Provincial Labor Mobility Dynamics in China

Provincial Labor Mobility Dynamics in China

著者 胡 水文

journal or

publication title

人間社会環境研究

volume 15

page range 125‑140

year 2008‑03‑27

URL http://hdl.handle.net/2297/9837

論文

人間社会環境研究第15号２００８．３ 1２５

ProvincialLaborMobilityDynamicsinChina

社会環境科学研究科国際環境科学専攻 ＨｕＳｈｕｉｗｅｎ

中国における省間の労働力移動の原動力

胡水文

要旨

本稿は，1995年から2000年にわたって31の直轄市，省および自治体の間の労働力移動のデー タを用いて，中国における労働力移動の状況を考察した。最初中国の労働力移動についての歴 史，現状および形態を調べたうえで，省間の労働力移動にとっての障害および影響要因を分析 した。その結果，移動規制が時間とともに緩めたとわかった。さらに労働力移動は経済的要因 に大きく左右され，省間距離，失業率，収入およびGDPの成長率と高い関係があることを明ら

かにされた。

キーワード：労働力移動，距離，失業率

internalmotionlaw§weestmateamulti-regional modelthatrestsonthelogicthatwolkersmoveto mcreasetheirmcomesWefbcusonhowmobility decisionsrespondtothecostsofmigrationandlabor marketconditionsSpecifically;wewanttodetennine whetherrelaxedrestrictionsdecreasedthemportance ofmigrationcostsmthemobilitydecisionand enhancedresponsivenesstoeconomicfactors(suchas labormarketconditions）Suchanevolutioniscrucial tonprovelabormarketefficiencyandsustam

successfilleconomictransition

Theremainderofthepaperisorgamzedasfbllows Section2brieHypresemssomestylizedfactson intemalmigrationmChmaSection3mvestigates somefbrcesoflabormigrationBasedonmodelof NewEconomicGeography（nlbuchiandThisse，

2002)andCrozet,Z004Lwemvestigatesomefactors oflabormobilitybwhichwasperceivedasakeyfactor ofinequalityinChinamsection4Section5 1.Introduction

Intensegionallabormobilityisoneofthekeyissues mChma,stransitiontoamarketeconomyTbmove awayfiFomthemdustrialstructureinheritedfiomits plannedeconomy9Chmamustrelocateitsactivities Thelackofmatureandintegratedcapitalmarketsand thepersistenceofimpedmentstomter-provmcial tradeflowsmakelabormobnityevenmorevitalto achieveasuccessfUltransitiontoamarketeconoｍｙ ｈＣｈｍａ,intemallabormigrationisstronglyrestricted Smcethemidl980s,barrierstomigrationhavebeen relaxedmlmewiththedeepenmgofrefbrmsandthe sUbsequentrapidgrowthandvastdemandfbrlaborm urbanareasAuandHenderson(2002)howeverargue thatrestrictionsonmigrationremamtight

ThepresentarticleamstosmdyhowmigratoIy fbrcesevolvedbetweenl995and2000mChinalb analyzethedetennmantsoftheChineseworkers,

人間社会環境研究第15号２００８．３

1２６

Theturningpointfbramorerestrictivesystem cameafewmonthsafterthepromulgationofthel955 regulation,WhenfbodrationmgbeganinurbanareaS Therationingwasinresponsetofbodshortagesmthe marketduetothedepressionoffbodprices-anaction takenbythegovenunenttosupportthe mdustrializationpolicy（ｓｅｅＬｍｅｔａｌｌ９９６）Smce urbanhukouwasthepre-reqUisitefbrobtammg low-pricedrationedfiood，theurban-ruraldivision begantosurfaccAsthegovemnentmadeclearits mtentiontoguaranteetheprovisionoflow-costlivmg necessitiestourbanresidents，citiesbecamemore attractivetotheruralpeoplelnl957thegovenment

hastenedtheruralcollectivizationmovementmorder

toenfbrcethemandatoryprocurementofagricultural products，raiseagriculturalproductivityandsupport themdustrializationAsaresult，alargenumberof fmmersdesertedthelandandthecollectives，

migratmgtocitiestoseekjobslnordertosafeguard

themovementtowardscollectivizationasweUasto

controlthetotaldemandbyurbanresidentsfbrfbod，

theStandmgConunitteeoftheNPCissued

"RegulationsonHukouRegistrationmthePeople,s RepublicofChma?，mJanuaryl958Accordmgtothe regulation，“Whenapersonmigratesfieomanrural areatoacity，ｈｅ/shemustapplyfbrmovmg-outatthe hukouregistrationagencyoftheplaceofresidency，

andpresentcertificateofemploymentfiomanurban bureauoflabor，admissionletterfi?omaschool，or certificateofmovmg-mfiFomanurbanhukou registrationagency.,,(NPC1958）

Thisregulationeffectivelycentralizedthepowerof controllmgmigrationintothehandsofurbanhukou administrators-thepoliceofTicesBefbrethen,urban unitshadrecruitedworkerswithoutmvolvmgthe policeAlthoughthegovenunenthadattemptedto controlmigrationusmgothermeansmcludmgmoral persuasion，expulsionandrepatriation，and discouragmgdirectrecruitmentbyftlctories，these effbrtshadnotbeensuccessfUlingeneralEvidenceof concludes．

2.TheHistoryofMigrationControl

Thestrictenfbrcementoftherural-to-urban

migrationcontrolstartedmtheearlyl960sasaresult ofthｅｄｅｖａｓｔａｔｍｇＧｒｅａｔＦａｍｍｅＴｈｅｆｂｏｄｓｈortage continuedintothel970sandendedmthel980sTb

controlthemigration,ahouseholdregistrationsystem wasutilizedmconjunctionwiththepracticesofjob assignmentandrationmgoflivmgnecessitiesmurban areasTbgetherwithruralcollectivefarmmg，the systemwaseffectivemcontrollmgmigration HoweveLastheeconomicrefblmswerecaIriedoutin theearlyl980s,thesecontrollmgapparatuseslosttheir

effectiveness

Theprimaryapparatusofcontrollingpopulation mobilitymChmahasbeenthehouseholdregistration system，orhukousystemmChmeseThesystem registerseachpersonataspecificplace(usuallythe birthplace)andsmcel958govenⅡnentapprovalhas beenrequiredfbrchangesofregistrationHowever，

whenthehukousystemwasstartedafterthe Conununisttake-over,itwasnotmtendedtorestrict populationmobility・Ｔｈｅ“LmporaryRegulationon UrbanHukouManagement,,issuedbytheMmistryof PublicSecurity(ＭOPS）mJulyl951stipulatedthat thepurposeoftheregulationwasto“protectsocial ordeLsafeguardthesafetyandthefireedomof residenceandmigration,,（MOPS1951；emphasis addedbytheauthor)TheregulationmerelyreqUired visitorstoregisterwiththedestinationgovenⅡnentif thedurationofstayexceededthreedays，butthe systemgraduallybecamerestrictivehJunel955,the StateCouncilissueda“DirectiveConcernmgthe EstablishｍｅｎｔｏｆａＰennanentSystemofHukou Registration,，,reqUirmgallchangesofresidencetobe registeredmbothorigmanddestmationgovemments;

ｇovenⅡnentapprovalwasnecessaryonlyfbrlandlords andparoledcrnmals(StateCouncill955）

ProvincialLaborMobilityDynamicsinChina 1２７

theseattemptsarefbundmtheStateCouncil,ｓ

"DirectiveonDissuadmgPeasantsftomBlmdInflux mtoCities，，ml953,“JointDirectivetoControlBlmd InHuxofPeasantsintoCities,,bytheMmistlyof lnteriorandMinistryofLaborml954Chengand Seldon（1994）ａｎｄthe`DirectiveonPreventmgthe BlmdOutHowofRuralPopulation，,bytheState Councilml956StateCouncil(1956）Withthehukou systemmstitutedasthemechamsmofcontrol,thesize ofdesirablemigrationcouldpotentiaUybeeasily controlledmthefbnnofinunigrationquotas、

