealed that occupant’s satisfaction depends a great deal on the
nearness to economic opportunities, appropriateness of site for erection of dwelling and ease
of access to public transportation. However, a variable that was found to have a low influence
on satisfaction were the aesthetics variables, such as the colour of the building, entrance / lobby
design, building height and the building forms, which concurred with Amole (2009) and
Aragones et al., (2002) study findings.
360
In conclusion therefore, the results seem to suggest that the attributes that bring about
residential satisfaction in South Africa low-income housing are similar to the determinants in
other cultural contexts. Further, residential satisfaction is assured if there is a consideration of
these factors in the development of subsidised low-income housing for the poor in South
Africa. Of particular importance are the factors of dwelling unit, housing physical
characteristics, household or personal characteristics, building quality features, community
services, neighbourhood facilities, needs and expectation and beneficiaries meaningful
participation, which have all been described as being of significant influence and having a high
impact in determining residential satisfaction.
Objective DSO2
The second objective of the Delphi Study was to establish the factors that make subsidised low-
income housing unsustainable in South Africa. This is because the improvement of adequate
housing rights and the quality of life for the poor and low-income groups in South Africa can
only be assured when the factors and challenges faced by the government in the provision of
housing in South Africa are identified.
The impact significance of a set of factors as identified from literatures were evaluated by the
experts as individual factors on a scale of 1-10; with 1-2, being no impact and 9-10, being very
high impact. The identified factors from the experts’ ratings that make subsidised low-income
housing unsustainable in South Africa are:
1. Housing backlog (VHI: 9.00);
2. Political involvement (VHI: 9.00) ;
3. High cost of construction (HI: 8.00);
4. Growing unemployment (HI: 8.00);
5. Bureaucratic capacity (HI: 8.00);
6. No solution to general housing problem (HI: 8.00);
7. Lack of skills to handle the production rate (HI: 8.00);
8. Maintenance cost to the beneficiaries (HI: 8.00);
9. Size of the national housing budget (HI: 8.00); and
10. Problem of under-spending (HI: 8.00).
Two other factors: (i) none involvement of the big contractors (MI: 6.00); and (ii) lack of skills
to handle the delivery demand (MI: 6.00), that were assessed to have a medium impact are the
361
most noticeable issues in the South African housing space. But these were considered by the
expert panellist’s to only have medium impact. This therefore indicates that when the small
and medium contractors are given the necessary support, attention and the right environment
to grow by the government, they will acquire the required capacity to meet the demand and
supply of low-income housing in South Africa. In additional, the government discussion
document on poverty reduction strategy, which came out in May 2008 refers to housing
delivery as asset transfer, representing a central component of the emerging approach (Toward
an anti-poverty strategy for South Africa: a discussion document). Together with the various
measures to invest in human capital and service infrastructure, the strategy document identifies
the housing asset as indispensable to economic participation for the poor. Stating that housing
access will lead to improved economic and social security providing economic engagement for
the citizens in the long run.
Also, the housing backlog was scored to have a very high impact (VHI), which is regarded as
one of the prominent housing delivery problems in South Africa. Moreover, subsidised low-
income housing was established by the government as a solution to the problem of housing
backlogs, which affect mostly black and other previously disadvantaged South Africans.
Besides, consequences of housing backlog are physically reflected in overcrowding, squatter
settlements and increasing land invasions in urban areas, and poor access to services in rural
areas. From a social and political view, housing backlog gives, both individual and communal
insecurity and frustration on a daily basis, and contributes significantly to the high levels of
criminality and instability prevalent in many communities in South Africa. Similarly, due to
the high rate of population growth and low rate of housing provision, it is estimated that the
housing backlog in the country is increasing at a rate of around 204 000 units per annum (South
Africa Yearbook, 2011) which is a serious challenge in the provision of low-income housing
in the country. However, in 2009, the DHS admitted that the data it relies on to estimate the
housing backlog in South Africa is most likely unpredictable. Consequently, in terms of
eliminating the housing backlog and delivering adequate housing to the low-income groups
especially, the department is not really clear where it stands. Therefore, the DHS further
indicated that in relation to its statistical collection and verification, they are hoping that the
2011 Census will give the department a better sense of the accuracy of the data they will need
to measure backlogs and access delivery. It is presumed that the estimated data might have
been under-estimated or over-estimated, thus revealing either a limited or over-blown idea of
the housing backlog situation.
362
Also, other factors as stated by the experts with very high impact, which contributes to
subsidised low-income housing sustainability in South Africa, are:
1. Culture of entitlement (VHI: 9.00);
2. Poor national planning regimes (VHI: 10.00);
3. Lack of strategic management by government (VHI: 9.00);
4. Corruption (10.00);
5. Delivery mechanism (VHI: 9.00);
6. Lack of adequately knowledgeable personnel in the housing departments (VHI: 9.00);
7. Partnership with the beneficiary (VHI: 10.00);
8. Dependency on government (VHI: 10.00);
9. Lack of project management skills (VHI: 10.00);
10. Inadequate beneficiary participation (VHI: 10.00);
11. Price escalation (10.00); and
12. Inflation (VHI: 9.00).
From the above listed factors of unsustainability, poor national planning regimes, corruption
and dependency on government for housing provision are issues that have also taken a centre
stage in the South African housing space. But, it should be clearly stated that these factors
correlate with the factor of the lack of adequately knowledgeable personnel in the housing
departments. Also, factors like beneficiary’s needs assessment, which was rated to have a low
impact in making subsidised housing unsustainable in the country, can be argued to have a very
high impact as the needs of the low-income groups should be known before houses are
designed, which should meet these needs.
In conclusion therefore, the results seem to suggest that the factors that make low-income
housing unsustainable in South Africa are inter-linked as some of the factors are a result of the
looseness from the others. Since the provision of low-income housing is a constitutional
mandate for the government, it is essential that these factors are known, the provision of low-
income housing to house the poor is assured if there is an adequate knowledge of the limiting
factors, which will enable the right solutions to be profiled. Of particular importance are the
factors of poor national planning regimes; corruption; dependency on government for housing
provision; housing backlog; political involvement; high cost of construction and growing
unemployment, which have been described as having a very high impact in making subsidised
363
low-income housing unsustainable in South Africa and thus, a key leading indicator for better
delivery of low-income housing.
Objective DSO3
The third objective of the Delphi Study was to identify the combination of housing policy
instruments that will better serve the South African low-income housing groups.
Housing policy instruments in South Africa are a set of systems by which the government
exercises their influence in an attempt to ensure support and effect or prevent the provision of
affordable housing to its citizens. Over time, a range of housing policy instruments have been
developed to respond to policy imperatives of low-income housing delivery in the country.
These housing policies instruments and models have been used as an attempt to try and address
housing problems, especially with respect to the low-income earners, with the view of helping
them to achieve access to adequate housing. From this perspective, this objective attempts to
establish the influence of the various housing instrument’s and their combinations towards low-
income housing development in South Africa. This is because in the past two decades, the
failure and the irresponsiveness of various housing policy instruments has hindered housing
delivery models and instruments to respond to the sheer scale of need of the urban poor raising
the question of whether the time has arrived to revisit the housing approaches currently in place
to more realistic approaches to meet the housing needs of the urban poor.
Nine housing policy instruments were identified from literature besides the established policy
instruments being used in South Africa, which were assessed in this study’s objective. These
were assessed in terms of their influence in serving the low-income housing need. Also the
combinations of the instruments that will best serve the low-income were assessed. The
identified instruments from literature include: incremental housing, interest subsidies for home
ownership, tax relief, housing allowance, social housing instrument, public housing subsidy
scheme amongst others (See Figure 8.7).
Findings revealed that incremental housing which is a highly projected policy instrument by
the Department of Human Settlement was not given the same accord by the expert panelist.
Despite incremental housing could help to reduce the ever growing housing budget by
recognizing that poor urban families can build and extend their own dwellings incrementally
in response to their needs and the availability of resources.
364
The basis of incremental housing being pushed by the South Africa DHS is to support low-
income families by providing appropriate legal and technical support in order for them to house
themselves. This could happen through the environmental upgrading of existing informal
settlements with adequate social services, such as safe water, sanitation, drainage, electricity
and access ways and provision of recognised title to new plots of serviced land – sites and
services on which households could build their own dwellings. Despite this background,
incremental housing is not the preferred method of housing delivery in the country; this was
evident from the experts’ assessment. Public housing (subsidy scheme) being the major method
of housing delivery to the poor in the country was therefore appraised by the experts as having
a very high impact in the delivery of low-income housing in the country. From literature and
DHS review documents, public housing (subsidy scheme) currently has the greatest impact in
housing delivery in the country being the major policy instrument of delivery. Regardless of
the projection of public housing subsidy scheme as the dominant policy been used, the experts
could not reach a consensus, which signifies a level of variability as further confirmed by the
SD score of 2.11, which is very high considering the number of responses.
Further findings revealed that the combinations of policy instruments, which will better serve
the need of the low-income groups as informed by the experts, are public housing subsidy
schemes, social (medium density) housing and incremental housing. This thus inferred that a
single policy instrument is not sufficient to respond to the needs of the low-income groups, and
a combination of different policy instruments would be better adapted to suit different low-
income groups across the country. Beside the combination of the policy instrument, findings
also revealed that PPP was also given a nudge as an option for low-income housing delivery,
with an undertone of thoughtfulness in order to meet the need of the poor.
Additional findings also revealed that the failure of the South African housing policy to
adequately respond to the need of the low-income groups is hinged on the dominance of the
subsidy scheme, being used as the main delivery instrument. This confirmed the above
statement that the reliance on a single policy instrument cannot adequately respond to the needs
of the low-income groups. The researcher is of the opinion that if the current South Africa
housing backlog which was raised as a problem in the delivery of low-income houses in the
country (Delphi Study Objective two, DSO-2) will be eradicated, the DHS must vigorously
implement more than one policy instrument as it delivery vehicle.
365
Also, the analysis of the National Housing Programmes as contained in the Housing Code of
2000, revealed that IRDP, a programme that was established to facilitate the development of
integrated human settlements in well-located areas that provide convenient access to urban
amenities, including places of employment and creation of social cohesion and SHP and ePHP
were scored as the ideal national housing delivery programmes that will better serve the need
of the poor. SHP facilitates the provision of secure, stable rental tenure for the lowest income
persons who are not able to be accommodated in the formal private rental and social housing
market. While ePHP is aimed at delivering better human settlement outcomes (at household
and at community level) based on community contribution, partnerships and the leveraging of
additional resources through partnerships. The foundation of this programme (ePHP) is based
on the achievement of developing livelihoods interventions leading to outcomes, such as job
creation, developing a culture of savings, skills transfer, community empowerment, building
of community assets and social security, and cohesion. The ePHP model is an improved version
of the basic incremental housing approach system with the element of empowerment. The
experts’ assessment was slightly skewed from the recommendation of the DH. DHS have
preferred the combination of IRDP, SHP and Informal Settlement Upgrading Programmes as
the ideal national housing delivery programmes. However, the finding supports numerous
research findings (Aigbavboa, 2010; Charlton, 2006; 2009; Ogunfiditimi, 2008; Nobrega,
2007). Synopsis from the referenced authors informs that housing should not only be provided
for the low-income groups without their involvement and empowerment to give them a head
start into the wider economy. The inclusion of IRDP reinforced one of the key lessons learnt
in the review of the outcomes of housing programmes since 1994, which owing to a variety of
reasons placed low income settlements on the urban periphery without the provision of social
and economic amenities and by and large constituted housing subsidised beneficiary islands.
This research finding was further supported by the outcome of the preferred housing delivery
model, which will better respond to the need of the poor. Once again, social housing, the public
housing subsidy and incremental housing models were rated as the preferred model to
adequately respond to the needs of the poor. This also infers that the government cannot be
held solely responsible for housing the poor and the low-income groups despite being the major
role players. But other housing delivery models should be run concurrently with the existing
model to serve a variety of needs in adequately housing the low-income groups. Findings also
revealed that the need of the urban poor is still better projected, as compared to the rural poor.
366
The rural housing model was assessed the least in the ranking score with a high level of
variability as evident from the SD score of 1.03.
Further findings relating to the envisaged pivotal context of the South African housing policy
in the next 10 years revealed that a South African housing policy that will respond to the
upgrading of informal settlements, spatial policy and spatial analysis of need, greater
beneficiary’s participation inclusive of their capacity-building in order to house themselves
amongst others were stated. The stated pivotal contexts are typical needs of any developing
states, as the problem plaguing their human settlement is the urgent need to respond to the
upgrading of informal settlement which has become a global phenomenon, much more of a
nightmare to the developing states. Also, the call for spatial policy and spatial analysis of needs
is in response to the peculiar situation of the South African housing space, being a space that
was subjected to racial discrimination, thus inhabiting the growth of housing for a greater
percentage of the population. Therefore, the IRDP will be a programme to respond to this call
to facilitate the development of integrated human settlements in well-located areas that provide
convenient access to urban amenities, including places of employment, creation of social
cohesion and integration of the poor and low-income into the wide society, thus creating
opportunity for development and participation.
Likewise, the inclusion of housing beneficiaries participation as a pivotal context in the next
10 years and above, is a reflection of the fact that development will be too remote to be truly
‘of, by and for the people’ without their involvement in the development that affects them.
Beneficiary’s participation in the housing process will ensure appropriate housing strategies
and policies are more efficiently evaluated, developed and implemented to guarantee the
satisfaction of the beneficiaries’, thus supporting the spatial analysis of needs. While
inadequate beneficiaries’ participation in the process can lead to community conflict or in worst
case scenario, anti-development initiatives and ultimately housing dissatisfaction, which
impact on the quality of life of the beneficiaries. A successful beneficiaries’ participation will
allow the poor and low-income population to be involved in defining their housing problem
and crafting practical solutions to house themselves which will in turn lead to greater capacity
building for them to meaningfully contribute to the economic growth. Which will in-turn move
the current South African housing delivery system from welfare to a developmental state,
where the poor and low-income groups are empowered to progressively meet their own needs
and contribute to the society.
367
In conclusion therefore, the results seem to suggest that a single policy instrument is not
sufficient to respond to the need of the low-income groups, and that a combination of different
policy instruments will be better adapted to suit different low-income groups across the
country. This approach will be particularly useful in the eradication of the current housing
backlog. To this end, IRDP, a programme that facilitates the development of integrated human
settlements in well-located areas and SHP, which assists in the provision of secure, stable rental
tenure for the lowest income persons and ePHP, which is aimed at delivering better human
settlement outcomes based on community contribution, partnerships and the leveraging of
additional resources through partnerships, are recommend as the ideal combination of national
housing delivery programmes that will better serve the need of the poor. Furthermore, the low-
income groups will be adequately housed when the pivotal context of the South African
housing policy responds to the needs of the informal settlement dwellers, spatial policy and
spatial analysis of needs and assurances of beneficiaries’ participation a security for sustainable
human development, which will ultimately bring about housing satisfaction.
Objective DSO4
The forth objective of the Delphi Study was to identify the critical factors affecting the delivery
of low-income housing and their effects on beneficiaries’ residential satisfaction.
The logic behind this question was to assess the hindrances that the DHS is facing in the
delivery of low-income housing, when fulfilling the constitutional mandate of adequate
housing provision. This is because access to adequate shelter is defined as a basic right for all
citizens in South Africa, as stipulated in the country’s constitution. For instance, Section 26 (1)
of the constitution states that everyone has the right to access adequate housing; Section 26(2)
also states that the state must make reasonable legislative and other measures, within its
available resources, to achieve the progressive realisation of this right. Hence, the provision of
adequate housing is a qualified right, but is subject to available public resources of which the
extent to which the state supports the provision of good quality low cost housing is
consequently a matter of on-going debate in South Africa.
The Delphi Survey findings for this objective revealed that limited budget allocation,
appropriate policy to handle informal settlement and poor planning and coordination from
national to local government levels were found to be the most critical factors affecting the
368
delivery of low-income housing in the country with an impact factor of 9.00. With particular
emphasis on the limited budget allocation, the government initially suggested in 1994 that 5%
of the budget would go to housing, but the actual budget allocation has never achieved this
figure. Hence, the housing budget has drifted between R3-5 billion when it could have been
within R10-11 billion in terms of the budget allocation. This thus suggests that the declining
budget has contributed to the limited provision of low-income housing for the poor, which has
caused the housing backlog to increase with a compromise on the quality of the produced
houses. Also, the dwindling budget can be recognized as the cause of the other critical factors
of poor planning and coordination from the national to local government and the
implementation of appropriate policy to handle informal settlements. This is because of the
limited budget of the DHS to satisfactorily train its workers and to engage in housing
development consultants in work that the DHS does not have capacity for. However, this can
be argued because when the appropriate policy to handle the development of settlements is in
place, regardless of the limitation from budget constraints, adequate housing can still be
provided, but will only be lacking in quantity and not in quality. Other factors deemed to be
critical in affecting housing delivery, which are also related to the problem of a dwindling tax
base, include beneficiary’s active participation and lack of capacity to enable the poor to build
their own houses.
Also, the experts’ consensus that the waiting time on the housing database has an impact on
the housing delivery and thus, affects beneficiaries’ residential satisfaction further suggests that
the listed critical factors should be considered and a remedy of sorts for improvement by the
DHS be included. The DHS created the housing database because of the high demand for
housing in South Africa. The housing waiting list was created for the poor and low-income
groups with an income as specified on the qualifying criteria’s (refer to Section 6.2.2.4.1) for
people applying for housing. This list is used to allocate public housing to beneficiaries’
according to their areas of registration when the development is completed. The waiting list is
usually divided into categories, with applications placed into the category, which best reflect
the urgency of the housing need, such as:
Category 1: applicants in urgent need of housing (such as the homeless) and are unable
to access private rental housing options;
Category 2: applicants who have high housing needs and who face long term barriers
to accessing other housing options; and
369
Category 3: applicants who pass the income and assets test but do not have a high
housing need (as per Categories 1 or 2).
Despite the categorization, there has been complaint from the housing beneficiaries’ and
NGO’s alike that the housing waiting lists are not followed as a result of inadequacies from the
local authority personnel’s. This has resulted in beneficiaries’ having to wait for housing for
longer periods than necessary and some not receiving any housing until they pass away. Hence
the experts’ consensus revealed that the waiting period on the housing database has an impact
on the housing delivery and satisfaction of housing to the low-income group.
In conclusion, the findings revealed that the daunting critical factor, which affects low-income
housing provision in South Africa, is financial limitations caused by the dwindling tax base.
This suggests that when this barrier is overcome other factors listed as critical would also be
overcome. It was further revealed that the waiting period on the housing database impacts on
the delivery of housing and in turn affects the housing satisfaction of the low-income groups.
Objective DSO5
The fifth objective of the Delphi Study was to predict the life span of the present South Africa
public housing subsidy delivery model.
The South African public housing subsidy scheme is a process by which grants are provided
by the government to qualifying beneficiaries for housing purposes. Government does not pay
the grants as cash to the beneficiaries. Rather, the grant is either paid to a seller of a house, or
in new developments or used to construct a house that complies with the minimum technical
and environmental norms and standards, which is then transferred to the qualifying beneficiary.
One of the Department of Human Settlements’ areas of responsibility in the delivery of human
settlements relates to the bottommost end of the market (poor and low-income), where the
department provides housing subsidies to the poorest of the poor. This is a critical area, as the
bulk of the housing backlog exists here, and affects mainly those who earn between R0 and R3
500 a month.
The total amount allocated to grants for the poor in 20010/11 reached R15 billion, which is
envisaged to rise to R17.9 billion in 2013/14, representing an average annual increase of 13%.
A large percentage of this amount is allocated to provinces in the form of housing development
370
grants. Hence, as a result of the dwindling tax base, this objective solicits and predicts from the
experts the expected time they envisage the present housing delivery model will be stopped by
the government. Findings indicated that a majority of expert’s foresee the programme ending
in 10 years and above time, representing 67.0%. This finding indicates that although there is
the urgent call for the incorporation of other models to support the delivery of low-income
housing; the current public housing subsidy delivery model is perceived as the dominant
housing delivery model for some time to come. Likewise, other responses indicated that the
model of housing delivery will never be stopped. This view is based on the prevalence of
economic inequality, unemployment and poverty, which will necessitate the government to
continue with the programme in order to fulfill its constitutional mandate to the citizens.
However, the experts disagreed that this model of housing delivery will not always be the
model used for housing provision asserting that government alone cannot simply provide
housing for the poor without some other intervention.
Therefore, when the experts were asked to predict the envisaged programme that will replace
the subsidised housing delivery scheme when it is eventually replaced; it was found that the
market and subsidy mix (social housing type models), assisted self-help housing, upgrading of
informal settlement amongst others were stated as the envisaged programmes. This is a further
indication that the current housing delivery process in South Africa will better respond to the
use of a mix of housing policy instrument.
In conclusion, the findings have shown that though subsidised low-income housing will not be
the major delivery model for low-income housing provision to the poor in the near future, the
end of the programme is not likely to come anytime soon either. This suggests that other models
should be given the same attention by the DHS in order to eradicate the daunting housing
backlog and to respond to the constitutional call of the citizens.
Objective DSO6
The sixth objective of the Delphi Study was to investigate the management issues affecting the
national, provincial and local government housing agencies in the development of housing in
South Africa.
In response to the open-ended question, it was found that 29 issues were raised as the issues
currently faced by the national, provincial and local government housing departments in the
371
delivery of low-income housing. From the listed issues, government capacities to facilitate
development and corruption in provincial and local government were ranked as the highest
constraints hindering the development of low-income housing in South Africa. Similarly,
budget constraints, poor political will amongst others were also listed as the current problems
hindering housing delivery. The findings suggest that despite the quantity of low-income
housing delivered in the country, there are still numerous problems, which hinder effective
housing supply. As previously discussed in Section 8.4.6, budget constraints is a major
hindrance to effective housing delivery in South Africa and in the countries of the global south
and in other developing nations like Brazil, and Jamaica, amongst others. However, the
inclusion of corruption as a challenge will further delay low-income housing provision as the
limited budget that will be unaccounted for increases the housing backlog and other associated
housing issues, like poor quality.
In contrast, an issue not raised by the exert panelist, which can be debated as a major hindrance
regardless of the limited budget is the issue of provincial and local government under spending.
The problem of under-spending can be directly linked to lack of capacity and poor management
skills to administer allocated grants. This view has been stressed by the National Department
of Human Settlement, prompting that provinces whose local government under-spend, will
forfeit such grants through redirection to Provinces that are performing in their mandate of
housing delivery. It can also be argued that the issues, as stated, are largely due to a lack of
project management experience and bulk infrastructure delivery, such as large-scale electricity
and water supply projects. These constraints currently faced by the DHS and the stance for
under-spending provinces to forfeit their budget will further widen the housing backlog gap.
Subsequently, when a prediction of the issues that will hinder housing delivery in the next 10
years or above were considered, it was found that a majority of the listed current constraints
were again considered as the future issues. However, urbanization and the cost of basic social
services like water and electricity, which were considered in the previous question as having
the least impact, were now considered as major future problems. According to the proceedings
from the first African Ministerial Conference on Housing and Urban Development
(AMCHUD) held in Durban, South Africa in 2005, it informs that in the next 30 years Africa’s
population will double from 888 million in 2005 to 1.77 billion. During the same period the
urban population will increase from 353 million, which is 39.7 percent, to 748 million
inhabitants at the rate of 4 to 5 percent per annum. Also, in the next 30 years, roughly 400
372
million people are projected to be added to the Africa urban population. According to the UN-
Habitat (2010), 60 per cent of all Africans will be living in cities (urban areas) by 2050.
As also highlighted in the 1994 Cairo International Conference on Population and
Development (ICPD), the process of urbanization is fundamental to economic and social
development when properly handled with the foresight of the advantages that go with it. While
it may be true that rapid urbanization in Africa has not been associated with corresponding
economic growth, in South African, urbanization has gone hand-in-hand with growth in real
gross domestic product over the past decade although both urbanization and economic growth
have been highly uneven and unequal over time (UN-Habitat, 2010). For instance, the UN-
Habitat (2010) report further informs that during the 2000-2010 decade, the Southern African
sub-region retained its position as the most urbanized on the continent, with the rate increasing
from 53.8 to 58.7 percent. The Southern African sub-region is projected to reach a two-thirds
urban majority sometime around 2025. The region has now entered a period where decade-
interval urbanization growth rates are expected to slow down, to 4.9 per cent for 2000-2010
and a continuing steady slowdown to 2.1 per cent in the 2040-2050 decade (UN-Habitat, 2010).
However, the Republic of South Africa is 62 per cent urbanized, with an annual rate of change
of 1.2 per cent, which is way ahead of the least-urbanized nations of the sub-regions. The
projected increase in urban populations will lead to an exponential increase in the demand for
shelter and services. Already, South African urban areas are inundated with slums and an
exponential urban growth of 1.2 per cent of the urban populations could spell disaster, unless
urgent and progressive action is initiated today. Against this backdrop, the experts projected
that urbanization and the demand for services such as water and electricity will be the major
management issues that South Africa provincial and local government will be faced with in the
next 10 years or in the future.
Also, the experts agree (agree: 40.0% and strongly agree: 27.0%) that the country’s current
unemployment condition has a huge impact on housing delivery and the eventual housing
satisfaction of the low-income groups. Since 1994, South Africa has made undeniable progress
across a number of critical areas. On the economic front, the government has pursued policies
that have restored and maintained macroeconomic stability in the context of a difficult global
environment. But despite these areas of success, there exists a widespread perception that South
Africa’s economic performance since 1994 has been disappointing. Real GDP growth has been
373
erratic, formal sector job losses have continued unabated, and the key objectives of poverty
reduction and improved service delivery remain largely unmet. According to Statistics South
Africa (2011), between the 2nd and 3rd quarters of 2011, the unemployment level declined by
96 000 resulting in a decline in the unemployment rate by 0,7 percentage pointing to 25,0%.
However, levels of unemployment still remain high as it is currently standing at 25.2% for the
first quarter of 2012, compared to the same quarter last year. Despite the marginal increment,
4,4 million people remain unemployed and just over 3,0 million (68,2%) have been
unemployed for a period of 1 year or more. Historically, from 2000 until 2012, the South Africa
Unemployment Rate averaged 25.5 Percent reaching an all-time high of 31.2 Percent in March
of 2003 and a record low of 21.9 Percent in December of 2008. Just like any other developing
country, the causes are almost in general terms, such as: lack of education, poverty, diseases
like HIV/AIDs, squalid living conditions and corruption from the top and lastly, influx from
foreigners who are either highly skilled and take all the top high paying jobs and cheap labour.
Although there would be disagreement that unemployment, is the most significant factor
affecting low-income housing delivery and it satisfaction by the beneficiaries in South Africa,
as already pointed out. Against this setting and the current unemployment rate, it therefore
suggests that the number of citizens who are not able to meet their own housing needs and who
are in the poor and low-income groups will keep increasing, thus expanding the housing
backlog. When houses are provided for the poor and low-income groups which do not meet
their housing needs and they are not able to expand the development to meet their need, there
will be dissatisfaction with the housing stock provided.
In conclusion, the findings have shown that there are dire issues facing provincial and local
government in their capacity to deliver low-income housing in order to fulfill the government
mandate of adequate housing for the poor. The investigated problems currently faced by the
department were also mirrored as the problem they will face in the future with a further problem
of urbanization and limited supply of basic social services for the growing population. These
findings thus suggest that the South African government via the National Department of
Human Settlement cannot afford to ignore the on-going urban transition taking place. To this
end, South African cities must become prioritized areas for public policies, with hugely
increased investments to build adequate governance capacities, provide equitable service
delivery, affordable housing provision and better wealth distribution. Therefore, the researcher
is of the opinion that plans should be put in place to help underperforming provinces and their
respective local governments to meet housing delivery targets.
374
Objective DSO7
The seventh objective of the Delphi Study was to investigate the effect of beneficiaries’
participation on their overall housing satisfaction.
The findings for this objective revealed that participation should not only be an essential
component for attaining sustainable development but also as a necessary precondition for
attaining sustainable development. The experts’ assessment of this objective suggests that
when beneficiaries participate in the housing process, residential satisfaction will be
guaranteed. Not only will the process guarantee their housing satisfaction, but it develops their
capacity for participation in subsequent projects. For instance, Gran (1983) argues that real
development, by definition, must involve beneficiaries in their own improvement and without
participation the people may benefit but not develop from a project, thus hindering their
capacity to participate in future development that will concern them. Therefore, the real
emphasis of beneficiaries’ participation should be on the satisfaction of basic human needs and
the meaningful participation of the masses in the shaping of their economic and social changes.
Also, the policy of self-reliance should be encouraged, with the emphasis on a self-confident
and creative use of local resources, manpower, technology, and knowledge (Finsterbusch &
Warren, 1987). One of the prime benefits of participation is better project design, because
participation ensures that felt needs are served through consultation with the end-users.
Presumably beneficiaries will shape the project to their specific needs in ways that outside
planners cannot. This is because the sense of immediate responsibility and ownership by the
beneficiaries puts pressure on a project to be truly worthwhile. However, a major obstacle to
participation is the difficulty of implementing it in practice. It takes additional time and
resources to mobilize less developed communities. Despite this constraint, there has to be
continuous consultation with the people and developmental projects should not be executed
without their involvement.
In conclusion, it can be argued that developmental efforts and initiatives should be directed not
just to achieve an improvement in the well-being of the low-income groups or communities
but also that the gains thereby made be retained and nurtured to greater levels. The finding also
suggest that the people who are seen as the beneficiaries of development must be capacitated
or helped to take charge of the processes and results of developmental interventions. This is
what is being referred to as participation in this study. For the beneficiaries of housing
375
development to continue realizing proceeds of developmental processes, they must be guided
to determine both processes, as well as end products of development, which will in turn bring
about their satisfaction with the housing development projects.
Objective DSO8
The last objective of the Delphi Study was to investigate the effect of meeting beneficiaries’
housing needs and expectations on their overall housing satisfaction.
Previous research has proven that expectations have a significant effect on overall satisfaction
of occupants; as satisfaction normally occurs based on a comparison of that which is expected
with that which is received (Caughey et al., 1998). Prior exposure to what is to be received has
the tendency to influence occupant’s satisfaction towards a housing product. While a negative
prior experience can generate a lower expectation, which will result in lower satisfaction.
Therefore, results for this objective revealed that prior exposure of beneficiaries to what they
will receive has a tendency to influence their satisfaction towards the housing product, as
revealed from 80.0% agreement to the statement. This finding suggests that when beneficiaries
are aware of what they will receive, they will either be satisfied based on the expected outcome
or dissatisfied. Research has shown that satisfaction with what is expected suggests that
satisfaction is the result of a comparison of that which was expected and that which was
received (Woodruff et al., 1983). A fundamental premise of dissatisfaction with prior exposure
(expectation) is that expectation is related to satisfaction. The result also suggests that in
addition to the influences from expected performance and subjective dissatisfaction, perceived
performance exerts direct influence on satisfaction (Tse & Wilston, 1998). Therefore it can be
asserted that beneficiary dissatisfaction with what has been received is a response to the
congruency between an individual’s expectations and the actual performance of the housing
product that was received. Hence, public housing beneficiaries’ satisfaction may be viewed as
a function of the interrelationship between what beneficiaries expect from the government and
their perceptions of the houses they have received; that is, the quality of the houses received
and the satisfaction derived from the housing meeting their needs. Although, dissatisfaction is
assumed to have a major effect on the users’ satisfaction, as research has shown dissatisfaction
is not the only direct outcome, but prior exposure to what is to be received (expectations) have
also been found to directly affect satisfaction (Reisig & Chandek, 2001). For instance,
individuals with lower expectations often report higher levels of satisfaction. Additionally,
376
Oliver (1981) maintains that as performance increases, so too do the levels of user satisfaction.
Expectations (prior exposure) and performance (quality), therefore, are believed to have both
direct and indirect effects on user’s satisfaction.
In response to the needs of the beneficiaries as stated in the research objective, Abraham
Maslow’s (1970) Needs Hierarchy Structure was used to evaluate the housing needs that should
be met for the low-income groups to be satisfied with their housing situation. This hierarchy
provides an interesting framework to categorize common needs into particular classifications.
According to this theory, human needs are unlimited and when one of them is met, another
follows it. In this process, complete satisfaction is not possible unless a need classified to be
important is first met. Individuals want what they do not have and the need once satisfied loses
its motivating power. Therefore, the expert panellists revealed that the prepotent needs of the
low-income groups in South Africa are esteem and physiological needs. This finding shows
that Maslow’s (1970) Hierarchical Needs order cannot be applied directly to the context of
South African low-income housing beneficiaries from the informed experts’ point of view.
Likewise, the classification of esteem and physiological needs as the greatest need is a
reflection of the inequality and past history of the South African state. This is because the poor
and low-income groups want to be recognised and respected because of their previously
disadvantaged backgrounds. The experts’ finding suggests that this should be in perfect
alignment with their physiological needs before consideration of social needs (love and
belongingness), self-esteem and ultimately self-actualization, to make them feel safe, secure,
and part of the society. This order will result in their satisfaction with the housing product that
will be allocated to them. Therefore, the new needs order from the experts’ responses
compressed Maslow’s Needs Hierarchical Theory into four classes in the context of South
African low-income housing.
Lastly, further findings revealed that housing was perceived by a majority of the expert
panellists as the paramount need of the low-income groups, which concurred with the previous
findings as both housing (physiological) and esteem needs were rated as the prepotent needs.
This finding suggests that shelter is still considered a most basic need for the low-income
groups. This emphasises the reason why every successive government in South Africa, since
the end of the Apartheid Government, has prioritized the provision of affordable housing for
the low-income group. This is because of the pivotal role played by housing in national
development and growth on one hand and it being a necessity in the life of the people, on the
377
other. This makes the right to housing one of the most important basic human rights recognised
in many international Human Rights Treaties, a necessity. Housing addresses basic human
needs at a primary level, Van Vliet (1998) says that housing acts as shelter, and Novick (1990)
further emphasized that housing is the environment, which exerts the greatest and most
immediate influence on the lives of the people, their health and well-being.
In addition, when the 13.0% expert panellists that rated housing as not a paramount need of the
poor and low-income groups were asked to state what their utmost need was; findings revealed
that employment, education and job security were ranked as the highest priorities with a very
urgent rating on the adopted scale of urgency, while, food security was ranked second, followed
by clothing, transportation and health security. Amongst the listed needs, clothing was
considered by the experts as not urgent on the scale of urgency in delivery.
Although, research has shown that housing is not the paramount need of the poor and low-
income groups. The most immediate needs from such research findings are usually, food,
employment with better pay (if they have jobs or employment if they do not have jobs) and
education for children. Also, in some surveys, housing does not feature in the top five. Other
surveys and findings have revealed that the poor come in different categories and the poor and
low-income groups with different characteristics have a different set of needs (International
Fund for Agricultural Development, 2009). The results therefore show that housing cannot be
regarded as the prime need of the poor and the low-income groups as the findings and literature
have shown.
In conclusion, the needs order theory postulates that the appearance of one need usually rests
on the prior satisfaction of another more prepotent need. Also, no need or drive can be treated
as if it were isolated or discrete; thus every drive is related to the state of satisfaction or
dissatisfaction of another drive. It is believed that once a need is satisfied, it ceases to motivate
behaviour. Also, an understanding of how resident’s expectations are formed is significant in
ascertaining how beneficiaries’ satisfaction is ultimately formed.
From all of the above, a number of factors that were considered to be important in determining
residential satisfaction have been identified and amplified by the Delphi Study. The factors
considered to be paramount determinants of residential satisfaction include: dwelling unit
features, neighbourhood features, building quality, services provided by government,
378
beneficiary participation, needs and expectations. These factors have been collectively
considered for the development of a holistic residential satisfaction model in this study. Four
of the factors have been previously considered in the development of residential satisfaction
model in other cultural contexts, but none of the existing models to date have included both
beneficiary participation, needs and expectations as inclusive factors to develop a model to
guide housing authorities in the construction of houses that will be satisfactory to the poor and
low-income groups.
8.5 CONCLUSION
This chapter presented a summary of results and discussions of the results from all the Delphi
rounds, first to the third round. Computation for each and every question element was made
for the influence and impact of the attributes in predicting residential satisfaction and
improvement of the low-income housing context in South Africa. Also, the influence or impact
of the absence or presence of a particular residential satisfaction element on the overall
residential satisfaction of the other elements was presented; likewise on the issues hindering
the effective delivery of low-income housing and the current problems faced by the DHS to
adequately house the poor and the low-income groups in South Africa. The chapter concluded
with a summative discussion of the findings based on the objectives of the Delphi study. The
findings from the expert participants revealed a coherent dialogue on low-income housing in
South Africa, with consensus being reached in most cases and in others with a discrete
conclusion. The result of the Delphi Study assisted in the determination of key factors and
constructs that are of critical significance (influence) to determine residential satisfaction in
subsidised low-income housing, which led to the development of the holistically integrated
conceptual model for residential satisfaction in South African low-income housing. The
evaluation of these factors and their interrelationships is presented in the next section.
379
CHAPTER NINE
THE CONCEPTUAL INTEGRATED RESIDENTIAL SATISFACTION
MODEL
9.1 INTRODUCTION
This chapter presents the discussion of findings from the review of literature and the Delphi
Study. This discussion forms the basis of the conceptual model’s theory. The hypothesised
integrated holistic residential satisfaction model is also presented in this chapter based on an
in-depth review of the previous models as presented in Chapter Two of this thesis. This chapter
also describes the integrated holistic model and the variables of the model in detail, except
beneficiaries’ participation, needs and expectations, which has already been discussed in
Chapter Three of this thesis, as variable constructs identified as gaps in residential satisfaction
research. Also presented in this chapter is the model identification and justification for the
selected variables, thereafter a conclusion is drawn for the chapter.
9.2 SELECTION OF VARIABLES FOR RESIDENTIAL
SATISFACTION
Most residential satisfaction study models have combined both objective and subjective
attributes for the assessment of residential satisfaction. For instance, Francescato, Weidemann,
and Anderson (1987) suggest that residence satisfaction with any residential dwelling depends
on three elements, which are: the design of the house, (i.e. the dwelling space organisation,
layout and facilities provided); the management practices; and the surrounding social aspects.
Varady and Carrozza (2000), Salleh et al. (2011), acknowledge that residential satisfaction
encompasses four distinct types of satisfaction, which include: satisfaction with the dwelling
unit; satisfaction with the services provided, including repair services; satisfaction with the
whole package received, as in the case of public housing, where no rent is paid (dwelling and
service); and satisfaction with the neighbourhood or area. These four constructs as proposed
by Varady and Carrozza (2000) is also supported and adopted for the current study.
Furthermore, Nurizan and Hashim (2001) maintain that outside the facilities in the house, other
basic facilities, such as shops, markets, schools, clinics, mailing system, community hall,
playgrounds, and others are important to support the daily life of the occupants, and enhance
380
their quality of life. Likewise, Oh (2000) states that there are three main qualities, which bring
about residential satisfaction, which are: the quality of the dwelling; the quality of the close
environment; and the quality of the urban site, which impacts on the quality of housing.
Therefore, based on the review of literature on variables that are likely to affect residential
satisfaction, the present study considers the residential satisfaction bundle in a typical low-
income housing development to contain the dwelling unit features with 17 variables;
neighbourhood and environmental features with 22 variables; services provided by the
government with 13 variables; building quality features with 16 variables. All these are the
constructs that have been frequently conceptualized in most residential satisfaction studies.
However, the present thesis brings into focus the impact of needs and expectations features
with 4 variables and beneficiaries’ participation features with 5 variables. These two additions
are the gaps identified from the review of literature, which were found peculiar to the
developing nation’s situation. The next section of this chapter will present a detailed
explanation of the four different constructs influencing the level of satisfaction towards
housing; the explanation for the two new added constructs have been discussed
comprehensively under the identified gaps and their treatment sections in Chapter 3.
9.2.1 Dwelling Unit Features (DUF)
In some situations, pleasant dwelling features are the main parameters used in describing
quality of housing. In order to distinguish if a house is of good quality or not, the dwelling
units features from the internal and external aspects and also the nearby area are used to
qualify the house (Salleh et al., 2011). Generally, dwelling unit features refer to the floor plan
of internal spaces within the dwelling unit and it includes the living room, dining room,
bedroom, kitchen, bathroom, toilet and drying areas, including ventilation of the house. These
can also be classified as internal dwelling unit features. Dwelling unit features also include:
dwelling size (Lu, 1999); size of the living room, bedroom, kitchen, toilets and washing area
(Opoku & Abdul-Muhmin, 2010; Salleh et al., 2011), number of bedrooms (Ibem & Amole;
2011; Ilesanmi, 2010), location of bedroom and size of dining room (Ukoha & Beamish, 1997)
laundry and washing area (Mohit, et al., 2010). Over time, researchers have combined some of
the above dwelling features to estimate the level of residents’ satisfaction with their dwelling
units, thus informing housing authorities of the necessary changes to be made in order for the
occupants to be satisfied with their houses. For instance, Husna and Nurizan’s (1987) study on
satisfaction on public housing was used by the Malaysian Government to redesign its public
381
housing dwelling unit feature to a new standard. Also, the number and location of the
bedrooms, availability of laundry facilities and the provision of a washing line are also dwelling
unit factors that are considered in order to bring about a quality housing units. Therefore, in
this present study, the dwelling unit features that have been hypothesised for the development
of a holistic residential satisfaction model are summarized in Table 9.1.
Table 9.1: Conceptual Model Latent Constructs
Latent Variable Constructs Measurement variables
Dwelling Unit Features (DUF) Location of bedroom
Number of bedrooms
Size of the bedroom
Location of living room
Location of dining room
Location of kitchen
Size of the kitchen
Size of bathroom(s)
Size of wardrobe/closet
Size of children’s play space
Size of children’s study space
Amount of privacy within the house
Amount of brightness / sunshine in the house
Quality of ventilation in the house
Quality of floor level in the house
Overall appearance of the house
Overall size of the house
Neighbourhood Features (NDF) Location of the dwelling unit in the
neighbourhood
Quality of relationship with neighbours
Quality of landscape in the neighbourhood
Quality of walkways
Ease of access to main roads
Amount of privacy from other neighbours
Quality of street lighting at night
Amount of security in the neighbourhood
Physical condition and appearance of the
neighbourhood
Cleanliness of the neighbourhood
Proximity of house to workplace
Proximity of house to shopping areas
Proximity of house to the nursery school
Proximity of house to the high school
Proximity of house to hospitals/clinics
proximity of house to place of worship
Proximity of house to police services
Proximity of house to parking facilities
Proximity of house to disabled facilities
382
Proximity of house to the community hall
Proximity of house to playground / recreational
facility
Proximity of house to public transportation and
services
Building Quality Features (BQF) External construction quality
Internal construction quality
Water pressure
Wall quality
Floor quality
Window quality
Door quality
Internal painting quality
External painting quality
Plumbing quality
The finished quality of sanitary system
Electrical wiring quality
Electrical fittings quality
Numbers of electrical sockets
Level of socket
Overall unit quality
Services Provided by Government The drainage system
(SPG) Garbage and waste collection
Fire protection services
Electricity supply
Water supply
Telephone service
Safety
How well resident complaints are handled
Housing department rules and regulations
Enforcement of rules by the Department of
Human Settlement (Housing)
Overall services provided by the government
Beneficiary Participation (BNP) Owners should be consulted about the housing
location
Owners should be consulted about the house
design
Owners should be consulted about the house
construction
Owners should be consulted about the internal
finishes of the house
Owners should be consulted about the external
finishes of the house
Needs and Expectation (NAE) Owners should be told beforehand the type of
house they will receive
Owners should be asked the type of house they
need
Owners expect good quality houses
Our houses should meet our family need
383
Residential Satisfaction (RS) I am satisfied living here
I am taking proper care of my house
I am taking proper care of my neighbourhood
I am constantly maintaining my house
I am not intending to move to another place in the
future
I like to live in another place like this
I will recommend to my friend to obtain a house
in the same way that I did
9.2.2 Neighbourhood Features (NDF)
The literature on neighbourhoods defines this concept in many ways. Brower (1996) informs
that its form is derived from a particular pattern of activities, the existence of a common visual
motif, an area with continuous boundaries or a network of often-travelled streets. Diverse
definitions serve different interests, so that the neighbourhood may be seen as a source of place-
identity, an element of urban form, or a unit of decision making. It is presumed that research
uses multiple definitions of a neighbourhood simultaneously to reflect the fact that
neighbourhood is not a static concept but rather a dynamic one (Talen & Shah, 2007). Likewise,
planners and designers have also thought of the neighbourhood setting as a fixed, controllable,
and imaginable physical area.
Researchers agree that a neighbourhood should comprise a walkable distance (the distance that
a person could pleasantly walk, a 3MPH pace in 5 minutes). However, the actual walkable
distance considered has varied from a quarter-mile to one mile from centre to edge
(Colabianchi et al., 2007; Talen & Shah, 2007).
Previous studies on housing satisfaction revealed that several features are required to determine
the housing satisfaction of a given household or individual. For instance, the availability of
desired features and structure types are related; accordingly, as different services are provided
by different structure types which also affect satisfaction with housing units (Johnson &
Abernathy, 1983). Also, the availability of space depends on the structure type, and the amount
of space in a dwelling unit correlates with the housing satisfaction level (Aigbavboa & Thwala,
2010; Kinsey & Lane, 1983). Satisfaction with neighbourhood features have been observed as
a vital determinant of residential satisfaction (Vrbka & Combs, 1991) to the extent that
residents are willing to compromise the inefficiencies within the dwelling unit because of the
satisfaction that is provided by the neighbourhood facilities and features (Ukoha & Beamish,
1997). Neighbourhood features refer to the location of the dwelling unit, neighbourhood
384
relations, distance to the shopping areas, distance to the workplace or school, distance to the
police services, distance to recreational facilities secure and clean environment, the building
image and parking facilities amongst others (Aigbavboa & Thwala, 2011; Awotona, 1991).
Hence residents of a given housing scheme are most likely to be dissatisfied with housing
facilities that require residents to travel or walk long distances to school; to workplaces,
shopping areas, medical centres and the geographical areas around their dwelling units. Easy
access to good public transportation, community and shopping facilities and physical
environment variables will provide residents’ satisfaction with their housing units.
For instance, research conducted by Bjorklund and Klingborg (2005) in eight Swedish
municipalities found the following top neighbourhood factors amongst others to be related to
residential satisfaction, these include proximity to commercial areas, building exteriors with
high aesthetic values, proximity to open spaces, less noisy environments with no traffic
congestion, good reputation, good quality along the housing surroundings, proximity to town
centres and a conducive environment. On the other hand, findings of a study conducted by
Abdul and Yusof (2008) on residential satisfaction shows that neighbourhood facility factors
are the most dominant factors in determining the level of satisfaction towards housing. The
study further revealed that factors of neighbourhood facilities that caused a low level of
satisfaction were poor public transport, lack of sport fields, lack of multi-purpose halls, lack of
parking areas and lack of safety facilities for the disabled. Also, Ramdane and Abdul’s (2000)
study on the factors of neighbourhood facilities to evaluate the level of residential satisfaction,
found that neighbourhood factors have a huge impact on the overall satisfaction with the
housing facilities. Research has pointed out the complex characteristics of neighbourhood
satisfaction (Amerigo & Aragones, 1997; Francescato, 2002; Marans & Rodgers, 1975; Marans
& Spreckelmeyer, 1981). It has also been identified that aesthetics, or pleasantness to the eye,
is one of the most important factors in neighbourhood satisfaction (Kearney, 2006; Sirgy &
Cornwell, 2002). Whilst, social and personal characteristics, such as neighbourhood cohesion,
or networks, were other factors associated with neighbourhood satisfaction (Chapman &
Lombard, 2006; Morrow-Jones, et al., 2005; Okun, 1993; Westaway, 2006). The
neighbourhood and environmental features, which are considered for the present study, are
summarized in Table 9.1.
385
9.2.3 Building Quality Features (BQF)
Residents’ satisfaction towards a given housing unit is also derived from the satisfaction with
the building quality and the housing condition features in the buildings (Ukoha & Beamish,
1997). Onibokun (1974) classified building condition features as dwelling subsystems to the
human habitat that influence the level of housing satisfaction. This position was further
supported by McCray and Day (1977) who states that low-income housing construction is
rarely developed to reflect the needs and types of families who are going to inhabit the houses,
as the building condition / quality elements are seldom considered in the establishment of
human habitats. Whereas, the quality of low-income housing should be a combination of both
the user’s requirements and the principles that define adequate housing. But because public
low-income housing is built for the poor and disadvantaged, with the cost being covered by the
government; the choices of design and materials used during construction are only based on an
affordable budget, which compromise best practices with regards to adequate housing for the
low-income groups. Hence, Kutty (1999) claims that a good building structure with good
quality is an important indicator that determines the residents’ satisfaction with the building
and the value they place on the dwelling.
According to Duncan (1971) and Ramdane and Abdullah (2000), three dimensions of housing
quality are usually considered with regards to dwellings, which include: internal aspects of a
dwelling unit, its external aspects, as well the surrounding area. Furthermore, Elsinga and
Hoekstra (2005) inform that the higher quality a dwelling is, the higher the resident’s
satisfaction with it. They state that housing quality and condition should not be assessed based
on one variable only, but from objective and subjective dimensions. Also, Kain & Quigley
(1970) divided housing quality into five critical factors namely: basic housing quality factor;
dwelling unit quality factor; surrounding property quality factor; non-residential land use
quality factor; and structural average quality factor. According to Kain & Quigley (1970), basic
housing quality factors refer to the index used to measure the surrounding areas and the external
physical quality of the unit. While the dwelling unit quality factor is assessed from the
structural aspects and internal hygiene of the dwelling unit; surrounding property quality factor
is assessed from the general cleanliness of the surrounding area, its ambience and landscaping.
The factor of quality for non-residential uses is measured from the effects of industrial and
commercial uses in residential areas. These effects are assessed based on the level of
discernible noise, air quality and traffic flow in the area. The structural average quality factor
is assessed based on the structural quality on the building facade.
386
In addition, the main source of occupants’ problems with their health, safety and physical issues
is caused most times by building quality (Salleh et al., 2011). Building quality factors that
contribute to residents’ satisfaction with their dwelling units include: wall, floor, window, roof,
door and painting quality, amongst others (Salleh et al., 2011; Ukoha & Beamish, 1997). For
instance, the materials used for wall construction give effect to the building temperature.
Besides, the quality of windows should provide good ventilation and air circulation in the units,
because if the windows are unable to open it will increase the heat in the unit, hence, residents
will be uncomfortable in their house. Therefore, the building quality features considered for
the present study are summarized in Table 9.1.
9.2.4 Services Provided by Government (SPG)
The relationship between the government departments responsible for public housing and the
occupants poses a large influence on their satisfaction with the housing units (James et al.,
2009). For instance, the response to occupants’ complaints; building defects and repair services
carried out by the government in a public housing environment is said to influence the level of
the occupants’ satisfaction towards their dwellings. Also, the time taken by the management in
handling the complaints is a major factor that influences occupant’s satisfaction with a given
public housing unit (Varady & Carrozza, 2000; Ukoha & Beamish, 1997). In addition, Husna
& Nurizan (1987) demarcate that prompt plumbing and building repairs, electrical wiring,
water supply, garbage disposal and security are vital services when provided, which influences
the level of satisfaction amongst residents of low-income housing. As already established,
services provided by the government in low-income housing plays an important role in
producing housing quality, which brings about satisfaction with the dwelling unit. The
indicator variables included in this component are summarized in Table 9.1.
9.3 MODEL SPECIFICATION AND JUSTIFICATION
This thesis aims to build a conceptual residential satisfaction model centered on the subsidised
low-income housing scheme. The theoretical conceptual framework for the current research
builds on the work of Marans and Rodger (1975) and Marans and Sprecklemeyer (1981)
models of satisfaction as discussed in Chapter Two (refer Section 2.3.6 - 2.3.7). Marans and
Rodger (1975) conceptualized that an individual’s overall satisfaction with housing depends
on their perception of the various neighbourhood characteristics and their assessment of them.
387
Marans and Rodgers’s (1975) model also conceptualized that both the perpetual evaluative
process and the overall satisfaction level are related to the residents’ own characteristics, such
as social class, housing status amongst others (Table 9.2). Similarly, Marans and
Sprecklemeyer (1981) further determined that residents’ satisfaction is a function of the
physical environment through one’s perception and beliefs of the physical environment. In this
particular model, housing satisfaction was derived as a result of an integrated relationship
between the environment and the human perception of beliefs. The three basic components of
the model were: the physical environment, the perception and attitude of residents toward their
housing environment and residents’ satisfaction. Based on the fundamental underpinning of
these two models, and the incorporated theoretical perspectives, which has been adopted in
other similar studies, they are therefore useful for conceptualizing the present study as a variety
of satisfaction studies with urban housing living being conceptualized within the broad
theoretical framework.
Therefore, the conceptual framework for this thesis is primarily based on the approach used by
Marans and Rodger (1975) when they view residential satisfaction as a criterion of evaluation
of residential quality and, at the same time, as a variable predicting certain behaviour. In this
regard, residential satisfaction was treated as a criterion variable and, therefore, as a dependent
variable. The approach was also used by Galster and Hesser (1981), Cutter (1982) and
Weidemann and Anderson (1985), which has also been adopted in the current study. Based on
the fundamental factors and constructs associated with all the previous models as revealed in
Table 10.1, the present model or conceptual framework model for the study looks at the
relationship of the dwelling unit, neighbourhood and environmental features, services provided
by government, building quality, which are the essential variables that have been measured in
a majority of the previous studies, with the inclusive consideration of the impact of needs and
expectations and beneficiaries participation; which have been classified as the exogenous
variables and their role in predicting overall beneficiary residential satisfaction, which is the
endogenous variable. These will in turn, predict the beneficiaries’ satisfaction towards the
housing stock, behaviour to maintain the housing stocks and their overall responsibility in the
low-income neighbourhood, or likelihood to move and eventually place attachment. The study
aims to forecast the relative predictive power of these different variables for beneficiaries
housing satisfaction in order to test/determine if residential satisfaction depends on the
supposed features of the variables, taking into account the effects of the beneficiaries needs,
388
expectations and meaningful participation prior to construction in alliance with the South
Africa Housing Policy and Codes, and as emphasized by other frameworks.
