social protection growth 2011-17

Upload: carlos-rivas

Post on 14-Apr-2018

215 views

Category:

Documents


0 download

TRANSCRIPT

  • 7/30/2019 Social Protection Growth 2011-17

    1/29

    1

    CSAEWorkingPaperWPS/201117

    SocialProtection,EfficiencyandGrowth

    StefanDercon1,UniversityofOxford

    September2011

    Abstract

    Socialprotectioncanplayanimportantroleinpovertyreductionandmakinggrowthinclusiveofthe

    poor.Attimes,itisalsoarguedthatsocialprotectioncandirectlycontributetogrowthandeconomic

    efficiency.Thepaperrevisitstheevidenceonthecostofsocialprotectiontoreducepoverty,andits

    contributiontoefficiencyandgrowth.Associalprotectionmayovercomemarketfailuresincreditand

    insurance,thepaperalsoconsiderstheroleofalternatives,suchasmicrocreditandmicroinsurance.

    Theevidenceonsocialtransfers(incashorinkind,conditionalornot)suggeststhatwhiletheyhave

    substantialpovertyandequityimpacts,theirefficiencyandgrowthimpactisunlikelytobehighnot

    dissimilartothelimitedgrowthimpactofmicrocredit.Theimplicationisthatthemainmotivationfor

    socialtransfersmustlieintheirequityorpovertyimpacts.Theevidenceoncontingenttransfers,made

    inresponsetoshockssuchasillness,droughtorunemployment,asinsocialinsurance,isthattheir

    contributiontoresolvingmarketfailuresmaybehigher,leadingtopotentiallymoresubstantialgains,

    especiallywherechildrenaretargeted.Giventheproblemswithdevelopingmarketbasedsolutionsvia

    microinsurance,thereisastrongcaseforsocialprotectioninitiativesinthisareafromanefficiency

    pointofview,tocomplementcontributionsbasedsocialinsuranceandmicroinsuranceinitiatives.

    Conditionsinconditionalcashtransferscanalsobeusedtoenhanceefficiencygains,forexampleif

    conditionstargetactivitiesorinvestmentswithclearsocialexternalities.Thepaperendswiththree

    areaswheretherecouldbepotentiallyhighgrowthimpacts:socialprotectionfocusingonchildren,

    especiallybeforetheageoffive;socialprotectionmeasurestomakemigrationsmootherandcities

    moreattractiveplacestoliveforlowskilledworkers,possiblyviaurbanworkfareschemesfocusingon

    urbancommunityassetbuilding;andsocialprotectiontargetedatadolescentsandyoungadults,

    includingtransfersconditionalontrainingfocusedonurbanlabourmarkettransitions.Inallthesecases,

    standardcashtransfersmaybetooblunttohavehighimpacts,suggestingtheneedformorecontext

    specificsmartersocialprotectionschemes.

    1ThispaperiswrittenfortheABCDEconferenceMay2011.ItdrawssomeinspirationfromtheEuropeanReport

    onDevelopment2010onSocialProtectionforInclusiveDevelopment,towhichIcontributedasacoreteam

    member.However,theviewsexpressedherearemine,asaretheerrorsandinaccuracies.

  • 7/30/2019 Social Protection Growth 2011-17

    2/29

    2

    1. IntroductionThereappearstobegeneralacceptancethatsocialprotectionhasaroletoplayinpoverty

    reductionandinmakinggrowthinclusive.Largecrisessuchasthefoodcrisisortheeconomic

    downturntendtomaketheprinciplemoreacceptable.Reasonablegrowthinthedeveloping

    andemergingeconomiescreatesthespacetolookforadditionalinstrumentstomakegrowth

    inclusive.Mostofthekeydecisionsonbroadeningsocialprotectionareinherentlypolitical:

    socialchoicesonthefairnessoftheallocationinsociety,ontherightsandentitlementsof

    thoselesssuccessfulinwealthcreationandthosefacinginsecurelivelihoods.Manyofthe

    challengesaremanagerialtheorganisationofinstitutionalisedsystemsofsocialprotection,

    thatoffersecureandcosteffectiveaccesstotheseentitlements.Thispaperrevisitsthe

    economicargumentsforandexperiencewithsocialprotection.Itsfocusisontheuseofsocial

    protectioninsomeofthepoorestsettings,notleastinAfrica,theregionwherearguablyits

    spreadisstillmostlimited.

    Inparticular,thepaperexplorestheevidencefortheeffectivenessofsocialprotectionin

    reducingpovertydirectly,butalsoviaimprovingefficiency(correctingmarketfailures)andby

    stimulatinggrowth.Inmuchdiscourseonsocialprotection,contributingtoincomechangeof

    thepoor,improvingefficiencyandstimulatinggrowthareoftenusedasiftheyarethesame;

    aswillbedevelopedfurther,theyarenot.

    Socialprotectionisabroadconceptwithmanymeanings;infact,noneofthewellknown

    dictionariesappeartohaveadefinition,illustratingthelackofasharedmeaning,anddifferent

    organisationstheWorldBank,ILO,UNRSID,DFIDallseemtouseadifferentdefinition.Inthis

    paper,adefinitionisusedasintheEuropeanReportonDevelopment(2010)onSocial

    Protection,whichreferstothreefunctions:offeringsocialinsurance,socialassistanceand

    effortstoensureaccesstosocialinsuranceandassistance.2Inthepaper,thefocusisonthe

    transactioninvolvedintheseschemes:offeringatransferincashorinkindtopoorand

    vulnerablegroups.Inthecaseofsocialinsurance,itisastatecontingenttransfer(i.e.

    dependentonsomethingoccurring,suchasillness,droughtorunemployment),andwhile

    usuallysocialinsuranceisunderstoodtobelinkedtocontributions,herewetreatissomewhat

    broader,andincludingnoncontributorystatecontingenttransfers.Socialassistanceismore

    simplyatransfer,irrespectiveofsomeeventoccurring.Bothcouldneverthelessbeconditional

    onsomething:somebehaviour(suchassendingachildtoschool,asinConditionalCash

    Transfers),someinitialconditions(suchasbeingpoorordisabled,asintargetedtransfers)or

    2ERDdefinesitasaspecificsetofactionstoaddressthevulnerabilityofpeopleslifethroughsocialinsurance,

    offeringprotectionagainstriskandadversitythroughoutlife;throughsocialassistance,offeringpaymentsandin

    kindtransferstosupportandenablethepoor,andthroughinclusioneffortsthatenhancethecapabilityofthe

    marginalisedtoaccesssocialinsuranceandassistance(ERD(2010,p.8)).

  • 7/30/2019 Social Protection Growth 2011-17

    3/29

    3

    havingpaidacontribution(asinmoreclassicsocialinsurance,suchasforhealth,

    unemploymentprotectionorapension).Thetransferscanbeincashsuchasinthecaseofa

    socialpension,orinkind,intheformofgoodsorservices,suchasfoodaid,schoolfeeding,supplementarychildfeeding,trainingwhenunemployed,ortreatmentatahealthfacilitywhen

    ill.

    Treatingsocial3protectionasessentiallyasetoftransferschemesforthepoorandvulnerable

    isarguablycontroversial,notleastwhenmanydefinitionsliketoemphasizenotsomuchthe

    transactionsinvolved,butratherthepurposeandentitlementsofthesystems(offering

    prevention,protection,andpromotionasinrecentWorldBankdocuments4,or

    transformativesocialprotectionasinSabatesWheelerandDevereux(2008),inwhichthe

    fundamentalchangeinpeoplelivesandpositionisemphasised,althougharguablythelatter

    wouldencompassafarwidersetofactionsandpoliciesthanenvisagedhere).Whilesomemay

    considerthisspin,itismorethanthat:itisastatementofintentofwhattheobjectivesofsocial

    protectionare,andthewayallpartsofthesocialprotectionsystemareintegratedtoachieve

    theseobjectives.Designingintegratedsystemswithdefinedambitionsisimportant,and

    objectiveswillgovernwhattypesandformsoftransferschemesarebeingoffered.These

    objectivesareinherentlydependentonsocialandpoliticalcontextsandchoices.

    Here,thefocusismorenarrowlyontheactualtransferofresources,andtheeconomic

    justificationforthesetransfers.Aretheyaffordable?Aretheythebestwaytospendresources

    toalleviatepoverty?Dotheycontributetoefficiency,viaaddressingmarketfailures?Howdotheycomparewithalternativemechanismstoachievetheseobjectivesforthepoorand

    vulnerable,suchasmicroinsuranceinsteadofsocialinsurance,ormicrocreditinsteadof

    transfers?Whatistheevidenceforgrowtheffects?Andhowcansocialprotectionbemore

    efficiencyorgrowthenhancing?

    Thefirstsectionwillbrieflyexplorethesizeofthechallengeandthecostinvolved.Italsolooks

    attheevidenceforthepovertyimpactsofsocialprotectiononincome,consumptionandother

    indicatorsinanumberofrecentsuccessstories.Asseenfromthatperspective,costsarerather

    substantialbutwithoutdoubt,successescanbeclaimed.Next,thepaperfocusesonthe

    contributionofsocialprotectiontoincreasedefficiencyandgrowth,andonhowtodeliver

    socialprotectionefficiently.Muchofthejustificationforusingsocialprotectioninsomeofthe

    poorestsettingsthesedaysappealstoefficiencyandgrowtharguments.Thetheoretical

    argumentsthatchangesinthewealthdistributionincludingviatransferscanhaveefficiency

    3Thesocialreferstotheroleofpublicaction;theprincipleisthatatleastpartlythesetransactionsoccuroutsidemarketmechanisms.Ofcourse,thereisnoreasonapriorithatmarketbasedprovisionandallocationaretotally

    excluded,andarolefortheprivatesectorispossible,suchasforexampleusingprivateinsuranceandhealth

    providers,asintherecentRSBYhealthinsuranceschemeinIndia.4Inparticular,intheconceptnotefortheconsultationforanewSocialProtectionandLaborStrategy20122022.

  • 7/30/2019 Social Protection Growth 2011-17

    4/29

    4

    gainsarebothwellfoundedandwellresearched.Marketfailuresininsurance,creditand

    labourmarkets,linkedtoinformationalandenforcementproblems,aswellasexternalitiesall

    leadtoinefficiencies,andinmanymodelconfigurations,notleastwhenaccompaniedsomenonconvexities(suchasindivisibilitiesintechnologies),redistributioncouldbewelfare

    enhancing,whilethepotentialgrowthconsequenceslinkedtothepoliticaldynamicsintheface

    ofinequalitysimilarlyofferefficiencygainsfromredistribution.5

    Butistheevidencereallytheretosupportthisasajustificationforsocialprotectionspending?

    Morespecifically,issocialprotectiontheappropriatesolution?Here,theconceptualarguments

    forusingsocialprotectionhavetobemoresubtle,andtheevidenceisoftennotencouraging.