ＴｈｅｐｏｗｅｒｏｆcontrOlwasnotimnediately exerciｓｅｄｄｕｅｔｏｔｈｅｌａｒｇｅｄｅｍａｎｄｆｂｒｗｏｒｋｅrsm urbanftlctoriesduringtheGreatLeapForwardwhen workerswereactivelyrecruitedAccordmgtonational statistics，thetotalpopulationofcitiesandtｏｗｎｓ ｍｃｒｅａｓｅｄａｔｔｈｅｒａｔｅｏｆｌＯｍｉｌｌｉｏｎｐｅｏｐｌｅｐｅｒｙｅａｒ ｆｉＳｏｍｌ９５７ｔｏｌ９59,whichwasthefastestgrowthrate smcｅｌ９４９（SSB1995,ｐ376)HoweveLtheGreat LeapForwardfiailedmiserablyうwhichwasmarkedby thetragicdeaihfi?omstarvationofatleast30million

people,mostlymruralareas,ｆｆｏｍｌ９５９ｔｏｌ９６１Ｉｔｉｓ ｗｏｒｔｈｎｏｔｍｇｔｈａｔａｔｔｈｅｔｉｍｅｏｆｔｈｅＧｒｅａｔＦａｍｍｅ ｕｒｂａnbiashadalreadytakenitsfbotholdItisevident fiBomthefactthatthestarvationwasnotonlyalmost entirelyaruralpenomenon,butthegovemnenttook actiontosendbackover20millionruralpeoplhewho hadpreviouslybeenrecruitedfiFomruralareastothe countlyside(SSB1995,p376）

Tbwardtheendofthefhmme,themigrationcontrol wasstrictlyenfbrcedSegregationbetweenruraland urbanareasappearedanditwastoremammeffectfbr morethanZOyearS

Despitetherecentprolifbratioｎｏｆｒｅsearchon migrationmChma,thereisstillmuchconfilsionabout themagnimdeofmigrationThisisduemparttothe exlstenceofmanydifferentconceptsandtennsrelated tomigrationandtheffeqUentchangesofdefinitionm censusandcensus-typesurveysmChma(Duanand Sun2006)Inthefbllowmg,weshallfbcusonﾉﾉzｲｄｍｇ '4e"AC〃ａｎｄ９ｊα"〕ノノ花"ko〃－ｂｙｆｈｒｔｈｅｔｗｏｍｏｓｔ ｃｏｎｕｎｏｎｌｙｕｓｅｄmeasuresofmigrationmChina

mHblelFloatingpopulationandmigrants

l990census Z000census

ｖｏｌｕｍｅ (million）

Percent

population＊

ｖｏｌｕｍｅ (million）

Percent

population Floatingpopulation(liudongrenkou）

Inter-county

lnter-county＋intra-county Migrants(qianyirenkou）

Inter-county

lnter-county＋intra-county hter-countymigrants Pennanentmigrants (PercenO

Temporalymigrants (Percent）

Interprovmcial (Percent）

htra-provmcial (Percent）

22.62 1．９７ ７８．７５

１４４．３９

６．３４ １１６２

3５．３３ 3．３９ ７９．０５

１２１．２１

６．７４ １０．３３

１９１３

(541）

１６．２０ (459）

１１．５３ (326）

２３．８０ (674）

1．８４ ２０．２２

(25.6）

５８．８４

(744）

３２．３０ (409）

４６．７５ (591）

1．７２

1．５６ 5．０１

1.11 2．７５

2．２９ 3９８

Formigrants,proportionofpopulationaged5＋、

Sources:l990census1％sample;LiangandMa(2004);PopulaUonCensusODHce(2002)．

人間社会環境研究第１５号２００８．３ 128

Thisnumberisconsistentwithmostpublishedsources，

Whichestimatethatthefloatmgpopulationwasabout 30millionmtheeadyl980s，70-80millionmthe earlyandmid-1990s，andbetweenlOOmillionand l40millionmthelatel990s(BaiandSong2002:４；

Jiao2002;Solingerl999:１８;Wan200LZhong2000）

The20050ne-PercentPopulationSampleSurvey reportedfUrthermcreaseofthefloatmgpopulationto

l4735million(NationalBureauofStatistics2006）

L加伽"g犯Ｍｏ"(Hoatingpopulation）

Ｌｊ"cﾉblZgだ"ﾉto"，generaUytranslatedasthe

"floatmgpopulation,,,isauniqUeconceptmChma andistiedtothe加肋〃system(GoodkmdandWest 200Z）Whenanmdividualisnotlivmgathis/her

〃ﾉtoW1ocationthens/heisconsidered“floatmg”This conceptisbasedonthenotionthattheh"AC〃location iswhereonebelongsandthatanymigrationisnot consideredofficialandpennanentuntUthemigrant，ｓ

ｈｚ"to〃locationisalsomovedtohis/herdestmation

FloatmgpopulationisastockmeasureRegardless ofwhenactualmigrationocculred,apersoniscounted aspartofthefloatmgpopulationsolongashis/her usualplaceofresidenceisdiffbrentfiomthe加肋〃

location

hpractice,atemporalcriterionusuallyqualifiesthe definitionoffloatmgpopulation，andthecriterion

variesfieomonesourcetoanother､Thel990census

specifiedthataperｓｏｎｍｕｓｔｈａｖｅｌｅｆｔｔｈｅｈ"ﾉto〃

1ocationfbratleastayearbefbres/hewouldbe consideredaspartofthefloatmgpopulationlnthe 2000census,thecriterionwasshortenedtosixmonths Thespatialcriterionalsochangedlnthel990census，

thefloatmgpopulationmcludedpersonswhohad movedffomonecounty(orcounty-levelcityorurban district)toanothercounty;whereasmtheZOOOcensus，

thespatialcriterionchangedtosub-countyunits,ｉｃ.，

townships,townsandstreets

lnotherwords，thel990censuscounted ner-countyfloatmgpopulationonlyWhilethe2000 censuscountedbothinter=countyandmtra-county floatmgpopulationhter-countyHoatmgpopulation mcreasedftom2Z62million，ｏｒ197％ｏｆｔｈｅ ｐｏｐｕｌａｔｉｏｎｍｌ９９０ｔｏ７８７５ｍｉｌｌion,ｏｒ６３４％ofthe populationm2000(Tnble612）Ｃｌearly，mobilityhad mcreasedconsiderablybetweenl990andZOOO

Combmmginter-countyandintra-countycounts，

theZOOOcensusreportedatotalofl4439million floatmgpopulation，accountmgfbrl162％ofthe nation,spopulation(PopulationCensusOfficeZOO2）

Qjtzlq〕ｗｅ"AC〃(migrants）

Theclosestequivalentstothetelmsmigrationand migrantsmtheChmeselanguageararespectively〉

9ｊα'Mand9jα〃ｊだ"AC"､Unlikefloatmgpopulation, 9/α〃ｊ”"肋〃isameasureofflowhthel990census，

9ｍ〃ｊ形"Aomwasdefinedasmdividualsfiveyearsor older(1)whohadmovedfieomonecountytoanother withinthepastfiveyears,ａｎｄ(2)(a)whosM"AC〃

locationhadmovedtothel990placeofresidenceor (b)whohadleftthe加肋"locationfbrmorethanone yea1aThosemgroup（a）constitutedpermanent migrantsbecausetheirハ"肋〃locationhadmovedto themigrationdestmation;whilethosemgroup(b)are refelredtoastemporarymigrantsbecausetheirh"AC〃

locationdifferedfiomthemigrationdestmationhthe 2000census，thespatialandtemporalcriteriawele changedrespectivelytosub-county-levelunitsandsix

months

Itshouldbenotedthatthetelmsh"c/O"ｇａｎｄ９ｊａｍノノ areoftenusedmterchangeablymtheChmese literatureandmediaAtthesametime,somescholars considerthetwotennsmutuallyexclusiveTbmakeit evenmoreconfnsmg，thetennsfloatmgpopulation andtemporarymigrantsarealsooftenused interchangeablyUnderstandmgthefloatmg populationasastoｃｋｍｅａｓｕｒｅａｎｄ９ｊα'2〕ﾉﾉだ"肋〃ａｓａ Ｈｏｗｍｅａｓｕｒｅｉｓｋｅｙｔｏｄｉｓｔｍｇｕｉｓｈmgbetweenthese

tennS

Usingthe9jα"Ｗ形"肋"criteriadescribedabove, thel990and2000censusesdocumenteｄａｔｏｔａｌｏｆ