It is apparent that some of the variables discussed above should be measured by objective
means, some by subjective means and some will include both forms of measurement. The
reason for combining both objective and subjective indicators within the proposed model is
supported by Campbell et al. (1976) and Falah, Al-Abeda and Wilda (1995) who stated that:
by themselves, objective indicators are often misleading and will remain so until indicators that
human beings attached to them, are obtained. Likewise, by themselves, subjective indicators
are insufficient as guides to policy.
Table 9.2: Factors of Residential Satisfaction
R e s i d e n t i a l s a t i s f a c t i o n
e l e m e n t s
Asp
ira
tio
n
Ex
pec
tati
on
s
Ph
ysi
cal
sett
ing
s
Per
cep
tion
Cu
ltu
re
Res
iden
tia
l
beh
av
iou
r
Per
son
al
or
resi
den
ts’
featu
res
Nei
gh
bou
rho
od
fea
ture
s
Nei
gh
bou
rho
od
&
env
iron
men
tal
fea
ture
s
Co
mm
un
ity
serv
ices
Dw
elli
ng
un
it
fea
ture
s
So
cio
eco
no
mic
cha
ract
eris
tics
Ho
usi
ng
ty
pe
Use
rs n
eed
Michelson (1977) -
Michelson’s Integrated Model X X X X X X X
Onibokun (1974) - Habitability
Model
X X X X
Marans and Rodger (1975) -
Marans-Rodger Model
X X X X
Hourihan (1984) - Path
Analysis Model X X X
Morris and Winter (1978) -
Housing Adjustment Model
X X X X X
Francescato et al. (1979) -
Francescato et al.’s Model of
Housing Satisfaction
X X X X
Weidemann and Anderson
(1985) - Integrated Conceptual
Model
X X X X X X
Marans and Sprecklemeyer
(1981) - Inclusive Model
(Basic conceptual Model)
X X X X X X
Source: Author’s Literature review
The conceptual model theorizes that residential satisfaction is established by the relationship
that exists between the exogenous variables, which include the basic elements by which the
subjective and objective measurements are linked. These variables identified from the review
of literature and from the Delphi Survey findings are considered the major determinants of
389
residential satisfaction in subsidised low-income housing. These have been adopted to fit with
the peculiar housing and other socio-economic characteristics of the South African society.
Hence, the combination of the objective and subjective measures will then produce a measure
of residential satisfaction for the beneficiaries of the low-income units, as defined in the
previous sections.
9.4 STRUCTURAL COMPONENT OF THE MODEL
The present conceptual model hypothesis that occupant (beneficiaries’) residential satisfaction
(RS) with publicly provided low-income housing in South Africa (developing countries) is
derived from residents’ overall satisfaction with their dwelling unit features (DUF),
neighbourhood and environment features (NDF), building quality features (BQF), services
provided by the government (SPG), beneficiaries’ needs and expectations (NAE) and the
assessment of the beneficiaries participation (BNP) in the housing process. The model to be
tested in the hypothesis postulates a priori that RS is a multidimensional structure composed
of DUF, NDF, BQF, SPG, BNP and NAE. This is presented schematically in Figure 9.1 (Model
1.0). The theoretical underpinning of this priori is derived from the works of Marans and
Rodger (1975) and Marans and Sprecklemeyer’s (1981) models of satisfaction and the
approach as adopted by Galster and Hesser (1981), Cutter (1982) and Weidemann and
Anderson (1985), as discussed above.
Inherent in the conceptualized model is the notion that satisfaction with housing provision is
related to the evaluation of many variables. While the principal variable under consideration is
the dwelling unit that has been received by the beneficiaries, it is difficult to discuss it without
reference to variables of neighbourhood; service provided by the government and inclusion of
the other exogenous variables. The satisfaction level is expressed by the households’ subjective
evaluation of their housing environment as defined by them. The evaluation will depend on the
beneficiaries’ assessment of several indicator variables under each of the exogenous variables.
Which attributes are most relevant, is an empirical question and may differ under different
circumstances. How households assess a particular aspect of their housing environment, for
instance, is considered to be dependent on their personal characteristics. This is meant to
include all characteristics and experiences of the beneficiaries that influence their evaluations.
For example, beneficiaries of different cultures, races, income or gender may have diverse
evaluation of the same dwelling unit they have received. As such, Ebong (1983) describes that
390
depending on an individual’s diverse value system, experiences and aspirations, certain
residential features can generate in residents feelings of convenience, beauty, health and
accessibility, or otherwise. Thus it is possible for one to see in any residential environment
what one is supposed to see, turning a blind eye to much that is actually present in that particular
environment. In this study, the objective evaluation of residential environment will be assessed
by measuring the actual condition of the housing environment (quality assessment of the
housing units), which is an exogenous variable in the model.
Figure 9.1: An Integrated Conceptual Model of Residential Satisfaction
(Model 1.0)
9.5 MEASUREMENT COMPONENT OF THE MODEL
The hypothesized measurement component of the model comprises of the following residential
satisfaction factors: DUF = 17 measurement variables; NDF = 22 measurement variables; BQF
= 16 measurement variables; SPG=13 measurement variables; BNP = 5 measurement
variables; NAE = 4 measurement and RS = 7 measurement manifest variables. In the present
model, it is therefore theorized that residential satisfaction is to be considered as a sufficient
391
indication to show the success of state subsidised housing provision in meeting the
beneficiaries’ needs and giving them a good start into the housing market.
9.6 CONCLUSION
In this chapter, a conceptual model was theorized, which postulates a priori that RS is a
multidimensional structure composed of six latent variables of dwelling unit features,
neighbourhood features, building quality features, services provided by the government,
beneficiary’s participation and needs and expectations of the beneficiaries’. These factors were
derived from the literature review and findings from the Delphi Study concurrently. Also
highlighted in this chapter is the theoretical framework for the explanation of the variables
selected for the construction of the integrated conceptual model of residential satisfaction.
Findings for the validation of the conceptual model developed in Chapter Nine will be
presented in the next chapter.
392
CHAPTER TEN
SURVEY RESULTS
10.1 INTRODUCTION
The last chapter discussed and presented the research’s conceptual model, as shown in Figure
9.1 (Model 1.0). The theory behind the hypothesised integrated holistic residential satisfaction
model is based on literature and on expert’s opinion acquired during the Delphi Study as
described in Chapter Nine. This chapter presents descriptive statistics, inferential statistics and
hypotheses testing results based on the questionnaire analysis.
Therefore, raw data from the questionnaire were entered into the Statistical Package for Social
Sciences (SPSS) software and was later exported to the Structural Equation Modelling (SEM)
software EQS Version 6.2 for analysis (Bentler, 1999). The sample used for the analysis of the
model was 751 cases. Sample size significantly affects model fit (Tong, 2007). Smaller sample
sizes contribute to greater model fit bias (Tong, 2007). Kline (2005:15) states that a sample
size of 751 is classified as large. A small sample of less than 100 cases tends to be challenging
when it comes to SEM analysis (Harris & Schaubroeck, 1990; Kline, 2005:15). The appropriate
sample size is dependent on observed variables (MacCallum, Browne, & Sugawara, 1996;
Tong, 2007). A general guideline is that in order to use SEM for the data analysis purpose, a
study should have more than 200 respondents with a certain number of observed variables
(Bentler & Chou, 1987; Bollen, 1989). Tong (2007) further suggested that the variable ratio of
an ideal SEM model should be at least 5:1. In other words, a SEM model with 10 observed
variables should have more than 50 respondents. The researcher has collected 751 responses,
which are considered appropriate for the present study with 78 hypothesised observed
variables. The variable ratio to sample size for this study is 10.45:1, which meets the literature’s
(Tong, 2007) recommendation.
10.2 DESCRIPTIVE STATISTICS
A total of 751 responses were realized for analysis in this study after the close of the survey.
Figure 10.1 is a breakdown of the survey locations. From the 751 responses, 24.37% were from
the City of Tshwane, 30.63% were from Ekurhuleni Metropolitan Municipality, 31.96% from
the City of Johannesburg and 13.05% from Mogale City, a district municipality.
393
Figure 10.1: Survey locations
Of the total 751 responses, 42.88% (N = 322) were males and 57.10% (N = 429) were females,
as shown on Table 10.1.
Table 10.1: Respondents’ Demographic and Socio-economic Characteristics
Socio-demographic characteristics Frequency (n = 751) Percent (%)
Gender Male 322 42.88
Female 427 57.1
Race African 645 85.89
Indian 6 0.80
Coloured 78 10.39
White 22 2.93
Age 19 - 24 48 6.4
25 - 30 104 13.9
31 - 40 216 28.8
41 - 50 193 25.7
51 - 60 106 14.1
61 - 70 66 8.8
> 71 17 2.3
Educational level None (Did not attend any school) 38 5.1
Primary (Grade 1-7) 141 18.8
Secondary (Grade 8-11) 237 31.6
Matric (Grade 12, Std 10) 240 32.0
Post Matric Diploma (Registered) 37 4.9
Post Matric Diploma (Completed) 38 5.1
Bachelor’s / Post-graduate (Registered) 5 0.7
Bachelor’s / Post-graduate (Completed) 7 0.9
Others 2 0.3
Current employment status Employed (full time) 191 25.4
Employed (Part-time) 92 12.3
Self employed 74 9.9
Unemployed, looking for work 268 35.7
Unemployed, not looking for work 38 5.1
Housewife 18 2.4
31.96%
13.05%
30.63%
24.37%
0 5 10 15 20 25 30 35
City of Johannesburg
District Municipality City (Mogale City)
Ekurhuleni Metropolitan Municipality
City of Tshwane
Percentage
Su
rvey
Lo
cati
on
s
394
Student 24 3.2
Retired 45 6.0
Others 1 0.1
Employment sector Government 80 22.6
Private sector 190 53.7
Self employed 80 22.6
Others 4 1.1
Marital status Married 213 28.4
Single (never married) 334 44.5
Single (married but separated from spouse) 38 5.1
Living together (co-habiting) 64 8.5
Divorced 36 4.8
Widow 50 6.7
Widower 15 2.0
Annual family income Less than R5 000 97 13.6
R5 000 – 10 999 92 12.9
R11 000 – R15 999 112 15.7
R16 000 – R20 999 182 25.5
> R20 999 231 32.4
Length of residency Less than 1 year 26 3.5
1 – 2 years 48 6.4
3 - 5 years 200 26.8
6 - 8 years 158 21.2
> 8 years 314 42.1
Numbers of rooms in the dwelling unit 1 room 297 39.7
2 rooms 444 59.4
3 rooms 6 0.8
4 rooms 1 0.1
Numbers of bedrooms in the dwelling unit 1 room 386 51.6
2 rooms 362 48.4
Dependents in the household - children < 19 years
None 10 1.7
1 child 182 30.8
2 children 219 37.1
3 children 97 16.4
4 children 49 8.3
5 children 23 3.9
6 children 7 1.2
7 children 3 0.5
- adults: 19-59 years None 2 0.3
1 person 120 16.5
2 persons 311 42.8
3 persons 170 23.4
4 persons 92 12.7
5 persons 19 2.6
6 persons 10 1.4
7 persons 2 0.3
395
None 32 17.0
1 130 69.1
2 25 13.3
3 1 0.5
4 32 17.0
Type of government grants received State pension 119 24.0
Child support grant 343 69.2
Disability grant 24 4.8
Foster care grant 10 2.0
Others 0 -
Greatest need Housing 101 13.4
Employment 361 48.1
Education 106 14.1
Safety 162 21.6
Privacy 21 2.8
The ethnic composition was 85.89% Africans, followed by 10.39% Coloured, 2.93% White
and 0.80% Indians as revealed in Table 10.1. Majority of the respondents (28.8%) were
between the ages of 41 – 50, followed by the age group of 41 – 50 (25.7%) and the aged (> 60)
constituted (11.1%) of the sample. The highest education level of the majority of the sample
respondents was Matric (32.0%; N = 240). The respondents’ employment status findings
revealed that the numbers of those employed both full and part-time were 37.7%, while those
unemployed and looking for work was 35.7%, unemployed and not looking for work 5.1%,
followed by 9.9% who had their own businesses. Majority of the respondents (53.7%) were
employed in the private sector, followed by the government sector (22.6%). Although families
with 2 children below the age of 19 years were dominant (37.1%; N = 219), 42.8% of the
respondents had 2 persons (adults) between the age of 19 – 59 years and 69.1% elderly (> 60
years) in their dwellings. Also, 59.4% respondents have a total of 2-rooms in their dwelling
and 51.6% have 1-bedroom. A large percentage of the respondents (42.1%; N = 314) have
stayed more than 8 years in their allocated dwelling units and 3.5% stayed for less than 1 year.
The mean annual family income of majority (32.4%) of the respondents was more than R20 999
($2 560), followed by 25.5% whose annual household earnings were between R16 000 ($1 951)
and R20 000 ($2 439). Also, a large percentage of the respondents (69.2%) receives child
support grant, 24.0% receives state pensions and 4.8% receive disability grants. The
respondents also informed that their greatest need is employment (48.1%), followed by need
for safety (21.6%), education need (14.1%) and housing need accounting for 13.4%, as shown
in Table 10.1.
396
Table 10.2: Available Dwelling Unit Features
Room/object (in house) Present Not present
Bedroom(s) Count 739 6
Percentage 99.2% 0.8%
Living room Count 203 542
Percentage 27.2% 72.8%
Dining room Count 111 632
Percentage 14.9% 85.1%
Kitchen Count 621 123
Percentage 83.5% 16.5%
Bathroom(s) Count 473 272
Percentage 63.5% 36.5%
Wardrobes Count 46 697
Percentage 6.2% 93.8%
Play space for children Count 38 705
Percentage 5.1% 94.9%
Study space for children Count 19 724
Percentage 2.6% 97.4%
From the survey assessment of the available dwelling unit features, majority (99.2%) of the
respondents informed that they have bedrooms in their dwelling unit, 0.80% said there is no
bedroom, as shown in Table 10.2. Also, 83.5% informed they have kitchen in their dwelling
unit, while 16.5% said they do not have one. This was followed by 63.5% who have bathrooms
inside their dwelling, whilst 36.4% do not have bathrooms in their dwellings.
Table 10.3: Available Dwelling Unit Services Features
Services (in house) Present Not present
Water for domestic use Count 744 4
Percentage 99.5% 0.5%
Sanitary fittings (e.g. shower,
bath, toilet, basin, taps)
Count 733 15
Percentage 98.0% 2.0%
Electricity Count 743 5
Percentage 99.3% 0.7%
Furthermore, when the in-house services available were assessed, the result revealed that
99.5% respondents have water for domestic use in their houses, while only 0.5% do not have
water. Although, respondents (98.0%) have sanitary fittings (meaning: shower, bath, toilet
basin, wash hand basin and taps), 2.0% do not have these in their houses, as shown in Table
10.3.
397
Table 10.4: Available Private / Public Neighbourhood Features
Facility (Private / Public services) Present Not present
Shopping area Count 463 285
Percentage 61.9% 38.1%
Place of worship Count 627 121
Percentage 83.8% 16.2%
Parking facilities Count 45 703
Percentage 6.0% 94.0%
Playground/recreational
facilities
Count 335 413
Percentage 44.8% 55.2%
Community hall Count 289 458
Percentage 38.7% 61.3%
Disabled facilities Count 42 705
Percentage 5.6% 94.4%
The assessment of the available private and public neighbourhood features revealed 61.9%
have a shopping mall within their neighbourhood, while 38.1% informed there is no shopping
mall in their neighbourhood. The majority (83.8%) have a place of worship in their vicinity,
while 16.2% do not. Likewise, 44.8% informed they have a playground / recreational facility
in their neighbourhood, while a majority (55.2%) said they do not have such a place in their
neighbourhood. Also, majority (94.4%) informed they do not have facilities for the disabled in
their neighbourhood, while only 5.6% said they have in their neighbourhood, as shown in Table
10.4.
Table 10.5: Available Government Neighbourhood Features
Service (Government services) Present Not present
Nursery school (Private or public) Count 572 179
Percentage 76.2% 23.8%
Primary school (Private or public) Count 511 240
Percentage 68.0% 32.0%
High school(Private or public) Count 385 366
Percentage 51.3% 48.7%
Hospital/clinic Count 276 475
Percentage 36.8% 63.2%
Police services Count 249 502
Percentage 33.2% 66.8%
Fire protection services Count 101 647
Percentage 13.5% 86.5%
Public transport Count 713 38
Percentage 94.9% 5.1%
Count 631 120
398
Drainage system (within
neighbourhood or outside)
Percentage 84.0% 16.0%
Garbage and waste collection Count 695 56
Percentage 92.5% 7.5%
Furthermore, when the presence or absence of some listed government services was assessed,
findings emanating from the survey revealed that a majority (94.9%) have access to public
transport, followed by 92.5% who informed they have access to garbage and waste collection,
84.0% have a drainage system (within neighbourhood or outside). However, the respondents
(86.5%) further indicated that they do not have fire protection services in their neighbourhood,
followed by 66.8% who do not have police services, 63.2% do not have access to hospital/clinic
in their neighbour, 48.7% do not have high school either private or public in their
neighbourhood and a combined response of 55.8% do not have primary / nursery schools
(either private or public) in their area, as shown in Table 10.5.
10.3 INFERENTIAL STATISTICS
10.3.1 Structural Equation Modelling (SEM)
SEM was preferred to other statistical approaches, such as ANOVA and regression because it
displayed better conditions to demonstrate causality. According to Hoyle (1995:10), there are
three necessary conditions to demonstrate causality, which are: association, isolation and
directionality. While SEM is not distinctive in the first aspect, for isolating and putative causes,
SEM is more flexible and comprehensive than any univariate or multivariate modelling
approaches, providing means of controlling not only for extraneous or confounding variables
but for measurement error as well. Directionality, finally, is often greatly misunderstood. It can
be shown with many statistical procedures because it comes from theory (research design) and
sample logic. However, when the model as a whole, produces a good-fit, the result greatly
supports the individual causal relationships within the model.
It is clear from the research that measuring residential satisfaction is a complex construct. This
coupled with the advantages of SEM (just cited) and others precipitated the decision in this
study to utilize SEM in conjunction with EQS to attempt to examine factors that determine
residential satisfaction in South African subsidised low-income housing. EQS Version 6.2
399
software was used to investigate the measurement model adequacy and structural model
goodness-of-fit.
Structural Equation Modeling Analytic Strategy
This study aims to test a model of residential satisfaction in a sample of government subsidised
low-income housing in three South African metropolitan municipalities and one district
municipality. The analyses conducted can be divided into two steps. First, a series of
Confirmatory Factor Analysis (CFAs) to test for the measurement equivalency for each of the
six proposed latent constructs and the manifest or composite variables of residential satisfaction
represented in the hypothesised model of residential satisfaction (Model 1.0). The CFA results
defined the relations between the observed and unobserved variables. In order words, it
provided the link between scores on a measuring instrument and the underlying constructs they
are designed to measure. This was done to reaffirm the factor structure of the observed and
unobserved variables, hence, the construct validity. Secondly, the fit of the entire measurement
model underlying the hypothesised structural model was tested. The structural model defined
the relationship amongst the different exogenous variables and specified the manner by which
each exogenous variable directly or indirectly influences the changes in the values of other
exogenous constructs in the model, thus, defining the endogenous variables (residential
satisfaction). All analyses were performed using EQations software (EQS), including testing
of the hypothesised Structural Equation Models. In SEM, a covariance matrix generated from
a particular sample is compared with the covariance matrix generated from the hypothesised
model and fit statistics are used to determine the acceptability of the solution obtained. Hu and
Bentler (1999) and other scholars have recommended using a combination of fit statistics to
evaluate the fit of models (refer to Chapter 7, Section 7.4.4.11), as adopted for the current
study.
Statistics on SEM Assumptions – Outliers and Missing Data
An inspection of the data sets revealed that some data sets had missing values. A detailed
examination of the pattern of missing data revealed that the missing data was missing at random
(MAR) and not missing completely at random (MCAR). According to McDonald and Ho
(2002:70), the condition that data was missing completely at random is a situation where the
presence or absence of the observation is independent of other observed variables and the
variable itself. Hence, McDonald and Ho (2002) posit that the condition MCAR is a very strict
assumption that may be difficult to justify in practice. Therefore, the assumption of the
400
condition MAR was adopted. Hence, the robust maximum likelihood estimation solution in
EQS was used to address the problem, as discussed in Chapter 7 (refer to Section 7.4.4.11).
This method produces better results compared to other methods (Kline, 2005). The assumption
in this method was that the means, variances and covariances were sufficient statistics.
Consequently, cases with missing variables were skipped and not included in the analysis.
Likewise, further examination of the data set revealed that there were a few outliers in the data.
The EQS result output included case numbers with the largest contribution to Mardia’s
Normalized Multivariate Kurtosis. Examination of these case numbers showed the case
numbers that include outliers and it was upon these inspections that the conclusion was reached
that there were a few outliers in the data. The chosen method of estimation namely, Robust
Maximum Likelihood (RML) was adequate in addressing the problems of outliers. Boomsma
(2000:469) states that RML method is reliable because it replaces ordinary sample covariances
with the robust estimates of the covariances.
Statistics on SEM Assumptions – Data Distribution Characteristics
The estimation method of maximum likelihood assumes multivariate normality. Hence, it was
necessary that the distribution characteristics of the data were established before model analysis
could commence. The EQS result output included univariate statistics such as mean, skewness,
Kurtosis and the respective standard deviations. Similarly, the multivariate Kurtosis formed
part of the result output. Analysis of the univariate statistics and Mardia based multivariate
Kurtosis suggested non-normality in the sample data set. All Mardia estimates of multivariate
Kurtosis were greater than the upper limit value of 3.0 (DeCarlo, 1997:292). Hence, the data
distributions were described as highly Kurtotic. The non-normality of the data led to the
adoption of the robust maximum likelihood estimation method of the postulated model. The
results in the following sections are reported using robust statistics for the chi-square (Satorra-
Bentler Scaled Statistics; Satorra & Bentler, 1988). In all models, the first item of each factor
is fixed to establish the factors’ scale. Mardia’s coefficient and other univariate statistics are
presented in Table 10.6.
Statistics on SEM Assumptions – Identifiability of the Model
A further requirement for SEM analysis is the identifiability of the structural model. In order
for a model to be analysed, it has to fulfill the conditions of model identification. It is the duty
401
of the researcher to examine whether a model is theoretically identified or not (Boomsma,
2000:466).
Table 10.6: Univariate and Mardia’s Normalized Multivariate Estimates
Latent
Constructs
Indicator
Variable
Mean
(x̅)
Skewness
(G1)
Kurtosis
(G2)
SD
(σx)
Mardia’s
coefficient
DUF1 2.5627 -0.0366 -1.3116 1.1593
DUF2 2.2583 0.4353 -0.9336 1.1014
DUF3 2.2440 0.3713 -0.9461 1.0567
Dwelling Unit DUF5 2.0549 0.3265 -0.9512 0.9418 26.5463
Features DUF9 2.0580 0.3509 -0.9005 0.9413
DUF12 2.2600 0.3679 -0.9478 1.0878
DUF16 2.6872 -0.1719 -0.8425 1.0197
DUF17 2.3688 0.1245 -1.1463 1.0713
NAE1 3.7710 -0.9388 0.1307 1.1608
Needs and NAE2 3.6059 -0.6563 -0.6639 1.2782 44.4301
Expectations NAE3 3.9640 -1.1267 0.5926 1.1408
NAE4 3.9574 -1.2076 1.1561 1.0650
BNP1 3.6804 -0.7626 -0.3352 1.1996
Beneficiary BNP2 3.6245 -0.6117 -0.4375 1.1765
Participation BNP3 3.5273 -0.4537 -0.6302 1.1906 56.0118
BNP4 3.6605 -0.6647 -0.2492 1.1515
BQF2 2.8797 -0.1177 -0.8336 1.1749
BQF3 3.7059 -0.9052 0.2404 1.0886
Building BQF4 3.0616 -0.3317 -0.8867 1.1768 27.4688
Quality BQF5 3.1004 -0.4277 -0.7469 1.1447
Features BQF10 3.2289 -0.3768 -0.7007 1.1852
BQF11 3.1744 -0.1771 -0.9111 1.1548
NDF1 3.3783 -1.0117 0.4860 0.9385
NDF3 3.2473 -0.8557 0.1551 0.9423
Neighbourhood NDF5 3.5456 -0.6751 -0.0309 0.9130 19.7754
Features NDF7 3.2021 -0.3421 -0.7902 1.1673
NDF10 2.8995 -0.3767 -1.1078 1.1073
SPG8 2.3573 0.2312 -0.7826 1.0259
Services SPG9 2.4826 -0.0094 -1.0952 1.0520 21.4192
provided by SPG12 2.6395 -0.0957 -0.8525 1.0398
Government SGP13 2.5912 -0.1056 -0.9775 1.0260
RS1 3.5047 -0.4573 -0.6733 1.1580
Residential RS3 3.8064 -0.6154 0.8174 0.7466 13.1652
Satisfaction RS5 3.1092 -0.1929 -0.7706 1.1414
RS7 3.5952 -0.6868 -0.4518 1.2431
Kline (2005:105) informs that a model is said to be identified if it is theoretically possible to
derive a unique estimate for each parameter. While, an identification test is presented in the
result section, identification is a property of the model and not the data. However, checking for
model identification is a requirement before model analysis can commence. Hence, a model
402
that is not identified remains so, no matter the sample size and any effort to analyse it may
prove unsuccessful (Kline, 2005:105).
Therefore, a model is said to be identified, if there are at least as many observations as free
model parameters (namely, the degree of freedom ≥ 0) and that every unobserved variable must
be assigned a scale (Kline, 2005:105). A model could be just-identified, over-identified or
under-identified (Byrne, 2006:31). Byrne (2006:31) further explains that an over-identified
model is one, in which the number of parameters to be estimated is less than the number of
data variances and covariances of the observed variables and therefore, results in a positive
degree of freedom. The significance of an over-identified model is that it allows for a model to
be rejected and therefore, rendering it of scientific value (Byrne, 2006:31). On the other hand,
a just-identified model cannot be rejected and it is impossible to obtain a solution for an under-
identified model. Examination of the EQS result outputs indicated that the lowest value for the
degree of freedom in the current study was 2 and the highest was 20. These scores indicated a
positive value of degree of freedom and therefore, were suggestive of an over-identified model.
10.3.2 Confirmatory Factor Analysis of the Latent Construct
To determine if the measures used for assessing the exogenous variables (dwelling unit
features, neighbourhood features, building quality, services provided by government,
beneficiary’s participation, needs and expectations) and the endogenous variable (residential
satisfaction) where sufficient indicators, a Confirmatory Factor Analysis (CFA) was conducted
in order to assess the coefficients and to reaffirm the factor structure of each construct. This
was in line with the recommendation of Byrne (2006) who states that the first step in assessing
measurement invariance is to conduct separate CFAs of the latent constructs.
Using EQS 6.2 Statistical Software, the measurement model was further explored using CFA
to assess the fit of the items to the latent variables. If the fit of each of these models is good
and the item loading is acceptable, it can be assumed that the indicators underlying the factor
are tapping into the construct at hand in each of the latent constructs. In keeping with the
practice established by McDonald and Ho (2002) and as echoed by the recommendations of
other experts in SEM, who attest that evaluation of models should be derived from an array of
criteria, rather than a single ‘magic index’ (Byrne, 2006; Kline, 2005). Therefore, various
goodness-of-fit indexes were considered in the study to determine the goodness-of-fit.
403
10.3.3 Fit Statistics on Measurement Models (CFA)
10.3.3.1 Measurement Model for Dwelling Unit Features (DUF) Construct
This section presents a unidimensional model for dwelling unit features. The number of cases
that were analysed for this construct (DUF) was 700 from a sample of 751. Fifty-one cases
were skipped (ignored / not used) because of missing variables. The initial model for this
construct consisted of 17 observed variables. However, from the preliminary CFA analysis,
nine indicator variables (DUF4, DUF6-8, DUF10-11 & DUF13-DUF15) had an unacceptably
high unstandardized and standardized residual covariance matrix (ranging from 2.88 – 3.20),
hence they were dropped. A residual covariance matrix value greater than 2.58 is described as
large (Byrne, 2006:94; Joreskog & Sorbom, 1988). In order for a variable to be included in a
CFA Analysis, thus enabling the model to be described as well-fitting, the distribution of
residuals covariance matrix should be symmetrical and centred around zero (Byrne, 2006:94;
Joreskog & Sorbom, 1988). The remaining eight-indicator model provides good measures of
residual matrix and evidence of convergent validity.
Figure 10.2: Measurement model of dwelling unit features
The CFA results further revealed that the dwelling unit features had 8 dependent variables, 9
independent variables and 16 free parameters. The number of fixed non-zero parameters was
9. The eight dependent indicator variables for the dwelling units were: location of bedroom,
number of bedrooms, size of bedroom(s), location of dining room, size of wardrobe/closet,
amount of privacy within the house, overall appearance of the house and overall size of the
E17
DUF1
DUF2
DUF3
DUF5
DUF9
DUF12
DUF16
DUF17
DUF
E1
E2
E3
E5
E9
E12
E16
404
house. These indicator variables are presented in Table 10.7. The dwelling unit features
measurement model shown in Figure 10.2 was analysed before it could be included in the full
latent variable model.
In order to establish how well the mode fit the sample data and the strength of the hypothesised
relationship between the variables, results on residual covariance matrix (unstandardized and
standardized), distribution of standardised residuals, fit statistics and statistical significance at
a probability level of 5% were examined. In addition, the Cronbach’s alpha and the Rho
Coefficient of Internal Consistency were examined to determine the score reliability. Results
of these statistics are presented in the next section for the dwelling unit features.
Table 10.7: Postulated Dwelling Unit Features Model
Latent constructs Indicator variables
(How satisfied or dissatisfied are you
with…)
Label
Dwelling Unit Features Number of bedrooms DUF1
(DUF) Location of bedroom(s) DUF2
Size of the bedroom DUF3
Location of dining room DUF5
Size of wardrobe/closet DUF9
Amount of privacy within the house DUF12
Overall appearance of the house DUF16
Overall size of the house DUF17
Diagnostic Fit Analysis: Analysis of Residual Covariance Estimate
The unstandardized and standardized absolute residual matrix values of the dwelling unit
features are presented in Table 10.8 and 10.9. The result reveals that all the absolute residual
values and the average off-diagonal absolute residual values were close to zero. The
unstandardized average off-diagonal residual was 0.0364 while the standardized average off-
diagonal residual was found to be 0.0349. A residual value greater than 2.58 is described as
large (Byrne, 2006:94). The results obtained for the Dwelling Unit Features Measurement
Model suggested a fairly acceptable fit to the sample data because the absolute residual were
all less than 2.58. In order for a model to be described as well-fitting, the distribution of
standardized residuals should be symmetrical and centred around zero (Byrne, 2006:94).
405
Table 10.8: Residual Covariance Matrix for Dwelling Unit Model (Unstandardized)
Unstandardized Residual Covariance Matrix
DUF1 DUF2 DUF3 DUF5 DUF9 DUF12 DUF16 DUF17
DUF1 0.000
DUF2 0.009 0.000
DUF3 0.033 0.019 0.000
DUF5 -0.024 0.026 0.017 0.000
DUF9 0.022 -0.046 -0.012 0.151 0.000
DUF12 0.014 -0.028 -0.012 -0.013 0.099 0.000
DUF16 -0.020 -0.054 -0.034 -0.050 -0.045 0.005 0.000
DUF17 -0.042 0.008 -0.018 -0.012 -0.045 0.008 0.154 0.000
Average absolute residual = 0.0283
Average off-diagonal absolute residual = 0.0364
% falling between -0.1 +0.1 = 99.99%
Further review of the frequency distribution reveals most residual values (99.99%) fall between
-0.1 and +0.1, which is in the acceptable range. Of the remaining residuals, 0.01% fell outside
the -0.1 to 0.1 ranges.
Table 10.9: Residual Covariance Matrix for Dwelling Unit Model (Standardized)
Standardized Residual Covariance Matrix
DUF1 DUF2 DUF3 DUF5 DUF9 DUF12 DUF16 DUF17
DUF1 0.000
DUF2 0.007 0.000
DUF3 0.027 0.016 0.000
DUF5 -0.022 0.025 -0.018 0.000
DUF9 0.020 -0.044 -0.012 0.171 0.000
DUF12 0.011 -0.023 -0.011 -0.013 0.096 0.000
DUF16 -0.017 -0.048 -0.031 -0.052 -0.047 0.004 0.000
DUF17 -0.034 0.007 -0.016 -0.012 -0.045 0.007 0.141 0.000
Average absolute residual = 0.0283
Average off-diagonal absolute residual = 0.0364
% falling between -0.1 +0.1 = 99.99%
From this information, the results suggested a measurement model that was well fitting albeit
minimal discrepancy in fit between the hypothesised model and the sample data. Therefore,
since this diagnostic fit analysis indicated a good fit; further tests of goodness-of-fit were
possible to conclusively make a decision on the fit and appropriateness of the measurement
model.
406
Goodness-of-Fit Statistics – Robust Maximum Likelihood (RML)
The analysis strategy of goodness-of-fit for the dwelling unit feature followed a two statistics
strategy of fit indexes as recommended by Hu and Bentler (1999).
The sample data on dwelling unit features measurement model yield the S – Bχ2 of 190.724
with 20 degrees of freedom (df) with a probability of p = 0.0000. Hence, the chi-square was
insignificant. This chi-square value indicated that the departure of the sample data from the
postulated measurement model was not significant and hence, indicative of an acceptable fit.
However, the chi-square test is very sensitive to sample size and is used more as a descriptive
index of fit rather than as a statistical test (Kline, 2005:136). Therefore the normed Chi-square
value is usually adopted by most researchers. The normed chi-square is the procedure of
dividing the chi-square by the degrees of freedom. The normed values of up to 3.0 or even 5.0
are recommended (Kline, 2005:137). From the above chi-square and degrees of freedom values
the ratio was found to be 9.54. This ratio was higher than the limit of 3.00 or 5.0 advocated for,
by some authors (Kline, 2005:137).
Further, the CFI was found to be 0.955 and the SRMR was found to be 0.046. The CFI value
was higher than the cut-off limit of 0.95 for a mode to be described as having a good fit.
Similarly, the absolute fit index SRMR value of 0.046 was less than the cut-off value of 0.09
for a good fitting model. Therefore, the result showed a good fit model with an SRMR value
of not more than 0.05. These fit indexes for the dwelling unit measurement model suggested
that the postulated model adequately describe the sample data and could therefore, be included
in the full latent variable model analysis (Table 410.10).
Table 10.10: Robust Fit Indexes for Dwelling Unit Features Construct
Fit Index Cut-off value Estimate Comment
S – Bχ2 190.724
df 0≥ 20 Acceptable
CFI 0.90≥ acceptable
0.95≥ good fit
0.955 Good fit
SRMR 0.08≥ acceptable
0.05≥ good fit
0.046 Good fit
407
Statistical Significance of Parameter Estimates
Inspection of the correlation values, standard errors and the test statistics in Table 10.11
revealed that all correlation values were not greater than 1.00, Z-statistics were greater than
1.96 and the signs were appropriate. The estimates were therefore deemed reasonable, as well
as statistically significant. The parameter with the highest standardized coefficient was the
indicator variable DUF2 (numbers of bedrooms). The parameter coefficient was found to be
0.872.
Table 10.11: Factor loading and Z-statistics of Dwelling Unit Features Measurement
Model
Indicator
Variable
Unstandardized
Coefficient (λ)
Standardized
Coefficient (λ)
Z- Statistics
R2
Significant
at 5%
level?
DUF1 0.967 0.834 37.160 0.696 Yes
DUF2 0.960 0.872 39.181 0.761 Yes
DUF3 0.916 0.868 36.641 0.753 Yes
DUF5 0.637 0.677 21.487 0.459 Yes
DUF9 0.544 0.578 17.741 0.335 Yes
DUF12 0.753 0.692 23.748 0.479 Yes
DUF16 0.682 0.669 20.237 0.447 Yes
DUF17 0.910 0.850 37.198 0.723 Yes
(Robust Statistical Significance at 5% level)
The variable DUF2, which asked the beneficiaries of their level of satisfaction with the
numbers of bedrooms in the dwelling unit, was found to associate more with the dwelling unit
features than the other variables. However, all parameter estimates had high correlations values
close to 1.00. The high correlation values suggest a high degree of linear association between
the indicator variables and the unobserved variable (dwelling unit features). In addition, the R2
values were also close to the desired value of 1.00 indicating that the factors explained more
of the variance in the indicator variables. The results therefore, suggest that the indicator
variables significantly predict the unobserved construct, because all the measured variables are
significantly associated with the dwelling unit features.
Internal Reliability and Validity of Scores
The internal consistency and reliability of scores for the dwelling unit features construct was
determined from the Rho and the Cronbach’s Alpha Coefficient. According to Kline (2005:59),
408
the reliability coefficient should fall between zero and 1.00. While, values close to 1.00 are
desired. The Rho Coefficient of Internal Consistency was found to be 0.920. This value was
above the minimum required value of 0.70. Likewise, the Cronbach’s Alpha was above the
minimum acceptable value of 0.70. The Cronbach’s Alpha was found to be 0.915 (Table
10.12). Both of these values revealed a high level of internal consistency and therefore
reliability, suggesting that the indicator variables represent the same latent construct (dwelling
unit features).
Table 10.12: Reliability and Construct Validity of Dwelling Unit Feature Model
Factor Indicator
Variable
Factor
Loading
Cronbach’s
Alpha
Rho
Coefficient
Dwelling DUF1 0.834
Unit
Features
DUF2 0.872
DUF3 0.868
DUF5 0.677 0.915 0.920
DUF9 0.578
DUF12 0.692
DUF16 0.669
DUF17 0.850
*Parameter estimates are based on standardized solutions
Further, construct validity was determined from the magnitude and reasonableness of the
parameter coefficients (factor loading). The parameter coefficients represent the magnitude of
correlation or covariance between an item and a construct. Higher parameter coefficients show
that the indicator variables have a stronger relationship with a construct and thus converge at a
common point. Parameter coefficients of greater than 0.5 indicate a close relationship between
the construct and an indicator variable. A parameter coefficient of 0.5 is interpreted as 25% of
the total variance in the indicator variable being explained by the latent variable (factor). Hence,
a parameter coefficient should be 0.5 or higher, and ideally 0.7 or greater to explain about 50%
of the variance in an indicator variable (Hair et al., 1998:111). The standardized parameter
coefficient presented in Table 10.12, revealed that all coefficients were significantly higher
with the lowest being 0.578 for the relationship between DUF9 and the dwelling unit feature.
This parameter estimate suggests that the dwelling unit construct accounts for 53% of the
variance in DFU9 (beneficiaries’ satisfaction with the size of the wardrobe/closet). The
magnitude of the parameter estimate was above the 50% minimum. This in addition, indicates
409
a strong relationship between the indicator variables and the factors of the dwelling unit
features construct.
Therefore the dwelling unit feature construct satisfied both internal reliability and the construct
validity criteria. The Rho value was above the minimum value of 0.70, the magnitude, signs
and statistical significance of the parameter estimates were appropriate (Table 412).
Summary on Dwelling Unit Feature Measurement Model
The CFA analysis revealed that the residual covariance estimates fell within the acceptable
range, likewise, the robust fit indexes met the cut-off index criteria and all the parameter
estimates were statistically significant and feasible. Considering these criteria, the
measurement model for the dwelling unit feature was found to adequately fit the sample data.
Therefore, there was no need to improve the measurement model before it could be included
in the full latent variable model. Hence, the dwelling unit feature construct was adequately
measured by the eight indicator variables and could be used in the analysis of the full latent
variable model.
10.3.3.2 Measurement Model for Need and Expectation (NAE) Construct
The number of cases that were analysed for the need and expectation (NAE) construct were
751 which is equivalent to the sample size of 751. No case was skipped because there were no
missing variables. Initial CFA analysis revealed that the residual covariance matrix were within
the accepted range, as recommended by Byrne (2006:94). In order for a variable to be included
in a CFA analysis, thus enabling the model to be described as well-fitting, the distribution of
residuals covariance matrix (factor loadings) should be symmetrical and centred around zero
(Byrne, 2006:94) and should not be greater than 2.58.
From the examination of the Bentler-Weeks Structure Representation, the NAE has 4
dependant variables, 5 independent variables and 8 free parameters. The number of fixed non-
zero parameters was 5. The hypothesis that the NAE construct is explained by indicator
variables NAE1 to NAE4, as shown in Table 10.13, was therefore evaluated.
410
Figure 10.3: Measurement Model of Needs and Expectation
The needs and expectations measurement model, as schematically shown in Figure 10.3, was
analysed before it could be included in the full latent variable model. In order to establish how
well the mode fit the sample data and the strength of the hypothesised relationship between the
variables, results on residual covariance matrix (unstandardized and standardized), distribution
of standardised residuals, fit statistics and statistical significance at a probability level of 5%
were examined. In addition, the Cronbach’s Alpha and the Rho Coefficient of Internal
Consistency were examined to determine the score reliability. Results of these statistics are
presented in the following section for the NAE variable.
Table 10.13: Postulated Needs and Expectation Model
Latent constructs Indicator variables
(Owners should be…)
Label
Needs and Expectation Told beforehand the type of house they will
receive
NAE1
(NAE) Asked the type of house they need NAE2
Owners expect good quality houses NAE3
Our houses should meet our family’s need NAE4
Diagnostic fit analysis: Analysis of residual covariance estimate
The average absolute residual values of the needs and expectation construct are presented in
Table 10.14 and 10.15. An examination of the unstandardized and standardized absolute
residual matrix values of the NAE reveals that all the absolute residual values and the average
off-diagonal absolute residual values were close to zero.
The unstandardized average off-diagonal residual was 0.0496 whilst the standardized average
off-diagonal residual was found to be 0.0375. These values were considered to be very small,
and therefore, acceptable. An absolute residual value is considered to be large, if it is more than
2.58 (Byrne, 2006:94). The results obtained for the NAE measurement model were suggestive
E4
NAE1
NAE2
NAE3
NAE4
NAE
E1
E2
E3
411
of an acceptable fit to the sample data since all residual values were below the 2.58 cut-off.
Besides, 100% of standardized residuals fell between -0.1 and +0.1, which is the acceptable
range.
Table 10.14: Residual Covariance Matrix for Needs and Expectation Model
(Unstandardized)
Unstandardized Residual Covariance Matrix
NAE1
NAE2
NAE3
NAE4
NAE1 0.000
NAE2 0.062 0.000
NAE3 -0.055 -0.017 0.000
NAE4 -0.021 -0.055 0.088 0.000
Average absolute residual = 0.0298
Average off-diagonal absolute residual = 0.0496
% falling between -0.1 +0.1 = 100%
In order for a model to be described as well-fitting, the distribution of standardized residuals
should be symmetrical and centred around zero (Byrne, 2006:94). Results suggest a
measurement model that has an adequate fit. Therefore, since the above examination of
residuals indicated a good fit; further tests of goodness-of-fit will now be presented in the next
sections.
Table10.15: Residual Covariance Matrix for Needs and Expectation Model
(Standardized)
Standardized Residual Covariance Matrix
NAE1
NAE2
NAE3
NAE4
NAE1 0.000
NAE2 0.042 0.000
NAE3 -0.042 -0.012 0.000
NAE4 -0.017 -0.040 0.072 0.000
Average absolute residual = 0.0225
Average off-diagonal absolute residual = 0.0375
% falling between -0.1 +0.1 = 100%
Goodness-of-Fit Statistics – RML
A two statistic strategy of fit indexes is reported as recommended by Hu and Bentler (1999).
The sample data on the NAE measurement model yields the S – Bχ2 of 22.812 with 2 degrees
412
of freedom (df) with a probability of p = 0.0001. The chi-square was insignificant. This chi-
square value indicated that the postulated model significantly differed from the sample data.
However, the chi-square test is very sensitive to sample size and therefore not very reliable.
The chi-square test tends to be affected by the sample size with a propensity to reject models,
if the samples are large. Therefore, a normed chi-square value is usually adopted by most
researchers (Kline, 2005:137). Normed Chi-square is the procedure of dividing the chi-square
by the degrees of freedom. The normed values of up to 3.0 or even 5.0 are recommended (Kline,
2005:137). From the above chi-square and degrees of freedom values the ratio was found to be
11.406. This ratio was higher than the limit of 3.00 or 5.0 advocated for, by some authors
(Kline, 2005:137) and therefore, the model fit may be described as not acceptable.
Hence, it was thus necessary to engage the use of other fit indexes in the determination of the
model’s goodness-of-fit. The CFI was found to be 0.980 and the RMSEA with 90% confidence
interval (lower bound value = 0.077 and upper bound value = 0.163) was found to be 0.118.
The CFI value was higher than the cut-off limit of 0.95 for a mode to be described as having a
good fit. Likewise, the RMSEA value of 0.118 was higher than the lower and the upper bound
cut-off value of 0.08 for an acceptable fit. The model could be accepted but the fit is an average
fit (MacCallum et al., 1996). The absolute fit index SRMR was found to be 0.033, which was
within the cut-off criteria for good fit.
Table 10.16: Robust Fit Indexes for Needs and Expectations Construct
Fit Index Cut-off value Estimate Comment
S – Bχ2 190.724
df 0≥ 20 Acceptable
CFI 0.90≥ acceptable
0.95≥ good fit
0.955 Good fit
SRMR 0.08≥ acceptable
0.05≥ good fit
0.046 Good fit
RMSEA 0.08≥ acceptable
0.05≥ good fit
0.118 Acceptable fit
RMSEA
90% CI
0.077:0.163 Slightly out of range
These fit indexes (Table 10.16) for the needs and expectation measurement model suggested
that the measurement model had an adequate fit to the sample data. Further, the fit statistics
indicated that the model was working properly and could be included in the full latent model
413
analysis. In addition, parameter estimates were scrutinized to determine whether the model
worked properly and was reasonable. This involved assessing the magnitude, signs and
statistical significance of the parameter estimates. These statistics are presented in Table 10.17.
Statistical Significance of Parameter Estimates
Inspection of the correlation values, standard errors and the test statistics in Table 10.17
revealed that almost all correlation values were not greater than 1.00, except the unstandardized
coefficient value for NAE2, which was 1.126, while the standardized coefficient value was less
than 1.00; Z-statistics were greater than 1.96 (p<0.05) and the signs were appropriate (positive).
The estimates were therefore found to be reasonable, as well as statistically significant. The
parameter with the highest standardized coefficient was the indicator variable NAE2 (owners
should be asked the type of house they need). The parameter coefficient was found to be 0.882.
Table 10.17: Factor loading and Z-statistics of Needs and Expectation Measurement
Model
Indicator
Variable
Unstandardized
Coefficient (λ)
Standardized
Coefficient (λ)
Z- Statistics
R2
Significant
at 5%
level?
NAE1 0.995 0.858 27.284 0.736 Yes
NAE2 1.126 0.882 38.755 0.777 Yes
NAE3 0.964 0.845 23.692 0.715 Yes
NAE4 0.873 0.820 20.153 0.673 Yes
(Robust statistical significance at 5% level)
The variable NAE2, which asked the beneficiaries their level of agreement, if an owner should
be asked the type of house they need, was found to be more closely associated with the needs
and expectation construct than all other variables. However, all standardized parameter
estimates showed high correlations values closer to 1.00, suggesting a high degree of linear
association between the indicator variables and the factor, needs and expectations. In addition,
the R2 values were found close to the desired value of 1.00, none were below 0.50, suggesting
that the factors explained more of the variance in the indicator variables. The results therefore
suggest that the indicator variables significantly predict the unobserved construct, because all
the measured variables are significantly associated with the needs and expectation variable.
414
Internal Reliability and Validity of Scores
In order to determine the internal consistency of the composition of the needs and expectation
measurement model, the Rho Coefficient and the Cronbach’s Alpha coefficient were examined
to establish reliability (Byrne, 2006:133). According to Kline (2005:59), the reliability
coefficient should be between zero and 1.00. Values close to 1.00 are desired. The Rho
Coefficient of internal consistency was found to be 0.915. This was above the minimum value
of 0.70. Likewise, the Cronbach’s Alpha was also found to be above the minimum value of
0.70 at 0.912. Both values showed a high level of internal consistency and therefore, reliability.
The construct validity was determined by examining the magnitude of the parameter
coefficients (factor loading). High parameter coefficients of greater than 0.5 indicate a close
relationship between the factor and an indicator variable. A parameter coefficient of 0.5 is
interpreted as 25% of the total variance in the indicator variable being explained by the latent
variable (factor). Accordingly, a parameter coefficient has to be between 0.5 - 0.7or greater to
explain about 50% of the variance in an indicator variable (Hair et al., 1998:111). Inspection
of the unstandardized parameter coefficient presented in Table 10.18, revealed that they were
significantly high with the minimum of 0.873 suggesting that the factor accounted for about
64% of the variance in NAE4. This value was however above the acceptable level. On the other
hand, all other parameter estimates were above 50% and therefore indicative of an adequate fit
between the indicator variables and the factor.
Table 10.18: Reliability and Construct Validity of Needs and Expectation Model
Factor Indicator
Variable
Factor
Loading
Cronbach’s
Alpha
Rho
Coefficient
Needs and
Expectation
NAE1 0.834
NAE2 0.872 0.912 0.915
NAE3 0.868
NAE4 0.677
*Parameter estimates are based on standardized solutions
Therefore, the factor needs and expectation satisfied both internal reliability and the construct
validity criteria. The Rho value was above the minimum value of 0.70 (Table 10.18) and the
construct validity criteria was justified by the magnitude, signs and statistical significance of
all parameter coefficients.
415
Summary on Needs and Expectation Measurement Model
The residual covariance estimates fell within the acceptable range, the robust fit indexes met
the cut-off index criteria, except for the RMSEA value and the RMSEA with 90% confidence
interval, which produced a poor fit, all other parameter estimates were statistically significant
and feasible. Based on these criteria, the measurement model for the needs and expectations
subscale was found to adequately fit the sample data. Further, there was no significant evidence
of model mis-specification. As a result, there was no need to modify a model that fits well,
because modification may only be fitting small characteristic features of the sample (Byrne,
2006:103). Therefore, the factor needs and expectations appeared to be explained by the
indicator variables NAE1 to NAE4 and hence adequately measured by the needs and
expectation constructs. The measurement model on needs and expectation could therefore be
used in the full latent variable model.
10.3.3.3 Measurement Model for Beneficiary Participation (BNP) Construct
The number of cases that were analysed for the beneficiary participation construct was 751
cases. No case was skipped. From the initial CFA statistical analysis one indicator variable
(BNP5) had an unsatisfactorily high residual covariance matrix factor loading (2.60), hence it
was dropped. Inspection of the Bentler-Weeks Structure representation for the construct
revealed that the BNP has 4 dependent variables, 5 independent variables and 8 free
parameters. The number of fixed non-zero parameter was 5.
Figure 10.4: Measurement Model of Beneficiary Participation
The 4 dependent indicator variables for the BNP, as presented in Table 10.19 and Figure 10.4
was analysed before it could be included in the full latent variable model. In order to establish
how well the mode fit the sample data and the strength of the hypothesised relationship between
the variables, results on residual covariance matrix (unstandardized and standardized),
E4
BNP1
BNP2
BNP3
BNP4
BNP
E1
E2
E3
416
distribution of standardised residuals, fit statistics and statistical significance at probability
level of 5% were examined. Additionally, the Cronbach’s Alpha and the Rho Coefficient of
internal consistency were examined for score reliability. Construct validity of the measurement
model was determined from model convergence and the magnitude of parameter coefficients.
Table 10.19: Postulated Beneficiary Participation Model
Latent constructs Indicator variables
(Owners should be consulted…
Label
Beneficiary
Participation
About the house location BNP1
(BNP) About the house design BNP2
About the house construction BNP3
About the internal finishes of the house BNP4
Diagnostic Fit Analysis: Analysis of Residual Covariance Estimate
The average absolute residual values of the BNP construct are presented in Table 10.20 &
10.21. Both unstandardized and standardized average absolute residual matrix values are
presented. Results revealed that all the absolute residual values and the average off-diagonal
absolute residuals, both unstandardized and standardized, were close to zero.
Table 10.20: Residual Covariance Matrix for Beneficiary Participation
(Unstandardized)
Unstandardized Residual Covariance Matrix
BNP1
BNP2
BNP3
BNP4
BNP1 0.000
BNP2 0.019 0.000
BNP3 -0.013 -0.001 0.000
BNP4 -0.005 -0.010 0.009 0.000
Average absolute residual = 0.0056
Average off-diagonal absolute residual = 0.0093
% falling between -0.1 +0.1 = 100%
The unstandardized average off-diagonal residual was 0.0093, while the standardized average
off-diagonal residual was found to be 0.0067. These values were smaller than the 2.58 upper
limits and therefore suggested that the model could have an adequate fit to the sample data. In
addition, 100% of standardised residuals fell between the acceptable range of -0.1 and +0.1
417
(Byrne, 2005:94). The favourable diagnostic fit analysis tests justified further tests of
goodness-of-fit on the beneficiary participation construct.
Goodness-of-Fit Statistics – RML
The sample data on beneficiary participation measurement model yield the S – Bχ2 of 2.104
with 2 degrees of freedom (N=751; p = 0.34917). The chi-square was insignificant. This chi-
square value indicated that the departure of the sample data from the postulated measurement
model was not significant and hence, indicative of an acceptable fit. The ratio of the chi-square
to the degrees of freedom was found to be 1.052. This ratio was lower than the upper limit of
3.0 or 5.0 (Kline, 2005:137). The measurement model was therefore considered to be of an
acceptable fit.
Table 10.21: Residual Covariance Matrix for Beneficiary Participation (Standardized)
Standardized Residual Covariance Matrix
BNP1
BNP2
BNP3
BNP4
BNP1 0.000
BNP2 0.013 0.000
BNP3 -0.009 -0.001 0.000
BNP4 -0.003 -0.007 0.007 0.000
Average absolute residual = 0.0040
Average off-diagonal absolute residual = 0.0067
% falling between -0.1 +0.1 = 100%
In addition to the chi-square test, the CFI was found to be 1.000. The CFI value was higher
than the minimum value of 0.95 set for good fit criteria. Moreover, the RMSEA (90% CI) was
found to be 0.008 (lower bound value = 0.000; upper bound value = 0.073). Likewise, the
RMSEA value of 0.008 was higher than the cut-off value of 0.05 for a good fitting model. It
fell within the acceptable range for a model to be considered fitting to the sample data. The
absolute fit index SRMR was found to be 0.006. This value met the cut-off of not exceeding
0.05 for a good fitting model. These fit indexes (Table 10.22) for the beneficiary participation
measurement model suggested that the model adequately fit the sample data and therefore,
could be included in the full latent variable model.