    Inshort,asimpleargumentthatsocialprotectionisanessentialpartofagrowthstrategy

    becauseithashighefficiencygainsinaworldofmarketfailureandcontributestogrowth

    directlyhastobetakenmorenuanced.Anumberofmorespecifictypesofsocialprotection

    interventions,especiallyinthespaceofsocialinsuranceandcontingenttransfers,wouldseem

    tosatisfythiscriterion,butnotall,atleastaccordingtotheavailableevidence.One

    interpretationoftheevidenceisthatmostsocialprotectionistooblunttogenerateoverall

    efficiencygains.Afinalsectiondevelopssomepossibleareasofdirectcomplementarity

    betweensocialprotectionandgrowth,focusingonsocialprotectionprogrammesthatcould

    contributetothevariousprocessesofeconomictransformationthatarerequiredtomake

    growthbothsustainableandmoreinclusive.Coreexamplesdevelopedarerelatedtosocial

    protectiontosupportinternalmigration,citydevelopment,childrenandyouthlabourmarkets.

    Standardsocialprotectioninstrumentsarelikelytobetooblunttobeeffectivetohandlethese

    challenges.Carefulcostbenefitanalysisisrequiredtoensure

    2.SocialProtectionandPovertyReduction

    Howbigisthechallengeforreducingpovertyviatransfers?Thescopefortransfersto

    contributetopovertyreductionisconvenientlyexpressedbyusingasimplepovertygap

    measureofpoverty.6Whenusingitonmonetaryvaluesofthestandardofliving,itgivesa

    senseofthesizeofthetransfersneededtobringeveryoneuptothepovertyline,andhow

    muchintotalwouldneedtoberaisedforthat.Forexample,usingChenandRavallion(2008)sdataonglobalpovertygaps,relativetoa$1.25dollaradaypovertyline,thepovertygapis

    about7.6%ofthepovertyline,or9.5centsperpersonperdayacrossallthedevelopingareas

    5ReviewsareinBardhanetal.(1999),theWorldBanksWorldDevelopmentReport(2006),andBowlesetal.

    (2006)6

    ,inwhichP1standsforthepovertygap,anditismeasuredastheaveragegapinasocietywith

    populationNbetweenthepovertyline(z)andactualincomelevels(y),expressedasafractionofthepovertyline,

    withthegapzeroforthose(NP)aboveit.

  • 7/30/2019 Social Protection Growth 2011-17

    5/29

    5

    theyconsider.7Whilethesemayseemsmallnumbers,theyareequivalenttodailytransfersof

    about520millionUSdollarsatPPPvalue,orabout190billionUSdollarsayearinPPPof2005.

    ForsubSaharanAfricaalone,(withahigherpovertygapof21%),thiswouldbejustover25centsadayperday,orthiswouldbeequivalentabout200milliondollarsaday,orabout73

    billionperyear.AnotherwaytolookatitinsubSaharanAfricaisthatitisabout5%ofGDP

    expressedatPPPof2005,oraboutathirdoftaxrevenue.Yetanotherwayofexpressingthisis

    thatthetotalresourcesneededtogetridofpovertyisnowforthefirsttimeinhistorymoreor

    lessequaltototalnetOfficialDevelopmentAssistance(atleastifwetrytocorrectitforPPP),

    bothforsubSaharanAfricaorthedevelopingworldasawhole.8Thearithmeticforacountry

    likeEthiopiaissimilar,withthepovertygapabout7.4%ofGDP,about40%oftaxrevenues,but

    withcurrentaidlevelsrecentlymovingwellabovetheequivalentlevelofGDP.9

    Thiswouldseemgoodnews:fightingpovertydoesnotseemoverlycostlyandaidhascaught

    up.10Anappropriatemixofsocialprotectionschemescouldresolvepoverty.Whatmixis

    required?Someofthispovertyhastransientelements,whileotherpartsaremorechronic.

    Onecaneventrytoquantifythis.UsingdatafromDerconetal.(2005),thiscanbedonefor

    Ethiopia.Usingdataonpaneldataonlivingstandards(proxiedbyconsumptionpercapita)and

    detaileddataonshocksfortheperiod1999to2004,acounterfactualcanbeexploredofhow

    highpovertywouldhavebeenifsocialprotectionwouldhaveremovedcontingentlosseslinked

    tospecified(andarguablyinsurable)shocks,basedonaregressionmodel.Startingfromabout

    44percentpoverty,thenifallshockswereinsured,theheadcountwouldbeabout31percentinotherwords,effectivesocialprotectionwouldhaveremoved13%orabout30%ofthepoor

    frompoverty.Fortherest,contingenttransferswouldnotbeenoughtoliftthemoutof

    poverty,andsocialassistancewouldberequired.Amoreappropriatecalculationrepeatsthis

    forthepovertygap,asthiswouldrepresenttheactualresourcesneeded.Herewefindaneven

    higherpercentagethatcanbecoveredbysocialinsurance:contingenttransfersinsuringall

    7WhichareallareasofthedevelopingworldasusedinWorldBankpublications,buteffectivelyAfrica,EastAsia

    andthePacific,SouthAsia,MiddleEastandNorthAfrica,CentralAsiaandLatinAmerica.8AbackoftheenvelopecalculationbasedonOECDfiguresoftotalODAtosubSaharanAfricasuggestsanincrease

    incurrentUSdollarsofabout36to48billionbetween2005and2009;in2009thiswouldbeequivalenttoabout

    75billionUSdollarsinPPP(of2005);asimilarcalculationforglobalODAtothedevelopingworldsuggestsafigure

    ofabout200billionUSdollarsin2005PPPby2009.Toobtainthisfigure,actualODAismultipliedwiththeratio

    ofGDPatPPPin2005andnominalGDPincurrentUSdollarsin2005.9Ethiopiahasarelativelylowpovertygapincomparisontoitspovertylevel,accordingtoofficialstatistics,

    reflectingrelativelylowinequalityamongthepoor,whilethePPPcorrectionmakespovertybelowtheAfrican

    average,contrarytomostotherlivingstandardindicatorsforEthiopia,suchasenrolment,malnutritionor

    mortality.Thepovertygapisabout10%.10Usingtaxrevenuestoarrangeforredistributivetransfersisnotquiteyetonthecardsinmostofthepoor

    countries,suchasinAfrica:Ravallion(2010)calculatesthemarginaltaxratesrequired,andtheywouldreach100%

    formanycountries.

  • 7/30/2019 Social Protection Growth 2011-17

    6/29

    6

    shockscouldcover40%ofthepovertygap,whiletheremaining60%wouldneedtobecovered

    bysocialassistancepayments.

    Ifonlyitwasthateasy.Experienceshowsthatfightingpovertyusingtransferswouldrequire

    substantiallymoreresourcesthanthepovertygap.First,incomeisonlyonedimensionof

    poverty,andwhileplayingaroleinachievingotheroutcomes,investmentsinhealthor

    education,tonamejusttwo,arerelevanttoo,andwouldrequireadditionalinvestmentsto

    improve.

    Therearewellknownotherproblems.Targetingisone,asnotallresourcescaneasilybe

    transferredtothepoororatzerocost,andtargetingerrorsarelikely.Meanstestingcanbe

    costly,andinmostofthepoorestcountries,someformofadministrativeorlocalidentificationofpovertyhastobeused.Whilelocalinvolvementinidentificationhasadvantages,itdoesnot

    avoidpoliticaleconomyproblems,oftenleadingtopoortargeting(ConningandKevane,2002).

    Definitionsandconceptscanbecontestedandidentificationofthepoormaysufferfrom

    patronageandcapture.ForexampletheBelowthePovertyLineassessmentsinIndia,thebasis

    ofgainingaccesstovariousprogrammes,arerifewithpoliticalmeddling(Besley,etal.2008).

    Localidentificationoftheneedymayalsoleadtopotentialexclusionfromsafetynetsof

    particulargroupsduringcrisis.Forexample,CaeyersandDercon(2011)showhowaccessto

    foodaidinEthiopiaduringthedroughtof200203washighlycorrelatedwithconnectiontothe

    localauthorityinchargeofdistributionwhetherthisisforpoliticalreasonsorjustreflectinglackofinformationbyotherscanbedisputed.Butintheend,targetingisimperfect.

    Thereisalsotheissuethateachdollarspentmayhavealowerreturn(intermsofpoverty)than

    thevaluethatreachesthehousehold.Thisisnotaboutthepoorspendingthismoneywrongly

    butthereareotherfactors.First,therecouldbecrowdingout:inresponsetoahousehold

    receivingapublictransfer,privatetransfersmaygodown(forexamplefromfamilyornetwork

    members).WorldBank(2001)hasarguedthatbetween20and91percentofprivatetransfers

    maybedisplacedifsocialassistancewereexpandedinthedevelopingworldwhichobviously

    wouldraisehugelythecostofremovingpoverty.Morerecentandarguablymorecareful

    evidenceismoremixed.Jensen(2003)estimatedthateachrandincreaseinpublicpension

    incomefortheelderlyinSouthAfricaleadstoa0.250.30randreductionintransfersmadeby

    children.Coxetal.(2004)suggestthatadeclineinprivatetransferby$0.4foreachdollarspent

    forthepooresthouseholdsinthePhilippines.MorerecentworkinAsiahasnevertheless

    suggestedthattheeffectsarenotsohigh,withatmosta0.08reductioninprivatetransfers

    (Gibsonetal.2006).Thesizeoftheeffectwilldependontheschemeandthecontextinany

    case,itwillmakefightingpovertymoreexpensive.

  • 7/30/2019 Social Protection Growth 2011-17

    7/29

    7

    Asecondwayinwhichthevalueintermsofincreasesinlivingstandardshouseholdsislower

    thantheamountactuallyreceivedhastodowithpovertytrapslinkedtothresholdsorsome

    nonconvexityintheopportunitysetforattaininghigherlivingstandards.Toillustratethis,supposethereisapovertytrapinthefollowingway,withadynamicprocessgoverninghow

    incomeequilibriaareobtainedasdescribedinfigure1.Incomeatt+1isontheYaxis,income

    attisontheXaxis.Theprocesstranslatingincomeattintoincomeatt+1isgovernedbytheS

    shapedfunction,withasteepconvexarearesemblingalmostathreshold;incomeskeepon

    changinguntilincomesattandt+1areequal(i.e.reachedequilibrium).Thekeyimplicationis

    thatifincomeisbetweenaandbatanymomentintime,thenincomewouldtendtowards

    equilibriumA,i.e.itsettlesatthelevela.Ifincomeisbetweenbandcatsomemomentintime,

    itwouldgrowtowardsequilibriumC,withincomec.Supposenowthatpovertyishaving

    incomesomewherejustbelowb.Ifsocialtransfersareusedtoliftpeoplebelowbuptothepovertyline,thenthesewillhaveazeroimpactonpoverty:anyspendingisirrelevant,asall

    thesepeoplewilldrifttowardsAwithincomea.Theseresourceswouldeffectivelybewasted,

    evenifthemoneyisputinpeoplespockets.Onlytransfersthatliftpeoplebeyondbwillhavea

    nonzeroreturn.Whethertrapslikethisexistisanempiricalanddisputedissue.Forexample,

    Dasgputa(1997)arguesforsuchtraps,basedonhowinsufficientnutritionbelowsomelevels

    translatesintoaninabilitytobeproductivelyemployedandquicklydrifttowardssuchlow

    equilibrium.Unlessthetransferisbigenough,themarginalreturntomorenutritionwouldbe

    insufficient.Ingeneraltheevidenceforactualtrapsisnotthatstrong,butatthesametime,the

    evidencepointstoslowdynamicprocessesthatwouldbeleadtoprocessesofpersistence,withsimilarpolicyimplications(Ravallion,2008).Wereturnbelowtotheideaofthresholds

    motivatingbutalsoconstrainingsocialprotectionschemes.