ProvincialLaborMobilityDynamicsinChina 129

InbleZInteILprovincialmigrationwithinandbetween１℃gions．

Proportionoftotalflows(％）

Destmation Eastern central Westem Sum

'９９０census Eastem Central Western Sum

2４４ １０．７ ３．７ ３８．８

21.0 ９．２ ４．１ ３４．３

1１．５ ６．３ 9.1 26.9

57.0 26.1 16.9 １００ ZOOOcensus

Eastern Central Westem Sum

1８．４ ３．８ 2.4 24.5

41.8 ４．０ 3.2 49.0

1８．２ ２．４ 5.8 26.5

７８．４ １０２ １Ｌ４ １００

Note:BecauseofdataUlnitations,Tibetisexcludedfmmthecomputation・BecauseChongqmgdidnotbecomeasepamte provincial-1evelunitunti11996,ｍtheanalysisitiscombinedwithSichuan、

SoulCes:StateStatisticalBureau(1992);NaUonalBureauofStatistics(2002)．

３５３３millionand7905mteFcountymigrants respectivelyうaccountmgfior339％ａｎｄ６７４％ofthe populationagedfiveandabovｅ(seeTnblel）These resultsagamsupporttheobservationthatmobilityhad mcreasedsignificantlybetweenthel985-1990and l995-2000periods(Fan2005a;Liang2001)Thesum ofinter-countyandintra-countymigrants,accordmgto the2000census,ｗａｓｌ２１２１ｍｉｌｌｉｏｎ,accountmgfbr

lO33％ｏｆｔｈｅ５＋population

Amongmter-countymigrants，respectively459％

and７４４％weretemporarymigrantsaccordmgtothe l990and2000censuses、T11evolumeofpennanent migrantshoverednear20millionbutthenumberof temporarymigrantsmcreasedby3､６ｔｍｅｓｆｉＦｏｍｌ６２０ ｍｉｌｌｉｏｎｔｏ５８８４ｍｉllionThedramaticsurgeof temporarymigrantsreflectsnotonlyh"AC〃refbnnbut alsomcreasedprommenceofmarketfbrcesm detennnngpopulationmovementsmChmaIn addition,thisresultmdicatesthatruralmigrants,ｗｈｏ constitutethebulkoftemporarymi厚ants,areplaymg anmcreasmglyimportantrolemshapmgChmese

cities

Breakmgmter-countymigrantsdowninto mter-provmcialandmtra-provmcialＣＯmponentssheds fhrtherlightmtomobilitychangeslnter-provmcial

migration，whichistypicallyoflongerdistances，

accountedfbr326％ａｎｄ４０９％respectivelyoftotal mter-countymigrationmthel990and2000censuses Thus,notonlyhadmobilitymcreased,butalsogreater proportionsofmigrantsmovedlongdistancesmthe l990sthanmthel980s（DuandGao2004Fan ZOO5b）

Tnble2showstheproportionsofmter-provmcial migrationattributabletomtra-regionaland mtelregionalflows，representedrespectivelyby diagonalandoffLdiagonalcells、Betweenthetwo censuses，mtra-regionalflowsdeclmedmrelative mportancewhilemtenFegionalproportionsmcreased Specifically〉ｔｈｅｓｕｍｏｆｏｆｆｎｉａｇonalproportions lncreasedfiFom5７３％ｔｏ７１８％・Ｔｈｉｓａｇａｍｓ叩ports theobservationthatmoremigrantstraveledlong

distancesmtheｌ９９０ｓｔｈａｎｍｔｈｅｌ９８０ｓ、Ofthesix offLdiagonalcells，onlytwo-central-to-easternand

westemto‐eastern-mcreasedbetweenthetwo

censuses,mdicatmganaccelerationofngrationflows ffomthetwonon-coastalregionstotheeasternregion TheflowfiFomthecentralregiontotheeasternregion isespeciallynoteworthy§mcreasmgffom210％ｔｏ

418％betweenthetwocensusesAllofthismdicates thatmter-provmcialmiHationisoverwhehmgly

人間社会環境研究第15号２００８．３

1３０

ＡｓＴｎｂｌｅ３ｓｈｏｗｎ，theprmarycharacteristicsof mter-provmcemigrationarethenetimmigrationto urbanareasandthenetemigrationofruralpopulation Ttlble614depictseachprovmcial-levelunitmand out-sharestothenationaltotalofmigrantsandtheir netmigrantsApparentlyうGuangdong，Shanghai，

BeijmgandJiangsu,etc,weretheprovmceswiththe largestsharesmtennsofinunigrants，Guangdong aloneaccommodatedl26millionmigrantsftomother provlnces，accountmgfbrmorethanone-tenthoｆｔｈｅ ftomimandtocoastalareasandthattheconcentration

ofmigrantsmtheeasternregion，whichisthemost urbanizedofthelhreeregions,ishighandmcreasmg TheprovmceofGuangdongalonereceived36１５％of allmter-provmcialmigrantsdurmgthel995-2000 periodRecentstudiesalsoconfinnthatprovmcialnet migrationvolumesandratesmcreasedbetweenthe twocensuses；ｔｈａｔｉｓ，sendmgprovinceslostmore migrantsandreceivmgprovlncesgamedmore migrantsmthel990sthanmthel980s(Fan2005）

Inble3IntelLprovmcerural-urbanmigrationinthe4thCensus^●

Netimmigrantstourbanareas Netimmigrationstoruralareas Netnngrations^■

Regions _{million ％ miIlion ％}

２６５１９４３４１５２４１５６３１Ｊ６０２０３１１３２０００００２－０宮０００００』一一。’

２１３２２１６５２６０１０５００００１２０２１０００●●■●●●●●巴●００００００００００００】』。□一『ロ

２６９７６４５９９１２２２８２６２２６６０３２６４０５２６２１１４２１１１４２３０ ２６１７１８７２２７４０４１００５１１２４３０３０００００００００００１０００

４７８５３７１７８１４１４６５１０２５１００００００００●の●Ｂ●●●Ｂ●●●ひ●●０００００００００１００００

Beijmg Tianjin Shanxi Liaoning Shanghai Jiangsu Fujian Shandong Hubei Guangdong Hainan Qinghai Ningxia

xmjiang 餌』』『一』』。’］。一一一００００－００００００００００ ２０１２０２０２３４９１１１１ ７８４２４７９７２６３７２４１ 』』』】□一口』一『一』。一一０００１１００００１１００００ ●●●●▲●●●巳●●●●●●５５９２３５３３６２０６３５６ ６９８１９７１８５８９７８３５ ４４２１７６３６５４９６６５２１ＯＯ１７６０２６２３８６２６０００１００００１１’一

６３２２１７２６７１８５９９８１００００００１０００００００００００００００００００００ １５２４３２７１５５５２３５３１０１２０００１２４８１０００００００一一０００００００００口』一』一一・一一』『『』