418
Table 10.22: Robust fit indexes for beneficiary participation construct
Fit Index Cut-off value Estimate Comment
S – Bχ2 2.104
df 0≥ 2 Acceptable
CFI 0.90≥ acceptable
0.95≥ good fit
0.955 Good fit
SRMR 0.08≥ acceptable
0.05≥ good fit
0.006 Good fit
RMSEA 0.08≥ acceptable
0.05≥ good fit
0.008 Acceptable fit
RMSEA
90% CI
0.000:0.073 Acceptable range
Furthermore, parameter estimates were analysed to determine whether the model worked
properly and hence was reasonable. This involved assessing the magnitude, signs and statistical
significance of the parameter estimates. These statistics are presented in Table 10.23.
Statistical Significance of Parameter Estimates
An assessment of the correlation values, standard errors and the test statistics in Table 10.23
show that 75% of the unstandardized correlation values were greater than 1.00, while 25% was
less than 1.00. However, all standardised coefficient values were not greater than 1.00, Z-values
were greater than 1.96 and the signs were appropriate and reasonable. All parameter estimates
were therefore considered to be reasonable as well as statistically significant. The parameter
with the highest standardised coefficient was the indicator variable BNP3. The parameter
coefficient was found to be 0.937. The indicator variable BNP3, which asked the beneficiaries
their level of agreement, if the owner should be consulted about the house construction, was
found to be more associated with the beneficiary participation factor than the other indicator
variable (BNP1, BNP2 & BNP4). Hence, all standardised parameter estimates had high
correlation values close to 1.00, suggesting that all indicator variables measured the beneficiary
participation factor.
The high correlation values suggest a high degree of linear association between the indicator
variables and the factor of beneficiary participation construct. In addition, the R2 values were
also found to be close to the desired value of 1.00 and hence indicated that the factors of
beneficiary participation explained the variance in the indicator variables. The results therefore
419
suggest that the indicator variables significantly predict the factor construct, because all the
measured variables are significantly associated with beneficiary participation.
Table 10.23: Factor loading and Z-statistics of Beneficiary Participation Measurement
Model
Indicator
Variable
Unstandardized
Coefficient (λ)
Standardized
Coefficient (λ)
Z- Statistics
R2
Significant
at 5%
level?
BNP1 0.984 0.821 25.908 0.674 Yes
BNP2 1.085 0.923 35.456 0.851 Yes
BNP3 1.115 0.937 39.528 0.878 Yes
BNP4 1.013 0.881 28.696 0.775 Yes
(Robust statistical significance at 5% level)
Internal Reliability and Validity of Scores
The Rho Coefficient and the Cronbach’s Alpha Coefficient were examined in order to establish
score reliability (Byrne, 2006:133). According to Kline (2005:59), the reliability coefficient
should fall between zero and 1.00. Values close to 1.00 are desired. The Rho Coefficient of
internal consistency was found to be 0.939. This value was above the minimum required value
of 0.70. Similarly, the Cronbach’s alpha was above the minimum acceptable value of 0.70 at
0.938. Both of these values indicated a high degree of internal consistency and homogeneity
(Table 10.24).
Construct validity was determined by examining the magnitude and signs of the parameter
coefficients. High parameter coefficients of greater than 0.5 indicate a close relation between
the factor and an indicator variable. A parameter coefficient of 0.5 is interpreted as 25% of the
total variance in the indicator variable being explained by the latent variable (factor). Therefore,
a parameter coefficient has to be greater than 0.5 or ideally 0.7 to explain about 50% of the
variance in an indicator variable (Hair et al., 1998:111). Inspection of the standardized
parameter coefficient in Table 10.24 revealed that all coefficients were significantly high with
the minimum factor loading being 0.821 for the relationship between BNP1 and the measured
factor. This parameter estimate of 0.821 suggested that the measured factor accounts for
62.15% of the variance in BNP1. Hence, the magnitude of the parameter estimate was above
420
the 50% minimum acceptable level, which indicates a strong relationship between the indicator
variables and the factors.
Table 10.24: Reliability and Construct Validity of Beneficiary Participation Model
Factor Indicator
Variable
Factor
Loading
Cronbach’s
Alpha
Rho
Coefficient
Beneficiary BNP1 0.821
Participation BNP2 0.923 0.938 0.939
BNP3 0.937
BNP4 0.881
*Parameter estimates are based on standardized solutions
As a result, the beneficiary participation construct satisfied both internal reliability and
construct validity criteria because the Rho value was above the minimum value of 0.70, the
magnitude, signs and statistical significance of the parameter estimates were fitting.
Summary on Beneficiary Participation Measurement Model
The measurement model for the beneficiary participation construct revealed an adequate fit to
the sample data. The residual covariance estimates fell within the acceptable range of -0.1 to
+0.1; the robust fit indexes (CFI = 1.00; RMSEA = 0.008 & SRMR = 0.006) met the cut-off
index criteria and the standardised parameter estimates were found to be statistically significant
at 5% level and were feasible. Consequently, there was no need to improve the measurement
model before including it in the full latent variable model.
10.3.3.4 Measurement Model for Building Quality Feature (BQF) Construct
The number of cases that were analysed for the beneficiary participation construct was 751
cases. The number of cases that were skipped was 6 because of missing variables. Primary
CFA results of the residual covariance matrix showed that ten indicator variables (BQF1, BQF6
– BQF9, & BQF12 - BQF16) had high values (3.20 - 4.52), hence they were dropped.
Examination of the Bentler-Weeks structure representation for the six variables that passed the
first CFA test revealed that the BQF construct has 6 dependent variables, 7 independent
variables and 12 free parameters. The number of fixed non-zero parameter was 7. The 6
dependent indicator variables for the BQF are presented in Table 10.25 and Figure 10.5. These
was analysed before it could be included in the full latent variable model.
421
Figure 10.5: Measurement Model of Building Quality Features
In order to establish how well the mode fit the sample data and the strength of the hypothesised
relationship between the variables, results on residual covariance matrix (unstandardized and
standardized), distribution of standardised residuals, fit statistics and statistical significance at
probability level of 5% were examined. Additionally, the Cronbach’s Alpha and the Rho
Coefficient of internal consistency were examined for score reliability. Construct validity of
the measurement model was determined from model convergence and the magnitude of
parameter coefficients. Results on these statistics are presented in the following section for the
BQF construct.
Table 10.25: Postulated Building Quality Features Model
Latent constructs Indicator variables
(How satisfied or dissatisfied are you with…
Label
Building Quality Internal construction quality BQF2
Feature (BQF) Water pressure BQF3
Wall quality BQF4
Floor quality BQF5
Plumbing quality BQF10
The finished quality of sanitary system BQF11
Diagnostic Fit Analysis: Analysis of Residual Covariance Estimate
The average absolute residual values of the building quality feature construct are presented in
Table 10.26 and 10.27. Results revealed that all the absolute residual values and the average
off-diagonal absolute residual values were close to zero. The unstandardized average off-
diagonal residual was 0.0451, while the standardized average off-diagonal residual was found
to be 0.0341.
E11
BQF2
BQF3
BQF4
BQF5
BQF10
BQF11
BQF
E2
E3
E4
E5
E10
422
Table 10.26: Residual Covariance Matrix for Building Quality Feature Model
(Unstandardized)
Unstandardized Residual Covariance Matrix
BQF2 BQF3 BQF4 BQF5 BQF10 BQF11
BQF2 0.000
BQF3 -0.085 0.000
BQF4 0.037 -0.014 0.000
BQF5 -0.007 -0.008 0.017 0.000
BQF10 -0.017 0.064 -0.032 -0.013 0.000
BQF11 -0.033 0.096 -0.060 -0.026 0.167 0.000
Average absolute residual = 0.0322
Average off-diagonal absolute residual = 0.0451
% falling between -0.1 +0.1 = 99.99%
These residual values were considered small as they were all less than 2.58 (Byrne, 2006:94).
In addition, 99.99% of the unstandardized and standardised residual fell within the acceptable
range of -0.1 and +0.1. The significance of this distribution is that for a model to be described
as well-fitting, the distribution of standardised residuals should be symmetrical and centred
around zero (Byrne, 2006:94).
Table 10.27: Residual Covariance Matrix for Building Quality Feature Model
(Standardized)
Standardized Residual Covariance Matrix
BQF2 BQF3 BQF4 BQF5 BQF10 BQF11
BQF2 0.000
BQF3 -0.067 0.000
BQF4 0.027 -0.011 0.000
BQF5 -0.005 -0.006 0.013 0.000
BQF10 -0.012 0.050 -0.023 -0.009 0.000
BQF11 -0.024 0.077 -0.044 -0.020 0.122 0.000
Average absolute residual = 0.0243
Average off-diagonal absolute residual = 0.0341
% falling between -0.1 +0.1 = 99.99%
From the above information, the results seemed to suggest that the model had a good-fit to the
sample data. Therefore, since this initial assessment of residuals indicated a good fit; a further
test of goodness-of-fit was justified.
423
Goodness-of-Fit Statistics – RML
The sample data on BQF measurement model yield an S – Bχ2 of 76.438 with 9 degrees of
freedom. The associated p-value was determined to be 0.0000 for the analysed sample of 751
cases. The chi-square value suggested that the difference between the sample data and the
postulated building quality features measurement model was insignificant. Additionally, the
ratio of S – Bχ2 to the degrees of freedom was determined to be 8.49, which was higher than
the upper limit value of 5.0 (Kline, 2005:137).
Similarly, other fit indexes indicated a good fit of the model to the sample (Table 10.28). The
robust CFI index of 0.964 was greater than the cut-off value for a good fitting model. A model
is said to be a good fit if the CFI is above the cut-off value of 0.95 (Hu & Bentler, 1999:27).
The robust RMSEA with 90% confidence interval (lower bound value = 0.080 and the upper
bound value = 0.121) was found to be 0.100. This value was a bit above the maximum value
of 0.08 for a good fit model; this is considered an acceptable average model fit (MacCallum et
al., 1996). In addition to the RMSEA value, the absolute fit index SRMR was found to be
0.039. This value indicated a very good fit because a good fitting model is expected to have an
SRMR index lower or equal to 0.05 while an index of 0.08 is sufficient to accept the postulated
model.
Table 10.28: Robust Fit Indexes for Building Quality Features Construct
Fit Index Cut-off value Estimate Comment
S – Bχ2 76.438
df 0≥ 9 Acceptable
CFI 0.90≥ acceptable
0.95≥ good fit
0.964 Good fit
SRMR 0.08≥ acceptable
0.05≥ good fit
0.039 Good fit
RMSEA 0.08≥ acceptable
0.05≥ good fit
0.100 Acceptable fit
RMSEA
90% CI
0.080:0.121 Slightly out of
range
The absolute fit index SRMR accounts for the average discrepancy between the sample and the
postulated correlation matrices and therefore, it represents the average value across all
standardised residuals and ranges between zero and 1.00 (Byrne, 2006:94). Evaluation of the
424
SRMR, RMSEA (90% CI) and the CFI fit indexes indicated a good fit of the measurement
model for the building quality features factor because those indexes met the condition for a
good fit (Table 10.28). Additionally, parameter estimates were analysed to determine whether
the model worked properly and was feasible. This involved evaluating the magnitude, signs
and statistical significance of the parameter estimates. These statistics are presented in Table
10.29.
Statistical Significance of Parameter Estimates
Apart from assessing the goodness-of-fit or the lack of it, feasibility of a model can be judged
by a further inspection of the obtained solution and this involves inspection of parameter
estimates, standard errors and the test statistics (Raykov, 1991:501). Estimates are said to be
unreasonable if in the standardised output there are estimates that have correlation values that
are greater than 1.00, have negative variances and the correlation or covariances are not definite
positive (Byrne, 2006:103). Also, the test statistics needs to be greater than 1.96 based on the
probability level of 5% before the hypothesis can be rejected (Byrne, 2006:103). The test
statistics reported in this study was the parameter estimate divided by its standard error and
therefore it functions as a Z-statistics to test that the estimate is statistically different from zero.
Table 10.29: Factor loading and Z-statistics of Building Quality Features Measurement
Model
Indicator
variable
Unstandardized
Coefficient (λ)
Standardized
Coefficient (λ)
Z- Statistics
R2
Significant
at 5%
level?
BQF2 0.903 0.769 24.599 0.592 Yes
BQF3 0.509 0.468 9.648 0.219 Yes
BQF4 1.049 0.893 37.724 0.797 Yes
BQF5 1.027 0.897 34.941 0.804 Yes
BQF10 0.957 0.809 26.869 0.654 Yes
BQF11 0.749 0.649 19.612 0.421 Yes
(Robust statistical significance at 5% level)
An assessment of the correlation values, standard errors and the test statistic in Table 10.29
revealed that majority of the standardized coefficient correlation values were not greater than
1.00; while 2 indicator variables BQF4 (unstandardized λ= 1.049) and BQF5 (unstandardized
λ= 1.027) coefficient value were greater than 1.00. Test statistics (Z-values) were greater than
425
1.96 (p<0.05) and the signs were appropriate (positive). The estimates were therefore found to
be reasonable, as well as statistically significant. The parameter with the highest standardized
coefficient was the indicator variable BQF5. The parameter coefficient was found to be 0.897.
The variable BQF5, which asked the respondents’ level of satisfaction, with the building floor
quality, was found to be more associated with the construct of building quality feature than all
other variables. Nevertheless, all standardized parameter estimates showed high correlations
values with about 83% above 0.50 and only 17% below 0.50, suggesting a high degree of linear
association between the indicator variables and the building quality factor. In addition, R2
values were found to be close to the desired value of 1.00. The only exceptions were the
indicator variables BQF3 and BQF11. The R2 values for these variables were below 0.50. The
results therefore suggest that the indicator variables predict the factor construct considerably.
Internal Reliability and Validity of Scores
In order to determine the internal consistency of the composition of the building quality feature
measurement model, the Rho Coefficient and the Cronbach’s Alpha coefficient were examined
to establish reliability (Byrne, 2006:133). According to Kline (2005:59), the reliability
coefficient should be between zero and 1.00. Values close to 1.00 are desired. The Rho
Coefficient of internal consistency was found to be 0.892. This was above the minimum value
of 0.70. Similarly, the Cronbach’s Alpha was also found to be above the minimum value of
0.70 at 0.885. Both values showed a high level of internal consistency, and therefore, reliability.
The construct validity was determined by examining the magnitude of the parameter
coefficients (factor loading). High parameter coefficients of greater than 0.5 indicate a close
relation between the factor and an indicator variable. A parameter coefficient of 0.5 is
interpreted as 25% of the total variance in the indicator variable being explained by the latent
variable (factor). Hence, a parameter coefficient has to be between 0.5-0.7 or greater to explain
about 50% of the variance in an indicator variable (Hair et al., 1998:111). Inspection of the
unstandardized parameter coefficient presented in Table 10.30, shows that they were
significantly high with the minimum of 0.509, which suggested that the factor accounted for
about 50.45% of the variance in BQF3.
426
Table 10.30: Reliability and Construct Validity of Building Quality Feature Model
Factor Indicator
Variable
Factor
Loading
Cronbach’s
Alpha
Rho
Coefficient
Building
quality
BQF2 0.769
Feature (BQF) BQF3 0.468 0.885 0.892
BQF4 0.893
BQF5 0.897
BQF10 0.809
BQF11 0.649
*Parameter estimates are based on standardized solutions
This value was however above the acceptable level. On the other hand, all other parameter
estimates were above 50% and therefore indicative of an adequate fit between the indicator
variables and the factor. Therefore the factor, building quality features satisfied both internal
reliability and the construct validity criteria. The Rho value was above the minimum value of
0.70 (Table 10.30) and the construct validity criteria was justified by the magnitude, signs and
statistical significance of all parameter coefficients.
Summary on Building Quality Feature Measurement Model
The CFA output results revealed that the residual covariance estimates fell within the
acceptable range, the robust fit indexes met the cut-off index criteria, except for the RMSEA
value and the RMSEA with 90% confidence interval which produced an average fit. All other
parameter estimates were statistically significant and feasible. It was therefore concluded
therefore, that the measurement model for the building quality feature, had an adequate fit to
the sample data. Consequently, there was no need to improve the measurement before it could
be included in the full latent variable model.
10.3.3.5 Measurement Model for Neighbourhood Features (NDF) Construct
The number of cases that were analysed for the neighbourhood feature construct was 749 cases
from a sample of 751. Two cases were skipped because of missing variables. Preliminary
observation of the data revealed that the residual covariance matrix scores for seventeen
indicator variables (NDF2, NDF4, NDF6, NDF8-NDF9, & NDF11-NDF22) had unacceptably
high scores (values ranged from 2.95-4.71); hence they were dropped from further CFA
analysis. Therefore only five indicator variables passed the test and were used for the
assessment of the measurement model goodness-of-fit. Examination of the Bentler-Weeks
427
structure representation for the approved construct revealed that the NDF construct has 5
dependent variables, 6 independent variables and 10 free parameters. The number of fixed non-
zero parameter was 6. The 6 dependent indicator variables for the NDF are presented in Table
10.31 and Figure 10.6. These was analysed before it could be included in the full latent variable
model. In order to establish how well the mode fit the sample data and the strength of the
hypothesised relationship between the variables, results on residual covariance matrix
(unstandardized and standardized), distribution of standardised residuals, fit statistics and
statistical significance at probability level of 5% were examined.
Figure 10.6: Measurement Model of Neighbourhood Features
In addition, the Cronbach’s Alpha and the Rho Coefficient of internal consistency were
examined for score reliability. Construct validity of the measurement model was determined
from model convergence and the magnitude of parameter coefficients. Results on these
statistics are presented in the following section for the NDF construct.
Table 10.31: Postulated Neighbourhood Features Model
Latent constructs Indicator variables
(How satisfied or dissatisfied are you with…
Label
Neighbourhood
Features (NDF)
Location of the dwelling unit in the
neighbourhood
NDF1
Quality of landscaping in the neighbourhood NDF3
Ease of access to main road NDF5
Quality of street lighting at night NDF7
Cleanliness of the neighbourhood NDF10
E10
NDF1
NDF3
NDF5
NDF7
NDF10
NDF
E1
E3
E5
E7
428
Diagnostic Fit Analysis: Analysis of Residual Covariance Estimate
Average absolute residual values of the building quality feature construct are presented in Table
10.32 and 10.33. Results revealed that all the absolute residual values and the average off-
diagonal absolute residual values were close to zero. The unstandardized average off-diagonal
residual was 0.0526 while the standardized average off-diagonal residual was found to be
0.0493. These residual values were considered small as they were all less than 2.58 (Byrne,
2006:94). In addition, 99.99% of the unstandardized and standardised residual fell within the
acceptable range of -0.1 and +0.1. The significance of this distribution is that for a model to be
described as well-fitting, the distribution of standardised residuals should be symmetrical and
centred around zero (Byrne, 2006:94).
Table 10.32: Residual Covariance Matrix for Neighbourhood Feature Model
(Unstandardized)
Unstandardized Residual Covariance Matrix
NDF1 NDF3 NDF5 NDF7 NDF10
NDF1 0.000
NDF3 0.075 0.000
NDF5 -0.023 -0.011 0.000
NDF7 -0.045 -0.041 0.043 0.000
NDF10 0.009 0.120 -0.015 0.144 0.000
Average absolute residual = 0.0350
Average off-diagonal absolute residual = 0.0526
% falling between -0.1 +0.1 = 99.99%
Table 10.33: Residual Covariance Matrix for Neighbourhood Feature Model
(Standardized)
Standardized Residual Covariance Matrix
NDF1 NDF3 NDF5 NDF7 NDF10
NDF1 0.000
NDF3 0.085 0.000
NDF5 -0.026 -0.012 0.000
NDF7 -0.041 -0.037 0.040 0.000
NDF10 -0.009 0.116 -0.014 0.112 0.000
Average absolute residual = 0.0329
Average off-diagonal absolute residual = 0.0493
% falling between -0.1 +0.1 = 99.99%
429
From the residual covariance matrix information, the results suggest that the model had a good-
fit to the sample data. Therefore, since this initial assessment of residuals indicated a good fit;
a further test of goodness-of-fit was justified.
Goodness-of-Fit Statistics – RML
The sample data on NDF measurement model yield an S – Bχ2 of 53.024 with 5 degrees of
freedom. The associated p-value was determined to be 0.0000 for the analysed sample of 749
cases. The chi-square value advocated that the difference between the sample data and the
postulated neighbourhood features measurement model was insignificant. Additionally, the
ratio of S – Bχ2 to the degrees of freedom was determined to be 10.61, which was higher than
the upper limit value of 5.0 (Kline, 2005:137).
Table 10.34: Robust Fit Indexes for Neighbourhood Feature Construct
Fit Index Cut-off value Estimate Comment
S – Bχ2 53.024
df 0≥ 5 Acceptable
CFI 0.90≥ acceptable
0.95≥ good fit
0.931 Good fit
GFI 0.90≥ acceptable
0.95≥ good fit
0.958 Good fit
SRMR 0.08≥ acceptable
0.05≥ good fit
0.050 Good fit
RMSEA 0.08≥ acceptable
0.05≥ good fit
0.113 Acceptable fit
RMSEA
90% CI
0.087:0.141 Slightly out of
range
Correspondingly, other fit indexes indicated a good fit of the model to the sample. The robust
Goodness-of-fit (GFI) index of 0.958 was greater than the cut-off value for a good fitting
model. Whilst the CFI index of 0.931 was slightly lower than the cut-off value for a good fitting
model. A model is said to be a good fit if the CFI and GFI are above the cut-off value of 0.95
(Hu & Bentler, 1999:27; Joreskog & Sorbom, 1996). With the drop (difference of 0.019) in the
CFI value, the model can be described to have an acceptable fit. The robust RMSEA with 90%
confidence interval (lower bound value = 0.087 and the upper bound value = 0.141) was found
to be 0.113. This value was above the maximum value of 0.08 for a good fit model. However,
this is considered an acceptable mediocre model fit (MacCallum et al., 1996).
430
In addition to the RMSEA value, the absolute fit index SRMR was found to be 0.05. This value
indicated a very good fit because a good fitting model is expected to have an SRMR index
lower or equal to 0.05, while an index of 0.08 is sufficient to accept the postulated model. The
absolute fit index SRMR accounts for the average discrepancy between the sample and the
postulated correlation matrices and therefore, it represents the average value across all
standardised residuals and ranges between zero and 1.00 (Byrne, 2006:94). Evaluation of the
SRMR, RMSEA (90% CI), GFI and the CFI fit indexes indicated an acceptable fit of the
measurement model for the neighbourhood features factor (Table 10.34). Furthermore,
parameter estimates were analysed to determine whether the model worked properly and was
feasible. This involved evaluating the magnitude, signs and statistical significance of the
parameter estimates. These statistics are presented in Table 10.35.
Statistical Significance of Parameter Estimates
Raykov (1991:501) recommends that further examination of factor loading (parameter
coefficients), standard errors and the test statistics should be conducted in addition to the
analysis of fit statistics before conclusion could be drawn about the appropriateness of the
postulated models. Therefore, these estimates were examined and are presented in this section.
Table 10.35: Factor loading and Z-statistics of Neighbourhood Features Measurement
Model
Indicator
Variable
Unstandardized
Coefficient (λ)
Standardized
Coefficient (λ)
Z- Statistics
R2
Significant
at 5%
level?
NDF1 0.674 0.719 15.910 0.517 Yes
NDF3 0.624 0.662 14.744 0.439 Yes
NDF5 0.624 0.684 18.429 0.467 Yes
NDF7 0.842 0.722 21.429 0.521 Yes
NDF10 0.444 0.401 12.822 0.161 Yes
(Robust statistical significance at 5% level)
Byrne (2006:103) informs that estimates are said to be unreasonable if they have correlations
values greater than 1.00, have negative variances and the correlation or covariances are not
definite positive. Furthermore, the test statistics need to be greater than 1.96 based on the
probability level of 5% before the hypothesis can be rejected. The test statistics in this study
was the parameter estimate divided by its standard error and therefore, it functions as a Z-
431
statistics to test that the estimate is statistically different from zero. The coefficient was
therefore referred to as the Z-statistics. Inspection of the correlation values, standard errors and
the test statistic in Table 10.35 reveal that all standardized coefficient correlation values were
not greater than 1.00; all test statistics (Z-values) were greater than 1.96 (p<0.05) and the signs
were appropriate (positive). The estimates were therefore reasonable, as well as statistically
significant. The parameter with the highest standardized coefficient was the indicator variable
NDF7. The parameter coefficient was found to be 0.722.
The variable NDF7, which asked the respondents’ about their level of satisfaction with the
quality of street lighting at night, was found to associate more with the construct of
neighbourhood feature than all other variables. Nonetheless, all standardized parameter
estimates showed high correlations values close to 1.00 suggesting a high degree of linear
association between the indicator variables and the latent construct. In addition, R2 values were
found to be close to the desired value of 1.00. The only exceptions were the indicator variables
NDF3, NDF4 and NDF10. The R2 values for this variable were below 0.50. Despite these
values, the overall results suggest that the indicator variables considerably predict the factor
construct.
Internal Reliability and Validity of Scores
In order to determine the internal consistency of the composition of the measurement model,
the Rho Coefficient and the Cronbach’s Alpha Coefficient were examined to establish
reliability. According to Kline (2005:59), the reliability coefficient should fall between zero
and 1.00. Values close to 1.00 are desired. The Rho Coefficient of internal consistency was
found to be 0.772. This was above the minimum value of 0.70. Equally, the Cronbach’s Alpha
was also found to be above the minimum value of 0.70 at 0.764. Both values showed a high
level of internal consistency and therefore reliability.
The construct validity was determined by examining the magnitude of the parameter
coefficients. High parameter coefficients of greater than 0.5, indicate a close relation between
the factor and an indicator variable. A parameter coefficient of 0.5 is interpreted as 25% of the
total variance in the indicator variable being explained by the latent variable (factor). Therefore,
a parameter coefficient has to be between 0.5 - 0.7or greater to explain about 50% of the
variance in an indicator variable (Hair et al., 1998:111). Inspection of the unstandardized
parameter coefficient presented in Table 10.35, shown that they were significantly high with
432
the minimum at 0.444. The parameter estimate of 0.444 meant that the neighbourhood factor
accounted for about 47% of the variance in NDF10 and therefore indicative of a good fit
between the indicator variable and the factor.
Therefore, the neighbourhood factor satisfied both internal reliability and the construct validity
criteria. The Rho value was above the minimum value of 0.70 (Table 10.36) and the construct
validity criteria was justified by the magnitude, signs and statistical significance of all
parameter coefficients.
Table 10.36: Reliability and Construct Validity of Neighbourhood Feature Model
Factor Indicator
Variable
Factor
Loading
Cronbach’s
Alpha
Rho
Coefficient
Neighbourhood NDF1 0.719
Feature (NDF) NDF3 0.662 0.764 0.772
NDF5 0.684
NDF7 0.722
NDF10 0.401
*Parameter estimates are based on standardized solutions
Summary on Neighbourhood Feature Measurement Model
The residual covariance estimates fell within the acceptable range; the robust fit indexes had
an acceptable fit, while the RMSEA value and the RMSEA with 90% confidence interval
produced an average fit. All other parameter estimates were statistically significant and
feasible. It was therefore concluded that the measurement model for the neighbourhood feature,
had an adequate fit to the sample data. Consequently, there was no need to improve the
measurement before it could be included in the full latent variable model.
10.3.3.6 Measurement Model for Service Provided by Government (SPG)
Construct
The number of cases that were analysed for the service provided by government latent construct
were 751 cases. No case was skipped because there were no missing variables. From the
initially hypothesized thirteen indicators, preliminary investigation of the CFA output revealed
that the residual covariance scores for nine indicator variables (SPG1 - SPG7, & SPG10 -
SPG11) had covariance matrices scores; hence they were dropped from further CFA analysis.
433
Figure 10.7: Measurement model of services provided by government
Examination of the Bentler-Weeks structure representation for the indicator variables that
passed the first statistical exploration analysis revealed that the SPG construct has 4 dependent
variables, 5 independent variables and 8 free parameters. The number of fixed non-zero
parameter was 5. The 4 dependent indicator variables for the SPG construct are presented in
Table 10.37 and Figure 10.7. These was analysed before inclusion in the full latent variable
model. In order to establish how well the mode fit the sample data and the strength of the
hypothesised relationship between the variables, results on residual covariance matrix
(unstandardized and standardized), distribution of standardised residuals, fit statistics and
statistical significance at probability level of 5% were examined.
Table 10.37: Postulated Services Provided by Government Model
Latent constructs Indicator variables
(How satisfied or dissatisfied are you with…
Label
Services Provided
by Government
How well residents’ complaints are handled SPG8
(SPG) Government response to building defects SPG9
Enforcement of rules by the Department of
Human Settlement (Housing)
SPG12
Overall services provided by the government SPG13
Likewise, the Cronbach’s Alpha and the Rho Coefficient of internal consistency were
examined for score reliability. Construct validity of the measurement model was determined
from model convergence and the magnitude of parameter coefficients. Results on these
statistics are presented in the following section for the SPG construct.
Diagnostic Fit Analysis: Analysis of Residual Covariance Estimate
Average absolute residual values of the services provided by government latent construct are
presented in Table 10.38 and 10.39. Results revealed that all the absolute residual values and
E13
SPG8
SPG9
SPG12
SPG13
SPG
E8
E9
E12
434
the average off-diagonal absolute residual values were close to zero. The unstandardized
average off-diagonal residual was 0.0262, while the standardized average off-diagonal residual
was found to be 0.0245.
Table 10.38: Residual Covariance Matrix for Services Provided by Government Model
(Unstandardized)
Unstandardized Residual Covariance Matrix
SGP8 SPG9 SPG12 SPG13
SPG8 0.000
SPG9 0.034 0.000
SPG12 -0.052 0.003 0.000
SPG13 0.006 -0.028 0.034 0.000
Average absolute residual = 0.0157
Average off-diagonal absolute residual = 0.0262
% falling between -0.1 +0.1 = 100%
Table 10.391: Residual covariance matrix for services provided by government model
(Standardized) Standardized residual covariance matrix
SGP8 SPG9 SPG12 SPG13
SPG8 0.000
SPG9 0.034 0.000
SPG12 -0.052 0.003 0.000
SPG13 0.006 -0.028 0.034 0.000
Average absolute residual = 0.0147
Average off-diagonal absolute residual = 0.0245
% falling between -0.1 +0.1 = 100%
These residual values were considered small as they were all less than 2.58 (Byrne, 2006:94).
In addition, 100% of the unstandardized and standardised residual fell within the acceptable
range of -0.1 and +0.1. The significance of this distribution is that for a model to be described
as well-fitting, the distribution of standardised residuals should be symmetrical and centred
around zero (Byrne, 2006:94). From the above information, the results suggest that the model
had a good-fit to the sample data. Therefore, since this initial assessment of residuals indicated
a good fit; a further test of goodness-of-fit was justified.
435
Goodness-of-Fit Statistics – RML
The sample data for SPG measurement model yield an S – Bχ2 of 21.019 with 2 degrees of
freedom (N = 751; p = 0.0003). The chi-square value revealed that the difference between the
sample data and the postulated neighbourhood features measurement model was insignificant.
Additionally, the ratio of S – Bχ2 to the degrees of freedom was determined to be 10.51 which
was higher than the upper limit value of 5.0 (Kline, 2005:137).
However, other fit indexes indicated a good fit of the model to the sample. The robust
Goodness-of-fit (GFI) index of 0.970 was greater than the cut-off value for a good fitting
model. Whilst the CFI index of 0.990 was also higher that the cut-off value for a good fitting
model. Hence, the model can be described to have an acceptable fit. The robust RMSEA with
90% confidence interval (lower bound value = 0.073 and the upper bound value = 0.159) was
found to be 0.113. This value was above the maximum value of 0.08 for a good fit model.
However, this is considered an acceptable mediocre fit model (MacCallum et al., 1996).
Table 10.40: Robust Fit Indexes for Services Provided by Government Construct
Fit Index Cut-off value Estimate Comment
S – Bχ2 21.019
df 0≥ 2 Acceptable
CFI 0.90≥ acceptable
0.95≥ good fit
0.990 Good fit
GFI 0.90≥ acceptable
0.95≥ good fit
0.97 Good fit
SRMR 0.08≥ acceptable
0.05≥ good fit
0.023 Good fit
RMSEA 0.08≥ acceptable
0.05≥ good fit
0.113 Acceptable fit
RMSEA
90% CI
0.087:0.141 Slightly out of
range
In addition to the RMSEA value, the absolute fit index SRMR was found to be 0.023. This
value indicated a very good fit because a good fitting model is expected to have an SRMR
index lower or equal to 0.05, while an index of 0.08 is sufficient to accept the postulated model.
The absolute fit index SRMR accounts for the average discrepancy between the sample and the
postulated correlation matrices and therefore, it represents the average value across all
standardised residuals and ranges between zero and 1.00 (Byrne, 2006:94). Evaluation of the
SRMR, RMSEA (90% CI), GFI and the CFI fit indexes indicated an acceptable fit of the
436
measurement model for the neighbourhood features factor (Table 10.40). Furthermore,
parameter estimates were analysed to determine whether the model worked properly and was
feasible. This involved evaluating the magnitude, signs and statistical significance of the
parameter estimates. These statistics are presented in Table 10.41.
Statistical Significance of Parameter Estimates
Raykov (1991:501) recommended that further examination of factor loading (parameter
coefficients), standard errors and the test statistics should be conducted in addition to the
analysis of fit statistics before conclusion could be made about the appropriateness of the
postulated models. Therefore, these estimates were examined and are presented in this section.
Table 10.41: Factor loading and Z-statistics of Services Provided by Government
Measurement Model
Indicator
Variable
Unstandardized
Coefficient (λ)
Standardized
Coefficient (λ)
Z- Statistics
R2
Significant
at 5%
level?
SPG8 0.812 0.792 27.482 0.627 Yes
SPG9 0.940 0.894 40.826 0.799 Yes
SPG12 0.891 0.858 33.433 0.736 Yes
SPG13 0.861 0.839 32.536 0.705 Yes
(Robust statistical significance at 5% level)
Also, Byrne (2006:103) informs that estimates are said to be unreasonable if they have
correlations values greater than 1.00, have negative variances and the correlation or
covariances are not definite positive. Furthermore, the test statistics need to be greater than
1.96 based on the probability level of 5% before the hypothesis can be rejected. Inspection of
the correlation values, standard errors and the test statistic in Table 10.41 reveal that all
standardized coefficient correlation values were not greater than 1.00; all test statistics (Z-
values) were greater than 1.96 (p<0.05) and the signs were appropriate (positive). The estimates
were therefore reasonable as well as statistically significant. The parameter with the highest
standardized coefficient was the indicator variable SPG9. The parameter coefficient was found
to be 0.894.
The variable SPG9, which asked the respondents about their level of satisfaction with
government response to building defects, was found to be more associated with the construct
437
of services provided by government than other variables. Nonetheless, all standardized
parameter estimates showed high correlations values close to 1.00 suggesting a high degree of
linear association between the indicator variables and the latent construct. In addition, R2 values
were found to be close to the desired value of 1.00 indicating that the factors explained more
of the variance in the indicator variables.
Internal Reliability and Validity of Scores
The internal consistency and reliability of scores for the measurement model were determined
from the Rho Coefficient and the Cronbach’s Alpha coefficient. According to Kline (2005:59),
the reliability coefficient should fall between zero and 1.00. Values close to 1.00 are desired.
The Rho Coefficient of internal consistency was found to be 0.910. This was above the
minimum value of 0.70. Equally, the Cronbach’s Alpha was also found to be above the
minimum value of 0.70 at 0.909. Both values showed a high level of internal consistency and
therefore reliability.
Table 10.42: Reliability and Construct Validity of Services Provided by Government
Model
Factor Indicator
Variable
Factor
Loading
Cronbach’s
Alpha
Rho
Coefficient
Services SPG8 0.792
Provided by SPG9 0.894 0.909 0.910
Government SPG12 0.858
(SPG) SPG13 0.839
*Parameter estimates are based on standardized solutions
The construct validity was determined by examining the magnitude of the parameter
coefficients. High parameter coefficients of greater than 0.5 indicate a close relation between
the factor and an indicator variable. A parameter coefficient of 0.5 is interpreted as 25% of the
total variance in the indicator variable being explained by the latent variable (factor). Therefore,
a parameter coefficient has to be between 0.5 - 0.7or greater to explain about 50% of the
variance in an indicator variable (Hair et al., 1998:111). Inspection of the unstandardized
parameter coefficient presented in Table 10.42, revealed that they were significantly high with
the minimum of 0.812. The parameter estimate of 0.812 meant that the services provided by
government factor accounted for about 61.89% of the variance in SPG8 and therefore
indicative of a good fit between the indicator variable and the factor.
438
Therefore, the services provided by government factor satisfied both internal reliability and the
construct validity criteria. The Rho value was above the minimum value of 0.70 (Table 10.42)
and the construct validity criteria was justified by the magnitude, signs and statistical
significance of all parameter coefficients.
Summary on Services Provided by Government Measurement Model
The residual covariance estimates fell within the acceptable range; the robust fit indexes had a
good fit, while the RMSEA value and the RMSEA with 90% confidence interval produced an
average fit. All other parameter estimates were statistically significant and feasible. It was
therefore concluded that the measurement model for the services provided by government, had
a good fit to the sample data. Consequently, there was no need to improve the measurement
model; hence, it was proper to be included in the full latent variable model.
10.3.3.7 Measurement Model for Residential Satisfaction (RS) Outcome Variables
The number of cases that were analysed for the residential satisfaction manifest construct was
751 cases. Two cases were skipped because of missing variables. From the initially
hypothesized seven indicators, primary observation of the CFA data revealed that the residual
covariance matrix scores for two indicator variables (RS2 & RS6) were more than the cut-off
of 2.58 (Byrne, 2006:94; Joreskog & Sorbom, 1988). Hence they were dropped from the CFA
analysis. Inspection of the Bentler-Weeks structure representation revealed that the remaining
RS manifest construct has 4 dependent variables, 5 independent variables and 8 free
parameters.
Figure 10.8: Measurement Model of Residential Satisfaction Manifest Construct
The number of fixed non-zero parameter was 5. The 4 dependent indicator variables for the RS
manifest variables are presented in Table 10.43 and Figure 10.8. These were further analysed
E7
RS1
RS3
RS5
RS7
RS
E1
E3
E5
439
before it could be included in the full latent variable model. In order to determine how well the
mode fit the sample data and the strength of the hypothesised relationship between the
variables, results on residual covariance matrix (unstandardized and standardized), distribution
of standardised residuals, fit statistics and statistical significance at probability level of 5%
were examined. Furthermore, the Cronbach’s Alpha and the Rho Coefficient of internal
consistency were examined for score reliability. Construct validity of the measurement model
was determined from model convergence and the magnitude of parameter coefficients. Results
on these statistics are presented in the following section for the RS construct.
Table 10.43: Postulated Residential Satisfaction Manifest Model
Latent
Constructs
Indicator Variables
Label
Residential I am satisfied living here RS1
Satisfaction I am taking proper care of my neighbourhood RS3
(RS) I am not intending to move to another place in the
future
RS5
I will recommend to my friends to obtain a house
in the same way that I did
RS7
Diagnostic Fit Analysis: Analysis of Residual Covariance Estimate
Average absolute residual values of the building quality feature construct are presented in Table
10.44 and 10.45.
Table 10.44: Residual Covariance Matrix for Residential Satisfaction Model
(Unstandardized)
Unstandardized Residual Covariance Matrix
RS1 RS3 RS5 RS7
RS1 0.000
RS3 0.024 0.000
RS5 -0.070 0.028 0.000
RS7 0.014 -0.052 0.067 0.000
Average absolute residual = 0.0254
Average off-diagonal absolute residual = 0.0424
% falling between -0.1 +0.1 = 100%
Results revealed that all the absolute residual values and the average off-diagonal absolute
residual values were close to zero. The unstandardized average off-diagonal residual was
440
0.0424 while the standardized average off-diagonal residual was found to be 0.0377. These
residual values were considered small as they were all less than 2.58 (Byrne, 2006:94). In
addition, 100% of the unstandardized and standardised residual fell within the acceptable range
of -0.1 and +0.1. The significance of this distribution is that for a model to be described as
well-fitting, the distribution of standardised residuals should be symmetrical and centred
around zero (Byrne, 2006:94).
Table 10.45: Residual Covariance Matrix for Residential Satisfaction Model
(Standardized)
Unstandardized Residual Covariance Matrix
RS1 RS3 RS5 RS7
RS1 0.000
RS3 0.027 0.000
RS5 -0.053 0.033 0.000
RS7 0.010 -0.056 0.047 0.000
Average absolute residual = 0.0226
Average off-diagonal absolute residual = 0.0377
% falling between -0.1 +0.1 = 100%
From the above information, the results suggest that the model had a good-fit to the sample
data. Therefore, since this initial assessment of residuals indicated a good fit; a further test of
goodness-of-fit was justified.
Goodness-of-Fit Statistics – RML
The sample data on RS measurement model yield an S – Bχ2 of 16.540 with 2 degrees of
freedom. The associated p - value was determined to be 0.00026 for the analysed sample of
751 cases. The chi-square value indicated that the difference between the sample data and the
postulated residential satisfaction manifest measurement model was insignificant.
Additionally, the ratio of S – Bχ2 to the degrees of freedom was determined to be 8.27 which
was higher than the upper limit value of 5.0 (Kline, 2005:137).
However, other fit indexes indicated a good fit of the model to the sample. The robust
Goodness-of-fit (GFI) index of 0.988 was greater than the cut-off value for a good fitting
model. Whilst the CFI index of 0.963 was also higher than the cut-off value for a good fitting
model. A model is said to be a good fit if the CFI and GFI are above the cut-off value of 0.95
441
(Hu & Bentler, 1999:27; Joreskog & Sorbom, 1996). The robust RMSEA with 90% confidence
interval (lower bound value = 0.059 and the upper bound value = 0.145) was found to be 0.099.
This value was above the maximum value of 0.08 for a good fit model; however, this is
considered an acceptable fit (MacCallum et al., 1996).
Table 10.46: Robust fit indexes for residential satisfaction construct
Fit Index Cut-off value Estimate Comment
S – Bχ2 16.540
df 0≥ 2 Acceptable
CFI 0.90≥ acceptable
0.95≥ good fit
0.963 Good fit
GFI 0.90≥ acceptable
0.95≥ good fit
0.988 Good fit
SRMR 0.08≥ acceptable
0.05≥ good fit
0.032 Good fit
RMSEA 0.08≥ acceptable
0.05≥ good fit
0.099 Acceptable fit
RMSEA
90% CI
0.059:0.145 Slightly out of
range
In addition to the RMSEA value, the absolute fit index SRMR was found to be 0.032. This
value indicated a very good fit because a good fitting model is expected to have an SRMR
index lower or equal to 0.05. The absolute fit index SRMR accounts for the average
discrepancy between the sample and the postulated correlation matrices and therefore it
represents the average value across all standardised residuals and ranges between zero and 1.00
(Byrne, 2006:94). Evaluation of the SRMR, RMSEA (90% CI), GFI and the CFI fit indexes
indicated an acceptable fit of the measurement model for the residential satisfaction factor
(Table 10.46).
Furthermore, parameter estimates were analysed to determine whether the model worked
properly and was feasible. This involved evaluating the magnitude, signs and statistical
significance of the parameter estimates. These statistics are presented in Table 10.47.
Statistical Significance of Parameter Estimates
Raykov (1991:501) recommended that further examination of factor loading (parameter
coefficients), standard errors and the test statistics should be conducted in addition to the
442
analysis of fit statistics before conclusion could be made about the appropriateness of the
postulated models. Therefore, these estimates were examined and are presented in this section.
Table 10.47: Factor loading and Z-statistics of Residential Satisfaction Measurement
Model
Indicator
Variable
Unstandardized
Coefficient (λ)
Standardized
Coefficient (λ)
Z- Statistics
R2
Significant
at 5%
level?
RS1 0.831 0.718 15.213 0.516 Yes
RS3 0.411 0.551 12.937 0.304 Yes
RS5 0.546 0.479 10.898 0.229 Yes
RS7 0.800 0.644 15.134 0.415 Yes
(Robust statistical significance at 5% level)
Parameter estimates are said to be unreasonable if they have correlations values greater than
1.00, have negative variances and the correlation or covariances are not definite positive Byrne
(2006:103). Furthermore, the test statistics need to be greater than 1.96 based on the probability
level of 5% before the hypothesis can be rejected. Inspection of the correlation values, standard
errors and the test statistic in Table 10.47 reveal that all standardized coefficient correlation
values were not greater than 1.00; all test statistics (Z-values) were greater than 1.96 (p<0.05)
and the signs were appropriate (positive). The estimates were therefore reasonable, as well as
statistically significant. The parameter with the highest standardized coefficient was the
indicator variable RS1. The parameter coefficient was found to be 0.718.
The variable RS1, which asked the respondents’ if they are satisfied living in their respective
low income houses, was found to be associated more with the construct of residential
satisfaction than all other variables. Nonetheless, all standardized parameter estimates showed
high correlation values close to 1.00 suggesting a high degree of linear association between the
indicator variables and the latent construct. In addition, R2 values were found to be close to the
desired value of 1.00. The only exceptions were the manifest variables RS3, RS5 and RS7. The
R2 values for this variable were below 0.50. However, the results therefore suggest that the
indicator variables considerably predict the factor construct.
443
Internal Reliability and Validity of Scores
In order to determine the internal consistency of the composition of the measurement model,
the Rho Coefficient and the Cronbach’s Alpha Coefficient were examined to establish
reliability. According to Kline (2005:59), the reliability coefficient should fall between zero
and 1.00. Values close to 1.00 are desired. The Rho Coefficient of internal consistency was
found to be 0.745. This was above the minimum value of 0.70. Correspondingly, the
Cronbach’s Alpha was also found to be above the minimum value of 0.70 at 0.715. Both values
showed a reasonable level of internal consistency and therefore reliability.
Table 10.48: Reliability and Construct Validity of Residential Satisfaction Model
Factor Indicator
Variable
Factor
Loading
Cronbach’s
Alpha
Rho
Coefficient
Residential RS1 0.718
Satisfaction RS3 0.551 0.715 0.745
(RS) RS5 0.479
RS7 0.644
*Parameter estimates are based on standardized solutions
The construct validity was determined by examining the magnitude of the parameter
coefficients. High parameter coefficients of greater than 0.5 indicate a close relation between
the factor and an indicator variable. A parameter coefficient of 0.5 is interpreted as 25% of the
total variance in the indicator variable being explained by the latent variable (factor). Therefore,
a parameter coefficient has to be between 0.5 - 0.7or greater to explain about 50% of the
variance in an indicator variable (Hair et al., 1998:111). Assessment of the unstandardized
parameter coefficient presented in Table 10.47, shown that they were significantly high with
the minimum of 0.411. The parameter estimate of 0.411 meant that the residential satisfaction
factor accounted for about 45.12% of the variance in RS3 and therefore indicative of a good fit
between the indicator variable and the factor.
Therefore, the neighbourhood factor satisfied both internal reliability and the construct validity
criteria. The Rho value was above the minimum value of 0.70 (Table 10.48) and the construct
validity criteria was justified by the magnitude, signs and statistical significance of all
parameter coefficients.
444
Summary on Residential Satisfaction Measurement Model
SEM findings revealed that the residual covariance estimates fell within the acceptable range;
the robust fit indexes had an acceptable fit, while the RMSEA value and the RMSEA with 90%
confidence interval produced a reasonable fit. All other parameter estimates were statistically
significant and feasible. It was therefore, concluded that the measurement model for the
residential satisfaction construct had an adequate fit to the sample data. Consequently, there
was no need to improve the measurement before it could be included in the full latent variable
model.
In general, the results of the measurement model analysis were mixed, although some variables
approached or met the threshold of ‘well-fitting’ and ‘good-fitting’ for various indexes based
on the statistical test and the significant of the parameter estimates. Despite this mixed results,
analysis of the full structural model will have to proceed based on a number of factors. First,
the internal consistency and reliability analyses conducted yielded acceptable results. The Rho
Coefficient of internal consistency was found to be above the minimum value of 0.70 (Table
10.48). Correspondingly, the Cronbach’s Alpha was also found to be above the minimum value
of 0.70 (Table 10.49). According to Kline (2005:59), the reliability coefficient should fall
between zero and 1.00. Values close to 1.00 are desired. Hence, the internal consistency and
reliability was met. Secondly, the indicator variables yielded high correlation values, which
suggested a high degree of linear association between the indicator variables and the factors.
In addition, the R2 values were also found to be close to the desired value of 1.00 and hence
indicating that the factors explained the variance in the indicator variables. This meant that the
results suggested that the indicator variables significantly predict the factor constructs, because
a majority of the measured variables are significantly associated with the factors. Lastly, the
construct validity as determined by examining the magnitude of the parameter coefficients
(factor loading) also revealed that the parameter coefficients (Z-statistics) indicated a close
relation between the factors and the indicator variable (Table 10.49). A parameter coefficient
of 0.5 is interpreted as 25% of the total variance in the indicator variable being explained by
the latent variable (factor). Hence, the reported parameter coefficient explained more than 25%
of the variance in the indicator variable, which were indicative of an adequate fit between the
indicator variables and the factors.
445
10.3.4 Structural Model – Testing Of the Hypothesised SEM Model
After showing that each of the six latent factors to be included in the full latent model was
working very well based on the test indexes and the statistically significance’s of the parameter
estimates, the full structural model was tested, which included all six factors (with tested
indicator variables) and the residential satisfaction outcome (manifest) variables for the study.
Once again, CFA measurement model for latent constructs were tested in order to confirm that
the indicators that have been used to measure one or more latent factors are indeed doing so;
thus loadings of the indicators on the specific factors are examined to see how well each factor
has been specified in the context of the others.
Table 10.49: Reliability and Construct Validity of the Latent Variables
Latent
(Exogenous)
Factors
No. of
Indicator
Variables
Indicator
Variable
Parameter
Coefficient
Rho
Coefficient
Cronbach’s
Alpha
DUF1 0.834
DUF2 0.872
DUF3 0.868
Dwelling Unit 8 DUF5 0.677 0.920 0.915
Features DUF9 0.578
DUF12 0.692
DUF16 0.669
DUF17 0.850
NAE1 0.858
Needs and 4 NAE2 0.882 0.915 0.912
Expectations NAE3 0.845
NAE4 0.820
BNP1 0.821
Beneficiary BNP2 0.923
Participation 4 BNP3 0.937 0.939 0.938
BNP4 0.881
BQF2 0.769
BQF3 0.468
Building 6 BQF4 0.893 0.892 0.885
Quality BQF5 0.897
Features BQF10 0.809
BQF11 0.649
NDF1 0.719
NDF3 0.662
Neighbourhood 5 NDF5 0.684 0.772 0.764
Features NDF7 0.722
NDF10 0.401
SPG8 0.792
Services 4 SPG9 0.894 0.910 0.909
446
provided by SPG12 0.858
Government SGP13 0.839
RS1 0.718
Residential 4 RS3 0.551 0.745 0.715
Satisfaction RS5 0.479
RS7 0.644
Covariances between the latent factors are added to the model for any relationship that will be
examined when the structural model is tested. Also, covariances between the latent factors and
outcome variables are also added to rule out the possibility that any of them may serve as an
indicator of any of the proposed factors.
As already indicated above (analysis of the measurement models), the measurement models
indicated that the models (latent variables CFA’s) worked well and it was therefore feasible to
test the full latent variable model. The question of whether measurement models should be
checked before analysing the full SEM, or not, is simply a strategy a researcher adopts (Hayduk
& Glaser, 2000:122). Similarly, the question of how many factors a construct should have is
also debatable (Bollen, 1989; Hayduk & Glaser, 2000:122). However, assessing the
measurement models first has an advantage. The first merit of analysing the latent variable
measurement models separately before analysing the full SEM model is that the research is
assured of a proper working measurement model before analysing the full SEM latent model.
Hence, the researcher avoids the frustration of re-specifying the full model if a solution cannot
be obtained. Therefore, Herting and Costner (2000:100) state that “if a CFA model cannot be
satisfactorily fitted moving to the structural model will provide no additional guidance or
benefit”. Moreover these observations as presented in the current study were a pure
confirmatory analysis and therefore recommendations were based on whether the postulated
priori model fit the sample data or not. Hence, not all the initially derived indicator variables
form the literature which were on the questionnaires were tested in the CFA, as the preliminary
residual covariance matrix (factor loadings) for some indicator variables of some latent
constructs were more than the recommended value. A residual covariance matrix value greater
than 2.58 is described as large (Byrne, 2006:94). In order for a model to be described as well-
fitting, the distribution of the residuals should be symmetrical and centred around zero (Byrne,
2006:94; Joreskog & Sorbom, 1988).
447
10.3.5 Hypothesised Relation for the Structural Model
The hypothesised model (Model 1.0) was tested, in which dwelling unit features,
neighbourhood features, building quality features, services provided by the government,
beneficiary participation, needs and expectations were expected to define residential
satisfaction. The hypothesised model was fitted to the data for the entire sample and, as is the
norm, covariances for all the exogenous factors and variables were specified. The six-factor
model was fitted to the data with the Robust Maximum Likelihood (RML) method of EQS and
the model converged. As with all of the analyses presented in this study, the testing of this
model was based on the Robust ML estimation and robust statistics were used to ascertain the
fit of the model. The robust solution adjusts for non-normality in the data. As is the norm in
SEM analyses (Kline, 2005), one variable loading per latent factor was set equal to 1.0 in order
to set the metric for that factor.
Figure 10.9: Hypothesised Model of Residential Satisfaction (Model 1.0)
10.3.6 Fit Statistics on the Structural Model
A confirmatory factor analysis of the full latent model was conducted. The full structural model
hypothesised that dwelling unit features, neighbourhood features, building quality features,
services provided by the government, beneficiary participation, needs and expectation define
residential satisfaction of occupants’ in subsidised low-income housing. The SEM Model is
presented in Figure 10.10 (Model 2.0). Model 2.0 is founded on the general hypothesis for the
study, which is based on the fact that overall residential satisfaction is directly related to the
influence of the exogenous variables’ in predicting / determining overall beneficiary
satisfaction. The theory and basis of the model was presented in Chapter 9 of the thesis.
