food

lifebetter

economics

economic

government

gdpdevelopm

ent

Green

rateequity

returns

index

exchangetax

bonds

businessmarket

gold

expectations

arbitrage

markets

gnilaw

gross

product

unionscale

asymmetric

money

efficiency

budget euro

Education

bank

demand

hedge

horizontal

human

expenditure

aid

altruismamortisation

Energy

antitrust

appreciation growth

pricing

theory

Agro - Processing

coefficient

assets

Tourisminformation

shockauctions

austrian

living

Broadband average

backwardation

balance

payments

balanced standards

services

bankruptcy

barriers

entry

exitbarter

cohesionbasis

point

bear

behavioural

beta

Safety

mac

financeblack

economy

black-scholes

boom

bust

bounded

rationality

brand

bretton

woods

bubble

bull

social

confidence

cycle

buyer

responsibility

gini TRT

Health

WiFi

food

lifebetter

economics

economic

government

gdpdevelopm

ent

Green

rateequity

returns

index

exchangetax

bonds

businessmarket

gold

expectations

arbitrage

markets

gnilaw

gross

product

unionscale

asymmetric

money

efficiency

budget euro

Education

bank

demand

hedge

horizontal

human

expenditure

aid

altruismamortisation

Energy

antitrust

appreciation growth

pricing

theory

Agro - Processing

coefficient

assets

Tourisminformation

shockauctions

austrian

living

Broadband average

backwardation

balance

payments

balanced standards

services

bankruptcy

barriers

entry

exitbarter

cohesionbasis

point

bear

behavioural

beta

Safety

mac

financeblack

economy

black-scholes

boom

bust

bounded

rationality

brand

bretton

woods

bubble

bull

social

confidence

cycle

buyer

responsibility

gini TRT

Health

WiFi

food

lifebetter

economics

economic

government

gdpdevelopm

ent

Green

rateequity

returns

index

exchangetax

bonds

businessmarket

gold

expectations

arbitrage

markets

gnilaw

gross

product

unionscale

asymmetric

money

efficiency

budget euro

Education

bank

demand

hedge

horizontal

human

expenditure

aid

altruismamortisation

Energy

antitrust

appreciation growth

pricing

theory

Agro - Processing

coefficient

assets

Tourisminformation

shockauctions

austrian

living

Broadband average

backwardation

balance

payments

balanced standards

services

bankruptcy

barriers

entry

exitbarter

cohesionbasis

point

bear

behavioural

beta

Safety

mac

financeblack

economy

black-scholes

boom

bust

bounded

rationality

brand

bretton

woods

bubble

bull

social

confidence

cycle

buyer

responsibility

gini TRT

Health

WiFi

food

lifebetter

economics

economic

government

gdpdevelopm

ent

Green

rateequity

returns

index

exchangetax

bonds

businessmarket

gold

expectations

arbitrage

markets

gnilaw

gross

product

unionscale

asymmetric

money

efficiency

budget euro

Education

bank

demand

hedge

horizontal

human

expenditure

aid

altruismamortisation

Energy

antitrust

appreciation growth

pricing

theory

Agro - Processing

coefficient

assets

Tourisminformation

shockauctions

austrian

living

Broadband average

backwardation

balance

payments

balanced standards

services

bankruptcy

barriers

entry

exitbarter

cohesionbasis

point

bear

behavioural

beta

Safety

mac

financeblack

economy

black-scholes

boom

bust

bounded

rationality

brand

bretton

woods

bubble

bull

social

confidence

cycle

buyer

responsibility

gini TRT

Health

WiFi

CITY OF TSHWANE

BETTER LIFE INDEX2016

One Nation, One Capital

Advancing Tshwane Vision 2055

CITY OF TSHWANE

CRANE INDEX2016

One Nation, One Capital

Advancing Tshwane Vision 2055

2 ©Copyright City of Tshwane 2016

3©Copyright City of Tshwane 2016

CiTy of Tshwane CRane inDeX

2016

4

eXeCUTiVe sUMMaRy

Following slow global growth rates in 2015, growth prospects in general across the globe remain subdued with South Africa’s growth forecast adjusted downwards by the IMF to 0.6 percent, down from its initial forecast in January 2016 of 0.7 percent.

The key drivers contributing towards slowing global growth include falling oil and commodity prices and slowing demand in China as well as other areas of the East that has led to a faster than expected slowdown in exports and imports in addition to opposing monetary policy movements in the US and Europe.

Within South Africa, we face our own set of challenges, with persistently high unemployment, low-growth and high-inflation characterising our economic landscape.

Tshwane’s performance in this regard is particularly important in the face of ever-increasing urbanisation, with the City receiving an estimated influx of between 50,000 and 60,000 new residents from across the country each year.

Economic indicators within the Capital paint a different picture with our estimated economic growth rate at 2.4 percent, which is considerably higher than the national average.

Furthermore, Tshwane’s unemployment rate is approximately 21.2 percent, also lower than the national average whilst the city-specific inflation figure is on par with that of the national average.

This resilient and above-average economic performance enables Tshwane to contribute at least 27 percent to Gauteng’s economy and over 9 percent to the national economy annually.

Beyond this technical criterion of economic development, the City of Tshwane has designed a South African Metropolitan Better Life Index based on the OECD Better Life Index. The index goes beyond the limits of the narrow technical criteria and considers broader social and developmental indicators.

Based on the Better Life Index, Tshwane is the quickest improving metro in South Africa with the highest growth rate in the index value over the period 2011 to 2015, implying that Tshwane is rapidly becoming the best place to live in South Africa. When we account for public transport and access to digital connectivity, our position within the index is significantly altered in relation to our peer metros.

To consolidate all these gains, Council will, at the end of April, deliberate and ratify the City’s sustainable inclusive growth strategy and its implementation plan. In addition to galvanising the City’s efforts towards accelerating economic transformation, the strategy will also guide and inform the City’s efforts to engage with all necessary stakeholders, including national and provincial government, as well as key industry players and other stakeholders.

5

The strategy prioritises three sectors, namely:

• Theeducationsector,whichhasthepotentialtocontributeanadditionalR37 billionin2030toTshwane’sGrossValueAdd,alsoknownasGVA,andcreatean additional160,000jobs;

• Theagri-businesssector,whichhasthepotentialtocontributeanadditionalR24 billionin2030toTshwane’sGVA,andcreateanadditional120,000jobs;and

• Thetourismsector,whichhasthepotentialtocontributeanadditionalR28billion in2030toTshwane’sGVA,andcreateanadditional70,000jobs.

In the context of sluggish economic performance and turbulent global economic dynamics, the need to leverage both private and public sector investment becomes considerable.

The Capital has seen tremendous growth in the number of building plans approved during the current term of office. The number of building plans approved has doubled between the 2010/11 and 2014/15 financial years, representing a growth rate of 103 percent. This amounted to R81 billion in approved building plans over the period, excluding the value of land, and is despite low economic growth rates characterising the country during this period. In addition, over R15 billion worth of building plans were approved between July 2015 and March 2016 alone. In the last three years, in excess of 8,000 buildings were completed per year in the City as a whole.

The private sector component of these approvals amounts to approximately R71 billion, thus nullifying the argument that Tshwane’s economy is solely driven by the public sector. The private property sector significantly employs over 98,000 people within the Capital.

The highest investment growth rates have been evident in Region 4, that is, Centurion and surrounds; followed by Region 1, that is, the North; and Region 3, namely the inner-city and its surrounds.

Notably, the revitalisation of the inner-city is also occurring with developments focussing on retail, offices and residential units. Since 2012, approximately 150,000m2 of office space has been constructed or is still under construction at a cost of R5 billion with the full implementation of the bus-rapid transit system, A Re Yeng, expected to play a further role.

Importantly, the City has since approved an institutional framework for processing large investments within the City in order to surmount current challenges experienced; inextricably resulting in significant delays for both private and public sector projects across the City.

The City has also approved a Development Investment Incentives Policy in May 2015 that aims to attract catalytic investments into the Capital to fund our strategic trajectory of “Building the City”, through offering a range of direct and indirect incentives to qualifying investments that enhance spatial planning initiatives within the Capital.

We have received applications for investment incentives related to catalytic investments amounting ```to more than R40 billion. Of these, direct investment incentives have already been approved for two major investment projects, representing new investment into the City in excess of R2 billion.

6

Given the crucial role that development and investment play in any economy the city identified the need to develop an index that captures the developement activity taking place in the city and quantifies it in a meaningful manner. Therefore, this document investigates the development of a crane index for the City of Tshwane, as a leading indicator of economic and construction activity. The principles of construction cost indices and its implications for construction activity in the wider economy are discussed. International literature is reviewed in order to determine what is the current use and applicability of cost indices and also to link this to the development of a crane index. It is aimed have a wide scope literature review to fully understand the usefulness of a crane index, as well as how such an index can be developed and tested for applicability. In addition, the principles of the link between space and capital markets, which dictates the demand for construction to take place, are discussed. Comments are also made on the implications for the use of a crane index within the ambit of market activity.

Data used to evaluate construction activity includes building plans approved and buildings completed. Both these are provided as the area in square meter as well as the Rand value. By dividing the one by the other gives an indication of the construction cost per square meter at any given time. This result is, however, a combination of construction cost as well as construction quality. From this, it would be possible to see construction cycles that is based on quality and cost, with the expectation that higher quality space is being provided in a better economic climate.

Cranes are not that often utilised and also not in all types of construction. As cranes are, however, used in prime construction, it is an indication that the market is at its highest levels and reached a point where substantial output both in quality and quantity is evident. A lack of cranes would, however, not indicate that the market is not performing or inactive, but merely that a specific type of construction is not taking place. For this purpose the Crane Index might prove valuable and can provide valuable information about the health of the market. It was found that a 12 to 8 month lag provides the closest correlation between the 2 data sets and is therefore considered indicative of the average time taken from approval of building plans to completion of construction.

Tshwane crane index as correlated to building plans passed are closest correlated with a 6 month lag. The result shows an R2 of 0.5989, which is considered a fair indication of construction activity, but with the lag being reduced from 12 months to 6 months, it is evidently a timelier, but less ac-curate measure of construction activity

7

8

9

Table of ConTenTs

1. spaCe anD CapiTal MaRkeTs......................................................13

1.1 Background ...................................................................................................14

1.2 The FDW model .............................................................................................14

1.3 The REEFM model ........................................................................................17

1.4 Development of a Crane Index ......................................................................21

2. inDiCes in GeneRal...........................................................................23

2.1 Introduction ....................................................................................................24

2.2 Types of Price indices ....................................................................................25

2.3 Indices in the Building Industry ......................................................................28

2.4 Factors Influencing the Composition of an Index ..........................................32

2.5 Use of Indices ................................................................................................36

2.6 Challenges associated with Indices ...............................................................38

2.7 Summary and Conclusion .............................................................................39

3. CRane inDeX...........................................................................................41

3.1 Introduction ....................................................................................................42

3.2 Methodology ..................................................................................................42

3.3 Advantages and Disadvantages ....................................................................47

3.4 Conclusion ......................................................................................................49

10

4.analysis of Tshwane speCifiC DaTa........................................50

4.1 Tshwane Crane Index ....................................................................................53

4.2 Tshwane Crane Index Compared to Construction Activity .............................56

4.3 Tshwane Crane Index Compared to Construction Cost .................................57

5. ConClUsion...........................................................................................60

6. RefeRenCes................................................................................................62

list of figuresFigure 1: FDW-model functions (Archour-Fischer, 1999: 38-39) .............................16

Figure 2: REEFM’s conceptual framework (Viezer, 1998: 107) ................................18

Figure 3: Cost trends incurred on a project (Eurostat, 2008) ..................................29

Figure 4: Source: Statistics South Africa ..................................................................43

Figure 5: Source: Statistics South Africa .................................................................43

Figure 6: Source: Authors’ calculations ....................................................................44

Figure 7: Source: Statistics South Africa .................................................................45

Figure 8: Source: Authors’ calculations ...................................................................45

Figure 9: Source: Authors’ calculations ...................................................................46

Figure 10: Source: Authors’ calculations ..................................................................46

Figure 11: Affordability vs Building Plan Lags .......................................................50

Figure 12: Corporate Savings vs Gross Fixed Capital Formation .........................51

Figure 13: Source: Authors’ calculations .................................................................51

Figure 14: Source: Authors’ calculations ..................................................................53

Figure 15: Source: RLB ...............................................................................................54

Figure 16: Source:(Authors’ calculations) ................................................................55

11

Figure 17: Source:(Authors’ calculations) ...............................................................56

Figure 18: Source:(Authors’ calculations) ..............................................................57

Figure 19: Source:(Authors’ calculations) ...............................................................57

Figure 20: Source:(Authors’ calculations) ...............................................................58

Figure 21: Source:(Authors’ calculations) ..............................................................58

list of TablesTable 1: Number of cranes 2011 to May 2016(Authors’ primary data collection) ..52

Table 2: : Index: 2011 = 100(Authors’ calculations) .................................................52

Table 3: Crane Index Year-on-YearPercentage Change (Authors’ calculations) ...53

Table 4: RLB Crane Index(RLB) ..................................................................................54

12 ©Copyright City of Tshwane 2016

13©Copyright City of Tshwane 2016

1. spaCe anD CapiTal MaRkeTs

14

1.1 baCkGRoUnD

The unique characteristics of real estate create, on the one hand, many opportunities for real estate investors, and, on the other, many difficulties. The different factors influencing the behaviour of real estate should therefore be investigated carefully.

