Economic-returns-to-schooling-in-urban-China--1988

JournalofComparativeEconomics33(2005)730–752∗,YaohuiZhaob,AlbertParkc,XiaoqingSongcaChineseUniversityofHongKong,Shatin,NT,HongKongbChinaCenterforEconomicResearch,PekingUniversity,Beijing,Chinac238LorchHall,611TappanStreet,UniversityofMichigan,AnnArbor,MI48109-1220,USAReceived10December2003Availableonline22August2005Zhang,Junsen,Zhao,Yaohui,Park,Albert,andSong,Xiaoqing—Economicreturnstoschool-inginurbanChina,1988to2001ThisstudyprovidesestimatesofthereturnstoschoolinginurbanChinaoveranextendedperiodofeconomicreforms.Wefindadramaticincreaseinthereturnstoeducation,fromonly4.0percentperyearofschoolingin1988to10.2percentin2001.Mostoftheriseinthereturnstoeducationoccurredafter1992andreflectedanincreaseinthewagepremiumforhighereducation.Theriseisobservedwithingroupsdefinedbysex,workexperience,region,andownership,andisrobusttotheinclusionofdifferentcontrolvariables.Thetimingandpatternofchangingschoolingreturnssuggestthattheywereinfluencedstronglybyinstitutionalreformsinthelabormarketthatincreasedthedemandforskilledlabor.JournalofComparativeEconomics33(4)(2005)730–752.ChineseUniversityofHongKong,Shatin,NT,HongKong;ChinaCenterforEconomicResearch,PekingUniversity,Beijing,China;238LorchHall,611TappanStreet,UniversityofMichigan,AnnArbor,MI48109-1220,USA.2005AssociationforComparativeEconomicStudies.PublishedbyElsevierInc.Allrightsre-served.JELclassification:J24;J31;P23Keywords:Economicreturnstoschooling;UrbanChina*Correspondingauthor.E-mailaddress:[e-mail](J.Zhang).0147-5967/$–seefrontmatter2005AssociationforComparativeEconomicStudies.PublishedbyElsevierInc.Allrightsreserved.doi:10.1016/j.jce.2005.05.008J.Zhangetal./JournalofComparativeEconomics33(2005)730–7527311.IntroductionInrecentyears,Chinahasexperiencedarapidriseinincomeinequality.Parketal.(2004)andKnightandSong(2003)assertthatrisingreturnstoeducationseemtobeanimportantpartofthestory.1Thisriseinthereturnstoeducationdeservesspecialattention.Foratleastadecadefollowingtheeconomicreformsinitiatedintheearly1980s,scholarshavefoundunusuallylowreturnstoschoolinginChina,inboththestate-dominatedurbaneconomyandamongruralhouseholds,comparedwithothercountriesatasimilarstageofdevelopment.Ifconfirmed,thisnewdevelopmentrepresentsasignificantchangeinChina’seconomicandsocialsystemsandhasimportantimplicationsforthedynamicsofeconomicandhumandevelopment.Returnstoeducationprovideimportantinformationabouttheincentivesforhumancap-italaccumulation,theefficiencyofresourceallocation,andthedistributionalconsequencesofdifferencesinhumancapital.2Thispaperhasthreeobjectives.First,weevaluatepaststudiesofthereturnstoeducationinChinatoestablishthefactthatreturnstoeducationwerelowinthe1980sandearly1990s.Second,usingarepeatedcross-sectionaldataset,weestimatewageequationsanddemonstratearapidincreaseinthereturnstoeducationinurbanChinafrom1988to2001.Weverifytherobustnessofthistrendbyexaminingthesensitivityofourresultstovariousspecificationsoftheregressionmodelsandtheinclu-sionofdifferentcontrolvariables.Weexaminechangesinthereturnstoeducationwithingender,experience,ownershipandregionalgroups.Theseresultsprovideinsightsintothenatureoflabormarketchangesandpossiblecausesoftheincreaseinreturnstoschooling.Finally,weassesstheextenttowhichsupplyanddemandfactors,especiallyinstitutionalreforms,canexplaintheincreasingtrendinthereturnstoeducation.Inthenextsection,webrieflydiscusstheevolutionofChina’slabormarketinstitutions,especiallythoselikelytoaffectwagedetermination.InSection3,weprovideacompre-hensivereviewofpastresearchonthereturnstoeducationinurbanChina.InSection4,wedescribeourdataandpresentdescriptivestatisticsforoursample.EstimatesofthereturnstoschoolingarepresentedinSection5.Section6examinesthecausesoftheincreaseinthereturnstoschoolingandSection7concludeswithpolicyimplicationsforthelabormarketinChina.2.LabormarketinstitutionalbackgroundinurbanChinaState-directedlaborallocationwasanintegralpartofthesystemofeconomicplanninginstitutedinthemid-1950s.Withtheparamountgoalofrapidindustrialization,thegov-ernmentsetthewagesofworkersatrelativelylowlevelsinordertoreducelaborcosts,andlaborallocationdecisionswerecentralizedintothehandsofeconomicplanners.Lowwagesweremadepossiblebystate-subsidizedfoodpricesandstateprovisionofnon-wage1Usingthesamedataset,Parketal.(2004)findthat,from1988to2001,theGinicoefficientincreasedfrom0.242to0.372andtheTheilentropymeasureincreasedfrom0.101to0.235.2Heckman(2003)arguesthatChinaspendstoolittleonhumancapitalinvestmentandtoomuchonphysicalcapitalbasedonthedifferencesinthereturnstohumanandphysicalcapital.732J.Zhangetal./JournalofComparativeEconomics33(2005)730–752benefitstoworkersandtheirfamilies,suchashousing,childcare,medicalinsurance,andpensions.Undertheplanningsystem,allworkersandemployerswerematchedtojobsbygov-ernmentlaborbureaus.Lifetimeemploymentwasguaranteed,butlittlelabormobilitywaspermitted,eithergeographicallyoracrossoccupations.Fromthelate1950stothelate1970s,theBureauofLaborandPersonnelcentra