  • 7/30/2019 Social Protection Growth 2011-17

    8/29

    8

    Allthesefactorssuggestthatthecostofreachingthepoorviatransfersisgoingtocostmore

    thanthecostofthepovertygapwouldimply.Evenabout190billiondollarperyearishardly

    appealingoreasilyaffordable,andgivensomanycompetingdemandsongovernment

    resourcesanddevelopmentspending,spendingonsocialprotectioninasmartwayisessential.

    Theremaybewaystodothisintermsofhowonespendsonsocialprotection(forexample,choosingwaysofspendingthatreducetransactionsortargetingcosts),spendingonitemsthat

    helpeconomicefficiency(forexample,maximizingpositiveexternalitiesorresolvingother

    marketfailures),orsocialprotectionspendingasinvestment,forexamplebybuildingassets

    withatleastasocialreturn.Inthenextsection,theseoptionswillbediscussed.

    Beforeturningtothis,itisworthreflectingonhoweffectivesomecountrieshavebeenin

    expandingtheirsocialprotectionprovisionbothintermsofcontingenttransfersasinterms

    ofsocialtransfers,andthecostinvolvedaswellanythingthatisknownabouttheiroverall

    impact.Table1alsoshowsthepovertygapasapercentageofGDPtobenchmarkthecostsinvolved,usingthe2dollaradayinPPPpovertyline(RavallionandChen2008)asformostof

    thecountriesinvolvedthiswouldbeamorerelevantpovertyline,asitwouldbeclosertothe

    nationalpovertylinesusedlocallyastheyareLatinAmericancountries,aswellasSouth

    Africa.ForIndiaandEthiopia,theearlierbenchmarkof$1.25perdayisused.First,wecansee

    thatforalltheseprogrammesthecostinvolvedisrelativelylargeinrelationtothepovertygap

    asa%ofGDPthesearebigprogrammes.Relativelyspeaking,theEthiopiaprogrammeisthe

    smallestinrelationtothepovertygap,butitisthelargestsocialtransferandworkfare

    programmesoperatinginsubSaharanAfrica,withapproximately8millionrecipients.The

    Incomeatt

    Incomeatt+1

    45O

    B

    C

    a b c

    Figure1PovertyTrapandTransfers

  • 7/30/2019 Social Protection Growth 2011-17

    9/29

    9

    IndianNREGA(NationalRuralEmploymentGuaranteeAct)programmecoveredabout34

    millionpeoplein200708,offering1.4billionworkingdaysofwork.Oportunidadescovered5

    millionhouseholds,offeringconditionalcashtransfers,andBolsaFamiliaofferedconditionalcashtransfersto11millionhouseholds2006.Themainpartofthesocialgrantsprogrammein

    SouthAfricaisthesocialpensioncomponent,reachingabout90%ofthoseeligibleonmeans

    testinggrounds;overallabout15%ofthepopulationgetonesocialgrantoranother.Finally,

    theJefesandJefasworkfareprogrammewasthekeypolicyresponsetothecrisisinArgentina

    in200203.

    Table1 Somewellknownlargesocialprotectionprogrammes

    Country Programme Cost

    (as%ofGDP)

    Povertygapusing

    $2/day(as%ofGDP)

    Impact?

    Brazil BolsaFamilia

    (transfer)

    0.4 0.6 Povertygapof

    thosereached

    wentdownby

    12%200105

    Mexico Oportunidades

    (transfer)

    0.4 0.4 19%povertygap

    declineinrural

    areas199606

    SouthAfrica SocialGrants

    includingsocial

    pensions

    3.0

    (2.0pensionsonly)

    1.7 Positivebenefits

    documentedbut

    notonoverallpoverty

    Ethiopia ProductiveSafety

    NetProgram

    (workfare)

    1.5 7.4

    (basedon1.25$

    perday)

    Limitedimpacton

    consumptionand

    nutrition;some

    asset

    accumulation

    Argentina JefeyJefas

    (workfare)

    0.8 0.2 Counterfactual

    poverty53%

    ratherthan67%

    amongrecipients,

    mainlystoppingpeoplefallinginto

    povertyduring

    crisis

    India NREGA

    (workfare)

    1.0(allSP

    programsabout

    2.0)

    2.2

    (basedon1.25$

    perday)

    Nosystematic

    impactevaluation.

    UsingdatafromWorldDevelopmentIndicators(2011),EuropeanReportonDevelopment(2010)andsources

    therein.Brazilrefersto2005povertydata,Mexicoto2000,Ethiopia2005,SouthAfrica2000,Argentina2000.

    Notethattheseestimatesareapproximateestimates,anddependonvariousassumptionsusedinthesources

    mentioned.Also,theinterpretationoftheimpactevidencecouldbecontested.

  • 7/30/2019 Social Protection Growth 2011-17

    10/29

    10

    Thetablealsoreportsonpovertyimpacts,althoughtakenfromacrossthevariousstudiesand

    programmes,itisnotalwayspossibletodistilcomparablefindings,whilemethodsarenotalwaysconvincing.Ingeneral,povertyimpactsareratherpositive,althoughinIndialittle

    systematicnationalevidenceisthusfaravailable.11Still,thedifficultiesofbringingdown

    povertygapsdowniswellillustratedbythescaleoftheimpactsreported.12Forthoseforwhom

    therearedataavailable,theymakepositiveevenifarguablystillmodestinroadsintopoverty.

    Ofcourse,iftheseprogrammeswerenotcredibleandpeopleperceivethatthisisonlyaone

    offopportunity,consumptionsmoothingwouldhaveencouragedpeopletosavemuchofthe

    contributions,soaonetooneimpactonpovertyshouldnotbeexpected.Overall,this

    illustrateswellthattryingtogetaneffectivesocialprotectionprogrammethatwillfightincome

    povertywillneverbecheap,howeversmallpovertygapnumbersmayseemtobe.Fightingpovertyisnotsimplyamatterofspending.

    Othershaveappearedtoarguethecontrary.TheILOhastriedtocalculatethecostofabasic

    packageofsocialprotectionincludingfreebasichealthcare,childbenefit,cashtransferstothe

    poorandunemployed,anddisabilityandoldagepensions,allatmodestlevelsneverpayinga

    directtransferofmorethan$1inPPPperday(ILO,2008).Theysimulatedthecostfora

    numberofcountries,suchasEthiopia,Kenya,Tanzania,Cameroon,BurkinaFaso,Indiaand

    Pakistan,andsuggestedthattoofferthispackagewouldcostabout3.7to10.6%ofGDP.13In

    relationtotypicaltaxreceiptsof1018%,thisissubstantial.ILO(2008)considersthisneverthelessafeasibleplanworthallocatingresourcestointheshortrun.

    Still,thisisnottosaythatsocialprotectionisjusttooexpensive.Povertyreductionisan

    importantgoalandmanyotherinterventionsmaywellhavealowerpovertyimpact.For

    example,inLatinAmericaorinSouthAfrica,itishardtothinkofalternative,defined

    interventionsoutsidesocialprotectionthathavehadorcouldhavethispovertyimpact.Butthis

    comesatacost,andrathercarefulthoughtneedstobegiventowhatworksandhowmostcan

    begainedfromsocialprotection.Intheend,politicalsupportforsuchschemesiskey,andthe

    wayneedsandrightsofthepoorarticulatedinpoliticalchoices.Belowwefocusoneconomic

    benefits,whichmayhelpinarguingforthesechoices.

    11AsmallscaleimpactevaluationstudybyRaviandEngler(2009)suggestedthatfoodexpenditureincreased15%

    forparticipants,andevenmorecomparedtothosewhowantedtoworkbutsomehowwereexcluded.12AnexamplefromChina,reportedinRavallion(2007)confirmsthisforaprogrammeaimingtobringallurban

    householdsaboveaparticularpovertyline,theMinimumLivelihoodGuaranteeScheme.By2003,22million

    peoplewerecovered,or6%ofurbanhouseholds.Althoughambitious,thepovertygapwasonlyreducedby12%.

    Thepovertyheadcountamongtherecipientswasreducedfrom57to45%duetotheprogramme.13TheEuropeanDevelopmentReport(2010)questionedtheassumptionsandsuggestedthatacostof5to12

    percentwasmorerealistic.

  • 7/30/2019 Social Protection Growth 2011-17

    11/29

    11

    3. Efficientsocialprotection

    Inthissection,weaddresstwoquestions.First,dotransfersorcontingenttransfersimproveeconomicefficiency?Andsecondly,aretherewaysofdeliveringtransfersthatpromotemore

    economicefficiencyaswell?Theargumenthereisnotthatallsocialprotectionmustimprove

    efficiencybutitwouldpossiblyhelp,inaworldofconstrainedresources,tospendonthose

    programmesthatmayhavetheseeffects.Intheprocess,wecontrastsocialprotectionto

    alternativemoremarketorientedinterventionsthataimtoresolvesimilarmarketfailures,

    microcreditandmicroinsurance,andforwhichsimilareconomicbenefitsasforsocial

    protectionarebeingclaimed.

    SocialassistanceandcreditmarketfailureIsthereareturntoofferingsocialtransfers?Therewouldbeoneifthosepeoplereceivingthe

    transferswouldhavebeenconstrainedintermsofinvestmentsinassetsandactivities.Ifcredit

    constraintsbind,thenprofitableopportunitiesmayhavebeenforegone.Asthepoorhavefew

    assets,theycannotoffercollateralbuttransfersmayimprovetheirassetpositionorofferthem

    directlythecashrequiredtodotheinvestments.Similarly,ifinsurancemarketsorinformal

    substitutesaremissing,thenshockswouldleadtoassetdepletion,andlowerfuturereturns

    wouldfollow,aprocessthatcanbeavoidedifcontingenttransfersofferprotection.

    Furthermore,withoutinsurance,fearofbadoutcomeswouldleadtolessrisktakinginactivity

    orassetportfolios,leadingalsotolowerreturns.Thepromiseofhavingaccesstocontingenttransferswouldalsoavoidthis.

    Howmuchevidenceistherefortheseprocesses?Quiteafewmodelspredictthattransfers,by

    boostingassetlevelsofthepoor(andreducinginequality)wouldleadtomoreaccesstocredit

    andthereforehelptoovercomethemarketfailure(BanerjeeandNewman(1993),Benabou

    (1996),DasguptaandRay(1986),Bardhanetal.(1999)).Introducingthresholdsor

    indivisibilities,sothatquitelargeinvestmentsumsareneededtoenterlucrativeactivities

    wouldsuggestpotentiallyevenlargereffects,providedtransfersarelargeenough(andhelpto

    overcomethethreshold,asdiscussedintheprevioussectionandfigure1).