ａ

ｏ２）２）血、⑪即ｇ ｏ

帥加川鼬輕．、螂鎚”》岬》、、１

池

Note:NSBthepopulation4uⅢCensus

ProvincialLaborMobilityDynamicsinChina 1３１

nationaltotalofmter-provmcialmigrantslnShenzhen，

oneofChma,sfburEconomicSpecificZonesml980s，

thel990CensusrecordedthattherewasatotaloflO2 millionfloatmgpopulation,orl6tmestheirresidents (O63million）Sichuanwasthemostimportantsendｅｒ ｗｉｔｈｌ３２ｍｉｌｌｉｏｎｅｍｉｇｒａｎｔｓｏｒｌ２％ofthenational totallntennsofnetmigFation,mostofthecoastal，

easternprovmces（suchasGuangdong，Shanghai，

BeijmgandJiangsu）gamedfTomprovmceslike SichuanandGuangpdmthecentralandwestern

reglons（ＫＷＣｈａｎ，1999b）Ｃｈａｎ（ＫＷＣｈａｎ，

1999a)mappedthe301argestlnter-provmcialflowsof non-migrantsbasedｏｎｔｈｅｌ９９０Ｃｅｎｓｕｓｄａｔａａｎｄｌ％

sampleltrevealsthatthemter-provmcialnon-h"ﾉto〃

migrationHowwereprmarilytowardtheeasternand southerncoastGuangdong，ShangdlaiandBeUmg

becametheconcentratedcentersofrural-urban mlgratlon

Wehavetousemdirectapproachestoestimatethe mter-provmcialmigrationmtheperiodofl990-2000 Inble4Estimatedinte限provincialmigrantsandInigrationgmowthrateinl990-2000

Estimated Annual

Migration

GrowthＲａｔｅ

Estimated Migration Without

Hukou Average

Annual GrowthRate

Average Natural GrowthRate

Population byHukou in2000 Population

in2000 Population inl990 Regions

Beljmg Tianjin Shanxi

Liaoning Shanghai Jiangsu FUjian Shandong

Hubei

Guangdong

Hainan Qinghai Ningxia Xinjiang

２１７８４８１９８２７８２・っ８０９３７３７７２４８１６２３０２２６４４００６７５５９１１３４１７３９６８１ ２９６６４６５９７３６６６６８７７４３００３９８５４６１０８８９３７０４３２６４４５１２３１６３８５６１

２．４８ １３１ １３８ ０．７２ ２．２９ １０４ １４５ ０．７３ Ｌ11 324 1.84 1.51 １．９ ２．４２

１４９５３８５８８２３４６２４６２６１８１０９３５５５５００１０００１０１１１１１ ７７９７６６３７３２１３４９０６００１１００１９３０３０２０００２０００１０００口』

1１１４ ９１９ ３１９６ 4135 1322 7069 3305 8975 5936 7499 ７６１ ４８０ ５５４ １７９２

８２１３２９６４２３６８８３６８００５６６０９４２３３２１１３３１１１１１

ａｏ２）ｇ）、ｎｏ、１且）

ｂｅｎ艸争Ⅷ、”鼬緬『『》ｍ

Ⅷ麺伽 伽

６６７１ ２３０１ ２６２７ ３６９８ ４５０１ ６２７８ ４１６４ ９５２７ ６５１５ 4724 11499 ３６７７ ４０７７ ２５１ ３５７２ ２５３４

３５１９６２４１５５０２１１３８７７０卓７９２７７３８５１１３２１１２－２－２’１２』。

Note:１)ＡＮ′erageannualpopulationgrowthlatciscalculatedbythefbllowiingfbnnula:に＝(Pqp2000/Pqpl999）－１(Umts:10,000_{ａｎｄ％）}

２)ＡＮ/emgenaturalgrowthrate＝(naturalgrowthlateml990＋natulalgrowthrateinl999)/２ Sources:ＳＳＢ,ChinaPopulationStaUsticsYearbook2000and2001．

人間社会環境研究第15号２００８．３

1３２

duetotheunavailabilityofthedetailedZOOOCensus onmigrationFirstly）theaverageannualpopulation growthrateofeachprovmceiscalculatedbythe geometricalaverageapproachandtheaverageannual mi厚ationgrowthrateofeachprovmceisestnatedby reducmgitsaveragenaturalgrowthrateSecond】y，the sizeofmigrantswithoutlocaM"わ〃fbreachprovince isestnatedbyreducmgitstotalpopulationmthe 200OCensustoitstotalpopulationof〃"んo〃regｉｓｔｌｙａｔ ｔｈｅｅｎｄｏｆ２０００ｂｅｃａｕｓｅｔｈｅｆbnnerfiguremcluded thosemigrantｓｗｈｏｈａｄｌｉｖｅｄａｔｔｈｅｄｉｓｔｒｉｃｔｏｒ ｔｏｗnshipfbrmorethanahalfyearbuttheirresident hz'AC〃wereregisteredatotherdistrictsortownships Shanghai，BeijmgandGuangdongarethema]or destmationprovmcesofrural-urbanmigrationTheir averageannualｇｒｏｗｔｈｒａｔｅｓａｎｄｓｉｚｅｏｆｎｏｎ－ｈ"肋〃

血grationml990-2000arethehighestmChina，

respectivelyShanghai（352millionand217％)，

Beijmg（Z68millionandZO7)，andGuangdong (１１４３millionandl9Z)Themajorsourceprovmces ofrural-urbanmigrationare:Anhui(-07％and2g2 mUlionofnetemigrants),Guizhou(-062％andl5Z millionofnetemigants),Henan（-051％and271 million)，GuangXi（-O47and235million)，Hunan (-O46andO75million),andSichuan(‐O28andO8 million）

Insumnary9thespatialpattemofmter-provmcial rural-urbanmigrationmChinahasbeenprmarily ftomthecentralandwestregionstotheeast(coastal）

regionandthePearlRiverDelta，theYangtzeRiver Delta,andBeijmghavebecomethemostimportant centerssmcethｅｌ９８０ｓ，whichisconsistentwiththe regionaldisparityofeconomicdevelopmentand householdincome、Thismdicatesthatnon-h"肋〃

rural-urbanmigrationmChmamtherefbnnerais primarilyeconomic-motivatedormarket-drivenThe mainobjectivefbrrurallaborerstocrosshundredsor eventhousandskilometerstocitiesmotherprovmces istoseekemploymentopportunitiesandbenefitfiFom largewagedifferentialS

3.DrMngfbrcesoflabormigration

Therestrictivesystemofhouseholdregistration becamethemstitutionalbasisfbrurbanbias､Whenthe govemmenttookontheresponsibilityofprovidingfbr urbanresidents，thesystemitselfautomatically acqUiredanelementofexclusiveneｓｓｂｅｃａｕｓｅｗｈｅｎ ｔｈｅｓｉｚｅｏｆｔｈｅｐｉｅｉｓfixed,tｈｅｏｎｌｙｗａｙtomaximize eachperson,sshareistorestrictthenumberofpeople eligiblefbrdistributionTheexclusivenessfiJrther madeitpoｓｓｉｂｌｅｔｏｅｎｌａｒｇｅｔｈｅｐｉｅｂｙｕｓｍｇｔｈｅ distributivepowerofthegovenⅡnentOvertme,more andmorebenefitswereprovidedtourbanresidents，

mcludmgfiFeeorcheaphousmg,medicalcare,pension，

livmgnecessitieS

Asaresultoftheurbanbiasmstitutionalizedby rural-urbansegregation，themcomegapwidened ShengandSun（1994）estmatedthatifthe govemmentsubsidiesofhousmg,healthcareandfbod aremcluded,theratioofurbantoruralmcomeswaｓ ３０９ｍｌ９８０Ｔｈｅｈｉｇｈｅｒｍｃｏｍｅｄｉｆｈｅｎｃe strengthenedthemcentivefbrmralpeopletomoveto

cities

HowevelBasurbanitesenjoyedmoreandmore govenmentsubsidies，betterprotectionandhigher lncomes，theyalsocametoperceivethemselvesas bemgsuperiortoruralpeopleThisbecamethe historicalandpsychologicalbasisfbrthe discriminationtowardruralpeopleLater，asrural peoplebegantomvadetheir‘Gtellitolies,,via spontaneousmigration,thediscrimmatio、wouldtum mtohostility

