多维幸福衡量实践：聚焦于改进全球多维贫困指标的综述（英）

PublicDisclosureAuthorized PublicDisclosureAuthorized PolicyResearchWorkingPaper10800 MultidimensionalWell-BeingMeasurementPractices AReviewFocusedonImprovingGlobalMultidimensionalPovertyIndicators BenoitDecerf DevelopmentResearchGroup&PovertyandEquityGlobalPracticeJune2024 PolicyResearchWorkingPaper10800 Abstract Multidimensionalwell-beingindicatorshavethepoten-tialtoreducethe“bias”associatedtomonetaryindicators.However,theyfacestringentdataconstraints.Thispaperstudiestheconstructionofindicatorsthatstrikeabalancebetween(i)reliabilityinapproximatingconceptuallysoundwell-beingcomparisonsand(ii)simplicityofapplicationandcommunication.Therecommendationsfocusonglobalmultidimensionalpovertymeasures.Thepaperidentifiesthreepotentialsourcesofimprovements:“wasting”lessdata,betterfilteringthedata,andfurtherdeveloping multidimensionalanalysis.Lessinformationwouldbe“wasted”byavoidingneedlesslydichotomizingallthevari-ables,usingtheavailablemortalitydata,andcombiningvariablesfromseparatesurveys.Tofilterthedatabetter,“equalweights”couldbereplacedbyweightsselectedfromexternalinformationonpreferences.Whenthedatapermit,theunitofanalysisshouldbeswitchedfromhouseholdleveltoindividuallevel.Finally,multidimensionalindi-catorsshouldbeusedtohelpmovebeyondasuboptimal“dimension-by-dimension”approachtopolicymaking. ThispaperisaproductoftheDevelopmentResearchGroup,DevelopmentEconomicsandthePovertyandEquityGlobalPractice.ItispartofalargereffortbytheWorldBanktoprovideopenaccesstoitsresearchandmakeacontributiontodevelopmentpolicydiscussionsaroundtheworld.PolicyResearchWorkingPapersarealsopostedontheWebathttp://www.worldbank.org/prwp.Theauthormaybecontactedatbdecerf@worldbank.org. ThePolicyResearchWorkingPaperSeriesdisseminatesthefindingsofworkinprogresstoencouragetheexchangeofideasaboutdevelopmentissues.Anobjectiveoftheseriesistogetthefindingsoutquickly,evenifthepresentationsarelessthanfullypolished.Thepaperscarrythenamesoftheauthorsandshouldbecitedaccordingly.Thefindings,interpretations,andconclusionsexpressedinthispaperareentirelythoseoftheauthors.TheydonotnecessarilyrepresenttheviewsoftheInternationalBankforReconstructionandDevelopment/WorldBankanditsaffiliatedorganizations,orthoseoftheExecutiveDirectorsoftheWorldBankorthegovernmentstheyrepresent. ProducedbytheResearchSupportTeam MultidimensionalWell-BeingMeasurementPractices:AReviewFocusedonImprovingGlobalMultidimensionalPovertyIndicators∗ BenoitDecerfDECPI,WorldBank JEL:I32. Keywords:MultidimensionalWell-Being,PovertyMeasurement,Dataconstraints. ∗Acknowledgments:IamgratefultoPaolaBallon,MeryFerrando,JedFriedman,BalintMenyhert andBerkOzlerforhelpfulsuggestions.IthankalltheparticipantsattwointernalseminarsoftheWorldBank.ThispaperwaspreparedaspartoftheResearchFrontiersinPovertyMeasurementseriesorganizedbytheGlobalSolutionsGroupDataforPolicyofthePovertyandEquityGlobalPracticeandDEC,ledbyUtzPapeandJedFriedman.Thefindings,interpretations,andconclusionsexpressedinthispaperareentirelythoseoftheauthorandshouldnotbeattributedinanymannertotheWorldBank,toitsaffiliatedorganizations,ortomembersofitsBoardofExecutiveDirectorsorthecountriestheyrepresent.TheWorldBankdoesnotguaranteetheaccuracyofthedataincludedinthispaperandacceptsnoresponsibilityforanyconsequenceoftheiruse. 1Introduction Well-beingindicatorsareimportanttoolsforevidence-basedpolicymaking.Theyallowtargetingtheworst-off,allocatingeffortstowardsregionswiththelowestwell-beingandmonitoringprogress.Thesetasksaremostoftendonewithmonetaryindicators.How-ever,monetaryindicatorstypicallyyieldbiasedwell-beingcomparisonsbecausetheydonotaccountforthemultidimensionalnatureofwell-being.Multidimensionalwell-beingindicators,whichalsoaccountfornon-monetaryoutcomes,canpotentiallyreducethisbias. Inpractice,theconstructionofmultidimensionalindicatorsfacesstringentdatacon-straints.Firstandforemost,therearenowell-definedpricesfornon-monetaryoutcomesor“achievements”.Asaresult,thepractitionertypicallyhaslimitedornoinformationonthepreferencethatisrelevanttorankachievementvectors.Second,thejointdistributionbetweenkeydimensionsofwell-beingisoftenmissing.Hence,thepractitionermayob-servethepartialdistributions,butdoesnotknowwhetherthesameindividualscumulatelowachievementsinseveraldimensions.Thesetwoconstraintspreventpractitionersfromconstructingconceptually-idealmultidimensionalindicators. Thesedataconstraintshaveledtotwooppositeextremereactions.Thefirstreactionconsistsinditchingapreference-basednotionofwell-beingaswellasnon-paternalism.Manypragmatictheoriesformultidimensionalindicatorsdealwiththeabsenceofinfor-mationontherelevantpreferencebysettingasidepreferencetheory.Thisisforinstancethecaseofthetheorygroundingthemultidimensionalpovertymeasuresmostusedinpractice.1Thesecondextremereaction,whichisperhapstriggeredbythefirstextremereaction,istodisqualifymultidimensionalindicators.Accordingtothisview,adashboardofindicatorsissufficientforpolicymakingandasummarymultidimensionalindicatorbringsnoaddedvalue.Admittedly,theconceptualfoundationsforpragmaticmultidi-mensi