DiPasquale and Wheaton (1992: 181) stated that analysing the market for real estate presents challenges because of the inter-relation of space- and asset markets.

The earliest recording of work that distinguishes between use decisions and investment decisions with respect to real estate was probably Weimer (1966), but Hendershott and Ling (1984) were the first to integrate space- and capital markets into real estate. According to Viezer (1999:504), Hendershott and Ling’s model evaluated investment value responses to tax code alterations in a dynamic programming algorithm that used a traditional discounted cash-flow equation with assumed parameters.

Corcoran (1987) graphed the space market and capital market of real estate separately, but interdependently, explicitly distinguishing between the short- and long-run supply of space. A similar model was published by Fisher (1992: 167). Fisher shows the equilibrium existing between the short- and long-run situations of the space and capital markets.

DiPasquale and Wheaton (1992) and Fisher, Hudson-Wilson and Wurtzebach (1993) further refined this model, which is referred to as the diagrammatic model by Viezer (1999: 504)). The model was officialized in a textbook on property economics by DiPasquale and Wheaton (1992) as the FDW-model, the most detailed treatment found in a seminal textbook.

Du Toit (2002) carried out research on the FDW-model and describes the principles of the model with an accompanying practical example of office space in Pretoria. The FDW-model conceptualizes the interrelationships between the market for space, asset valuation, construction sector and stock adjustment.

Viezer (1998) developed a completely new model that similarly describes the space and asset markets in the property sector, but this model is of an econometric rather than diagrammatic nature. Viezer refers to it as the Real Estate Econometric Forecast Model (REEFM), and uses statistical principles to explain the property market, in contrast with the diagrammatical FDW-model.

1.2 The fDw MoDel

1.2.1 The fDw-MoDel DefineD

Archour-Fischer (1999:33) states that the Fischer-DiPasquale-Wheaton model is an elegant metaphor that integrates the different markets in the built environment, with specific reference to the property market, the capital market and construction activity. Du Toit (2002: 10) describes the FDW model as being a static quadrant model that has the ability to trace the relationships between real estate market and asset market variables. Archour-Fischer also suggests that it is

15

a dynamic model (Archour-Fischer, 1999: 40-42), in which the parameters of the model can be changed to determine the influence in the different markets represented by the model, although Viezer (1998) criticizes the application of the model (see section 2.3).

Taking into consideration the flow of real estate as discussed by DiPasquale and Wheaton, it is evident that the depreciation of real estate and the subsequent replacement of such depreciation is an output of the model, as it is a reduction in the stock level seen in quadrant four of the model. The reduction and replacement cause a shift in the supply and demand patterns so that the market reacts to it. It thus acts as an input to the rest of the model. The model reacts to the changes and further depreciation takes place, resulting in a change in the then present stock level.

Figure 1 show a graphical illustration of the model, which consists of four quadrants, and represents the following (Archour-Fischer, 1999: 34 – 37):

Quadrant 1 – Demand function on the market for space;

Quadrant 2 – The valuation function;

Quadrant 3 – The construction function;

Quadrant 4 – The adjustment supply.

Quadrant 1 indicates the demand function on the market for space demanded by users. With a static supply, the price of space or rent level will increase when demand increases, and conversely. In equilibrium, the supply of property should be equal to the demand at various price levels.

In Quadrant 2 the rent level applicable to the equilibrium level of demand is discounted at the capitalisation rate, which is illustrated in Figure 1 as the slope of the asset valuation curve, to arrive at the asset value, represented by the function P = R/i .

Quadrant 3 represents the construction activity, which is a function of the asset value. When the asset value is higher than construction costs, new F construction will be triggered, otherwise construction will come to a halt. Thus, P = f(C).

The level of construction activity is carried over to Quadrant 4, the adjustment of supply, and is given by the function S = C /d, or ∆S = C – dS.

16

Figure 1: FDW-model functions (Archour-Fischer, 1999: 38-39)

Rent Asset valuation P = R/i

Price (R)

2

1

Market for space D(R, Econ.) = S

Stock (sqm)

Construction sector P = f(C)

3

4

Stock adjustment

S = C/d Construction

(sq2)

Figure 1: Diagrammatic FDW - model

Quadrant 1: Demand for space

R = f 1 (S) S = E(b – a . R)

Quadrant 2: Determination of value P = f 2 (R) P = R/i

Quadrant 3: Construction function

C = f 3 (P) C = (P – β)/α

Quadrant 4: Stock adjustment function S = f 4 (C) S = C/d

Where:

R = rent per unit S = supply E = the number of office workers a and b = demand parameters P = price or value per unit i = the capitalisation rate C = construction α and β = construction parameters d = a depreciation rate

1.2.2 ReMaRks on The fDw MoDel

The FDW model seems to offer an acceptable interpretation of the property market using a diagrammatic model, which is mathematically explained by the developers of the model. However, Viezer (1998) points out that the FDW model is of little value as an investment tool, and he develops a Real Estate Econometric Forecast Model (REEFM) that is able to forecast implicit market returns. The REEFM seems to be of much more value as an investment tool, as it can be used for calculating historical returns and forecasted returns.

17

According to Archer and Ling (1997), a multi-factor asset pricing model should be used to determine the discount rate, which in turn would determine both the market value and the cap rate, rather than assuming that the cap rate is exogenously determined. Viezer (1998) developed an econometric model for the integration of real estate’s space and capital markets – the Real Estate Econometric Forecast Model (REEFM). In his research, he answers the above comment by Archer and Ling by including a stochastic equation for, inter alia, the cap rate. Viezer’s equation contains five predetermined variables four of which are taken, with some modifications, from the pre-specified Arbitrage Pricing Theory (APT) model by Chen, Roll and Ross (1986).

The FDW model is interpreted by Viezer (1998) to suggest that equilibrium is a natural state where all values are determined simultaneously, but in reality there are lags in the adjustment process.Viezer (1999: 507) also modifies the DiPasquale-Wheaton (1992) model by positing that real construction costs are a function of the lagged net change in stock, rather than a current-period new construction

Viezer further points out that the FDW model is of little value in terms of practical advice, and is limited to forecasting the changes in the direction of real estate markets and general levels of return. He maintains that the model should be estimated statistically in individual markets if it is to be useful to the practitioner. The REEFM integrates real estate’s space and capital markets econometrically rather than diagrammatically. The model also links the short- and long-run markets, and calculates implicit market returns for property markets (Viezer, 1998: 143). The model can therefore be used as an effective investment or forecast tool. The only forecast inputs needed are the local economic variables, and national financial variables (Viezer, 1998: 144).

1.3 The ReefM MoDel

1.3.1 pRinCiples of ReefM

The conceptual framework of REEFM is illustrated in Figure 2 (Viezer, 1998: 107). REEFM is a recursive model, containing six stochastic equations (occupancy, real rents, capitalisation rate, market value per unit, change in stock, and real construction costs) and seven deterministic equations (a net operating income proxy, market value per unit, stock of space, vacancy rate, implicit appreciation market return, implicit income market return and implicit total market return) (Viezer, 1998: 134-5).

The six stochastic equations given by Viezer are all in the format:

Yi = αi + β1Xi + εi

where:

αi = Y intercept for the population;

β1 = slope for the population;

εi = random error in Y for observation i.

This relationship is confirmed by two sources, with different formats for the same equation:

18

Yi = β0 + β1Xi + εi (Levine, Berenson & Stephan, 1998: 538)

and;

μy = α + βx (Steyn, Smit & Du Toit, 1989: 378),

which is the function of a straight line relationship between x and y.

Each of the stochastic equations is a variation of the above equation, allowing for the different variables that influence the Yi - factor. In all six equations the variable:

Σt=1T-1δtYRDUMt

is also added, which is missing data indicators to be used in estimating the unbalanced panel (Viezer, 1998: 115).

The deterministic equations are in different formats, calculating a specific result in each case by combining the results from the stochastic equations.

Figure 2: REEFM’s conceptual framework (Viezer, 1998: 107)

D

Local and National

Economic Factors

Occupancy

Occupancy Vacancy

Rent

$ S

SF

Space market

Rent

NOI

CAPITAL MARKET

Discount Rate

Cap

Rate

lag current

Dividend Discount

Model

Market Value

Determination of Construction

Market Value versus Cost

Stock versus Occupancy

Change in Stock

Construction Cost

Arbitrage Pricing Theory

In both the stochastic and deterministic equations, the variables are given in the format Vp.m.t., which in this case would indicate a variable (V) for property type p, in metro area m, at time period t.

19

The different equations will be discussed in the text to follow, and the similarities and differences relating to the FDW model will be explained.

Short-run asset market

OCCt = αt + β1RNT$t-1 + γ1ECONt + Σt=1T-1δtYRDUMt + εt 1

(Viezer, 1998: 115)

RNT$t = αt + β1VACt-1 + Σt=1T-1δtYRDUMt + εt 2

(Viezer, 1998: 115)

Viezer explains occupancy (OCCp.m.t.) as a function of lagged real rent, (RNT$p.m.t, nominal rent deflated by the Consumer Price Index) and an economic variable. The economic variable in the case of office space would be office employment. Real rents, in turn, respond with a lag to vacancies in the market (Viezer, 1998: 114-15). On the contrary, the FDW model only takes the demand as equal to the supply of office space, using only the equation S = E(b – a.R) to calculate this (see section 2.2.2). Rent levels are taken as a given, and are not calculated as above. This means that the variables are calculated according to a much less scientific method, limiting the capabilities of the model for the historical explanation of the market, as well as possibilities for forecasting, which would be much more useful.

The first deterministic equation is a proxy for the net operating income (NOIp.m.t). The net operating income is determined by multiplying the occupancy by the rental levels. However, the rental levels are calculated for real rent in terms of equation 2 and should therefore be inflated by the Consumer Price Index (CPI). The equation for the net operating income is therefore:

NOIp.m.t = OCCp.m.t xRNT$p.m.t x (CPI/100) 3

(Viezer, 1998: 117)

Short-run capital market

The capital market attempts to translate the results of the short-run space market into asset prices. “The reasonably calculated expected future net income flow of an investment property discounted to its present value, when capitalised at the prevailing rate sought by prudent investors, represents the estimated capitalised value of the property at that time” (SAIV, 1999: 6-4). When considering the future income stream, it increases approximately in line with inflation, or the country’s CPI. The income stream can therefore be capitalized by dividing the first year’s income by the capitalisation rate, which is the discount rate minus CPI. As the discount rate is not determined, the Cap rate cannot be determined from the discount rate. However, Viezer determines the Cap rate with the equation:

CAPp.m.t = αp.m.t + β1RISKt + Φ1TERMt + γ1INFLt + η1%ΔECONp.m.t + ξRNTp.m.t-1/MSFp.m.t-1 + Σt=1

T-1δtYRDUMt + εp.m.t 4

(Viezer, 1998: 123)

20

The variable RNTp.m.t-1/MSFp.m.t-1 considers the backward-looking comparisons of appraisers and therefore takes into consideration historical data. %ΔECONp.m.t is the percentage change in the economic variable as used in the equation for occupancy. INFLt is the current inflation rate, while RISKt and TERMt are risk variables used by Viezer as the difference between the corporate Baa bond rate and the 10-year Treasury bond rate, and the difference between the 10-year Treasury bond rate and the 3-month Treasury bill rates respectively (Viezer, 1998: 123, 124).

With the Cap rate established, it is possible to calculate the market value per unit for the metro property stock with the equation:

MSFEp.m.t = NOIp.m.t x (1 + (PASSp.m.t x INFLt)) ÷ STKp.m.t 5 CAPp.m.t

(Viezer, 1998: 124)

The PASS variable indicates the extent to which the inflation rate is passed through to the property appreciation. The MSFE variable is then regressed to determine the actual property value per unit:

MSFp.m.t = αp.m.t + β1MSFEp.m.t-1 + Σt=1T-1δtYRDUMt + εp.m.t 6

(Viezer, 1998: 124)

While the above equations are used to determine the per unit market value of property, the FDW model divides the demand, multiplied by the rate per unit, by the cap rate to get to the market value. While REEFM takes into consideration different risk factors as well as economic variables for calculating the cap rate, the FDW model does not indicate how this is calculated (Du Toit, 2002: 31). From this it is also taken that REEFM calculates the market value of property in a much more scientific way, which creates an opportunity for explaining the current market as well as forecasting future trends.

Long-run space market

The long-run space market is the addition of new stock or construction and the removals of stock or depreciation. These construction and removals are a function of the difference in quantities (STK – OCC) and real prices (MSF$ – CST$) (Viezer, 1998: 132). The asset market is expressed in quantities and the capital market is expressed in prices with the following equations:

Asset market:

NEWp.m.t– RMVp.m.t. = αp.m.t + β1(STKp.m.t-L – OCCp.m.t-L) + γ1(MSF$p.m.t-L – CST$p.m.t-L) + Σt=1

T-1δtYRDUMt + εp.m.t 7

(Viezer, 1998: 132)

The stock of space in the current period takes into consideration the stock in the previous period, plus the current period’s construction, minus the current period’s removals:

21

STKp.m.t = STKp.m.t-1 + NEWp.m.t – RMVp.m.t 8

(Viezer, 1998: 132)

Capital market:

CST$p.m.t. = αp.m.t + β1NEWp.m.t-1 – RMVp.m.t-1+ Σt=1T-1δtYRDUMt + εp.m.t 9

(Viezer, 1998: 132)

In the above, the real construction costs are indicated as being a function of the lagged net change in stock (Viezer, 1998: 132).