Themassivelabormigrationafterl980canbe broadlyattributedtothefbllowmgfactors:thefierce pushingfbrcesffomruralareatotransfbnnsurplus rurallaborunleashedbythedecollectivization programml97athestrongpullmgfbrcesthat resultedfiomrapidindustrializationandffoma contmumglargemcomedisparitybetweenruraland urbanresidentsaswellasinlandandcoastalregions；

ProvincialLaborMobilityDynamicsinChina １３３

ａｎｄｔｈｅｒｅｆｂｎｎａｎｄｉｍｐｒｏｖｅｍｅｎｔｏｆｓｕpportmg mstitutionalaIrangementsonmigrationcontrol，fbod andlabormarket,etc

Theadoptionofthehouseholdresponsibilitysystem (HRS)mruralareasmtheearlylg80ssolvedChma,s fbodsholtageproblemAsshownmLm(1992),gram outputmcreasedby40-50percentbetweenl979and l983、Withfbodsholtagealleviated,thegovemnent startedtorelaxsomeaspectsofmigrationbuttriedto maintamoverallcontrolInthelatel970sandearly l980s,itallowedurbanyouthsthatweresenttolural areasdurmgtheCulturalRevolutiontoretumtotheir citiesoforigmltalsoallowedsomespousesofurban workerstomovetocitiesThemcreasedsupplyof fbodgavensetotheemelgenceandexpansionofthe fiFeemarketfbrgrainmurbanareas，Rationcoupons wereabandonedmearlyl990s，butfbodhadbeen availableontheffeemarketlongbefbrethatThe marketizationoffbodmadeitpossiblefbrtemporary migrantstolivemcities

lntheple-refbrmperiod，ruralpeoplewere membersofvarious“People，sconmnunes，,，which werecollectiveeconomicorgamzationsThey olganizedeconomicactivitiesaccordingto govemnentplanningandweregenerallynot encouragedtoengagemnon-agriculturalproductio、

thatwasassignedtotheurbanregistered

GGnon-agriculturalpopulation，，、R1Jrallaborerswere restrictedffommovingtourbanareasbytightcontrols ofemploymentopportunities，residenceregistration andrationｍｇｏｆｇｒａｍａｎｄｏｔｈｅｒｐｒｏｄｕｃｔｓＣｈｍａ,s agriculturalsectorandruralareａｗｅｒｅｕｓｅdtofUnction asareservoirfiorresidualrurallabors

WiththerapidexpansionofChma，srurallabor fbrcesandthecontmumgdecreaseofcultivatedland，

alargeportionofrurallaborhasbecome underemployedorsurplusmcomparisontoavailable agriculturalresourcesAsElble5shown，cultivated landperrurallaborerdeclmedsharpｌｙｆｏｍＯ６２ｈａｍ

ｌ９６０ｔｏＯ３１ｈａｍｌ９８０ａｎdfUrthertoO21haml995・

Accordmgtoasamplesurveyofllmillionrural householdsm2,468townshipsandtownsm773 counties，SSBestimatedthattherewere60million

surplusrurallaborersml982(SSB,1988),andsome l20-150millionmthemiddleofthel990s(Fan,ＸＹ，

1997)Zhangcalculatedthattherewereabout70 millionsurpluslurallaboLaccountmgfbrl8％oftotal rurallaborfbrcemthebeginnmgofthel980s;about l30millionsurplusnⅢ.allaboEaccountmgfbr28％of totalrurallaborfbrcemthebeginｎｍｇｏｆｌ９９０ｓ (Zhang,ＣＹ,1994)Fuestimatedtheshareofsurplus nuallaborwas31％inl990s(Hu,ＷＬ,1999）

hble5AveragecultivatedlandperrurallaborinChinainl955-ZOO4 Areaofcultivatedland

（nllionha）

Rurallaborerfbrce

（million） Areapercapita

（ha/person）

Ｙｅａｒ

１９５５ １９６０ １９６５ １９７０ １９７５ １９８０ １９８５ １９90 1995 2000 2004

1１０１６ １０４．８６ １０３．５９ １０１１３ ９９．７１ ９９．３１ 96.85 95.67 94.97 128.24 122.44

１８６．９２ １７０．１９ ２３３．９８ ２７８１４ ２９４．５９ ３１８．３６ ３７０．６５ 420.1 450.42 47962 49695

９２４６４１６３１７５５６４３３３２２２２２●■●ｅ■●●●●●●０００００００００００

Soume:ＳＳＢ,2000,2005,NewChina1sAgricultureStatisticalDatafbr50years

人間社会環境研究第15号２００８．３ 1３４

Withthemtroductionofthe‘`ruralhousehold

responsibilitysystem”andtheruraldecollectivization m1978,itbecameessentialtotransfbnnsurplusrural laborfiFomagriculturetonona2ﾛｺﾞculturalsectors Generallyspeakmg，therewerethreemodesfbr transfbnnmgsurplusrurallabormChma：力"ko〃

rural-urbanmigration;“leavmgthelandbutnotthe villages,,（ノノ、６〃ノノ川'Zg）orruralurbamzation locally;and"leavmgboththelandandtheviUages"(ﾉﾉ ﾉ"ｙｏ〃ﾉﾉ川"g)ornon-hukourural-urbanmigration Thesecondmode,leavmgthelandbutnotthevillages，

ｄｏｍｍａｔｅｄｍｔｈｅｌ９８０ｓａｎｄｍａｄｅａｖｅｒｙｉmportant contributiontothetransfbnnationofsurplusrural

laborfbrce

Despitetherestrictivepoliciestoward rural-to-urbanmigration，ｓｍｃｅｔｈｅｏｌｄｓｙｓｔｅｍｏｆ ｃｏｎｔｒｏｌｌｍｇｍｏbilityhadbecomelesseffective，rural peoplewerestillabletorespondtonewemployment opportunitiesinurbanareasOnesuchopportunity becameavailablemtheconstructionmdustrySmce theearly-tomid-1980s,urbanstateownedenterpnses weregivenfinancialautonomy，whichfUeleda constructionboommurbanareasSmceconstruction jobsareconsideredonerousanddangerous,itwas difficulttofindwillingworkersamongurban residems;ruralworkersreadilyfilledthevacancies AccordmgtoBanisterandTtlylor(1989),therewere aheady5millionruralconstructionworkersmurban

areasml988

Anotherfactorthatpromptedruralworkerstoseek employmentmcitiesistheslowdownofjobcreation mruralentelprisesmthelatel980sBetweenl984 andl98aruralentelpnsesprovidedanaverageof lO8millionnewjobsperyearHowever,thegrowth rateofnon-ftlrmemploymentdecreasedtoabouthalf thatnumberfbrl988-1994Smcethestatisticsmclude ruralpeopleworkmgmbothruralandurbannon-farm jobs,theslowdownofemploymentcreationmrural enterprisesshouldbemoreｔｈａｎｗｈａｔｔｈｅｄａｔａｓｈｏｗ Ｔｈｅｏｔｈｅｒｍａｍreasonswillbediscussedmnext

sectlon

Anotherimpoltantdrivmgfbrceofrural-urban migrationisthehugerural-urbandisparityand inland-coastaldisparityofmcomeandeconomic developmentlevelApartftomtheprivnegesof enjoyingthevariousstate-subsidizedwelhlresonfbod supply，education,employment,medicalserviceMtc，

urbanresidentshadamuchhighermcomethantheir ruralcounterpartsmthepre-refbnnerahl978，the PerCapitaAnnualDisposablehcｏｍｅｏｆＵｒｂａｎ Ｈｏｕｓｅｈｏｌｄｓｗａｓ２６ｔｎｅｓthePerCapitaAnnualNet lncomeofRuralHouseholdsatcomparableprice・As Chapter3Figure32shows，thisratiogradually declinedtothelessthanllml985,butbegantorise agamthereafterltwas22ml994andZOOOThis demonstratesacontmumglargemcomedisparity betweentheruralandtheurbanhouｓｅｈｏｌｄｓｍＣｈｍａ