448
The number of cases that were analysed for the full latent variable Model 2.0 was 751. Out of
the total sample size, all 751 cases had positive weights, while 73 had missing variables, which
were corrected with the maximum likelihood method of missing data correction. The model
had 36 dependent variables and 42 independent variables. It also had 90 free parameters and
44 numbers of fixed nonzero parameters. The covariance matrix of the model was analysed
using the robust maximum likelihood estimation method. Raw data was used for the analysis.
The raw data was not transformed since data transformation can provide an incorrect
specification (Shook, Ketchen, Hult & Kacmar, 2004:399). One alternative to transformation
is to use an estimation approach available in EQS (robust maximum likelihood) as already
discussed, which adjusts the model fit chi-square test statistics and standard errors of individual
parameter estimates.
449
Figure 10.10: Model 2.0 - Integrated Holistic Residential Satisfaction Model
Model Parameters (from left to right): Residential Satisfaction (Endogenous variable); Exogenous variables: DUF (8 indicator variables),
NDF (5 indicator variables), BQF (6 indicator variables), SPG (4 indicator variables), BNP (4 indicator variables), and NAE (4 indicator
variables).
D7
RS1RS3
RS5RS7
F7
E139E141
E143E145
DUF1DUF2
DUF3DUF5
DUF9
DUF12
DUF16
DUF17
F1
E61E62
E63E65
E69E72
E76E77
NDF1NDF3
NDF5NDF7
NDF10
F2
E78E80
E82E84
E87
BQF2BQF3
BQF4BQF5
BQF10
BQF11
F3
E101E102
E103E104
E109E110
SPG8SPG9
SPG12
SPG13
F4
E124E125
E128E129
BNP1BNP2
BNP3BNP4
F5
E130E131
E132E133
NAE1NAE2
NAE3NAE4
F6
E135E136
E137E138
450
Figure10.11: Model 2.0 - Integrated Holistic Residential Satisfaction Model Covariances Association
Covariance Relationship (from left to right): NAE (4 indicator variables), BNP (4 indicator variables), SPG (4 indicator variables), BQF (8
indicator variables), NDF (5 indicator variables), DUF (8 indicator variables) & RS (4 indicator variables).
D7
RS1
RS3
RS5
RS7
RS
E139
E141
E143
E145
DUF1
DUF2
DUF3
DUF5
DUF9
DUF12
DUF16
DUF17
DUF
E61
E62
E63
E65
E69
E72
E76
E77
NDF1
NDF3
NDF5
NDF7
NDF10
NDF
E78
E80
E82
E84
E87
BQF2
BQF3
BQF4
BQF5
BQF10
BQF11
BQF
E101
E102
E103
E104
E109
E110
SPG8
SPG9
SPG12
SPG13
SPG
E124
E125
E128
E129
BNP1
BNP2
BNP3
BNP4
BNP
E130
E131
E132
E133
NAE1
NAE2
NAE3
NAE4
NAE
E135
E136
E137
E138
451
10.3.6.1 Analysis of Residual Covariance Estimate
Investigation of the average absolute residual values of the structural model revealed that all
the absolute residual values and the average off-diagonal absolute residual values were close
to zero. The unstandardized average off-diagonal residual was 0.0820 while the standardized
average off-diagonal residual was found to be 0.0700 (refer Appendix H). These residual values
were considered small as they were all less than 2.58 (Byrne, 2006:94). In addition, 99.99% of
the residuals fell within the acceptable range of -0.1 and +0.1. The significance of this
distribution is that for a structural model to be described as well-fitting, the distribution of
residuals should be symmetrical and centred around zero (Byrne, 2006:94; Joreskog & Sorbom,
1988), which the analysed data has displayed. From the above information, the results suggest
that the hypothesised structural model had a good-fit to the sample data; overall, the model as
a whole appears to be quite well fitting. Therefore, since this initial assessment of the structural
model residuals indicated a good fit; further tests of goodness-of-fit were justified.
10.3.6.2 Structural Model Goodness-of-Fit statistics – Robust Maximum
Likelihood
The test of the hypothesis that occupants’ residential satisfaction is a six factor structure as
depicted in Figure 10.10 (Model 2.0) via the sample data on the model yielded a robust
Likelihood Ratio Test (S – Bχ2) of 2762.558 with 540 degrees of freedom. The associated p-
value was less than 0.0001 (p = 0.0000) with a sample of 751 cases. The chi-square index
suggested that the difference between the hypothesised model and the sample data matrix was
significant, but not entirely adequate. Interpreted literally, this test statistics indicates that given
the present sample data, the hypothesis bearing on occupants’ residential satisfaction (RS)
relates as summarised in the model, represents an unlikely event (i.e. occurring less than one
time in a thousand under the hypothesis) and should be rejected. However, the chi-square test
(Likelihood Ratio Test) of fit is very sensitive and therefore, could not be relied upon to
determine model fit. The chi-square test tends to be affected by the sample size with a
propensity to reject models if the samples are large (Joreskog & Sorbom, 1993). Yet, the
analysis of covariance structure (SEM) is grounded in large sample size theory (Byrne,
2006:96). As such, large sample sizes are critical to obtaining precise parameter estimates, as
well as to the tenability of asymptotic distribution approximations (Byrne, 2006:96;
MacCallum et al., 1996). Therefore a normed Chi-square value is usually adopted by most
researchers (Bentler, 2005; Byrne, 2006; Kline, 2005:137; MacCallum et al., 1996). Normed
chi-square is the procedure of dividing the S – Bχ2 by the degrees of freedom. The normed
452
values of up to 3.0 or even 5.0 are recommended (Kline, 2005:137). From the above chi-square
and degrees of freedom values the ratio was found to be 5.12:1. This ratio was within the limit
of 5.0 advocated for by some authors (Byrne, 2006; Kline, 2005:137) and therefore indicative
of a reasonable fit of the model.
Similarly, other fit indexes indicated a good fit of the model to the latent variables. The robust
CFI index was found to be 0.955. The CFI index was less than 0.90, which is the lower limit
value for model acceptance. However, a two strategic approach is considered satisfactory to
accept or reject a mode (Hu & Bentler, 1999:28). Hence, RMSEA and SRMR statistics were
further used to decide on the acceptability of the model.
Table 10.50: Robust Fit Indexes for Structural Model 2.0
Fit Index Cut-off value Model 1.0 Comment
S – Bχ2 2762.558
df 0≥ 540 Acceptable
Normed χ2 3 or 5 5.12:1 Acceptable
CFI 0.90≥ acceptable
0.95≥ good fit
0.955 Good fit
GFI 0.90≥ acceptable
0.95≥ good fit
0.981 Good fit
SRMR 0.08≥ acceptable
0.05≥ good fit
0.042 Good fit
RMSEA 0.08≥ acceptable
0.05≥ good fit
0.074 Acceptable fit
RMSEA 90% CI 0.071:0.077 Acceptable range
The robust RMSEA with 90% confidence interval (lower bound value = 0.071 and the upper
bound value = 0.077) was found to be 0.074. The RMSEA index was just above the upper limit
of 0.05 for the model to be described as good. However, the value of 0.074 indicated that the
model has an acceptable fit. MacCallum et al. (1996) informed that an RMSEA of between
0.08 and 0.10 provides a poor fit and below 0.08 shows a good fit. Also, Hooper et al (2008)
posits that RMSEA with 90% confidence interval in a well-fitting model lower limit should be
close to 0.00 while the upper limit should be less than 0.08. Hence, the structural model
RMSEA with 90% CI met the above criteria and the model could be considered good fit. In
addition, the absolute fit index, SRMR, was found to be 0.042, while the GFI was found to be
0.981. The SRMR and the GFI absolute fit index indicated an adequate fit of the full structural
model to the sample data. Therefore, the goodness-of-fit statistics indexes (CFI, GFI, SRMR,
453
RMSEA, RMSEA @ 90% and S – Bχ2) met the condition for model acceptance (Table 10.50).
However, the Lagrange Multiplier (LM) test conducted on the full latent model sample data
did not reveal any significant indicators of model mis-specification of the hypothesised
parameters. In EQS, a model can be said to be mis-specified if there are any mis-fitting
parameters using the LM test (Byrne, 2006:112). The criterion is to identify any significant
drop in the χ2 values of parameters. Also, in univariate and multivariate analysis, the probability
that a parameter estimate is equal to zero should be less than 0.05 in order to be rejected. This
is also an indication of mis-specification according to Byrne (2006:112). Hence, inspection of
the LM test output revealed that there were no significant mis-fitting variables that would have
warranted mode re-specification.
10.3.6.3 Internal Reliability and Construct Validity of the SEM Model
The Rho Coefficient and the Cronbach’s Alpha Coefficient were examined in order to establish
score reliability for the SEM model. According to Kline (2005:59), the reliability coefficient
should fall between zero and 1.00. Values close to 1.00 are desired. The Rho Coefficient of
internal consistency was found to be 0.954. This value was above the minimum required value
of 0.70. Similarly, the Cronbach’s Alpha was above the minimum acceptable value of 0.70 at
0.909. Both of these values indicated a high degree of internal consistency and homogeneity
(Table 89). These findings informed that the degree to which responses are consistent across
all indicator variables was statistically significant, indicating that the measures of the latent
variables total scores are the best possible unit of analysis for the exogenous variables which
thus predictor the endogenous variable (residential satisfaction).
The construct validity for the SEM model was determined by examining the magnitude of the
parameter coefficients. High parameter coefficients of greater than 0.5 indicate a close relation
between the factor and an indicator variable. A parameter coefficient of 0.5 is interpreted as
25% of the total variance in the indicator variable being explained by the latent variable
(factor). Therefore, a parameter coefficient has to be between 0.5 - 0.7or greater to explain
about 50% of the variance in an indicator variable (Hair et al., 1998:111). Inspection of the
standardized parameter coefficient presented in Table 10.52, show that they were significantly
high with a maximum of 0.924. The parameter estimate of 0.924 meant that the residential
satisfaction accounted for about 64.89% of the variance in BNP3 and therefore indicative of a
good fit between the indicator variable and the factor, likewise in the other factors.
454
10.3.6.4 Structural Model Hypothesis Testing
Besides assessing the goodness-of-fit of the structural model, feasibility of a model can be
judged by a further inspection of the obtained solution and this involves inspection of the
statistical significance of the parameter estimates, standard errors and the test statistics
(Raykov, 1991:501). Therefore, the rejection of the hypothesis depends on how reasonable
parameter estimates were in terms of their magnitude, signs and statistical significance. In
addition, if the output showed estimates that had correlation values greater than 1.00, had
negative variances and the correlation or covariances were not definite positive, then they were
said to be displaying unreasonable estimates (Byrne, 2006:103). Likewise, the test statistics
had to be greater than 1.96 based on the probability level of 5% before the hypothesis can be
rejected (Byrne, 2006:103). The test statistics reported was the parameter estimate divided by
its standard error and therefore it functions as a Z-statistics to test that the estimate is
statistically different from zero. Hence, the test was used to evaluate the hypothesis.
Testing the influence of the exogenous variables on overall beneficiary’s residential
satisfaction
It was general hypothesised that beneficiary’s overall residential satisfaction with their housing
product is directly related to the influence of the exogenous variables’ in predicting overall
beneficiaries’ satisfaction in public funded housing scheme in developing countries using
South Africa as a case study.
Results from the SEM analysis yielded support for the hypothesis. The hypothesised
relationships between all exogenous factors and the endogenous factor were found to be
significant and they all had definite positive directions. Inspection of the correlation values,
standard errors and the test statistic in Table 10.51 revealed that all standardized coefficient
correlation values were not greater than 1.00. All test statistics (Z-values) were greater than
1.96 (p<0.05) and the signs were appropriate, they all have positive values (refer to Table 10.51
& 10.52), suggesting that all latent variables measured the overall beneficiary residential
satisfaction. The estimates were therefore reasonable as well as statistically significant.
Therefore, the general hypothesis that beneficiary’s overall residential satisfaction with their
housing product is directly related to the influence of the exogenous variables’ in predicting
overall beneficiary satisfaction in publicly funded housing schemes in South Africa could not
be rejected. The relationship between residential satisfaction and beneficiary’s participation
455
indicators was found to be the most significant. The parameter with the highest standardised
coefficient for this factor was the indicator variable BNP3. The parameter coefficient was
found to be 0.924. The indicator variable BNP3, which asked the beneficiaries their level of
agreement, if an owner should be consulted about the house construction; was found to be most
associated with overall beneficiary residential satisfaction than any other indicator variable.
However, in order to determine if each exogenous variable considerably predicted the
endogenous construct, an inspection of the interfactor correlation (R2) values will be examined,
thus establishing the exogenous variables direct influence on the dependent variable (presented
in the subsequent sections). However, the overall results therefore suggest that the exogenous
variables considerably predict the endogenous variable (residential satisfaction)
Also, assessment of the outcome variables of overall residential satisfaction revealed that all
standardized factor loadings values were generally large and statistically significant (values
ranged from 0.391 to 0.797). However, the interfactor correlation (R2) values were not all
statistically significant (values ranged from 0.0.153 to 0.635) as shown in Table 10.52. The
variance accounted for in each measure by the endogenous variable revealed that the scores
were significance at 5% level. The score results suggested that the interfactor relationship
between the manifest variables is weak and does not have significant level of correlations.
Testing the direct influence of dwelling unit features on overall beneficiary’s residential
satisfaction
Results from the confirmatory factor analysis of the full structural model, presented in Table
10.51 and 10.52 yield support for the general hypothesis. The relationship between the factors
and the endogenous variable (dependent variable) was found to be statistically significant at
5% probability level. On the other hand, all standardized parameter estimates showed high
correlations values close to 1.00 suggesting a high degree of linear association between the
indicator variables and the endogenous construct.
Inspection of the R2 values for the dwelling unit indicator variables revealed that the values
were close to the desired value of 1.00. The only exceptions were the indicator variables DUF5
(R2 = 0.457), DUF9 (R2 = 0.328), DUF12 (R2 = 0.478) and DUF16 (R2 = 0.453). The R2 values
for this variable were below 0.50 suggesting that these indicator variables did not considerably
predict the endogenous factor construct. DUF9 (R2 = 0.328) had the weakest association
amongst these variables. Despite the none coherent level of the interfactor correlation within
456
the indicator variables, the direct influence of dwelling unit factor on overall residential is
statistically significant as the degree of variances accounted for in each measure was adequate
as shown in Table 10.52.
Testing the direct influence of neighbourhood features on overall beneficiary’s residential
satisfaction
Examination of the R2 values for the neighbourhood feature indicator measures revealed that
only two indicator values were close to the desired value of 1.00. The other variables were
weak in predicting the endogenous variable. However, a further assessment of the variance
accounted for in each measure by the endogenous variable revealed that the scores were
significance at 5% level. The values were above the minimum required value of 25%.
Therefore, from the statistical assessment, score results suggested that the influence of this
factor on the endogenous variable was weak (indirect). However, the total variance accounted
for revealed that it has a good indirect association with the other latent variables in the
prediction of overall residential satisfaction.
Table 10.51: Model 2.0 Factor Loadings and Z-statistics
Indicator
Variable
Unstandardized
Coefficient (λ)
Standardized
Coefficient (λ)
(Z-values)
Significant
at 5%
level?
DUF1 ** 0.830 ** Yes
DUF2 0.998 0.872 40.601 Yes
DUF3 0.951 0.866 36.595 Yes
DUF5 0.662 0.676 20.491 Yes
DUF9 0.559 0.573 17.523 Yes
DUF12 0.782 0.691 23.543 Yes
DUF16 0.712 0.673 19.399 Yes
DUF17 0.951 0.855 33.512 Yes
NDF1 1.00 0.723 ** Yes
NDF3 0.919 0.671 22.368 Yes
NDF5 0.859 0.647 16.957 Yes
NDF7 1.290 0.761 20.580 Yes
NDF10 0.612 0.380 11.849 Yes
BQF2 ** 0.774 ** Yes
BQF3 0.529 0.441 8.551 Yes
BQF4 1.157 0.894 28.198 Yes
BQF5 1.126 0.894 28.944 Yes
BQF10 1.058 0.813 25.657 Yes
BQF11 0.827 0.651 18.623 Yes
SPG8 ** 0.785 ** Yes
457
SPG9 1.165 0.891 29.087 Yes
SPG12 1.112 0.862 25.896 Yes
SGP13 1.075 0.844 27.549 Yes
BNP1 ** 0.835 ** Yes
BNP2 1.082 0.922 41.280 Yes
BNP3 1.098 0.924 40.218 Yes
BNP4 1.020 0.888 36.359 Yes
NAE1 ** 0.869 ** Yes
NAE2 1.120 0.884 41.553 Yes
NAE3 0.941 0.832 29.545 Yes
NAE4 0.863 0.818 27.866 Yes
RS1 ** 0.797 ** Yes
RS3 0.418 0.510 13.527 Yes
RS5 0.491 0.391 9.122 Yes
RS7 0.842 0.617 14.956 Yes
(Robust Statistical Significance at 5% level)
** SEM Analysis Norm (Kline, 2005) - One variable loading per latent factor is set equal to
1.0 in order to set the metric for that factor.
Testing the direct influence of building quality feature features on overall beneficiary’s
residential satisfaction
Similarly, inspection of the R2 values for the building quality feature indicators revealed that
four out of the six indicator variables that were used to measure the latent factor, had values
close to the desired value of 1.00. Two other variables (BQF3 & BQF11) were weak in
predicting the endogenous variable (Table 10.52). BQF3, which measured the occupants’ level
of satisfaction with the water pressure in their unit had the lowest R2 value. This suggests that
the interfactor relationship of these variables and other indicators in determining overall
residential satisfaction is minor. Further, assessment of the variance accounted for in each
measure by the endogenous variable revealed that all scores were significance at 5% level. The
reported parameter coefficient explained more than 25% of the variance in the latent variable,
which were indicative of an adequate fit between the latent variables and the endogenous
construct. Thus, the score results suggested that the influence of this latent factor on the
endogenous variable was direct and significant.
Testing the direct influence of services provided by government on overall beneficiary’s
residential satisfaction
Inspection of the score values for this factor revealed that all standardized factor loadings were
generally large and statistically significant (values ranged from 0.785 to 0.891). Likewise, the
458
interfactor correlation (R2) values was also large and statistically significant (values ranged
from 0.616 to 0.794) as shown in Table 10.52. Also, the variances accounted for in each
measure by the endogenous variable revealed that the scores were significance at 5% level. The
values were above the minimum required value of 25%. Hence, the score results suggested that
the influence of the services provided by the government on the endogenous variable was direct
and statistically significant.
Testing the direct influence of beneficiary participation on overall beneficiary’s residential
satisfaction
An assessment of the standardized factor loadings revealed that all values were generally large
and statistically significant (values ranged from 0.835 to 0.924). Also, the interfactor
correlation (R2) values were large and statistically significant (values ranged from 0.698 to
0.854) as shown in Table 10.52. The total variances accounted for in each indicator variables
by the endogenous variable revealed that the scores were significance at 5% level. The score
results suggested that the influence of beneficiary participation in determining beneficiaries
overall satisfaction with their subsidised dwelling units was direct and statistically significant.
Table 10.52: Model 2.0 factor loadings, Z-statistics, Variance accounted for & reliability
and construct validity
Indicator
Variable
Standardized
Coefficient (λ)
(Z-values)
R2
Total Variance
Cronbach’s
Alpha
Rho
Coefficient
DUF1 0.830 ** 0.689 62.41%
DUF2 0.872 40.601 0.761 63.56%
DUF3 0.866 36.595 0.751 63.40%
DUF5 0.676 20.491 0.457 57.48%
DUF9 0.573 17.523 0.328 53.36%
DUF12 0.691 23.543 0.478 58.02%
DUF16 0.673 19.399 0.453 57.37%
DUF17 0.855 33.512 0.731 63.31%
NDF1 0.723 ** 0.523 59.12%
NDF3 0.671 22.368 0.450 57.30%
NDF5 0.647 16.957 0.419 56.41%
NDF7 0.761 20.580 0.579 60.35%
NDF10 0.380 11.849 0.145 43.18%
BQF2 0.774 ** 0.599 60.75%
BQF3 0.441 8.551 0.195 46.87%
BQF4 0.894 28.198 0.799 64.13%
BQF5 0.894 28.944 0.799 64.13% 0.909 0.954
BQF10 0.813 25.657 0.661 61.92%
BQF11 0.651 18.623 0.424 56.56%
459
SPG8 0.785 ** 0.616 61.09%
SPG9 0.891 29.087 0.794 64.06%
SPG12 0.862 25.896 0.743 63.29%
SGP13 0.844 27.549 0.713 62.80%
BNP1 0.835 ** 0.698 62.55%
BNP2 0.922 41.280 0.850 64.84%
BNP3 0.924 40.218 0.854 64.89%
BNP4 0.888 36.359 0.788 63.98%
NAE1 0.869 ** 0.755 63.48%
NAE2 0.884 41.553 0.782 63.87%
NAE3 0.832 29.545 0.692 62.46%
NAE4 0.818 27.866 0.668 62.06%
RS1 0.797 ** 0.635 61.45%
RS3 0.510 13.527 0.260 50.50%
RS5 0.391 9.122 0.153 43.88%
RS7 0.617 14.956 0.381 55.24%
(Robust Statistical Significance at 5% level)
Testing the direct influence of needs and expectation on overall beneficiary’s residential
satisfaction
Similarly, the inspection of the standardized factor loadings revealed that all values were
generally large and statistically significant (Table 10.52). The R2 values were large and
statistically significant (values ranged from 0.668 to 0.782). This suggests that the interfactor
relationship between the variables is significant. The variance accounted for in each measure
by the endogenous variable revealed that the scores were significance at 5% level. The score
results suggested that the direct influence of beneficiary participation in determining
beneficiaries overall satisfaction with their subsidised dwelling unit is statistically significant.
10.3.6.5 Summary on SEM Model
Results from the EQS output revealed that the robust fit indexes, CFI, GFI, SRMR and the
RMSEA values met the cut-off index criteria and the parameter estimates were found to be
statistically significant and reasonable. The postulated model, which hypothesised that overall
residential satisfaction, is directly related to the influence of the exogenous variables in
predicting / determining overall beneficiaries’ satisfaction fit the sample data adequately. In
view of the fact that the analysis was confirmatory of the priori model, there was no need to
further improve the structural model. Investigation of alternative models, such as the reduction
of latent variables could be a matter for further studies as the current study was a confirmatory
analysis of the priori. However, the Lagrange Multiplier test did not unveil significant
indication of model mis-specification demanding a re-specification. Byrne (2006:112) informs
460
that for most models, model enhancement is purely a process that attempts to fine-tune small
features of the sample and does not essentially add value to an already fitted model, like the
present model. Likewise, MacCallum et al. (1992:501) cautioned that “when an initial model
fits well, it is probably unwise to modify it to achieve even better fit because modifications
may simply be fitting idiosyncratic characteristics of the sample”. Hence, the presented model
(Model 2.0) was therefore accepted with its level of fit. The lines of covariances (Figure 10.11)
indicates that the integrated holistic influence of the latent variables determines overall
residential satisfaction because they were all statistically significant.
10.4 CONCLUSION
The postulation for the overall model was that overall residential satisfaction is directly related
to the influence of the exogenous (latent) variables’ in predicting / determining overall
beneficiary satisfaction. In this chapter, the SEM results of the measurement model were
presented. These results were obtained from an analysis (SEM) to determine whether the
indicator variables (questionnaire items) actually measured the constructs that they were
supposed to measure. In addition, results were presented in order to establish whether the
statistically significant number of factors for the latent models were feasible. Likewise, the
measurement model reliability and construct validity were also reported.
The analysis of the structural model (full latent model-SEM) was conducted, which validates
the hypothesised integrated holistically residential satisfaction model. The influence of the
latent variables on the endogenous variable was also reported. Considering the feasibility and
statistical significance of all parameter estimates; the substantially good fit of the model, with
particular reference to the CFI (0.955), SRMR (0.042) and RMSEA (0.074) values; and lack
of any substantial evidence of model misfit, it was concluded that there is therefore, no need to
further improve the fit of the structural model. Further findings from the SEM results revealed
that the exogenous variables influences’ determine beneficiary’s residential satisfaction with
their housing units. Further, it was found that five of the exogenous variables have a significant
direct influence on the endogenous variables, while only one had a weak (indirect) influence
in determining residential satisfaction in subsidised low-income housing in South Africa.
Adhering to this caveat, it is therefore concluded that the six-factor model schematically
portrayed in Figure 10.10 (Model 2.0) represents an adequate description of residential
satisfaction in subsidised low-income housing in South Africa.
461
CHAPTER ELEVEN
DISCUSSION OF RESULTS
11.1 INTRODUCTION
The last chapter discussed and presented the quantitative research findings, with reference to
the descriptive and inferential statistics. Also, the research hypotheses were tested based on the
SEM result analysis validating the assumption that subsidised housing occupants’ residential
satisfaction is a six-factor model schematically portrayed in Figure 10.10 (Model 2.0). The
findings from the SEM analysis which model subsidised housing occupants’ residents’
satisfaction as a six-factor model showed that the factors of dwelling unit features,
neighbourhood features, services provided by the government, building quality features,
beneficiary participation, needs and expectation were found to have a significant influence in
predicting the occupants’ residential satisfaction. However, the neighbourhood feature had a
weak (indirect) influence in predicting residential satisfaction; nevertheless, the covariation
with the other exogenous construct to determine residential satisfaction was found to be
statistically significant.
11.2 QUESTIONNAIRE SURVEY RESULTS
The structural model results of the thesis hypothesis testing revealed that the general
hypothesis, which states that dwelling unit features, neighbourhood features, building quality
features, services provided by the government, beneficiary participation, needs and
expectations jointly predict residential satisfaction of occupants in subsidised low-income
housing could not be rejected. In view of the hypothesis, the discussion section will be
structured in order to respond to the research’s sub-questions.
Lazenby (1988:55) informs that, “housing satisfaction can be defined as the level of satisfaction
with a specific house within a chosen residential, physical and social environment, as well as
its specific housing attributes.” The first observation that must be made is that the respondents
in this study are first-time homeowners, who most probably lived in shacks or as backyard
squatters in their previous accommodation, as revealed by findings from the length of residency
of the respondents (Table 10.1). Findings from this aspect revealed that 42.1% have been living
in the units for more than eight years, while a combined percentage of 48.0% have been living
462
in the units between 3-8 years. Hence, it is evident the respondents who participated in the
survey have a good knowledge of their residential dwelling and of the neighbouring
environment. Therefore, their opinion could thus be used to make an inference.
11.2.1 Dwelling Unit Features Influence on Beneficiary’s Residential Satisfaction
The research’s sub-question RO1 was posed to determine the extent that beneficiary’s RS is
influenced by the dwelling unit features in the subsidised low-income houses.
First, a descriptive assessment of the available dwelling unit features revealed that 99.2% of
the respondents informed they have bedrooms in their dwelling unit. Also, 83.5% informed
they have a kitchen in their dwelling unit, while 16.5% said they do not have. This was followed
by 63.5% who have bathrooms inside their dwelling, while 36.4% informed they do not have
bathrooms in their dwellings. A further assessment of the services available inside the units
revealed that 99.5% beneficiaries (respondents) have water for domestic usage in their houses;
98.0% also indicated that they have sanitary fittings (meaning: shower, bath, toilet basin, wash
hand basin and taps) installed in their units; while 99.3% informed they have access to
electricity. Further, results from the structural model revealed that the relationship between the
dwelling unit features and the endogenous variable (residential satisfaction) was found to be
statistically significant at 5% probability level. On the other hand, all standardized parameter
estimates showed high correlations values, suggesting a high degree of linear association
between the indicator variables and the endogenous construct. Also, the interfactor values for
this variable were considerate; suggesting that more than 50.0% of the latent variable
considerably predicted the endogenous factor construct. The summarised result for this variable
revealed that the latent factor has a direct influence in determining overall residential
satisfaction.
The results suggests that most variables included in the model have a significant effect on
residential satisfaction and the effects are generally in line with findings reported in the
previous research (Lu, 1999; Mohit et al., 2010; Salleh, 2011; Ozo, 1990). The findings further
suggest that dwelling unit features is a significant determinant of housing satisfaction. Hence,
the lower the level of dwelling unit features, the less likely the residents will be satisfied with
the housing units. Additionally, the results also suggest that the residents were dissatisfied with
features, such as location of the dining room, size of the wardrobe/closet, amount of privacy
within the house and the overall appearance of the house. These findings concurs with the work
463
of Moolla et al. (2011), which was conducted in Braamfischerville, Soweto, which was one of
the areas where data for the present study was collected. The studies found that occupants of
publicly provided low-income houses were dissatisfied with the amount of privacy within their
units and the overall appearance of the houses. Hence, Salleh et al. (2011) claims that pleasant
dwelling features are the main parameters used in describing residential satisfaction and in
order to distinguish if a house is of good quality or not, the dwelling units’ features from the
internal and external aspects and also the level of perceived privacy are used to qualify the
house. Moreover, Ukoha and Beamish (1997) states that satisfaction towards housing is related
to dwelling units features, while Kutty (1999) indicated that dwelling unit features are an
important indicator determining the residents’ satisfaction. However, the present work did not
concur with the work of Husna and Nurizan (1987) study where it was found that a large
number of their study’s respondents were dissatisfied with the characteristics of their dwelling
units.
Besides, the findings revealed a different stance towards non-satisfaction with the number of
bedrooms and size of bedrooms as previous studies has shown (Aigbavboa, 2010; Aigbavboa
& Thwala, 2011; Moolla et al., 2011; Ogunfiditimi, 2008). One of the major sources of housing
dissatisfaction in South African low-income housing is the limited number of bedrooms and
their sizes. A probable reason for respondents’ satisfaction with these specific dwelling unit
features as indicated by this study can be attributed to the number of dependents in the
household. Descriptive findings (Table 10.1) revealed that a majority of the households had
between 1 - 2 dependents with 99.1% respondents indicating that they have between 1 - 2
rooms in their dwelling units. In substantiating this further, Mohit et al. (2010), Salleh, et al.
(2011), Ibem and Amole (2011) and Ilesanmi (2010) noted that the ratio of the housing
occupants to the number of rooms (bedrooms) constitutes a major source of satisfaction to
public housing occupants. Likewise, Lane and Kinsey (1980) posit that the number of rooms
per family and the availability of other rooms for different uses influence satisfaction with
housing (Kaitilla, 1993).
The implication of these findings is that overall residential satisfaction is a product of the direct
influence of dwelling unit features. Hence, the residential satisfaction of South African low-
income housing occupants’ can be enhanced through improvement of the dwelling unit
features, such as the amount of privacy in the units, increase of the size of the wardrobes, which
is related to the increase of the size of the units and the improvement of the overall appearance
464
of the houses. Research in South African low-income housing has shown that despite the
acknowledged significance of the dwelling unit features to the subsidised housing occupants,
the Department of Human Settlement and other stakeholders responsible for the provision of
these houses have not responded significantly to rectify this challenge (Aigbavboa & Thwala,
2011; Charlton & Kihato, 2006; Mkuzo, 2011). The findings emanating from the dwelling unit
feature assessment were therefore significant, and when attention is given to the issues of
dissatisfaction regarding the dwelling unit features, the much desired housing satisfaction of
the low-income group residing in the subsidised houses will be realised. Furthermore, the
findings make it possible for policy makers to address factors of dwelling unit’s most
importantly the amount of privacy within the house and the overall appearance of the houses,
in such a way that it will ensure the occupants satisfaction with their dwelling units.
11.2.2 Building Quality Features Influence on Beneficiary’s Residential
Satisfaction
The findings suggested that building quality features have a direct influence on the prediction
of residents’ housing satisfaction. Findings from the interfactor relationship revealed that
building quality features had significant association with the latent variables in predicting the
endogenous variable (Table 10.52). From the assessment of the variance accounted for in each
measure by the endogenous variable revealed that all scores were statistically significance at
5% level. The reported parameter coefficient explained more than the baseline level of the
variance in the latent variable, which were indicative of an adequate prediction of the
endogenous construct. Hence, these results suggested that the influence of this latent factor on
the endogenous variable was direct and statistically significant.
The six indicator variables used in measuring building quality features construct were highly
causative to the endogenous variable as shown in Table 10.52. The wall and floor quality were
the most highly causative items followed by the plumbing quality indicator. The lowest
causative items were the water pressure, the quality of finishing of the sanitary system and the
internal construction quality. The findings suggest that the residents were satisfied with most
of the building quality features, while they were dissatisfied with the finish quality of the
sanitary system and the internal construction quality of the units. These finding support the
study done by Abdul Ghani (2008) who found that residents of low cost housing in Malaysia
were partially satisfied with the building quality features. It also concurs with the findings of
study done by Salleh et al. (2011) and Ukoha and Beamish (1997), which found that most
465
respondents (low-income occupants) were dissatisfied with the qualities of the internal
construction of their units. However, during the questionnaire survey, the researcher observed
that most of the units had problems with the quality of their housing units. Complaints from
the occupants varied from roofs and doors that were improperly built due to poor craftsmanship
to doors that did not open or close properly. The lack of ceilings in most units also led to high
levels of dissatisfaction because residents complained about extreme in temperature conditions
during seasonal changes. According to Turner (1972), the value of a house is of greater
importance to a person than the appearance of the housing unit. In addition, the structure of the
house, even if the building material was of a lower standard, would not affect the person’s
perception if value could be attached to the unit. This is clearly evident from the results of this
study; although the respondents’ found individual aspects of the housing units’ satisfactorily,
the overall level of dissatisfaction with the total house was high.
Literature informs that good building structure with good quality is an important indicator that
determines the residents’ satisfaction with the building and the value they place on the dwelling
(Kutty, 1999). Also, Duncan (1971) and Ramdane and Abdullah (2000), stated that the internal
construction quality of a dwelling unit are usually considered with regards to its overall
satisfaction, and when adequate, the occupants will be satisfied with their housing product.
Findings on this exogenous construct revealed that the occupants were not totally satisfied as
the measure of covariance and interfactor association with other indicators was average.
Furthermore, Elsinga and Hoekstra (2005) inform that the higher quality a dwelling is, the
higher the resident’s satisfaction towards it. They posit that housing quality and condition
should not be assessed based on one variable only, but from the objective and subjective
dimensions as assessed in the present study.
The implication of these findings are that overall residential satisfaction is a product of the
direct influence of building quality features and that the residential satisfaction of South Africa
low-income housing occupants can be enhanced through improvement of the building quality
features. The findings originating from the building quality feature assessment were therefore
significant because when attention is given to issues of dissatisfaction regarding the building
features various aspects, housing satisfaction of the low-income group residing in the
subsidised houses will be realised. Besides, the findings make it possible for policy makers to
address factors of building quality in a way that it will bring about occupants’ satisfaction with
their houses.
466
11.2.3 Neighbourhood Features Influence on Beneficiary’s Residential
Satisfaction
The SEM analysis for this variable (neighbourhood feature) indicator revealed that only two
indicators were closely associated with the dependent variable. The other variables were weak
in predicting the endogenous variable. However, a further assessment of the variance accounted
for in each measure by the endogenous variable revealed that the scores were significant, as
the values were above the minimum required value. The statistical assessment suggests that the
direct influence of this factor on the endogenous variable was weak (indirect). However, the
total variance accounted for revealed that it has a good indirect association with the other latent
variables in the prediction of overall residential satisfaction.
Similarly, descriptive assessment of the available neighbourhood features revealed that the
following were not present in the neighbourhood, parking facilities (94.0%),
playground/recreational facilities (55.2%), community hall (61.3%) and the presence of
disabled facilities (94.4%). Additional neighbourhood features that were lacking in a majority
of the neighbourhoods surveyed include: hospital/clinic (63.2%), Police services (66.8%) and
fire protection service (86.5%). However, the respondents had access to public transportation
(94.9%), nursery schools (76.2%), primary schools (68.0%), shopping malls (61.9%), and
places of worship (83.8%). These findings agrees with the work of Mohit et al. (2010), Parkes
et al. (2002) and Chapman and Lombard (2006) that indicated that most respondents in their
study were not satisfied with the security and crime prevention features in a low-income
residence because of the lack of a permanent policing facility in their neighbourhood. Also,
Zack and Charlton’s (2003) work, which was a South African based study, found that crime
and safety concerns, and the lack of adequate public transport feature strongly in beneficiary
complaints about their neighbourhoods, which ultimately lead to their dissatisfaction with the
neighbourhood and the housing unit.
Satisfaction with neighbourhood features have been observed as a vital determinant of
residential satisfaction (Vrbka & Combs, 1991) to the extent that residents are willing to
compromise the inefficiencies within the dwelling unit because of the satisfaction that is
provided by the neighbourhood facilities and features (Aigbavboa & Thwala, 2011; Onibokun,
1974; Ukoha & Beamish, 1997). Residents of the low-income housing scheme are most likely
to be dissatisfied with housing facilities that require residents to travel or walk long distances
467
to school; to the workplace, shopping areas, medical centres and the geographical areas around
their dwelling units, as most of them do not have access to the means of mobility. Easy access
to good public transportation, community and shopping facilities and physical environment
variables will provide residents’ satisfaction with their housing units.
Also, the research conducted by Bjorklund and Klingborg (2005) in eight Swedish
municipalities found the following top neighbourhood factors amongst others to be related to
residential satisfaction, these include proximity to commercial areas, access to playground and
recreational amenities, good quality walkways and access to community halls. The present
findings did not support the findings of a study conducted by Abdul (2008) where it was found
that neighbourhood facilities factors are the most dominant (direct influencing) factors in
determining the level of satisfaction towards housing. Also, Ramdane and Abdul’s (2000)
study on the factors of neighbourhood facilities, which found that neighbourhood factors have
a huge (direct) impact on the overall satisfaction with the housing facilities was not supported
by the present study findings on the neighbourhood feature construct.
There is broad consensus in the South African subsidised low-income landscape that many of
the neighbourhoods in which subsidised low-income housing is located, are not adequate and
do not offer a full range of amenities (Charlton, 2004). This is despite an obvious recognition
that the environment within which a house is situated is recognized as being equally as
important as the house itself in satisfying the needs and requirements of the occupants
(Charlton, Silverman, & Berrisford, 2003; National Department of Housing, 2000). Over and
over again in South African low-income housing development, provisions are made in most
township layout for essential facilities, and the land set aside, but for several years or long after
the housing has been developed in those areas, the amenities remain as undeveloped. For
instance, the Public Services Commission (2003) during the Housing Subsidy Scheme Review
in 2004 noted that access to schools was generally reasonable in new housing projects but that
a range of other facilities were often lacking, which supports this assessment. Thus, the South
African Department of Human Settlement acknowledged that most low-income residential
areas have been developed without the necessary social and other amenities and this “detracts
from the ideal to establish habitable, viable and sustainable human settlements” (Department
of Housing’s Overview of Achievements and Challenges Report of 2003:28). Hence, Charlton
(2004) posits that many housing projects have manifested as low density and mono-functional
neighbourhoods, lacking in integrated, holistic development. This circumstance does not
468
facilitate the economic growth or socio-economic development of low-income communities so
necessary to metropolise development. Likewise, it runs counter to the intention that ultimately,
the housing process must make a positive contribution to a non-racial, non-sexist, democratic
and integrated society (Charlton, 2004).
The findings on the hypothesis that neighbourhood features influence residential satisfaction
entails therefore that the minimum that the housing delivery stakeholder could do in order to
significantly influence residential satisfaction determination is to have a procedure in place.
This can be implemented through the strict enforcement of the Breaking New Ground
Programme (Integrated Residential Development Programme) by developing housing
neighbourhoods, which are comprehensive with the presence of all amenities and
infrastructures. The findings offer a minimum requirement that could be used by the
Department of Human Settlement to influence residential satisfaction. A checklist of items
defining the factors of neighbourhood features could ensure that stakeholders meet the basic
required criteria to influence residential satisfaction through the housing development
neighbourhood environment.
11.2.4 Beneficiaries Participation Influence on Residential Satisfaction
It will be necessary to inform the reader that beneficiary participation is the involvement of the
citizenry (community, public, beneficiaries, etc.) in the affairs of planning, governance and
overall development programmes at local or grass roots level in housing development that
concerns them. It involves how and why members of a community are brought into these affairs
(Chamber, 1995; 1997; Choguill, 1996). Public participation has been described as one of the
key elements of sustainable development (Roodt, 2001). The importance of including this
variable in the model is to determine the influence of beneficiary’s participation in public
housing occupants’ satisfaction and to ascertain if participation in the housing development
process will have positive spin-offs with regard to sustainable housing delivery (residential
satisfaction with the housing product). This is because government’s accountability can easily
be measured by the extent to which it involves beneficiary participation in decision-making
and having control over resources that affect their lives. This is one of the reasons why planners
and policy-makers should not ignore them in the development process (World Bank, 1996:121-
179).
469
The SEM results for this exogenous variable revealed that the standardized factor values and
interfactor correlations for the beneficiary participation latent factor were large and statistically
significant (Table 10.52). Inspection of the total variances accounted for in each measure by
the endogenous variable revealed that the scores were also significant. The relationship
between beneficiary’s participation indicators and residential satisfaction was found to be the
most significant amongst all latent variables. The parameter with the highest standardised
coefficient for this factor was the indicator variable BNP3. The parameter coefficient was
found to be 0.924. The indicator variable BNP3, which asked the beneficiaries of their level of
agreement, if owners should be consulted about the construction, was found to be mostly
associated with overall beneficiary residential satisfaction than all other indicator variables.
Thus, the overall results suggested that the influence of beneficiary participation in determining
beneficiaries overall satisfaction with their subsidised dwelling units is direct and statistically
more significant than any other factor.
Findings suggest that when beneficiary participation is incorporated into the housing
development process, the outcomes are more likely to suit local circumstances, ensure
community ‘ownership’, and increase the sustainability and eventually the satisfaction with the
housing development. Developing and maintaining the participation of beneficiaries can often
be a challenge requiring various strategies and considerations. However, participation can
encompass many activities. It can be beneficiary involvement in the initial planning stages of
a project, the development of action plans, or being a member of working groups, reference
groups and focus groups. It could mean receiving project updates in the form of a newsletter,
or providing reflections or feedback about the implementation of a project strategy from a
project recipient’s point of view. Most times, the promotion of beneficiary participation in both
project planning and implementation are implemented especially with regard to: project
location; type of land tenure; type and level of services; house design; position of the house
within the housing location; choice of material supply; and house construction methods, etc.
The SEM results advanced that the respondents participation in a housing development project
that concerns them, could potentially lead to the implementation of appropriate responses
through their being consulted about the housing location, house design, house construction and
selection of the internal finishes of the house, which can be incorporated into the development
process and eventually lead to their satisfaction and sustainability (continuity of what the
community has started) of the housing project.
470
The study findings did not concur with the work of Lizarralde and Massyn (2008) where it was
found that the performance (satisfaction) of low-cost housing projects does not depend on
community participation and that some of the mechanisms and advantages of community
participation need to be reconsidered in low-cost housing projects. Also, Davidson et al. (2007)
previously found that community participation can easily become rhetoric in its
implementation if it is not well guided. However, in the study done by Lizarralde and Massyn
(2008), it is further reported that community participation was wrongly implemented in the
reported case studies. In other words, the principle is ‘good’ but the implementation failed,
hence the statement that the performance of low-cost housing projects does not really depend
on community participation. In reality, the performance of low-cost housing projects depends
on a complex interaction of participants’ interests, objectives, resources and processes that go
beyond the benefits of the participation of the beneficiaries alone. Hence, it should be stated
that the participation of the beneficiaries is not positive; in fact it is crucial.
When the community meaningfully participates, they understand what the housing project
entails which limits misunderstandings with regards to the overall project aims. Hence, time
reduction occurs in explanations because people understand and know what is going on as they
now have a stake in the process. With meaningful beneficiary participation, the people take
responsibility for the project. However, the only way to encourage participation in housing
development is to educate the beneficiaries’ on various issues such as sustainability,
empowerment, capacity building, self-reliance and effectiveness. Government support is
critical in starting any form of beneficiary participation policies; which should be drafted to
create a regulatory framework and an enabling environment that facilitate active participation
(World Bank, 1996:121). In South Africa for instance, citizen participation is the principle
upon which democracy is founded, and it has been firmly entrenched by legal frameworks,
such as the White Paper on Reconstruction and Development (1994), the Constitution (1996),
the While Paper on Local Government (1998), the Municipal Structures Act (1998), the
Municipal Systems Act (2000), the Municipal Finance Management Act (2003) and the
Municipal Property Act (2004). Amongst these, the Municipal Systems Act (2000) stands out
as of particular relevance for promoting public participation, in that it requires municipalities
to develop a culture of public participation by building the capacity of local communities.
The SEM results thus suggest that when beneficiaries have control over resources affecting
their lives, it can lead to changes in knowledge and skill and their needs and expectations would
471
have been taken care of through their activate participation in the development process.
11.2.5 Needs and Expectation Influence on Residential Satisfaction
The finding was that housing needs and expectations have a direct positive influence on
beneficiary residential satisfaction. Beneficiary’s needs and expectations were defined by the
following indicator: owners should be told beforehand the type of house they will receive,
owners should be asked what type of house they need, owners expect good quality houses and
the houses should meet the needs of the family.
The findings suggested that beneficiary’s housing needs and expectations have a direct
influence on overall residential satisfaction. The finding was consistent with Hablemitoglu et
al.’s (2010) study, which found that meeting respondents’ housing needs set forth a dimension
of needs and the current satisfiers determine a set of requirements for their satisfaction.
According to Maslow’s (1980) Needs Theory, human needs are unlimited and when one of
them is met, another follows suit. In this process, complete satisfaction is not possible unless a
need classified to be important is first met. Individuals want what they do not have and the
need satisfied loses its motivating power. This is because housing occupants’ are only satisfied
when their current housing needs and expectations are satisfied. However, it must be noted that
residents’ satisfaction will not stay unchanged, because soon, there will be other higher order
needs and expectations that will have to be satisfied. More so, households that are dissatisfied
are likely to consider some form of adjustment. Morris and Winter (1975) and Hamnett (2001)
inform that residents may attempt to make adjustments to reduce dissatisfaction by revising
their needs and expectations to reconcile their incongruity, or by improving their housing
conditions through remodeling. They may also move to another place to bring their housing
into conformity with their needs and expectations. However, both mobility and adjustments are
subject to the constraints posed by financial resources at one’s disposal and by information
regarding alternative adaptation opportunities.
The most significant findings from the SEM results highlighted the fact that owners should be
told beforehand the type of house they will receive and should also be asked what type of house
they need. This significance is highlighted by the fact that the gratifications of occupants
housing needs and expectations should have noteworthy prominence. Because people with
different housing needs and expectations, the same housing condition could bring different
satisfaction levels because their needs and expectations are different. Hence, unless the level
472
one need is sufficiently satisfied, they will remain in the occupant’s consciousness and will
thus become the prime determinants of housing behaviour. In earnest, the living condition that
is currently pursued forms the housing expectation of the individual, which is highly related to
the overall residential satisfaction. The current study finding on this aspect concur with
Marcus’s (1995) study which found that housing is like a mirror, which has a powerful effect
on our sojourn toward a state of wholeness (satisfaction). Hence, all over the world, and in
South Africa, there is a growing consensus that meeting residents’ housing needs and
expectations constitutes a significant new dimension of community development and
establishment. This is because housing needs as a shelter are mostly a concern for those who
struggle for these needs, such as the homeless, those previously disadvantaged from owning
property, as a result of government policy of the past.
Likewise, previous research (Caughey et al., 1998) has shown that expectations have a
significant effect on overall satisfaction of housing occupants. This is because satisfaction
normally occurs based on a comparison of that which is expected with that which is received.
Similarly, prior exposure to that which is to be received has the tendency to influence
occupant’s satisfaction towards a housing product. While a negative prior experience can
generate a lower expectation, which will result in lower satisfaction. Further findings revealed
that when beneficiaries of the housing units have an expectation of what they will receive, they
will either be satisfied based on the expected outcome or be dissatisfied. Satisfaction with what
is expected suggests that satisfaction is the result of a comparison of that which was expected
and what was received (Caughey et al., 1998; Woodruff et al., 1983). Also, Tse and Wilston
(1998) posit that a fundamental premise of dissatisfaction with prior exposure (expectation) is
that expectation is related to satisfaction. The result also suggests that in addition to the
influences from expected performance and subjective dissatisfaction, perceived performance
exerts direct influence on satisfaction. Therefore, it can be asserted that when beneficiaries are
dissatisfied with what has been received it is in response to the congruency between their
expectations and the actual performance of the housing product that was received.
Hence, subsidised public housing beneficiaries satisfaction may be viewed as a function of the
interrelationship between what beneficiaries expect from the government in relation to their
housing need and their perceptions of the houses they have received (i.e. the quality of the
houses received and the satisfaction derived from the housing meeting their needs). It should
473
be noted that satisfaction is not the only direct outcome, but prior exposure to what is to be
received have also been found to directly affect dissatisfaction.
11.2.6 Service Provided by Government Influence on Residential Satisfaction
The finding was that services provided by the government had a direct positive influence on
the satisfaction level of the residents of government subsidised housing. Services provided by
the government (SPG) was defined by four indicator variables: satisfaction with how residents’
complaints are handled, government response to building defects, enforcement of rules by the
Department of Human Settlement and the overall services provided by the government.
The findings suggested that SPG have a direct positive influence on the beneficiary’s overall
residential satisfaction. The finding was consistent with that of Salleh et al. (2011) and
Ahlbrandt and Brophy (1976) who found that residents’ satisfaction had a direct positive link
with the services provided by the management (government) of a public housing. In other
words, the findings revealed that improved services by the government being the managers
(administrators) of subsidised low income housing in South will influence and increase
residents’ overall satisfaction. This is apparent from the standardized factor loadings and
interfactor correlations with other indicator variables, which were statistically significant
(Table 89). Also, the variances accounted for in each measure by the endogenous variable
revealed that the scores were statistically significance and the values were above the minimum
required value of 25% to be ascribed an influence on residents’ satisfaction with their housing
units.
The findings of this study corresponds with the work of Husna and Nurizan (1987) and Salleh
(2011) which find that most residents of a low-income public housing in Kuala Lumpur were
satisfied with the services offered to them by the government. Findings from the current study
strongly correspond with the aspect of residents’ concern with the time taken by management
to respond to their complaints; response to building defects and the enforcement of rules by the
management as found in the previous studies (Ukoha & Beamish, 1997; Husna & Nurizan,
1987; Salleh et al., 2011). Similarly, Varady and Carroza (2002) and Ukoha and Beamish
(1997) also reported corresponding results where they find that the most causative factor of
residents’ satisfaction regarding management services were housing management’s prompt
feedback to the residents, response to building defects and the enforcement of rules by
management. Therefore, the finding was significant in that it provides the South African
474
Department of Human Settlement and other stakeholders of subsidised low-income housing in
the country with knowledge of the fact that the services they provide to the low-income
beneficiaries influence their satisfaction with the housing units. This is not only in a positive
sense, but also in a negative way. For instance, from the descriptive survey results, the
respondents informed that they have shopping facilities; place of worship, drainage system
(within and outside the neighbourhood) within their reach as well as having access to garbage
and waste collection.
Similarly, the findings are significant because subsidised low-income housing stakeholders
could constitute a feasible checklist of services and features that could be provided in low-
income locations, in order to ensure the residents’ satisfaction with the houses. Not that there
are no existing guidelines, but a feasible level of services to be provided should be specifically
determined before the construction of the houses to make the projects an integrated
development. Also, by using the checklist, the stakeholders would have adequate knowledge
of services that have to be provided in order to ensure an acceptable level of services in the
housing locations. These services to be provided could also constitute leading indicators for all
stakeholders involved in low-income housing projects either subsidised or in private
developments.
11.2.7 Extent the Hypothesised Integrated Model Fit the Identified Factors
The SEM finding is that five of the identified factors (exogenous variables) had a direct positive
influence, while only one had an indirect influence in determining residential satisfaction. The
general hypothesis that residents’ overall residential satisfaction with their housing product is
directly related to the influence of the exogenous variables in predicting overall beneficiary
satisfaction in publicly funded housing schemes in developing countries using South Africa as
a case study is sustained. The findings supported previous research studies which informed that
housing satisfaction is a multidimensional construct (Amerigo, 1990; Amole, 2009; Amerigo
& Aragones, 1990; Campbell et al., 1976; Francescato et al., 1987; Hourihan, 1984; Lu, 1999;
Marans & Rodger, 1975; Marans & Sprecklemeyer, 1981; Michelson, 1977; Weidemann &
Anderson, 1985). Therefore, Lu (1999) specifically informs that the formation of residential
satisfaction is not simply based upon freedom from dissatisfaction; it is more complex.
The postulated relationships between all exogenous factors and the endogenous factor were
found to be statistically significant. The result suggested that all latent factors adequately
475
measured the overall resident satisfaction. The relationship between residential satisfaction and
beneficiary participation was found to be the most significant. This suggested that when
beneficiaries are made to participate in the housing development process, their eventual
satisfaction with the project will be sustained and hence, develop their capacity to further
participate in future development projects as advanced by previous researches (Chamber, 1995;
1997; Choguill, 1996; Lizarralde & Massyn, 2008).
Findings on the outcome variables of residential satisfaction for the study reveal that the direct
and the interfactor influence of the latent variables will manifest in an objective and subjective
behaviour from the beneficiaries such that they will be:
satisfied living here (in the housing unit and neighbourhood);
will take proper care of their neighbourhood; and
will recommend to their friends to obtain a house in the same way that they did.
Findings also revealed that the outcome of residential satisfaction will not influence the
beneficiaries not to move to another place in the future. This was consistent with the housing
adjustment theory as proposed by Morris and Winter (1976), which informs that if a
household’s current housing meets the norms, the household is likely to express a high level of
satisfaction with housing and neighbourhood. However, an incongruity between the actual
housing situation and the housing norm results in a housing deficit, which in turn gives rise to
residential dissatisfaction. Therefore, households with a housing deficit who are dissatisfied
are likely to consider some form of housing adjustment to meet the known norm. They may
attempt to make adjustments to reduce dissatisfaction by revising their needs and expectations
to reconcile the incongruity, or by improving their housing conditions through remodeling (Lu,
1999). According to Morris and Winter (1976), they may also move to another place to bring
their housing into conformity with their needs and expectations. Though, both mobility and
adjustments are subject to the constraints posed by financial resources at one’s disposal and by
information regarding alternative adaptation opportunities (Lu, 1999). Thus moving behaviour
is only one type of adjustment residents perform during the time of dissatisfaction of housing
needs; but in the case of the low-income group, it might not be possible, as most cannot access
housing on their own and the subsidised houses received might be their only life time
opportunity to access housing.
