the impact of natural disasters on income and poverty ... · poverty: framework and some evidence...

BARD meets with UnionBARD meets with UnionBARD meets with UnionBARD meets with UnionBARD meets with UnionParishad on local povertyParishad on local povertyParishad on local povertyParishad on local povertyParishad on local povertymonitor ingmonitor ingmonitor ingmonitor ingmonitor ing 5 5 5 5 5

66666

The impact of natural disasters on income andpoverty: framework and some evidence fromPhilippine households♣

www.pep-net.orgISSN 1908-6889Volume XI, Number 1December 2013

IntroductionAlthough understanding the impact ofnatural disasters on income and povertyat the household level is important fordisaster-prone and poverty-strickencountries, empirical works on thesubject have been limited so far. Amajor reason behind this is thatstandard national household surveysgenerally do not include the collectionof data and information on naturaldisasters (De la Fuente et al. 2009). Inthe Ph i l i pp ines , howeve r , theCommunity-Based Monitoring System(CBMS) Household Profile Survey hasstarted to include selected naturaldisaster-related data and informationamong those that it col lects andmonitors.

This Policy Note provides a framework foranalyzing the impact of natural disasterson household income and householdpoverty and empirically estimates theeffects of natural disasters on householdincome using 2011 CBMS data for PasayCity, Metro Manila. The Note summarizessome of the results and findings of arecent study entitled “Disasters, Povertyand Coping Strategies: The Frameworkand Empirical Evidence from Micro/Household Data - Philippine Case”conducted by the Philippine Institute forDevelopment Studies (PIDS) and fundedby the Economic Research Institute forASEAN and East Asia (ERIA).

CBMS Philippines to holdCBMS Philippines to holdCBMS Philippines to holdCBMS Philippines to holdCBMS Philippines to hold10th National Conference10th National Conference10th National Conference10th National Conference10th National Conference 88888

CBMS Team in PakistanCBMS Team in PakistanCBMS Team in PakistanCBMS Team in PakistanCBMS Team in Pakistanassured of support byassured of support byassured of support byassured of support byassured of support bydistrict officials in Punjabdistrict officials in Punjabdistrict officials in Punjabdistrict officials in Punjabdistrict officials in Punjab

InsideInsideInsideInsideInsideFramework of analysisA general framework for studying theimpact of natural disasters at the householdlevel can be developed based on Lindell andPrater (2003). In summary, the physicalimpact of a natural disaster on householdsis determined by geographical andnatural factors, hazard agentcharacter ist ics, hazard mit igationpractices, and emergency preparednesspractices (Figure 1). The framework furtherexplained that after the physical impact,

TRAIL OF DEVASTATION. Super Typhoon Haiyan/Yolanda struck Eastern Visayas onNovember 8, 2013 and left a trail of destruction and tragedy in its wake includingtotally washed-out homes in the coastal barangays of the municipality of Balangkayan,Eastern Samar.

♣ Reprinted with permission from the authors. This article first appeared in the November 2013edition of the Policy Notes of the Philippine Institute for Development Studies (PIDS).

Danilo C. Israel and Roehlano R. Briones

CBMS Team in BoliviaCBMS Team in BoliviaCBMS Team in BoliviaCBMS Team in BoliviaCBMS Team in Bolivialaunches websitelaunches websitelaunches websitelaunches websitelaunches website 77777

CBMS results galvanizeCBMS results galvanizeCBMS results galvanizeCBMS results galvanizeCBMS results galvanizeofficials into actionofficials into actionofficials into actionofficials into actionofficials into action 77777

2 CBMS Network UpdatesDecember 2013

Research ResultsResearch ResultsResearch ResultsResearch ResultsResearch Results

the social impact of a natural disaster onthe households fol lows, which isdetermined by the physical impact, aswell as household recovery resources orcoping mechanisms, extra assistance,and sociodemographic and economicfactors.

The said framework can also be extendedto allow the identification of the linksamong natural disasters, householdincome, and household poverty (Figure2). Specifically, the social impact ofnatural disasters can be classified intofour types: socioeconomic,sociodemographic, sociopolitical, andpsychosocial factors. Individually andcollectively, these factors are likely toreduce household income in general andworsen poverty among poor householdsin particular (with the possible exceptionof politically favored and well-organizedpoor households actually benefiting fromoutside assistance after the occurrenceof natural disasters). Consequently,reduced incomes and increased povertyamong poor households result in theirreduced access to adequate nutrition,education, health and other socialservices, and overall welfare.

