英国发展研究所-针对长期危机：尼日尔案例研究（英）-2023.12_市场营销策划_重点报告20230.docx

uaidfrom the British peopteBETTERASSISTANCEINCRISESRESEARCHTargetinginProtractedCrises:NigerCaseStudyFredMerttens5LouisHodeyandAlexandraDoyleDecember2023ImplementedbyinstHutofdvlop<nfftttudiBetterAssistanceinCrises(BASIC)Research(fundedbyUKaid)aimstoinformpolicyandprogrammingonhowtohelppoorandvulnerablepeoplepebetterwithcrisesandmeettheirbasicneedsthroughmoreeffectivesocialassistance.AllcostsrelatedtoBASICResearcharecoveredbytheUKForeign,Commonwealth&DevelopmentOffice(FCDO).Formoreinformationabouttheprogramme,visitBASlCReSearChLinkedmandBASlCReSearChOPenDoCSCollection.SummaryTargetingsocialassistanceinsituationsofprotractedconflict,displacementorrecurrentclimateshockssothatitreachesthosemostinneedinatimelyandeffectivemanner,andwithoutdoingfurtherharm,isacomplextechnicalandpoliticalchallengefordevelopmentandhumanitarianactorsacrossgovernmentandnon-governmentsectors.Trade-OffSinvolvingcostsbeyondtheeconomic-suchasriskofexclusion,andconcernsoverprotectionandsocialcohesion-raisekeyquestionsaboutwhototarget,howtotargetorwhethertotargetatall(i.e.throughuniversalcoverageorlotteries).Whiletargetingeffectivenessisfairlywellresearchedinstabledevelopmentcontexts,thereismuchlessunderstandingandevidenceaboutwhatworksbestinprotractedcrisissettings.ThispaperisoneofthreecountrycasestudiescoveringEthiopia,Niger,andNigeria.Eachcasestudyfocusesonaspecifictimeperiodtodemonstratehowthechallengeoftargetingmightbeapproachedinsituationsofprotractedcrisesbasedonactualhistoricalexamples.Theobjectiveofthisapproachistogenerateinsightsonthetopicoftargetinginacomparativewayandbyconsideringvulnerabilityasadynamiccondition.ThecasestudieswillbeusedtoproduceasynthesispaperthatwillbringfindingstogethertodrawoutgenerallessonsIortargetingsocialassistanceincrisiscontexts.Thiscasestudyfocusesonthe2012Saheldrought,whichproducedafoodsecurityemergencyinNiger.Usingtwoyearsofpaneldata(2011and2014)fromeithersideofthe2012droughtcrisisintheSahel,weanalysewhowasaffectedbytheshockandhowtheywereimpacted.Weusethedatatomodelthenotionalperformanceofdifferentpotentialtargetingapproaches-hadtheybeenusedatthetime-toindicatethetypesofchoicesandtrade-offsthatmaybenecessarywhenselectingappropriatetargetingcriteriaforresponsestosimilarcrisesinthefuture.Finally,wealsoconsiderthestatusofenablingconditions(suchasmobilephonenetworkandelectricitycoverage)forimplementingdifferenttargetingapproachesinshock-responsivesocialassistance.AbouttheauthorsFredMerttensisaseniorresearcherspecialisedinpovertyanalysisandsocialprotectionpolicy.Hehasover16years1experiencedesigningandleadingmultidisciplinaryappliedresearchprojectsforpublicpolicyreformforarangeofinternationaldevelopmentorganisations,governmentsandnon-governmentalorganisationsinAfrica,theMiddleEast,CentralAsiaandEurope.FredisTechnicalLeadonthetargetingworkstreamfortheBASICResearchprogramme.1.ouisHodeyisadevelopmentresearcherspecialisinginrurallivelihoodstudies,povertyanalysis,impactevaluations,socialprotection,agriculturalcommercialisationandclimatechange.Hehasovertenyears,experienceindesigningandimplementingmultidisciplinaryresearchindevelopingcountriesincludingBurkinaFaso,Colombia,Ethiopia,Ghana,Mali,Niger,NigeriaandZimbabwe.Mostrecently,LouisworkedasaPrincipalResearchAssistantattheInstituteofStatistical,SocialandEconomicResearch,UniversityofGhana.LouisisaquantitativeresearcheronthetargetingworkstreamfortheBASICResearchprogrammeattheInstituteofDevelopmentStudies.AlexandraDoyleisaninternationaldevelopmentresearcherspecialisinginsocialprotectionpolicyandmonitoringandevaluation.ShehasworkedonarangeofprojectsacrossAfrica,AsiaandtheMiddleEast,includingtheGambia,Kenya,Lebanon,Mozambique,Nepal,Nigeria,SierraLeone,SouthAfrica,Thailand,ZambiaandZimbabwe.MostrecentlyAlexwasMonitoringandEvaluationDeputyDirectorfortheWesternCapeGovernment,SouthAfrica.ExecutivesummaryIntroductionTheBetterAssistanceinCrises(BASIC)Research,systemsfordesignanddelivery*themeincludesatargetingworkstreamthatlooksathowtomakethedesignanddeliveryofsocialassistanceprogrammesmoreeffectiveandresponsiveincrisiscontexts.Thisworkstreamaimstoproducefouroutputs,comprisingthreecountrycasestudies-coveringEthiopia,Niger,andNigeria-andasynthesispaperthatwillbringthefindingsfromthecasestudiestogethertodrawoutgenerallessonsTortargetingsocialassistanceincrisiscontexts.Thefindingsfromthecasestudiesareintendedtoinformgloballearning,anddecisionmakersingovernments,developmentpartners,internationalfinancialinstitutions,andhumanitarianactorswhoareinvolvedinthedesignanddeliveryofsocialassistanceprogrammesinsituationsofprotractedcrises.Eachcountrycasestudywillfocusonaspecificperiodtodemonstratehowthechallengeoftargetingmightbeapproachedinsituationsofprotractedcrisesbasedonanactualhistoricalexample.TheobjectiveofthisapproachistogenerateinsightsonthetopicOftargetinginacomparativewayandbyconsideringvulnerabilitytosuchcrisesasadynamiccondition.ThiscountrycasestudyreportfocusesonNiger.Itconcentratesonthespecificfoodinsecuritycrisisthataffectedthecountryasaresultofthe2012Saheldrought,whichitselfwasjustoneincidentinalongerprotractedcrisisinducedbyclimatevulnerability.The2012SaheldroughtandfoodcrisisIn2012,forthethirdtimeintenyears,Nigerfacedamajorfoodsecuritycrisis,initiallycausedbydrought.Anestimated6.4millionpeoplewereaffectedbyfoodinsecurity,particularlyagro-pastoralisthouseholdsin"11llab6ri,TahouasZinder,andDiffaregions.The2012rainyseasonthenbroughtabove-averagerainfallthatledtofloodingintheNigerRivervalley.Bytheendof2012,thefloodswereestimatedtohaveaffectedmorethan530,000people,destroyinginfrastructureandfloodingagriculturalfields,furthercontributingtoharvestlossandfoodinsecurity.Severalfactorscompoundedthefoodsecurityemergency,includingdisease,pestilence,conflict,populationdisplacement,andinflation.HouseholdsinNigerwerestillrecoveringfromadroughtandfoodcrisisin2010andwerethereforeparticularlyvulnerableeventosmallshocks.Further,thereturnofNigerienmigrantsinlate2011andarrivalofMalianrefugeesin2012strainedlocalfoodsuppliesinhostcommunitiesandplacedpressureonreceivinghouseholds.ThesocialassistanceandhumanitariancrisisresponseGivenNiger,sexposuretoclimaticshocksandvulnerabilitytofoodinsecurity,theGovernmentofNiger(GoN)preparesanAnnualPlan(Plandesoutien)tosupportfood-insecurehouseholds.TheplanisdevelopedbytheGoNbutisprimarilydeliveredbyUNagenciesandotherhumanitarianactorsduetothevastresourcerequirements.Basedonfindingsfromavulnerabilityassessment,whichidentifiedapproximately7,000localitiesinneedofsupport,theGoNbegantoimplementtheplanfromOctober2011,sixmonthsearlierthanpreviousdroughtresponses.Conditionalandunconditionalcashtransfersaccountedfor45percentoftheoverallbudgetofthe2011/12annualplan(WFP2013).Socialassistanceprogrammeslaunchedinresponsetothecrisismainlyusedacombinationofgeographictargetingandcommunity-basedhouseholdtargeting.Thedecisiontotarget,ratherthanprovideauniversalresponsetofoodinsecureareas,wasdrivenbyresourceconstraints.Mostcashtransferprogrammesweredesignedandimplementedasemergencyinitiatives,piggybackingonexistinginfrastructurewherepossible.However,atthetimeofthecrisis,routinesocialassistanceprogrammeswereintheirinfancyandtheinfrastructuretoswiftlyscaleupassistancetocovertheaffectedpopulationviahorizontaland/orverticalexpansionwaslimited.AffectedpopulationsandwelfaretrendsTwowavesoftheNationalSurveyonHouseholdLivingConditionsandAgriculture(EnqueteNationalesurIesConditionsdeViedesMenagesetAgriculture-ECVMA),thenationalsurveythatprovidesconsumptionandpovertyestimatesforthecountry,conductedintheyears2011and2014,respectively,siteithersideofthe2012droughtcrisis.Inwhatfollows,weusethisdatatounderstandwhowaslikelytohavebeenaffectedbytheshockandhowtheywereimpacted.AffectedpopulationsandtheircharacteristicsThenationalsurveydatashowsthatdroughtanddrought-relatedshocksnegativelyimpactabroadanddiversepopulationgroup,withdifferentsub-populationsaffectedindifferentwaysandtodifferentdegrees.Corroboratingtheliterature,ouranalysisindicatesthatthepeoplemostaffectedbythe2012droughtwereconcentratedinTillaberiandTahouaregions,and,toalesserextent,Maradi,Zinder5andDiffa.TillaberiandTahouahadthehighestproportionofpastoralistsandagro-pastoralists,aswellasthelowestproportionofpeoplelivinginurbanareas.Atnationallevel,thedatashowsthatthewelfaredistributionofthepopulationwasveryflat,withlimiteddifferencebetweenconsumptionlevelsforlargepartsofthepopulation,particularlythoseinruralareas.Overall,householdsintheaffectedregionsdemonstratedsignificantvulnerabilitiesintermsofpoverty,labourcapacityandqualityOflivingconditions(e.g.accesstosanitationorelectricity),bothinabsolutetermsandrelativetootherpartsofthepopulation.Welfaretrendspre-andpost-shockWefindavariedpicturewhenanalysingtrendsinindicatorsofpoverty,foodsecurityandvulnerabilitytoshocksbetween2011and2014acrossvariousgroups.Ingeneral,therewasahighlevelofdynamismintermsofwelfarestatusovertime.Analysisindicatesthathouseholdsacrossthepopulationwerehighlyvulnerabletofoodinsecurityatanygivenmoment.Further,inboth2011and2014,thesituationremainedsimilarintermsofhumancapitalandlivingconditions.Thelackofdevelopmentinthesedimensionssuggeststhatthepopulationwashighlylikelytoremainvulnerabletopoverty,foodinsecurityandshocks.TargetingsimulationsUsingtheECVMA5wemodelthenotionalperformanceofdifferenttargetingapproachestoindicatethetypesofchoicesandtrade-offsthatmaybenecessarytomakewhenselectingappropriatetargetingcriteriaforresponsestosimilarcrisesinthefuture.Theapproachesselectedforthemodellingexerciseinclude'categorical'approaches(e.g.basedontheageofrecipients)andformula-basedapproaches(e.g.asimpleproxymeanstest(PMT).PerformanceatnationallevelWefirstcomparetheperformanceofthepossibletargetingapproachesatthenationallevel.Theresultsshowthatthecoveragerateofthetargetpopulationisprimarilydrivenbytheoverallcoveragerate,whichinthiscaseishighestwhentargetingchildrenagedunderfiveyearsandlowestwhentargetingolderpeopleover64.Thedatashowsthatinclusionandexclusionerrorstendtobehigh(over40percent)nomatterwhichselectionmethodisadopted.Thisisbecausethetargetpopulationisabroadanddiversegroup,whichisnotcomprehensivelycapturedbyanysinglehousehold-orindividual-levelcharacteristic.Unsurprisingly,theformula-basedapproachperformsmarginallybetterthancategoricalselectioncriteriaintermsofselectingpoorpeople.DisaggregatingtargetingperformancebygeographiclocationWealsointroduceageographicelementtothetargetingtoassesshowthisaffectstargetingperformance.Inurbanareas,inclusionandexclusionerrorsaregenerallyhigherthanatthenationallevel,asfewerurbandwellersareinthetargetpopulation.ThePMTselectionmethodalsoperformsbetterinurbanareasthannationallyinidentifyingpoorpeople,asthewelfaredistributionislessflatinurbanareas.Inruralareas,incontrast,inclusionandexclusionerrorsacrossallselectionmethodstendtobemarginallysmallerthanatthenationallevel,reflectingthehighershareofthetargetpopulationlivinginruralareascomparedtourbanareas.Atthesametime,duetotheflatterwelfaredistributioninruralareas,thePMTperformsslightlylesswellatselectingpoorpeople,andallselectionmethodsperformlittlebetterthanrandomwhenitcomestoidentifyingthosevulnerabletofoodinsecurity.Finally,lookingattargetingperformanceinselectedregionsofthecountry,weseethataddingageographicelement(region)tothetargetingapproachcanreduceinclusionandexclusionerrors.Overall,thesesimulationsshowthattherearenosingleorsimplecriterionthatcomprehensivelycapturesthepopulationthatisespeciallyvulnerabletodroughtordrought-relatedshocks,thoughcombininghousehold-orindividual-levelcharacteristicswithgeographictargetingcanincreasecoverageandreduceinclusionandexclusionerrors.Community-basedtargetingSomeformofcommunity-basedtargeting(CBT)wasthemainmethodactuallyusedbymostmajorprogrammesinresponsetothe2012crisis(albeitcombinedwithgeographictargeting).WhileitisnotpossibletoreplicatetheresultsofCBTselectionmethodsinthedata-aswehavenowayofknowingwhocommunitieswouldselectinpractice-studiesfromNigerandelsewhereindicatethatCBTislikelytoincorporatesimilarlevelsofinclusionandexclusionerrors.WhileitmaybeexpectedthattheparticipatorynatureofCBTwouldrenderitcomparativelylegitimateasatargetingapproach,evidencefromNigerandelsewheresuggestthatformula-basedorcategoricaleligibilitycriteriamayalsogarnerhighlevelsoflegitimacy,perhapsevenmoresothanCBT.Forexample,researchershavefoundthatformula-basedmethodswereperceivedtobemorelegitimatethanCBTinNigerduetoperceivedmanipulationbyCBTcommitteemembersandinformationimperfectionsaffectingtheimplementationofCBT.OperationalcontextAtthelevelofoperations,keyinformantsidentifiedthemainchallengeatthetimeofthe2012droughtastheneedtoestablishinstitutionsandcoordinationmechanisms.TheECVMAandtheJointNationalSurveyonHouseholdVulnerabilitytoFoodInsecurityinNiger(EnqueteNationaleConjointesurIaVulnerabilitealns6curiteAlimentairedesMenagesauNiger-EVIAM)wereexistingsystemsthatprovidedusefulinformationonthelevelanddistributionofneed,butthestrengtheningand/orcreationofcoordinationandgovernancestructuresatbothnationalandsub-nationallevelswasstillrequired.Inadditiontotheinstitutionalarchitecture,deliverysystemsrelyonunderpinninginfrastructuresuchasroads,electricity,communicationsinfrastructure,financialservicesinfrastructureandcivilregistrationsystems,alongsidehumancapitalembodiedintheskilllevelsofimplementingagentsandeducationallevelsofthepopulation.The2018ECVMAsurvey(ECVMA2018)andotherdataindicatethat,inNiger,theseunderpinninginfrastructuresarenothighlydeveloped.ConclusionsInthecontextofhighlevelsofpovertyandvulnerabilitytoshocks,thereisaninherentchallengeinreachingabroadanddiversepopulationcharacterisedbyintersectingvulnerabilitiestopoverty,foodsecurity,andbothcovariateandidiosyncraticshocks.Socialassistanceprogrammingthereforehastocontendwithseveraldifferentdimensionswhenselectinganappropriateapproachtotargeting.Alignthetargetingapproachtoclearpolicyobjectives.Thereareimportantdifferencesbetweenroutinesocialassistanceandemergencyresponse,particularlyintermsofobjective(i.e.mitigatingpovertyandlifecyclerisksversusaddressingimmediate,life-threateningrisks).Nevertheless,thereareoverlapsbetweenthetwopolicydomains,especiallyincontextssuchasNigerwherechronicvulnerabilitytoshocksresultsinprotractedexposuretoEmergency*situationssuchassevereandrecurrentfoodinsecurity.Inthesecircumstances,socialassistancepoliciesandemergencyresponsepoliciesneedtobeclearlydelineatedandplaycomplementaryroles.Thetargetingcriteriaforanygivensocialassistanceoremergencyresponsepolicyhavetoclearlyalignwiththestatedpolicyobjectives.Forexample,iftheprimaryobjectiveofthepolicyistopreventmalnutrition,targetingyoungchildrenwillbemoreappropriatethantargetingpoorhouseholds.If,however,thepolicyisintendedtomitigatepoverty,formula-basedapproachestargetingmonetarilypoorpeoplemaybemoreappropriate.Aparticularconsiderationthatmaybeimportantintheselectionoftargetingcriteriaforemergencyresponseisthespeedatwhichitcanbedelivered.Communitiesneedtoclearlyunderstandandaccepttherationaleandtargetingcriteria.Peo