476
11.3 QUESTIONNAIRE AND DELPHI SURVEY RESULTS
Findings from the Delphi Study were that the factors considered to be the principal
determinants of residential satisfaction, include dwelling unit features, neighbourhood features,
building quality, services provided by government, beneficiary participation, needs and
expectations. The results state that beneficiaries are very likely to be satisfied with their housing
units with the influence of the identified factors.
The Delphi Study was validated by a Field Questionnaire Survey. The results suggested that,
the identified factors from the Delphi Study have direct and indirect influence in determining
residential satisfaction. In the questionnaire survey, the hypothesis that the exogenous factors
had a direct and positive influence on beneficiaries’ residential satisfaction could not be
rejected. The exogenous variables: dwelling unit features, neighbourhood features, building
quality / housing condition, services provided by government, beneficiary participation and
needs and expectations were found to have a statistically significant influence in predicting
resident satisfaction. The findings from both the Delphi and the Questionnaire Survey
therefore, suggested that the exogenous variables influenced the determination of the the
endogenous variable (occupants’ residential satisfaction).
The merit of using Structural Equation Modeling to validate the Delphi findings was that it was
possible to specifically ascertain which of the exogenous factors had significant influence on
residents’ satisfaction. Therefore, instead of making a general blanket statement that the
exogenous variables had influence on determining the beneficiary’s residential satisfaction, it
was possible to precisely state that the factors: dwelling unit features, building quality, services
provided by government, beneficiary participation, needs and expectations had a direct
/(stronger) statistically significant influence on the beneficiaries, while neighbourhood features
had an indirect (weak) influence in determining residents’ satisfaction with their housing units.
Beneficiary participation has been shown to exert a more profound influence on residential
satisfaction. It appears to enhance residential satisfaction significantly, based on the SEM
modeling results of this study. Needs and expectations were also strongly associated with
increased satisfaction.
11.4 CONCLUSION
In conclusion, the findings from the Questionnaire Survey generally supported the predictions
477
that were made by the experts from the Delphi Study. The validated predictions were that
dwelling unit features, neighbourhood features, building quality / housing condition, services
provided by government, beneficiary participation, needs and expectations influence the
prediction of public housing occupants’ residential satisfaction. In addition, the existing
literature lends support to the findings of the current study. The supported findings were that
dwelling unit features, neighbourhood features, building quality / housing condition, services
provided by government, beneficiary participation and needs and expectations, are
fundamental to subsidised housing scheme residents’ satisfaction with their dwelling units.
478
CHAPTER TWELVE
CONCLUSIONS
12.1 INTRODUCTION
The general overall objective of the current study was to develop residential satisfaction model
for publicly funded subsidised low-income housing in developing countries using South Africa
as a cases study, and to specifically identify the determinant attributes which collectively
predict residents’ satisfaction with their low-income houses.
Hence, in order to achieve the general objective, the study adopted a Mixed Methodology of
conducting an extensive literature review, a Delphi Study and a Field Questionnaire Survey
which was analysed using Structural Equation Modelling. The Field Questionnaire Survey was
conducted in order to validate findings from the Delphi Study with regards to the factors which
predict residents’ satisfaction. Conclusions regarding the study are presented relative to the
objectives of the study in the next sections.
12.1.1 Research Objective RO1
The first objective of the study was to establish the factors that determine residential
satisfaction in low-income housing, based on a literature review. In order to achieve this
objective, a review of literature was conducted. Findings are that residents’ satisfaction with
their houses is not a product of only one attributed, but a multi-faceted construct. Further
findings revealed that residential satisfaction has been a major topic in various disciplines, such
as sociology, psychology, planning and geography. In addition, it is found that an
understanding of the low-income’s satisfactory evaluation with their housing product will bring
improvements, which could thus be employed for future developments, which will improve the
effectiveness of the service provision. Also, it is found that residential satisfaction research
deals with the housing products’ consumer satisfaction, and aims to inform policy and planning
intervention.
The literature also informs that residential satisfaction is recognized as an important component
of an individual’s general quality of life; arguing that for most people, housing is the largest
consumption item in their lifetime because the house when transformed to become a home is
479
the setting where one finds refuge, rest and satisfaction. Therefore, pointing out that the degree
to which individuals’ needs and aspirations are met by their housing condition is a concern for
researchers but most importantly for housing developers, planners and the Departments of
Human Settlement. Similarly, it is found that the factor which brings about residential
satisfaction encompasses satisfaction with the dwelling unit, satisfaction with the dwelling
neighbourhood and the entire neighbourhood, amongst others. Findings from the literature
were that more research and effort is required to try and address the problem of residents’
satisfaction with their dwelling units.
12.1.2 Research Objective RO2
The second objective of the research was to establish the current theories and literature that
have been advanced on residential satisfaction and to identify the gaps that needed
consideration. A review of literature was carried out to achieve this objective. The findings
revealed that residential satisfaction research has not been studied with an all-inclusive
construct in the development of the previous models and theories. The identified gaps from the
extensive literature review were beneficiaries’ meaningful participation in the housing process,
in terms of their needs and expectations. The identified gaps formed the new constructs in the
current study conceptual framework (Model 1.0). These gaps were considered essential
because people have different housing needs and expectations that cannot be satisfied with the
same housing condition; and this could bring about different satisfaction levels because their
needs and expectations are different. Also, their participation in the housing process will bring
about ownership of the project and enable them to contribute to the process, which will thus
bring about sustainability of the project. The current study offers a synthesized classification
of the constructs, which should be collective considered to predict residents’ satisfaction. From
the synthesized literature, this current study argues that residential satisfaction is a six-factor
construct.
12.1.3 Research Objective RO3 & RO4
The third objective of the study was to determine the main and sub-attributes that bring about
residential satisfaction and to examine if the attribute that determines satisfaction in other
cultural contexts is the same within South Africa. The fourth research objective (RO4) was to
evaluate the critical factors and issues that affect the delivery of low-income housing in South
Africa.
480
A Delphi Study was conducted in order to achieve these objectives. Findings were that a
number of factors that were considered to be important in determining residential satisfaction
were identified and amplified by the Delphi Study. The factors considered to be paramount
determinants of residential satisfaction were dwelling unit features, neighbourhood features,
building quality, services provided by government, beneficiary participation and needs and
expectations. The findings suggested that the attributes that brings about residential satisfaction
in South African low-income housing are similar to the determinants in other cultural contexts.
Further, residential satisfaction is assured if there is a consideration of these factors in the
development of subsidised low-income housing for the low income groups in South Africa.
These factors were collectively considered for the development of the all-inclusive (integrated
holistic) residential satisfaction model.
Findings further reveal that the daunting critical factors, which affects, low-income housing
provision in South Africa, which is equally applicable to other developing countries is financial
limitation caused by the dwindling tax base. Also, it is revealed that waiting time on the housing
database impacts on the delivery of low-income housing and in turn affects the housing
satisfaction of the low-income groups. Also, it was found that developmental efforts and
initiatives should be directed not just at achieving the set objectives for the well-being of the
low-income groups or communities but also that the gains should be retained and nurtured to
greater levels. The findings also reveal that beneficiaries of the housing development must be
capacitated or assisted/supported to take charge of the processes and the results of developmental
interventions (participation). Because for the beneficiaries of public housing development to
continue realizing proceeds of the developmental processes, they must be guided to determine both
process, as well as end products of development, which will in turn bring about their satisfaction
with the housing development project.
12.1.4 Research Objective RO5
The fifth research objective of the study was to develop an integrated residential satisfaction
model for subsidised low-income groups based on both literature and the Delphi Study.
A synthesis of the reviewed literature together with the findings from the Delphi Study was
used to achieve this objective. The conceptual model theorized that subsidised occupants’
residential satisfaction is a six-factor construct. These factors were: dwelling unit features,
neighbourhood features, building quality features, services provided by the government,
481
beneficiary participation, needs and expectation, which jointly predict residential satisfaction
of occupants in subsidised low-income housing. This hypothesis was validated through a
Structural Equation Modeling of data from the Field Questionnaire Survey.
12.1.5 Research Objective RO6
The sixth and final research objective of the study was to test and validate the conceptually
integrated residential satisfaction model by conducting a Questionnaire Survey and analysing
it using Structural Equation Modeling (SEM). A Questionnaire Survey and analysis of the
results using SEM software, EQS Version 6.2 was conducted in order to achieve this objective.
Findings from the SEM analysis, which model subsidised housing occupants’ residential
satisfaction as a six-factor model showed that the factors of dwelling unit features,
neighbourhood features, services provided by the government, building quality features,
beneficiary participation and needs and expectations were found to have a significant influence
in determine the occupants’ residential satisfaction. However, the neighbourhood feature had
a weak (indirect) influence in predicting residential satisfaction, nevertheless, it covariances
with the other exogenous construct to determine residential satisfaction, which was found to
be statistically significant. These findings validated the conceptually integrated holistic model
developed from literature and the Delphi Study.
12.2 CONTRIBUTION AND VALUE OF THE RESEARCH
The value and contribution of the current research is described at three levels. These are the
theoretical, methodological and practical levels of the research findings. However it is pertinent
to note that the outstanding contribution of the study is the revelation and validation of the
influences of beneficiary participation, needs and expectations in predicting residents’
satisfaction.
12.2.1 Theoretical Contribution and Value
The results of the SEM analysis indicated that subsidised low-income housing occupants’
residential satisfaction is a six-factor model. The researcher could not find evidence of a similar
study that has been conducted in the low-income housing context. The study is also significant
because it addresses the lack of theoretical information about which factors are most significant
in predicting resident satisfaction in subsidised low-income housing.
482
The SEM results also indicated that the factors of dwelling unit features, neighbourhood
features, services provided by the government, building quality features, beneficiary
participation, needs and expectation were found to have a significant influence in determining
occupants’ residential satisfaction. Only one had a weak (indirect) influence in determining
residential satisfaction in subsidised low-income housing in South Africa. The findings
enforced the theory that low-income housing occupants’ satisfaction is multifaceted. The SEM
findings also show that the latent variables lead to residential satisfaction outcome variables,
which could be used for residential satisfaction measurement.
Also, the literature review did not reveal evidence of a similar study to the current one and
therefore, suggested that this type of research has not yet been conducted in housing studies,
and especially, in South Africa. Moreover, there was no evidence that suggested that a Mixed
Method of using Delphi and SEM had been used in housing studies in South Africa. Therefore,
this study may offer a base for other researchers to use for other follow-up studies. Likewise,
the current study modeled residential satisfaction as a six-factor construct with the inclusion of
two new variables: beneficiary participation and their needs and expectations. Previous studies
have tried to model satisfaction using other variables without the inclusion of these two
additional constructs. This study has shown that there is more than one factor that can influence
residents’ satisfaction with their dwelling units and most especially, the new additions should
always be considered as beneficiaries’ meaningful participation in the housing process and the
incorporation of their needs assessments and expectations will predict the resultant output of
the housing process. Apart from the study contributing to theoretical knowledge, it also
contributed to methodological advance in terms of the approach used in conducting the
research.
12.2.2 Methodological Contribution and Value
Most studies have used univariate statistical methods such as ANOVA, MANOVA or
Regression Modeling to model residential satisfaction. However, the current study used SEM,
which is more robust and superior to the methods mentioned to determine causality of factors
in a model and their direction of influence (Kline, 2005; Musonda, 2012:252). With SEM
analysis, it was possible to identify the factors of residential satisfaction, which had significant
effect and hence influence residents’ satisfaction with their houses as opposed to a general
blanket statement that there are numerous constructs, which influence residents’ satisfaction.
483
The Questionnaire Survey instrument had high internal reliability values and therefore, could
be used in similar studies to validate the current study or for similar purposes. Findings from
the Delphi Study and the conceptual model developed from both the literature review and the
Delphi Study was validated by conducting a Questionnaire Survey. Hence, data from the
questionnaire survey was analysed using SEM software, EQS Version 6.2. As a result of this
Mixed Method, a parsimonious model was developed. Aside from this contribution and value
to the body of knowledge in terms of the methodological approach, a contribution to practice
and the housing industry was also achieved.
12.2.3 Practical Contribution and Value
The significance of citizen participation on developmental projects has been well expanded in
literature, likewise on the People’s Housing Project in programmes in South Africa. Little has
been report on its relevance to subsidised housing in South Africa. However, Delphi results
have indicated that beneficiaries of these housing projects will be better served if they are made
to participate in the housing process. Further, SEM results indicated that beneficiary
participation has direct significant influence on overall beneficiary satisfaction with the
housing units. Similarly, the influence of beneficiaries’ needs assessment and expectations
before the construction of the houses thus significantly influence their perception towards the
finished product.
Besides, the knowledge of the influence of the six-factor construct could help subsidised low-
income housing development stakeholders to plan, organize, coordinate and control all aspects
relating to the housing development. The Department of Human Settlement (DHS) could use
this knowledge to help with decisions on how to best allocate finances towards the development
of an inclusive low-income settlement. The practical significance of the study are further
elaborated as follows:
Significance to Planning
The provision of adequate housing in planning is important as it contributes towards a quality
living environment and directly supports the concept of sustainable development. Studying
housing satisfaction of low income urban dwellers has provided information on the deficiencies
of their housing units, facilities and the housing environment. This information will be useful
in planning future housing in order to ensure a better living environment particularly of the low
income groups, which has always been the marginalized group.
484
Significance to Community
The objective of good development is to create a more just and united society besides
maintaining social stability and effective economic management; as well as to changes lives.
The Delphi and SEM results have shown that beneficiary satisfaction in housing means the
fulfillment of housing needs. By so doing, the welfare of the low-income community is taken
care of and improvement can be done to upgrade their quality of life. In return, the low-income
groups will be satisfied and perform better in their work and social life, as they will not feel
excluded from the development of the society.
Significance to the Department of Human Settlement
The study results have also demonstrated the level of satisfaction of beneficiaries towards the
current condition of their low-income houses. The output of the study will help the Department
of Human Settlement in making decisions about the criteria to be given priority in providing
good, quality housing. The findings will help DHS to plan programmes for the housing
communities, as well as planning effective housing management and deployment processes.
The output of the study will further help in providing feedback on housing design and will
assist DHS in future housing decisions and policies. As affirmed by Amerigo & Aragone
(1997), beneficiaries’ housing satisfaction is important because it broadens the understanding
of how and why beneficiaries respond to certain factors in the environment in which they live
as well as to ascertain housing types and living conditions. This will ultimately enable DHS to
know the vital areas to commit resources to, so that the quality of life and well-being of the
disadvantaged and low-income groups can be assured. In order to ensure that housing is not
just a home but a home in livable neighbourhood, the government should monitor low-income
housing programmes to ensure the needs of the low-income groups are met.
The integrated holistic residential satisfaction model should be used as a guide to ensure that
all elements necessary for an all-inclusive development is in place to ensure an acceptable low-
income housing standard. The study offers an opportunity for further research to improve the
model developed in this study and probably refine indicator variables to suit specific
environments. Therefore the recommendations and policy implications for practice of all these
areas in which the current study may add value and contribute are presented below.
485
12.3 RECOMMENDATIONS
Recommendations are made from the methodological, theoretical and practical points of view.
12.3.1 Methodological
It is recommended that a similar study should be conducted with a different populations and
samples (in other metropolitan municipalities) to improve the application in the South Africa
low-income housing space. Also, further research should be conducted on the indicator
variables to establish any improvement in model fit, as the current study was purely a
confirmatory factor analysis. There is the possibility that residential satisfaction could be
defined by more indicator variables. Recognition should be made however that there is no such
a thing as a perfect model. However, there should be a move to try and improve on the current
model rather than invent a new model.
Findings from the current study recommend that the Mixed Method of using a Delphi and a
Questionnaire Survey be encouraged in studies, such as the current one where a test – retest
methodology may not be feasible to validate a study. This situation is common in most social
science studies and most studies end at Questionnaire Survey or Delphi Study and as such
renders generality of conclusions especially on causality to be questionable. The recommended
method could commence with a Delphi Study followed by a questionnaire survey or vice-versa
in order to validate a study and therefore, improve its generalisability.
Previous research studies in the social science and most especially in housing studies try to
establish cause and effect relationships between different latent variables. But, most of these
studies use inadequate analytical methods, such as ANOVA and Multiple Regressions.
ANOVA or MANOVA, which are basically standard statistical procedures, do not offer an
appropriate and a straightforward way to test a hypothesis at a higher level of abstraction.
Therefore, for similar studies, such as the current thesis, Structural Equation Modeling is
recommended to be used as the analysis technique for better results and abstraction.
12.3.2 Theoretical
It was observed from the literature that there were still different definitions and an
understanding of how residents’ satisfaction is formed. This has led in the past to a limited
view and narrow conceptualization of residential satisfaction. Besides, there has not been a
486
consensus on how residential satisfaction in subsidised low-income housing in South Africa
should be measured. However, in the current study, literature was reviewed and synthesized on
the determinants of residential satisfaction. In conjunction with the experts’ knowledge
obtained through the Delphi Study, a six-factor residential satisfaction model was arrived at for
subsidised low-income housing. These factors were identified as dwelling unit features,
neighbourhood features, building quality features, services provided by the government,
beneficiary participation, needs and expectations. It is therefore recommended that the
developed model and theory of residential satisfaction, with particular emphasis on
operationalization it, should form the basis for further refinement of the concept and thereby
making it beneficial to the South African low-income housing space and other developing
countries. It is further recommended that the influence of beneficiary participation, needs and
expectations should be integrated into existing models, as proposed in other studies that have
been developed.
12.3.3 Policy Implication and Practical Recommendation
As a result of the identified contributions that the current study makes, as revealed by the
findings, the following policy implications and practical recommendations have been
identified:
The policy implication suggests that residential satisfaction of beneficiaries of publicly
funded low-income housing can be enhanced through the improvement of the dwelling
unit features, neighbourhood features, building quality features, services provided by
government, involving the beneficiary in the housing development process and the
assessment of the beneficiaries’ needs and expectations.
Also, the DHS and other stakeholders responsible for low-income housing provision in
South Africa and other developing countries can adopt proper management measures
in order to improve the residents’ housing environment and the resulting dwelling units.
Most importantly, the neighbourhood features, as findings indicated that this aspect was
considered to have a weak impact in predicting residents’ satisfaction.
Likewise, the future of subsidised low-income housing in South Africa should be
responsive to the six-factor model and especially to beneficiary participation and the
assessment of their needs, as these are considered vital in the total housing provision.
Thus, the development of low-income housing projects should take into account the
needs of the residents more than their effective demand for a house.
487
Another policy implication is that stakeholders planning subsidised low-income
housing development should embrace the integrated housing development option, as
contained in the Breaking New Ground Housing Policy to enable the creation of an all-
inclusive neighbourhood that will meet the wide needs of the low-income group.
Similarly, low-income housing development should be integrated with other urban land
use so that an efficient social infrastructure provision system could be effectively
implemented within the framework of public housing delivery system, in order to
enhance the qualitative adequacy of public housing in South Africa.
Another policy implication of the study is that the DHS should adopt the criteria of
family size projection (housing life cycle) in determining the low-income house size,
particularly, the number of bedrooms.
Also, future low-income housing design should be responsive to low-income residents’
need for safety, security, thermal comfort, and job creation amongst others. This is
because the respondents’ socio-demographic findings reveal that most of the
respondents’ are unemployed, lacked education and desired better security protection.
12.4 LIMITATIONS
Interesting and valuable findings have emerged from this study, however, the following
limitations regarding the current study should be considered. Firstly, the research was only
conducted in the Gauteng Province of South Africa as indicated in Chapter 1, Section 1.3.5.
This is because the Gauteng Province has delivered more subsidised low-income housing in
South Africa. Given enough resources, it would be preferable to conduct a similar research
study with the entire metropolitan and district municipalities in South Africa where subsidised
low-income houses have been built. Also, the consideration of other developing countries could
be included. Secondly, the ethnic representation in the study was not comparable: African
(85.89%), Indian (80%), Coloured (10.39%) and White (2.93%). However, it does sufficiently
reflect the representation of publicly subsidised low-income housing population in South
Africa. Thirdly, the SEM Methodology used in data analysis may be construed as a limitation.
The results presented herein are based on the analysis of a causal model with raw data. Hence,
the results are intended to support the priori causal model. Thirdly, the use of additional items
or constructs might improve the inherent reliability and validity of the measures used. Fourthly,
several nested models especially for the measurement models, could have been evaluated to
check out the suitability of other alterative models. The current study was purely confirmatory
488
in nature. Fifthly, although the internal reliability tests indicated high internal consistency and
therefore a well-constructed research tool, some constructs revealed high correlational values.
This may be due to the fact that only one questionnaire was used to collect information in all
the low-income housing locations. A review of the research tool would have benefited findings
in this study. A final limitation is related to the sample, in addition to the aforementioned
limitations the study has shown that some of the SEM measures may have been influenced by
the sample size of the study. All empirical studies are limited by the nature of the sample
studied. The exploration of the dependent variable (residential satisfaction), has shown that it
has a very complex organisation (multi-faceted), and claims for further interpretations.
12.5 RECOMMENDATIONS FOR FURTHER RESEARCH
The following suggestions for further studies have been identified:
Further studies should examine factors related to the limitations of the current study.
Firstly, more rigorous and detailed testing of measurement scales in South African and
other developing countries would further the knowledge of low-income residents’
satisfaction. It is possible that some scales developed in Western culture and in the
current study may not be suitable for other cultural contexts.
The model did not include culture as a separate construct, and including this may have
influenced the research results. Although, inherent in the current study is the notion that
individuals make their decisions based on their cultural values. However, it is
recommended that future studies should consider culture as a separate construct to
confirm the influence of the ‘low-income culture’ on their satisfaction level with the
subsidised houses, as the houses are not culturally oriented, but based on a premeditated
design, as approved by the Department of Human Settlement.
These results also need to be replicated with other populations. Important features of
residential satisfaction may vary between different regions and cultures, while some
might remain in common. Equally important are youths and the elderly that may have
different perspectives on the provided houses, and those groups of individual
characteristics were not considered in the conceptualization and validation of the
current model.
A validation of the Aigbavboa Integrated Holistic Residential Satisfaction Model
presented in Figure 10.10 (Model 2.0) is recommended.
489
12.6 CONCLUSION
An integrated residential satisfaction model for subsidised low-income residents’ was
developed using existing residential satisfaction and other theories grounded in housing
studies. It was postulated that overall subsidised low-income residents’ satisfaction is directly
related to the influence of the exogenous (latent) variables’ in predicting / determining overall
housing satisfaction. The postulated model was analysed with the use of Structural Equation
Modeling software EQS Version 6.2. The fit statistics for the measurement and structural
models had an adequate fit to the sample data. The finalized empirical model revealed that the
exogenous variables (dwelling unit features, neighbourhood features, building quality features,
services provided by government, beneficiary participation, needs and expectations) had a
statistically significant influence in determining subsidised low-income housing occupants’
satisfaction. Specifically, the exogenous variables such as: neighbourhood features had a weak
(indirect) influence on determining residential satisfaction in subsidised low-income housing
in South Africa. Adhering to this caveat, it is therefore concluded that the six-factor model
schematically portrayed in Figure 10.10 and 10.11 (Model 2.0) represents an adequate
description of residential satisfaction in subsidised low-income housing in South Africa.
The results of this study have theoretical, methodological and policy (practical) values because
respondents for the Delphi Study were drawn from academics’, housing practitioners and the
Department of Human Settlement personnel’s. While the respondents for the Questionnaire
Survey was the low-income housing occupants. Furthermore, the respondents had a good
working knowledge of the studied environment. In addition, the Questionnaire Survey whose
results were modeled using Structural Equation Modeling was a validating study of a
conceptual model developed from synthesized theories established from literature and more
importantly from the Delphi Study. Hence, it is considered that the presented model for
subsidised low-income housing satisfaction interpretation maintains its validity.
The result of the study provided information that can inform governmental, corporate,
institutional and community policy-makers, as they plan for and implement subsidised housing
programmes designed to enhance the quality of life of the poor and low-income groups.
Secondly, the study provides indicators that will be a baseline for assessing low-income
housing in developing countries. This is because an increase in housing satisfaction improves
people’s quality of life, thus, directly affecting people’s satisfaction with their lives. Housing
490
that is satisfactory and pleasant for people increases their self-worth and personal fulfillment,
and helps them to be successful in life. For this reason, the results of this study should constitute
a reference guide in developing countries low-income housing policies, and the factors that
increase user satisfaction should be taken into consideration in future planning. Consequently,
housing planners, designers and other stakeholders will be able to contribute to the ways of
solution to increase the low-income groups’ quality of life and level of satisfaction by carefully
considering the factors that determine residents’ satisfaction in housing. Stakeholders and
institutions who are involved in the planning process should wield the contemporary factors
revealing user-preferences about housing satisfaction, as part of planning input so as to increase
the level of user satisfaction. As a result, the public’s (low-income) requirements and
expectations will have been taken into consideration, and members of the public (the poor and
the low-income) will be able to participate in the planning process. Also, the validated
conceptual model of occupants’ satisfaction in housing, which has been formulated in this
study, will provide a reference to researchers who will study housing satisfaction in the future.
The current study, lends support to other studies that have utilised alternative methods to
establish the factors which influences low-income groups residential satisfaction with their
houses. These studies have concluded that low-income residents’ housing satisfaction is multi-
faceted as also claimed in the current study. The current study utilizes a more robust modeling
method of SEM. By adopting the methodology, the current study was able to model the
influence of the selected multi-faceted variables and the constructs, which were statistically
significant. The practical implication is that residential satisfaction of beneficiaries of publicly
funded low-income housing schemes can be enhanced through improving the dwelling unit
features, neighbourhood features, building quality features, services provided by government,
encouraging their active involvement in the housing development process and undertaking a
careful assessment of the beneficiaries’ needs and expectations. Also, the DHS and other
stakeholders responsible for low-income housing provision in South Africa can adopt proper
management measures in order to improve the residents’ neighbourhood environment and the
dwelling unit features. Most importantly, the neighbourhood features, as findings indicated that
this aspect was considered to have a weak impact in predicting residents’ satisfaction.
Likewise, the future of subsidised low-income housing in South Africa should be responsive
to the six-factor model and especially, to the beneficiaries’ participation and the assessment of
their needs, as these are considered vital in total housing provision. Thus, the development of
491
low-income housing projects should take into account the needs of the residents more than their
effective demand for housing.
492
REFERENCES
Aaker, A., Kumar, V. & George, S. (2009). Marketing research. 10 edn. New York: John Wiley
and Sons, Inc.
Abbott, J. (1996). Sharing the city: community participation in urban management. London:
Earthscan.
Abdul Ghani, S. (2008). Neighbourhood factors in private low cost housing in Malaysia.
Habitat International, Vol. 32, No. 4. (December 2008), pp. 485-493
Abdul, G.S. & Yusof, N.A. (2006). Residential satisfaction in low-cost housing in Malaysia.
Report of Research. 1-69.
Abrams, C. (1971). The Language of Cities: A Glossary of Terms. New York: The Viking
Press.
Adam, J. (1984). The meaning of housing. America Annals of the Association of American
Geographers, 74(4):515-526.
Addison, T. (2003). E-commerce project development risks: Evidence from a Delphi survey.
International Journal of Information Management, 23(1):25-40.
Ademiluyi, A.I. & Raji, B.A. (2008). Public and private developers as agents in urban housing
delivery in sub-Saharan Africa: The situation in Lagos state. Humanity & social Sciences
Journal, 3(2):143-150.
Ademiluyi, I.A. (2010). Public housing delivery strategies in Nigeria: A historical perspective
of policies and programmes. Journal of Sustainable Development in Africa, 12(6):153-
161.
Adeniyi, E.O. (1985). Housing in Nigeria’s national development in Housing in Nigeria: A
book of readings., A.G. Onibokun (ed), Niser: Ibadan, pp. 80093.
Adeyemo, J. & Dekolo, T. (2000). Housing situation of the poor in Lagos metropolis. The
professional builder, (June/July):20-30.
Adisa, A.L., Agunbiade, O.M. & Akanmu, O.E. (2008). House ownership as a well-being index
among retirees in Osun state, Nigeria. Journal of international social research, 1(5):30-
46.
Adler, M. & Ziglio, E. (1996). Gazing into the Oracle: The Delphi method and its application
to social policy and public health. London: Kingsley Publishers.
Agbola, T. (1998). The housing of Nigerians: A review of policy development and
implementation. Ibadan, Nigeria: Development Policy Centre.
493
Agbola, T. & Alabi, M. (2000). Sustainable Housing Delivery, Lesson from International
Experience. National workshop on Sustainable Housing Delivery in Nigeria; Challenges
for public/private partnership, 5-7 June 2000 2000 Sheraton Hotel Abuja, Nigeria.
Agbola, T. & Jinadu, A.M. (1997). Forced eviction and forced relocation in Nigeria: The
experience of those evicted from Maroko in 1990. Environment and urbanization,
9(2):271-288.
Agyemang, K.K. (2001). The Political Economy of Housing and Urban Development in Africa-
Ghana‘s Experience from Colonial Times to 1998. Westport, UK: Praegers Publishers.
Ahadzie, D. K. (2008). Management practices in Ghanaian house building industry.
Unpublished material.
Ahmad, H.H. (1994). Residential satisfaction and social integration in public low cost housing
in Malaysia. Birmingham: University of Birmingham.
Ahmed, I. (2011). An overview of post-disaster permanent housing reconstruction in
developing countries. International Journal of Disaster Resilience in the Built
Environment, 2(2):1-27.
Aigbavboa, C.O. (2010). An evaluation of the post occupancy experience of housing subsidy
beneficiaries in South Africa: A case study of Gauteng. Johannesburg: University of
Johannesburg.
Aigbavboa, C.O. & Thwala, W.D. (2011). Housing experience of South African low-income
beneficiaries’. The built & human environment review, 4(0):1-13
Aigbavboa, C.O. & Thwala, W.D. (2011). An overview of human settlement in Nigeria: A ray
of hope for the slum dwellers? In: Laryea, S., Leiringer, R. & Hughes, W. ed. West Africa
Built Environment Research (WABER) Conference, 19-21, Accra, Ghana, 167-179.
Aigbavboa, C.O. & Thwala, W.D. (2009). Problems facing the Implementation of Low Cost
Housing Schemes in Nigeria: 6th post graduate conference on construction industry
development. Midrand, Johannesburg: Pages 1-16.
Ajanlekoko, J.S. (2001). Sustainable Housing Development in Nigeria - The financial and
infrastructural Implication. International Conference on Spatial Information for
Sustainable Development, 2–5 October, Nairobi, Kenya.
Akinjo, T. (1984). Nigerian land policies: Implications for housing and physical planning.
Ibadan, Nigeria: CURP, University of Ibadan.
Akintokunbo, A.A. (2008). Social Housing in Nigeria - An Imminent Mass Housing
Revolution. Available from: http://www.nigeriansinamerica.com/articles/2982/1/Social-
494
Housing-in-Nigeria--An-Imminent-Mass-Housing-Revolution/Page1.html. (Accessed 14
May 2011).
Akpomuvie, O.B. (2010). Self-help as a strategy for rural development in Nigeria: A bottom-
up approach. Journal of alternative perspectives in the social sciences, 2(1):88-111.
Alder, G. (2002). Ownership is not a priority among the urban poor: The case of informal
settlements in Nairobi. Habitat Debate, 5(3):1-5
Alonso, W. (1964). Location and Land Use Cambridge (USA). Harvard: University Press.
Althaus, C., Bridgman, P. & Davis, G. (2007). The Australian Policy Handbook. Sydney: Allen
& Unwin.
Amara, R. (1975). Some Methods of Futures Research Menlo. Park: Institute for the Future.
Amaratunga, D., Baldry, D., Sarshar, M. & Newton, R. (2002). Quantitative qualitative
research in the environment. Work study, 51(1):17-31.
Amerigo, M.A. (1992). A model of residential satisfaction in Socio-Environmental
Metamorphoses: Builtscape, Landscape, Eethnoscape. Euroscape. Karaletsou, Salonica:
Aristotle University of Thessaloniki.
Amerigo, M.A. (1990). The perception of residential environment and environment role.
Culture, Space and History, eds. R. Pamir, I. V & N. Teymur, Ankara, M.E.T.V .: Faculty
of Architecture, V.
Amerigo, M.A. & Aragones, I.J. (1997). A theoretical and methodological approach to the
study of Residential satisfaction. Journal of Environmental Psychology, 17: 47-57.
Amoa-Mensah, K. (2003). Housing in Ghana: A Search for Sustainable Options as the Way
Forward for Enhanced Output- Year 2003 and Beyond. International Building Exhibition
Seminar. Accra, Ghana.
Amole, D. (2009). Residential satisfaction in students’ housing. Journal of environmental
psychology, 29: 76-85.
Anderson, C. (2005). What’s the Difference Between Policies and Procedures? Available from:
http://www.bizmanualz.com/information/2005/04/26/what’s-the-difference-between-
policies-and-procedures.html. (Accessed 13 July 2011).
Anderson, E.W. & Fornell, C. (1994). A customer satisfaction research prospectus in Service
quality: New direction in theory and practice, eds. R.T. Rust & R.L. Oliver, CA. Sage, pp.
241-68.
Anderson, J. & Weidemann, S. (1997). Developing and utilizing models of resident satisfaction
in Advances in Environment Behavior Design. G. Moore & R. Marans, New York:
Plenum; pp. 287-314.
495
Anderson, R. (1973). Consumer dissatisfaction: The effect of Diperformance. Journal of
marketing research, 10(2):38-44.
Andrews, F. & Withey, S. (1976). Social Indicators of Well-Being: America’s Perceptions of
Life Quality. Plenum Press: New York.
Aragones, J.I., Francescato, G. & Garling, T. (2002). Evaluating residential environments:
Choice, satisfaction and behaviour. J.I. Aragones, G. Francescato & T. Garling, London:
Bergin and Garvey, pp. 1–14.
Arbuckle, J.L. (1996). Full information estimation in the presence of incomplete data in
Advanced structural equation modeling. G.A. Marcoulides & R.E. Schumacker, Erlbaum,
Mahwah, NJ, pp. 243–277.
Argyle, M. (1987). The Psychology of Happiness. London: Methuen & Co. Ltd.
Aribigbola, A. (2008). Housing policy formulation in developing countries: Evidence of
programme implementation from Akure, Ondo State, Nigeria. Journal of human ecology,
23(2):125–134.
Arnstein, S.R. (1969). A ladder of citizen participation. Journal of the American institute of
planners, 35(4):216-224.
Arrow, K.J. (1962). Economic Welfare and the Allocation of Resources for Invention in the
Rate and Direction of Inventive Activity: Economic and Social Factors, National Bureau
of Economic Research. Priceton, New Jersey: Princeton University Press.
Arrow, K.J. & Hahn, F.H. (1971). General Competitive Analysis. Edinburgh: Oliver & Boyd.
Atkinson, G.A. (1960). Mass housing in rapidly developing tropical areas. The Town Planning
Review, 31(2 (July):85-102.
Atkinson, R. & Cope, S. (1997). Community participation and Urban regeneration in Britain,
in Contested communities’. P. Hoggett, Policy Press: Bristol, pp. 201-221.
Awotona, A. (1991). Nigerian government participation in housing: 1970-1980, Nigeria. Social
indicators research, 25: 63-98.
Awotona, A. (1990). Nigerian government participation in housing: 1970- 1980. Habitat
International, 14(10):17-40.
Bagozzi, R.P. & Yi, Y. (2012). Specification, evaluation, and interpretation of structural
equation models. Journal of the academy of marketing science, 40(1):8-34.
Baillie, S.T. & V, Peart. (1992). Determinants of housing satisfaction for older married and
unmarried women in Florida. Housing and society, 19(2):101-116.
Ball, M. (1977) Differential Rent and the Role of Landed Property. International Journal of
Urban and Regional Research, 3(3), pp. 380-402.
496
Ball, M. & Harloe, M. (1992). Rhetorical barriers to understanding housing provision: What
the ‘provision thesis’ is and is not. Housing studies, 7(1):3-15.
Ball, M. (1996). Housing and Construction: A Troubled Relationship. Policy Press, Bristol.
Bank of Ghana (2007). The Housing Industry in Ghana: Prospects and Challenges. Policy
Brief. Available from:
http://www.bog.gov.gh/privatecontent/File/Research/PolicyBrief/pbrief-housing-
new(1).pdf (Accessed 30 July 2011).
Bardo, J.W. & Hughey, J.B. (1984). The structure of community satisfaction in a British and
an American community. Journal of social psychology, 124: 151–157.
Barrett, P. & Sutrisna, M. (2009). Methodological strategies to gain insights into informality
and emergence in construction project case studies. Construction management and
economics, 27: 935-948
Bartholomew, J., Loukas, A., Jowers, E.M. & Allua, S. (2006). Validation of the physical
activity self-efficacy scale: Testing measurement invariance between Hispanic and
Caucasian children. Journal phys act health, 3: 70-8.
Batho Pele Policy Review (2003). Final Report and Recommendations.
http://www.sarpn.org/documents/d0000875/docs/BathoPelePolicyReviewFinalReport&
Recommendations.pdf. (Accessed 6 September 2012).
Beales, R. (2005). Delphi creates fresh CDS challenge: A new cash-settlement method for
simplifying credit derivatives faces a stern test with the latest bankruptcy, writes Richard
Beales (capital markets & commodities). The financial times, 43.
Bell, S. (1996). Learning with information systems: learning cycles in information systems
development. 1st edn. New York: Routledge.
Bengtsson, B. (2001). Housing as a social right: Implications for welfare state theory.
Scandinavian political studies, 24(4):255–275
Benjamin, C. (2007). A Brief History of Housing in Ghana. Available from:
www.thestatesmanonline.com. (Accessed 20 August 2010).
Bentler, P.M. (1988). Causal modeling via structural equation systems in Handbook of
multivariate experimental psychology, Perspectives on individual differences, eds. J.R.
Nesselroade & R.B. Cattell, Plenum Press, New York, NY, US, pp. 317-335.
Bentler, P.M. & Chou, C.P. (1987). Practical issues in structural modeling. Sociological
methods and research, 16(1):78-117.
Bentler, P.M. & Speckart, G. (1979). Models of attitude–behavior relations. Psychological
Review, 86:452–464.
497
Bentler, P.M. & Wu, E.J. (2002). EQS 6 for Windows user’s guide. Encino, CA: Multivariate
Software, Inc.
Bentler, P.M. & Wu, E.J. (2005). EQS 6.1 for windows. Encino, CA: Multivariate Software
Inc. p. 1-26.
Berry, L. (1995). Ghana. A country study. Area handbook series. Lanham, MD: Bernan.
Bitner, M. (1987). Contextual cues and consumer satisfaction: The role of physical
surroundings and employee behaviors in service settings. University of Washington.
Bjorklund, K. & Klingborg, K. (2005). Correlation between negotiated rents and
neighbourhood quality: A case study of two cities in Sweden. Housing studies, 20(4):627-
647.
Bollen, K.A. (1989). Structural Equations with Latent Variables. New York: John Wiley &
Sons, Inc.
Bolnick, J. (1996). The grass speaks: People's dialogue and the South African homeless
people’s federation (1994-1996). Environment and urbanization, 8(2):153-70.
Bonaiuto, M., Aiello, A., Perugini, M., Bonnes, M. & Ercolani, A.P. (1999). Multimimensional
perception of residential quality and neighbourhood attachment in the urban environment.
Journal of environmental psychology, 19(4):331-352.
Bond, P. (2001). Against Global Apartheid: South Africa meets the World Bank, IMF and
International Finance. London and New York: University of Cape Town Press. p. 239.
Bonnes, M., Bonaiuto, M. & Ercolani, A.P. (1991). Crowding and residential satisfaction in
the urban environment: A contextual approach. Environment and behavior, 23(5):531-
552.
Boomsma, A. (2000). Reporting analyses of covariance structures. Structural equation
modeling, 7(3):461-83.
Boote, D.D. & Beile, P. (2005). Scholars before researchers: On the centrality of the
dissertation literature review in research preparation. Educational researcher, 34(5):3-15.
Brace, I. (2008). Questionnaire design: How to plan, structure and write survey material for
effective market research. United States: Kogan Page Publishers.
Bramley, G. & Morgan, J. (2003). Building competitiveness and cohesion: The role of new
housebuilding in central Scotland’s cities. Housing studies, 13(4):447-471.
Brandsen, T. (2001). Bring actors back in: Towards an institutional perspective. Housing,
theory and society, 18:2–14.
Brennan, G. & Moehler, M. (2010). Neoclassical Economics. Available from:
http://www.moehler.org/files/Neoclassical%20Economics.pdf. (Accessed 24 June 2011)
498
Bret, T.J. (2002). The human right to adequate housing: Tool for promoting and protecting
individual and community health. American journal of public health, 92(5):712–715.
Brill, J., Bishop, M. & Walker, A. (2006). The competencies and characteristics required of an
effective project manager: A web-based Delphi study. Educational technology research
and development, 54(2):115-40.
Brower, S. (1996). Good neighborhoods: Study of in-town and suburban residential
environments. Westport, CT: Praeger Publishers.
Brown, G. & Plenert, G. (2006). Gap Analysis. Encyclopedia of Management. M. Helms, Gale:
Detroit, pp. 319-321.
Bruce, J. (1998). Review of tenure terminology. University of Wisconsin.
Bruin, M. & Cook, C. (1997). Understanding constmints and residential satisfaction among
low-income single-parent families. Environment and behavior, 29(4):532-553.
Brieschke, P.A. (1992). Reparative Praxis: Rethinking the catastrophe that is Social Science.
Theory into Practice, 31(2), 173-180.
Bryman, A. (2001). Social Research Methods. Oxford University Press: Oxford, UK.
Buckley, C. (1995). Delphi: A methodology for preferences more than predictions. Library
management, 16(7):16-19.
Buckley, C.C. (1994). Delphi technique supplies the classic result? Australian library journal,
43(3):158-64.
Burrell, G., & Morgan, G. (1994). Sociological Paradigms and Organizational Analysis.
Heinemann, 6th Edition, 1-37.
Burns, A.C. & Bush, R.F. (2002). Marketing research: Online research applications. New
Jersey: Prentice Hall.
Burns, D., Hambleton, R. & Hoggett, P. (1994). The politics of decentralization revitalizing
local government London: Palgrave Macmillan.
Byrne, B.M. (2010). Structural Equation Modelling with AMOS. Basic Concepts, Applications
and Programming. New York/London: Routledge.
Byrne, B.M. (2006). Structural equation modelling with EQS- Basic concepts, Applications
and programming. Lawrence Erlbaum Associates, Mahwah.
Cadotte, E., Woodruff, R. & Jenkins, R. (1983). Expectations and norms in models of consumer
satisfaction. Journal of marketing research, 8(3):305-314.
Cameron, R. (2011). Mixed Methods Research: The Five Ps Framework. The Electronic
Journal of Business Research Methods, 9(2), 96-108.
499
Campbell, A., Converse, P.E. & Rogers, W.J. (1976). The quality of the America life:
Perceptions, evaluations, and satisfaction. New York: Russell Sage Foundation.
Canter, D. & Ress, K.A. (1982). Multivariate model of housing satisfaction. International
review of applied psychology, 32:185–208.
Cardozo, R. (1965). An experimental study of customer effort, expectation, and satisfaction.
Journal of marketing research, 2(8):244-249.
Carlsmith, J. & Aronson, E. (1963). Some hedonic consequences of the confirmation and
disconfirmation of expectations. Journal of abnormal and social psychology, 66(2):151-
156.
Carvalho, M. (1995). Residential satisfaction in Condominios Exclusivos (gate-guarded
neighborhoods) in Brazil. EDRA Proceedings, pp. 169.
Carvalho, M., George, V.R. & Anthony, K.H. (1997). Residential satisfaction in conominos
exclusivos (gate-guarded neighborhoods) in Brazil. Environment and behavior, 29:734–
768.
Caughey, C., Francis, S. & Kolodziej, A. (1998). Effects of expectations on user satisfaction
with a remodeled university dining facility. Journal of consumer satisfaction,
dissatisfaction and complaining behavior, 11:180-185.
Cavalli-Sforza, V. & Ortolano, L. (1984). Delphi forecasts of land use - transportation
interactions. Journal of transportation engineering, 110(3):324-39.
Central Intelligence Agency World Factbook (2011). The work of a nation. The centre of
intelligence. Available from: Https://www.cia.gov/library/publications/the-world-
factbook/geos/ke.html. (Accessed 21 July 2011).
Chambers, R. (1995). Poverty and livelihoods: Whose reality counts? Environment and
Urbanization, 7(1):173-204.
Chapman, D.W. & Lombard, J.R. (2006). Determinants of neighborhood satisfaction in fee-
based gated and nongated communities. Urban affairs review, 41(6):769-799.
Charlton, S. (2009). Housing for the nation, the city and the household competing rationalities
as a constraint to reform. Development Southern Africa, 26(2): 301-315
Charlton, S. (2004). An overview of the housing policy and debates, particularly in relation to
women (or vulnerable groupings). Johannesburg, South Africa: Centre for the Study of
Violence and Reconciliation
Charlton, S. & Kihato, C. (2006). Reaching the Poor: An analysis of the influences on the
evolution of South Africa’s housing programme. Democracy and Delivery: Urban Policy
500
in South Africa, eds. U. Pillay, R. Tomlinson & J. du Toit, HSRC Press, Cape Town, South
Africa, pp. 263.
Charlton, S., Silverman, M. & Berrisford, S. (2003). Taking stock: A review of the Department
of Housing’s Programmes, Policies and Practices 1994-2003. Pretoria: National
Department of Housing.
Chereni, S. (2010). Dynamics of housing land allocation in Bulawayo: implications for low-
cost housing. Available from:
http://www.urbanlandmark.org.za/conference/2010_abstracts/abstract_chereni.pdf.
(Accessed 26 June 2011).
Chetty, R. & Szeidl, A. (2004). Consumption commitments and asset prices. Paper presented
at the 2004 SED Meeting. Harvard University Working Paper.
Choguill, M.B. (1996). A ladder of community participation for underdeveloped countries.
Habitat international, 20(3):431-444.
Chou, C.P. & Bentler, P.M. (1995). Estimates and tests in structural equation modelling. In
Structural equation modeling: Concepts, issues, and applications. R.H. Hoyle, Sage
Publications, Inc.: Thousand Oaks, CA.
Chowdhury, A.N. (1996). Let Grassroots Speak: People's Participation, Self-Groups and
NGOs in Bangladesh. University Press Limited: Bangladesh.
Churchill, G. & Iacobucci, D. (2004). Marketing research: Methodological foundations Ohio:
Thomson South-Western.
Churchill, G. & Surprenant, C. (1982). An investifation into the determinants of customer
satisfaction. Journal of marketing research, 19(4):491-504.
Churchill, G.A. (2001). Basic Marketing Research. Fort Worth: The Dryden Press.
Clapham, D. (2009). Social Constructionism and Beyond in Housing Research. Paper for ISA
Housing Conference Glasgow.
Clapham, D., Franklin, B. & Saugeres, L. (2000). Housing management; the social construction
of an occupational role. Housing theory and society, 17(2):68-82.
Coates, J.F. (1975). Review of Sackman Report- Technological Forecasting and Social Change
- Vol. 7 No. 2 New York: American Elsevier Publishing Co.
Colabianchi, N., Dowda, M., Pfeiffer, K.A., Porter, D.E., Almeida, M.J. & Pate, R.R. (2007).
Towards an understanding of salient neighborhood boundaries: Adolescent reports of an
easy walking distance and convenient driving distance. International journal of behavioral
nutrition and physical activity, 1(3): S99-117
501
Comrey, A.L. & Lee, H.B. (1992). A first course in factor analysis. New Jersey: Lawrence
Erlbaum Associates.
Cook, C.C. (1988). Components of neighbourhood satisfaction: Responses from urban and
suburban single-parent women. Environment and behavior, 20(2):115-149.
Cook, C.C., Bruin, M.J. & Laux, S. (1994). Housing assistance and residential satisfaction
among single-parent women. Housing and society, 21(2):62-75.
Cooper, D. & Schindler, P. (2006). Marketing Research. New York: McGraw-Hill.
Cooper, D.R. & Emory, C.W. (1995). Business Research Methods US: Irwin.
Creswell, J. (2010). Mapping the developing landscape of mixed methods research. Sage
Handbook of Mixed Methods in Social & Behavioral Research, A. Tashakkori & C.
Teddlie, Sage, California, pp. 45-68.
Creswell, J. (2003). Research design: Qualitative, quantitative, and mixed methods
approaches. Thousand Oaks: Sage.
Creswell, J.W. (1994). Research Design: Qualitative, Quantitative, and Mixed Methods
Approaches. Thousand Oaks: Sage.
Creswell, J.W., Clark, V.L., Gutmann, M.L. & Hanson, W.E. (2003). Advanced mixed
methods research designs. Handbook of mixed methods in social and behavioral research,
eds. A. Tashakkori & C. Teddlie, Sage, Thousand Oaks, CA, pp. 209-240.
Creswell, J. W., Tashakkori, A., Jensen, K. D., & Shapley, K. L. (2003). Teaching mixed
methods research: Practices, dilemmas, and challenges. Handbook of mixed methods in
social & behavioural research. In A. Tashakkori & C. Teddlie (Eds.), Thousand Oaks,
CA: Sage, pp. 619–637.
Crisp, J., Pelletier, D., Duffield, C., Adams, A. & Nagy, S. (1997). The Delphi method? Nursing
Research, 46:116-8.
Critcher, C. & Gladstone, B. (1998). Utilizing the Delphi technique in policy discussion: A
case study of a privatized utility in Britain. Public administration, 76(3):431-49.
Cronbach, L.J. (1951). Coefficient alpha and the internal Structure of tests. Psychometrika,
16(3), 297-334
Cross, C. (2008). Profiling housing demand: Settlement typology and survey results. Report
prepared for the demonstrator Toolkit for Integrated Planning, for the Department of
Science and Technology. CSIR Report no: CSIR/BE/PSS/ER/2008/0048/B.
Crotty, M. (1998). The Foundations of Social Research: Meaning and Perspective in the
Research. Process London: Sage Publications Ltd.
502
Crull, S.R., Bode, M.E. & Morris, E. (1991). Two tests of the housing adjustment model of
residential mobility. Housing and society, 18(3):53-64.
Cudjoe, F. (2010). Ghana: How affordable is the STX-Ghana affordable housing project?
IMANI - the foreign policy magazine of influential think tank in Africa, (February).
Cuhls, K. (2003). Delphi method. Available from:
http://www.unido.org/fileadmin/import/16959_DelphiMethod.pdf. (Accessed 26
October 2011)
Cutter, S. (1982). Residential satisfaction and the suburban homeowner. Urban geography,
3(4):315-327.
Czinkota, M.R. & Ronkainen, I.A. (1992). Global marketing 2000: A marketing survival guide.
Marketing management, January/February, 37-44.
Dahmann, D.C. (1985). Assessments of neighbourhood quality in metropolitan America.
Urban Affaris Quarterly, 20(4):511-535.
Dahmann, D.C. (1983). Subjective assessments of neighborhood quality by size of place.
Urban studies, 20(1):31-45.
Dalkey, N.C. & Helmer, O. (1963). An experimental application of the Delphi method to the
use of experts. Journal of the institute of management sciences, 9:458-467.
Darkwa, I. (2006). Post-occupancy evaluation of state-subsidised housing units in kayamandi,
stellenbosch. Unpublished material.
David, H. (2007). The Toughest of Chores: policy and practice in children collecting water in
South Africa. Policy Futures in Education, 5(3), 315-326.
Davidson, C.H., Johnson, C., Lizarralde, G., Dikmen, N. & Sliwinski, A. (2007). Truths and
myths about community participation in post-disaster housing projects. Habitat
International, 31(1):100-115.
Davis, G. & Roizen, R. (1970). Architectural determinants of student satisfaction in college
residence halls. Environment design and research association, 2:28-44.
Davy, J. (2006). Assessing public participation strategies in low-income housing: The mamre
housing project. Stellenbosch University, Stellenbosch.
Dawes, R., Singer, D. & Lemons, P. (1972). An experimental analysis of the contrast effect
and its implications for intergroup communication and indirect assessment of attitude.
Journal of personality and social psychology, 21(3):281-295.
Day, L.L. (2000). Choosing a house: The relationship between dwelling type, perception of
privacy and residential satisfaction. Journal of planning education and research, 19:265–
275.
503
Day, R. (1980). How satisfactory is research on consumer satisfaction? Advances in Consumer
Research., ed. J. Olson, Ann Arbor: Association for Consumer Research., pp. 593-597.
De Loor, J.H. (1992). Housing in South Africa: Report by the task group on national housing
strategy. Pretoria: Republic of South Africa, Department of Housing.
De Soto, H. (2000). The Mystery of Capital. London: Bantam.
DeCarlo, L.T. (1997). On the meaning and use of kurtosis. Psychological methods, 2:292-307.
Delbecq, A.L., Van de Ven, A.H. & Gustafson, D.H. (1975). Group Techniques for Program
Planning: A Guide to Nominal Group and Delphi Processes. Glenview, IL: Scott,
Foresman & Company.
Department of Housing (2004). Annual report of department of housing: Gauteng provincial
government. Johannesburg, South Africa: Department of Housing.
Department of Housing (2000). Annual report of department of housing: Gauteng provincial
government. Johannesburg, South Africa: Department of Housing
Department of Housing (1994). White Paper: A new Housing Policy and Strategy for South
Africa, December.
Department of Human Settlement (2011). Statement by Human Settlements Minister, Mr Tokyo
Sexwale during Human Settlements Vision 2030 Youth Summit. Durban: Department of
Human Settlement.
Department of Human Settlement (2009). Technical and General Guidelines - Part A of Part
3 Vol 2 of the National Housing Code (2009) 21. Pretoria: Department of Human
Settlement.
Dewilde, C. & Raeymaeckers, P. (2008). The trade-off between home-ownership and pensions:
Individual and institutional determinants of old-age poverty. Ageing and society,
28(6):805-830.
Diaz-Serrano, L. (2006). Housing satisfaction, homeownership and housing mobility: A panel
data analysis for twelve EU countries.
Dion, P.A. (2008). Interpreting structural equation modelling results: A reply to martin and
cullen. Journal of business ethics, 83:365-368.
Djebuarni, R. & Al-Abed, A. (2000). Satisfaction level with neighbourhood in low- income
public housing in Yemen. Property management, 18(4):230-242.
Duncan, T.L.C. (1971). Measuring housing quality: A study of methods. Centre for Urban and
Regional Studies: University of Birmingham.