Ｔｈｅｒｅｉｓａｌｓｏａｂｉｇｒｅｇionaldisparityofeconomic develoｐｍｅｎｔｍＣｈｍａｍＺＯＯＯ・ＰｅｒCapitaGDPat provmcial-levelreducessharplyftomthecoastal regionstothemiddleandwesternmandregions

Accordmgtothe4thNationalCensusml990，a

totalof34millionmi印antscrossmgcounty-level unitswererecordedovertheperviousfive-yearperiod (1985-1990),about70％or24millionareidentifiedto bemtra-provmcialmigrantswhiletheother３０％ｏｒlO millionbelongtolnter-provmcemigrationThe dominanceofmtra-provmcemigrationisconfirmed bySSB,ssamplesurveyonrurallabormigrationThe proportionofruralsurpluslaboremployedwithintheir ownprovmcetothetotaltransfbnnedmcreasedfmm ６８％ｉｎｌ９９８ｔｏ７９％ml999Further,accordingtothe sthCensus，ａｍｏｎｇｔｈｅｔｏｔａｌｌ２ＬＯ７ｍｉｌｌｉｏｎ ｎｍｉgrantswithouth"AC〃fbrmorethan6months，

６５％or7865millionmigratewithintheiｒｏｗｎ

ｐｒｏｖｌｎｃｅ

4.TheoreticalEvidence：Modelofmigration Therearetwobasicstrandsofmigrationliterature．

ProvincialLaborMobilityDynamicsinChina 1３５

ＴｈｅｆｉｒｓｔｏｎｅｉｓｔｈｅＬｅｗｉｓｆｔａｍｅｗｏｒｋＴｈｅＬｅｗｉｓ

（1954）model（whichwaslaterextendedand fbrmaliｚｅｄｂｙＲａｍｓａｎｄＦｅｉ（1961)）becausethe dommantparadigmmthelａｔｅｒｌ９５０ｓａｎｄｌｎｔｈｅｌ９６０ｓ ＴｈｅＬｅｗｉｓffameworkconsideredmtemalmigrationas anaturalanddesirableprocessmwhichsurpluslabor wasgraduallyshiftedfiFomtheagriculturetoprovide neededmanpowerfbrurbanSector・Theprocesswas deemedsociallybeneficialbecausehumanresources werebemgshiftedfiFomthesectorwheretheirsocial margmalproductwasoftenassumedtobevirtually zerotothesectorwherethismargmalproduct significantlypositiveandpositiveandrapidlygrowmg TheLewismodeldoesnotexplicitlyconsiderutility (orexpectedearnmgs)maximizationofagentsThus，

withitsadhocstructure，itf1ailstothrowmuchlight onthemigrationdecisionandthefimctionmgoflabor marketsetcmaddition,ａｓＲａｙ（1998,ｐ367）states，

theLewismodelnplicitlyassumesthatfamilyfiarms aretaxedaslaboriswithdrawnfiFomagriculture Keepmgagriculturalpercapitamcomeｃｏｎｓｔａｎｔａｎｄ ａｌｌｏｗｍｇｔｈｅｓ叩plycurveoflabortomdustlyto remamperfectlyelastic

TheHanis-Tbdarｏ（H-T）fiFamework（ie，Tbdaro (1969)andHalTis-Tbdaro（1970)modelslaswellas subsequentworkthatextendedthesemodels)，

elaboratedmanyfeaturesofmigrationthatwent unnoticedwithinthegeneralandadhocstructureof theLewisfmmeworkTheH-Tftamework(whichis clearimprovementovertheLewisfiameworkmmany ways）becausethemamstreamparadigmininternal migrationliteraturesmcetheearlyl970sltpostulated thatmigrationproceedsmresponsetourban-rural difTerencesmtheexpectedeammgsmurbanrather andtheactualearnmgsinrural、hthatfiFamework，

givenapolitically-detennmedminimumurbanwage thatexceedstheagriculturalwage,theprobabilityof gettmgafbnnaljoｂｉｓｄｅｔｅｒｍｍｅｄｂｙｔｈｅｒａｔioof availablefbnnalsectorjobstojobseekers Consequentlyうmequilibrium，theexpectedfblmal

sectorwage(1ｅ,wagetmestheprobabilityoffindmg suchajob)isequaltotheruralwage

ltisnotdifficulttoseethatthecoreassumptionof theH-Tffamework，‘thepresenceofanexogenous urban-Iuralmcomegap,，makesthefiamework problematicbothattheempiricalandtheoretical levelsConsequently;therehavebeensomenotable attemptstoprovideanendogenousexplanationfbrthis urban-ruralmcomegapＴｈｅｆｉｒｓｔａｎｄｍｏｓｔｎｏｔａｂｌｅ ｏｎｅｉｓｂｙＳtiglitz（1974）Itsbasicideawasto mtroducelabortumoveLwhichiscostlytourban employersThepaceofwolkintheurbansectormay slowdownduetotumover・Supposethatafarm payingahigherwagecanfhcealowlabortumover rataonecanthenexplamwhyurbanemployersmay notlowerthewagestheypaydespitethepIesenceof unemploymentThepapersthatendogemzedthe urban-ruralmｃｏｍｅｇａｐｏｆｔｈｅＨ－Ｔｆｉａｍｅｗｏｒｋ，

howeveLmamtamedotherexogenousfeaturesofH-T fmmeworkFormstance，pricesofagriculturaland manuftlｃｔｕｒｍｇｇｏｏｄｓａｒｅｔａｋｅｎａｓｕｎｉｔｙｍａｌｌofthe abovepapers;thisisaconseqUenceoftheirsmallｏpen economyassumptionThisassumption，ontheother hand,isnotrealisticconsidermgthatmanycountries havebeenexperiencmgsignificantmtemalmigration levelssmcel950s,andmanyofthesecountrieswere notexactlyopeneconomiesespeciallydurmgthatera，

replymgheavilyonimportsubstitutionpolices ThemodelofmiHationmthispaperfbllowsthatof nbuchiandThisse（2002）andCrozet（2004）We consideramobileworkerkfTomprovmcejandhis locationdecisionamongRprovmces（mcludmgj）

Migrantschoosetheirdestmationthroughthe comparisonoftheperceivedqualityoflifemthe variouslocationsMigrationdecisionsarebasｅｄｏｎ：

migrationcosts，expectedrealmcome，andthe probabilityoffindmgajobmthedestmationThis macro-economicsimplificationissensiblemｏｕｒcase，

asourfbcusisspecificallyonthemvestigationofthe responsivenessofChmesemigrationdynamicsto

人間社会環境研究第15号２００８．３ 1３６

bordersarefbundtohavelargetradeimpedmgeffects

mChma

Individualｋｗｉｌｌｃｈｏｏｓｅｔｏｌｏｃａｔｅｍｐｒｏｖｍｃｅｉｉｆ

嗽〉J')|i,Ｗ≠jWilhconvementassumptionson thedistributionofgj,theprobabilityofchoosmg provmceiisgivenbythelogitfUnction

P(Ｍｊ#)=e端／ZiLle端 ^（２）

Theexpectedmigrationflowfromprovmcejtoiis mjgri,＝L/Ｐ(Ｍ,),withLjthepopulationofprovmcej SmilarlyぅthetotaloutHowffomjisL/[１－Ｐ(Ｍ１)L theshareofemigrantsfiFomregionjmovmgtolis：

mJgr/，ｅ噺 _（３）

Ｚ,鬚’蝿rノザエ,e'ｿＬｅ噺

costsandothermacro-economicfactorslndeed，our mtentionisnottoidentifythevariousmicro-economlc detennmantsofmigrationFollowmgCrozet(2004)，

weconsiderthatmigrationdecisionsaredesignedto maxmizethefbllowmgobjectivefUnction：