The last equation to close the loop for the model is the vacancy rate:

VACp.m.t = 1 – (OCCp.m.t / STKp.m.t) 10

(Viezer, 1998: 133)

The two mentioned models investigate the space and capital markets in real estate.The value of this is in the possibility of applying the model in the South African context in order to monitor property behaviour more closely. As such it could be used successfully to explain specific property economics with regards to geographical areas or different types of property, or even to valuate property in general. A specific application exists in the relationship of property activity to construc-tion cost and how this could result in property cycles or construction cycles.

1.4 DeVelopMenT of a CRane inDeX

In the previous section it was highlighted that construction cost plays an important role in the development of new properties. A distinction should, however, be made between construction costs and construction quality. When considering the FDW model, an increase in construction cost would imply an increase in terms of Quadrant 3 of the model, which is not associated with an increase in the value of the space being created, typically due to higher profit margins, or higher input cost such as material and labour as a result of economic circumstances. Construction qual-ity on the other hand might also result in higher construction cost, but is associated with increased value of space based on the higher specification levels and quality of materials and eventually end-product produced.

Data used to evaluate construction activity includes building plans approved and buildings com-pleted. Both these are provided as the area in square meter as well as the Rand value. By dividing the one by the other gives an indication of the construction cost per square meter at any given time. This result is, however, a combination of construction cost as well as construction quality. From this, it would be possible to see construction cycles that is based on quality and cost, with the expectation that higher quality space is being provided in a better economic climate. This should, however, be evaluated to measure the time-series data of construction against long term space markets. By comparing this to a Crane Index provides the opportunity to evaluate if the market reached a specific point of optimism due to the type of construction that is performed where cranes are being utilised. This might be termed a Crane Status of the construction sector,

22

rather than a Crane Index, due to the fact that cranes are not that often utilised and also not in all types of construction. As cranes are, however, used in prime construction, it is an indication that the market is at its highest levels and reached a point where substantial output both in quality and quantity is evident. A lack of cranes would, however, not indicate that the market is not performing or inactive, but merely that a specific type of construction is not taking place. For this purpose the Crane Index might prove valuable and can provide valuable information about the health of the market.

23

2. inDiCes in GeneRal2.1]

24

2.1 inTRoDUCTion

Changes occur on a daily basis in every person’s life. These changes can be on demographic, economic, as well as social levels. According to Steyn, Smit, du Toit and Strasheim (2007), price changes have an effect on the lives of every family. These price changes, also termed inflation, causes a constant rise in cost of living which, in turn, leads to higher salary demands which, in turn, would increase production costs that can only be recovered by a rise in sales prices. Steyn et al. (2007) state that the calculation of indices, or index numbers, can be used to quantify such changes in a standardised manner. The definition of an index, as given by Steyn et al. (2007) is as follows: “An index is a ratio that measures a relative change”.

Other definitions for index numbers are given by the Steel and Engineering Industries Federation of South Africa (SEIFSA) (undated), which states that “An index is a numerical scale representing a relative level of price at a particular date, compared with the price ruling at some other date”. Flemming and Tysoe (1991) define an index as follows: “Index numbers of costs and prices provide a convenient means of expressing changes over time in the cost or prices of a group of related products in a single summary measure”.

Another way of explaining index numbers, as indicated by Agarwal (2009), is that index numbers are intended to show the average percentage changes in the value of certain products at a spe-cific time, place or situation, when compared to those at any other time, place or situation. Akin-toye (1991) further stated that indices express the current price or quantity, as a percentage of the level at some reference point in the past. This is taken as 100; and consequently, in essence, indices provide a measure of trends.

An example of a typical index that is used world-wide is the consumer price index (CPI), which measures changes in the prices of goods and services consumed by households (Agarwal, 2009).

Marx (2005) is of the opinion that indices only measure relative numbers; and at best they can only give an indication of the measure with which a variable, compared to an earlier period, has changed. Therefore, an index figure cannot reflect any information on the actual level of a vari-able; and it cannot be very accurate. According to Marx (2005), therefore, index figures are in essence arbitrary; and they should be interpreted with circumspection.

Another warning is given by Agarwal (2009), who states that before work is started on the construction of an index, the objective of the index numbers must be clearly defined. If this is not done, the whole labour of constructing the index might be wasted; since no index can be an all-purpose index. According to Agarwal (2009), the primary requirement is to decide what must be measured; and what it is going to be used for.

All the steps that would normally follow in constructing the index, such as the choice of the popula-tion, the base period, the basket of goods, the formula, etc., depend on the objective of the index.

2.2 Types of pRiCe inDiCes

The International Labour Organisation (ILO) (2004b) states that many different kinds of math-ematical formulae have been compiled and proposed over the past 200 years. The ILO (2004b) further mentions that, while there is no single formula that would be preferred in all circumstances,

25

economists and compilers of consumer price indices world-wide seem to be in agreement that the preferred formulae should belong to a small group of indices that are called superlative indices. The ILO (2004b) states that a characteristic feature of such superlative indices is that they treat the prices and quantities that are being compiled in both periods of the indices similarly.

However, before one can look at a group of superlative indices; it should be necessary to go back to the basic compilation of these indices. Steyn et al. (2007) are of the opinion that one must distinguish between simple and composite indices on the one hand, and un-weighted and weighted indices, on the other hand. Steyn et al. (2007) argue that a simple index is used to represent the price change of a single commodity. A composite index, on the other hand, represents the price changes of more than one commodity.

Further, when an un-weighted composite price index is calculated, the price changes of all the commodities are to be regarded as equally important; while in a weighted composite price index, different weights are allocated to the different commodities according to the relative importance of each (Steyn et al., 2007).

As all of the important indices that are being used in the construction industry are weighted com-posite price indices, both simple and un-weighted indices will be ignored in this study. The general formula for a weighted composite price index, where the price of a specific commodity in the base period and in the current period are indicated by po and pn respectively, and with a weighted series (w), is given by:

P = pn

w x 100 10

pow

Much has been written in the past on price indices; and a number of indices have been constructed over time that can be used for tracking the movement of prices of goods over a certain time period. For the purposes of this study, only the most important and the most frequently used of these indices will be discussed. As different authors uses different notations when discussing the formulae, the meaning of the notation used in this study is as follows:

po = price in base year

pn = price in current year

qo = weight in base year

qn = weight in current year

All of these indices are known by the special names of their authors. The first index is called the Laspeyres index; and it was developed in 1871 by Etienne Laspeyres, a French statistician (Diewert, 2001). According to Yu and Ive (2008), the Laspeyres price index is a base weight index where the relative quantities of the base period provide the weighting for the respective prices. The formula for calculating the Laspeyres index (Pl) is:

26

PP = poqo

x 100 11

11

poqo

Pf =

pnqo

x pnqn

x 100 poqo poqo 13

PP = pnqn

x 100 12

11

poqo

Pf = 1 {

pnqo +

pnqn } x 100 2 poqo poqo 12 14

Akintoye (1991) was of the opinion that the Laspeyres index assumes that people are currently buying the same quantity as they bought in the base year, hence the reference to a base-weighted index. In the ideal situation, according to Yu and Ive (2008), the goods (or items) found in the base period are matched with the exact items found in the reference period; therefore the Laspeyres index (sometimes also referred to as a base weighted, match- item index), has a good control of the quality of the items being indexed.

The ILO (2004a) states that the advantage of applying the base weight, or fixed weighted, method is that it is consistent with the method used for other consumption goods and services, as well as the fixed basket-index formula. Therefore, according to the ILO (2004a), the Laspeyres index has until recently been used world-wide as the intellectual basis for the calculation of consumer price indices.

Marx (2005) is of the opinion that the Laspeyres index is the most popular composite index. The main advantage of this index, as stated by Marx (2005), is that a series of Laspeyres indices can be compared; as all they have a common denominator. Another advantage is that a Laspeyres index needs less information, when compared to other indices (Marx, 2005). As already mentioned, in the Laspeyres index, the weights are determined in the base year, and kept the same for the calculation of the index in ensuing years.

Yu and Ive (2008) state this as being a problem; since it does not take into account the quantities in the reference period, where people tend to substitute a cheaper item for a more expensive one, where there is a relative price change. The Laspeyres index is, therefore, criticised for its failure to capture the substitution effects (also called substitution bias). And, it may, consequently, overstate the inflation (Yu &Ive, 2008).

Marx (2005) confirms this problem, when he states that because fixed weights are used, the particular basket of items becomes unsuitable over time. Too much value may be attached to items of which the value has decreased: either because these items have become more expensive; or because they have disappeared from the market. The latter reason can be because the items or products have been replaced by newer, cheaper alternatives. Marx (2005) states an example of such in the building industry is where vinyl floor tiles have replaced wooden block flooring.

The second type of index is called the Paasche price index, named after the German economist, Herman Paasche, who developed this index in 1874 (Diewert, 2001). According to Akintoye (1991), the Paasche index assumes that people are buying the same quantity of items in the current year, as they were buying in the base year; and therefore, this is called a current weighted-price index. The Paasche index (Pp) is represented as follows:

PP = poqo

x 100 11

11

poqo

Pf =

pnqo

x pnqn

x 100 poqo poqo 13

PP = pnqn

x 100 12

11

poqo

Pf = 1 {

pnqo +

pnqn } x 100 2 poqo poqo 12 14

27

The major advantage of current weighted indices over base weight indices, as indicated by Akintoye (1991), is that the items are weighted, in accordance with their current importance. There is, therefore, no danger that the index number could be misleading, due to the use of out-dated weights. Yu and Ive (2006) are of the opinion that the Paasche index is more suitable for deflating output than the Laspeyres index, because of the use of current outputs as the weightings.

Yu and Ive (2006), however, also mention that the Paasche index is criticized as understating inflation, because the choice of goods is not reflected under the base-period prices. Another problem identified with the Paasche index was indicated by van der Walt (1992), who concluded that the compiling of a Paasche type index normally requires more work than the Laspeyres index, because the weights must be determined every time from the start.

Van der Walt (1992) also stated that, when compared with the Laspeyres index, the Paasche index does not offer any additional statistical advantages. The ILO (2004a) mentions that statistical agencies do not produce Paasche type indices, because the lack of information on current period quantities prevents them from doing so, on a timely basis.

The third type of index is called the Irving Fisher price index. It is also called the Ideal index; the idea being that if one index (Laspeyres) overstates inflation, and another (Paasche) understates inflation, the answer would be to take the average of the two indices as a true measure of inflation (Yu &Ive, 2006). This is what Fisher, an American economist, did in 1921, when the index originated. It is the geometric mean of the Laspeyres and the Paasche indices; and it is expressed as follows:

PP = poqo

x 100 11

11

poqo

Pf =

pnqo

x pnqn

x 100 poqo poqo 13

PP = pnqn

x 100 12

11

poqo

Pf = 1 {

pnqo +

pnqn } x 100 2 poqo poqo 12 14

Fisher dubbed this index as the “best form of index” (Yu &Ive, 2008); and in theory, it should be a better measure of inflation. Yu and Ive (2006), however, mention that this advantage comes with a cost, because the calculation of the index requires the information of quantities at both base and reference periods. Akintoye (1991) was also of the opinion that although the Fisher ideal index is theoretically an excellent index, the amount of information required to implement it makes it difficult to use as a general purpose index.

Another index that aims to rectify the difference between the Laspeyres and the Paasche indices is the Drobisch index. The difference between this index and the Irving Fisher index is that the Drobisch index uses the arithmetic mean of the two indices (Steyn et al., 2007). This index is expressed as follows:

PP = poqo

x 100 11

11

poqo

Pf =

pnqo

x pnqn

x 100 poqo poqo 13

PP = pnqn

x 100 12

11

poqo

Pf = 1 {

pnqo +

pnqn } x 100 2 poqo poqo 12 14

28

It is the opinion of Yu and Ive (2008) that all three of the above methods assume that complete data for price and quantity are available. However, as mentioned before, new goods enter the market, and old goods drop out, on a regular basis. Yu and Ive (2008) state that when calculating any of the indices it is less than straight-forward to ascertain how much, for example, a lap top computer should have been priced in 1900, as well as how much a Model T Ford should cost in 2013.

Other indices that are found in some of the literature (for example the ILO, 2004b) are the Young index, the Walsh price index and the Törnqvist price index. Very little evidence could, however, be found in the literature that any of these indices are being used on a regular basis in the construc-tion industry to calculate the movement of prices; and therefore, for the purpose of this study, this index will not be discussed.

2.3 inDiCes in The bUilDinG inDUsTRy

2.3.1 baCkGRoUnD

The following is a discussion on how indices, as described above, can generally be applied in the building industry (the specific need for and the use of indices in the building industry will be discussed later on in the study). According to Seeley (1996) the labour and material content of every building differs; and these cost variations must be taken into account when cost planning for buildings is done. The best way to adjust the available data is through the compilation of an index of building cost. One of the problems encountered in the literature is that different authors use different terminology to describe what essentially are the same concepts.

Flemming and Tysoe (1991) and Davis Langdon Management Consultancy (2008), for example, state that there are three main types of indices used in the construction industry, namely:

• Buildingcostindices

• Tenderpriceindices;and

• Outputindices

In contrast, Eurostat (2008), as well as Statistics Norway (2008), mention that construction price indices can be grouped into three main types:

• Inputindices

• Outputindices;and

• Sellers’indices

To clarify these conflicting terms, Eurostat (2008) state that the terms “cost index” and “price index” should be considered from the point-of-view of a contractor, as depicted in Figure 1.