    Thereisplentyofsuggestiveevidencethatcreditconstraintsmatterformanypoor,even

    thoughhardevidenceisdifficulttocomeby,asotherexplanationsforobservedbehaviours

    cannotbeeasilydismissed(Ray(2008)).Still,overallevidenceseemstosuggestthatcredit

    constraintsmatterformanyinthedevelopingworld(BanerjeeandDuflo(2005);WorldBank

    (2006),Ravallion(2006)).Evidenceonpovertytrapsishardertocomeby,butsomedynamic

    behavioursconsistentwithpovertytrapsarefoundinKenyaandIndia(Lybbertetal.,(2006),

  • 7/30/2019 Social Protection Growth 2011-17

    12/29

    12

    DerconandOutes(2011)).Asseenfromthisperspective,offeringtransfersaspartofsocial

    protectioncouldhavehighreturns.

    Isthereevidencefromsocialtransfersprogrammethatthetransfersappeartoovercomecredit

    marketfailureandleadpeopleintovirtuouscycleofproductiveinvestments?Here,the

    evidenceisnotverystrong,atleastforphysicalassets.Formostprogrammes,thereportsshow

    someinvestmentinproductiveinvestmentswithhighreturns.Gertler,etal.(2006)foundthat

    12%ofPROGRESA(thepredecessorofOportunidades)beneficiariesinvestedsomeoftheircash

    intotheproductiveactivitiessuchasmicroenterprisesandagriculture.Theyreported30to

    50%returnsontheseassets.Whilethesereturnsarehigh,thenumberofinvestorsinvolved

    (12%)issmallifcreditconstraintsforproductiveinvestmentshadbeenhigh,thensurely

    morewouldhavetakenadvantage.InEthiopiasPSNP,whereassetbuildingbytheseruralhouseholdswasacentralobjective,andassetpovertytrapshadbeenarguedtoexistasa

    motivationfortheprogramme,therewassomeimpactonassetaccumulation,butwellbelow

    expectations.Gilliganetal.(2009)reportedthatlivestockincreasedamongstthebeneficiaries

    byabout9%,butformost,assetlevelsremainedwellbelowthelevelsinitiallysetfor

    graduationacrossthepovertythreshold.Initialtargetshadsuggestedthatallbeneficiaries

    weremeanttograduateby2009throughthevirtuousgrowthprocessesofincomesandassets

    linkedtoaccessofthevariouscomponentsoftheprogramme.Butofthemorethan8million

    households,only1.3%hadgraduatedby2009.14

    Inshort,theseefficiencygainsviatransfersappeartoberelativelysmallinsomeofthelarger

    programmes.Maybetransfersarenotthebestwayofovercomingthis?Infact,afarmore

    commoninterventiontoovercomecreditmarketconstraintsinmanycountriesismicrocredit,

    andatleastintermsofuptake,thatwouldseemaverysuccessfulactivity.Wouldthatoffera

    betteralternativetogenerateefficiencygainsthanoffercashtransfers?

    Infact,therelativelylowreturnstocashtransfersformosthouseholdsareratherconsistent

    withtheemergingevidenceonmicrocredit.Foralongtime,ithasbeenhardtofindclearand

    undisputedevidencethatmicrocreditoffersthetransformationoflivespromised(Armanderiz

    deAghionandMorduch,2005).Morecarefulevaluationstudieshavebeenimplementedin

    recentyears,andgenerally,theyarenotfindinglargeeffectsonpoverty,assetholdingsand

    ambiguouswelfareeffects.AnumberofrecentRCTsinurbanareashavefocusedonwhether

    microcreditleadstohighreturnfrombusinessinvestmentinIndiaandthePhilippines,

    targetingareaswherehighreturnsshouldhavebeenexpected.Theyhaverecentlyfoundthat

    theaveragereturnwaszeroforfemalesreceivingloans(Banerjeeetal2010,KarlanandZinman

    14Thisdoesnotbodewellforcurrentplans,wherewithinthenextfiveyear,beneficiarieswillbereducedtobelow

    1.5million,meaningthattherestisexpectedtograduate.

  • 7/30/2019 Social Protection Growth 2011-17

    13/29

    13

    2010).MirroringthesestudiesbutnowofferinggrantstowomeninSriLankaandGhana,itwas

    foundinSriLankathatthereturnwasalsozero,andinGhanathatonlyforthosewitha

    businessinthetop40%ofthedistributionofinitialfirmsize,didthecashtransferhaveapositivereturn(deMeletal2008,Fafchampsetal.2011).InIndia,Banerjeeetal.(2010)find

    thatthosealreadyinvolvedinentrepreneurshipinIndiausetheloantoexpandbusiness,

    withoutnecessarilyguaranteeingareturn.Overall,Banerjeeetal.(2010)concludethat

    microcreditisnotforeveryhousehold,orevenmosthouseholds,,anditdoesnotleadto

    themiraculoussocialtransformationsomeproponentshaveclaimed.Thesameappearstobe

    trueforsocialtransfersaspartofsocialprotectionspending:formany,thereturnstothese

    grantsareunlikelytobehigh,andtheirtransformativepowerisnotobvious.

    However,itcanbecorrectlyarguedthatthenonmonetaryreturnstothesecashtransfers(orindeedmicrocredit)maybeasimportant.Justasmicrocreditsmainresultappearstobe

    impactsonfemaleempowermentandnonmonetaryeffects(ArmanderizdeAghionand

    Murdoch,2005),thesameseemstoapplytosocialtransfers.Transfersmaybeinvestedin

    otherassets,suchashealthoreducationofchildren.Itmaymoregenerallyleadtosocial

    emancipation.Thereisindeedevidenceofsomesucheffects.

    Conditionalcashtransferstendtoimposeconditionsintermsofschoolattendanceorhealth

    clinicattendanceofchildrenonparentsaccesstocashtransfers.Whilemanyofthese

    schemes,suchastheProgresa(Oportunidades)havefoundpositiveimpactsonhealthandeducation,itwouldnotbequiterighttodirectlyattributethesetothecashpartofthetransfer.

    However,exploitingvariationinthecashtransfergiven,Gertleretal,(2009)havearguedthat

    thecashpartwasrelevantforimprovedhealthoutcomes.Thesocial(oldage)pensionscheme

    inSouthAfricaalsoappearstohavefurthereffectsbeyondcashreturns:Duflo(2003)finds

    positivehealthandnutritioneffectsonyoungchildrenofthefamily,whileEdmonds(2006)

    reportsreducedchildlaboursupply.15SomeoftheschemesinLatinAmerica,suchasJuntosin

    Peru,canbecreditedwithofferingmarginalizedgroupssuchasAndeanpeasantssomething

    concretefromthestate,ratherthanexperiencingitsoppressionandviolence.

    Finally,ifmicrocreditandcashtransfershavesimilarreturns,whynotpromotemicrocredit,as

    itsurelywillbecheaperifsomeresourceswerebeingrepaidwithlittlelossofthebenefits?Itis

    notsosimple.Onekeyargumentformoresocialassistanceisthatmicrocreditprogrammesfind

    itnotoriouslydifficulttoreachthepoorest(ArmanderizdeAghionandMurdoch,2005).And

    creatingobligationstoberepaid,ifreturnstoassetsarenothighformany,wouldmakecredit

    15Socialprotectionishowevernotnecessarilyleadingtobeneficialchildlaboureffects:Woldehanna(2009)finds

    thatchildlabouractuallyincreasedamongbeneficiariesofthePSNP,asparentsleavechildrentomindthe

    householdwhiletheytakeontheworkfare.

  • 7/30/2019 Social Protection Growth 2011-17

    14/29

    14

    notasuitablesolutionforthepoorest.Nevertheless,thesuccessofmicrocreditprogrammesof

    reachingmillionsacrosstheworldillustratesapossiblelessonforsocialassistance

    programmes.Microcreditinvolvesaclearcontractbetweenproviderandborrower,withclearlyspelledoutrightsandobligations.Socialprotectionoftenlacksthiscontractualaspect.

    Contingenttransfersandinsurance

    Thediscussionthusfarhasfocusedonsocialassistancetransfers;butwhatabouttransfers

    contingentonparticularevents,suchasharvestlossesorillness?Mostpoorliveinrisky

    environmentsthatconstantlythreatentheirlivelihoods.Themajorityofthepoorareself

    employed,inagricultureorintheinformalsector;othersareininsecurecasualwagelabour

    jobs.Healthrisksarerifeinurbanandruralareas;rurallivelihoodsstronglydependonagro

    climaticconditions.Mutualsupportandrisksharingmechanismsarewidespread,whilehouseholdsalsouseavarietyofselfinsuranceandriskmanagementstrategies.However,

    extensiveresearchinthelast20yearsorsosuggeststhatatbestallthesesophisticated

    livelihoodstrategiesofferonlypartialinsurancetothepoor(Dercon,2004).Typically,they

    cannotoffersufficientprotectionagainsteconomicdownturns,climateshocksandserious

    healthshocks.

    Fundamentally,thisislinkedtomissinginsurancemarkets,orfurthermarketfailuresinsavings

    andcreditmarkets,thetypicalmarketbasedsubstitutesforlackofinsurance.Welldesigned

    socialprotectionofferingcontingenttransferscouldfillthisgap.Examplesareworkfare

    programmestohandleeconomicdownturns(suchasJefesyJefasinArgentina)orthe

    employmentguaranteescheme(NREGA)inIndia,offeringupto100daysatrelativelylowpay

    toanyone.Freeorsubsidizedhealthcaretospecificgroupswouldbeameanstooffera

    contingenttransferinkind;morestandardsocialinsuranceforhealthcouldtaketheformof

    publiclyrunorsupportedhealthinsuranceschemes,suchasthestronglysubsidizedRSBY

    hospitalizationinsuranceinIndia.Welldesigneddroughtreliefandotherfoodaidschemes

    couldalsohavetheirrole,forexampleinthefaceofweatherorpriceshocks.

    Socialprotectionmayresolvesomeofthemostpertinentinsurancemarketfailures.However,

    whilenodoubtmuchhumanhardshipcouldberesolved,wouldthisofferasufficientlystrong

    economicargumenttospendresourcesonit?Thiswoulddependonthesizeofthelosses.

    Therearetwotypesoflosseslinkedtoriskworthdistinguishing.First,thereisthelossof

    variousassetsandhumancapital.Secondly,therearetheprofitableopportunitiesmisseddue

    toriskashouseholdsandfirmsneedtoavertriskduetolackofprotectionagainstdownside

    risk.

  • 7/30/2019 Social Protection Growth 2011-17

    15/29

    15

    Largeorrepeatedshocks,withoutgoodinsuranceorsocialprotection,forcepeopletoselloff

    assetstosustainnutritionandconsumptionwithrisksofpersistentpoverty:astateofdeep

    povertywithlittlehopeoropportunityofescape.Thenarrativeofassetpovertytrapsinspiredbytheanalyticsinfigure1appearsconsistentwithmanyofthelifehistoriesofthosein

    persistentpoor:theyfacedepletedproductiveassets,withsuchlowreturnsthattheyarelikely

    toremainstuckindeeppovertyforever.Inahighriskenvironment,onecouldeasilyfallwhen

    aseriousshockstrikes,butthenfinditveryhardtoescape.Iftheselossesarelargeand

    widespread,theefficiencylossesmaybeveryhigh.