Some evidence: the case of Pasay CityAn attempt was made to empiricallyestimate the relationships among naturaldisasters, household income, andhousehold poverty given available data(Israel and Briones 2013). The 2011 CBMS

Figure 1. Model of disaster impact at the household level Household Profile Survey was used becauseit contained data and information that can beutilized to measure the occurrence andfrequency of natural disasters, householdincome, and other household variables.Pasay City was selected because theauthors were granted permission by thecity government to use its data.

Furthermore, the city is among thehardest hit areas by floods, includingthose caused by typhoons, in MetroManila. While the CBMS survey gathereddata and information on other forms ofnatural disasters, typhoons and floodswere considered because they are moreapplicable for Pasay City. The sampleconsisted of 70,326 households.

Figure 2. Social impact of natural disasters at the household level

Social

impact

Socio-economic

Socio-demographic

Socio-political

Psycho-social

Direct

Indirect

Property loss

Loss of life

Loss of work

Others

Higher prices

Scarcer food

Others

Dislocation of households

Destruction of dwelling

Others

Rise of conflicts

Rise in favoritism/ patronage

Others

Mental stress

Emotional breakdown

Others

Household income is likely

lower; poor households are

likely to get poorer; poverty

is likely to worsen

Poor households

are likely to be dislocated; poverty is likely to worsen.

Politically less connected, less organized poor households are

likely to lose; poor

households headed by women and

minorities are likely to lose.

Poor households

who have less resources for

medication are likely to recover slower and thus

work less; poverty is likely

to worsen.

Reduced

household

welfare

Geographical and natural factors, hazard agent characteristics

Hazard mitigation practices

Emergency preparedness

practices

Physical impact

Household recovery resources or coping

mechanisms and extra assistance

Social impact at the household

level

Sociodemographic and economic factors

CBMS Network Updates 3December 2013


The following equations were estimatedwhile the definitions of the variables andthe expected signs of the coefficients arepresented in Table 1. The choice of thevariables was limited by the availabledata from the CBMS survey.

LnINCOMEi = a0 + a1SEXi + a2LnAGEi +a3STATUSi + a4LnEDUCATIONi + a5LnSIZEi +a6OFWi + a7PROGRAMi + a8CALAMITYi + ei

and

LnINCOMEi = a0 + a1SEXi + a2LnAGEi +a3STATUSi + a4LnEDUCATIONi + a5LnSIZEi +a6OFWi+ a7PROGRAMi + a8CALAMITY1i + ei

The descriptive statistics of the variablesconsidered are shown in Table 2. Theaverage annual income of the householdswas PHP 96,509.34; 76 percent of thehousehold heads were male; the averageage of the household heads was42.88years; 51 percent of the householdheads were married; the average numberof years of schooling of the householdheads was 10.93 years; the averagehousehold size was 3.85 members; 3percent of the households received OFWcash remittance; 49 percent of thehouseholds were recipients of agovernment program; and 29 percent ofthe households were affected by atyphoon and/or flood in the last 12months. Furthermore, of the households,71 percent were not affected by a typhoonor flood; 5 percent were affected once bya typhoon and/or flood; 7 percent wereaffected twice by a typhoon and/or flood;and 17 percent were affected more thantwice by a typhoon and/or flood.

The regression results using CALAMITY asthe natural disaster occurence variableare presented in Table 3 (first threecolumns). All the coefficients of theindependent variables are of the expectedsigns and levels of significance, exceptthat for SEX. The adjusted R-squared is0.178 which is low but is expected incross-section data. The results forCALAMITY indicate that households thatwere affected by a typhoon/flood in thelast 12 months tend to have annual percapita incomes that are significantlylower, by 6.72 percent, than thosehouseholds that were not affected.

Table 1. Definitions of variables and expected signs of coefficients

Table 2. Descriptive statistics of variables used in the estimation

Table 3. Regression results using OLS and Hubert-White estimator(with CALAMITY as independent variable)


Given the cross-section nature of thedata, one may suspect thathomoscedasticity of the error term neednot hold. The Cook-Weisberg test indeedrejected the nul l hypothesis ofhomoscedastic error term at 1-percentsignificance level. Therefore, standarderrors were re-estimated with the samecoefficient values applying the Hubert-White correction. The t-values consistentwith the robust standard errors are alsoshown in Table 3 (last two columns).Specifically, the corrected t-valuesremain statistically significant at lowprobability thresholds.

The results using CALAMITY1 as thenatural disaster frequency variable arepresented in Table 4. The findings aresimilar to those obtained using CALAMITYas the natural disaster var iable.CALAMITY1 also has the expected signand is statistically significant, even withrobust standard error (which is preferredas homoscedasticity is again rejected at5-percent level of significance). Thecoefficient value implies that eachincrement in the CALAMITY1 indexreduces annual per capita householdincome by about 2.18 percent.