29

Figure 3: Cost trends incurred on a project (Eurostat, 2008)

Construction cost =

factor prices

Productivity

Profit margins

Other

Energy

Materials

Labour

Plant and equipment

Transport

Output prices = producer

prices

VAT

Site

Architect’s fee

Other costs

Client costs

Profit margins

Selling price = cost to final

client

Contractor (A) Client (B) Final Owner

(C)

In terms of Figure 2, the costs incurred by the contractor in carrying out the construction process (A in the diagram) can be referred to as the construction-input-price index; while the output-price index (B in the diagram) refers to the prices paid by the client. These two indices can be distin-guished from the selling price index (C in the diagram), which measures any changes in the prices paid by the final owner. These terms will now be discussed in more detail.

2.3.2 inpUT-pRiCe inDiCes

Input-price indices, also called building-cost indices (Ashworth, 1991), or construction-cost indices (Wang and Wei, 1998) are, according to McCabe, O’Grady and Walker (2002), as well as Statistics Norway (2007), representative of the construction-process inputs, such as materials, equipment, labour, machinery, transport, energy and other costs. An input index is a weighted index of price indices for a representative selection of basic input elements in the construction process (Mohammadian and Seymour, 1997).

McCabe et al. (2002) state that the primary objective of an input-price index is to reflect the local market prices; and it should not, according to the Statistics Directorate, European Union (1997), be used to provide information on price movement for finished construction works; as these prices do not reflect the complete range of influences that impact on market prices. Prism Economics and Analysis (2001) mention that they do not contain any assumptions or adjustments related to productivity, skills levels, or local code requirements.

Mohammadian and Seymour (1997) also add to this list, the contractor’s overheads and profit margins that are generally excluded.

McCabe et al. (2002) are of the opinion that such indices could be useful for construction compa-nies that move around the country between projects, and only need to know the differences in the

30

cost of materials. According to Marx (2005), such indices can also be used to determine contract price adjustments - after a tender has been awarded. Such adjustments reimburse the contractor in respect of any cost increases of material and labour. SEIFSA (undated) states that a cost index also represents any changes in the cost of a commodity, and sites as an example: their published index of labour.

This index is compiled from a breakdown of the actual labour costs experienced by construction companies; and it is therefore a direct reflection of the cost of labour, and not of necessarily the price per hour that is charged to the end-user.

The charge to the end-user, according to SEIFSA (undated), includes profit; and, in a free market system, it is up to each supplier to determine his or her own profit margin.

2.3.3 TenDeR pRiCe / oUTpUT-pRiCe inDiCes

From the literature, it is evident that tender-price indices and output-price indices are essentially the same concept. Both McCabe et al. (2002) and Prism Economics and Analysis (2001) state that output-price indices attempt to measure the total cost of construction of a completed structure in each location, meaning that these indices reflect the local conditions specific to each project. Statistics Norway (2009) confirms this, and adds that output-price indices also take into account any changes in productivity and contractors’ profit margins, in addition to the input costs, as discussed previously. Marx (2005) who calls these indices contract-price indices, indicates that this monitors market prices, i.e. the price that a developer pays a contractor for erecting a building.

Marx (2005) further states that the factors influencing such indices are the contractor’s profit and overhead costs, as well as his/her competition in the tender market, as this has an influence on the profit margin of tenderers. Seeley (1996), Beeston (1983), Ashworth (1991) and others, all define a tender-price index as being an attempt to represent the level of prices agreed upon between the clients and the contractors.

These prices move, according to any changes in the contractors’ costs, as well as allowances for market conditions and profit. From this definition, it is clear that, according to the literature, output-price indices and tender-price indices are basically the same concept.

Because these indices take into account the tendering market, they are much more useful when updating prices for a design budget (Ashworth, 1991). Where input price indices measure changes in basic building costs, tender-price indices, according to Seeley (1996), indicate what the feeling of the building industry is about the current and future workload. Mohammadian and Seymour (1997), therefore, state that when output-price indices fall, or remain constant during an economic recession, input-price indices have been shown to rise during the same period. The reason for this tendency, according to Mohammadian and Seymour (1997), is that output-price indices reflect market prices, which can be much more volatile than basic input costs, such as labour and material. Beeston (1983) confirms this by saying that contractors slash their profit margins, when they are competing for scarce work, or via an increase, when they find orders easy to obtain, and when there is a glut of work.

Kirkham (2007) as well as Flemming and Tysoe (1991), site several advantages that tender- price indices have from the point-of-view of the client’s quantity surveyor:

31

• Itindicatesthemovementofthecosttotheclient,ofprojectsovertime,ratherthanthe changeincosttothecontractor;

• Itallowscomparisonofthepriceobtainedforaspecifictenderwiththenationalor regionaltenderpricetrend;

• Itallowsforthecostrelationshipbetweendifferentbuildingtypestobeplotted;

• Itgivesanindicationofthetenderingclimateatthedateoftender,becauseittakesinto considerationthevariationsinfactorcosts,aswellastheeffectofcurrenteconomic conditions;

• Asitisnotbasedonotherindices,anyinherentinaccuraciesarenotcompounded;and

• Costplanningcanbeimproved;asthecostofknownschemes,andotherhistoricaldata arebroughttoacommonlevel.

According to the Statistics Directorate, European Community (1997) the formulae that are most commonly used by European Union member countries for compiling output-price indices are the fixed base-weighted Laspeyres and the current weight Paasche indices. In the Laspeyres index, the weights are determined (fixed) in the base year, and kept as such for the calculation in ensuing years (Marx, 2005). Kirkham (2007) states that instead of analysing all the rates in bills of quantities, the amount of work can be reduced by selecting only the few largest items in each work section as weights.

Yu and Ive (2006) are of the opinion that by matching the goods found in the base period with the exact goods found in the reference period, the Laspeyres index has a good control of the quality of the goods being indexed. An example of a Laspeyres index is the BER building-cost index currently being used in South Africa.

For the Paasche index, being a current-weighted index, the priced bills of quantities can also be used (Beeston, 1983). The Building Cost Information Services (BCIS) in the United Kingdom use this method when compiling a tender-price index. According to Akintoye (1991), when compiling this index, a project index is first produced by taking the priced bills of quantities for a project, and then re-pricing any significant items. This is done by selecting the items in each trade that represent 25% of the value of the work in that trade, and re-pricing these items by using a schedule of standard base rates, from what is called a “price book” (Beeston, 1983).

The published tender-price index is an average of several individual project-index figures, calculated for the same period as that described above.

32

2.3.4 selleR’s pRiCe inDeX.

The third type of index, the seller’s price index is not much used in the building industry. The Statistics Directorate of the European Community (1997) indicated that these indices measure changes in the prices of construction output that is paid for by the final owner (or purchaser) of the construction product. According to Statistics Norway (2007), seller’s’ prices include not only all the cost of the completed construction project, such as the cost of labour and the cost of the materials paid by the contractor, but also the cost of land, direct and indirect selling expenses, finance costs, professional fees to architects, engineers, quantity surveyors, etc., value-added tax, as well as the seller’s profit.

2.4 faCToRs inflUenCinG The CoMposiTion of an inDeX

2.4.1 inTRoDUCTion

When constructing an index, decisions have to be made on a number of factors. While Seeley (1996) and Flemming and Tysoe (1991) each name four such factors, Akintoye (1991) is of the opinion that the following six factors all have an influence on the composition of an index:

• Purposeoftheindex;

• Availabilityofthedata;

• Selectionofitemstobeincluded;

• Choiceofthebaseperiod;

• Choiceofweights;

• Methodofconstruction.

Because of the importance that the above six factors have on the composition of an index, they will have to be taken into account when a new tender-price index is constructed. These factors will each be discussed in more detail.

2.4.2 pURpose of The inDeX

According to Akintoye (1991), the purpose of the index, or the use for which the index is intended, should be established upfront, before any attempt is made to construct an index. The statement of the purpose of the index would influence all the factors involved in the construction of the index. Hassanein and Khalil (2006) share this opinion - that the correct interpretation of the index can only be made when its purpose is fully understood. The purpose of an index could, for example, be to measure the movement of tendered prices of commercial, industrial and public buildings over a period of time in South Africa.

33

2.4.3 aVailabiliTy of The DaTa

Akintoye (1991) concluded that it is always important to ensure that there will be enough data in the right format available on a continual basis to construct an index. If not, this might distort the future usefulness and reliability of the index. Flemming and Tysoe (1991) state that instead of using information on the total price of a contract, it is possible to use information of the unit rates that were used for the different categories of work that make up the total contract price. Information on unit rates is readily available from priced bills of quantities; since a contractor who was the successful tenderer on a contract must submit his priced bills of quantities, containing all the unit rates for the project, before the start of the contract.

Yu and Ive (2008) confirm this, by saying that bills of quantities provide a rich source of information on the prices, as well as the quantities of the various trades that are measured for tenders of building projects. Akintoye (1991) also cites a study by Bowley and Corlett (1970), who recognised that a construction-price index that is based on a short list of items, based on priced bills of quantities, reflects the trend in prices shown by a full re-pricing of the bills of quantities.

Flemming and Tysoe (1991) state that this method requires access to a reasonably large number of representative bills of quantities; because the rates used by different contractors can vary considerably. The reason for this is because of the difference in the levels of efficiency, as well as the differences in the labour, materials and plant costs used by contractors’ estimators.

Furthermore, different contractors adopt different practices in arriving at the final tender price (Flemming and Tysoe, 1991). Luckily, according to van der Walt (1992), tender prices for different types and sizes of buildings are obtained on a regular and wide-spread basis in South Africa, by utilising bills of quantities; and it would, therefore, be meaningful to use this method to construct a tender-price index.

2.4.4 seleCTion of iTeMs

Flemming and Tysoe (1991) mention that selecting the items for inclusion in an index can be one of the more difficult problems when formulating an index, especially when there are a lot of possible items that could be included. Akintoye (1991), as well as Steyn et al. (2007), conclude that when constructing a composite index, such as the consumer price index, it is practically impossible and often unnecessary to include all consumer goods. Although selecting a large number of items can be construed as being more representative, Hassanein and Khalil (2006) acknowledge that collecting such large numbers would be very costly.

Steyn et al. (2007), therefore, suggest that a representative sample of items should be selected from the survey population and that only those items be used in the construction of the index. Steyn et al. (2007) further state that by using this method, there will be a time saving without any undue loss of accuracy of information. According to the Statistics Directorate of the European Community (1997), the decision on which actual items to include in the index is largely a matter of judgement; and it depends on the impact that these items may have on the total price of the project.

For the purposes of this study, the objective will, therefore, be to identify the minimum number of indicator items from priced bills of quantities that collectively represent a high proportion of the total value of various construction projects.

34

2.4.5 base peRioD

For the construction of indices, a decision has to be made on a base period (also called a ref-erence period) with which the indices can be compared at a particular time (Agarwal, 2009). Ashworth (1991) stated that normally for general purposes, the cost at the base period is usually given the arbitrary value of 100 in order to allow for both increases, as well as decreases in the value of the data - without having to deal with negative numbers, where values fall below the base index number.

Statistics South Africa (2009) further mentions that the chosen period should preferably cover a seasonal cycle, typically a calendar year. A number of authors (Akintoye, 1991; Flemming and Tysoe, 1991; Agarwal, 2009; Swarup, undated and Steyn et al., 2007) conclude that generally one should choose a base period of reasonable economic stability, as well as a period that is not too distant in the past. Such a “normal” year is explained by Akintoye (1991) as a period of average, steady inflation -without any unusual occurrences. Unusual occurrences are further defined by Steyn et al. (2007), as war years, abnormal climatic conditions, such as droughts or floods, as well as industrial strikes and serious recessions.

The reason why the base period should not be too distant in the past, according to Steyn et al. (2007), is because the prices of commodities might change considerably, the specification of products may be upgraded; and some products might even disappear from the market altogether, if the time between the base period and the reference period is too long.

2.4.6 .ChoiCe of weiGhTs

As discussed before (2.4.4) it is possible to achieve a reliable index by significantly reducing the number of items in the index. Marx (2005) states that this makes the process of gathering market-related items easier and more manageable. Steyn et al. (2007) conclude that when an un-weighted composite price index is calculated, the price changes of all the commodities are regarded as being equally important. However, when items that are being considered for an index are not of equal importance, then the choice of weights for the different items becomes very important.

It is the opinion of Akintoye (1991) that the weights assigned to the various items must reflect their relative importance, and should be carefully chosen, in order to avoid biased and misleading results. Ashworth (1991) mentioned that the majority of the index numbers used in the construc-tion industry include weighted items, according to their importance in the index, hence these indi-ces being called weighted-composite indices (Steyn et al., 2007).

This principle, as stated by Ashworth (1991), is commonly known as “the basket of goods”. When using bills of quantities to determine the selection of items to be used in the index, Seeley (1996) concluded that the major items incorporating the largest price extensions in each trade of the bills of quantities, should be included in the index. Agarwal (2009) calls this system of weighting “value-weighting”. Marx (2005) cites Mitchell (1971), who has shown that by selecting various items, which represent as little as 25% of the sample of the total contract value, an accepted level of reliability can be achieved.