    Thereisevidenceforsuchprocessesoperatinginsomeofthepoorestsettingsintheworld,in

    theformofassetorincomepovertytrapsinKenyaandIndia(Lybbertetal.,(2006),Derconand

    Outes(2011))andincomedynamicsleadingtopovertypersistenceinChinaandEthiopia(JalanandRavallion(2004),Dercon(2004)).Thesizesofthesedynamiceffectsfortheaffectedgroups

    canbelarge,andattimesmanyareaffected:forexample,alargepartoftheruralpopulation

    wasaffectedbythefamineof198485inEthiopia,andinthe1990s,itwasfoundthatduringa

    periodofgrowthandrecovery,theyhadabout6percentlowergrowthinincomeperyearthan

    thosenotaffected(Dercon2006).InTanzaniaandEthiopia,incomesofthoseaffectedby

    droughtsmorethan10yearsearlierwerestillfoundtobelowercomparedtounaffected

    communities(Beegleetal.2008,Dercon,2006).

    Thesizeoftheseeffectsisevenbetterillustratedbytheimpactoflargedroughtandconflictshocksonhumancapital,intheformofhealthandeducation.Catastrophichealthshockswere

    foundtohavelargeeffectsonconsumptioninIndonesia(GertlerandGruber,2002).More

    dramaticconsequenceswerefoundfromcrisesaffectingyoungchildren.Evidencefrom

    ZimbabweandEthiopiafromcrisesinthe1980sfoundthatupto20yearlater,youngchildren

    affectedbythesecriseswerefoundtohavepermanentlossesintermsofheightandeducation,

    suggestingpermanentincomelossesofbetween7and15%(AldermanandHoddinott,2004;

    DerconandPorter(2010)).MaternalorphanhoodinTanzaniawasshowntoleadtosubstantial

    heightandeducationlosses,andabout9%lowerearningsatadulthood(Beegleetal.2010).In

    short,aseriouscrisisatanearlyagehaslifelongimpactsonproductivityandincomes.

    Someofthekeyimpactshereworkthroughnutritionaldeprivationrelativelyearlyinlife.The

    crisesdiscussedinmostofthesepapersarelargelycovariate,inthatalargepercentageofthe

    populationisaffected.Buttheseeffectsareunlikelytocomeaboutonlyfromcovariateshocks.

    Stuntingisstillveryhighmorethan40%ofchildreninsubSaharanAfricaandSouthAsiaare

    stunted,linkedtodeprivationsinearlychildhood.Causallinksbetweenstuntingandcognitive

    development,andthereforeearningspossibilitiesarewellestablished(GranthamMcGregoret

    al.,2007).

  • 7/30/2019 Social Protection Growth 2011-17

    16/29

    16

    Inshort,theavailableevidencesuggestslongtermefficiencylossesfromtwokeysources:at

    thelevelofhouseholds,largeuninsuredshocksinincomestendtohavelongtermincomeconsequences,linkedtolossesofassets,whileespeciallychildrenbearanimportantcostinthe

    longrun,intermsofeducationandhealth,affectingtheirproductivityinlaterlife.

    Hardertoquantifybutpossiblysubstantiveaswellarethecostsofforegoneopportunitiesdue

    torisk.ThereisevidenceofactivityandassetportfoliochoicesinagricultureinIndiaand

    Tanzania,suggestingthatreturnstoassetsareabout25to50%lowerforthosenotableto

    protectthemselvesagainsttheconsequencesofbadharvests(BinswangerandRosenzweig,

    1993;Dercon,1996).Thereisalsoevidenceofloweruptakeinagricultureofprofitableinputs

    andtechnologiesinIndiaandEthiopia,linkedtothelackofprotection(DerconandChristiaensen,2011;Morduch1995).Dercon(2006)hasmoreexamples.

    Thiswouldsuggestaratherstrongcaseforsocialprotectionmechanismsthathelptoavoid

    consequencesofclimaticriskorserioushealthshocks.Fewofthelargesocialassistance

    schemesinoperationhaveasystemofcontingenttransfersincluded,forexamplewhere

    transfersincreaseorcoverageisbroadenedwhenincomeshocks,suchaslinkedtopoor

    harvestsoccur.DeJanvryetal.(2006)neverthelesssuggestthatProgresahelpedchildrento

    stayinschoolwhenharvestsfailed,althoughtheyalsoarguethatamorestatecontingent

    systemoftransferswouldmakethesystemmoreeffective.

    Inanumberofcountries,effortsareunderwaytoexpandhealthinsurancesystemsfor

    hospitalizationcosts.InGhana,theNationalHealthInsuranceschemeistryingtooffermore

    widelyaccesstohealthinsurancetoreduceoutofpocketexpenditures,althoughevidence

    suggeststhatthepoorestremainexcluded(EuropeanReportonDevelopment2010).InIndia,

    theRSBYlowcosthospitalizationinsuranceschemeoffersanambitiouslargescaleattemptto

    ensurethepooravoidlargehealthexpendituresandprotecttheirhealth.Whileenrolment

    appearstobehighandrathersuccessful(WorldBank,2011),earlyevidenceonitsfunctioning

    israthermoredamningaboutitssuccess.Rajasekharetal.(2011)reporthowinKarnataka,85%

    oftheeligiblehouseholdsknewaboutthesystem,and68%hadbeenenrolled.Butinpractice,

    aftersixmonths,utilizationrateswerevirtuallyzero,withbeneficiariesnotreceivingcardsor

    informationwheretogetcare,andwithenrolledhospitalsnothonouringtheircommitments,

    askingforcashorturningawaypatientswithRSBYcards.

    TheNREGAemploymentguaranteeschemeinIndiawouldinprincipleofferoneidealformof

    contingenttransfers,byofferingpermanentlyaccesstoupto100daysofworkatwagesthat

    makeitonlyofinteresttoeitherthepoorestorthosetemporarilyfacinglowincomes,asit

  • 7/30/2019 Social Protection Growth 2011-17

    17/29

    17

    wouldguaranteeasourceofincomewhenruralincomesarelowduetopoorharvestsand

    limitedrurallabourmarketopportunities,whilealsocontributingtobuildupcommunityassets

    (Ravallion,2008).Itsimpactisnotentirelyclearatthemoment,andtheprogrammeisrifewithimplementationproblems.Forexample,DrezeandKhera(2009)findintheirsurveythat98%of

    thepotentialworkersstatedthattheywerereadytoworkfor100daysintheyear,whichisthe

    upperlimitoftheNREGA,butonly13%reportedtobeabletogetthisnumberofdays(seealso

    WorldBank(2011)).

    Ifuninsuredriskiscostly,andfewprogrammesappeartofunctionwell,couldinsurancemarket

    developmentnotofferanalternative?Inmostricheconomies,publicorsubsidisedsolutions

    forhealth,unemploymentanddroughtinsurancearestandard,butthisdoesnotmeanthatin

    settingswithlimitedresources,theremaynotbeefficiencygainsfromusingmoremarketbasedsolutions.Indeed,alotofeffortisputintotryingtodevelopmicroinsurancesolutions

    forthepoor,usingexactlythesamemotivationasusedforthedevelopmentofsocial

    protection(e.g.seeDerconandKirchberger,2009).Itisdefinitelyworthwhiletoexplorewhat

    theplaceofeitherwouldbeinanefficientsystemgearedtowardsreducingvulnerabilityofthe

    poor.

    Experiencewithdevelopingmicroinsurancemarketshastaughtusalotoflessons.Inpractice,

    whiletherearealotofmicroinsuranceinitiatives,coverageisstillverylow.Forexample,Matul

    etal.(2010)estimatedthatonly2.6%ofthepopulationunder$2adayinsubSaharanAfrica

    hadaninsurancepolicyfrominsurancecompanies,NGOsandcommunitybasedsystems,and

    morethanhalfofthesewereinSouthAfrica,wherefuneralandlifeinsuranceiswidespread.

    Morecarefulimpactevaluationsofdroughtorhealthinsuranceproductsalsofindlowuptake

    (Coleetal.2010,Derconetal.2011).Thesemarketsareclearlyhardtodevelop,forvarious

    reasons.First,insuranceisadifficultconceptandthereforeadifficultproducttosell

    anywhereintheworld(Coleetal,2010).Furthermore,trustisakeyissue(Caietal,2010,

    Derconetal.2011).Notethekeydifferencewithsellingmicrocreditproducts:inthecaseof

    insurance,thesellerfirstcollectsmoneyandthecustomerhastotrustthatthesellerwill

    honourthecontractifabadeventoccurs;incredit,thesellergivesmoneyandhastofinda

    waytoensurethecustomerrepays.Partingwithmoneyasinthecaseofinsurancewillbeseen

    ashighlyriskybythepoor,addingtotheirvulnerability.

    Thesewouldseemreasonsforinvolvementofthestate,especiallytoensurethatsomeofthe

    poorestarecoveredaswell;contributionbasedsocialinsuranceschemesclearlycanhavetheir

    role,withpossiblysubsidisedaccessforsomeofthepoorest.Butifunderstandingandtrust

    areproblemsforprivatecontracts,thesewouldsimilarlybeproblemsforstaterun

  • 7/30/2019 Social Protection Growth 2011-17

    18/29

    18

    programmesiftrustingovernanceofpublicschemesisnotpresentwhichcouldbeaproblem

    inmanysettings.

    Therecouldbeotherreasonswhysocialprotection,possiblyworkingwithprivateinsurance

    systems,hasaclearspace.Insuranceproductsforsomeofthemostcatastrophicand/or

    covariateshockstendtoberatherexpensive,withcostlyreinsuranceforprivateinsurance

    contracts.Renewalsbycustomersarealsoattimeshardtoachieveifnoshocksoccurredinthe

    lastyear.Finally,privateinsurancemarketshaveincentivestoguardthemselvesagainst

    adverseselection,bykeepingsomeofthehighestrisksoutoftheriskpool,leadingtoexclusion

    ofsomeofthepoorest.Involvementofthestate,includingviaguaranteesandsubsidiescould

    ensurethepoorestarecovered,evenforthelargestrisks.Morestandardcontributionsbased

    socialinsurancecouldthenbefurtherdevelopedfromspecificpublicsocialprotectionactivities.

    EfficiencyinDeliveryofSocialProtection

    Withlimitedbudgetsandpotentiallyhighcosts,findingwaysofdeliveringsocialprotectionat

    lowcostmustalwaysbeapriority.Manyoftheseissuesrelatetodesigningmanagerialand

    governancesystems.However,therearecertainelementsinthedesignoftheproducts

    involvedthatprovidepotentialefficiencygains.