Dunn, W.N. (1994). Public policy analysis: An introduction. Englewood Cliffs, NJ: Prentice
Hall.
504
Duran, C.O. (1995). The impact of current global housing strategies on the development of the
housing sector In Colombia. Available from:
http://www.ucl.ac.uk/dpu/k_s/publications/working_papers/k-o/WP71.pdf. (Accessed
5 July 2011).
Duruzoechi, N.F. (1999). Housing development and public policy, Owerri, Nigeria: Alphabet
Nigeria publishers.
Dwijendra, N.K. (2007). Quality of low cost housing settlement project. Architecture
Department, Faculty of Engineering, Udayana University.
Eboh, E. (2010). MDGs-based Planning in Africa: Lesson, Experiences and Challenges: A
Case Study of Nigeria: Economic commission for Africa, MDGs/LDCs section, EDND
case studies. Available from: http://www.uneca.org/ednd/mdgs/WorkshopJune15-
16/documents/nigeria.pdf. (Accessed 25 July 2011).
Ebong, M.O. (1983). The perception of residential quality: A case study of Calabar, Nigeria.
Third world planning review, 5(3):273-85.
Ellram, L. (1996). The use of the case study method in logistics research. Journal of business
logistics, 17(8):93-138.
Elmore, P.E. & Beggs, D.L. (1975). Salience of concepts and commitments to extreme
judgments in response patterns to teachers. Educational, 95(4):325-334.
Els, D.A. & Delarey, R.P. (2006). Developing a holistic wellness model. South African Journal
of Human Resource Management, 4(2):46-56.
Elsinga, M. & Hoekstra, J. (2005). Homeownership and housing satisfaction. Journal of
Housing and the Built Environment, 20:401-424.
Engels, F. (1970). On the Housing Question. Moscow: Progress Publishers.
Engel, J., Kollat, D. & Blackwell, R. (1968). Consumer behavior. New York: Holt, Rinehart
and Winston.
Erevelles, S. & Leavitt, C. (1992). A comparison of current models of consumer
satisfactio/Dissatisfaction. Journal of consumer satisfaction, dissatisfaction and
complaining behavior, 5:104-114.
Eriksson, H. (2002). Benefits from TQM for organisational performance. Lulea: Lulea:
University of Technology.
Esping-Andersen, G. (1990). The Three Worlds of Welfare Capitalism. Cambridge: Polity
Press.
505
Eto, H. (2003). The suitability of technology forecasting/ foresight methods for decision
systems and strategy. A Japanese view, in: Technological Forecasting and Social
Change, no. 70: 231-249.
Falah, M., Al-Abed, A. & Stan, W. (1995). A model for assessing the effectiveness of public
housing in Sana’a (Republic of Yemen). Construction management and economics,
13:457-465.
Fallis, G. (1985). Housing Economics. Toronto: Butterworths.
Farzana, F. (2004). Shortages of middle-income owner-occupied housing in Dhaka - Failures
of Government or market? Available from:
http://scholarbank.nus.edu.sg/bitstream/handle/10635/14687/FarzanaF.pdf?sequence=
1. (Accessed 9 July 2011).
Federal Housing Authority (2011). Delivery models and targets in Nigeria. Available from:
http://www.fha.gov.ng/index.php/structure. (Accessed 28 July 2011).
Federal Republic of Nigeria. (2006). National housing policy. Abuja: Chebychev Ventures Ltd.
Federal Republic of Nigeria (2004). Draft national housing policy. Abuja: Chebychev ventures
Ltd. pp 9-22.
Federal Republic of Nigeria (1999). Economic policy direction for Nigeria, 1999-2003. Abuja:
Federal ministry of information. .
Federal Republic of Nigeria (1991). National housing policy. Federal ministry of works and
housing, Lagos.
Federal Republic of Nigeria (1981). Fourth national development plan 1981-85. Lagos.
Ferguson, B. (2001). Housing Policy in the New Millennium Conference. Conference
proceedings of the U.S. Department of Housing and Urban Development. Washington,
D.C.
Festinger, L. (1957). A theory of cognitive dissonance. Stanford: Stanford Press.
Finsterbusch, K. & Warren, V. (1987). The contribution of beneficiary participation to
development project effectiveness. Public administration and development,
7(January/March):1-23.
Fishbein, M. & Ajzen, I. (1975). Belief, attitude, intention and behavior: An introduction to
theory and research. Reading Mass:Addison-Wesley.
Flick, U. (2009). An introduction to qualitative research. Thousand Oaks, CA: Sage
Publications Ltd.
506
Forman, A. (1986). The impact of purchase decision confidence on the process of consumer
satisfaction/dissatisfaction. Unpublished Ph.D. Dissertation, Knoxville: The University of
Tennessee.
Fornell, C., Johnson, M., Anderson, E., Cha, J. & Bryant, B. (1996). The American customer
satisfaction index: Nature, purpose, and findings. Journal of marketing, 60(4):7-18.
Forzano, G. (2008). Research methods for the behavioural sciences. United States: Cengage
Learning EMEA.
Francescato, G. (2002). Residential satisfaction research: the case for and against. In
Residential environments: Choice, satisfaction and behaviour. Bergin and Garvey
(London).
Francescato, G., Weidemann, S., Anderson, J., et al. (1974). Evaluating residents' satisfaction
in housing for low and moderate income families: a multi- method approach. In D. H
Carson (ed.) Man-Environment interactions: Evaluation & Applications, Environmental
Design.
Francescato, G., Weidemann, S. & Anderson, J.R. (1989). Evaluating the built environment
from the users’ point of view: An attitudinal model of residential satisfaction. New York:
Plenum Press.
Francescato, G., Weidemann, S. & Anderson, J.R. (1987). Residential satisfaction: its uses and
limitations in housing research in Housing and neighbourhood: Theoretical and empirical
contributions, W.V. Vliet, H. Choldin, W. Michelson & P. Popene, Greenwood Press:
Westport, CT, pp. 43-57.
Francescato, G., Weidemann, S., Anderson, J.R., et al. (1979). Satisfaction in HUD-assisted
housing: Design and management Washington, DC: U.S Department of Housing and
Urban Development.
Francescato, G., Weidemann, S. & Anderson, J.R. (1986). Residential satisfaction and
residential quality: An overview of recent applications. 21st International Congress of
Applied Psychology, Jerusalem, Israel.
Freire, P. (2000). The Pedagogy of the Oppressed. New York: The Continuum International
Publishing Group Ltd. p. 181.
Galster, G.C. (1987). Identifying the correlates of dwelling satisfaction: An empirical critique.
Environment and behavior, 19(5):539-568.
Galster, G.C. & Hesser, G.W. (1981). Residential satisfaction: An empirical critique.
Environment and behavior, 13(6):735-758.
507
Galster, G. (1997). Comparing demand-side and supply-side housing policies. Housing studies,
12(4):561.
Garland, R. (1991). The mid-point on rating scale: Is it desirable? Marketing bulletin,
2(May):66-70.
Garrod, B. (2008). The Delphi Technique. Institute of Rural Sciences. University of Wales:
Aberystwyth, UK.
Garson, G.D. (2009). Structural equation modelling. Available from: http:/
/faculty.chass.ncsu.edu/garson/PA765/structur.htm. (Accessed 2 July 2012).
Gbolagade, M. O. (2005). An appraisal of maintenance of privately owned and publicly owned
multi-tenanted estate in Lagos metropolis. Unpublished material.
Ghana Statistical Service (2011). 2010 population and housing census. Available from:
http://www.statsghana.gov.gh/. (Accessed 5 August 2011).
Giddens, A. (2002). Runaway World: How Globalization is Reshaping Our Lives. London:
Profile Books.
Gifford, R. (1997). Environmental psychology: Principles and practices. USA: Allyn and
Bacon.
Gilbert, A. (2004). Helping the poor through housing subsidies: Lessons from Chile, Colombia
and South Africa. Habitat international, 28(1):13-40.
Glossop, C. (2008). Housing and Economic Development: Moving Forward Together. London:
Centre for Research and Market Intelligence, Housing Corporation.
Goldstein, K. (1934). The organism: A holistic approach to biology derived from pathological
data in main. New York: Zone Books, 1995.
Gomez-Jacinto, L. & Hombrados-Mendieta, I. (2002). Multiple effects of community and
household crowing. Journal of environmental psychology, 22:233-46.
Goodman, C.M. (1987). The Delphi technique: A critique. Journal of advanced nursing,
12:729-734.
Government of Ghana (2010). The Coordinated Programme of Economic and Social
Development Policies, 2010 – 2016. Available from:
www.ghana.gov.gh/documents/coordinatedprogramme.pdf. (5 August 2011)
Government of Ghana (2005). Draft national housing policy. Ministry of Water Resources,
Works and Housing. Accra, Ghana: Government of Ghana
Government of Ghana (1993). National shelter strategy, policy planning and evaluation unit.
Accra: Government of Ghana
508
Graham, T. & Korboe, D. (1998). Housing policy in Ghana: Towards a supply-oriented future.
Habitat international, 22(3):245-257.
Gran, G. (1983). Development by People: Citizen Construction of a Just World. New York:
Praeger.
Green, B., Jones, M., Hughes, D. & Williams, A. (1999). Applying the Delphi technique in a
study of GP’s information requirements. Health and social care in the community,
17(3):198-205.
Greenberg, P. (1999). Improving neighbourhood quality. A hierarchy of needs. Housing policy
debate, 10(3):601-624.
Greenwood, J. & Hercowitz, Z. (1991). The allocation of capital and time over the business
cycle. Journal of political economy, 99:1188-1214.
Grisham, T. (2008). Cross cultural leadership. Doctor of Project Management, School of
Property, Construction and Project Management , RMIT, Melbourne.
Guney, Y. (1997). The evaluation of high-rise residents' satisfaction in Turkey. Proceedings,
EDRA 28, Space Design and Management for Place Making, University of Quebec,
Montreal.
Guy, S. & Henneberry, J. (2002). Developers and Development: Perspectives on Property.
Oxford: Blackwell Sciences/RICS Foundation.
Ha, S. (2008). Social housing estates and sustainable community development in South Korea.
Habitat International, 32:349-363.
Hablemitoglu, S., Özkan, Y. & Purutçuoglu, E. (2010). The assessment of the housing in the
theory of Maslow’s hierarchy of needs. European journal of social sciences, 16(2):214-
220.
Hackl, P. & Westlund, A. (2000). On structural equation modelling for customer satisfaction
measurement. Total quality management, 11(4-6):820-826.
Häder, M. & Häder, S. (1995). Delphi und kognitionspsychologie: Ein zugang zur
theoretischen fundierung der Delphi-Methode. ZUMA-nachrichten, 37(19):12.
Hair, J., Anderson, R.E., Tatham, R.L., et al. (1998). Multivariate Data Analysis. Upper Saddle
River, NJ: Prentice Hall.
Hair, J., Bush, R. & Ortinau, D. (2003). Marketing research: Within a changing information
environment. New York: McGraw-Hill/ Irwin.
Hallowell, M. & Gambatese, J. (2010). Qualitative research: Application of the Delphi method
to CEM research. Journal of construction engineering and management, 136 (Special
Issue: Research Methodologies in Construction Engineering and Management):99-107.
509
Hamdi, N. (1991). Housing without Houses: Participation, Flexibility, Enablement. New
York:Van Nostrand Reinhold.
Hamnett, C. (2001). Social segregation and social polarization. Handbook of urban studies, R.
Paddison, London: Sage.
Hansemark, O.C. & Albinson, M. (2004). Customer satisfaction and retention: The experiences
of individual employees. Managing service quality, 14(1):40-57.
Hardy, J.D., O’Brien, A.P. & Gaskin, C.J. (2004). Practical application of the Delphi technique
in a bicultural mental health nursing study in New Zealand. Journal of advanced nursing,
46(1):95-109.
Harloe, M. (1995). The People’s Home? -Social Rented Housing in Europe & America.
Oxford: Blackwell.
Harris, M.M. & Schaubroeck, J. (1990). Confirmatory modeling in OB/HRM: Technical issues
and applications. Journal of management information systems, 16:337-360.
Harris, R. (1998). The silence of the experts: “Aided self-help housing”, 1939–1954. Habitat
International, 22(2): 165-189.
Harris, R. & Giles, C. (2003). A mixed message: The agents and forms of international housing
policy, 1945-1973. Habitat International, 27(2):167-191.
Harvey, D. (1982). The Limits to Capital. Oxford: Basil Blackwell.
Hasson, F., Keeney, S. & McKenna, H. (2000). Research guidelines for the Delphi survey
technique. Journal of advanced nursing, 32(4):1008-15.
Hatzichristos, T. & Giaoutzi, M. (2005). Landfill siting using GIS, fuzzy logic and the Delphi
method (author abstract). International journal of environmental technology and
management, 6(1/2):218.
Hauptmann, E. (2001). Can less be more? Leftist deliberative democrats’ critique of
participatory democracy. Polity, 33(3):397–421.
Haworth, A., Manzi, A. & Kemeny, J. (2004). Social constructionism and international
comparative housing research. In Jacobs, Kemeny and Manzi (eds) op cit.
Hayduk, L.A. & Glaser, D.N. (2000). Jiving the four-step, waltzing around factor analysis, and
other serious fun. Structural equation modeling: A multidisciplinary journal, 7(1):1-35.
Hayes, B.E. (1998). Measuring customer satisfaction: survey design, use and statistical
analysis methods. Milwaukee: ASQ Quality Press.
He, X. (2009). Residential satisfaction with home location: Examination of the relationship
between location- embedded benefits and risk perception. Texas State University-San
Marcos. An unpublished doctoral dissertation.
510
Healey, P. & Barret, S. (1990). Structure and agency in land and property development
processes: Some ideas for research. Urban studies, 27(1):89-103.
Helmer, O. (1977). Problems in futures research: Delphi and causal cross-impact analysis.
Future, 9(1):25-52.
Heppner, P.P. & Heppner, M.J. (2004). Writing and publishing your Thesis, Dissertation and
research- A guide for students in the Helping Professions Belmont: Brooks & Cole-
Thomson Learning.
Herting, J.R. & Costner, H.L. (2000). Another perspective on the proper number of factors and
the appropriate number of steps. Structural equation modeling, 7(1):92-110.
Hill, S. (1995). The social organization of boards of directors. The British Journal of Sociology,
46(2), 245–279.
Hishamuddin, I. (2007). Empirical analysis on factors influencing customer loyalty in
Malaysian telecommunication industry. Multimedia University, Malaysia.
Hodgson, G.M. (1998). The approach of instituional economics. Journal of economic
literature, 36:166-192.
Hodgson, G.M. (1997). The ubiquity of habits and rules. Cambridge journal of economics,
21:663-684.
Hodgson, G.M. (1988). Economics and Institutions: A Manifesto for a Modern Institutional
Economics Cambridge: Polity Press.
Holey, E.A., Feeley, J.L., Dixon, J. & Whittaker, V.J. (2007). An exploration of the use of
simple statistics to measure consensus and stability in Delphi studies. BMC medical
research methodology, 7(52):1-10.
Hong Kong Census and Statistics Department (2006). Population by-census office. Census and
Statistics Department.
Hooley, G.J. & Hussey, M.K. (1994). Qualitative methods in marketing Press. London: The
Dryden.
Hooper, D., Coughlan, J. & Mullen, M. (2008). Structural equation modelling: Guidelines for
determining model fit. Electronic journal of business research methods, 6(1):53-60.
Hourihan, K. (1984). Context-dependent models of residential satisfaction: An analysis of
housing groups in cork Ireland. Environment and behavior, 16:369-393.
Hovland, C., Harvey, O. & Sherif, M. (1957). Assimilation and contrast effects in reaction to
communication and attitude change. Journal of abnormal and social psychology,
55(7):244-252.
511
Hoyer, W.D. & MacInnis, D.J. (2001). Consumer Behaviour. Boston: Houghton Mifflin
Company.
Hsu, C.C. & Sandford, B.A. (2007). The Delphi technique: Making sense of consensus.
Practical assessment, research and evaluation, 12(10):1-8.
Hu, L. & Bentler, P.M. (1999). Cutoff criteria for fit indexes in covariance structure analysis:
Conventional criteria versus new alternatives. Structural equation modeling, 6(1):1-55.
Huchzermeyer, M. (2004). Unlawful occupation: Informal settlements and urban policy in
South Africa and Brazil. Trenton: Africa World Press.
Hunt, H. (1977). CS/D: Bits and pieces: Consumer satisfaction/ dissatisfaction and
complaining behavior. Bloomington, IN: Indiana University Conference April 1977. Pp.
38-41.
Husna, S. & Nurizan, Y. (1987). Housing provision and satisfaction of low-income households
in Kuala Lumpur. Habitat international, 11(4):27–38.
Iacobucci, D. & Oston, A. (1995). Distinguishing service quality and customer satisfaction:
The voice of the customer. Journal of consumer psychology, 4(3):277-303.
Ibem, E.O. (2010). An assessment of the role of government agencies in public-private
partnerships in housing delivery in Nigeria. Journal of construction in developing
countries, 15(2):23–48.
Ibem, E.O. & Amole, O.O. (2011). Evaluation of public housing programmes in Nigeria: A
theoretical and conceptual approach. The built & human environment review, 3:88-117.
Ibem, E.O., Anosike, M.N. & Azuh, D.E. (2011). Challenges in public housing provision in
the post-independence era in Nigeria. International journal of human sciences, 8(2):421-
443.
Ifesanya, O. (2003). Developing affordable housing Delivery in Nigeria Ado-Ekiti, Nigeria:
Department of Quantity Surveying School of Environmental Studies Federal Polytechnic.
Ikejiofor, U. (1998). If past traditions were building blocks: A perspective on low income
housing development in Nigerian cities. Building and environment, 34(2):221-230.
Ilesanmi, A.O. (2010). Post-occupancy evaluation and residents’ satisfaction with public
housing in Lagos, Nigeria. Journal of building appraisal, 6(2):153-169.
International Fund for Agricultural Development (2009). Understanding poor people and their
livelihoods. Available form: http://www.ifad.org/english/institutions/guidance/2.pdf.
(Accessed 18 July, 2012).
International Fund for Agricultural Development (2008). Improving access to land and tenure
security. Available from:
512
http://reliefweb.int/sites/reliefweb.int/files/resources/09C0395A8FE2D522492575C2000
6E337-Full_Report.pdf. (8 July 2012).
Iqbal, S. & Pipon-Young, L. (2009). The Delphi method. The psychologist, 22(7):598-600.
Jack, E.P. & Raturi, A.S. (2006). Lessons learned from methodological triangulation in
management research. Management research news, 29(6):345-357.
Jacobs, K., Kemeny, J. & Manzi, M. (2004). Social Constructionism in Housing Research.
Aldershot: Ashgate.
Jacobs, K. & Manzi, T. (2000). Evaluating the social constructionist paradigm in housing
research. Housing, theory and society, 1735-42.
James, R.N., Carswell, A.T. & Sweaney, A.L. (2009). Sources of discontent: Residential
satisfaction of tenants from an internet ratings site. Environment & behavior, 41(1):43-59.
Jean, K. (1992). Livelihood strategies among farm youth in Rwanda. Michigan: Michigan State
University.
Jenkins, P. (1999). Difficulties encountered in community involvement in delivery under the
new South African housing policy. Habitat international, 23(4):431-446.
Jiboye, A.D. (2011). Achieving sustainable housing development in Nigeria: A critical
challenge to governance. International journal of humanities and social science, 1(No. 9
Special Issue – July 2011):1-7.
Jimoh, R.A. & Olayiwola, S.J. (2008). Managing safety on construction sites. Environmental
technology & science journal, 3(1):29-34.
Jinadu, A.M. (2004). Understanding the Basics of Housing: A Book of Study Notes for Students
in Tertiary Institutions. Minna, Nigeria: King James Publishers.
Johnson, P.J. & Abernathy, T.J. (1983). Sources of urban multifamily housing satisfaction.
Housing and society, 10:36-48.
Johnson, R. B., Onwuegbuzie, A. J., & Turner, L. A. (2007). Toward a definition of mixed
methods research. Journal of Mixed Methods Research, 1(2), 112–133.
Jolaoso, B. A., Musa N.A. & Oriola, O. A (2008). Self-help Contribution: a Viable Source of
Financing Low-income Housing. The 3rd Built Environment conference, Cape Town,
South Africa.
Joreskog, K.G. & Sorbom, D. (1988). PRELIS: A program for multivariate data screening and
data summarization. A preprocessor for LISREL. , 2nd edn. Mooresville, IN: Scientific
Software.
Jyh-Bin, Y. & Sheng-Chi, P. (2006). Development of a customer satisfaction evaluation model
for construction project management. Building and environment, 43:458–468.
513
Kabir, B. & Bustani, S.A. (2009). A Review of Housing Delivery Efforts in Nigeria. ISA
International Housing Conference, University of Glasgow, UK.
Kahana, E., Lovegreen, L., B., K. & Kahana, M. (2003). Person, environment, and person–
environment as an infuences on residential satisfaction of elders. Environment and
behavior, 35:434–453.
Kain, J.F. (1962). The Journey to Work as a Determinant of Residential Location. Papers and
Proceedings, Regional Science Association: 137-160.
Kain, J.F. & Quigley, J.M. (1970). Measuring the value of housing quality. Journal of the
American statistical association, 65(330):532-48.
Kaitilla, S. (1993). Satisfaction with public housing in Papua, New Guinea: The case of West
Taraka housing scheme. Environment and behavior, 25:514-545.
Kassim, N.M. (2001). Determinants of customer satisfaction and retention in the cellular
phone market of Malaysia. Lisbon: Southern Cross University.
Kaya, N. & Erkip, F. (2001). Satisfaction in a dormitory building: The effects of poor height
on the perception of room size and crowding. Environment and behavior, 33:35–53.
Keeney, S., Hasson, F. & McKenna, H.P. (2001). A critical review of the Delphi technique as
a research methodology for nursing. International journal of nursing studies, 38(2):195-
200.
Keivani, R. & Werna, E. (2001). Modes of housing provision in developing countries. Progress
in planning, 55(2):65-118.
Kemeny, J. (2004). Extending constructionist social problems to the study of housing
problems. Social Constructionism in Housing Research, eds. J. Keith, K. Jim & M. Tony,
Ashgate: Aldershot, pp. 49-70.
Kemeny, J. (1992). Housing and Social Theory. London: Routledge.
Kemeny, J. & Lowe, S. (1998). Schools of comparative housing research: From convergence
to divergence. Housing studies, 13(2):161-96.
Kerlinger, F. & Lee, H. (2000). Foundations of behavioural research. Fort Worth, TX:
Harcourt.
Khan, Z and Houpt, T. (2006). Community participation- a necessary element of community
development project. Acta structilia. 13(2), 39-61.
Kilpatrick, D.G. (n.d.). Definitions of Public Policy and the Law. Medical University of South
Carolina: National Violence against Women Prevention Research Center.
514
Kim Sung-H, O. (1997). Modeling resident’s satisfaction: Comparison of the francescato and
fishbein-ajzen model. Department of Urban and Regional Planning, University of Illinois
at Urbana-Champaign. An unpublished doctoral dissertation.
Kim, J.E., Nesselroade, J.R. & Featherman, D.L. (1996). The state component in self-reported
worldviews and religious beliefs of older adults: The MacArthur successful aging studies.
Psychology and aging, 11(3):396-407.
Kim, S. (1997). Outdoor environment satisfaction: Contributions of landscape design to multi-
family housing residents' satisfaction. Environmental design and research association,
University of Illinois at Urbana-Champaign.
Kim, S. & Anderson, J. (1997). Modeling residents’ satisfaction: Comparison of the
francescato and Fishbein and Ajzen models. Environmental design and research
association., University of Illinois at Urbana-Champaign
Kincaid, H. (1998). Positivism in the social sciences: Routledge encyclopedia of philosophy.
London: Routledge.
King, P. (2004). Relativism, subjectivity and the self: A critique of social constructionism in
Jacobs, Kemeny and Manzi (eds) op cit.
Kinsey, J. & Lane, S. (1983). Race, housing attributes and satisfaction with housing. Housing
and society, 1098-116.
Kleinhans, R. (2007). Does Social Capital Affect Residents’ Propensity to Move from
Restructured Neighbourhoods? ENHR 2007 International Conference on Sustainable
Urban Areas. Holland, W05 – Poverty neighbourhoods.
Kline, R.B. (2010). Principles and practice of structural equation modeling. 3rd Edition. New
York: Guilford Press.
Kline, R.B. (2005). Principles and practice of structural equation modeling. 2nd edn. New
York: Guilford Press.
Kline, R.B. (1998). Principles and practice of structural equation modeling. 1st edn. New
York: Guilford Press.
Kotler, P. (2000). Marketing management: Analysis, planning, implementation and control.
Upper Saddle River: New Jersey: Prentice Hall.
Kotler, P., Siew, M.L., Swee, H.A., et al. (1996). Marketing Management: An Asian
Perspective Singapore: Prentice Hall.
Kotze, D.A. & Kellerman, G.E. (1997). Participation and managerial approaches to
development. Development, administration and management: A holistic approach, ed.
D.A. Kotze, Pretoria: Van Schaik, .
515
Kutty, N.K. (1999). Determinants of structural adequacy of dwellings. Journal of housing
research, 10(1):1-27.
Kwofie, T. E., E. Adinyira, et al. (2011). Historical overview of housing provision in pre and
post-independence Ghana. West Africa Built Environment Research (WABER)
Conference. S. Laryea, R. Leiringer and W. Hughes. Accra, Ghana: 541-557.
Landeta, J. (2006). Current validity of the Delphi method in social sciences. Technological
forecasting and social change, 73(5):467-482.
Landman, K. (2004). Gated communities in South Africa: A Review of the relevant policies and
their implications. CSIR Building and Construction Technology.
Landman, K. & Napier, M. (2010). Waiting for a house or building your own? Reconsidering
state provision, aided and unaided self-help in South Africa. Habitat international,
34(3):299-305.
Lane, S. & Kinsey, J. (1980). Housing tenure status and housing satisfaction. Journal of
consumer affairs, 14:341-365.
Lang, R.E. & Hornburg, S.P. (1998). What is social capital and why is it important to public
policy? Housing policy debate, 9(1):1-16.
Lang, T. (1995). An overview of four futures methodologies (Delphi, environmental scanning,
issues management and emerging issue analysis). The manoa journal of fried and half-
fried ideas (about the future), 7 (Occasional Paper Number Seven) - Hawaii Research
Center for Futures Studies, Honolulu.
Lansing, J., R.W. Marans & R.B. Zehner (1970). Planned residential environment. Ann Arbor:
Institute for Social Research, The University of Michigan.
Lazenby, K. (1988). Bewoningsbevrediging in die blanke woonhuissek-tor binne die
munisipaliteit van bloemfontein: n studie van proses, patrone en strategie. Bloemfontein:
University of the Orange, Free State.
Lawhon, L. L. (2009). The Neighborhood Unit: Physical Design or Physical Determinism?
Journal of Planning History, 8(2), 111-132.
Leder, D. & Sayre, T. (1989). Adapting the bedroom of individuals with severe handicaps:
Improving life satisfaction and self-sufficiency. EDRA Proceedings, pp. 320.
Lee, T. (1968). Urban neighbourhood as a socio-spatial schema. Human relations, 21:241–267.
Leedy, P.D. & Ormrod, J.E. (2005). Practical research: Planning and design. 8th edn. Prentice
Hall: Upper Saddle River, NJ.
Leigh, W. & Mitchell, M. (1980). Public housing and the black community. The review of
black political economy, 11(1):53-75.
516
Li, Z. (2002). Development and contradictions of wuhan’s public housing system. Singapore:
National University of Singapore.
Lincoln, Y.S. & Guba, E.G. (1985). Naturalistic Inquiry. London: Sage.
Lindberg, E., Garling, T. & Montgomery, H. (1988). Subjective belief-value structures as
determinants of preferences for and choices among housing alternatives. Umea:
University of Umea, pp. 17.
Lindberg, E., Garling, T., Montgomery, H. & Waara, R. (1987). People’s evaluation of housing
attributes. A study of underlying beliefs and values. Scandinavian housing and planning
research, 481–103.
Linstone, A. & Turoff,M. (1975). The Delphi method: techniques and applications Reading,
MA: Addison-Wesley.
Linstone, H.A. (1978). The Delphi technique. Handbook of futures research, ed. J. Fowlers,
Greenwood Press, Westport, CT, pp. 273-300.
Linstone, H.A. & Turoff, M. (2002). The Delphi method: techniques and applications.
Available from: from www.is.njit.edu/pubs.php. (Accessed 12 December 2011).
Lisle, J. (2011). The benefits and challenges of mixing methods and methodologies: Lessons
Learnt from Implementing Qualitatively Led Mixed Methods Research Designs in
Trinidad and Tobago. Caribbean Curriculum, 18, 87–120.
Little, T.D. (1997). Mean and covariance structures (MACS) analyses of cross-cultural data:
Practical and theoretical issues. Multivariate Behavioral Research, 32:53-76.
Lizarralde, G. & Massyn, M. (2008). Unexpected negative outcomes of community
participation in low-cost housing projects in South Africa. Habitat international, 32(1):1-
14.
Loo, R. (2002). The Delphi method: A powerful tool for strategic management, policing. An
international journal of police strategies and management, 25(4):762-769.
Lu, M. (1999). Determinants of residential satisfaction: Ordered Logit vs. regression models.
Growth and change, 30(Spring):264-287.
Lucko, G. & Rojas, E.M. (2010). Research validation: Challenges and opportunities in the
construction domain. Journal of construction engineering and management, 136(1):127-
135.
Lundequist, J. (1999). Tools of Scientific Thinking. Royal Institute of Technology: Stockholm.
Lux, M. (2003). Housing Policy: An end or a new beginning? Budapest: Open Society Institute.
Maart, L. & Soa, S. (1996). From political to developmental practice: Facing the challenges
of fieldworker development and training Durban: Olive Subscription Service.
517
Mabogunje, A. L. (2003). The new mass housing and urban development policy: Social and
economic impact. Unpublished material.
MacCallum, R.C., Browne, M.W. & Sugawara, H.M. (1996). Power analysis and
determination of sample size for covariance structure modeling. Psychological methods,
1:130-149.
Maclennan, D. (1982). Housing Economics: An Applied Approach New York: Longmans,
London.
Mafukidze, J.K. & Hoosen, F. (2009). Housing shortages in South Africa: A discussion of the
after-effects of community participation in housing provision in Diepkloof. Urban forum,
20:379–396.
Maguire, M. (1987). Doing participatory research: a feminist approach. Amherst, MA: The
Center of International Education, University of Massachussetts.
Mahama, C. A. (2004). Institutional and Legal Arrangements for Land Development in Ghana.
England: University of Cambridge.
Malhotra, N.K. (1999). Marketing research: An applied orientation. 3rd edn. New Jersey:
Prentice Hall.
Malpass, P. & Murie, A. (1999). Housing Policy and Practice. 5th edn. Houndmills,
Basingstoke: Palgrave.
Malpezzi, S. (1990). Urban housing and financial markets: Some international comparisons.
Urban studies, 27(6):971-1022.
Mandler, G. (1984). Mind and body: Psychology of emotion and stress, Norton: New York.
Mangan, J., Lalwani, C. & Gardner, B. (2004). Combining quantitative and qualitative
methodologies in logistics research. International journal of physical distribution &
logistics management, 34(7):565-578.
Manikela, S.J. (2008). Understanding the Peripheralisation of low-income housing delivery in
the Mbombela Local Municipality in Faculty of Engineering and the Built Environment,
University of the Witswatersrand, Johannesburg, South Africa.
Manu, P., Ankrah, N., Proverbs, D., & Suresh, S (2010). Exploring the influence of
construction project features in Accident causation. CIB 2010 World Congress, Salford,
CIB.
Marans,R. & Rogers,S. (1975). Toward an understanding of community satisfaction.
NewYork: Halstead Press.
Marans, R. & Sprecklemeyer, K. (1981). Evaluating Built Environment: A Behavioral
Approach Michigan: The University of Michigan: Ann Arbor.
518
Marcus, C.C. (1995). Home as a mirror of self; exploring the deeper meaning of home.
Berkeley Calif: Conari press.
Marshall, G. (1998). A Dictionary of Sociology, second edition Oxford: Oxford University
Press.
Marston, G. (2002). Critical discourse analysis and policy-oriented housing research. Housing,
theory and society, 19:82-91.
Marx, K. (1976). Capital. New York: International Publishers.
May, T. (2001). Social research - Issues, methods and process. 3rd Edition. Open University
Press. Buckingham & Philadelphia
Mayring, P. (2007). Introduction: Arguments for mixed methodology. Mixed methodology in
psychological research. In P. Mayring, G. L. Huber, L. Gurtler, & M. Kiegelmann (Eds.),
Rotterdam/Taipei: Sense Publishers, pp. 1–4.
Masini, E. (1993). Why Futures Studies? London: Grey Seal.
Maslow, A.H. (1998). Towards a psychology of being. New York: J. Wiley and Sons.
Maslow, A.H. (1970). Motivation and Personality. New York: Harper and Row.
Masser, I. & Foley, P. (1987). Delphi revisited: Expert opinion in urban analysis. Urban
studies, 24(3):217-224.
Mastura, J., Noor, L. H., Osman, M., & Ramayah, T (2007). The determinants of housing
satisfaction level: a study on Residential development project by Penang Development
Corporation (PDC). Available from: www.fppsm.utm.my/.../73-the-determinants-of-
housing-satisfaction-level-a-study-on-residential-development-project-by-penang.html.
(Accessed 30 December, 2010).
Mathbor, G.M. (2008). Effective Community Participation in Coastal Development. Chicago,
IL: Lyceum Books.
Mayavo, P. (2002). Non-citizens in a democratic space: Perspectives on human security in
Zimbabwe’s large-scale commercial agriculture under the land reform programme:
1980–2002. Available from: http://www.accord.org.za/ajcr/2004-1/AJCR%20vol4-
1%20pg45-63.pdf. (23 May 2011).
Mbanjwa, X. (2009). Sexwale declares war on housing crooks: iOL. Available from:
http://www.iol.co.za/news/politics/sexwale-declares-war-on-housing-crooks-
1.463450. (Accessed 14 September 2011).
McClave, J.T., Benson, P.G. & Sincich, T. (2008). Statistics for business and economics. 10th
edn. United States: Pearson Prentice Hall.
519
McCray, J.W. & Day, S.S. (1977). Housing values, aspirations, and satisfactions as indicators
of housing needs. Home economics research journal, 5:244-254.
McDonald, R.P. & Ho, M.-.R. (2002). Principles and practice in reporting statistical equation
analyses. Psychological methods, 7(1):64-82.
McIntyre, L.J. (1999). The practical skeptic: Core concepts in sociology. CA: Mayfield
Publishing: Mountain View.
McKenna, H. (1994). The Delphi technique: A worthwhile research approach for nursing?
Journal of advanced nursing, 19(6):1221-1225.
McMaster, R. & Watkins, C. (1999). The economics of housing: The need for a new approach.
PRRES/AsRES/IRES Conference, Kuala Lumpur
McQuitty, S., Finn, A. & Wiley, J.B. (2000). Systematically varying customer satisfaction and
its implications for product choice. Academy of marketing science review of regional
studies.
Meredith, W. (1993). Measurement invariance, factor analysis and factorial invariance.
Psychometrika, 58: 525-543.
Meyer, I. & Theron, F. (2000). Workbook: Public participation in local government. A
framework for action. Bellville: SOPMP.
Michelson, W. (1977). Environment Choice, Human Behaviour, and Residential Satisfaction.
Oxford University Press: UK.
Miller, M.K. & Crader, K.W. (1979). Rural-urban differences in two dimensions of community
satisfaction. Rural sociology, 44:489–504.
Miller, M.M. (1993). Enhancing regional analysis with the Delphi method. Review of regional
studies, 23(2):191-212.
Miller, R.L. & Brewster, J.D. (2003). The A-Z of Social Research London: Sage publications.
Mills, E.S. (1972). Studies in the Structure of the Urban Economy Baltimore: Johns Hopkins
Press.
Mitchell, J.C. (1983). Case and situation analysis’. Sociological review, 31(2):187-211.
Mohd. Zulfa,A. (2000). Kajian kepuasan penghuni dan persekitarannya; kajian kes: Taman
perumahan permin jaya, cendering, kuala terengganu: Thesis ijazah sarjana sains
(perumahan), universiti sains malaysia: Pusat pengajian perumahan, bangunan dan
perancangan. (Computer software).
Mohit, M.A., Ibrahim, M. & Rashid, Y.R. (2010). Assessment of residential satisfaction in
newly designed public low-cost housing in Kuala Lumpur, Malaysia. Habitat
International, 34(1):18-27.
520
Møller, V. & Saris, W. (2001). The relationship between subjective well-being and domain
satisfactions in South Africa. Social indicators research, 55(1):97-114.
Monk, S., Tang, C. & Whiteheadat, C. (2010). What does the literature tell us about the social
and economic impact of housing? Available from: www.scotland.gov.uk/socialresearch.
(Accessed 01 July 2011).
Montero, M. (1991). Residential satisfaction in low interest housing in Mexico city.
Environmental design and research association, 22:68-74.
Moolla, R., Kotze, N. & Block, L. (2011). Housing satisfaction and quality of life in RDP
houses in Braamfischerville, Soweto: A South African case study. Urbani Izziv,
22(1):138-143.
Moote, M.A., McClaran, M.P. & Chickering, D.K. (1997). Research theory in practice:
Applying participatory democracy theory to public land planning. Environmental
management, 21(6):877–889.
Morris, D. & Heathcote, J. (2003). Housing and the business cycle. Available from:
http://www.federalreserve.gov/pubs/feds/2004/200411/200411pap.pdf. (01 July, 2011).
Morris, E.W. (1972). Measuring the quality of housing. Land and economics, 48:383-387.
Morris, E.W. & Jakubczak, M. (1988). Tenure-structure deficit, housing satisfaction and the
propensity to move: A replication of the housing adjustment model. Housing and society,
15(1):41-55.
Morris, E.W. & Winter, M. (1978). Housing, Family and Society. John Wiley and Sons: New
York.
Morris, E.W. & Winter, M. (1975). A theory of family housing adjustment. Journal of
marriage and the family, 37:79-88.
Morris, E.W., Crull, S.R. & Winter, M. (1976). Housing norms, housing satisfaction and the
propensity to move. Journal of marriage and family, 38(2):309-320.
Morrow-Jones, H.Wenning, M.V. & Li, Y. (2005). Differences in neighborhood satisfaction
between African American and White homeowners. Association of Collegiate Schools of
Planning (ACSP46). Kansas City, MO.
Morshidi, S., Abdul Fatah, C. H., Abdul Rashid, A. A., Alip, R., Halim, S., and Usman, Y.
(1999). Low-Cost Housing. In Urban-Industrial Centres of Malaysia: Issues and
Challenges, Penang: Universiti Sains Malaysia Bookshop Ltd.
Muellbauer, J. & Murphy, A. (2008). Housing markets and the economy: The assessment.
Oxford review of economic policy, 24(1):1-33.
521
Mukiibi, S. (2011). An Evaluation of Factors that have Influenced Housing Policy
Development in Uganda. Second International Conference on Advances in Engineering
and Technology, Uganda.
Musonda, I. (2012). Construction health and safety (H&S) performance improvement- A client-
centred model. Johannesburg, South Africa: University of Johannesburg.
Muth, R.F. (1969). Cities and Housing. Chicago: University Press.
Nachmias, D. & Nachmias, F. (1997). Research Methods in the Social Sciences. St. Martins:
New York.
Naidu, E. & Isaacson, M. (2009). How to build quality houses for R55 000: iOL News.
Available from: http://www.iol.co.za/news/south-africa/how-to-build-quality-houses-for-
r55-000-1.466848. (Accessed 14 September 2011).
Nathan, V. (1995). Residents’ satisfaction with the sites and services approach in affordable
housing. Housing and society, 22(3):53-78.
National Department of Housing (2007). Framework for an inclusionary housing policy in
south africa. Pretoria: Department of Housing
National Department of Housing (2007). South Africa: Millennium development goals mid-
term country report. Pretoria: Department of Housing
National Department of Housing (2003). ABC of housing statistics. Pretoria: Department of
Housing.
National Department of Housing (2000). A draft housing strategy for the new millennium.
Pretoria: Department of Housing
National Department of Housing (2000). The user friendly guide to the National Housing Code.
Pretoria: Department of Housing.
National Department of Housing (1995). A new housing policy and strategy for South Africa.
Pretoria: Department of Housing.
National Department of Human Settlement (2011). History of the Department. Available from:
http://www.dhs.gov.za/Content/The%20Department/History.htm. (Accessed 13 August
2011).
National Economic Empowerment and Development Strategy (2004). Federal republic of
Nigeria. Abuja: National planning commission.
National Treasury (2009). Chapter 6: Human Settlements' in Provincial Budgets and
Expenditure Review 2005/06 - 2011/12. Available from:
http://www.treasury.gov.za/publications/igfr/2009/prov/06.%20Chapter%206%20-
%20Human%20Settlements.pdf. (Accessed 3 September 2011).
522
Nelson, D. & Ayeh, S. (2009). Provision of Affordable Housing in Ghana; The Realities.
National Housing Conference.
Neuman, W.L. (2006). Social research methods qualitative and quantitative Approaches. 6th
edn. Boston: Pearson.
Newman, O. (1972). Defensible space: Crime prevention through urban design. New York:
Macmillan.
Nguluma, H.M. (2003). Housing themselves: Transformations, modernisation and spatial
qualities in informal settlements in Dar es Salaam, Tanzania. Kungl Tekniska Hogskolan
Royal Institute of Technology. Stockholm.
Nobrega, C. (2007). The challenge of delivering quality housing in the Eastern Cape: A Case
Study into government-subsidised housing at the Ngqushwa Local Municipality.
North, D.C. (1993). Institutional Change: A Framework of Analysis. Institutional Change:
Theory and Empirical Findings New York: Cambridge University Press. p. 35-46.
North, D.C. (1990). Institutions, Institutional change and Economic Performance. New York:
Cambridge University Press.
North, D.C. (1984). Transaction costs, institutions, and economic history. Journal of
institutional and theoretical economics, 1:407-17.
Novick, R.E. (1990). The Poor die young: Housing and Health in The Third World cities
London: Earthscan Publication Ltd.
Nurizan, Y. and Hashim, A. H. (2001). Perumahan dan Kediaman. Malaysia, Universiti Putra
Malaysia.
Nwaka, G.I. (2005). The urban informal sector in Nigeria: Towards economic development,
environmental health and social harmony. Global urban development magazine, 1(1):1-
11
Oakley, P. (1991). Projects with people: The practice of participation in rural development
Geneva: International Labour office.
Ogu, V.I. (2002). Urban residential satisfaction and the planning implications in a developing
world context: The example of Benin City, Nigeria. International planning studies,
7(1):37-53.
Ogu, V.I. & Ogbuozobe, J.E. (2001). Housing policy in Nigeria: Towards enablement of
private housing development. Habitat international, 25(4):473-492.
Ogunfiditimi, O. & Thwala, W.D. (2008). The people’s Housing Process (PHP) Scheme in
Gauteng province of South Africa- lessons learnt. 5th Post Graduate Conference on
523
Construction Industry Development; J. J. Vester and H. J. Marx. Bloemfontein, South
Africa.
Oh, L. S. (2000). Housing satisfaction of middle income households in bandar baru bangi,
Selangor, Universiti Pertanian Malaysia.
Okoli, C. & Pawlowski, S. (2004). The Delphi method as a research tool: An example, design
considerations and applications. Information Management, 42:15-29.
Okpala, D. (1986). Aspects of urban housing and human settlements policies and strategies in
Africa. Habitat international, 10(3):203-223.
Okun, M.A. (1993). Predictors of volunteer status in a retirement community. International
journal of aging and human development, 36(1):57-74.
Olatubara, C.O. (2002). Housing policy and its impact on the populace: The elusive solution
to housing problem. Continuing Professional Development Workshop on Housing Policy
and Its Impact on the Populace. Ogun State, Nigeria, Nigeria Institution of Estate
Surveyors and Valuers, Ogun State Branch.
Oliver, R.L. (1993). A conceptual model of service quality and service satisfaction: Compatible
goals, different concepts. Advances in services marketing and management, 2(65-85).
Oliver, R.L. (1989). Processing of the satisfaction response in consumption: A suggested
framework and research propositions. Journal of satisfaction, dissatisfaction and
complaining behavior, 21-16.
Oliver, R.L. (1981). Measurement and evaluation of satisfaction process in retail setting.
Journal of retailing, 57:25-48.
Oliver, R.L. (1980b). A cognitive model of the antecedents and consequences of satisfaction
decisions. Journal of marketing research, 17(11):460-469.
Oliver, R.L. (1980a). Theoretical bases of consumer satisfaction research: Review, Critique
and Future Direction in Theoretical Developments in Marketing, eds. W.L. Charles &
M.D. Patrick, American Marketing Association., Chicago.
Oliver, R.L. (1979). Product satisfaction as a function of prior expectation and subsequent
disconfirmation: New evidence, Bloomington: Indiana University.
Oliver, R.L. (1977b). Effect of expectation and disconfirmation on post-exposure product
evaluations: An alternative interpretation. Journal of applied psychology, 62(8):480-486.
Oliver, R.L. (1977a). A Theoretical Reinterpretation of Expectation and Disconfirmation
Effects on Postexposure Product Evaluations: Experience in the Field, in Ralph L. Day
(ed.), Consumer Satisfaction, Dissatisfaction, and Complaining Behavior, Bloomington,
IN: Indiana University Division of Business Research, 2-9.
524
Olivier, A. (2003). A critical review of public participation in development planning within
South African local governments. Organisation Development Africa.
Olotuah, A.O. (2010). Housing development and environmental degeneration in Nigeria. The
built & human environment review, 342-48.
Olotuah, A.O. (2006). Housing quality in suburban areas (An empirical study of Oba–Ile,
Nigeria). DIMENSI (Jurnal Teknik Arsitektur), 34(2):133-137.
Olshavsky, R.W. & Miller, J.A. (1972). Consumer expectation, product performance and
perceived product quality. Journal of marketing research, 9(February):19-21.
Olson, J.C. & Dover, P. (1979). Disconfirmation of consumer expectation through product
trial. Journal of Applied Psychology, 641:79-89.
Olson, M. (1965). The Logic of Collective Action. Cambridge, Mass: Harvard University Press.
Omenya, A. (2002). Sustainable Self-help Housing in South Africa? Conference of Housing
and Urban Development in Sub Saharan Africa. Accra, Ghana.
Omole, F.K. (2001). Basis issues in housing development Ondo, Nigeria: FemoBless
Publishers.
Omoniyi, M.I. (1994). Housing crisis in Nigeria: A critical analysis. The professional builders:
Lagos (March/April).
Onibokun, A.G. (1985). Housing in Nigeria. Ibadan: Nigerian Institute for Social and
Economic Research (NISER).
Onibokun, A.G. (1976). Social system correlates of residential satisfaction. Environment and
behavior, 8(3):323-344.
Onibokun, A.G. (1974). Evaluating consumers’ satisfaction with housing: An application of a
system approach. Journal of American Institute of Planners, 40(3):189-200.
Onukwugha, V.C. (2000). Agricultural - villages: A pragmatic Co-operative Approach to
poverty Alleviation Housing schemes in Nigeria. 4th International Conference on
Housing. Abuja, Nigeria.
Opoku, R.A. & Abdul-Muhmin, A.G. (2010). Housing preferences and attribute importance
among low-income consumers in Saudi Arabia. Habitat international, 34(2):219-227.
Oseland, N. & Raw, G. (1996). Satisfaction with privacy in modern owner-occupied UK
homes. Environment design and research association, 27:106-112.
Oseland, N.A. (1990). An evaluation of space in new homes. Proceedings of the IAPS
Conference Ankara, Turkey, 322-331.
Oxford Library of Words and Phrases. (1993). Oxford University Press.
525
Oxley, M. (2009). Financing Affordable Social Housing in Europe. The Human Settlements
Financing Tools and Best Practices Series, ed. X.Q. Zhang, United Nations Human
Settlements Programme, Nairobi, Kenya.
Ozo, A.O. (1990). Low cost urban housing strategies in Nigeria. Habitat International,
14(1):41-54.
Parasuraman, A. (1991). Marketing research. 2nd edn. Massachusetts: Addison Wesley.
Parasuraman, A., Zeithaml, V. & Berry, L. (1985). A conceptual model of service quality and
its implications for future research. Journal of marketing, 49(4):41-50.
Parente, R.J., Hiob, T.N., Silver, R.A., Jenkins, C., Poe, M.P. & Mullins, R.J. (2005). The
Delphi method, impeachment and terrorism: Accuracies of short-range forecasts for
volatile world events. Technological forecasting and social change, 72(4):401-11.
Paris, E. & Kangari, R. (2006). Multifamily affordable housing: Residential satisfaction.
Journal of performance of constructed facilities, 19(2):138-45.
Parker, C. & Matthews, B.P. (2001). Customer satisfaction: Contrasting academic and
consumers' interpretations. Marketing intelligence and planning, 19(1):38-44.
Parkes, A., Kearns, A. & Atkinson, R. (2002). What makes people dissatisfied with their
neighborhoods? Urban studies, 39(13):2413-2438.
Peck, C. & Stewart, K.K. (1985). Satisfaction with housing and the quality of life. Family and
consumer sciences research journal, 13363–372.
Peyton, R.M., Pitts, S. & Kamery, R.H. (2003). Consumer Satisfaction/Dissatisfaction (CS/D):
A review of the literature prior to the 1990s (pp. 41-46).
Phago, K. G. (2010). Effects of the development and implementation of the National Public
Housing Policy in South Africa with specific reference to The Gauteng province. Pretoria,
University of South Africa, Submitted in accordance with the requirements for the degree
of Doctor of administration.
Phillips, D.R., Siu, O.L., Yeh, A.G. & Cheng, K.H. (2005). The impacts of dwelling conditions
on older persons' psychological well-being in Hong Kong: The mediating role of
residential satisfaction. Social science & medicine, 60(12):2785-97.
Phillips, R. (2000). New applications for the Delphi technique. Annual San Diego: Pfeiffer and
Company.
Pinsonneault, A. & Kraemer, K.L. (1993). Survey research methodology in management
information systems: An assessment. Journal of management information systems,
10(2):75-105.
526
Poisz, T.B.C. & Van Grumbkow, J. (1988). Economic wellbeing, job satisfaction, income
evaluation and consumer satisfaction: an integrative attempt in Handbook of Economics
Psychology, eds. W.F. Raaij, G.M. Veldhoven & K.E. Warnyerd, Kluwer Academic
Publishers: The Netherlands.
Popper, K. (1974). Replies to My Critics: The philosophy of K. popper. La Salle Illinois: Open
Court
Porter, L.W. (1961). A study of perceived need satisfaction in bottom and middle management
jobs. Journal of applied psychology, 4:51-10.
Potter, J., Chicoine, L. & Speicher, E. (2001). Predicting Residential Satisfaction: A
Comparative Case Study in EDRA 32 Proceedings University of Nebraska - Lincoln.
Prinsloo, L. (2010). NHBRC to crack down on unregistered home builders: Engineering news.
Available from: http://www.engineeringnews.co.za/article/nhbrc-to-crackdown-on-
unregistered-home-builders-2010-05-12. (Accessed 10 September 2011).
Public Services Commission (2003). Evaluation of the national housing subsidy scheme.
Pretoria: National Department of Housing.
Pugh, C. (1986). Housing theory and policy. International journal of social economics,
13(4/5):3-104
Putnam, R.D. (1995). Bowling alone, revisited. Responsive community, 5(2):18-33.
Quattrone, P. (2000). Constructivism and accounting research: Towards a trans-disciplinary
perspective’. Accounting, Auditing & Accountability Journal, 13(2), 130- 155.
Rahnema, M. (1992). Participation in The development dictionary - A guide to knowledge and
power, ed. W. Sachs, Zed Books Ltd: London, pp. 116-131.
Raji, O. (2008). Public and private developers as agents in urban housing delivery in Sub-
Saharan Africa. The situation in Lagos State. Humanity of social sciences journal,
3(2):143-150.
Ramdane, D. & Abdullah, A.A. (2000). Satisfaction level with Neighborhood’s in low-income
public housing in Yemen. Property management, 18(4):230-242.
Randava, H. (1979). Origin of Urban Planning: Architectural Handbook. New York: McGraw
Hill.
Rapoport, A. (1977). Human Aspects of Urban Form. Oxford, UK: Pergamon.
Raskin, M.S. (1994). The Delphi study in field instruction revisited: Expert consensus on issues
and research priorities. Journal of social work education, 30:75-89.
Rayens, M.K. & Hahn, E.J. (2000). Building consensus using the policy Delphi method. Policy
politics nursing practice, 1(2):308-315.
527
Raykov, T., Tomer, A. & Nesselroade, J.R. (1991). Reporting structural equation modeling
results in psychology and aging: Some proposed guidelines’. Psychology and aging,
6:499-503.
Reid, N.G. (1988). The Delphi technique, its contributions to the evaluation of professional
practice. Professional Competence and Quality Assurance in the Caring Professional, ed.
R. Ellis, Croom Helm, Beckenham, pp. 230-62.
Reise, S.P., Widaman, K.F. & Pugh, R.H. (1993). Confirmatory factor analysis and item
response theory: Two approaches for exploring measurement invariance. Psychological
bulletin, 114(3): 552-566.
Reisig, D.M. & Chandek, M.S. (2001). The effects of expectancy disconfirmation on outcome
satisfaction in police-citizen encounters. An international journal of police strategies &
management, 24(1):88-99.
Rent, G.S. & Rent, C.S. (1978). Factors related to residential satisfaction. Environment and
behavior, 10:459-488.
Republic of Ghana (2011). Ghana at a Glance. Know More About Ghana. Available from:
http://www.ghana.gov.gh/index.php?option=com_content&view=category&layout=bl
og&id=92&Itemid=238. (Accessed 5 August 2011).
Republic of South Africa (1999). Local government: Municipal structures act [no. 117 of
1998]. Cape Town: Government Gazette, Vol. 402, No. 19614.
Republic of South Africa. (1996). Constitution of the Republic of South Africa [no. 108 of
1996]. Available from: www.info.gov.za/documents/constitution/1996/a108-96.pdf.
(Accessed 10 March, 2011).
Republic of South Africa. (1994). Reconstruction and development program document. ANC
housing policy framework. Johannesburg: Umanyano Publications.