35

By using a fixed-weight series, it can be used for as long as the base period is left unchanged (Steyn et al., 2005). Marx (2005), however, cautions by saying that sooner or later, the items and weights will have to be revised, because of changes in the quality of the materials, improved construction techniques, etc. This would mean that a fundamental shift of the base period would also be necessary. This observation is confirmed in the literature on consumer price indices, where the ILO (2004a) mentions that weights should be updated once every five years, in order to ensure their relevance; while in the literature on construction price indices, Statistics Finland (2001) also quotes a figure of five-year intervals for the revision of their weighting system.

The Statistics Directorate, European Community (1997), on the other hand, is of the opinion that weights used for construction price indices need to be reviewed from time to time; and where necessary, these should be revised at least every five to ten years.

2.4.7 MeThoD of ConsTRUCTion

According to Akintoye (1991), the method of construction relates to the choice between a number of formulae that are available and; furthermore, it can be used to monitor the movement of prices in the building industry. Akintoye (1991) further states that the choice of a particular formula should be based on practical considerations. As discussed before, the most frequently used indices in the building industry are the Laspeyres index, where the base year weightings are being used for calculating the index, and the Paasche index, where the index uses weightings obtained from the current year or period in time that is under consideration.

2.4.8 RaTes

Although not mentioned by Akintoye (1991), as one of the six factors that influence the composition of a tender-price index, the use and selection of unit-rates should also be discussed. The source of prices used in tender-price indices, are priced bills of quantities, where the measured quantities are multiplied by the unit rates inserted by the contractor, in order to obtain quantifiable items. As discussed before, the Laspeyres and Paasche indices are most frequently used for calculating these tender price indices.

Flemming and Tysoe (1991) argued that in practice the Laspeyres index measures the changes in the rates quoted in current tenders compared with the base-period tenders (by using base weights); while in the Paasche index, the items with the biggest value in current tenders are re-priced with rates obtained from a schedule of base-year rates. The United Nations’ economic and social council (2003) mentions that where an index is calculated for a relatively large geographical area, the price collection should be carried out in such a way that these prices (or rates in the case of tender-price indices) must be representative of the complete geographical area; as it is possible that there could be significant differences in price movements between dif-ferent areas.

2.5 Use of inDiCes

From the literature, it is evident that construction-price indices can be used for a variety of pur-poses; and, as indicated by Mohammadian and Seymour (1997), by a number of role- players in the building industry such as producers and purchasers of construction projects, suppliers and manufacturers of construction products, designers, quantity surveyors, cost estimators and budget managers. Some of the particular purposes, for which such indices could be utilised, are as follows:

36

2.5.1 CosT planninG

Flemming and Tysoe (1991) argued that in the process of cost planning, cost information that is available for use in cost planning, and that is based on past projects, would become out of date, if not updated on a regular basis. For this reason, cost indices, according to Kirkham (2007), are fundamental to cost planning, as they provide valuable insight into the changes in cost over time. This can be done for an item, or groups of items from one point in time to another (this is also called a “time series”). Ashworth (1991) agreed by saying that the process of cost planning requires the use of large amounts of historical cost data; and in order for them to be used effectively, the data would need to be updated by the use of indices.

2.5.2 foReCasTinG

Flemming and Tysoe (1991) stated that indices can play an important role in the forecasting of cost trends, which could be beneficial to both clients and building contractors. A lot of research that is done on construction price indices, as stated by Yu and Ive (2008), is about forecasting - by using time-series techniques, in which the past values of price indices are used to forecast their future values, and thereby to determine price inflation. The opinion of Ashworth (1991) was that, although the pattern of existing indices can be extended to a date in the future, the extrapolation of existing indices should be done with caution, because subjective allowances must be made to allow for the differences in the varying conditions between the past and the future.

Under stable conditions, the projection of indices is a simple matter; but the erratic behaviour of inflation, as experienced in recent years, has made accurate forecasting very difficult.

2.5.3 UpDaTinG CosT esTiMaTes

The updating of elemental cost-analyses numbers is, according to Ferry et al. (2003), perhaps the most common use of index numbers by quantity surveyors. In this way, the information on past projects can be brought up to current costs. Brook (1974) also stated that from a cost-estimating point of view, a common use of price indices is to update estimates to current costs; but this method can also be used to project estimates into the future. Both Brook (1974) and Ferry et al. (2003) caution that care should be taken, when updating information beyond a period of two years, because of the fluctuating climate of the construction industry.

2.5.4 UpDaTinG of TenDeRs

As with estimates, tenders can be updated by contractors through the use of indices (Van der Walt, 1992). Seeley (1996) concluded that building-cost indices are used by contractors to assess the differences in levels of tenders at varying dates. In this way, a tender figure for a previous simi-lar project can be brought up-to-date with current prices for comparison purposes. By calculating the percentage change between the building cost-index figure for a specific quarter of a specific year, and the same quarter of a previous year, the increases in building costs can be determined as a percentage per annum (Segalla, 1991).

37

2.5.5 MoniToRinG pRiCe MoVeMenTs

Because of its weighting system, the building-cost index provides useful tools for monitoring and analysing the movement of - not only building costs in general, - but also the prices of separate components (Statistics Finland, 2001). An index, as indicated by Marx (2005), measures the changes in the cost of an item, or groups of items, over time and therefore, according to Ferry et al. (2003), through cost indices it should be possible to see the changes in the relationship between different cost components over time, for example structural steel versus reinforced con-crete. This relationship can then be used as a possible solution for design problems, when one option appears to be a better proposition from a cost point- of-view.

2.5.6 ReplaCeMenT CosT of bUilDinGs

A number of authors (Segalla, 1991; Kilian, 1980 and Akintoye et al., 1998) are of the opinion that construction-price indices are used by the insurance brokering fraternity to calculate the replace-ment value of buildings. This is done by using an index value to calculate the percentage change in building costs, since the last valuation - up to the present time. This percentage change is then applied to the original replacement value of the building, in order to obtain the current replace-ment value.

2.5.7 MoniToRinG The naTional eConoMy

Some countries use index numbers to monitor aspects of the national economy. According to Statistics Finland (2001), building cost indices are used as an indicator in the national economic policy, as well as in economic policy-making research. Likewise, Statistics Norway (2007) indi-cates that construction-cost indices can be used to estimate the output prices to deflate national accounts, and to estimate the national output of construction activities.

2.5.8 neGoTiaTion of ConTRaCTs

Brook (1974), as well as Statistics Finland (2001), concludes that an important function of con-struction-cost indices is in tying long-term building contracts, where it can be difficult to control escalation costs to the satisfaction of all parties, to the index. This can be done by linking the ne-gotiation to a specific starting date, and then calculating the increase in cost, at regular intervals, by using the published index numbers.

2.6 ChallenGes assoCiaTeD wiTh inDiCes

There are some inherent challenges with the use of tender price indices. The following are some of the major challenges, as sourced in the literature:

2.6.1 aCCURaCy of The inDeX

As mentioned by Ashworth (1991), index numbers can at best only provide a general indication of the changes in the value of building projects; and therefore, they cannot be considered to be very precise. Van der Walt (1992) agreed by saying that an index is relative; in other words it is not its absolute value that matters, but its tendency over time. Hassanein and Khalil (2006) cite Chase

38

and Nichols (1992), who mentioned that a perfect forecast is typically impossible; because there are too many factors in the construction environment that cannot be predicted with any degree of certainty.

One of these factors is that an index for a typical or model building is more of an economic model, indicating a general trend; and it may not be measuring the change over time for a particular pro-ject that is being developed (Kirkham, 2007).

2.6.2 saMple size

The Statistics Directorate of the European Community (1997) stated that, to determine an ad-equate number of respondents, is largely a sampling question. The larger the number of respond-ents, the more detailed would the indices need to be that are produced. In practice, however, according to the Statistics Directorate of the European Community (1997), the size of the sample is largely a trade-off against cost and the quality of the data. Ferry et al. (2003) mention that in tender-based indices, a good sample of priced bills of quantities is needed to avoid bias caused by regional variations, and building functions that might distort the results.

Van der Walt (1992) agreed with this opinion, stating that tenders of only a small number of tenders would probably result in unstable indices.

The literature either differs, or is vague, about the actual number of priced bills of quantities that are required to construct a reliable tender-price index. Most authors that comment on the BCIS’s tender-price index (such as Yu and Ive, 2006; Kirkham, 2007 and Ferry et al., 2003) conclude that the BCIS aims at sampling 80 projects in each quarter; because it believes that if 80 projects are sampled, approximately 90% of the indices of individual projects would fall within about 2,8% of the average.

Yu and Ive (2006), however, indicate that this requirement is seldom met; and between 1990 and 2004, the BCIS’s average quarterly sample size was 67. The reason for this drop in the number of priced bills of quantities can be attributed to a clear shift of British procurement methods between 1985 and 2004 - from traditional procurement with bills of quantities to lump-sum design and build (Yu &Ive, 2008). Akintoye (1991) was also of the opinion that a drop in construction activities would inevitably result in difficulty in meeting the requirement of 80 projects. In South Africa, Kilian (1980) argued that in the case of the BER index, it was found that about 40 projects are sufficient to have a stable index; while, according to Brook (1985) it was found that when the index value is based on less than seven projects, it cannot be regarded as being statistically stable.

2.6.3 ChanGes in qUaliTy

Another problem, as indicated by Kirkham et al. (2007), is that changes in technology may make the mix of base weights atypical. Simonton (2004) states that building quality and specifications have improved steadily over time, because of advances in building technology, design and in pro-cesses. As a typical index would be measuring the trend of building costs for a typical or model building over time, these changes in quality would not necessarily be taken into account. Marx (2005) mentions an example where, if a standard interior door that is used today is of a better quality than the typical door that is used in the base period, the increase in the rate for the door is made up not only of the inflation in the building industry, but also of the improved quality of the product.

39

Kirkham et al. (2007) add to this by saying that as new techniques and materials come into greater use, their prices tend to decrease proportionally; while the cost of obsolescent technology tends to rise faster than the general rate of increase. One way of overcoming this problem, according to the Statistics Directorate of the European Community (1997), is to review the basket of items on a regular basis, so that the components and their weights reflect any of the changes in technical standards, new construction methods, new technology, new building materials, etc.

2.6.4 UniT RaTes

Kirkham et al. (2007) state that one of the limitations of the tender-based index is the question-able validity of the rates of bills of quantities. Marx (2005) agrees with this, by saying that the unit rates in bills of quantities can differ markedly, because of the different approaches by tenderers to determine such rates. Marx (2005) cites Seeley (1996), who found that the standard deviation of individual tariffs of the average tariff of an item can be as much as 15%. Van der Walt (1992) further mentioned that, apart from different techniques that contractors use in building up their rates, such rates can sometimes be estimates of expected prices in the case of a fixed price contract. The reason for this is because the contractor would then have to allow for expected inflation for the duration of the contract in his rates; e.g. for example: what could he expect the price to be at the time when the work is finally executed?

2.7 sUMMaRy anD ConClUsion

The use of index numbers is a well-established practice in different spheres of life, including the building industry. Although any index must be constructed with care, taking into account a number of considerations, it should be possible to have a tender-price index that serves its purpose, - i.e. to measure the general movement of building costs over time.

Given the discussion on the different construction price indices, it is clear that a tender-price index (or output-price index) would be the best option, because it measures the movement of changes in contractors’ prices, taking into account all factors, such as the contractors’ input costs, as well as allowances for market conditions and profit.

Another issue that influences this choice is the availability of a sufficient number of priced bills of quantities in the South African building industry on a regular basis for the foreseeable future, to use as a basis for compiling such an index.

Another decision that has to be made is what formula to use. As stated before, the most well-known and often-used formulae are the Laspeyres and Paasche indices, with the Irving- Fisher index also being option. The biggest factor that counts against the use of the Paasche formula is the fact that, in order to use such an index, a schedule of standard base rates (or a price book) would be needed. As there is currently nothing of such a nature available in South Africa, a new schedule of base rates would have to be developed. Considering the amount of work involved, not only in developing such a schedule, but also in the execution of the index, where a project index must be compiled for every project. Therefore, a more logical choice would be to use the Laspeyres formula for composing a new tender price index for use in South Africa.

This would involve the analysis of bills of quantities in order to arrive at a set of fixed base weights, or a basket of items. Another consideration for using the Laspeyres formula is that the BER index that is currently in use in South Africa is also a Laspeyres-type index; and therefore, any compari-son would be easier to make.

40

41

3. CRane inDeX

42

3.1 .inTRoDUCTion

There is not a great deal of literature available on crane index methodology. Currently two global companies involved with the construction and property industries were found to conduct and publish information on crane indices, viz. Rider, Levett, Bucknall (RLB) and Deloitte (Real Estate). RLB is an independent, global property and construction practice with a world-wide network of offices on all continents. It publishes a crane index on a regional basis such as the Middle East, Australia/New Zealand and North America. Each region is subsequently broken down into the major cities of that region, e.g. the cities in Australia that are used are Adelaide, Brisbane, Canberra, Darwin, Melbourne, Perth and Sydney, while North America consist of Boston, Calgary, Chicago, Denver, Honolulu, Los Angeles, New York, Phoenix, Portland, San Francisco, Seattle, Toronto and Washington DC.

Deloitte refers to Deloitte Touche Tohmatsu Limited, A United Kingdom (UK) private company with a network of firms in various countries. Deloitte’s crane survey concentrate on the UK and separate reports are published for London, Leeds, Birmingham, Manchester and others. Deloitte’s crane index for London is further sub-divided into a general and office crane survey.