    Iftargetingcanbedonemoreefficiently,reachingthetargetpopulationwillbemorecost

    effective.Imposingstricttargetingruleswouldmeanthatnoonenotdeservingthetransfer

    receivesit.Ingeneral,targetingusingmeanstestingoradministrativemeanstendstobe

    expensiveorsystemstendtomakeerrors.Theseerrorsareoftwotypes:includingthosein

    schemesnotdeservingthetransferandexcludingthosethatshouldgetthetransfer.Ifone

    wantstomakesurethatpovertyreductionisparamount,thenexcludingdeservinggroups

    wouldcomeatahighpriceinsocialwelfare.Universaltransfers(insteadoftargetedtransfers)

    havethereforebeenproposed,butobviouslywouldcomeatahighcost.

    Onemeanstoexcludeanynontargeted(nonpoor)householdsistosetupselftargeting

    schemes:socialprotectionprogrammesdesignedinsuchawaythatonlythosereallyneedingit

    enrolortakeadvantageoftheprogramme(Ravallion,2006).Imposingconditionstoreceivea

    transfersuchasinconditionalcashtransferschemescanencourageselftargetingifthe

    conditionwouldbeconsidereddemanding,e.g.intime,tastesorevenshameforrichergroups.

    AworkfareprogrammesuchastheemploymentguaranteeschemeinIndiaisoneexample,as

    thebenefitrequireshardwork.Usinginferiorcommoditiesforpricesubsidies,whose

    consumptiondecreaseswhenincomerisesisanotherexample.Itisnotalwaysclearthatthis

  • 7/30/2019 Social Protection Growth 2011-17

    19/29

    19

    willresultinbettertargeting:oneproblemisthatnotallthepoorcansatisfythecriterion.For

    example,notallcanoffertowork,comingatthecostofexcludingsomegroups.

    Whilebettertargetingoffersameanstodeliverschemesatlowercosts,thisisnotthesameas

    providingefficiencygainstotheeconomyfromsocialprotection.Asthediscussioninthis

    sectionhasargued,cashtransferstomostofthepoorasoperatingcurrentlyarenotlikelyto

    havelargereturnforgrowthandefficiencyperse,althoughcontingenttransfersmayhave

    largerefficiencygains.Conditionalcashtransfersmayoffersomefurtherbenefits,astheynot

    onlytransfercashortypicallyfungiblecommodities,butalsoimposeratherpreciseconditions.

    Conditioningtransferscanhavebenefitsforefficiencyiftheconditionimposedoffersdirect

    efficiencygains.Forexample,onewayofreadingthebenefitsofconditionalcashtransfersasin

    operationinmanyLatinAmericansettingsisthattheyresolveinefficienciesinhouseholdallocation:childrensbargainingpowerisinsufficientandtheirintertemporalbenefitsfrom

    betterhealthandeducationareinsufficientlyvalued.Conditionsensurethatmoreofthegains

    accruetothechildren(Dasetal.2005).Furthermore,therecouldbedirectbenefitsforthe

    economyaswell.Ifinvestinginhealth,educationortraininghasexternalitiesfortheeconomy

    asawhole,provide(private)transferstohouseholdswithconditionsforhealthandeducation

    (asinOportunidades),orfortraining(JefesyJefas,Argentina)ensuremoreinvestmentthan

    whattheunconditionaluseofthetransferhadoffered.Thesizeoftheexternalitiesremainas

    disputedastheexactroleofhumancapitalforgrowth,whichespeciallyatlowerskilllevelsis

    notresolved(GrowthCommission,2008).

    Othergrowthexternalities

    Therearealsopossiblyotherexternalitiesfromredistribution,thatmaybeimportantfor

    growthbutworkmuchmoreindirectlyordynamically.Thistouchesonthemoremacrolevel

    literatureontheroleinequalityongrowth,viapoliticaleconomyeffects.Thereisrelatively

    strongevidencethatinequalityisbadforgrowth(WorldBank,2006).Theorymodelshave

    suggestedvariousmechanismsincludingviapressuresofredistributionindemocraticsocieties

    ortheimpactofsocialconflictallleadingtolowergrowth(PerssonandTabellini1994;Alesina

    andPerotti1996).Importantly,whileitislessdisputedatpresentthatseriousinequalityleads

    tolowergrowth,inmodelssuchasinPerssonandTabellini(1994)itisbecauseredistribution

    leadstolowergrowth.TheevidenceinBanerjeeandDuflo(2000),thatchangesininequalityin

    eitherdirectionisassociatedwithreductionsingrowth,mayfurtherlimittheroleoflargescale

    socialtransfersasofferingadirectrouteforinequality,evenifinequalityinitselfisbadfor

    growth.Ofcourse,itwouldbehardtoarguethatcurrentsocialprotectionschemesinmostof

    thepoorerdevelopingcountrieshavethescalerequiredtoattractanylargenegativedynamic

    growtheffects,whilethecontributiontoreducingsocialconflicttensionsinhighinequality

    societiessuchasinLatinAmericamaywellbelarge.Aswithallregressionwork,the

  • 7/30/2019 Social Protection Growth 2011-17

    20/29

    20

    associationsimpliedareaveragesacrossaworldwithrelativelyfewcountriesofferingdata

    pointsandhugeheterogeneityinsocialandpoliticalcontexts.Inthisreview,wehavelessto

    sayondynamicgrowtheffectsviatheimpactofsocialprotectiononinequalitybylackofevidenceeitherway.

    4. SocialProtectionandgrowth?

    Thediscussionabovesuggeststwobroadtypesofsocialprotectionthatcanhavesubstantial

    efficiencyorgrowthbenefits.First,giventheproblemswithriskandinsurancemarkets,

    developingsocialprotectionspecificallytodealwithshockshasastrongappeal,notleastgiven

    thelongtermcoststhelackofinsuranceappearstoimposeonparticulargroups.Secondly,

    whilebroadbasedcashtransfersareunlikelytohaveahighreturnbeyondtheir(important)

    redistributiverole,theideathateithertargetingspecificgroupsorincludingspecific(non

    fungible)conditionstosocialprotectionseemsappealingtoachieveefficiencyandgrowth

    goals.Alltargetingrisksexcludingcertaingroupsasaconsequence,soatradeoffbetween

    efficiencyandequityconsiderationsloomslargeinthiscase.Belowwesuggesthowtransfers

    tothreespecificgroupsthatcouldhaverelativelyhighefficiencygains,aswellasthepossible

    typesoftransfersthatcoulddeliverthesebenefits.

    TargetingYoungChildren

    Thecasefortargetingyoungchildreniswellrecognised.Theevidencequotedaboveonthe

    longtermdamagethatcanbedonefromnutritionalandotherdeficienciesinearlychildhood

    arewelldocumented.Infact,eventheGrowthCommission(2008)recognisedtheneedfor

    spendingonearlychildhoodprotectivemeasuresforgrowthreasons.Cashtransfersperseare

    bluntinstrumentstoensurethatspendingbyfamiliesreachesthechildrenproperly.

    Furthermore,themostcriticalagesforchildrenarepreschool,soschoolbasedsystemswould

    notworkeither.Acombinationofconditionalcashtransfersaswellasformsofcontingent

    transfersduringcrisestargetedatchildrenmayofferasolution,eventhoughtheyarelikelyto

    beratherbluntinstrumentstoensurethatnolongtermlossesemergefromshortterm

    deprivation.

    Someoftheevidenceavailablesuggeststhatstandardconditionalcashtransfersareindeedtoo

    blunttohaveaconsiderableimpactonearlychildhoodindicators.Eventhoughsomeimpacts

    werefound,Schady(2006)argues,usingevidencefromGertlerandFernald(2004)(seealso

    Fernaldetal.2009)thatProgresa/Oportunidadeswastooblunttohavemuchimpact.Targeted

    foodsupplementstochildren,suchasaprogrammeinPeruofferingmilktochildrenalsomake

    littleinroadsinnutritionaldeprivationandotheroutcomes(StifelandAlderman,2003).

    Strongerimpactsarefoundfromspecialprogrammesencouragingattendanceatdaycareor

  • 7/30/2019 Social Protection Growth 2011-17

    21/29

    21

    preschoolfacilitiessuchasinthePIDIprogrammeinBoliva(Behrmanetal.2004).Themost

    encouragingresultscomefromtheworkinJamaicabyMcGregorandcolleagues(1991,1997),

    inwhichnutritionalsupplementsandchildstimulationwereoffered,withespeciallythelatterofferingveryhighreturnsinthemediumrun.Howtoscaleupsuchinterventionsinpoor

    settingssuchasinAfricaisobviouslyaseriouschallenge.

    MigrationandCities

    Duringperiodsofgrowth,livelihoodsrarelychangesmoothly.Formany,itinvolvestakingrisks,

    includingmigratingandenteringintoactivitiespreviouslynotperformed.Suchchangesare

    essentialtoallowthepoortotakepartandbenefitfromeconomictransformationbutasthe

    fastgrowingeconomiesinAsiaandLatinAmericahaveshown,whileimprovingmanylives,

    migrationtendstoinvolveserioushardshipforsome,eveniftemporarily,andevenleadingtopersistentpoverty.Thiswillmakeothersreluctantparticipants,slowingpovertyreduction

    duringgrowth.Atthesametime,theeconomictransformationrequiresareadysupplyof

    workers,anddespiteperceptionstothecontrary,migrationinmanydevelopingcountriesis

    oftentooslowleadingtoinefficiencies,andthelackoftheappropriateskillmixforlocallabour

    markets.AuandHenderson,(2006)arguedthatlimitationsonmigrationinChinaleadtocities

    thataretoosmallforgrowthandefficiency.Beegleetal.(2011)findurbanruralgapsinthe

    standardoflivinginTanzaniaconsistentwithundermigration;similarevidenceexistsforIndia.

    Internalmigrantsareoftenperceivedtoliveinitiallyinsqualorsufferingpoverty.Intermsof

    averagemateriallivingconditions,theevidencesuggeststhecontrary,atleastcomparedto

    conditionsleftbehind;however,therisksofmigration,leadingtoheterogeneityinliving

    conditions,aswellasthecostsofcongestion,andpoorpublicservicesintheformofaccessto

    health,sanitationandeducationreducetheattractivenessofmigration.Astherewouldbe

    efficiencygains,theeconomiccaseforofferingmigrantsandotherurbangroupsbettersocial

    protectiontolowertherisksinmigrationmaybehigh,evenifcontroversial.Mostworkfare

    programmesinthepoorestcountriesareruralbased,atbestbuildingupruralinfrastructure,

    ofteninareaswithlimitedfutureforagricultureoremploymentinthelongrun.Introducing

    workfareprogrammesinurbanareas,bothinsmallandlargercities,focusingonbuildingup

    communityassets,suchassanitationorbetterhousing,couldhavesubstantialbenefits,andcontributetomakingurbanareasmoreattractiveforworkers,contributingtoamoreready

    supplyofworkerstofeedintotherequiredeconomictransformation.Atpresentmanyofthe

    ruralcommunityassetbuildingprojectssuchasinthePSNPorNREGAmaynothaveproduced

    highqualityassets;itwouldbeachallengetoensurethiswouldhappeninurbanworkfare

    programmes.