Republic of South Africa (1994). White paper on housing: A new housing policy and strategy
for South Africa, Pretoria: Government Printer.
Ria, V.W. & Bontle, M. (2004). A post-occupancy evaluation of the hope city housing complex
of the greater middelburg housing association, mpumalanga. University of Stellenbosch.
Available from:
www.dhs.gov.za/Content/Research/.../29%20Van%20Wyk%20Moja.pdf. (Accessed 17
July 2008).
Rifkin, S.B. & Kangere, M. (2002). What is participation? A participatory strategy in Africa,
ed. S. Hartley, London: University College London.
528
Robin, R., Brian, O. & Kingstone, M. (2007). Measuring quality of life in informal settlements
in South Africa. Social indicators research, 81:375–388.
Robinson, R. (1979). Housing Economics and Public Policy. Basingstoke: Macmillan.
Robson, C. (2002). Real word research: a research for social scientists and practitioner-
research. Oxford: Blackwell Publishing.
Rogers, C. (1951). Client-centered therapy: Its current, implications and theory. London:
Constable.
Rogers, M.R. & Lopez, E.C. (2002). Identifying critical cross-cultural school psychology
competencies. Journal of school psychology, 40(2):115-141.
Rohe, W.M. & Basolo, V. (1997). Long-term effects of homeownership on the self-perceptions
and social interaction of low- income persons. Environment and behavior, 29(6):793-819.
Roodt, M.J. (2001). Participation, civil society and development. Development theory, policy
and practice, eds. J.K. Coetzee, J. Graaff, F. Hendricks & G. Wood, Oxford University
Press, Cape Town.
Rory, C., Maarten, V.H. & Peteke, F. (2010). A longitudinal analysis of moving desires,
expectations and actual moving behaviour. Forschungsinstitut, zur Zukunft der Arbeit,
Institute for the Study if Labor.
Rosenberg, M.J. & Hovland, C.I. (1960). Cognitive, effective, and behavioral components of
attitude. Yale University Press: New Haven.
Rowe, G., Wright, G. & Bolger, F. (1991). Delphi - A re-evaluation of research and theory.
Technological forecasting and social change, 39:238-251.
Rubin, S., Brian, T., Fong, C. & Brian, K. (1998). Research directions related to rehabilitation
practice: A Delphi study. Journal of rehabilitation, 64(1):19-26.
Rust, K. (2006b). Analysis of South Africa's housing sector performance. Johannesburg:
FinMark Trust.
Rust, K. (2006a). Analysis of South Africa’s Housing Sector Performance. Johannesburg:
FinMark Trust.
Ruyter, J.C. & Scholl, N.B. (1998). Positioning qualitative market research: Reflections from
theory and practices. Qualitative market research journal, 1(1):7-14.
Sackman, H. (1974). Delphi Assessment: Expert Opinion, Forecasting, and Group Process
Santa Monica, CA: Rand Corporation. p. R-1283-PR, April.
Saizarbitoria, I.H. (2006). How quality management models influence company results –
conclusions of an empirical study based on the Delphi method (ISO 9000 and EFQM
529
quality management standards of Europe). Total quality management and business
excellence, 17(6):775-94.
Salleh, A., Yusof, N.A., Salleh, A.G. & Johan Noraire, D. (2011). Tenant satisfaction in public
housing and its relationships with rent arrears: Majlis Badaraya Ipoh, Oerak, Malaysia.
International journal of trade economics and finance, 2(1):10-18.
Sarantakos, S. (2005). Social research. 3rd edn. Basingstoke: Palgrave Macmillan.
Satsangi, M. & Kearns, A. (1992). The use and interpretation of tenant satisfaction surveys in
British social housing. Environment and planning, 10(4):317-331.
Saugeres, L. (1999). The social construction of housing management discourse: Objectivity,
rationality and everyday practice. Housing, theory and society, 16:93-105.
Savasdisara, T., Tips, W.E.J. & Sumannodom, S. (1989). Residential satisfaction in private
estates in Bangkok: A comparison of low-income housing estates and determinant factors.
Habitat International, 13(1):65-73.
Scheele, D.S. (2002). Reality construction as a product of Delphi interaction: The Delphi
method: Techniques and applications.
Schoemaker, P.J.H. (1993). Multiple scenario development: Its conceptual and behavioral
foundation. Strategic management journal, 14(3):193-213.
Schreiber, J.B., Stage, F.K., King, J., Nora, A. & Barlow, E.A. (2006). Reporting structural
equation modeling and confirmatory factor analysis results: A review. The journal of
education research, 99(6):323-337.
Schreiber, J.B., Nora, A., Stage, F.K., Barlow, E.A. & King, J. (2006). Reporting structural
equation modeling and confirmatory factor analysis results: A review. The journal of
educational research, 99(6):323-338.
Schumacker, R.E. & Lomax, R.G. (2004). A Beginner’s Guide to Structural Equation
Modelling Mahwah, New Jersey: Lawrence Erlbaum Associates, Inc.
Schuster, W. M. (2008). For the Greater Good: The Use of Public Policy Considerations in
Confirming Chapter 11 Plans of Reorganization. Houston Law Review, Vol. 46, p. 467.
Sekaran, U. (2000). Research method for business: A skill building approach. New York: John
Wiley and Sons, Inc.
Sexwale, T. (2010). Parliamentary Question No 2692 and reply from the Minister of Human
Settlements. Available from: http://www.polity.org.za/article/sa-has-21m-housing-
backlog-sexwale-2011-07-21. (Accessed 6 September 2011).
530
Sewale, T. (2011). South Africa has 2.1m housing backlog. Available from:
http://www.polity.org.za/article/sa-has-21m-housing-backlog-sexwale-2011-07-21.
(Accessed 6 September 2011).
Shakantu, W.M. (2004). An investigation into building material and waste logistics: A case of
Cape Town. Britain: Glasgow Caledonian University.
Sharipah, N. (2007). Quality affordable housing: A theoretical framework for planning and
design of quality housing. Journal of Techni-Social, 2(1):1-10.
Shook, C.L., Ketchen, D.J., Hult, T. & Kacmar, J.M. (2004). An evaluation of structural
equation modeling in strategic management research. Strategic management journal,
25:397-404.
Shu’aibu, M. U. (2007). Housing policy implementation problems in Nigeria. Daily triumph
newspaper.
Shyllon, Y. (1999). National housing policy in perspective, a building. Manufacturers view
point. The professional builders, June/July. Lagos.
Sidjak, C. (1995). Are mini-suites a viable housing alternative? Two independent studies on
residential satisfaction with small, self-contained apartments. EDRA Proceedings, pp. 169.
Simmonds, C. (1977). The Nature of Futures Problems in Futures Research: New Directions.
H. Linstone & C. Simmonds, Addison-Wesley Publishing Company: London.
Sirgy, M.J. & Cornwell, T. (2002). How neighborhood features affect quality of life. Social
indicators research, 59(1):79-114.
Skinner, J. (1989). Housing wealth and aggregate saving. Regional science and urban
economics, 19(2):305-24.
Skulmoski, G.J., Hartman, F.T. & Krahn, J. (2007). The Delphi method for graduate research.
Journal of information technology education, 6:1-21.
Smith, H. (1999). Theoretical framework: Networks and spaces of negotiation in the provision
of housing. Networks and spaces of Negotiation in Low-income Housing: The case of
Costa Rica. PhD Thesis, ed. H. Smith, School of Planning and Housing, Heriot Watt
University, Edinburgh, pp. 14-21; 41-44; 2.
Smith, M.A. (1997). Perceptions of quality in journalism and communications education: A
Delphi study. Journal of the association for communication administration, 1:32-50.
Smith, M.H. (2004). A Sample/Population size activity: Is it the sample size of the sample as
a fraction of the population that matters? Journal of statistics education, 12(2):1-12.
531
Smith, M.K. (2001). Kurt Lewin, groups, experiential learning and action research: The
encyclopedia of informal education. Available from: http://www.infed.org/thinkers/et-
lewin.htm. (Accessed 23 December 2011).
Smith, N. (1996). The urban frontier: Gentrification and the revanchist city. London: New
York: Routledge.
SOA (1999). Final report of the 1999 Delphi study Schaumburg, IL: Society of Actuaries.
South Africa at a Glance (2011). Provinces & Metropolitan Municipalities. Available from:
http://www.southafricaataglance.com/south-africa-provinces.html. (Accessed 17 September
2011).
South Africa Yearbook 2010/2011 (2011). Human Settlements. Available from:
http://www.gcis.gov.za/sites/default/files/docs/resourcecentre/yearbook/chapter13.pdf.
(Accessed 9 August 2012).
Speare, A. (1974). Residential satisfaction as an intervening variable in residential mobility.
Demography, 11:173-188.
Spencer, C. & Barneji, N. (1985). Strategies for sharing student accommodation: A comparison
of male and female student responses to single and shared rooms. Architecture and
behaviour, 2:123–135.
Spinelli, T. (1983). The Delphi decision-making process. Journal of psychology, 113:73-80.
Spreng, R., MacKenzie, S. & Olshavsky, R. (1996). A reexamination of the determinants of
consumer satisfaction. Journal of marketing, 60(July):15-18.
Statistics South Africa (2011). General Household Survey. Available from:
http://uscdn.creamermedia.co.za/assets/articles/attachments/34446_ghs_2010_p0318june
2010.pdf. (Accessed 16 June 2011).
Statistics South Africa (2009). General Household Survey. Available from:
http://www.statssa.gov.za/publications/P0318/P0318June2009.pdf. (Accessed 6
September 2011).
Steiger, J.H. (2007). Understanding the limitations of global fit assessment in structural
equation modelling. Personality and individual differences, 42(5):893-98.
Stenbacka, C. (2001). Qualitative research requires quality concepts of its own. Management
decision, 39(7):551-555.
Stitt-Gohdes, W.L. & Crews, T.B. (2004). The Delphi technique: A research strategy for career
and technical education. Journal of career and technical education, 20(2):55-67.
Streeck, W. & Thelen, K. (2005). Introduction in Beyond Continuity. Oxford: University Press.
532
Streiner, D.L. (2006). Building a better model: An introduction to structural equation
modelling. The Canadian Journal of Psychiatry, 51(5):317–324.
Sullivan, T. & Gibb, K. (2006). Housing Economics and Public Policy. Great Britain:
Blackwell Science Ltd.
Sundstrom, E., Bell, P.A., Busby, P.L. & Asmns, C. (1996). Environmental Psychology 1989-
1994. Annual review of psychology, pp. 485-521.
Tabachnick, B.G. & Fidell, L.S. (2006). Using multivariate statistics. 5th edn. Boston: Allyn
and Bacon.
Tabachnick, B.G. & Fidell, L.S. (2001). Using multivariate statistics. 4th edn. Boston: Allyn
& Bacon.
Talen, E. & Shah, S. (2007). Neighborhood evaluation using GIS: An exploratory study.
Environment and behavior, 39(5):583-615.
Tamblyn, A. & Shelton, D. (1996). Market Research Manual for Providers of Vocational
Education and Training Melbourne: Victoria Office of Training and Further Education.
Tapscott, C. & Thompson, L. (2010). Participatory Development in South Africa - Between
Rhetoric and Practice. 14th International Research Society for Public Management
Conference. Berne Switzerland: 1-19.
Tashakkori, A. & Teddlie, C. (2003). Handbook of Mixed Methods in Social & Behavioral
Research. Thousand Oaks: Sage Publication.
Tashakkori, A. & Teddlie, C. (1998). Mixed methodology: Combining qualitative and
quantitative approaches: Applied social research methods, no.46. Thousand Oaks, CA:
Sage Publication.
Taylor, P. (1995a). Social and environmental contexts of aging in place. EDRA Proceedings,
pp. 107-112.
Taylor, R.B. (1995b). The impact of crime on comunities. Annals of the American academy of
political and social science, 539:28-45.
Teddlie, C. & Tashakkori, A. (2009). Foundations of Mixed Methods Research Thousand
Oaks, CA: Sage Publications.
Thach, E.C. & Murphy, K.L. (1995). Training via distance learning. Training and development,
49(12):44.
Thangaratinam, S. & Redman, C.W.E. (2005). The Delphi technique. The Obstetrician &
Gynaecologist, 7(2):120-125.
533
The Oxford Library of Words and Phrases (1993). The Oxford Library of Words and Phrases:
Concise Oxford Dictionary of Proverbs, Concise Oxford Dictionary of Quotations and
Concise Oxford Dictionary of Word Origins. United Kingdom: BCA.
The State of African Cities (2007). The Millennium Development Goals and Urban
Sustainability: 30 Years of Shaping the Habitat Agenda. UK & USA: Earthscan. Nations
Human Settlements Programme (UN-HABITAT).
Theodori, G.L. (2001). Examining the effects of community satisfaction and attachment on
individual well-being. Rural sociology, 66(4):618–628.
Thorne, M. & Giesen, M. (2002). Statistics for the behavioral sciences. New York: McGraw-
Hill.
Thwala, W.D. (2009). Experiences and challenges of community participation in urban renewal
projects: The case of Johannesburg, South Africa. Journal of construction in developing
countries, 14(2):37-54.
Tissington, K. (2011). A Resource Guide to Housing in South Africa 1994-2010 Legislation,
Policy, Programmes and Practice. Socio-Economic Rights Institute of South Africa
(SERI): Johannesburg.
Tissington, K. (2010). A Review of Housing Policy and Development in South Africa since
1994. Socio-Economic Rights Institute of South Africa (SERI): Johannesburg.
Tissington, K., Rust, K., Mcgaffin, R., Napier, M, and Charlton, S. 2010). Let’s see the real
value in RDP houses. Available from:
http://www.businessday.co.za/articles/Content.aspx?id=119621. (Accessed 10 September
2011).
Tomlinson, R. (1999). Urbanization in a Post-Apartheid South Africa. London: Unwin Hyman.
Tong, D.Y. (2007). An empirical study of e-recruitment technology adoption in Malaysia:
Assessment of modified technology acceptance model. Multimedia University, Malaysia.
Trochim, W. & Donnelly, J. (2007). The Research Methods Knowledge Base. 3rd edn.
Cincinnati, OH: Atomic Dog Publishing.
Truett, L.J. & Truett, D.B. (1987). Macroeconomics St Louis: Timer Mirror/Mosby College
Publishing.
Tse, D.K. & Wilton, P.C. (1988). Models of consumer satisfaction formation: An extention.
Journal of marketing research, 25(2):204-212.
Turkoglu, H.D. (1997). Residents’ satisfaction of housing environments: The case of Istanbul,
turkey. Landscape and urban planning, 39:55–67.
534
Turner, J.F.C. (1972). Housing as a verb. Freedom to build, dwellers control of the housing
process. eds. J.F.C. Turner & R. Fibhter, New York: M Macmillan.
Turner, M.A. (2005). Landscape preferences and patterns of residential development. Journal
of urban economics, 57:19-54.
Turoff, M. (1970). The design of a policy Delphi, technological forecasting and social change.
2(2):140-71.
Ukoha, O.M. & Beamish, J.O. (1997). Assessment of residents’ satisfaction with public
housing in Abuja, Nigeria. Habitat International, 21(4):445-460.
Umeh, L.C. (2004). Towards improving degraded urban neighbourhoods in Nigeria: Prospects
for residential participation. Management of Environmental Problems and Hazards in
Nigeria, ed. In Mba, H.C., Uchegbu, S.N., Udeh, C.A. and Muoghalu, L.N.(2004) (eds.),
Ashgate Publishing Ltd, Hants, pp. 285-297.
UNCHS (1995). The future of human settlements: Good policy can make a difference. Nairobi:
UNCHS. (A/CONF.165/PC.3/CRP.2).
UNDP (1998). UNDP and governance: Experiences and lessons learned. Oslo: UNDP.
UN-HABITAT (2010). Housing as a Strategy for Poverty Reduction in Ghana. Available
from: www.chfinternational.org/.../1428_file_Ghana_Housing_Report_PDF_.pdf.
(Accessed 30 July 2011).
UN-Habitat (2010). The State of African Cities 2010: Governance, Inequality and Urban Land
Markets Nairobi: United Nations Human Settlements Programme (UN-HABITAT).
UN-Habitat (2006). Shelter for All: The Potential of Housing Policy in the Implementation of
the Habitat Agenda. HS/488/97 E. United Nations: Earthscan.
UN-Habitat (1994). Case studies of Innovative Housing Finance Institutions. Nairobi: United
Nations Centre for Human Settlements.
UN-Habitat (United Nations Human Settlements Programme) (2003). The Challenge of Slums:
Global Report on Human Settlements 2003. London: Earthscan.
United Nations (1995). World urbanization prospects: The 1994 revision: Estimates and
projections of urban and rural populations and of urban agglomerations.
ST/ESA/SER.A/150. New York: United Nations.
United Nations (1978). The role of housing in promoting social interaction. Department of
Economics and Social Affairs. New York: United Nations.
United Nations Centre for Human Settlements (2000). Homelessness: A proposal for global
definition and classification. Nairobi: United Nations
535
United Nations Department of Economic & Social Affairs (1961). Community development in
urban areas. New York: United Nations
United Nations Development Programme (1998). Governance for Sustainable Human
Development. New York: United Nations.
United Nations Human Settlements Programme. (2008). Un-habitat and the Kenya slum
upgrading programme strategy document. May 2008 HS: 1010/08E. New York: United
Nations.
Urban LandMark & Social Housing Foundation (2010). Small -Scale Private Rental: A
Strategy for Increasing Supply in South Africa. Available from:
http://www.urbanlandmark.org.za/downloads/small_scale_rental_booklet_2010.pdf.
(Accessed 3 September 2011).
Uysal, M. & Crompton, J. (1985). An overview of approaches used to forecast tourism demand.
Journal of travel research, 23(4):7-15.
Van Vliet, W. (1998). The encylopedia on housing. London: Sage Publications.
Varady, D. (1983). Determinates of residetial mobility decisions. Journal of the American
planning association, 491:84-99.
Varady, D.P. & Carrozza, M.A. (2000). Towards a better way to measure customer satisfaction
levels in public housing: A report from Cincinnati. Housing studies, 15(6)797-825.
Varady, D.P., Walker, C.C. & Wang, X. (2001). Voucher recipient achievement of improved
housing conditions in the US: Do moving distance and relocation services matter? Urban
studies, 38(8):1273-1305.
Veblen, T. (1900). Preconceptions of economic science. Quarterly journal of economics,
14(February):261.
Vedung, E. (1998). Policy instruments: typologies and theories. In Carrots, Sticks & Sermons:
Policy Instruments & Their Evaluation, eds. M. In: Bemelmans-Videc, R.C. Rist & E.
Vedung, Transaction Publishers, New Brunswick, NJ, pp. 21–58.
Vrbka, S.J. & Combs, E.R. (1991). Predictors of Neighborhood and Community Satisfaction.
Referred papers of the American Association of Housing Educators Annual Conference.
Durham, NH.
Wang, L. (2004). A social perspective on the reformed urban housing provision system in
china: Three cases in Beijing, Xi’an and Shenzhen. Trondheim: Norwegian University of
Science and Technology.
Wates, N. (2005). The Community Planning Handbook. London: Earthscan.
536
Weidemann, S. & Anderson, J. (1982). Residents’ perceptions of satisfaction and safety: A
basis for change in multifamily housing. Environment and behavior, 14(6):695-724.
Weidemann, S. & Anderson, J.R.A. (1985). A conceptual framework for residential
satisfaction. In Home environment. I. Atman & R. Werner, Plenum Press, New York, .
Wessels, L. (2010). Black market highlights RDP cracks. Available from:
http://www.fin24.com/Business/Black-market-highlights-RDP-cracks-20100331.
(Accessed 10 September 2011).
Westaway, M.S. (2006). A longitudinal investigation of satisfaction with personal and
environmental quality of life in an informal South Africa housing settlement, Doornkop,
Soweto. Habitat International, 30:175-189.
Westbrook, R. & Reilly, M.D. (1983). Value percept disparity: an alternative to the
disconfirmation of expectation theory of consumer satisfaction. Advances in Consumer
Research, eds. R.P. Bagozzi & A.M. Tybout, Association for consumer Research, Ann
Arbor, MI, pp. 256-61.
Westergaard, K. (1986). People’s participation, local government and rural development: The
case of west Bengal, India. Copenhagen: Centre for Development Research.
Whitman, N.I. (1990). The committee meeting alternative: Using the Delphi technique. The
Journal of Nursing Administration, 51(1):57-68.
Widaman, K.F. & Reise, S.P. (1997). Exploring the measurement invariance of psychological
instruments: Applications in the substance use domain. The science of prevention:
Methodological advances from alcohol and substance abuse research, eds. K. J. Bryant,
M. Windle & S.G. West, American Psychological Association, Washington, DC, pp. 281–
324.
Wiesenfeld, E. (1994). Construction of the meaning of a in divergent neighborhoods: The case
of a Caracas Barrio. Unpublished material.
Wilcox, D. (1999). A to Z of Participation. New York: Joseph Rowntree Foundation.
Wilkinson, P. (1998). Housing policy in South Africa. Habitat International, 22(3):215-229.
Williams, J.J. (2006). Community participation: Lessons from post-apartheid South Africa.
Policy studies, 27(3):197-217.
Wilson-Doenges, G. (2000). An exploration of sense of community and fear of crime in gated
communities. Environment and behavior, 32(5):597-611.
Wilton, P. & Nicosia, I. (1986). Emerging paradigms for the study of consumer satisfaction.
European research, 14(1):4-11.
537
Wiredu, G. O. (2000). An assessment of the post-commercialisation performance of the state
housing company limited (SHC ltd.) in the solution of housing problems in Ghana. Focus
on Kumasi. Unpublished material.
Wong, T.C. (1999). Marketing research. Oxford, UK: Butterworth-Heinemann.
Wong, Y.H. & Chan, R.Y.K. (1999). Relationship marketing in china: Guanxi, favouritism and
adaptation. Journal of business ethics, 22:107-118.
Woodruff, R., Cadotte, E. & Jenkins, R. (1983). Modeling consumer satisfaction processes
using experience-based norms. Journal of marketing research, 20:296-304.
World Bank (2008). World development report. Washington: World Bank.
World Bank (1996). World Bank Participation Sourcebook. Washington: World Bank.
World Bank (1993). Housing: Enabling markets to work. World Bank: Washington, DC.
World Bank (1988). World development report 1988. Oxford university press for the World
Bank: New York.
World Bank (1984). Housing and Financial institutions in developing countries. Washington
DC: World Bank
World Bank (1975). Housing sector policy paper. World Bank: Washington DC.
Woudenberg, F. (1991). An evaluation of Delphi. Technological forecasting and social
change, 40:131-150.
Yeh, S.H.K. (1972). Homes for the people: A study of Tenant’s views on public housing in
Singapore. University of Singapore: Economic Research Center.
Yi, Y. (1990). A critical review of consumer satisfaction. Review of Marketing, American
Marketing Association, Chicago, IL.
Yin, R.K. (1994). Case Study Research: Design Methods. San Francisco: Sage Publications.
Yiping, F. (2005). Residential satisfaction conceptual framework revisited- A study on
redeveloped neighbourhood in inner-city Beijing. University of Colorado at Denver. An
unpublished doctoral dissertation.
Yong, R.R. (2008). Housing satisfaction perceived by the residents of projek perumahan rakyat
(ppr) sungai bonus in setapak, kuala lumpur. Kulliyyah of Architecture and
Environmental Design International Islamic University Malaysia. An unpublished masters
dissertation.
Young, I. (1990). The Ideal of Community and the Politics of Difference.
Feminism/Postmodernism, ed. N. Linda, New York: Routledge, , pp. 300-23.
Young, J. & Enrique, M. (2009). Helping researchers become policy entrepreneurs. London:
Overseas Development Institute.
538
Yousuf, M.I. (2007). The Delphi technique. Essays in education, 20:80-9.
Yuen, Y.Y. (2007). An ergonomic study of mobile phone usage in Malaysia. Multimedia
University, Malaysia.
Zehner, R. (1977). Indicators of the quality of life in new communities Cambridge: Ballinger.
Zikmund, W.G. (2000). Exploring marketing research. 7th edn. Fort Worth: Dryden Press.
539
APPENDIX A
INVITATION LETTER TO PARTICIPATE IN A DELPHI STUDY
30 November 2011
Dear Sir,
Clinton Aigbavboa is registered for a Ph.D. in the Faculty of Built Environment at the
University of Johannesburg under the supervision of Prof Wellington Didibhuku Thwala
(Masters Programme Co-ordinator, Faculty of Engineering and the Built Environment,
University of Johannesburg, South Africa, Vice-President of The South African Council for
Project and Construction Management Professions - SACPCMP).
The area of his research is on developing a residential satisfaction model on public subsidised
housing in Developing countries: A case study of South Africa. He will be using a Delphi
approach and needs to compile a panel of experts in the field to participate in this process. It
would be appreciated if you would consent to participating in the study in this capacity.
Kind regards
Wellington Didibhuku Thwala PhD (Eng) Pr CPM MCIOB Pr. Pln A/1272/2003 MSAPI
Professor of Construction Project Management
Masters Programme Co-ordinator
Vice-President of The South African Council for Project and Construction Management
Professions (SACPCMP)
Tel: +27 (0)11 559 6048
Fax: +27 (0) 11 559 6630
Fax to email:086 219 1096
Mobile: +27 83 383 5537
Email: [email protected]
Department of Construction Management and Quantity Surveying, Faculty of Engineering and
the Built Environment, P.O. Box 17011, Doornfontein, 2028, South Africa
540
APPENDIX B
REQUEST FOR EXPERT’S CURRICULUM VITAE
07 December 2011
Dear Sir/Madam
I would like to thank you for accepting the invitation to participate as an expert in our project
to develop a residential satisfaction model on public subsidised housing in Developing
countries: A case study of South Africa.
The process of collecting input from the expert panel will probably involve no less than three
rounds. The first round will be in the month of December 2011, the second round will be held
at the end of January 2012 and the third round is anticipated to be held at the end of March
2012. We understand that your time is important. Each round of the Delphi process will take
approximately 30 minutes to complete. A more complete description of the Delphi process is
attached for your information.
To start with, I would like to request for your curriculum vitae for our records and to confirm
your area of expertise. We would appreciate your response by the 15th of December 2011.
Kind regards
Clinton Aigbavboa PhD Candidate
University of Johannesburg
Faculty of Engineering and the Built Environment
Tel: +2711 559 6398
Mobile: +27787958231
Email: [email protected] / [email protected]
541
APPENDIX C
DELPHI METHOD AND APPLICATION TO THIS STUDY
BACKGROUND INFORMATION
The Delphi Method
The Delphi method is a process to collect data and information to solve non-analytical
problems. Used as a research tool, the process gathers knowledge from individuals (experts),
analyses and combines the information to obtain a group consensus.
The Process
The information is gathered in a series of questionnaires or surveys called rounds. The first
round is exploratory in nature and presents the participant with a standard questionnaire. The
second round will present the participant with the group response with his or her response from
the first round. Each experts (participants) member has the opportunity to alter his or her
answer or to voice his or her opinion about new issues collected in the previous survey. The
third round, if necessary, will finalise the statistical response of the group to form a consensus.
The Advantages of Delphi
Delphi has three features over other data collection methods: (1) anonymity, (2) controlled
feedback of results, and (3) statistical group response or consensus.
Anonymity – The members of the process are unknown to other members. This feature will
help minimise the “bandwagon effect.” In public group meetings, one participant, possibly
less knowledgeable, may be more vocal during discussion, potentially persuading more
knowledgeable panellists. Also by keeping the participants unknown, one participant may
change his or her answer to one question without publicly admitting that he or she has done so.
Controlled Feedback – The benefit to the participant of Delphi is gained by feedback of results
collected in earlier rounds. The participant will be sent the group response of colleagues and
other experts in the industry.
Statistical Group Response – The goal of Delphi is to move towards a group consensus.
However, the end result, undoubtedly will display a range of opinions. The statistical group
response is created to assure that the opinions of all participants in the surveys are represented.
The Application of Delphi
You have been asked to participate in the Delphi process for a doctoral study on Housing
Satisfaction of the low-income group, a case study of South Africa to help assess and to
prioritise the vital housing features, policy and management issues in the housing development
industry that can help bring about housing satisfaction to the disadvantaged and low-income
group. The goal of this project is to identify the determinant factor which brings about housing
satisfaction in the low-income housing context to inform stakeholders on the features to give
proper attention when considering building houses for this group. If you agree to participate,
you will be sent the first round questionnaire. The first questionnaire will ask you, in your
opinion, to evaluate issues relating to housing satisfaction and other housing related concerns
affecting the industry. The second questionnaire will follow and will display the group
response of the first questionnaire and new issues collected. You will be able to compare your
response with the opinions displayed by the rest of the participants. The third questionnaire
542
will be used, only if necessary, to eliminate gaps in the information collected in earlier rounds
and to finalise the group response.
Time Commitment
The time commitment is minimal. Each round or questionnaire should take approximately 30
minutes to complete and submit.
Schedule
If you agree to participate in this research project, the first-round questionnaire will be sent (by
e-mail) in December 2011. The researcher will analyse the results and send out the second-
round questionnaire in January 2012. If a third and final-round questionnaire is necessary to
achieve a consensus on the some key issues, this questionnaire will be sent in February 2012
and final results will be available to you in March 2012.
543
APPENDIX D
DELPHI INSTRUCTIONS FOR ROUND 1 AND QUESTIONNAIRE
DELPHI SURVEY – ROUND 1 (Q1)
Thank you for accepting to serve on the Delphi panel for this research. Your acceptance for
participation is greatly appreciated.
This first Round survey is intended to be completed in approximately 25-30 minutes.
Subsequent surveys will require significantly less time to complete. When you have finished
answering all of the questions, please email your response, in Word format, to
[email protected] or [email protected] by 10 January, 2012.
You will be given the opportunity to change your response later on after all Delphi participants
have completed the first Round survey and results have been analysed. Results will be in simple
statistics e.g. median response, average, range and percentage.
INSTRUCTIONS
1. Please answer all of the following questions to the best of your ability.
2. Please indicate your response by placing an ‘X’ in the appropriate boxes. The survey
requests that you rate the prospect of the elements influencing housing satisfaction,
development and advancement in South Africa; the impact of other factors in predicting
residential satisfaction of the low-income group South Africa.
3. Experts are also required to state their levels of agreement using a 5-point Likert Scale with
certain statement and to support their choices where necessary with regards to South Africa
housing subsidy, policy issues and the future of public low-income housing in order to arrive
at a consensus.
4. The influence (probability) scale is presented below and only a number should be used for
a probability range. For instance, if you consider the influence (probability) range to be
between 61 & 70% of the feature’s influence then you should mark ‘X’ under the box ‘7’.
If the impact is considered to be high, then the ‘X’ should be marked under the ‘7’ or ‘8’
box depending on whether your opinion is inclined more towards high or very high impact.
(See below).
Please use your experience, expertise and judgement to rate what you believe the average
negative or positive influence of the various features are on housing satisfaction and on the
South Africa low-income at large would be if the described elements were lacking or present.
INFLUENCE SCALE (probability in percentage)
0-10% 11-
20%
21-
30%
31-
40%
41-
50%
51-
60%
61-
70%
71-
80%
81-
90%
91-
100%
1 2 3 4 5 6 7 8 9 10
X
544
IMPACT SCALE
No impact
Low impact Medium
impact
High impact Very high
impact
1 2 3 4 5 6 7 8 9 10
X
Q1.1 RESIDENTIAL SATISFACTION ATTRIBUTES: To identify the main attributes
that brings about residential satisfaction and to examine if the attribute that determine
satisfaction in other cultural context is the same with South Africa.
Residential satisfaction attributes What is the influence of the following attributes on residential
satisfaction in South Africa low-income housing beneficiaries’?
(1=low influence, 10=high influence)
1 2 3 4 5 6 7 8 9 10 Rank
Dwelling unit X
Housing physical characteristics X
Household or personal
characteristics
X
Housing condition or quality of
building
X
Social features X
Economic features X
Community services X
Neighbourhood and environmental
facilities
X
Culture X
Tenure X
Homeownership X
Needs and expectation X
Beneficiaries meaningful
participation
X
Personality variables X
Aesthetics X X
Location X
Health (personal and
environmental)
X
Safety and Security X
Psychological factors X
Q1.2 DWELLING UNIT FEATURES: This refers to the floor plan of internal spaces
within the dwelling unit and it includes the living, dining, bedroom etc.
Residential satisfaction attributes What is the Impact of each of the listed dwelling unit features
on Overall residential satisfaction in South Africa low-
income housing beneficiaries’ if the listed features are lacking?
(1=no impact, 10=very high negative impact)
545
No impact Low impact Medium
impact
High impact Very High
impact
1 2 3 4 5 6 7 8 9 10
The number of bedrooms X
The size and the location of the
spaces
X
Location of living room X
Location of dining room X
Location of Kitchen X
Size of the bedrooms X
Size of the Kitchen X
Size of the bathrooms X
Size of the wardrobes or closet X
Size of the dining room X
Space for children to play X
Space for children to study X
Balcony X
Privacy within the house X
Brightness and sunshine X
Ventilation in the house X
The floor level X
Overall appearance of building X
Interior design X
Overall size of House X
Washing room area X
Q1.3 NEIGHBOURHOOD AND ENVIRONEMTNAL CHARATERISTICS: Refers to
the position of the housing area with respect to work place and other facilities such as
distances to town centre, school etc.
Residential satisfaction attributes What is the Impact of each of the listed neighbourhood and
environmental features on Overall residential satisfaction in
South Africa low-income housing beneficiaries’ if the listed
features are lacking? (1=no impact, 10=very high negative
impact) No impact Low impact Medium
impact
High impact Very High
impact
1 2 3 4 5 6 7 8 9 10
Location of dwelling unit X
Friends and neighbours X
Closeness to workplace X
Closeness to shopping areas X
Closeness to schools X
Closeness to hospitals/clinics X
Closeness to the place of worship X
Public transportation and services X
Landscape of the neighbourhood X
Adequacy of on-street parking
(bays)
X
Parking facilities X
546
Walkways and access to main
roads
X
Privacy from other neighbours X
Closeness to playground and other
recreational facilities
X
Street and highway noise X
Smoke or odours X
Street lighting at night X
Secure environment X
Physical condition and appearance
of the neighbourhood
X
General cleanliness of the
neighbourhood
X
Proximity to Police services X
Police protection X
Incidence of burglary activities X
Elderly centres X
Community hall X
Facilities for the disabled X
Q1.4 HOUSEHOLD CHARACTERISTICS: This refers to the demographic and economic
characteristics of the household head and includes information about sex, education etc.
Residential satisfaction attributes What is the Impact of each of the listed household
characteristics on Overall residential satisfaction in South
Africa low-income housing? (1=no impact, 10=very high
negative impact) No impact Low impact Medium
impact
High impact Very High
impact
1 2 3 4 5 6 7 8 9 10
Gender (sex) X
Marital status X
Employment and welfare X
Number of children X
Age X
Occupation X
Education X
Household structure X
Race X
Ethnicity X
Tenureship of residence X
Payments for own house X
Length of residency X
Family income X
The amount of rent X
Location of last residence X
Tenureship of last residence X
547
Q1.5 SOCIAL FEATURES: This refers to the social features or the occupants’ social
environment which are likely to impact housing satisfaction which includes variables such as
noise, crime, community relations etc.
Residential satisfaction attributes What is the Impact of each of the listed social features on
Overall residential satisfaction in South Africa low-income
housing beneficiaries’ if the listed features are lacking? (1=no
impact, 10=very high negative impact) No impact Low impact Medium
impact
High impact Very High
impact
Privacy from neighbours X
Interactions with neighbours X
Security around neighbourhood X
Safety at home X
Security provision of flat
(collapsible, sliding front gate,
window burglary etc.).
X
Density of population X
Freedom of choice X
Social relations (social networks) X
Adequacy of escape route in case
of fire
X
Attachment of the communities X
Anticrime measures (report
centres)
X
Special requirement for disabled X
Accident situation X
Community relations X
Q1.6 HOUSING CONDITION OR QUALITY OF BUILDING: This refers to the
observable physical attributes of the housing unit, the housing complex and the community.
Residential satisfaction attributes What is the Impact of each of the listed housing condition or
the quality of building on Overall residential satisfaction in
South Africa low-income housing beneficiaries’ if the listed
features are lacking? (1=no impact, 10=very high negative
impact) No impact Low impact Medium
impact
High impact Very High
impact
The water pressure X
The quality of exterior construction X
The quality of walls X
The quality of interior construction X
The quality of the floors X
The quality of the windows X
The quality of the doors X
The quality of the interior painting X
The quality of the exterior painting X
The quality of the plumbing works X
The quality of the sanitary
finishing
X
548
Functioning of the plumbing
fixtures
X
Plumbing repairs X
Air quality X
Electrical wiring quality X
Rooms and others spaces lighting X
Electrical fittings quality X
Numbers of electrical sockets X
Level of sockets X
Clothes-line facilities X
Overall quality of the unit X
Q1.7 ECONOMIC FEATURES: This refers to the economic value of the housing unit and
the neighbourhood.
Residential satisfaction attributes What is the Impact of each of the listed economic features on
the Overall residential satisfaction of South Africa low-
income housing? (1=no impact, 10=very high negative impact) No impact Low impact Medium
impact
High impact Very High
impact
Home value X
Neighbourhood socio-economic
status
X
Community cost of living X
Q1.8 COMMUNITY SERVICES OR SERVICS PROVIDED BY THE GOVERNMENT:
This refers to the services provided by the government to each housing development areas and
the institutional arrangement under which public housing is administered.
Residential satisfaction attributes What is the Impact of each of the listed services provided by
the government on the Overall residential satisfaction of
South Africa low-income housing beneficiaries’ if the listed
features are lacking? (1=no impact, 10=very high negative
impact) No impact Low impact Medium
impact
High impact Very High
impact
Drainage system X
Garbage and waste collection
system
X
Owned houses X
Security system X
Fire protection X
Maintenance and repair services X
Convenience of bus and public
transportation
X
Electricity supply X
Water supply X
Telephone service X
Handling of residents’ complaints X
549
Management responds to necessary
repairs
X
Housing department officials
treatment of beneficiaries
X
Housing department rules and
regulations of the development
X
Enforcement of rules by the
Department of housing
X
Participation by the community X
Q1.9 PERSONALITY VARIABLES: This refers to the personality variables of the
occupants towards the government and their neighbours.
Residential satisfaction attributes What is the Impact of each of the listed personality variables
on the Overall residential satisfaction of South Africa low-
income housing beneficiaries’? (1=no impact, 10=very high
negative impact) No impact Low impact Medium
impact
High impact Very High
impact
Mistrust of authority X
Negative emotions X
Pessimism X
Q1.10 AESTHETICS: This refers to the building form such as the appearance, height, etc.
Residential satisfaction attributes What is the Impact of each of the listed building aesthetics on
the Overall residential satisfaction of South Africa low-
income housing beneficiaries’? (1=no impact, 10=very high
negative impact) No impact Low impact Medium
impact
High impact Very High
impact Rank
Building forms X
Building height X
External appearance (compare with
others in the neighbourhood)
X
Entrance / lobby design X
Colour of the building X
Q1.11 LOCATION: This refers to the area where the housing units are situated.
Residential satisfaction attributes What is the Impact of the following building location variables
on the Overall residential satisfaction of South Africa low-
income housing beneficiaries’? (1=no impact, 10=very high
negative impact) No impact Low impact Medium
impact
High impact Very High
impact
Size of housing development X
Ease of access by public transport X
Appropriateness of site for erection
of residential building
X
Nearness to slums X
Nearness to economic
opportunities
X
550
Q1.12 HEALTH (PERSONAL AND ENVIRONMENTAL): This refers to the adequacy of
features like ventilation, acoustics that affect the health of the occupants’.
Residential satisfaction attributes What is the Impact of each of the listed personality variables
on the Overall residential satisfaction of South Africa low-
income housing beneficiaries’? (1=no impact, 10=very high
negative impact) No impact Low impact Medium
impact
High impact Very High
impact
Adequacy of daylight distribution
in the units
X
Adequacy of natural ventilation in
the units
X
Acoustic quality in the units X
Water quality (cleanliness, etc.) X
Cleanliness of public areas X
Q1.13 Are there attributes that in your opinion affect satisfaction that has not been addressed? If
any, please state the attributes below
Attributes not listed
Q1.14 SUBSIDISED PUBLIC HOUSING: To identify the factors that makes subsidised
public housing unsustainable in South Africa.
Attributes What is the Impact of the following listed factors on
Public housing delivery in South Africa? (1=no impact,
10=very high negative impact) No impact Low impact Medium
impact
High impact Very High
impact
1 2 3 4 5 6 7 8 9 10
High cost of construction X
None involvement of the big
contractors
X
No solution to general housing problem X
Undesirable X
Lack of skills to handle the production
rate (skills shortage)
X
Lack of resources to handle the
production rate
X
Maintenance cost to the beneficiaries X
Lack of maintenance plan
Illiteracy of the beneficiaries X
Lack of housing education X
Growing unemployment X
Political involvement X
Housing backlog X
Size of the national housing budget X
Bureaucratic capacity X
551
Problem of under-spending X
Q1.15 Are there factors that in your opinion IMPACT public housing delivery in South Africa that
has not been addressed? If any, please state the factor and rank the IMPACT.
Attributes not listed:
Q1.16 HOUSING POLICY INSTRUMENTS: To identify the combination of Housing
Policy instrument that will better serve the South Africa low-income group.
Residential satisfaction attributes What is the influence of the following housing policy
instruments on Public housing delivery in South Africa?
(1=low influence, 10=high influence)
1 2 3 4 5 6 7 8 9 10
Public housing (subsidy schemes) X
Social (Medium-density) Housing
Instrument
X
Rent regulation X
Allocation and rental policies in
current social housing
X
Support for the construction of new
social housing flats
X
Housing allowances X
Tax relief X
Interest subsidies for
homeownership
X
Incremental housing X
Q1.17. Are there any housing policy instruments that in your opinion IMPACT public housing
delivery in South Africa that has not been addressed? If any, please state the factor and rate the
IMPACT
Housing Policy
Instruments not
listed:
-
Q1.18 Does the failure of the South Africa Housing Policy to adequately respond to the needs
of the poor and low-income hinge on the available housing policy instruments in place?
Agreement:
Strongly disagree
Disagree
No opinion
X Agree
Strongly Agree
Q1.19 Please identify any critical issues affecting the provision of housing for the low-income
group in South Africa that have been omitted from the questions above.
Critical issues:
552
Q1.20 Waiting time on the housing database has an impact on the housing satisfaction of the
low-income group?
Agreement: Strongly disagree
Disagree
No opinion
X Agree
Strongly Agree
Q1.21 Which of the following housing delivery programmes will better deliver houses that
will be satisfactory to the housing subsidy beneficiaries’ in South Africa?
Housing delivery programmes Rank
Integrated Residential Development Programme
(IRDP)
1 1
Enhanced People’s Housing Process (ePHP) 2 0
Informal Settlements Upgrading Programme (UISP) 3 0
Consolidation Subsidies 4 0
Emergency Housing Assistance 5 2
Institutional Subsidies 6 0
Social Housing Programme (SHP) 7 3
Community Residential Units (CRU) 8 0
Rural Subsidy: Informal Land Rights 9 4
Farm Residents Housing Assistance Programme 10 0
Comment:
Q1.22 Which of the following housing delivery models will best respond to the needs of the
low-income group in South Africa?
Housing delivery models Rank
Public housing (through the provision
of free subsidy)
1
Self-help housing 2 X – to a
certain
degree
Enabling the market to work 3 X
Provision of social housing (rental
option)
4 X
Q1.23 State subsidised (financed) housing will always be the major housing delivery models
to provide housing to the poor and low-income group in South Africa?
Agreement:
Strongly disagree
Disagree
No opinion
X Agree
Strongly Agree
553
Q1.24 When do you predict government will end the current model of free housing delivery
to the poor and low-income group in South Africa?
Agreement:
6 months’ time
1 - 2 years’ time
3 – 5 years’ time
6 – 8 years’ time
9 - 10 years and above
10 years and above
X – Not in the time that I can think off – the poverty seems to perpetuate itself –
hence no reason to stop the free housing delivery.
Q1.25 What do you envisage will be the replacement for the current government housing
subsidy scheme when it is eventually stopped in the near future?
Comments:
Q1.26 What do you envisage will be the pivotal context of the South Africa Housing policy
in the next 10 years?
Pivotal
context:
Q1.27 Please list and rank the housing development and management issues that affect the
National, Provincial, and the local government housing agencies today in South Africa (1 being
most important, 5 being least important).
Rank
____ A.) _______________________________________________________________
_______________________________________________________________________
_____B.) _________________________________________________________________
_______________________________________________________________________
____ C.) _______________________________________________________________
_______________________________________________________________________
____ D.) _______________________________________________________________
_______________________________________________________________________
____ E.) _______________________________________________________________
_______________________________________________________________________
554
Q1.28 Please list and rank the housing development and management issues that will affect the
National, Provincial, and the local government agencies in the next 10 years or in the future (1
being most important, 5 being least important).
Rank
____ A.) _______________________________________________________________
_______________________________________________________________________
____ B.) _______________________________________________________________
_______________________________________________________________________
____ C.) _______________________________________________________________
_______________________________________________________________________
____ D.) _______________________________________________________________
_______________________________________________________________________
____ E.) _______________________________________________________________
_______________________________________________________________________
Q1.29 Unemployment has an impact on the housing satisfaction of the low-income group?
Agreement:
Strongly disagree
Disagree
No opinion
Agree
X Strongly Agree
Q1.30 Prior exposure to what is to be received has a tendency to influence beneficiaries’
satisfaction towards a given housing product?
Agreement:
Strongly disagree
Disagree
No opinion
X Agree
Strongly Agree
Q1.31 Which of the following housing order needs should be met in order to satisfy the
South Africa low-income housing beneficiaries’?
Housing needs order Rank
Self-actualisation 1 X
Esteem Needs 2 X
555
Social Needs 3 X
Safety Needs 4 X
Physiological 5 X
Q1.32 The provision of housing is the paramount need of the poor and low-income group in
South Africa?
Agreement:
Strongly disagree
Disagree
No opinion
Agree
X Strongly Agree
Q1.33 If you disagreed to the previous statement, what do you think should be the paramount
needs of the poor and low-income group in South Africa that the government should
make a priority?
Principal
needs:
Q1.34 Participation in the housing development process potentially leads to the
implementation of appropriate responses through the involvement of locals in collective
decision making – through the assessment of their needs and expectations?
Agreement:
Strongly disagree
Disagree
No opinion
X Agree
Strongly Agree
Q1.35 Can the current South Africa housing system be referred to as developmental or
welfare?
Agreement:
Strongly disagree
Disagree
No opinion
X Agree
Strongly Agree
Q1.36 The current urban and housing planning system in South Africa favours only the
bourgeois interest in the society?
Agreement:
Strongly disagree
Disagree
X No opinion
Agree
Strongly Agree
556
PERSONAL INFORMATION OF EXPERT PANEL MEMBER
Title (Mr., Mrs., Ms., Dr., Prof.) Mrs
Highest qualification MSc in International Construction Management
Field of specialisation Education & Training in Construction/Housing
Professional registration (CIH,
FCIOB etc.)
ICIOB
Years of experience (housing
studies, development studies,
policy etc.)
7yrs
Current employer Tshwane University of Technology
Position Lecturer
Province and Metropolitan
Municipality currently residing
Tshwane
Have you lived in other
Metropolitan Municipality (s)
before
Yes
If yes, kindly state Ethekwini
Thank you for taking your time to fill out this first round survey. The second round of the
Delphi process will begin on January 25, 2012.
Please do not hesitate to contact me or my promoter Professor Wellington Didibhuku Thwala
if you have any questions about this survey or about the research project in general. Kindly see
contact details below.
Contact details:
Clinton Aigbavboa
Ph.D. Candidate
Dept. of Construction Management and Quantity Surveying, University of Johannesburg
Doornfontein Campus 2028;
Johannesburg, South Africa.
Tel.: +27115596398;
Mobile: +27787958231
Email: [email protected]; [email protected]
Promoter
Professor Wellington Didibhuku Thwala PhD (Eng.) Pr CPM, MCIOB, Pr. Pln, MSAPI
Professor of Construction Project Management
Masters Programme Co-ordinator
Department of Construction Management and Quantity Surveying, Faculty of Engineering and
the Built Environment, University of Johannesburg, South Africa.
Vice-President of The South African Council for Project and Construction Management
Professions (SACPCMP)
Tel: +27 (0)11 559 6048, Fax: +27 (0) 11 559 6630, Fax to email: 086 219 1096, Mobile: +27
83 383 5537, Email: [email protected]
557
APPENDIX E
DELPHI INSTRUCTIONS FOR ROUND 2 AND AN EXAMPLE OF COMPLETED
QUESTIONNAIRE WITH GROUP MEDIAN
DELPHI SURVEY – ROUND 2 (Q2)
Thank you for completing Round 1 of the Delphi survey. We recognise that the survey required
a significant time investment to complete thoughtfully. We appreciate your time and effort.
This Round 2 survey continues the Delphi process for this study. The purpose of Round 2 is to
provide you with the opportunity to change your response, if desired, given the median group
response for each question and element.
The second round survey is intended to take approximately 30 minutes as you are only being
asked to review your previous responses given the collective group median. When you have
finished answering all of the questions, please email your response to [email protected] by
Friday, February 24, 2012.
INSTRUCTIONS
For each element you will see 2 values: your response from the Round 1 survey (indicated with
a yellow highlighted box), and the group median from the Round 1 survey indicated in the
column to the far right hand of each table. Please take one of the following three actions for
each category:
1. Accept the group median response by leaving the field completely unchanged.
2. Maintain your original response by placing an ‘X’ in the highlighted field*.
3. Indicate a new response by placing an ‘X’ in the appropriate field*.
* *If your response is more than ten percent (one unit) above or below the group median
please provide a reason for your outlying response in the field provided.
* * Please kindly review the questions without an initial response from the Round 1 survey
(indicated with an orange highlighted shading); – Indicate a response by placing an ‘X’ in the
appropriate field; or accept the group median response by leaving the field completely
unchanged.
* * New issues identified from the Round 1 survey are also included for your response –
these are indicated with a green shading.
Q2.1. RESIDENTIAL SATISFACTION ATTRIBUTES: To identify the main
attributes that brings about residential satisfaction and to examine if the attribute that
determine satisfaction in other cultural context is the same with South Africa.
Residential satisfaction
attributes
What is the influence of the following attributes on residential
satisfaction in South Africa low-income housing beneficiaries’?
(1=low influence, 10=high influence)
1 2 3 4 5 6 7 8 9 10 Median
Dwelling unit 9
Housing physical characteristics 7
558
Household or personal
characteristics
6
Housing condition or quality of
building
8
Social features 6
Economic features 6
Community services 7
Neighbourhood and
environmental facilities
7
Culture 6
Tenure 8
Homeownership 8
Needs and expectation 7
Beneficiaries meaningful
participation
6
Personality variables 4
Aesthetics 5
Location 9
Health (personal and
environmental)
7
Safety and Security 8
Psychological factors 7
Comment:
Q2.2. DWELLING UNIT FEATURES: This refers to the floor plan of internal spaces
within the dwelling unit and it includes the living, dining, bedroom etc.
Residential satisfaction
attributes
What is the Impact of each of the listed dwelling unit features
on Overall residential satisfaction in South Africa low-income
housing beneficiaries’ if the listed features are lacking? (1=no
impact, 10=very high negative impact) No impact Low impact Medium
impact
High
impact
Very High
impact
1 2 3 4 5 6 7 8 9 10 Median
The number of bedrooms X 8
The size and the location of the
spaces (rooms)
7
Location of living room 5
Location of dining room 5
Location of Kitchen 5
Size of the bedrooms 7
Size of the Kitchen 6
Size of the bathrooms 5
Size of the wardrobes or closet X 5
Size of the dining room 5
Space for children to play 5
Space for children to study 6
Balcony X 3
559
Privacy within the house 6
Brightness and sunshine 6
Ventilation in the house 6
The floor level 5
Overall appearance of building 5
Interior design 5
Overall size of House 6
Washing room area 5
Comment:
Q2.3. NEIGHBOURHOOD AND ENVIRONEMTNAL CHARATERISTICS: Refers to
the position of the housing area with respect to work place and other facilities such as
distances to town centre, school etc.
Residential satisfaction
attributes
What is the Impact of each of the listed neighbourhood and
environmental features on Overall residential satisfaction in
South Africa low-income housing beneficiaries’ if the listed
features are lacking? (1=no impact, 10=very high negative
impact) No impact Low impact Medium
impact
High
impact
Very High
impact
1 2 3 4 5 6 7 8 9 10 Median
Location of dwelling unit 8
Friends and neighbours 7
Closeness to workplace X 9
Closeness to shopping areas X 8
Closeness to schools 8
Closeness to hospitals/clinics X 8
Closeness to the place of
worship
7
Public transportation and
services
9
Landscape of the neighbourhood X 6
Adequacy of on-street parking
(bays)
X 5
Parking facilities X 4
Walkways and access to main
roads
7
Privacy from other neighbours X 7
Closeness to playground and
other recreational facilities
6
Street and highway noise X 6
Smoke or odours X 7
Street lighting at night X 8
Secure environment 9
Physical condition and
appearance of the
neighbourhood
X 7
560
General cleanliness of the
neighbourhood
7
Proximity to Police services 7
Police protection 8
Incidence of burglary activities 9
Elderly centres 5
Community hall 7
Facilities for the disabled X 6
Comment:
Q2.4. HOUSEHOLD CHARACTERISTICS: This refers to the demographic and
economic characteristics of the household head and includes information about sex,
education etc.
Residential satisfaction
attributes
What is the Impact of each of the listed household characteristics
on Overall residential satisfaction in South Africa low-income
housing? (1=no impact, 10=very high negative impact) No impact Low impact Medium
impact
High impact Very High
impact
1 2 3 4 5 6 7 8 9 10 Median
Gender (sex) 5
Marital status 5
Employment and welfare 8
Number of children 7
Age 6
Occupation 6
Education 6
Household structure 6
Race 5
Ethnicity 6
Tenureship of residence 7
Payments for own house 8
Length of residency 8
Family income 8
The amount of rent 9
Location of last residence 6
Tenureship of last residence 6
***Disability X
Comment:
Q2.5. SOCIAL FEATURES: This refers to the social features or the occupants’ social
environment which are likely to impact housing satisfaction which includes variables
such as noise, crime, community relations etc.