3.2 MeThoDoloGy

The basic methodology of the RLB crane indices is that the number of cranes is tracked on a ````````bi-annual basis in a city. Firstly the number of cranes in the city’s CBD is calculated, and then the number of cranes that are visible within a 5km radius from the city centre. A brief report is also given with each publication on the location of the major sites in each city where these cranes are situated. The movement of the numbers are also reported i.e. whether there was either an increase or a decrease in the number of cranes. A summary of all the cities are also given in each bi-annual publication with a legend indicating either a significant increase, a significant decrease or a steady number of cranes. With this summary a comparison can also be made of the general movement of building activity between various cities in the country.

The Deloitte survey, which is published on a quarterly basis, is more comprehensive, not only indicating an increase/decrease in the number of cranes in a city, but reports also contain an indication of the area of new buildings that have started to be constructed since the previous survey, subdivided into residential, office, hotel and educational space. This information is ````````summarised in a table indicating developments by name, developer, total size (in floor area), completion date and comments.

If a crane index is therefore published from information based on the movement of a number of cranes only, it will be classified as a simple index because, as indicated previously under general index theory, it represents the changes in a single commodity.

In order to develop a crane index, the first step will be to develop a construction cost index, based on data that are supplied by metropolitan areas in South Africa. The data are provided in value terms, as well as size of planned construction. Figure 4 provides the total value of building plans passed as well as buildings completed as reported by larger municipalities at national level.

43

Figure 4: Source: Statistics South Africa

0

20,000,000

40,000,000

60,000,000

80,000,000

100,000,000

120,000,000

2000

/01/

01

2000

/12/

31

2001

/12/

31

2002

/12/

31

2004

/01/

01

2004

/12/

31

2005

/12/

31

2006

/12/

31

2008

/01/

01

2008

/12/

31

2009

/12/

31

2010

/12/

31

2012

/01/

01

2012

/12/

31

2013

/12/

31

2014

/12/

31

2016

/01/

01

R'00

0

Year

Value of Buildings at Current Prices

Buildings completed Building plans passed

The total size of construction in square meter of space developed or planned to be developed is provided in Figure 5.

When the value of building plans approved is divided by the size of planned development, a construction cost index of such planned construction is obtained. Similarly the value of buildings completed divided by the size of buildings completed would provide the cost index for such completed buildings. These two indices are provided in Figure 6.

Figure 5: Source: Statistics South Africa

0

500,000

1,000,000

1,500,000

2,000,000

2,500,000

2000

/01/

01

2000

/12/

31

2001

/12/

31

2002

/12/

31

2004

/01/

01

2004

/12/

31

2005

/12/

31

2006

/12/

31

2008

/01/

01

2008

/12/

31

2009

/12/

31

2010

/12/

31

2012

/01/

01

2012

/12/

31

2013

/12/

31

2014

/12/

31

2016

/01/

01

Squa

re M

eter

Year

Square Meter of Buildings Planned or Developed

Buildings completed Building plans passed

0

1,000

2,000

3,000

4,000

5,000

6,000

7,000

2000

/01/

01

2000

/12/

31

2001

/12/

31

2002

/12/

31

2004

/01/

01

2004

/12/

31

2005

/12/

31

2006

/12/

31

2008

/01/

01

2008

/12/

31

2009

/12/

31

2010

/12/

31

2012

/01/

01

2012

/12/

31

2013

/12/

31

2014

/12/

31

2016

/01/

01

Rand

Year

Construction Cost at Current Prices

Buildings completed Building plans passed

44

Figure 6: Source: Authors’ calculations

0

500,000

1,000,000

1,500,000

2,000,000

2,500,000

2000

/01/

01

2000

/12/

31

2001

/12/

31

2002

/12/

31

2004

/01/

01

2004

/12/

31

2005

/12/

31

2006

/12/

31

2008

/01/

01

2008

/12/

31

2009

/12/

31

2010

/12/

31

2012

/01/

01

2012

/12/

31

2013

/12/

31

2014

/12/

31

2016

/01/

01

Squa

re M

eter

Year

Square Meter of Buildings Planned or Developed

Buildings completed Building plans passed

0

1,000

2,000

3,000

4,000

5,000

6,000

7,000

2000

/01/

01

2000

/12/

31

2001

/12/

31

2002

/12/

31

2004

/01/

01

2004

/12/

31

2005

/12/

31

2006

/12/

31

2008

/01/

01

2008

/12/

31

2009

/12/

31

2010

/12/

31

2012

/01/

01

2012

/12/

31

2013

/12/

31

2014

/12/

31

2016

/01/

01

Rand

Year

Construction Cost at Current Prices

Buildings completed Building plans passed

Although the indices provide some information of cost over time and it would be possible to con-sider the overall increase in construction cost from year to year, the short term variability in the indices should be further investigated. In order to do this, the same information as in Figure 4 is considered, but at constant prices. This provided in Figure 7. A construction cost index could then be calculated that indicates construction over time at constant prices. This provides an indication of how construction prices vary from year to year, based on factors such as type of construction, quality of buildings, level of finishes, changes in contractors’ profit margins, etc. These indices for planned development and buildings completed are displayed in Figure 8.

In addition to the construction cost indices provided in Figure 8, the data also contains information with regards to the differences between planned construction and actual construction that took place. It is evident in all graphs that there is a similarity in the movement of development planned and buildings completed, but there appears to be a lag. Figure 9 provides the information where this lag is tested, with a zero lag, as well as where buildings completed lags behind planned de-velopment by 6 months, 12 months, 18 months and 24 months. The results of this are as follows:

• Zerolag - R2of0.5887

• 6Monthlag - R2of0.7981

• 12Monthlag- R2of0.9205

• 18Monthlag- R2of0.9168

• 24Monthlag- R2of0.7727

From this, the closest correlation is at 12 months, followed shortly by an 18 month lag. It is thus evident that there is a lag of approximately 12 to 18 months from planned construction up to com-pletion of such construction. It should, however, be emphasised that this would vary for different types of construction, quality of building and possibly during different market cycles.

45

In addition to the calculated lag, it is evident from Figure 9 that the minimum and maximum val-ues for developments planned and actual construction that took place, differ. The minimum value for planned development is 1,122,164m2, while the maximum value is 2,186,590m2. The cor-responding values for actual construction that took place have respective values of 691,607m2 and 1,345,668m2. Figure 10 investigates this relationship by comparing the 12 month lagged completed construction to the building plans approved. This provides an indication of how much throughput is obtained from planned construction to actual buildings completed, which is indi-cated to vary between 54% and 73%.

The information provided in these graphs is, as indicated earlier, at national level. To further in-vestigate the development of a crane index, specifically for the Tshwane Metropolitan Municipal-ity, similar information is obtained at local level in order to attempt to explain the local economic variables better.

Figure 7: Source: Statistics South Africa

0

20,000,000

40,000,000

60,000,000

80,000,000

100,000,000

120,000,000

2000

/01/

01

2000

/12/

31

2001

/12/

31

2002

/12/

31

2004

/01/

01

2004

/12/

31

2005

/12/

31

2006

/12/

31

2008

/01/

01

2008

/12/

31

2009

/12/

31

2010

/12/

31

2012

/01/

01

2012

/12/

31

2013

/12/

31

2014

/12/

31

2016

/01/

01

)secirp 0102 tnatsnoC( 000'R

Year

Value of Buildings at Constant Prices

Buildings completed Building plans passed

0

1,000

2,000

3,000

4,000

5,000

6,000

2000

/01/

01

2000

/12/

31

2001

/12/

31

2002

/12/

31

2004

/01/

01

2004

/12/

31

2005

/12/

31

2006

/12/

31

2008

/01/

01

2008

/12/

31

2009

/12/

31

2010

/12/

31

2012

/01/

01

2012

/12/

31

2013

/12/

31

2014

/12/

31

2016

/01/

01

)secirp 0102 tnatsnoC( dnaR

Year

Construction Cost at Constant Prices

Buildings completed Building plans passed

Figure 8: Source: Authors’ calculations

0

20,000,000

40,000,000

60,000,000

80,000,000

100,000,000

120,000,000

2000

/01/

01

2000

/12/

31

2001

/12/

31

2002

/12/

31

2004

/01/

01

2004

/12/

31

2005

/12/

31

2006

/12/

31

2008

/01/

01

2008

/12/

31

2009

/12/

31

2010

/12/

31

2012

/01/

01

2012

/12/

31

2013

/12/

31

2014

/12/

31

2016

/01/

01

)secirp 0102 tnatsnoC( 000'R

Year

Value of Buildings at Constant Prices

Buildings completed Building plans passed

0

1,000

2,000

3,000

4,000

5,000

6,000

2000

/01/

01

2000

/12/

31

2001

/12/

31

2002

/12/

31

2004

/01/

01

2004

/12/

31

2005

/12/

31

2006

/12/

31

2008

/01/

01

2008

/12/

31

2009

/12/

31

2010

/12/

31

2012

/01/

01

2012

/12/

31

2013

/12/

31

2014

/12/

31

2016

/01/

01

)secirp 0102 tnatsnoC( dnaR

Year

Construction Cost at Constant Prices

Buildings completed Building plans passed

46

Figure 9: Source: Authors’ calculations

1,000,000

1,200,000

1,400,000

1,600,000

1,800,000

2,000,000

2,200,000

600,000 800,000 1,000,000 1,200,000 1,400,000

devorppa snalp gnidliuB

Buildings completed

Lag -Building Plans Approved vs. Buildings Completed

Zero Lag 6 Month Lag 12 Month Lag 18 Month Lag 24 Month Lag

0%

10%

20%

30%

40%

50%

60%

70%

80%

2001

/12/

01

2002

/07/

01

2003

/02/

01

2003

/09/

01

2004

/04/

01

2004

/11/

01

2005

/06/

01

2006

/01/

01

2006

/08/

01

2007

/03/

01

2007

/10/

01

2008

/05/

01

2008

/12/

01

2009

/07/

01

2010

/02/

01

2010

/09/

01

2011

/04/

01

2011

/11/

01

2012

/06/

01

2013

/01/

01

2013

/08/

01

2014

/03/

01

2014

/10/

01

2015

/05/

01

2015

/12/

01

% Throughput of Building Plans Approved

Buildings completed

Figure 10: Source: Authors’ calculations

1,000,000

1,200,000

1,400,000

1,600,000

1,800,000

2,000,000

2,200,000

600,000 800,000 1,000,000 1,200,000 1,400,000

devorppa snalp gnidliuB

Buildings completed

Lag -Building Plans Approved vs. Buildings Completed

Zero Lag 6 Month Lag 12 Month Lag 18 Month Lag 24 Month Lag

0%

10%

20%

30%

40%

50%

60%

70%

80%

2001

/12/

01

2002

/07/

01

2003

/02/

01

2003

/09/

01

2004

/04/

01

2004

/11/

01

2005

/06/

01

2006

/01/

01

2006

/08/

01

2007

/03/

01

2007

/10/

01

2008

/05/

01

2008

/12/

01

2009

/07/

01

2010

/02/

01

2010

/09/

01

2011

/04/

01

2011

/11/

01

2012

/06/

01

2013

/01/

01

2013

/08/

01

2014

/03/

01

2014

/10/

01

2015

/05/

01

2015

/12/

01

% Throughput of Building Plans Approved

Buildings completed

3.3 aDVanTaGes anD DisaDVanTaGes

Advantages

The obvious advantage of a crane index is that it is a fairly simple way to gauge the movement of building activity in a particular city over time, e.g. a significant increase in the number of cranes from one period to the next can be an indication that there was also an increase in the building activity in the city over that period. It therefore provides a quick and easy way to provide interested parties with a broad overview of the movement in building activity in a city over a certain period of time.

47

Disadvantages

The most obvious disadvantage is that such an index, because of its simplicity, cannot provide accurate figures and is merely a rough indication that there was either an increase or decrease in building activity from one period to the next. This index can also be distorted by a concentration of cranes on one or two developments. This is currently the situation in Tshwane where there are a large number of cranes situated in three building sites in the east of the city (Menlyn Main pre-cinct, Menlyn shopping centre and the office development across Menlyn shopping centre). This can give an indication that there was a large jump in building activity across Tshwane compared with the previous 6 months, while it is actually true of only a small concentrated area.

Another disadvantage is that cranes are mostly used on high rise developments. The counting of cranes on these developments will therefore excludes building activity on low rise developments such as residential estates, single storey strip malls, single or double storey office parks, etc.

3.4 ConClUsion

A crane index can be a useful tool for gauging movement of building activity over time. As described above, such an index, based on the number of cranes alone, will only provide a very broad indication of such activity and, therefore, if the aim is a more detailed report of the actual building activity, additional information such as the value of approved plans, etc. during the same period, will have to be used to support the crane index calculations.

48

49

4. analysis of Tshwane speCifiC DaTa

50

In order to understand the movement of building plans passed and ultimately buildings completed as discussed earlier in this document, it is necessary to investigate the drivers of construction activity.

Boshoff (2010) found that there is a direct link between disposable income of households, which are capitalised with interest rates in order to calculate an affordability index, and house prices. Figure 11 provides a correlation between such an affordability index for Gauteng vs. building plans passed. From Figure 11 it is evident that a number of data points appear to provide a close correlation, but others are moving outwards and distorting such a possible correlation. Upon closer investigation, all these outliers are after 2009 and appear to have caused a structural change in the relationship. It is also so that the further away from the pre 2009 relationship, the longer the period after 2009 that the data point is observed. The cause of this is unsure, but might be related to the National Credit Act, the global financial crises at the time or subsequent political and other economic influences in South Africa. This would, however, need further analysis. The important implication is that affordability, driven by disposable income and interest rates, influence construction activity in the residential market through the equilibrium as discussed in space and capital market theory early in this document. It is, however, evident that a lag exist which causes uncertainties in future analysis, i.e. it takes longer to see evidence of the extent to which households plan to build new houses with a further lag in the evidence of how many of this actually materialise as new construction. These lags make it difficult to perform accurate forecasts of economic activity specifically in the construction sector.