  • 7/30/2019 Social Protection Growth 2011-17

    22/29

    22

    Adolescentsandyoungadults

    Anothergroup,largelyforgotteninsocialprotection,butbecominganeverlargerchallengefor

    socialandeconomicpolicyarerelativelypooradolescentsandyoungadults.Themassiveriseof

    enrolmentintoeducationandtheeconomictransformationtakingplacehasraisedthe

    expectationsofthisgroup.Giventhewaydemographictrendshaveevolved,inmanycountries,

    theyarecurrentlyorinthenextdecadeorso,thelargestsingledemographicgroup.Inmany

    developingcountries,educationhasnotdeliveredguaranteedjobsinthesamewayasitwould

    havebeenfortheirparents.Atthesametime,concernsaboutthequalityofeducationinmany

    developingcountriesmeansthateventheircognitiveskillsarelimited;labourmarketsdooften

    notappeartoconsiderthemeasilyemployable.

    Aseconomictransformation,andtheinclusivenessofgrowthdependsontheextenttowhich

    labourmarketscanabsorbtheserelativelylowskilledworkers,theywoulddeservespecific

    attentiononefficiencygroundsfromsocialprotectionsystems.Designingconditionalcash

    transfersconditionalonadditionaltrainingandlabourmarketpreparation,orapprenticeships,

    maywellhavehighlongtermreturns,andhelpingtobridgetheabsorptionofthisgroupinthe

    labourmarket.Itwouldinanycasebepreferabletousingsocialprotectionfundsforlargescale

    selfemploymentorientedmicrobusinesstrainingwithcredit,afavouriteintervention:asthe

    evidenceabovehassuggestedthisislikelytobebeneficialtoonlyalimitedandwelltargeted

    groupofpotentialentrepreneurs.Preparinganarmyofworkersfortheinformalsectorwould

    inthelongrunnotbetheappropriaterecipe.Asthehistoryoftheeconomictransformationin

    richeconomieshasshown,relativelyfewarecalledtobesuccessfulinselfemployment,and

    mostwillendupwageworkersandtogetherthisformsasuccessfulformulaforpoverty

    reduction.

    5. ConclusionEconomicgrowthandsustainedjobcreationareessentialforlargescalepovertyreductionin

    thepoorerregionsoftheworld.Higherincomeswillmeanthatincreasinglyfewerpeoplewill

    bestuckinpovertybutmorewillbeabletowithstandshocks.Therefore,doappealstoincreasesocialprotectiontoreducepovertynotsimplyshowthatthisprocessisnottakingplace?There

    isdefinitelysometruthinthisargument.Manyofthepoorarestuckinpovertybecausethey

    havelimitedopportunities.Muchoftheirvulnerabilitytoshocksandpersistentpovertyis

    closelylinkedtotheirlivelihoodopportunities:workingonfarmsinhighlyriskyagricultureor

    beingselfemployedinasmallbusinessinariskymarketenvironment.

    Povertyreductionthroughouttheworldischaracterisedbytheabsorptionoflargepartsofthe

    labourforceinstablewagejobs.Higherincomesforthoseremaininginselfemploymentand

    agriculturewouldmeanopportunitiestobuildupassetsorothermeanstowithstandshocks

  • 7/30/2019 Social Protection Growth 2011-17

    23/29

    23

    andmisfortune.Formany,thethreatofpersistentpovertyandassetpovertytrapswouldbe

    unravelled.Whilesomeparticularlyvulnerablegroupswillalwaysrequireformsofsocial

    protection,thescaleofthetaskwouldshrinkwithsustainedgrowth.Socialprotectionisnosubstituteforthisprocess.

    Nevertheless,socialprotectionhasanimportantroleinthisdevelopmentagenda.Itisone

    mechanismformakinggrowthpropoor.Itoffersadirectandsimplemeansofredistributing

    someofthegainsfromgrowthincludingtothosenotabletoproductivelycontributetothe

    economysuchastheelderlyordisabledormoreingeneral,thosethatriskstaybehind.It

    alsoofferswaystoensurethatshocksdontpushpeopleback.

    Welldesignedsocialprotectionmaycontributetogrowth.Socialtransfersandothersocial

    assistancecanoffertheproductiveassetsthepoorneedtoengageproductivelyinthe

    economy,andallowthemtograduatefromdependence.Publicworksprogrammescanalso

    buildrelevantpublicgoodsandinfrastructureinlocalcommunities,contributingtogrowth.

    Welldesignedsocialinsurancecanpluggapsinprivateinsurancemarketsandcomplement

    communitybasedsystems.Byovercomingmarketfailures,itcancontributetoefficiency,

    allowinghouseholdstousetheirresourcesmoreeffectively,andencouragetherisktakingand

    innovationessentialforgrowth(Ravallion,2006;Dercon,2005).Butdontexpectsocial

    protectiontobeadriverofgrowth.Itwouldbenaivetoexpectthatthemacroeconomiceffects

    viathedemandincreasesfromcashtransfers,withoutproductionandefficiencyincreases,

    wouldbeanythingelsebutinflationary.Andtheevidenceonthecontributiontoefficiencyand

    growthofsocialprotectionisjustnotthatstrong.

    Socialprotectionisnotsimplythepanaceatomakegrowthinclusive.Itcouldaddressmarket

    failuresbutitisnotnecessarilyoralwaysthebestinstrumenttodoso.Itwillsupportequity

    objectives,andtheyremainparamount.Tradeoffsarethebreadandbutterofwelfare

    economics,andexposestheneedforotherstomakethepoliticalchoiceswheresimple

    economiclogicdoesnotofferanexactsolution.Expandingsocialprotectionwillremaina

    matterofsocialtastes,butitdoesnotmeanthatitcannotbedone,northatitnecessarilywill

    hinderefficiencyandgrowthprocesses.Whileitisnotasimpleawinwinsituation,choosing

    bothdesignsandthefocalareascarefullywillhelptomaximizetheirreturntogrowthandpovertyreduction.

    Butthereareareaswhereitcanhavestrongeffectsandatleastprovidetheoiltomakethe

    growthengineworkmoresmoothly.Weidentifiedthreeareas.First,bothsocialinsuranceand

    socialassistancecanprotectfamilyinvestmentsinhumancapital,suchaseducationandhealth,

    byensuringthatchildrenstayinschoolorthatnutritiondoesnotsufferwhenafinancialshock

    hits.Becausetheselossesinnutritionandeducationareoftenirreversible,theyimplythat

    earliersocialinvestmentsarewasted,whichcouldhavebeenavoidedwithappropriatesocial

  • 7/30/2019 Social Protection Growth 2011-17

    24/29

    24

    protection.Thisisespeciallysointhecaseofveryyoungchildrenandseriouscrisesany

    nutritionallosshaslargeandirreversibleimpactsontheirabilitytocontributeproductivelyin

    adultlife.Secondly,socialprotectioncouldbetargetedtohelptheeconomictransformationbecomesufficientlylabourintensive.Areasofimportancewouldbesupporttomigrantsand

    migration,andanythingthatcanhelpcitiesgrowinaninclusiveway.Thirdly,andrelated,use

    socialprotectiontomakeadolescentsmoreemployable,andofinteresttofirmstoabsorb

    themintotheeconomy.Inallthesecases,somesmartmechanismswouldberequired,

    probablywellbeyondtheratherbluntstandardsocialprotectionmeasuresandcannotbedone

    inisolationfromotherpoliciesrelatedtohealth,educationandemploymentskillsformation.

    References

    Alderman,H.,J.HoddinottandB.Kinsey,2006."Longtermconsequencesofearlychildhood

    malnutrition,"OxfordEconomicPapers,OxfordUniversityPress,vol.58(3):450474.

    Alesina,A.andR.Perotti.1996.IncomeDistribution,PoliticalInstability,andInvestment,

    EuropeanEconomicReview,40(6):120328.

    ArmanderizdeAghion,B.andJ.Murdoch2005.TheEconomicsofMicrofinance.MITPress.

    Au,CCandJ.V.Henderson.2006b.AreChinesecitiestoosmall?ReviewofEconomicStudies

    73(3):549576

    Banerjee,A.,E.Duflo,R.GlennersterandC.Kinnan.2010.Themiracleofmicrofinance?

    Evidencefromarandomizedevaluation.BREADWorkingPaperno278.

    Banerjee,A.,E.Duflo,R.ChattopadhyayandJ.Shapiro.2010bTargetingtheHardCorePoor:An

    ImpactAssessment,MITmimeo.

    Banerjee,A.andE.Duflo,2005.GrowthTheorythoughttheLensofEconomicDevelopment.In

    Aghion,P.AndS.DurlaufHandbookofEconomicGrowthvol1A,chapter7.

    BanerjeeA.V.andE.Duflo,2003.InequalityandGrowth:WhatCantheDataSay,Journalof

    EconomicGrowth,8(3):267299.

    Banerjee,A.V.andA.F.Newman,1993."OccupationalChoiceandtheProcessof

    Development,"JournalofPoliticalEconomy,101(2):27498,April.

    Bardhan,P.,S.BowlesandH.Gintis.1998."WealthInequality,WealthConstraintsand

    EconomicPerformance,inA.B.AtkinsonandF.Bourguignon(eds.),2000.HandbookofIncome

    Distribution,Elsevier,volume1,no.1.

  • 7/30/2019 Social Protection Growth 2011-17

    25/29

    25

    Beegle,K.,J.DeWeerdtandS.Dercon.2011."MigrationandEconomicMobilityinTanzania:

    EvidencefromaTrackingSurvey",ReviewofEconomicsandStatistics.

    Beegle,K.,J.DeWeerdtandS.Dercon2010.Orphanhoodandhumancapitaldestruction:is

    therepersistenceintoadulthood?.Demography,Vol.47,No.1,Februarypp.163180.

    Beegle,K.,J.DeWeerdtandS.Dercon.2008.AdultMortalityandConsumptionGrowthin

    Tanzania,2008,EconomicDevelopmentandCulturalChange,vol56,2pp.299326.

    Behrman,J.,Y.Cheng,andP.Todd.2004.EvaluatingPreschoolProgramswhenLengthof

    ExposuretotheProgramVaries:ANonparametricApproach.ReviewofEconomicsand

    Statistics86(1):10832.

    Behrman,J.R.,S.W.ParkerandP.E.Todd.2004.MediumTermEffectsoftheOportunidades

    ProgramPackage,IncludingNutrition,onEducationofRuralChildrenAge08in1997.

    Unpublishedmanuscript.

    Besley,T.,R.PandeandV.Rao.2008.JustRewards?LocalPoliticsandPublicResource

    AllocationinSouthIndia.STICERD,LSE,discussionpaper49.

    Bowles,S.,S.DurlaufandK.Hoff.2006.PovertyTraps,PrincetonUniversityPress,2006

    Caeyers,B.andS.Dercon,2011.PoliticalConnectionsandSocialNetworksinTargeted

    TransferProgrammes:EvidencefromRuralEthiopia,EconomicDevelopmentandCultural

    Change,forthcoming.

    Cai,H.,Y.Chen,H.FangandL.A.Zhou,2010.2010.Microinsurance,TrustandEconomic

    Development:EvidencefromaRandomizedNaturalFieldExperiment,BREADWorkingPaper

    279,January.

