亚开行-进入壁垒：消除巴基斯坦城市求职中的性别差距（英）-2023.12_市场营销策划_重点报告20.docx

BARRIERSTOENTRYDECOMPOSINGTHEGENDERGAPINJOBSEARCHINURBANPAKISTANElisabettaGentile7NikitoKohli,NivedhithaSubramanian,ZuniaTirmazee,andKateVybornyNO. 707December 2023AdbeconomicsWORKINGPAPERSERIESADBEconomicsWorkingPaperSeriesBarrierstoEntry:DecomposingtheGenderGapinJobSearchinUrbanPakistanElisabetta Gentile, Nikita Kohli, NivedhithaSubramanian, Zunia Tirmazee1 and Kate VybornyNo. 707 I December 2023The ADB Economics Working Paper Series presents research in progress to elicit comments and encourage debate on development issues in Asia and the Pacific. The views expressed are those of the authors and do not necessarily reflect the views and policies of ADB or its Board of Governors or the governments they represent.Elisabetta Gentile (egentileadb.org) is a senior economist at the South Asia Department, Asian Development Bank. Nikita Kohli (nikita.kohli() duke.edu) is a PhD student and Kate Vyborny (katherine.vybornyduke.edu) is associate director ofthe DeVLabDUke and a research associate at the Department of Economics, Duke University. Nivedhitha Subramanian (nsubramabates.edu) is an assistant professor at Bates College. Zunia Tirmazee (zuniatirmazee) is an assistant professor at Lahore School of Economics.ICreativeCommonsAttribution3.0IGOlicense(CCBY3.0IGO)©2023AsianDevelopmentBank6ADBAvenue,MandaluyongCity,1550MetroManila,PhilippinesTel+63286324444;Fax+63286362444Somerightsreserved.Publishedin2023.ISSN2313-6537(print),2313-6545(electronic)PublicationStockNo.WPS230551-2DOI:TheviewsexpressedinthispublicationarethoseoftheauthorsanddonotnecessarilyreflecttheviewsandpoliciesoftheAsianDevelopmentBank(ADB)oritsBoardofGovernorsorthegovernmentstheyrepresent.ADBdoesnotguaranteetheaccuracyofthedataincludedinthispublicationandacceptsnoresponsibilityforanyconsequenceoftheiruse.ThementionofspecificcompaniesorproductsofmanufacturersdoesnotimplythattheyareendorsedorrecommendedbyADBinpreferencetoothersofasimilarnaturethatarenotmentioned.Bymakinganydesignationoforreferencetoaparticularterritoryorgeographicarea,orbyusingtheterm"untry"inthispublication.ADBdoesnotintendtomakeanyjudgmentsastothelegalorotherstatusofanyterritoryorarea.ThispublicationisavailableundertheCreativeCommonsAttribution3.0IGOlicense(CCBY3.0IGO)Byusingthententofthispublication,youagreetobeboundbythetermsofthislicense.Forattribution,translations,adaptations,andpermissions,pleasereadtheprovisionsandtermsofuseathttps:/www.adb.org/terms-use#openaccess.ThisCChtmaterialsintispublication.Ifthematerialisattributedtoanothersource,pleasentactthecopyrightownerorpublisherofthatsourceforpermissiontoreproduceit.ADBcannotbeheldliableforanyclaimsthatariseasaresultofyouruseofthematerial.Pleasentactpbsmarketing(g)adb.orgifyouhavequestionsorcommentswithrespecttoComenLorifyouwishtoobtaincopyrightpermissionforyourintendedusethatdoesnotfallwithintheseterms,orforpermissiontousetheADBlogo.CorrigendatoADBpublicationsmaybefoundatNote:Inthispublication,*PRs"referstoPakistanrupeesande$"referstoUnitedStatesdollars.ADBrecognizes"China"asthePeople'sRepublicofChina.ABSTRACTGendergapsinlabormarketoutcomespersistinSouthAsia.Anopenquestioniswhethersupplyordemandsideconstraintsplayalargerrole.WeinvestigatethisusingmatcheddatafromthreesourcesinLahore,Pakistan:representativesamplesofjobseekersandemployers;administrativedatafromajobmatchingplatform;andanincentivizedresumeratingexperiment.Employers1genderrestrictionsarealargerconstraintonwomen,sjobopportunitiesthansupply-sidedecisions.Athigherlevelsofeducation,demand-sidebarriersrelax,allowingwomentoqualifyformorejobsbutatlowersalaries.Onthesupplyside,educatedwomenbecomemoreselectiveintheirsearch.Keywords:gender,discrimination,jobsearch,jobsplatform,vacancies,applicationsJELcodes:J16,J22,J23,R23Wearegratefulforvaluablefeedbackfromseminar,conference,andworkshopparticipantsattheAsianDevelopmentBank,DukeUniversity,theDiversityandHumanCapitalWorkshopatExeter,theEuropeanEconomicAssociation,theIZAMatchingWorkersandJobsOnlineWorkshop,theLACDevConference,theADEConferenceattheLahoreSchoolofEconomics,theNorthEastUniversitiesDevelopmentConsortiumConference,theMaineEconomicsConference,theYaIe-ADBIGender-SensitiveEconomicRecoveryConference,andtheYaleEGCworkshop.WethankLiviaAlfonsi,EmilyConover,RohiniPande,DanilaSerra,andNiharikaSinghforcomments.WethanktheentireJobTalashteamatCERPandDukeforexcellentresearchassistanceandprogrammaticsupport.WethankEricaFieldandRobGarlickforhelpfulcommentsonthedraftandcollaborationonthebroaderJobTalashprogramofresearch.WegratefullyacknowledgefundingforthisprojectfromADB1andforthebroaderJobTalashresearchprogramfromtheGLM-LIC13IE,IGC,J-PAL1PEDL,andtheNationalScienceFoundation(SES1629317).ThisresearchreceivedethicsapprovalfromDukeUniversity(2019-0067).Previouslypresentedas,ALeakyPipeline:DecomposingthegendergapinjobsearchinurbanPakistan/1 IntroductionVastgendergapsinemployment,stemmingfromlowlevelsofwomen'semployment,persistinmanylow-andmiddle-incomeeconomies,particularlyinSouthAsia,theMiddleEast,andNorthAfrica(Addatietal.,2016).Agrowingliteraturedocumentsthatsupplysidefactorssuchasself-selectingintooccupationsthatconformtogenderidentity,differingpreferencesforjobattributes,andgenderedsocialnormsabouttimeuseconstrainwomen'slaborsupplyinthesecontexts.(AkerlofandKranton,2000;CortesandPan,2017;Delfino,2022;DeanandJayachandran,2019;Subramanian,2021;MasandPallais52017;Fletcheretal.12018).However,demand-sideratherthansupply-sidefactorsmightformthebindingconstrainttowomen'slaborforalargerfractionofthepopulationincontextswithlowfemalelaborforceparticipation(FLFP).Indeed,asmallerbodyofworkdemonstratesthatdemand-sidefactorssuchasfirmgenderpreferencescancontributetogendergapsinemployment(KuhnandShen,2013;GoldinandRouse,2000;Hangartneretal.t2021;Ozenetal.i2019).Despitethewealthofresearchonlowfemaleemployment,wehavelimitedevidencequantifyingtherelativesizeofdemandandsupplyfactorsthatgiverisetothisphenomenon.Inaddition,muchoftheexistingevidencefocusesonspecificpopulations(e.g.educatedwomen)orsectors(e.g.governmentwork).Knowingwhethersupplyordemandconstraintsarebindinginthebroaderlabormarketisimportanttotargetpolicytoeithersupply-sideinterventions,suchasprovisionoffinancialservicestowomenorexposuretofemalerolemodels(Fieldetal.;Ahmedetal.,forthcoming),ortodemand-sideinterventionssuchaseliminatinggendercriteriainjobads,hiringquotas,wagesubsidies,physicalinfrastructuresuchaswomen,stoilets,oremployer-basedchildcare(Cardetal.,2021;Milleretal.,2022;KuhnandShen,2023).WeaddressthisgapintheliteraturebycombiningdatafromajobsearchplatformandincentivizedresumeratingexperimentfromLahore,Pakistan,toquantifytherelativeimportanceofdemand-sideandsupply-sidesourcesofthegendergap.Wedocumentthatatloweducationlevelsdemand-sideconstraintsaremuchlargerthansupply-sidefactors,butthisdemand-sidegapinquantityofjobopportunitiesclosesaseducationlevelsincreaseandjobsbecomemore,lwhite-collar.Ourempiricalapproachusesanovelcombinationofmatcheddatafromrepresentativesurveys,administrativedataonjobsearch,andexperimentaldatatoovercomefourkeychallengestoquantifyingthesizeofdemandandsupplysidefactorscontributingtothethegendergapinthelabormarket.Thefirstempiricalchallengeisthatsurveydatafromrepresentativesamplesofhouseholdsorfirmscanbeusedtoquantifytherealizedgendergapinequilibrium,butdonotallowtheanalysttodecomposehowmuchofthisgapcomesfrommen,sandwomen'swillingnesstosupplylaborversusfirms*demandformaleandfemalelabor.Second,alternativedatasourcessuchasjobplatformdataallowresearcherstoobservethedetailsofsearchactivitybyjobseekersandfirms;however,suchdataoftenhavelimitations(Nomuraetal.,2017).Theytypicallydonotallowtheresearchertoobservejobseekerpreferencesdirectly,onlytoinferthemfromapplicationchoices,whicharealsoinfluencedbyotherfactorssuchasthevacanciesavailableontheplatform.Similarly,theytypicallydonotallowtheresearchertoobserveawell-definedchoicesetofvacanciesobservedbythejobseeker,makingitdifficulttodisentanglewhetherthedecisiontoapplytoagivenvacancyisafunctionofthecharacteristicsofthatvacancyorofsearcheffortinbrowsingtheplatform(e.g.Belotetal.(2018);Wheeleretal.(2021);JonesandSen(2022);Banfietal.(2019);Matsudaetal.(2019),Third,selectionofbothfirmsandworkersintosearchandtheuseofjobsearchplatformslimitstheextenttowhichresultscanbeextrapolatedtothepopulationasawhole(Kurekovaetal.,2015).Addressingthesefirstthreechallenges,ourresearchpartnersattheCentreforEconomicResearchinPakistandevelopedanewjobmatchingplatform,JobTalash,andoffereditasafreeservicetorepresentativelistingsofthousandsofhouseholdsandthousandsoffirmsinasingleurbanlabormarket.Weemphasizetheimportanceofstudyingtheformergroup,economicallyinactive"latentworkers/*whomightbeinterestedinworkingbutareeconomicallyinactiveduetolackofopportunities,whorepresentthepopulationwiththelargestpotentialbenefitfromreductionsinlabormarketbarriers.ThisgroupisparticularlyimportantforunderstandinggenderdifferencesinsettingssuchasPakistan,wheresurveydatasuggestthatahighfractionofthefemalepopulationarelatentworkers.FemalelaborforceparticipationinPakistanwas21%in2020comparedtoamalelaborforceparticipationrateof78%(InternationalLabourOrganization,2019a,b);however,aquarterofwomenarenotworkingbutreporttheywouldliketoworkiftheycouldfindasuitablejob(FieldandVyborny,2016).Theresearchdesigninvolveddevelopment,piloting,andrefinementofahigh-frequencyjobmatchingservicethatlistsjobsanddeliversinformationtorespondentsaboutthemthroughtextmessageandacallcenter.JobTalashallowsustopreciselyobserveeachstepofjobsearchactivityonthesupplyanddemandside(FieldandVybornyj2022;Subramanian,2021;Fieldetal.,2023).Theplatformworksbymatchingeachjobseekertoopenvacanciesbasedonwhethertheysatisfyminimalcriteriasetbythefirmforthevacancy,andoccupationalpreferencessetbythejobseeker.Theplatformsendsinformationtothejobseekeraboutallthevacanciesthatmeetallcriteria(werefertosuchapairingthatsatisfiesallcriteriaasa“match")，andthejobseekercandecidewhethertoapplytoeachone.Thus,theplatformgenerateshigh-frequency,detaileddataonboththesupplyanddemandsidesofthelabormarketformillionsofpotentialjobmatchesbetweenfirmandrespondent.Theplatformdoesnothaveasearchfunction,whichmeansthatweobservethefullsetofvacanciesthatthejobseekersees,andthefullsetofcandidatessenttothefirm.Becauseweprovideinformationtobothsides,weobserveexactlythesameinformationasbothsidesofthemarketuptothepointofaninterview.Therepresentativerecruitmentofbothjobseekersandfirms(rareintheliterature(Kurekovaetal.,2015)andtherichadministrativedataobservedontheplatformprovideusuniqueleveragetohelppinpointthesupply-sideversusdemand-sideconstraintstowomen,sversusmen'sjobsearch.Thefourthchallengeinquantifyingsupplyversusdemandsideconstraintsisthateveniftheinitialsampleofjobseekersandfirmsisrepresentative,whenobservingdownstreamoutcomesinthejobsearchprocesssuchasinterviewsandhires,selectionproblemsariseagain,becausemaleandfemalecandidateswhochoosetoapplyforagivenvacancymaydiffersystematicallyfromeachotherandfromnon-applicants.Toaddressthischallenge,wecombinetheadministrativedatawithanincentivizedresumerating(IRR)experimentwhichweconductedwithfirmsintheJobTalashsample(Kessleretal.,2019).Weshowemployersontheplatformaseriesofpairsofcurriculumvitae(CV)andineachpairasktherespondenttoselecttheonethattheywouldbemostlikelytohire,withtheincentivethatthiscouldhelpinformtheapplicantpoolsenttothemthroughtheJobTaIashplatform.CVsforthisexercisewereconstructedusingtheactualjobapplicantdatafromtheJobTalashpool,makingthemarealisticrepresentationofthecandidatesthefirmmightseeontheplatform;werandomlyvariedthegenderoftheapplicantontheCVtoidentifyfirmpreferencesovergender,holdingconstantpotentialConfounderssuchasdifferencesinlevelsofeducationandexperiencebetweenmenandwomeninthepool,anddifferentialselectionintoapplication.Ourfirstkeyfindingisthatgendergapsinemploymentaregreaterinmagnitudethangendergapsinsearch.Thiscomplementstheliteratureshowingthatwomen'slaborsupplyiselastictotheintroductionofgoodjobsevenintheSouthAsiancontext,whereoverallfemaleemploymentratesarelow(Jensen,2012;HeathandMobarak,2015).Womenare89%lesslikelythanmentobeworkingatbaseline;however,theyareonly53%lesslikelytocompletesign-upfortheJobTalashplatform,aninvestmentoftimeinthetelephonebasedsign-upprocesswhichindicateswillingnesstosearch.Thegendergapinbothworkandwillingnesstosearchnarrowsaseducationlevelsrise.Athighereducationlevels,thegendergapincompletingthesignupprocessfallsby65%.Thesefindingssuggestthatmanywomen,particularlyeducatedwomen,are"latentworkers*'pointingtokeyconstraintsonthelabordemandside.Oursecondkeyfindingisthatforlesseducatedjobseekers,firmgendercriteria,anentirelydemand-sideconstraint,aremorebindingforwomenthanmen,andarealsoalargerconstraintthansupply-sidedecisions.Womeninoursettingare53%lesslikelythanmentosatisfytheexplicitgenderrequirementsforanygivenvacancy.Thesepatternspersistevenwhenwerestricttovacancieswheretheindividualmettheeducationandexperiencecriteria,andexpressedinterestintheoccupation:demand-sidecriteriaarethebindingconstraintonopportunitiesavailabletowomen.Infact,inthesetofvacancieswhereindividualssatisfiedallbasiccriteriaandwereeligibletoapply,womenapplyatahigherratethanmen,overall.Ourthirdkeyfindingisthatthedemandsidegapinquantityofjobopportunitiessubstantiallyclosesaseducationlevelsrise,whileonthesupplysidewomenbecomemoreselective.Thegendergapinsatisfyingthegendercriteriaforapositionshrinksby70%fortheminorityofwomenwithsecondaryeducationandeffectivelydisappearsforthethirdofwomenwithatertiaryeducation.WedocumentthisthroughboththeadministrativedataandtheIRRexperiment.Wefindthatfirms*gendercriteriaandtheeducationalrequirementsofthejobaremirrored;vacancieswith,4bluecollar*1characteristicssuchasmanuallaborandlongerandlateworkhoursaremorelikelytoexcludewomenandmorecommonamongjobswithloweducationrequirements,evenconditionalonindustryandoccupationfixedeffects.Additionally,firms'gendercriteriaandtheeducationleveltheyseektohirereflectexistinginfrastructureatthefirm:firmsthathaverestroomsoraseparateprayerspaceforwomenarebothmorelikelytobewillingtohirewomenandmorelikelytobehiringatahigheducationlevel.Thisconnectstoabroaderliteraturepointingtotheroleofon-wagecharacteristicsoffirmsandvacanciesingendergapsinthelabormarket(MasandPallais,2017;Floryetal.,2015;FieldandVyborny,2022;GoldinandKatz,2016;ChiplunkarandGoldberg,2021;Milleretal.,2022).Strikingly,amongthosewithatertiaryeducation,womenaremoreselectivethanmenintheirjobsearch.Atthishigheducationlevel,womenareslightlylesslikelythanmentohaveselectedtheoccupationofagivenvacancy,andareslightlylesslikelytoapplytoavacancy.Butthisislikelydrivenbydifferencesinthequalityofvacanciesbygender;indeed,wefindthatamongthosewithatertiaryeducation,womenaremorelikelythanmentoqualifyforthevacanciesatthelowestquintileofthesalarydistribution.Weadvancetheliteratureintwokeyways.First,ournovelcombinationofmatcheddataallowustoseparatelyquantifytheroleofdemandandsupplysidedecisionsinthegendergapinjobopportunities.Sec