Figure 11: Affordability vs Building Plan Lags

0

1,000,000

2,000,000

3,000,000

4,000,000

5,000,000

6,000,000

0 2,000,000 4,000,000 6,000,000 8,000,000 10,000,000 12,000,000

Affor

dabi

lity

Building Plans Passed

Affordability vs 6 Month Lag Building Plans Passed

On the commercial property side, the importance of income drivers and existing property stock in the market should be considered as discussed in the literature on space and capital markets earlier. Figure 12 also shows some evidence of corporate savings that might influence construction activity. This might be due to surpluses that are available by corporate institutions that are used to expand business infrastructure in order to improve future prospects, which are done through investment in property and subsequent new construction taking place.

51

Figure 12: Corporate Savings vs Gross Fixed Capital Formation

R² = 0.4808

0

50,000

100,000

150,000

200,000

250,000

0 2,000 4,000 6,000 8,000 10,000 12,000

(R'0

00 0

00 -

cons

tant

201

0 pr

ices

)

(R'000 000 - constant 2010 prices)

Corporate Savings vs. Gross Fixed Capital Formation: Construction

Although the data in Figure 12 is for South Africa as a whole, Gross Capital Formation in South Af-rica is also investigated for the extent that it influenced construction activity in Gauteng specifically. The relationship of this is provided in Figure 13. It is again evident that there are some close correlating data points and a number of outliers. It is again found that the outliers are for data points after 2009.

Figure 13: Source: Authors’ calculations

R² = 0.3315

0

2,000,000,000

4,000,000,000

6,000,000,000

8,000,000,000

10,000,000,000

12,000,000,000

0 5,000,000 10,000,000 15,000,000 20,000,000 25,000,000 30,000,000 35,000,000

Gros

s Cap

ital F

orm

ation

Buildings Completed

Gross Capital Formation vs Buildings Completed

From the above, it is evident that certain economic variables provides a view of construction activity, but due to the delay in information becoming available and the lags between the processes , it is difficult to interpret this for an understanding of immediate circumstances or even future forecasting. It is specifically for this reason that a crane index is considered to provide a quick view of construction activity that have realised and also to identify specific high growth nodes through the concentration of cranes.

It was noted earlier that the crane index would be analysed to determine if such an index can provide a more timely result of planned construction that actually materialised and also to

52

compare the crane index to a construction cost index to determine the relationship between construction cost and type of construction that takes place. In order to perform such comparisons, it is important to first determine the specific details of the crane index as it pertains to Tshwane construction activity.

4.1 Tshwane CRane inDeX

The accepted methodology normally followed in executing a crane index, as previously indicated in the literature review, is to physically count all cranes over a given period (say twice a year) in order to establish the number of cranes erected at different building sites in a city. This exercise is then repeated over time to establish an index. With this study it was not possible to follow the above methodology as the research into the establishing of a crane index for the Tshwane Metro was initiated in 2016.

This challenge was addressed by obtaining historical information from contractors in the Tshwane region, in order to establish an estimate of the numbers of cranes that were erected in the past five years.

The population for this study was all the contractors in Tshwane, but the majority of contractors who are active in the city are medium to small contractors who seldom or never make use of tower cranes. A purposive sample was therefore used to select contractors who are known to be involved with bigger projects where tower cranes are used, currently or in the past .

4.1.1 ResUlTs

Of the 10 contractors that were contacted, 7 responded with information and 1 contractor indicat-ed that they did not have any cranes on sites during this period. No information was received from the other two contractors. The result of the 7 contractors that provided information is indicated in Table 1, which are then provided as an index in Table 2 and graphically illustrated in Figure 14.

Table 1: Number of cranes 2011 to May 2016(Authors’ primary data collection)

2011 2012 2013 2014 2015 2016 (May)Contract 1 4 4 5 4 8 5Contract 2 1 1 2 2 2 1Contract 3 2 3 1 2 3Contract 4 1 2 3Contract 5 2 1Contract 6 1 1 1Contract 7 1 1Total 5 7 11 10 19 14

Table 2: : Index: 2011 = 100(Authors’ calculations)

Tshwane 100 140 220 200 380 280

53

Figure 14: Source: Authors’ calculations

5

7

1110

19

14

0

2

4

6

8

10

12

14

16

18

20

2010 2011 2012 2013 2014 2015 2016

Num

ber o

f cra

nes

Year

Crane movement as per author count

4.1.2 DisCUssion

The information that was received from 7 contractors for the period in question is indicated in Table 2. From this it can be observed that there seem to be a steady increase in the number of cranes from 2011, but this trend can also be attributed to incomplete information supplied by contractors (two contractors indicated that they did not have sufficient time to look into their archives to go back the full five years).

Because of the relatively low numbers that are involved in the survey, it can be seen that a small change in the number of cranes for one period to the other can have a large influence on the movement of the index, e.g. a 90% increase between 2014 and 2015 ( Table 3).

Table 3: Crane Index Year-on-YearPercentage Change (Authors’ calculations)

2011 2012 2013 2014 2015 2016 (May)*40.00% 57.14% -9.09% 90.00% -26.32%

4.1.3 Rlb CRane inDeX foR soUTheRn afRiCa

It was subsequently found that RLB has started to publish a crane index for Southern Africa, simi-lar to those in other countries as discussed in the literature survey. In this publication, which was started in the 4th quarter of 2014, the number of cranes in Cape Town, Johannesburg, Pretoria, Mozambique and Mauritius is currently counted. The results for Tshwane (indicated as “Pretoria” in the publication) are provided in Table 4 and Figure 15.

54

Table 4: RLB Crane Index(RLB)

Q4: 2014 Q2: 2015 Q4: 2015 Q2: 2016No. of cranes 26 21 38 33% change from previous

0 -19.23 80.95 -13.16

Index 100 80.77 146.15 126.92

Figure 15: Source: RLB

26

21

38

33

0

5

10

15

20

25

30

35

40

2014: Q4 2015: Q2 2015: Q4 2016: Q2

Num

ber o

f cra

nes

Year

RLB Crane movement

From this information, it can be observed that in the first quarter of 2016, there were 33 cranes in the Tshwane area. This information was verified by a personal count for this period. The big discrepancy between this figure and that of the historical information, can be attributed to the fol-lowing:

• Thehistoricalinformationwasbasedonasampleofcontractorsandthereforeallcranes inTshwanearenotreflected.Itcanbenotedinthisregardthatinformationwasnot forthcomingfromthecontractorontheMenlynMainprecinct,wheretherearecurently approximately10cranesinvolved.

If the information by RLB is compared in more detail to the data obtained by the authors, it should be considered that the RLB index is bi-annually, while the authors only obtained annual data. With the time taken to complete a construction project, it could also be assumed that to convert the RLB info to an annual index for comparison, the two quarter’s information of 2015 cannot be added together, as that would most probably result in a double counting of some of the data points. For this reason, the Q2:2015 is just left out and the other three remaining points are taken as the results for the three respective years. When then comparing it to the author’s data, it is established that the author’s crane count is 38%, 50% and 42% of the RLB count for the respec-tive periods. Although this is a very limited comparison, the general co-movement of the data is considered to confirm that the data obtained by the authors is a representative sample and due to the longer period of data obtained, offers a better means of comparison to other activity than the RLB data.

55

4.2 Tshwane CRane inDeX CoMpaReD To ConsTRUCTion aC-TiViTy

In order to determine if the crane index is a useful tool for indicating construction activity, it should be investigated if it provides more timely information than other information available. For this purpose, the accuracy of the building plans passed and the crane index is measured and it is also measured to what extent a lag is visible between building plans and crane movement. The most accurate measure for actual construction activity that resulted from building plans passed is the buildings completed statistics as provided by municipalities, similar to the building plans passed statistic. This lag between the buildings completed and building plans passed is expected to be longer than the lag of the crane index, as the crane index can pick up construction activity while it is still under way.

Figure 16 indicates the relationship between building plans passed in Gauteng against the actual buildings completed 12 months later. It was found that this time lag provides the closest correla-tion between the 2 data sets and is therefore considered indicative of the average time taken from approval of building plans to completion of construction.

Figure 16: Source:(Authors’ calculations)

R² = 0.8401

2,000,000

2,500,000

3,000,000

3,500,000

4,000,000

4,500,000

5,000,000

5,500,000

6,000,000

6,500,000

7,000,000

4,000,000 5,000,000 6,000,000 7,000,000 8,000,000 9,000,000 10,000,000 11,000,000

)retem erauqS( detelp

moc sgnidliuB

Building plans passed (Square meter)

Gauteng Building Plans Passed (sq.m) vs. 12 Month Lag Buildings Completed (sq.m)

R² = 0.5989

0

50

100

150

200

250

300

350

400

4,500,000 4,750,000 5,000,000 5,250,000 5,500,000 5,750,000 6,000,000 6,250,000

Cran

e In

dex

Building plans passed (Square meter)

Crane Index vs. 6 Month Lag Building Plans Passed

Although the crane index has a much more limited time frame than the data provided in Figure 16, it is still considered useful at least in motivating the rationale for such an index. Figure 17 provides the Tshwane crane index as correlated to building plans passed, 6 months prior to the cranes that were counted. The result shows an R2 of 0.5989, which is considered a fair indication of construction activity, but with the lag being reduced from 12 months to 6 months, it is evidently a more timely measure of construction activity.

56

Figure 17: Source:(Authors’ calculations)

R² = 0.8401

2,000,000

2,500,000

3,000,000

3,500,000

4,000,000

4,500,000

5,000,000

5,500,000

6,000,000

6,500,000

7,000,000

4,000,000 5,000,000 6,000,000 7,000,000 8,000,000 9,000,000 10,000,000 11,000,000

)retem erauqS( detelp

moc sgnidliuB

Building plans passed (Square meter)

Gauteng Building Plans Passed (sq.m) vs. 12 Month Lag Buildings Completed (sq.m)

R² = 0.5989

0

50

100

150

200

250

300

350

400

4,500,000 4,750,000 5,000,000 5,250,000 5,500,000 5,750,000 6,000,000 6,250,000

Cran

e In

dex

Building plans passed (Square meter)

Crane Index vs. 6 Month Lag Building Plans Passed

4.3 Tshwane CRane inDeX CoMpaReD To ConsTRUCTion CosT

The construction cost of planned construction and of buildings completed is determined by dividing the value of buildings planned on the one hand and of completed construction on the other, by the respective square meters of planned construction and completed construction. This provides an indication of the cost of the buildings at the time, which might be influenced by the type of construction.

When considering the comparison of the crane index to a construction cost index, it is also important to consider the fact that the crane index should be compared to the cost of planned construction in order to be meaningful. It is, however, the actual construction activity and the cost thereof that affects economic activity. It is thus firstly important to ensure that the construction cost index of planned construction is similar to the construction cost index of completed buildings. Figure 17 provides the two respective construction cost indices, both at constant 2010 prices and this provides an indication of changes in the cost of construction due to the quality of buildings or other influences than general price increases. Figure 18 then provides a correlation of these indices, where it is found that a close correlation exists between the two construction cost indices, keeping in mind the 12 month lag between the two datasets.

57

Figure 18: Source:(Authors’ calculations)

-

20.00

40.00

60.00

80.00

100.00

120.00

1998 2000 2002 2004 2006 2008 2010 2012 2014 2016

Gauteng Construction Cost Index2011 = 100

Buildings Completed Plans Passed

In Figure 19, the construction cost index is displayed against the Tshwane crane index. It should be borne in mind that the two indices are on different scales and is therefore only indicative of general direction of movement.

Figure 19: Source:(Authors’ calculations)

R² = 0.9614

50.00

60.00

70.00

80.00

90.00

100.00

110.00

50.00 60.00 70.00 80.00 90.00 100.00 110.00

xednI tsoC( detelpmoc gnidliuB

-201

1 =

100)

Building plans passed (Cost Index - 2011 = 100)

Gauteng Building Plans Passed Cost Index vs. 12 Month Lag Buildings Completed Cost Index

97.00

98.00

99.00

100.00

101.00

102.00

103.00

104.00

105.00

0

50

100

150

200

250

300

350

400

2010 2011 2012 2013 2014 2015 2016

Cons

tructio

n Co

st In

dex

Cran

e In

dex

Year

Construction Cost Index vs Crane Index

Tshwane crane index Building plans cost index

58

Figure 20: Source:(Authors’ calculations)

R² = 0.9614

50.00

60.00

70.00

80.00

90.00

100.00

110.00

50.00 60.00 70.00 80.00 90.00 100.00 110.00

xednI tsoC( detelpmoc gnidliuB

-201

1 =

100)

Building plans passed (Cost Index - 2011 = 100)

Gauteng Building Plans Passed Cost Index vs. 12 Month Lag Buildings Completed Cost Index

97.00

98.00

99.00

100.00

101.00

102.00

103.00

104.00

105.00

0

50

100

150

200

250

300

350

400

2010 2011 2012 2013 2014 2015 2016

Cons

tructio

n Co

st In

dex

Cran

e In

dex

Year

Construction Cost Index vs Crane Index

Tshwane crane index Building plans cost index

Figure 21 provides the correlation between the two indices. The construction cost index again is lagging 6 months behind the crane index. It is evident that although not a very high correlation, it does provide some similarity in movement. The main cause for the difference between the two and thus the lower R2, is that the crane index is substantially more volatile. This could be indicative that it is only affected by certain types of construction and not all construction activity in general.