    Chen,S.andM.Ravallion.2008.TheDevelopingWorldisPoorerthanweThought,butnoLess

    SuccessfulintheFightAgainstPoverty.WorldBankPolicyResearchWorkingPaperno2008.

    Cole,S.,X.Gine,J.Tobacman,P.Topalova,R.Townsend,J.Vickery,2010."Barrierstohousehold

    riskmanagement:evidencefromIndia,"PolicyResearchWorkingPaperSeries5504,TheWorld

    Bank.

    Conning,J.andM.Kevane.2002."CommunityBasedTargetingMechanismsforSocialSafety

    Nets:ACriticalReview,"WorldDevelopment,Elsevier,vol.30(3),pages375394,March.

    Cox,D.,B.HansenandE.Jimenez,E.2004.Howresponsiveareprivatetransferstoincome?

    Evidencefromalaissezfaireeconomy,JournalofPublicEconomics.

    Das,J.QT.DoandB.Ozler.2005.ReassessingConditionalCashTransfers,WorldBank

    ResearchObserver,vol20:5780

  • 7/30/2019 Social Protection Growth 2011-17

    26/29

    26

    Dasgupta,P.1997.PovertyTrapsinD.KrepsandK.Wallis(eds)AdvancesinEconomicsand

    Econometrics:theoryandapplicationsCambridge:CambridgeUniversityPress.

    deMel,S.,D.McKenzieandC.Woodruff.2008.Returnstocapital:Resultsfromarandomized

    experiment.QuarterlyJournalofEconomics.123(4):13291372

    deJanvry,E.,andE.Sadoulet,2002,AlternativeTargetingandCalibrationSchemesfor

    EducationalGrantsPrograms:LessonsfromProgresa.Mimeo.

    Dercon,S.,J.HoddinottandT.Woldehanna.2005.ShocksandConsumptionin15Ethiopian

    Villages,19992004,JournalofAfricanEconomies14.4:55985

    Dercon,S.andL.Christiaensen.2011.ConsumptionRisk,TechnologyAdoptionandPoverty

    Traps:evidencefromEthiopia.JournalofDevelopmentEconomics

    Dercon,S.2008.FateandFear:RiskanditsConsequencesinAfrica,JournalofAfrican

    Economies,vol.17(Supplement2):ii97ii127

    Dercon,S.2004.GrowthandShocks:evidencefromruralEthiopia.JournalofDevelopment

    Economics,August,vol74(2),pp.30929.

    Dercon,S.2002.IncomeRisk,CopingStrategiesandSafetyNets,(2002),WorldBankResearch

    Observer,December,Vol.17(2),pp.14166.

    Dercon,S.1996.Risk,CropChoiceandSavings.EconomicDevelopmentandCulturalChange.

    Dercon,S.andI.Outes.2011.IncomeDynamicsinRuralIndia19752006.Mimeo.

    Dercon,S.andC.Porter.2010.Liveaidrevisited:longtermimpactsofthe1984Ethiopian

    famineonchildren.CentrefortheStudyofAfricanEconomies,WPS/201038

    Dercon,S.,J.W.GunningandA.Zeitlin.2011.Thedemandforinsuranceunderlimited

    credibility:evidencefromKenya,mimeoOxfordUniversity.

    Dercon,S.andM.Kirchberger.2009.Literaturereviewonmicroinsurance,Microinsurance

    Paper1,InternationalLabourOrganisation.

    Devereux,S.R.SabatesWheeler,2008.TransformativeSocialProtection:thecurrencyofsocial

    justiceinA.BarrientosandD.Hulme,SocialProtectionforthePoorandPoorest.Palgrave.

    Dreze,J.andR.Khera.2009.Thebattleforemploymentguarantee,Frontline,January16.

    Duflo,E.2003.GrandmotherandGranddaughters:theeffectoftheoldagepensiononchild

    healthinSouthAfrica.WorldBankEconomicReview17:115.

    Edmonds2006.ChildlaborandschoolingresponsestoanticipatedincomeinSouthAfrica.

    JournalofDevelopmentEconomics.81(2):386414.

  • 7/30/2019 Social Protection Growth 2011-17

    27/29

    27

    EuropeanReportonDevelopment.2010.SocialProtectionforInclusiveDevelopmentAnew

    perspectiveinEUcooperationwithAfrica.EU.

    Fafchamps,M.,D.McKenzie,S.QuinnandC.Woodruff.2011.Wheniscapitalenoughtoget

    femalemicroenterprisesgrowing?EvidencefromarandomizedexperimentinGhana.Oxford

    mimeo.

    Fernald,L.C.H.,Gertler,P.J.,andNeufeld,L.M.2009.TenyearimpactofOportunidades

    Mexico'sconditionalcashtransferprogramonchildgrowth,cognition,languageandbehavior.

    TheLancet374(9706):19972005.

    Fernald,L.C.H.,Gertler,P.J.,andNeufeld,L.M.2008.TheImportanceofCashinConditional

    CashTransferProgramsforChildHealth,GrowthandDevelopment:ananalysisofMexicos

    OportunidadesTheLancet371(9615):828837.

    Gertler,P.J.andL.C.Fernald.2004.TheMediumTermImpactofOportunidadesonChild

    DevelopmentinRuralAreas.Unpublishedmanuscript.

    Gertler,P.,andJ.Gruber.2002.InsuringConsumptionagainstIllness.AmericanEconomic

    Review92(1):5170

    Gertler,P,S.MartinezandM.RubioCodina.2006.InvestingCashTransferstoRaiseLongTerm

    LivingStandards,WorldBankPolicyResearchWorkingPaperNo.3994

    Gibson,J.,S.OliviaandS.Rozelle.2006.HowWidespreadAreNonLinearCrowdingOut

    Effects?TheResponseofPrivateTransferstoIncomeinFourDevelopingCountries.

    DepartmentofAgriculturalandResourceEconomics,UniversityofCalifornia,Davis

    Gilligan,D.,J.Hoddinott,andA.SeyoumTafesse.2009.TheImpactofEthiopia'sProductive

    SafetyNetProgrammeanditsLinkages,"TheJournalofDevelopmentStudies,Taylorand

    FrancisJournals,vol.45(10):16841706

    GranthamMcGregor,S.M.,C.A.Powell,S.P.WalkerandJ.H.Himes.1991.Nutritional

    Supplementation,PsychosocialStimulation,andMentalDevelopmentofStuntedChildren:The

    JamaicanStudy.Lancet338:15.

    GranthamMcGregor,S.M.,S.P.Walker,S.M.ChangandC.A.Powell.1997.EffectsofEarly

    ChildhoodSupplementationwithandWithoutStimulationonLaterDevelopmentinStunted

    JamaicanChildren.AmericanJournalofClinicalNutrition66:247253.

    GranthamMcGregor,S.,Y.B.Yeung,S.Cueto,P.Glewwe,L.RichterandB.Strupp.2007.

    InternationalChildDevelopmentSteeringGroup,Lancet,6369(9555):6070.

    GrowthCommission2008.TheGrowthReport:StrategiesforSustainedGrowthandInclusive

    Development,CommissiononGrowthandDevelopment,TheWorldBank.

  • 7/30/2019 Social Protection Growth 2011-17

    28/29

    28

    ILO(2008),Canlowincomecountriesaffordbasicsocialsecurity?SocialSecurityPolicy

    Briefings,PaperNo3,SocialSecurityDepartment,ILO,Geneva.

    Jalan,J.andM.Ravallion.2004.HouseholdIncomeDynamicsinRuralChinainS.Dercon,

    InsuranceagainstPoverty,OxfordUniversityPress.

    Jensen,R.(2003)DoPrivateTransfersDisplacetheBenefitsofPublicTransfers?Evidencefrom

    SouthAfrica.JournalofPublicEconomics88,pp.89112

    Karlan,D.andJ.Zinman(2010)Expandingmicroenterprisecreditaccess,IPAworkingpaper.

    Lybbert,T.,C.Barrett,S.Desta,andD.LayneCoppock.2004."StochasticWealthDynamicsand

    RiskManagementAmongaPoorPopulation",TheEconomicJournal114:75077

    Matul.M.,M.McCord,C.PhilyandJ.Harms.2010TheLandscapeofMicroinsuranceinAfrica,

    ILOMicroinsurancefacilityworkingpaperno.4

    Morduch,J.1995.Incomesmoothingandconsumptionsmoothing.JournalofEconomic

    Perspectives.9(3):103114.

    Omamo,S.W.,U.GentiliniandS.Sandstrm.2010.Revolution:FromFoodAidtoFoodAssistance

    InnovationsinOvercomingHunger,WorldFoodProgram

    Persson,T.andG.Tabellini,1994.IsInequalityHarmfulforGrowth?AmericanEconomic

    Review,84:60021

    Rajasekhar,D.,E.Berg,M.Ghatak,R.ManjulaandS.Roy.2011.ImplementingHealthInsurance:

    TheRolloutofRashtriyaSwasthyaBimaYojanainKarnataka.EconomicandPoliticalWeekly,

    XLVI(20):5663.

    Ravallion,M.2010."DoPoorerCountriesHaveLessCapacityforRedistribution?,"Journalof

    GlobalizationandDevelopment,BerkeleyElectronicPress,vol.1(2):1.

    Ravallion,M.2008.BailingouttheWorldsPoorest.WorldBankPolicyResearchWorking

    Paperno4763.

    Ravallion,M.2007.DiBao:aguaranteedminimumincomeinurbanChinaWorldBankPolicy

    ResearchWorkingPaperno4305.

    Ravallion,M.2006.TransfersandSafetyNetsinPoorCountries:RevisitingtheTradeoffsand

    PolicyOptionsinA.Banerjee,R.BenabouandD.Mookherjee,UnderstandingPoverty,Oxford

    UniversityPress.

    Ravi,S.andM.Engler.2009.WorkfareinLowIncomeCountries:AnEffectiveWaytoFight

    Poverty?TheCaseofIndiasNREGS,mimeo.

  • 7/30/2019 Social Protection Growth 2011-17

    29/29

    29

    Rosenzweig,MandH.Binswanger,1993.Wealth,WealthRiskandCompositionand

    ProfitabilityofAgriculturalInvestments,EconomicJournal103:5678.

    Schady,N.2006.EarlyChildhoodDevelopmentinLatinAmericaandtheCaribbean

    EconomaVolume6,Number2,Spring2006,pp.185225(previouslyWorldBankPolicy

    ResearchWorkingPaper3869)

    VandeWalle,D.1998.TargetingRevisited.WorldBankResearchObserver13(2):231248

    Woldehanna,T.2009.ProductiveSafetyNetProgrammeandChildrensTimeUseBetween

    WorkandSchoolinginEthiopia,YoungLivesWorkingPaperno.40.

    WorldBank.2001.AttackingPoverty:WorldDevelopmentReport2000/2001OxfordUniversity

    PressfortheWorldBank.

    WorldBank.2006.WorldDevelopmentReport:EquityandDevelopment.OxfordUniversity

    PressfortheWorldBank.