Residential satisfaction
attributes
What is the Impactof each of the listed social features on Overall
residential satisfactionin South Africa low-income housing
beneficiaries’ if the listed features are lacking? (1=no impact,
10=very high negative impact)
561
No impact Low impact Medium
impact
High impact Very High
impact
1 2 3 4 5 6 7 8 9 10 Media
n
Privacy from neighbours 7
Interactions with neighbours 7
Security around
neighbourhood
8
Safety at home 8
Security provision of flat
(collapsible, sliding front
gate, window burglary etc.).
8
Density of population 6
Freedom of choice 7
Social relations (social
networks)
7
Adequacy of escape route in
case of fire
5
Attachment of the
communities
7
Anticrime measures (report
centres)
7
Special requirement for
disabled
5
Accident situation 6
Community relations 6
Comment:
Q2.6. HOUSING CONDITION OR QUALITY OF BUILDING: This refers to the
observable physical attributes of the housing unit, the housing complex and the
community.
Residential satisfaction
attributes
What is the Impact of each of the listed housing condition or the
quality of building on Overall residential satisfaction in South
Africa low-income housing beneficiaries’ if the listed features are
lacking? (1=no impact, 10=very high negative impact) No impact Low impact Medium
impact
High impact Very High
impact
1 2 3 4 5 6 7 8 9 10 Median
The water pressure 6
The quality of exterior
construction
7
The quality of walls 7
The quality of interior
construction
7
The quality of the floors 6
The quality of the windows 6
The quality of the doors 6
562
The quality of the interior
painting
6
The quality of the exterior
painting
6
The quality of the plumbing
works
7
The quality of the sanitary
finishing
7
Functioning of the plumbing
fixtures
7
Plumbing repairs 7
Electrical wiring quality 7
Rooms and others spaces
lighting
6
Electrical fittings quality 5
Numbers of electrical sockets 6
Level of sockets 5
Clothes-line facilities 5
Overall quality of the unit 7
Comment:
Q2.7. ECONOMIC FEATURES: This refers to the economic value of the housing unit and
the neighbourhood.
Residential satisfaction
attributes
What is the Impact of each of the listed economic features on the
Overall residential satisfaction of South Africa low-income
housing? (1=no impact, 10=very high negative impact) No impact Low impact Medium
impact
High impact Very High
impact
1 2 3 4 5 6 7 8 9 10 Median
Home value 6
Neighbourhood socio-
economic status
6
Community cost of living X 7
Comment:
563
Q2.8. COMMUNITY SERVICES OR SERVICS PROVIDED BY THE
GOVERNMENT: This refers to the services provided by the government to each
housing development areas and the institutional arrangement under which public
housing is administered.
Residential satisfaction
attributes
What is the Impact of each of the listed services provided by the
government on the Overall residential satisfaction of South Africa
low-income housing beneficiaries’ if the listed features are lacking?
(1=no impact, 10=very high negative impact) No impact Low impact Medium
impact
High impact Very High
impact
1 2 3 4 5 6 7 8 9 10 Median
Drainage system 7
Garbage and waste collection
system
7
Owned houses 7
Security system 6
Fire protection X 6
Maintenance and repair
services
X 6
Convenience of bus and
public transportation
8
Electricity supply X 8
Water supply 9
Telephone service X 5
Handling of residents’
complaints
7
Management responds to
necessary repairs
7
Housing department officials
treatment of beneficiaries
7
Housing department rules
and regulations of the
development
7
Enforcement of rules by the
Department of housing
6
Participation by the
community
7
Comment:
564
Q2.9. PERSONALITY VARIABLES: This refers to the personality variables of the
occupants towards the government and their neighbours.
Residential satisfaction
attributes
What is the Impact of each of the listed personality variables on the
Overall residential satisfaction of South Africa low-income
housing beneficiaries’? (1=no impact, 10=very high negative
impact) No impact Low impact Medium
impact
High impact Very High
impact
1 2 3 4 5 6 7 8 9 10 Median
Mistrust of authority 7
Negative emotions 7
Pessimism 7
Comment:
Q2.10. AESTHETICS: This refers to the building form such as the appearance, height, etc.
Residential satisfaction
attributes
What is the Impact of each of the listed building aesthetics on the
Overall residential satisfaction of South Africa low-income
housing beneficiaries’? (1=no impact, 10=very high negative
impact) No impact Low impact Medium
impact
High impact Very High
impact
1 2 3 4 5 6 7 8 9 10 Median
Building forms 6
Building height 5
External appearance
(compare with others in the
neighbourhood)
6
Entrance / lobby design 5
Colour of the building 4
Comment:
Q2.11. LOCATION: This refers to the area where the housing units are situated.
Residential satisfaction
attributes
What is the Impact of the following building location variables on
the Overall residential satisfaction of South Africa low-income
housing beneficiaries’? (1=no impact, 10=very high negative
impact) No impact Low impact Medium
impact
High impact Very High
impact
1 2 3 4 5 6 7 8 9 10 Median
Size of housing development 6
Ease of access by public
transport
9
Appropriateness of site for
erection of residential
building
6
Nearness to slums 6
Nearness to economic
opportunities
9
565
Comment:
Q2.12. HEALTH (PERSONAL AND ENVIRONMENTAL): This refers to the adequacy
of features like ventilation, acoustics that affect the health of the occupants’.
Residential satisfaction
attributes
What is the Impact of each of the listed personality variables on the
Overall residential satisfaction of South Africa low-income
housing beneficiaries’? (1=no impact, 10=very high negative
impact) No impact Low impact Medium
impact
High impact Very High
impact
1 2 3 4 5 6 7 8 9 10 Median
Adequacy of daylight
distribution in the units
6
Adequacy of natural
ventilation in the units
5
Acoustic quality in the units 5
Water quality (cleanliness,
etc.)
9
Cleanliness of public areas 6
Comment:
Q2.13. Are there attributes in your opinion that affects satisfaction that has not been
addressed? If any, please state the attributes below. See listed attributes below.
Attributes not listed Frequency
Size of land 1
Location of site 1
Lack of beneficiary participation at the grassroots 1
Suppression of vermin like rats 1
Location of kitchen sink 1
Service delivery 1
Temperature / insulation 1
Presence of ceiling 1
Quality (especially regularity and reliability) of sewerage and
garbage removals services
1
Physical safety (in terms of protected drain covers; fenced off
streams, rivers and railway lines as well as road safety issues
especially with regards to children’s and elderly’s safety)
1
566
Q2.14. SUBSIDISED PUBLIC HOUSING: To identify the factors that makes subsidised
public housing unsustainable in South Africa.
Attributes What is the Impact of the following listed factors on Public
housing delivery in South Africa? (1=no impact, 10=very high
negative impact) No impact Low impact Medium
impact
High
impact
Very High
impact
1 2 3 4 5 6 7 8 9 10 Median
High cost of construction 8
None involvement of the big
contractors
6
No solution to general housing
problem
7
Undesirable 6
Lack of skills to handle the
production rate (skills shortage)
6
Lack of resources to handle the
production rate
7
Maintenance cost to the
beneficiaries
7
Lack of maintenance plan X 6
Illiteracy of the beneficiaries 6
Lack of housing education 6
Growing unemployment 8
Political involvement 9
Housing backlog 9
Size of the national housing
budget
7
Bureaucratic capacity 8
Problem of under-spending 7
Comment:
Q2.15 Are there factors that in your opinion IMPACT public housing delivery in South
Africa that has not been addressed? If any, please state the factor and rank the
IMPACT.
See listed FACTORS below. Please kindly rank the IMPACT. (1 being most important, 5
being least important)
Attributes not listed Frequency Impact
Culture of entitlement 1 1
Poor national planning regimes 1 1
Lack of strategic management by government 1 2
Corruption 2 1
Delivery mechanism 1 3
Lack of adequately knowledgeable personnel in the housing
departments
1 1
Public private Partnership in housing delivery 1 2
567
Dependency on government 1 2
Lack of project management skills 1 2
Inadequate beneficiary participation 1 3
Nepotism 1 1
Undue process 1 3
Unnecessary delay 1 2
Price escalation 1 4
Inflation 1 4
Land market / land price 1 4
NIMBYism (Not-in-my-backyard) 1 3
Continuity of unfulfilled high expectations 1 2
Ignorance of bureaucrats dealing with housing 1 1
Q2.16. HOUSING POLICY INSTRUMENTS: To identify the combination of Housing
Policy instrument that will better serve the South Africa low-income group.
Residential satisfaction
attributes
What is the influence of the following housing policy instruments
on Public housing delivery in South Africa?(1=low influence,
10=high influence)
1 2 3 4 5 6 7 8 9 10 Median
Public housing (subsidy
schemes)
9
Social (Medium-density)
Housing Instrument
7
Rent regulation 6
Allocation and rental policies
in current social housing
7
Support for the construction
of new social housing flats
7
Housing allowances 7
Tax relief 6
Interest subsidies for
homeownership
7
Incremental housing 6
Comment:
Q2.17. Are there any housing policy instruments that in your opinion IMPACT public
housing delivery in South Africa that has not been addressed? If any, please state the
factor and rate the IMPACT
See listed instruments below. Please kindly rank their IMPACT. (1 being most important, 5
being least important)
Housing Policy Instruments not listed Frequency Impact
Partnership with the beneficiary 1 2
Needs assessment 1 1
Coordination of housing delivery 1 3
Community rental units 1 2
568
Lack of protection for vulnerable recipients of state built housing
such as RDPs
1 2
Q2.18. Does the South Africa Housing Policy fail to adequately respond to the needs of the
poor and low-income?
Agreement and Median value Round 1
Agreement 1 - Strongly disagree 12.0%
2 - Disagree 29.0%
3 - No opinion 0.0%
4 - Agree 41.0%
5 - Strongly agree 18.0%
Comments:
Q2.18 (a). Is (Does) the failure hinge on the lack of an adequate housing policy
instruments in place?
Agreement:
Strongly disagree
Disagree
No opinion
x Agree
Strongly Agree
Q2.19. Please identify any critical issues affecting the provision of housing for the low-
income group in South Africa that have been omitted from the questions above. Please
kindly rank their IMPACT. (1 being most important, 5 being least important)
Critical issues Frequency Impact
Lack of active participation of beneficiaries in the development
of housing
1 3
Dwindling tax base (limited budget) 1 4
Limited budget 1 1
Poor planning and coordination from National to local
government levels
1 1
Housing delivery mechanism 1 2
Enabling the poor to solve their own housing problem 1 2
The inability of relevant state authorities to consult adequately
with affected local communities to seek joint solutions to housing
crisis.
1 2
Appropriate policy to handle informal settlement 1 2
Comments:
Q2.20. Waiting time on the housing database has an impact on the housing satisfaction of the
low-income group?
569
Agreement and Median value Round 1
Agreement 1 - Strongly disagree 0.0%
2 - Disagree 6.0%
3 - No opinion 6.0%
4 - Agree 53.0%
5 - Strongly agree 35.0%
Comments:
Q2.21. Which of the following housing delivery programmes will better deliver houses that
will be satisfactory to the housing subsidy beneficiaries’ in South Africa? (Rank - 1
being most important, 10 being least important) – See appendix one for definition of
terms
Housing delivery programmes Rank Median
Integrated Residential Development
Programme (IRDP)
1 2
Enhanced People’s Housing Process (ePHP) 3 3
Informal Settlements Upgrading Programme
(UISP)
4 3
Consolidation Subsidies 5 5
Emergency Housing Assistance 2 6
Institutional Subsidies 4 3
Social Housing Programme (SHP) 3 3
Community Residential Units (CRU) 5 4
Rural Subsidy: Informal Land Rights 4 6
Farm Residents Housing Assistance
Programme
8 7
Comment:
Q2.22. Which of the following housing delivery models will best respond to the needs of the
low-income group in South Africa? (Please Rank - 1 being most important, 5 being
least important)
Housing delivery models Rank Media
n
Public housing (through the provision of free subsidy) 2 2
Self-help housing 3 2
Social and Rental Housing model (Programmes facilitating access
to Rental Housing opportunities, supporting Urban Restructuring
and Integration, e.g. Social housing, Institutional Subsidies etc.)
1 2
Incremental Housing model (Programmes facilitating access to
housing opportunities through a phased process, such as Informal
Settlement Upgrading, Consolidated Subsidy, IRDP etc.)
5
Rural Housing model (Programmes facilitating access to housing
opportunities in Rural areas - Rural Subsidy: Informal Land Rights)
4
Comment:
570
Q2.23. State subsidised (financed) housing will always be the major housing delivery models
to provide housing to the poor and low-income group in South Africa?
Agreement and Median value Round 1
Agreement 1 - Strongly disagree 6.0%
2 - Disagree 41.0%
3 - No opinion 0.0%
4 - Agree 29.0%
5 - Strongly agree 24.0%
Comments:
Q2.24. When do you prediction government will end the current model of free housing
delivery to the poor and low-income group in South Africa?
Agreement and Median value Round 1
Agreement 1 – 6 months’ time 7.0%
2 – 1-2 years’ time 7.0%
3 – 3-5 years’ time 13.0%
4 – 6-8 years’ time 13.0%
5 – 9-10 years’ time 0.0%
6 – 10 years and above 53.0%
7 – It will never be stopped 7.0%
Comments:
Q2.25. What do you envisage will be the replacement for the current government housing
subsidy scheme when it is eventually stopped in the near future?
See responses for question below.
Issue Statement Frequency Rank
Market and subsidy mix (social housing-type models) 3 3
Assisted Self-help 4 5
Upgrading of informal settlement 2 3
Providing leasing agreements with informal settlers 1 4
Community housing 1 1
Community based initiatives 1 1
Private buying and financing (much like the old urban foundation
system)
1 1
Rent to buy scheme 1 1
Public private Partnership in housing delivery 1 1
Comments: We can only wait & see.
Q2.26. What do you envisage will be the pivotal context of the South Africa Housing policy
in the next 10 years?
See listed pivotal context - (Please Rank - 1 being most important, 5 being least important)
571
Pivotal context Frequency Rank
Upgrading of informal settlement 3 3
Spatial policy and spatial analysis of need 1 2
Beneficiaries will be encouraged to participate effectively. 1 1
The policy will seek to build the capacity of beneficiaries in order
for them to build their own housing.
1 1
Affordability 1 1
Homelessness for the poor 1 1
Fairness of allocation of housing 2 3
Access to adequate services (both urban and rural) 1 1
Provision of low-income rental stocks 1 2
Delivery of sustainable and energy efficient low cost housing 1 1
Public private Partnership in housing delivery 1 1
Spatial policy and spatial analysis of need 1 1
Informal settlement upgrading 1 3
CRU 1 2
Global and local economic conditions; political will; better
quality control of construction tenders with less corruption
1 1
Acceptance of informal settlements as part of the urban areas
and working with them to improve their lives and not their
homes
1 1
Q2.27. Please list and rank the housing development and management issues that affect the
National, Provincial, and the local government housing agencies today in South Africa
(1 being most important, 5 being least important).
See listed housing development and Management Issues and their rankings
Issue Statement Frequency Rank
Government capacity to facilitate development 9 1
Corruption in provincial and local departments 7 2
Budget constrains 4 3
Poor political will 4 4
Access to well-located land for housing (availability of vacant land) 3 5
Legislature (Planning, Environmental and Heritage) 3 5
Housing backlog 3 5
The current structure of government (Provincial government sphere
is not necessary)
3 5
Government bureaucracy in the housing system 2 5
Sustainability 1 5
Passive beneficiaries (entitlement culture) 1 5
Lack of technical skills (Engineers and Artisans, etc.) 1 5
Unaccountability of government employees 1 5
Urbanisation 1 5
Informal settlements 2 5
Lack of national vision for housing 1 5
Poor planning and coordination 1 5
Ideological constraints 1 5
572
Lack of Participation from communities and CBOs 1 5
Cost of basic services (water, electricity etc.) in RDP houses, social
housing etc.
1 5
Limited participation of the private sector’s involvement in low cost
housing development
1 5
Public private Partnership in housing delivery 1 5
Lack of project management skills / ineptitude 1 5
Nepotism and cadre deployment 1 5
Salary and benefit expectations of executives 1 5
Bloated and duplicated bureaucracies 1 5
Lack of appropriate skills 1 5
Absence of reflection or serious acknowledgement of problems that
affect the poor
1 5
Continued economic resources 1 5
Better regulation and quality control of construction tenders 1 5
Q2.28. Please list and rank the housing development and management issues that will affect
the National, Provincial, and the local government agencies in the next 10 years or in
the future (1 being most important, 5 being least important).
See listed housing development and Management Issues and their rankings
Issue Statement Frequency Rank
Budget constrains 5 1
Government capacity 4 1
Corruption 3 1
Poor political will 3 3
The current structure of government (Provincial government sphere
is not necessary)
2 4
Availability of vacant land 2 5
Legislature (Planning, Environmental and Heritage) 2 5
Lack of national vision for housing 2 5
Unaccountability of government employees 2 1
Passive beneficiaries (entitlement culture) 1 5
Lack of technical skills (Engineers and Artisans, etc.) 3 5
Urbanisation 1 5
Informal settlements 1 5
Poor planning and coordination 1 1
Ideological constraints 1 5
Environmental impact of subsidy housing in South Africa 1 5
Limited participation of the private sector’s involvement in low
cost housing development
1 3
Sustainability of low cost housing programmes 1 5
Housing backlog 1 5
Public private Partnership in housing delivery 1 5
Resources 1 5
Staff expectations 1 4
Challenges from the dissatisfied public, including legal cases 1 2
Supply of water and electricity 1 5
573
Willingness to work with flexible, multiple housing delivery
programmes suited to local circumstances
1 5
Q2.29. Unemployment has an impact on the housing satisfaction of the low-income group?
Agreement and Median value Round 1
Agreement 1 - Strongly disagree 11.0%
2 - Disagree 17.0%
3 - No opinion 0.0%
4 - Agree 44.0%
5 - Strongly agree 28.0%
Comments:
Q2.30. Prior exposure to what is to be received has a tendency to influence beneficiaries’
satisfaction towards a given housing product?
Agreement and Median value Round 1
Agreement 1 - Strongly disagree 0.0%
2 - Disagree 0.0%
3 - No opinion 6.0%
4 - Agree 76.0%
5 - Strongly agree 18.0%
Comments:
Q2.31. Which of the following housing order needs should be met in order to satisfy the
South Africa low-income housing beneficiaries’? (Rank - 1 being most important, 5
being least important)
Housing needs order Rank Median
Self-actualisation 1 1
Esteem Needs 2 3
Social Needs 3 2
Safety Needs 4 2
Physiological 5 3
Comment:
Q2.32. The provision of housing is the paramount need of the poor and low-income group in
South Africa?
Agreement and Median value Round 1
Agreement 1 - Strongly disagree 12.0%
2 - Disagree 24.0%
3 - No opinion 0.0%
4 - Agree 18.0%
5 - Strongly agree 47.0%
Comments:
574
Q2.33. If you disagreed to the previous statement, what do you think should be the
paramount needs of the poor and low-income group in South Africa that the
government should make a priority?
See listed paramount need of the poor and low-income group in South Africa
Issue Statement Frequency Rank
Employment 3 2
Food security 3 2
Clothing 1 3
Job creation 2 1
Health security 1 1
Education 1 1
Transportation 1 2
Comment:
Q2.34. Participation in the housing development process potentially leads to the
implementation of appropriate responses through the involvement of locals in collective
decision making – through the assessment of their needs and expectations?
Agreement and Median value Round 1
Agreement 1 - Strongly disagree 12.0%
2 - Disagree 6.0%
3 - No opinion 0.0%
4 - Agree 59.0%
5 - Strongly agree 24.0%
Comments:
Q2.35. The current South Africa public housing system can be referred to as:
Agreement and Median value Round 1
Agreement 1 - Developmental 0.0%
2 - Welfare 93.0%
3 - No opinion 7.0%
Comments:
Q2.36. The current urban and housing planning system in South Africa favours only the
bourgeois interest in the society?
Agreement and Median value Round 1
Agreement 1 - Strongly disagree 19.0%
2 - Disagree 31.0%
3 - No opinion 19.0%
4 - Agree 25.0%
5 - Strongly agree 6.0%
Comments:
575
Thank you for taking your time to fill out this second round survey. The third round of the
Delphi process, which is a feedback of the response from this Round (Q2) will be sent to you
by March 23, 2012.
Please do not hesitate to contact me or my promoter Professor Wellington Didibhuku Thwala
if you have any questions about this survey and the research project in general. See contact
details below.
Contact details:
Clinton Aigbavboa
Ph.D. Candidate
Dept. of Construction Management and Quantity Surveying, University of Johannesburg
Doornfontein Campus 2028; Johannesburg, South Africa.
Tel.: +27 11 559 6398, Mobile: +27 78 795 8231, Email: [email protected];
Promoter
Professor Wellington Didibhuku Thwala PhD (Eng.) Pr CPM, MCIOB, Pr. Pln, MSAPI
Professor of Construction Project Management
Masters Programme Co-ordinator
Vice-President of The South African Council for Project and Construction Management
Professions (SACPCMP)
Tel: +27 (0)11 559 6048, Fax: +27 (0) 11 559 6630, Fax to email: 086 219 1096, Mobile: +27
83 383 5537, Email: [email protected]
576
APPENDIX F
INSTRUCTIONS TO EXPERTS ON DELPHI STUDY ROUND 3
DELPHI SURVEY – ROUND 3 (Q3)
Thank you for completing Round 2 of the Delphi survey. I acknowledge your busy schedule
and the significant time that you have invested in this survey. I appreciate your time and effort.
This Round 3 survey concludes the Delphi process for this study. The purpose of Round 3 is
to provide you with a final chance to change your response, if desired, given the group median
response of items that falls away by two units, calculated using a 10 point scale.
This is intended to take approximately 15-20 minutes as you are only being asked to review
your previous responses on items that have not attain consensus, i.e. falls out of the collective
group median. Please e-mail your completed survey to [email protected] by the 13 April
2012.
INSTRUCTIONS
The instructions for this Round 3 survey are nearly identical to that of the Round 2 survey. The
difference between this survey and the Round 2 survey is that only items which current falls
two or more units from the grouped median are only presented. In Round 2 all panelist
were asked to provide reasons if their responses were more than one unit above or below the
group median of each statement.
In this Round 3 survey, for each element you will see 2 values: your response from the Round
2 survey (indicated with a yellow highlighted box), and the group median from the Round 2
survey indicated in the column to the far right hand of each table. Please take one of the
following three actions for each category:
1. Accept the group median response by leaving the field completely unchanged.
2. Maintain your original response by placing an ‘X’ in the highlighted field*.
3. Indicate a new response by placing an ‘X’ in the appropriate field*.
* *If your response is more than twenty percent (two units) above or below the group median
please provide a reason for your outlying response in the field provided.
* * Please kindly review the questions without an initial response from the Round 2 survey
(indicated with an orange highlighted shading); – Indicate a response by placing an ‘X’ in the
appropriate field; or accept the group median response by leaving the field completely
unchanged.
We URGE you to review and consider the median and the responses provided by the other
expert panelist when considering your final responses for each element.
577
APPENDIX G
RESEARCH INTRODUCTION LETTER AND QUESTIONNAIRE
UNIVERSITY OF JOHANNESBURG
FACULTY OF ENGINEERING AND THE BUILT ENVIRONMENT
May 2012
Dear Respondent,
The Department of Construction Management and Quantity Surveying at the University of
Johannesburg is undertaking a research project to develop a residential satisfaction model for
government subsidised low-income housing in developing country, a case study of South
Africa. To this end we kindly request that you complete the following short questionnaire. It
should take no longer than 30 minutes of your time. Your response is of the utmost importance
to us.
To protect your anonymity, please do not enter your name or contact details on the
questionnaire.
Summary results of this research will be available in the Department of Construction
Management and Quantity Surveying in December 2012.
Should you have any queries or comments regarding this survey, you are welcome to contact
us telephonically at 011 559 6048, 083 383 5537 or email us at [email protected].
Yours sincerely,
Aigbavboa C. O.
University of Johannesburg
Tel: +27 11 559 6398, Mobile: +27 78 795 8231, email: [email protected]
578
QUESTIONNAIRE TO EVALUATE RESIDENTIAL SATISFACTION IN
PUBLICLY FUNDED HOUSING SUBSIDY SCHEME IN DEVELOPING
COUNTRIES: A CASE STUDY OF SOUTH AFRICA
PLEASE ANSWER THE FOLLOWING QUESTIONS BY CROSSING (X) THE
RELEVANT BLOCK OR WRITING DOWN YOUR ANSWER IN THE SPACE
PROVIDED.
SECTION A: BIOGRAPHICAL DATA
1. Metropolitan / District Municipality City:__________________________
2. Subsidised housing location:_____________________________________
3. What is this area called:_________________________________________
4. What is your Gender?
Male 1
Female 2
5. What is your marital status?
Married 1
Single (never married) 2
Single (married but separated from spouse) 3
Living together (co-habiting) 4
Divorced 5
Widow 6
Widower 7
6. What is your current employment status?
Employed (full time) 1
Employed (Part-time) 2
Self employed 3
Unemployed, looking for work 4
Unemployed, not looking for work 5
Housewife 6
Student 7
Retired 8
Other, specify 9
7. If employed, in what sector of the South Africa economy are you presently employed?
Government 1
Private sector 2
Self employed 3
Other, specify __________
4
579
8. How many employed adults live in this household?________
9. How many unemployed adults live in this household?______
10. What is your age in years?____________
11. What is your ethnicity?
1
2
3
4
5
African Indian Coloured White Other
12. What is your highest level of education?
None (Did not attend any school) 1
Primary (Grade 1-7) 2
Secondary (Grade 8-11) 3
Matric (Grade 12, Std 10) 4
Post Matric Diploma (Registered) 5
Post Matric Diploma (Completed) 6
Bachelor’s / Post-graduate (Registered) 7
Bachelor’s / Post-graduate (Completed) 8
Other, please specify 9
13. What is the annual income for this household (in South African Rand)?
1
2
3
4
5
Less than
R5 000
R5 000 –
10 999
R11 000 –
R15 999
R16 000 –
R20 999
Over
R20 999
14. Does anyone in this household receive any of the following government grants?
Type of grant Yes No
State pension
Child support grant
Disability grant
Foster care grant
Other grant (specify) ___________________
15. What is your greatest need?
Housing 1
Employment 2
Education 3
Safety 4
Privacy 5
16. How long have you been staying in this house?
1
2
3
4
5
Less than 1 year 1 – 2 years 3 - 5 years 6 - 8 years Over 8 years
580
17. How many children (younger than 19 years) live in this household?______
18. How many adults (19-59 years) live in this household?______
19. How many elderly (60 years and older) live in this household?______
20. How many rooms do you have in this house?_________
21. How many bedrooms do you have in this house?________
SECTION B: PRESENCE/ABSENCE OF ROOMS, ITEMS, SPACES, AND SERVICES
22. For each of the attributes listed below please indicate whether or not it is present
in your house:
Room/object (in house) Present Not present
Bedroom(s)
Living room
Dining room
Kitchen
Bathroom(s)
Wardrobes
Play space for children
Study space for children
23. For each of the services listed below please indicate whether or not it is present in
your house:
Service Present Not present
Water for domestic use
Sanitary fittings (e.g. shower,
bath, toilet, basin, taps)
Electricity
24. For each of the facilites listed below please indicate whether or not it is in the
neighbourhood/close to your house:
Facility (Private services) Present Not present
Shopping area
Place of worship
Parking facilities
Playground/recreational facilities
Community hall
Disabled facilities
581
25. For each of the services listed below please indicate whether or not it is accessible
in your area:
Service (Government services) Present Not present
Nursery school (Private or public)
Primary school (Private or public)
High school(Private or public)
Hospital/clinic
Police services
Fire protection services
Public transport
Drainage system (within neighbourhood or outside)
Garbage and waste collection
SECTION C: BENEFICIARY LEVELS OF HOUSING SATISFACTION
Below is a list of attributes relating to room, facility, or service which can be used to
evaluate quality of housing. Using the scale provided, please indicate your level of
satisfaction/dissatisfaction for each attribute (whether it is present/close to your house, or
not).
26. DWELLING UNIT FEATURES (DUF)
Code
How satisfied or dissatisfied are
you with:
Extent to which you are satisfied
1
Very
dissatisfied
2
Dissatisfied
3
Neither
satisfied nor
dissatisfied
4
Satisfied
5
Very
satisfied
DUF1 Location of bedroom
DUF2 Number of bedrooms
DUF3 Size of the bedroom(s)
DUF4 Location of living room
DUF5 Location of dining room
DUF6 Location of kitchen
DUF7 Size of the kitchen
DUF8 Size of bathroom(s)
DUF9 Size of wardrobe/closet
DUF10 Size of children’s play space
DUF11 Size of children’s study space
DUF12 Amount of privacy within the
house
DUF13 Amount of brightness / sunshine
in the house
DUF14 Quality of ventilation in the house
DUF15 Quality of floor level in the house
DUF16 Overall appearance of the house
DUF17 Overall size of the house
582
27. NEIGHBOURHOOD FEATURES (NDF)
Code
How satisfied or dissatisfied are you with:
Extent to which you are satisfied
1
Very
dissatisfied
2
Dissatisfied
3
Neither
satisfied
nor
dissatisfied
4
Satisfied
5
Very
satisfied
NDF1 Location of the dwelling unit in the
neighbourhood
NDF2 Quality of relationships with neighbours
NDF3 Quality of landscaping in the
neighbourhood
NDF4 Quality of walkways
NDF5 Ease of access to main roads
NDF6 Amount of privacy from other neighbours
NDF7 Quality of street lighting at night
NDF8 Amount of security in the neighbourhood
OR Quality of security in the
neighbourhood
NDF9 Physical condition and appearance of the
neighbourhood
NDF10 Cleanliness of the neighbourhood
NDF11 Proximity of house to workplace
NDF12 . . . of house to shopping areas
NDF13 . . . of house to the nursery school
NDF14 . . . of house to the high school
NDF15 . . . of house to hospitals/clinics
NDF16 . . . of house to place of worship
NDF17 . . . of house to Police services
NDF18 . . . to parking facilities
NDF19 . . . of house to the disabled facility
NDF20 . . . of house to the community hall
NDF21 . . . of house to playground / recreational
facilities
NDF22 . . . of house to Public transportation and
services
28. BUILDING QUALITY / HOUSING CONDITION (BQF)
Code
How satisfied or dissatisfied are you
with:
Extent to which you are satisfied
1
Very
dissatisfied
2
Dissatisfied
3
Neither
satisfied
nor
dissatisfied
4
Satisfied
5
Very
satisfied
BQF1 External construction quality
BQF2 Internal construction quality
BQF3 Water pressure
BQF4 Wall quality
BQF5 Floor quality
BQF6 Window quality
BQF7 Door quality
BQF8 Internal painting quality
583
BQF9 External painting quality
BQF10 Plumbing quality
29. BUILDING QUALITY / HOUSING CONDITION (BQF) Continue . . .
Code
How satisfied or dissatisfied are you
with:
Extent to which you are satisfied
1
Very
dissatisfied
2
Dissatisfied
3
Neither
satisfied
nor
dissatisfied
4
Satisfied
5
Very
satisfied
BQF11 the finished quality of sanitary system
BQF12 Plumbing quality
BQF13 electrical wiring quality
BQF14 Electrical fittings quality
BQF15 Numbers of electrical sockets
BQF16 Level of socket
BQF17 Overall unit quality
30. SERVICES PROVIDED BY GOVERNMENT (SPG)
Code
How satisfied or dissatisfied are you with:
Extent to which you are satisfied
1
Very
dissatisfied
2
Dissatisfied
3
Neither
satisfied
nor
dissatisfied
4
Satisfied
5
Very
satisfied
SPG1 The drainage system
SPG2 The Garbage and waste collection
SPG3 The fire protection services
SPG4 Electricity supply
SPG5 Water supply
SPG6 Telephone service
SPG7 Safety
SPG8 How well resident complaints are
handled
SPG9 Government response to building defects
SPG10 Housing Department (Human
Settlement) officials treatment of
beneficiaries
SPG11 Housing Department rules and
regulations
SPG12 Enforcement of rules by the Department
of Human Settlement (Housing)
SPG13 Overall services provided by the
government
584
31. BENEFICIARY PARTICIPATION (BNP)
Code
Beneficiary (owner) participation
Owners should be consulted . . .
Extent to which you agree or disagree
1
Strongly
Disagree
2
Disagree
3
Neither
agree nor
disagree
4
Agree
5
Strongly
Agree
BNP1 . . . about the housing location
BNP2 . . . about the house design
BNP3 . . . about the house construction
BNP4 . . . about the internal finishes of the house
BNP5 . . . about the external finishes of the house
32. NEEDS AND EXPECTATION (NAE)
Code
Needs and expectation
Owners should be . . .
Extent to which you agree or disagree
1
Strongly
Disagree
2
Disagree
3
Neither
agree nor
disagree
4
Agree
5
Strongly
Agree
NAE1 . . . told beforehand the type of house they
will receive
NAE2 . . . asked the type of house they need
NAE3 Owners expect good quality houses
NAE4 Our houses should meet our family need
33. RESIDENTIAL SATISFACTION (RS)
Code
Residential satisfaction
Extent to which you agree or disagree
1
Strongly
Disagree
2
Disagree
3
Neither
agree nor
disagree
4
Agree
5
Strongl
y Agree
RS1 I am satisfied living here
RS2 I am taking proper care of my house
RS3 I am taking proper care of my neighbourhood
RS4 I am constantly maintaining my house
RS5 I am not intending to move to another place in the
future
RS6 I like to live in another place like this
RS7 I will recommend to my friends to obtain a house
in the same way that I did
Thank you for your contribution. We really value your contribution and time spent on
completing this questionnaire. If you have any queries, please do not hesitate to contact the
undersigned.
585
Clinton Aigbavboa
University of Johannesburg
Tel: +27 11 559 6398; Mobile: +27 78 795 8231; email: [email protected];
586
APPENDIX H
MODEL 2.0 RESIDUAL COVARIANCE MATRIXES (S-SIGMA)
Residual covariance matrix for the full structural model (Unstandardized)
DUF1 DUF2 DUF3 DUF5 DUF9
V61 V62 V63 V65 V69
DUF1 V61 0.000
DUF2 V62 0.013 0.000
DUF3 V63 0.039 0.020 0.000
DUF5 V65 -0.015 0.026 -0.012 0.000
DUF9 V69 0.028 -0.042 -0.007 0.156 0.000
DUF12 V72 0.018 -0.027 -0.010 -0.008 0.103
DUF16 V76 -0.025 -0.060 -0.039 -0.056 -0.047
DUF17 V77 -0.043 0.003 -0.021 -0.019 -0.042
NDF1 V78 0.141 0.056 0.032 -0.023 0.085
NDF3 V80 0.100 0.017 0.038 -0.063 -0.051
NDF5 V82 -0.008 -0.109 -0.094 -0.138 -0.019
NDF7 V84 -0.120 -0.143 -0.126 -0.158 0.029
NDF10 V87 -0.015 0.006 -0.037 -0.045 0.141
BQF2 V101 0.100 0.085 0.124 -0.051 -0.050
BQF3 V102 -0.090 -0.175 -0.084 -0.292 -0.277
BQF4 V103 0.041 0.010 0.100 -0.096 -0.096
BQF5 V104 -0.028 -0.053 0.041 -0.129 -0.122
BQF10 V109 -0.157 -0.164 -0.098 -0.209 -0.226
BQF11 V110 -0.254 -0.313 -0.211 -0.338 -0.223
SPG8 V124 -0.057 0.055 -0.020 -0.001 -0.078
SPG9 V125 -0.220 -0.050 -0.134 -0.027 -0.127
SPG12 V128 -0.174 -0.004 -0.082 0.000 -0.099
SPG13 V129 0.047 0.161 0.055 0.111 0.002
BNP1 V130 0.132 0.012 0.056 -0.011 0.105
BNP2 V131 0.068 -0.027 0.051 -0.033 0.098
BNP3 V132 0.001 -0.067 -0.005 -0.050 0.076
BNP4 V133 0.092 -0.035 0.021 -0.039 0.095
NAE1 V135 0.153 -0.001 0.053 -0.024 0.080
NAE2 V136 0.020 -0.122 -0.055 -0.131 0.037
NAE3 V137 0.152 -0.034 0.036 -0.064 0.004
NAE4 V138 0.053 -0.069 -0.030 -0.055 0.030
RS1 V139 -0.119 -0.083 -0.113 -0.159 -0.081
RS3 V141 -0.094 -0.113 -0.098 -0.091 -0.038
RS5 V143 0.235 0.271 0.262 0.168 0.287
RS7 V145 0.207 0.197 0.184 0.114 0.177
DUF12 DUF16 DUF17 NDF1 NDF3
V72 V76 V77 V78 V80
DUF12 V72 0.000
DUF16 V76 0.000 0.000
DUF17 V77 0.005 0.144 0.000
NDF1 V78 0.150 0.197 0.074 -0.026
NDF3 V80 0.082 0.250 0.119 0.061 0.000
NDF5 V82 -0.018 0.084 -0.058 -0.011 0.003
NDF7 V84 0.014 0.129 -0.094 -0.093 -0.081
NDF10 V87 0.109 -0.001 0.022 0.000 -0.110
BQF2 V101 0.026 0.442 0.163 0.002 0.131
BQF3 V102 -0.184 0.039 -0.111 -0.234 -0.015
BQF4 V103 -0.012 0.414 0.132 -0.069 0.129
BQF5 V104 -0.063 0.402 0.077 -0.056 0.175
587
BQF10 V109 -0.140 0.276 -0.066 -0.005 0.195
BQF11 V110 -0.171 0.100 -0.226 0.066 0.192
SPG8 V124 0.052 0.221 0.112 -0.055 0.032
SPG9 V125 -0.086 0.177 -0.013 -0.058 0.059
SPG12 V128 -0.065 0.201 0.016 -0.030 0.032
SPG13 V129 0.068 0.284 0.184 -0.015 0.045
BNP1 V130 0.060 0.130 -0.057 0.037 0.017
BNP2 V131 0.036 0.093 -0.085 -0.026 -0.074
BNP3 V132 -0.017 0.043 -0.118 -0.063 -0.092
BNP4 V133 -0.001 0.098 -0.069 -0.009 -0.031
NAE1 V135 0.101 0.197 -0.013 -0.007 -0.041
NAE2 V136 -0.006 0.085 -0.122 -0.063 -0.090
NAE3 V137 0.016 0.146 -0.066 0.031 0.008
NAE4 V138 0.025 0.125 -0.029 0.068 0.063
RS1 V139 -0.010 0.085 -0.095 0.005 -0.007
RS3 V141 -0.026 -0.111 -0.158 -0.029 -0.069
RS5 V143 0.334 0.090 0.263 0.010 -0.115
RS7 V145 0.164 0.296 0.194 0.063 0.028
NDF5 NDF7 NDF10 BQF2 BQF3
V82 V84 V87 V101 V102
NDF5 V82 0.000
NDF7 V84 0.042 0.000
NDF10 V87 0.013 0.144 0.000
BQF2 V101 -0.089 -0.039 -0.060 0.000
BQF3 V102 -0.064 -0.334 -0.247 -0.062 0.000
BQF4 V103 -0.036 -0.034 -0.140 0.029 0.015
BQF5 V104 -0.041 0.019 -0.159 -0.010 0.024
BQF10 V109 -0.047 -0.008 -0.185 -0.027 0.090
BQF11 V110 0.044 0.126 -0.023 -0.039 0.117
SPG8 V124 -0.069 0.038 0.006 -0.039 -0.245
SPG9 V125 -0.063 0.091 -0.021 0.028 -0.260
SPG12 V128 -0.078 0.114 -0.031 0.013 -0.226
SPG13 V129 -0.087 0.003 -0.049 0.101 -0.187
BNP1 V130 0.032 0.110 0.055 0.101 -0.204
BNP2 V131 -0.006 0.027 0.057 0.052 -0.229
BNP3 V132 -0.040 0.080 0.076 0.013 -0.302
BNP4 V133 0.021 0.077 -0.043 0.054 -0.247
NAE1 V135 0.004 0.066 -0.028 0.049 -0.234
NAE2 V136 -0.048 0.013 -0.054 -0.072 -0.280
NAE3 V137 0.057 -0.091 -0.219 -0.007 -0.122
NAE4 V138 0.128 0.070 0.000 0.040 -0.142
RS1 V139 -0.010 0.122 0.042 0.020 -0.162
RS3 V141 0.007 -0.008 0.085 -0.111 -0.105
RS5 V143 -0.051 -0.002 0.154 -0.209 -0.403
RS7 V145 -0.010 0.105 -0.041 0.035 -0.250
BQF4 BQF5 BQF10 BQF11 SPG8
V103 V104 V109 V110 V124
BQF4 V103 0.000
BQF5 V104 0.019 0.000
BQF10 V109 -0.039 -0.014 0.000
BQF11 V110 -0.064 -0.026 0.161 0.000
SPG8 V124 -0.057 -0.091 -0.004 -0.010 0.000
SPG9 V125 0.039 -0.035 0.061 0.014 0.044
SPG12 V128 0.019 -0.008 0.075 -0.005 -0.049
SPG13 V129 0.062 0.003 0.047 -0.075 0.008
BNP1 V130 0.048 0.042 0.135 0.197 -0.017
BNP2 V131 -0.044 -0.038 0.052 0.165 -0.081
588
BNP3 V132 -0.072 -0.061 0.023 0.137 -0.054
BNP4 V133 -0.035 0.002 0.113 0.159 -0.071
NAE1 V135 0.009 0.032 0.108 0.175 -0.012
NAE2 V136 -0.178 -0.115 0.041 0.157 -0.095
NAE3 V137 -0.040 -0.004 0.117 0.124 -0.125
NAE4 V138 0.018 0.042 0.175 0.194 -0.070
RS1 V139 0.091 0.069 0.122 0.253 0.011
RS3 V141 -0.110 -0.130 -0.083 0.008 -0.109
RS5 V143 -0.239 -0.257 -0.273 -0.146 -0.074
RS7 V145 0.028 0.106 0.063 -0.013 -0.036
SPG9 SPG12 SPG13 BNP1 BNP2
V125 V128 V129 V130 V131
SPG9 V125 0.000
SPG12 V128 0.002 0.000
SPG13 V129 -0.030 0.027 0.000
BNP1 V130 0.054 0.162 0.042 0.000
BNP2 V131 -0.007 0.051 -0.078 0.001 0.000
BNP3 V132 0.017 0.068 -0.066 -0.016 0.017
BNP4 V133 -0.015 0.056 -0.047 -0.031 -0.018
NAE1 V135 0.077 0.150 0.035 0.135 0.033
NAE2 V136 -0.005 0.026 -0.120 0.055 0.011
NAE3 V137 -0.051 0.022 -0.051 0.019 -0.012
NAE4 V138 -0.001 0.073 -0.021 0.010 -0.065
RS1 V139 0.079 0.132 -0.011 0.086 -0.006
RS3 V141 -0.099 -0.064 -0.125 0.111 0.052
RS5 V143 -0.174 -0.145 -0.078 -0.070 -0.157
RS7 V145 0.005 0.065 0.065 -0.022 -0.101
BNP3 BNP4 NAE1 NAE2 NAE3
V132 V133 V135 V136 V137
BNP3 V132 0.000
BNP4 V133 0.016 0.000
NAE1 V135 -0.029 0.026 0.000
NAE2 V136 -0.009 0.035 0.044 0.000
NAE3 V137 -0.091 0.048 -0.052 -0.002 0.000
NAE4 V138 -0.061 0.014 -0.029 -0.054 0.104
RS1 V139 0.027 0.032 0.082 0.040 -0.010
RS3 V141 0.072 0.046 0.033 0.024 0.002
RS5 V143 -0.119 -0.122 -0.131 -0.156 -0.226
RS7 V145 -0.098 -0.035 -0.053 -0.188 -0.103
NAE4 RS1 RS3 RS5 RS7
V138 V139 V141 V143 V145
NAE4 V138 0.000
RS1 V139 0.094 0.043
RS3 V141 0.053 0.021 0.003
RS5 V143 -0.069 -0.019 0.084 0.004
RS7 V145 -0.010 -0.013 -0.012 0.163 0.011
Average absolute residual = 0.0775
Average off-diagonal absolute residual = 0.0820
589
Residual Covariance Matrix for the full structural model (Standardized)
DUF1 DUF2 DUF3 DUF5 DUF9
V61 V62 V63 V65 V69
DUF1 V61 0.000
DUF2 V62 0.010 0.000
DUF3 V63 0.032 0.017 0.000
DUF5 V65 -0.014 0.025 -0.012 0.000
DUF9 V69 0.026 -0.041 -0.007 0.177 0.000
DUF12 V72 0.014 -0.023 -0.009 -0.008 0.101
DUF16 V76 -0.021 -0.054 -0.036 -0.059 -0.049
DUF17 V77 -0.034 0.003 -0.019 -0.019 -0.042
NDF1 V78 0.130 0.055 0.032 -0.027 0.097
NDF3 V80 0.092 0.016 0.038 -0.071 -0.058
NDF5 V82 -0.008 -0.109 -0.098 -0.160 -0.022
NDF7 V84 -0.089 -0.112 -0.102 -0.144 0.027
NDF10 V87 -0.012 0.005 -0.032 -0.043 0.136
BQF2 V101 0.074 0.066 0.100 -0.047 -0.046
BQF3 V102 -0.071 -0.147 -0.073 -0.285 -0.271
BQF4 V103 0.030 0.008 0.081 -0.087 -0.087
BQF5 V104 -0.021 -0.042 0.034 -0.120 -0.114
BQF10 V109 -0.114 -0.126 -0.079 -0.188 -0.203
BQF11 V110 -0.190 -0.246 -0.173 -0.311 -0.206
SPG8 V124 -0.048 0.049 -0.019 -0.001 -0.081
SPG9 V125 -0.180 -0.043 -0.120 -0.027 -0.129
SPG12 V128 -0.145 -0.003 -0.074 0.000 -0.102
SPG13 V129 0.040 0.142 0.051 0.115 0.002
BNP1 V130 0.095 0.009 0.045 -0.010 0.094
BNP2 V131 0.050 -0.021 0.041 -0.029 0.089
BNP3 V132 0.001 -0.051 -0.004 -0.045 0.068
BNP4 V133 0.069 -0.027 0.017 -0.036 0.088
NAE1 V135 0.114 0.000 0.043 -0.022 0.073
NAE2 V136 0.013 -0.087 -0.041 -0.109 0.031
NAE3 V137 0.115 -0.027 0.030 -0.060 0.003
NAE4 V138 0.043 -0.059 -0.027 -0.055 0.030
RS1 V139 -0.089 -0.065 -0.093 -0.146 -0.075
RS3 V141 -0.109 -0.137 -0.124 -0.130 -0.054
RS5 V143 0.178 0.216 0.218 0.156 0.268
RS7 V145 0.144 0.144 0.140 0.097 0.152
DUF12 DUF16 DUF17 NDF1 NDF3
V72 V76 V77 V78 V80
DUF12 V72 0.000
DUF16 V76 0.000 0.000
DUF17 V77 0.004 0.132 0.000
NDF1 V78 0.148 0.206 0.073 -0.029
NDF3 V80 0.080 0.261 0.118 0.069 0.000
NDF5 V82 -0.018 0.091 -0.059 -0.013 0.003
NDF7 V84 0.011 0.108 -0.075 -0.085 -0.074
NDF10 V87 0.090 -0.001 0.019 0.000 -0.105
BQF2 V101 0.020 0.370 0.130 0.002 0.119
BQF3 V102 -0.155 0.035 -0.095 -0.230 -0.015
BQF4 V103 -0.009 0.346 0.105 -0.063 0.116
BQF5 V104 -0.051 0.345 0.063 -0.052 0.163
BQF10 V109 -0.109 0.229 -0.052 -0.005 0.175
BQF11 V110 -0.137 0.085 -0.183 0.061 0.176
SPG8 V124 0.047 0.212 0.102 -0.057 0.034
SPG9 V125 -0.076 0.165 -0.011 -0.059 0.060
590
SPG12 V128 -0.058 0.190 0.015 -0.031 0.032
SPG13 V129 0.061 0.272 0.168 -0.015 0.046
BNP1 V130 0.046 0.107 -0.044 0.033 0.015
BNP2 V131 0.028 0.078 -0.067 -0.023 -0.067
BNP3 V132 -0.013 0.036 -0.093 -0.057 -0.082
BNP4 V133 -0.001 0.084 -0.056 -0.009 -0.029
NAE1 V135 0.080 0.167 -0.011 -0.006 -0.038
NAE2 V136 -0.004 0.066 -0.089 -0.053 -0.074
NAE3 V137 0.013 0.126 -0.054 0.029 0.008
NAE4 V138 0.022 0.115 -0.025 0.068 0.063
RS1 V139 -0.008 0.072 -0.077 0.004 -0.006
RS3 V141 -0.032 -0.146 -0.197 -0.041 -0.098
RS5 V143 0.270 0.077 0.215 0.010 -0.107
RS7 V145 0.121 0.234 0.146 0.054 0.024
NDF5 NDF7 NDF10 BQF2 BQF3
V82 V84 V87 V101 V102
NDF5 V82 0.000
NDF7 V84 0.040 0.000
NDF10 V87 0.013 0.112 0.000
BQF2 V101 -0.084 -0.028 -0.046 0.000
BQF3 V102 -0.064 -0.263 -0.206 -0.049 0.000
BQF4 V103 -0.033 -0.025 -0.108 0.021 0.012
BQF5 V104 -0.039 0.014 -0.125 -0.007 0.019
BQF10 V109 -0.043 -0.006 -0.142 -0.019 0.070
BQF11 V110 0.042 0.094 -0.018 -0.029 0.093
SPG8 V124 -0.074 0.032 0.005 -0.032 -0.220
SPG9 V125 -0.066 0.074 -0.018 0.023 -0.227
SPG12 V128 -0.083 0.094 -0.027 0.011 -0.200
SPG13 V129 -0.093 0.003 -0.043 0.084 -0.168
BNP1 V130 0.029 0.079 0.042 0.072 -0.156
BNP2 V131 -0.005 0.020 0.044 0.038 -0.179
BNP3 V132 -0.036 0.058 0.058 0.009 -0.234
BNP4 V133 0.020 0.058 -0.034 0.040 -0.198
NAE1 V135 0.004 0.049 -0.022 0.036 -0.186
NAE2 V136 -0.041 0.009 -0.038 -0.048 -0.201
NAE3 V137 0.055 -0.069 -0.173 -0.005 -0.098
NAE4 V138 0.132 0.056 0.000 0.032 -0.123
RS1 V139 -0.010 0.091 0.033 0.015 -0.129
RS3 V141 0.011 -0.009 0.103 -0.127 -0.130
RS5 V143 -0.049 -0.002 0.122 -0.156 -0.325
RS7 V145 -0.009 0.073 -0.030 0.024 -0.185
BQF4 BQF5 BQF10 BQF11 SPG8
V103 V104 V109 V110 V124
BQF4 V103 0.000
BQF5 V104 0.014 0.000
BQF10 V109 -0.028 -0.011 0.000
BQF11 V110 -0.047 -0.019 0.118 0.000
SPG8 V124 -0.047 -0.078 -0.003 -0.009 0.000
SPG9 V125 0.031 -0.029 0.049 0.011 0.040
SPG12 V128 0.016 -0.006 0.061 -0.004 -0.046
SPG13 V129 0.052 0.003 0.039 -0.064 0.008
BNP1 V130 0.034 0.031 0.095 0.143 -0.014
BNP2 V131 -0.032 -0.028 0.037 0.122 -0.067
BNP3 V132 -0.051 -0.045 0.017 0.100 -0.044
BNP4 V133 -0.026 0.002 0.083 0.119 -0.061
NAE1 V135 0.007 0.024 0.079 0.131 -0.010
NAE2 V136 -0.119 -0.079 0.027 0.106 -0.073
591
NAE3 V137 -0.030 -0.003 0.086 0.094 -0.107
NAE4 V138 0.015 0.035 0.139 0.158 -0.065
RS1 V139 0.067 0.052 0.089 0.189 0.010
RS3 V141 -0.126 -0.153 -0.094 0.009 -0.142
RS5 V143 -0.178 -0.197 -0.203 -0.111 -0.063
RS7 V145 0.019 0.075 0.043 -0.009 -0.028
SPG9 SPG12 SPG13 BNP1 BNP2
V125 V128 V129 V130 V131
SPG9 V125 0.000
SPG12 V128 0.002 0.000
SPG13 V129 -0.028 0.025 0.000
BNP1 V130 0.043 0.130 0.035 0.000
BNP2 V131 -0.005 0.042 -0.065 0.001 0.000
BNP3 V132 0.013 0.055 -0.054 -0.012 0.012
BNP4 V133 -0.013 0.047 -0.040 -0.022 -0.013
NAE1 V135 0.063 0.124 0.029 0.097 0.024
NAE2 V136 -0.004 0.020 -0.091 0.036 0.008
NAE3 V137 -0.042 0.019 -0.044 0.014 -0.009
NAE4 V138 -0.001 0.066 -0.019 0.008 -0.052
RS1 V139 0.065 0.109 -0.009 0.062 -0.004
RS3 V141 -0.127 -0.082 -0.163 0.125 0.059
RS5 V143 -0.145 -0.122 -0.067 -0.051 -0.117
RS7 V145 0.004 0.050 0.051 -0.015 -0.069
BNP3 BNP4 NAE1 NAE2 NAE3
V132 V133 V135 V136 V137
BNP3 V132 0.000
BNP4 V133 0.012 0.000
NAE1 V135 -0.021 0.020 0.000
NAE2 V136 -0.006 0.024 0.030 0.000
NAE3 V137 -0.067 0.037 -0.039 -0.002 0.000
NAE4 V138 -0.048 0.011 -0.024 -0.040 0.086
RS1 V139 0.020 0.024 0.061 0.027 -0.007
RS3 V141 0.082 0.053 0.038 0.025 0.002
RS5 V143 -0.088 -0.093 -0.099 -0.107 -0.174
RS7 V145 -0.066 -0.025 -0.037 -0.118 -0.073
NAE4 RS1 RS3 RS5 RS7
V138 V139 V141 V143 V145
NAE4 V138 0.000
RS1 V139 0.076 0.032
RS3 V141 0.066 0.025 0.005
RS5 V143 -0.057 -0.015 0.099 0.003
RS7 V145 -0.008 -0.009 -0.013 0.115 0.007
Average absolute standardized residual = 0.0662
Average off-diagonal absolute standardized residual = 0.0700