Figure 21: Source:(Authors’ calculations)

R² = 0.6313

97.00

98.00

99.00

100.00

101.00

102.00

103.00

104.00

105.00

0 50 100 150 200 250 300 350 400

eulaV xednI tsoC noitcurtsnoC

Crane Index Value

Crane Index vs. 6 Month Lag Construction Cost Index

59

5. ConClUsion

60

The following can be concluded:

• Thecraneindexisfoundtobeveryvolatileandwouldthushavelimitedcapabilityto predictthemovementinbuildingactivitybecause,aspreviouslyindicated,arelative smallchangeinnumberscanhavealargeinfluenceontheindex.Evenwherea physicalcountisdoneandwheretheintervalbetweenthesecountsismorefrequent, thisphenomenonispresent,e.g.intheRLBindexwheretherewasan81%increase betweenquarters2and4in2015.

• Inordertoobtainmoreaccurateresults,itissuggestedthataphysicalcountmustbe doneonabi-annualbasis,whichwouldgiveabetterindicationofcranemovement.

• Ifpossible,thenumberofcranescanbesubdividedintodifferentactivitysectors wherethecranesaresituated,suchascommercial,residential,education,hotel,etc. Thiswillprovideanindicationinwhichsectorsthemajorityofbuildingactivitytakes place.

• Thebiggestbenefitofthecraneindexisseeninitsabilitytoindicateaconcentrationof highdensitydevelopment.Asthecraneindexisgenerallyassociatedwithhighdensity development,itisalsogenerallyfoundthatthesecranesarecountedincloseproximity toeachother.

61

6. RefeRenCes

62

AgarwaL B.L. 2009. Basic statistics. New age international, (P) Ltd. Publications, New Delhi, In-dia.

Akintoye A., Bowen P. and Hardcastle C. 1998. Macro-economic leading indicators of construc-tion contract prices. Construction management and economics. Vol.16, 1998.

Akintoye S.A. 1991. Construction tender price index: modelling and forecasting trends. Unpub-lished PhD thesis, University of Salford, UK.

Archer, W R & Ling, D C 1997. The three dimensions of real estate markets: Linking space, capi-tal, and property markets, Real Estate Finance, 14:7-14.

Archour-Fischer, D 1999. An integrated property market model: A pedagogical tool, Journal of Real Estate Practice and Education, 2 (1).

Ashworth A. 1991. Cost studies of buildings. Longman Group Ltd, UK.

Beeston D.T. 1983. Statistical methods for building price data. E & FN Spon, London.

Boshoff, D.G.B. 2010. The impact of affordability on house price dynamics in South Africa, Acta Structillia, 17(2):126-148

Boshoff, D.G.B. 2013. Empirical Analysis of Space and Capital Markets in South Africa - a review of the REEFM- and FDW models, South African Journal of Economics and Management Sci-ences, 16(4)

Brook D.E. 1974. Building cost index. How it works and how to use it. BER Building Survey no.21, p.5-18.

Brook D.E. 1985. The BER Building cost index: recent refinements. BER Building Survey no. 65, 46-65.

Building Cost Information Services. 2009. BCIS international cost elements enquiry report. Lon-don, October 2009.

Chen, N, Roll, R & Ross, S A 1986. Economic forces and the stock market, Journal of Business, 59:383-403.

Corcoran, P J 1987. Explaining the commercial real estate market, Journal of Portfolio Manage-ment, 13:15-21.

Davis Langdon Management Consultancy. 2008. Report: Review of BERR construction price and cost indices. London, March 2008.

Diewert W.E. 2001. The consumer price index and index number theory: a survey. Discussion paper, Department of Economics, The University of British Columbia. Vancouver, Canada.

Dipasquale, D & Wheaton, W C 1992. The markets for real estate assets & space: A conceptual framework, Journal of the American Real Estate and Urban Economics Association, 20 (1): 187-97.

63

Du Toit, H 2002. Appraisal of the Fischer-DiPasquale-Wheaton (FDW) real estate model and de-velopment of an integrated asset market model (IPAMM). Unpublished treatise submitted in part fulfilment of the requirements for the MSc (Real Estate), University of Pretoria.

Eurostat. 2008. European price statistics: an overview. European Commission. Eurostat statisti-cal books, Luxembourg. 2008 edition.

Ferry D.J., Brandon P.S. and Ferry J.D. 2003. Cost planning of buildings. 7th Edition. Blackwell Publishing, UK.

Fisher, J D 1992. Integrating research on markets for space and capital, Journal of Real Estate and Urban Economics Association, 20(1): 161-80.

Fisher, J D, Hudson-Wilson, S &Wurtzebach, C H 1993. Equilibrium in commercial real estate markets: Linking space and capital markets, Journal of Portfolio Management, 19:101-107.

Flemming M.C. and Tysoe B.A. 1991. Spon’s construction cost and price indices handbook. 1st Edition. E & F.N. Spon, London.

Hassanein A.A.G. and Khalil B.N.L. 2006. Building Egypt – a general indicator cost index for the Egyptian construction industry. Engineering construction and architectural management. Vol.13 no.5. Emerald Group Publishing Ltd.

Hendershott, P H & Ling, D C, 1984. Prospective changes in tax law and the value of depreciable real estate, Journal of the American Real Estate and Urban Economics Association, 12:297-317.

International Labour Organisation. (2004a). Consumer price index manual: theory and practice. Geneva, Switzerland. ISBN 92-2-1136999-X.

International Labour Organisation. (2004b). an introduction to consumer-price index methodol-ogy. Geneva, Switzerland.

Kilian W.F. 1980. An assessment of the BER building cost index. Building survey no. 44, January 1980. University of Stellenbosch, Stellenbosch, South Africa.

Kirkham R. 2007. Ferry and Brandon’s cost planning of buildings. 8th Edition. Blackwell publish-ing, UK.

Levine, D M, Berenson, M L & Stephan, D 1998. Statistics for managers. New Jersey: Prentice Hall.

Marx H.J. 2005. Correctness of the BER building cost index. Research report: Department of Quantity surveying and Construction Management, University of the Free State, Bloemfontein. June 2005.

McCabe B.Y, O’Grady J.O. and Walker F.A. 2002. A study of construction cost sources. Annual conference of the Canadian Society for Civil engineering. Montreal, Canada. June 2002.

Mohammadian R. and Seymour S. 1997. An analysis of some construction price methodologies. Statistics Canada, price division. Catalogue no. 62, F0014, MPB, no. 2. April 1997.

64

Prism Economics and Analysis and The Department of Civil Engineering, University of Toronto. 2001. A guide to construction cost sources: What can they tell us about competitiveness? Univer-sity of Toronto, Canada. November 2001.

Rider, Levett, Bucknall (RLB). Available at http://rlb.com/

Seeley. I.H. 1996. Building economics (4th edition). Macmillan Press Ltd. London, UK.

Segalla U. 1991. The BER building cost index – an overview. Studies in economics and econo-metrics. Vol. 15, No. 1. ISSN 0379-6205.

Simonton W.E. 2004. Creating construction cost indices and examining their cyclic behaviour. Un-published MSc (Real estate development) treatise, Massachusetts Institute of Technology, USA.

Statistics Directorate: European Community, organisation for economic co-operation and devel-opment. 1997. Sources and methods: construction price indices. Statistical office of the European Community, Luxembourg.

Statistics Finland. 2001. Building cost index 2000 = 100: User’s handbook. Helsinki, Finland.

Statistics Norway. 2007. Construction price indices and house and property price indices 2006: results and methods. Oslo, Norway. ISBN 978-82-537-7157-1.

Statistics Norway. 2014. Construction cost index for residential buildings, March 2014. http://www.ssb.no/en/priser-og-prisindekser/statistikker/bkibol/maaned/2014 Accessed: 11 April 2014.

Statistics South Africa. 2009. The South African CPI sources and methods manual. Release v.1. February 2009. Pretoria, South Africa.

Steel and Engineering Industries Federation of South Africa (SEIFSA). Undated. Theory and cal-culation of contract price adjustment. Johannesburg, South Africa.

Steyn A.W.G., Smit C.F., du Toit S.H.C. and Strasheim C. 2007. Modern statistics in practice. J.L. van Schaik Publishers, Pretoria, South Africa. 7th impression.

Steyn, A G W, Smit, C F & Du Toit, S H C 1989. Modernestatistiekvir die praktyk (Fourth ed). Pre-toria: J L van Schaik.

Swarup N. Undated. Construction cost indices. Presentation to the Construction Industry Devel-opment Council, India.

Van der Walt P. 1992. ‘nKontrakprysindeksvirgeboue in die Republiek van SuidAfrika. Unpub-lished PhD thesis, University of Pretoria, South Africa.

Viezer, T W 1998. Statistical strategies for real estate portfolio diversification. Doctoral Disserta-tion, Ohio State University, Columbus, OH.

Viezer, T W 1999. Econometric integration of real estate’s space and capital markets, Journal of Real Estate Research, 18(3): 503-19.

65

Wang C. and Wei Y. 1998. Model for forecasting construction cost indices in Taiwan. Construction Management and Economics. Vol. 16, 1998.

Weimer, A M 1966. Real estate decisions are different, Harvard Business Review, 44:105-112.

Yu M.K.W. and Ive G. 2006. A survey of building price index compilation methods in Britain. Proceedings of the annual research conference of the Royal Institution of Chartered Surveyors. September 2006, University College of London, UK.

Yu M.K.W. and Ive G. 2008. The compilation methods of building price indices in Britain: a critical review. Construction Management and Economics. July 2008. P.693 – 705.

66

67

food

lifebetter

economics

economic

government

gdpdevelopm

ent

Green

rateequity

returns

index

exchange

tax

bonds

business

market

gold

expectations

arbitrage

markets

gnilaw

gross

product

unionscale

asymmetric

money

efficiency

budget euro

Education

bank

demand

hedge

horizontal

human

expenditure

aid

altruismamortisation

Energy

antitrust

appreciation growth

pricing

theory

Agro - Processing

coefficient

assets

Tourisminformation

shockauctions

austrian

living

Broadband average

backwardation

balance

payments

balanced standards

services

bankruptcy

barriers

entry

exitbarter

cohesion

basis

point

bear

behavioural

beta

Safety

mac

finance

black

economy

black-scholes

boom

bust

bounded

rationality

brand

bretton

woods

bubble

bull

social

confidence

cycle

buyer

responsibility

gini TRT

Health

WiFi

food

lifebetter

economics

economic

government

gdpdevelopm

ent

Green

rateequity

returns

index

exchange

tax

bonds

business

market

gold

expectations

arbitrage

markets

gnilaw

gross

product

unionscale

asymmetric

money

efficiency

budget euro

Education

bank

demand

hedge

horizontal

human

expenditure

aid

altruismamortisation

Energy

antitrust

appreciation growth

pricing

theory

Agro - Processing

coefficient

assets

Tourisminformation

shockauctions

austrian

living

Broadband average

backwardation

balance

payments

balanced standards

services

bankruptcy

barriers

entry

exitbarter

cohesion

basis

point

bear

behavioural

beta

Safety

mac

finance

black

economy

black-scholes

boom

bust

bounded

rationality

brand

bretton

woods

bubble

bull

social

confidence

cycle

buyer

responsibility

gini TRT

Health

WiFi

food

lifebetter

economics

economic

government

gdpdevelopm

ent

Green

rateequity

returns

index

exchange

tax

bonds

business

market

gold

expectations

arbitrage

markets

gnilaw

gross

product

unionscale

asymmetric

money

efficiency

budget euro

Education

bank

demand

hedge

horizontal

human

expenditure

aid

altruismamortisation

Energy

antitrust

appreciation growth

pricing

theory

Agro - Processing

coefficient

assets

Tourisminformation

shockauctions

austrian

living

Broadband average

backwardation

balance

payments

balanced standards

services

bankruptcy

barriers

entry

exitbarter

cohesion

basis

point

bear

behavioural

beta

Safety

mac

finance

black

economy

black-scholes

boom

bust

bounded

rationality

brand

bretton

woods

bubble

bull

social

confidence

cycle

buyer

responsibility

gini TRT

Health

WiFi

food

lifebetter

economics

economic

government

gdpdevelopm

ent

Green

rateequity

returns

index

exchange

tax

bonds

business

market

gold

expectations

arbitrage

markets

gnilaw

gross

product

unionscale

asymmetric

money

efficiency

budget euro

Education

bank

demand

hedge

horizontal

human

expenditure

aid

altruismamortisation

Energy

antitrust

appreciation growth

pricing

theory

Agro - Processing

coefficient

assets

Tourisminformation

shockauctions

austrian

living

Broadband average

backwardation

balance

payments

balanced standards

services

bankruptcy

barriers

entry

exitbarter

cohesion

basis

point

bear

behavioural

beta

Safety

mac

finance

black

economy

black-scholes

boom

bust

bounded

rationality

brand

bretton

woods

bubble

bull

social

confidence

cycle

buyer

responsibility

gini TRT

Health

WiFi

CITY OF TSHWANE

BETTER LIFE INDEX2016

One Nation, One Capital

Advancing Tshwane Vision 2055