"》=γ;’十s｝=1,1Ｍ(｡川bF1D-Al十s｝ｉ[1,R］（１）

wherepiistheprobabilityoffindmgajobm destmationlocationi,のiistherealincomemi,[qＭ１

＋6局/＋cＪ１/)]~lco江espondstomigrationcostsand g'isastochasticcomponentcapturin9ｋ,spersonal perception,softhecharacteristicsofprovmceiThis specificationallowsustomvestigatethevariouscost componentsthatdiscouragemigrationwithdij denotmgthedistancebetweenhomeandhost locations,Fijbemgadumnyvariableequaltooneif theorigmanddestmationprovmcesshareacommon border,andzerootherwiseProportionaldistancecosts typicallymcludethephysicalcostsofmovmgthat mcreasewithdistance，reductionsmthequahtyand amountofinfbnnationasdistancemcreases，and networksofcontactsandsupportbasedonpast migrationflows（Helliwell，1998）Wetestwhether thesemigrationcostsaremitigatedifthetwolocations belongtoneighbormgprovmces，thatisiffinancial andmoralcostswhichaffectmigrants，mcomeand satisfactionarelowermthecasewherethedestmation

provmcesharesacomnonborderwiththeprovmceof origmMoreoverweallowfbrextracoststｏａｐｐｌｙｔｏ ｍｉｇｒａｎｔｓｔｈａｔｍｏｖｅｏｕｔｏｆtheirprovmce，thereby engagmgminter-provmcialratherthanin lntra-provmcialmigrationsWeanticipatethatthe extracostsofmovmgoutofandfilrtherawayfiFomthe departurelocation，ｂａｎｄｃａｒｅｔｈｕｓｅｘｐｅｃｔｅｄｔｏｂｅ ｓｔｒｉｃｔｌｙｎegativecoefficients

Thesehypothesesrelatetotheinferencethatmany ofthenetworksofknowledgamstitutionsandshared valuesthattendtofacilitatetradeareatleastpartly detennmedbybordersandthatnetworkssupportmg migrationcanbeexpectedtobesmilar、Provincial

UsmgthedefinitionofVijk，thissharecanbe

C

rewrlttenas：

川G,＝'､の,+'Ｍ+'､レル凪)1入+aノ（４）

'､乙篝ｊ"恥，

ｗｉｔＭ=-'n(ヱゾーe1di）

Equation（４）fbnnalizesthetradeofTpotential mlgrantsareconfmntedwithwhentheychooseamong severalpotentialdestmationsTheshareofmigrants fiomagivenprovmcewhodecidetomovetoprovlnce iisexplamedbytheexpectedmcomemtheprovmce，

whichisapositivefUnctionoftherealmcomeandof theprobabilityofbemgemployedthereandanegatlve fUnctionofmobilitycosts

Despiteitsapparentcomplexity§EqUation（４）

correspondstoarathersimplegravity-typeequationof migration．

h鏡〒β,hMhwAMM小門(５）

Meanwhilaasmotionedmsection3,theprovmcial growth(gi)andtheshareofruralpopulation(Pri)are alsomportantfactorstoinfluencethechoiceof migrantsIfweaddedthemtoequation(1),itcould

ProvincialLaborMobilityDynamicsinChina 1３７

expressedbythefbllowmgeqUation： migratoryflowsthatoccurwithmsUb-periods

l995-2000

ourestimationsrenectthecostsofmigration throughheemdicatorsThefirstmdicator colTespondstothemigratorydistanceandreHectsthe costsproportionaltothedistancebetweenthe departureanddestmationlocations

ThesecondmdicatorｉｓａｄｕｎⅡnyvariablethat equalsonewhentheｏｒｉｇｍａｎｄｄestmationprovmces shareaconⅡnonborderltcapturestheextracostof movmgfhrtherawayfiFomthedepartureprovmcethan toaneighboringprovmce

Thethirdindicatorrelatestotheliteratureonborder efTectsfbrmigrationOurdatasetallowsusto measureadditionalcoststhatapplytomigrantswhｏ ｍｏｖｅｏｕｔｏｆｔｈｅｉｒｐｒｏｖｍｃｅｓｍｃｅｗｅａreablecompare thefluidityofinter-provmcia1migrationwiththatof mtra-provmcialmigrationThebordereffect,captured byadununyvariableequaltoonefbrmigrations betweendifferentprovmcesandzerofbr mtra-provmcialmigrations，answersthefbllowing question：fbreveryimmigrantmaChineseprovmce whocamefiFomanotherdistrictoftheprovmcewithm thelast5years，howmanyinunigrantscamefiFom anotherprovmceofsimilareconomicconditionsand migrationcosts､Thisvariablemeasurestheextracost ofengagmgmanmter-provmcialmigrationm comparisontoanintra-provmcialmigFation

WeestmateeqUation(5)and(7)onapaneldataof migratolyflowsbetweenandwithm31Chmese provmceswithfixedeffectsrelativetodeparture regionstocontrolfbrtheirspecificfeatures

hnlble6，ＴｈｅｆＩｒｓｔｔｗｏｃｏｌｕｍｎｓ，theresultsof SandraProncet（2006)，reportestmationson intePprovmcialmigrationflows，ml985-l990and l990-1995respectivelylnthefbllowmgcolumns，

mter-provmcialmigrationsml995-2000arealso

mtroduced

Distanceenterswiththeexpectednegativesign：

migrationHowsdecreasesignificantlywiththe

砺ﾀｰﾂﾞ;+91=l､し`gwBルル凪)-ルド（６）

andthewayうtheequation(5)couldbechangedto：

Ⅲ翁=βＭβ鬘hwAhw

β４Ｍ,＋β5F},＋βＧＢ,＋αｊ＋s,（７）

WeusedataftomthefburUhNationalPopulation Censusof2000NationalPopulationSurveybasedon asamplesizeofl/１００

Thedatareporttheplaceofresidenceatthetimeof thesurvey（2000）aswellastheanswerofthe respondentstothequestioTWhatwasyourplaceof residenceatthelstofJulyl99５（thatisfiveyears befbrethesurvey)?，Weconsiderthosewhochange theirplaceofresidence（evenwithinthesame provmce)tobemigrants

Wefbcusexclusivelyoninter=provmcialmigrations，

thatis,mdividualsthatwithmtheprevious5yearsof thesurvey）１ｅｆｔｏｎｅｐｒｏｖｍｃｅｃｏｎｍｕｎｅｔｏｇｏｔｏ ａｎｏｔher・Ourmodelestimatestheextracostofcrossmg provmcialbordersfbrmigrantsbasedonthe comparisonofmter-provmcialandintra-provmcial migrationflows・

Distance2ismeasuredasrealdistancebyroadm

kilometersfbllowmgthequickestroutebasedon detailedmapslnter-provmcialdistancescorrespondto distancesbetweentheircapitalcities・Considermgthe situationofChma,ｗｅｕｓｅｔｈerailwaydistantasthe prmWvariableofdistant，becausetramisthe representativetransportationmChina，especiallym themoveacrossprovmcesMeanwhile，Haman provmceandculTentTibetwasappliedfbrtheairlme distantduetonoavailabilityofrailway

Smceourfbcusisonmter-provmcialmigrations，

ｗｅｕｓｅＧＤＰｐｅｒｃａｐｉｔａｔｏｐｒｏｘｙｆｂｒｗａｇesmthe destmationprovmceiWeusetheaveragevaluefbraU eXplanatolyvariablesoverl995-2000toexplam