Conclusion and recommendationsBoth the occurrence and frequency oftyphoons and/or floods have significantand negative effects on household per

capita income as the case of Pasay Cityhas shown. These effects would havesignificant and negative impacts on poorhouseholds in the city whose povertysituation would likely worsen with naturaldisasters. The results support the findingsof previous studies (e.g., GOP 2009a, GOP2009b).

The finding that poor households lose asignificant portion of about 7 percent oftheir per capita incomes due to naturaldisasters hopeful ly would helpgovernment decide more accurately thelevel of funding it has to allocate forshort-term disaster relief and long-termrehabilitation. The result that incomelosses differ depending on the frequencyof natural disasters impl ies thatassistance to poor households may bedifferentiated based on the number oftimes they are affected by disasters.

If and when data are available, studieswith a larger regional or national coveragefurther elucidating the effects of naturaldisasters on household income andpoverty in terms of the type of impact(e.g. , socioeconomic andsociodemographic), location (e.g., urbanand rural areas), sector group (e.g.,agricultural and nonagricultural), andother relevant classifications may beconducted. A study on the reverse impactof household poverty on natural disasters

is also worth pursuing. The use ofhousehold income as the variable ofinterest in the study is necessitated byavailable data and resource constraints.In future studies, other poverty-relatedindicators can be considered, includinghousehold consumption, education,health, nutrition, and other similarvariables.

ReferencesDe la Fuente, A., L.F. Lopez-Calma, and

A. Revi. 2009. Assessing therelationship between naturalhazard and poverty: aconceptual and methodologicalproposal. Bangkok: UnitedNations DevelopmentProgramme.

Government of the Philippines (GOP).2009a. Typhoons Ondoy andPepeng: post-disaster needsassessment. Joint initiative ofthe Government of the Republicof the Phi l ippines,Representatives of the PrivateSector and Civi l SocietyOrganizations, Mult i lateralDevelopment Partners, andBi lateral DevelopmentPartners.

———. 2009b. Typhoons Ondoy andPepeng: post-disaster needsassessment sector reports.Joint init iat ive of theGovernment of the Republic ofthe Phi l ippines,Representatives of the PrivateSector and Civi l SocietyOrganizations, Mult i lateralDevelopment Partners, andBi lateral DevelopmentPartners.

Israel, D.C. and R.M. Briones. 2013.Disasters, poverty and copingstrategies: the framework andempirical evidence from micro/household data - Philippinecase. Final report submitted tothe Economic ResearchInstitute for ASEAN and EastAsia (ERIA).

Lindell, M.K. and C.S. Prater. 2003. Assessingcommunity impacts of naturaldisasters. Natural Hazards Review4(4):176–186.

Table 4. Regression results using OLS and Hubert-White estimator(with CALAMITY1 as independent variable)


❋


News UpdatesNews UpdatesNews UpdatesNews UpdatesNews Updates

BARD meets with Union Parishadon local poverty monitoring

BMS project officials from theBangladesh Academy for RuralDevelopment (BARD) met withthe representatives of

Mohammadpur (West) Union Parishad onNovember 29, 2013 to discuss projectactivities and details of the revised workplan for the implementation of the LocalLevel Poverty Monitoring System (LLPMS).

Thirteen elected representatives and theSecretary of the Union Parishad attendedthe meeting, which was presided byShamsul Hoq Sarker, chair ofMohammadpur (West) Union Parishad.

The meeting agreed on the following:

• The Chairman of Mohammadpur(West) Union Parishad hasformed a CBMS projectimplementation committee toassist the BARD officials inundertaking the var iousactivities of the CBMS project.Members of the di f ferentstanding committees of UnionParishad will be the generalmembers of the committeewhile the Chairman of UnionParishad wi l l chair thecommittee and the UnionParishad Secretary will be thecommittee secretary.

• A project brigade was formed inUnion Parishad for CBMS–Bangladesh. Brigade memberswill include all the 12 electedmembers of the Union Parishad.Their basic functions are (i) tomonitor data col lect ion(including tablet-based datacollection), and (ii) to assist inthe preparation of digitalmapping. The data will be usedin the preparation of the WardInformation Book (WIB). Aftercompletion of the WIB, thebrigade wi l l identi fy the

problems of the respectivewards with emphasis on youthemployment andentrepreneurship. It will alsoorganize ward meetings to getfeedback and inputs from thepeople before finalizing theUnion Development Plan.

• The Union Parishad is still usingthe previous year’s (2004) WIBsin its decisionmaking on suchmatters as select ion ofbeneficiaries for governmentsafety-net programs. Afterupdating the WIBs throughtablet-based and paper-baseddata collection, the UnionParishad will use the data forbetter t ransparency andaccountability in its decision-making process.

In his closing remarks, the chairman ofthe meeting said that in 2004 they were

still not fully aware of the importance ofthe database. But when they startedusing the database they benefitedsignificantly and got maximum supportfrom the government because the Upazila(sub-district) Parishad were informedthat Mohammadpur (West) Union had itsdatabase and WIBs for makingdevelopment-related decisions. Finally,he thanked the CBMS Network and BARDofficials for selecting their Union Parishadas a project partner.

The CBMS project t i t led“Institutionalizing Local Level PovertyMonitoring System in Bangladesh” isbeing implemented by BARD. The projectaims to develop an inst itut ionalmechanism for implementing acommunity-based poverty monitoringsystem with emphasis on youthemployment at the local level, and toensure the effective use of CBMS informationby the local government functionaries inpreparing development plans.

C

❋

CONSULTATIVE & ADVOCACY MEETING. The CBMS-Bangladesh team has completed itsfirst consultative and advocacy meeting with Mohammadpur (West) Union Parishadofficials who agreed to form a brigade to implement the CBMS in Bangladesh.



CBMS Team in Pakistan assured of supportby district officials in Punjab

he CBMS team in Pakistan hasreceived assurance of support fortheir project from the District

Coordinating Offices (DCOs) of Faisalabadand Mandi Bahauddin and the TehsilMunicipal Administration during a series ofconsultative meetings held in November2013.

The consultative meetings, conductedahead of the local elections scheduled inJanuary 2014 in the Punjab Province,introduced the community-basedmonitoring system (CBMS) methodologyto district and local governments,discussed the list of indicators for the firstand second phase of the CBMS project,and gathered suggestions from projectstakeholders. The CBMS team alsoreceived endorsement letters from districtand local governments to carry out fieldactivities, and were helped to understandthe new delimitations of the unioncouncils in rural areas and municipalcommittees/corporations in urban areas

the community development initiatives thatthey can take up as topics of research. Theproject had attracted the interest of economicand sociology departments, which are themain source of interviewers for the project’sdata collection activities.

Meanwhile, the CBMS team has alsojumpstarted its resource mobilizationactivities by exploring potential collaborationson youth employment and entrepreneurshipwith organizations that have the sameobjectives. These organizations include theErwing Marion Kauffman Foundation, theUSAID Ambassador Program, the RoyalEmbassy of Norway with its projects andprogram support, and the Embassy of Denmarkin Pakistan.

The CBMS project, which started in September2013, is implemented by the ResearchAnalytics International (Pvt.) Limited incollaboration with the Pakistan Institute forDevelopment Economics. The project aims towork closely with local stakeholders, includingeducational institutions, to engage incommunity research and to help identify keyindicators that require specific policy attentionat the grassroots level, particularly poverty-related and youth employment indicators.

based on the new Local GovernmentOrdinance 2013 of the PunjabGovernment.

The project team also visited local educationalinstitutes to introduce the project as part of

MEETING WITH OFFICIALS OF THE FAISALABAD DISTRICT COORDINATING OFFICE. TheCBMS-Pakistan team, represented by Nadeem Akhtar and Shujaat Farooq, meets withofficials of the District Coordinating Office in Faisalabad, province of Punjab.

CBMS PROJECT SITE. A sketch map of Ward 28 in the district of Mandi Bahauddin.

❋

T



CBMS Team in Bolivia launches website

he CBMS Project Team fromFundacion ARU has recentlyunveiled its website: http://www.comunidad-i.info/, which is

devoted to presenting and promoting theobjectives, outputs and impacts of thecommunity-based monitoring system(CBMS). The website also features a demoof the Android application that the teamis going to use in collecting household-level data.

Fundacion ARU has also conducted atraining in October 2013 for the technicalstaff of the Cruceño Institute of Statistics(ICE) on procedures and protocol of CBMSdata collection. The training was part of aset of activities related to the pilot-testing of the CBMS methodology in thelocal government of Santa Cruz de laSierra, the capital of the Santa Cruzdepartment in eastern Bolivia.

Aside from strengthening the institutionalcapacity of ICE, Fundacion ARU and thelocal government also agreed to replicatethe CBMS methodology in other locations.

The main objective of the CBMS project in Boliviais to analyze the individual and social risks thatrecur more often among informal workers, withemphasis on the characteristics andcircumstances of workers who face differentsocial or occupational risks, the vulnerability of

T

workers, and institutions that contribute toaddress those risks. The analysis of risks, througha profile of workers and social protectionmechanisms, is expected to contribute to theformulation of social protection interventionsassociated with informal employment. ❋

he results of the community-based monitoring system (CBMS)census conducted in the GreaterTzaneen Municipality in the

province of Limpopo in South Africa have

raised a lot of interest from its localofficials.

According to the CBMS Team in SouthAfrica, the officials were alarmed by the

fact that the majority of the household heads inthe area are unemployed and that only 0.2percent of individuals earn above R50,000 perannum. As a result, their immediaterecommendation was to ensure that the localeconomic development strategy and otherrelated programs must be geared towardsimproving the above figures through the creationof decent work and sustainable livelihoods.

In addition, the Greater Tzaneen Municipalityhas committed to use the results to deal withextreme poverty through integrated deliveryof services and partnership with structures inWard 1. The process will help mobilizestakeholders and resources in a more systematicway. The local officials have also committed touse the CBMS results in empowering communitiesand in facilitating a participatory and evidence-based approach in the formulation of integrateddevelopment plans.

CBMS results galvanize officials into action

DAYCARE. Children in a daycare center run by the municipality of Greater Tzaneen.

❋

T

HOME PAGE. This photo is part of the initial or main webpage of comunidad-i.info.


The CBMS Network UpdatesCBMS Network UpdatesCBMS Network UpdatesCBMS Network UpdatesCBMS Network Updates is thequarterly newsletter of the CBMS Networkof the PEP Project. This work was carried outby the Angelo King Institute for Economicand Business Studies with financial supportfrom the International DevelopmentResearch Centre (IDRC) and the UnitedKingdom Department for InternationalDevelopment (DFID).

The UpdatesUpdatesUpdatesUpdatesUpdates may be downloaded freefrom the Project's website:http://www.pep-net.org.

For inquiries , please write or call:PEP Asia & CBMS International Network

Coordinating TeamAngelo King Institute for

Economic and Business StudiesDe La Salle University-Manila

10th Flr. Angelo King International CenterEstrada cor. Arellano Sts., Malate,

Manila 1004, PhilippinesTel No: (632) 523-8888 loc. 274DL/Telefax No: (632) 526-2067E-mail: [email protected]

[email protected]

Celia M. ReyesEditor-in-Chief

Marsmath A. Baris, Jr.Managing Editor

Joel E. BancolitaSteffie Joi I. Calubayan

Novee Lor C. LeysoAnne Bernadette E. Mandap

Jasminda A. QuilitisResearchers/Writers

Editorial Staff Editorial Staff Editorial Staff Editorial Staff Editorial Staff


CBMS NETWORK UPDATESPEP Asia and CBMS International Network Coordinating TeamAngelo King Institute for Economic & Business StudiesDe La Salle University-Manila10th Flr. Angelo King International CenterEstrada cor. Arellano Sts., Malate, Manila 1004, Philippines

NO ONE TO RECEIVE

INSUFFICIENT ADDRESS

RESIGNED

UNLOCATED ADDRESS

MOVED OUT

UNKNOWN ADDRESSEE

REFUSED TO ACCEPT

❋

CBMS Philippines to hold 10th National Conference

T

RESEARCH RESULTS. Members of the CBMSResearch Team present preliminary resultsduring the workshop facilitated by Dr.Panagiotis Karfakis of the Food andAgriculture Organization (FAO).

he CBMS Philippines is set to holdits 10th National Conference onMarch 24-26, 2014 at the CrownePlaza Manila Galleria in Pasig City.

The 3-day event will gather key officials andrepresentatives of national governmentagencies, local government units (LGUs),nongovernment organizations (NGOs), anddevelopment partner agencies to discusslatest developments in the implementationand uses of the community-basedmonitoring system (CBMS) in thePhilippines.

This year’s conference will take up, amongothers, the uses of CBMS for bottom-upbudgeting, disaster risk reduction and

management, local governance,vulnerability mapping, program design andtargeting, and impact monitoring.Highlights of the conference will includethe annual recognition of good practicesand innovations among LGUs in theinstitutionalization and use of CBMS in theirareas.

The annual CBMS Philippines nationalconference is organized by the PEP Asia-CBMSNetwork Office of the De La Salle University(DLSU)-Angelo King Institute (AKI).

For more information, please visit:http://www.pep-net.org/events/event/article/10th_cbms_philippines_national_conference/

BOOK LAUNCHING. One of the highlights of the 9th CBMS Philippines NationalConference held last year was the launching of the 4th volume of The Many Faces ofPoverty.

the impact of natural disasters on income and poverty ... · poverty: framework and some evidence...

Documents