人間社会環境研究第15号２００８．３ 1３８

mble6TheRegmssionofmigratiolBanditsfactors

Thispaper SandraProncet(2006）

Variable

1995-2000 1990-1995

1985-1990

－９４２(2.09）

Ｌ６８(224）

-115(286）

－０２７(105）

143(498）

199(411）

Constant

m〕PPercapita(①）

UnemPlOyment(p）

Distant(｡）

NeighboringProvince(F）

GDPGrowth(9) Ruralpopu(pr）

ＯｂｓＮｏ

Ｒ２

244**＊

-035*＊

-0.93＊*＊

０．９２**＊

172**＊

－０２７**＊

-1.07**＊

１１８**＊

－１３０(3.57）

０６０(153）

－０４５(063）

９３１ 0.114 ０．６３(１５）

６５２ ０．３６

９３１ ０．１１ ８２０

０．３８

Note:Heteroskedasticconsistentstandａｒｄｅ汀ors,ｗｉｔｈ＊*＊＊＊ａｎｄ＊dcnomgsigmificanceat１％,５％ａｎｄ10％confidencelevelinSandra^, Proncet(2006),andtwaluemparcnthesesmthispapeⅢ

ｒａｔｅｉｎＣｈｍａｌｔｉｓｗｅｌｌｋｎｏｗｎｔｈａｔｐｒｏvmceswith highruralpopulationratetendtohavemore abandonedlabor，Whichwouldbethemam compositionofmigrationWealsointroducedGDP growthasasupplementvariablefbrtheGDPper capita,ａｓｗｅｋｎｏｗｎ,themorerapid]ygrowthmone area，themoreemploymentopportunitymorebe offered,anditwouldeperceivedasafactortoprompt themigration

Resultsreportedmcolumns3and4emphasizethat thehighertheruralpopulationrateofthedestmation provmcerelativetothedepartureprovmce,thehigher themigratolyHowsThesignificantpositivesign highhghtstheexistenceofacomplementary relationshipbetweenruralpopulationrateand migrationwithmChmaTheseresultsemphasizethat migrationflowsareattractedtoprovmcesthathave higherruralpopulationratethantheprovmceof

orlgm

distancebetweendepartureanddestmationlocations ThemHuenceofdistanceisqUitehighasthe coeBHcientsareclosetoonemabsolutevalue

Thetwovariablesthatrelatetotraditionaleconomic detennmantsofmigrationenterwiththeexpected signsandaresignificant，thatispositivefbrreal mcomeandnegativefbrunemploymentTherateof unemploymententerswiththeexpectednegativesign Thesimultaneoussignificanceoftherevenueandthe unemploymentvariablesisexceptional，asmmost estimationthesevariablesfailtoentersignificantlyin asimultaneousway(Crozet,2004）

ＴｈｅｍＨuenceofbothfactorsappearstomcrease betweenthetwosUb-periods,attestmgtothegrowmg mpoltanceofanexpectedmcomedifferentialln migratorydecisions、Thecoefficientonrealincome mcreasessignificantly（atthe５％confidencelevel）

betweenthetwosub-periods，Ourresultsemphasize，

thatallelseequal,migFantstendtofavordestmations ftomwhichtheyexpectmorefavorableprobabilities offindmgemploymentandhighermcomedifferentials Thisfeatureunderlmesthe印eaterresponsivenessof mobilitydecisionofruralworkerstoeconomic conditionsandpointstothe血provementmefficiency

ofthelabormarket

Inthelasttwocolumnswemvestigatetheexistmg relationshipbetweenmigrationandruralpopulation

6.Conclusions

Afterintroductionofsituation,sourceandpattemof labormobility，wegetaclearsketchoflabormarketm ChmaMeanwhilethefbrcesoflabormobilitywere alsomvestigatedMamythispaperstudiesthe npactsoftheHZJAo〃systemofpennanentregistration

ProvincialLaborMobilityDynamicsinChina 1３９

onlncomemequalityandlabormigraｔｉｏｎｍＣｈｉｎａｌｔｓ ａｍｉｓｔｏｕｓｅａｎｕｍｅｒicalmodeltohelpassessthe contributionofvariouspoliciesandotherfactorsm Chmaineithercontributmgtoorretardmgmequality mChinaWeuseaneconometricmodeltomvestigate thefIactorsoflabormobilitymChma，whichwas consideredtobeakeyftlctortothemequalityinChina

ThebasemodelistakenfiFomTnbuchiandThisse

(2002)andCrozet(2004),Whoevaluatethenpactsof crosscountryimmigrationrestrictionsonglobal meqUalityandmigrantschoosetheirdestmation throughthecomparisonoftheperceivedqUalityoflifb mthevariouslocationsMigrationdecisionsarebased on：migrationcosts，expectedrealmcome，andthe probabilityoffindmgajobmthedestmationOur extensionallowsfbrthemvestigationofthe responsivenessofChmesemigrationdynamicsto costsandothermacro-economicfactors,suchasGDP growthandruralpopulationratio

hequalityinChinasmcel978Jb"”α/q/Ｐｅａｓα"/肋ａｊｅｓ，

21,41-70

DeniseHareandLoraineAWest（1999)SpatialPattemsin China,sRuralIndustrialGrowthandProspectsfbrthe AlleviationofRegionalIncomelnequalityJb"r"αノｑ／

ＣＯ'"ｐａｒａｒｊｖｅＥｍ"o"zjcs,27,475-497

FanZhanandZhiWang（2002）WTOAccession，Rural LabourMigrationandUrbanUnｅｍｐｌｏｙｍｅｎｔｉｎＣｈｍａ Ｕ>bα〃団"ぬＣｓ,３９(12),2199-2217

FengWangandXuQjinZuo（1999）InsideChina,sCities：

InstimtionalBalTiersandOpportunitiesfbrUrban Migrants・Ａｍｅ７ｊｃα〃ＥＣＯ"ｏｍｊｃＲｅＷｅｗ：AEAPapersand Proceedings,８９(2),276-280

ＧｅｎｅＨＣｈａｎｇ（2002）ＴｈｅＣａｕｓｅａｎｄＣｕｒｅｏｆＣｈｉｎａ，s WideninglncomeDisparity・DepartmentofEcononncs，

UniversityofTbledohttp:"umedu/homepage/elliottp／

cerl34changpdf

JianChen(1996)RegionallncomelnequalityandEconomic GrowthinChinaJb"r"α/q/CO"ZP["αm’e此o"omcs,２２，

l41-164

JyotsnaJalanandMartmRavallion(1998)TransientPoverty inPostrefbrmRuralChina,1952-1985．

九"7"αﾉq/CO/"ＰａｒαﾉﾙeEco"omjcs,26,338-357

Kaj-yuenTM（1991）China，sRegionalInequality，

l952-1985JD"r"αノｑ／ＣＯ〃α､!"veEco"０ｍ/Ｃｓ，２９１５，

１－２１

Kai-yuenTEui（1993）DecompositionofChina，sRegional lnequalities力”"αノｑ／ＣＯ/"ｐａ７ａｒｊｖｅＥｃｏ"omjcs，17,

600-627

Kai-yuenT1sui（1996)EconomicRefbnnandhterprovincial hequalityinChinaJD"7"α/q/、ｅｙｅ/Opme"'ＥＣＯ"omJcs，

50,353-368

Kaj-yuenTM（1998）FactorDecompositionofChinese RuralIncomelnequality：NewMethodology，Empirical Findings，andPolicylmplicationsJb”"αノｑ／

ＣＯ"、α7αノルＣＥＣＣ"omjcs,26,502-528

Kaj-yuenTEui（1998）TiPendsandInequalitiesofRural WelfareinChina：EvidencefiPomRuralHouseholdsin GuangdongandSichuanノリ"'"αノｑ／ＣＯ〃”αﾉﾙｅ Ｅｍ"omjCF,26,783-804

KamWingChanandLiZhang（1999)TheHukouSystem andRural-UrbanMigrationinChina：Processesand Changesd2j"αｇ"α"ｅ７か,160,Decemberl999,818-855 MatthieuCrozet（２００４）Domigrantsfbllowmarket

potentia1s？Anestimationofaneweconomicgeography modeL1'ノリ"γ"α/q/ＥＣＯ"ｏｍｊｃＧｅｏｇｍｐ/1ﾘﾉ,ｖｏＬ４(4),pages

439-458

RaviKanburandXiaoboZhang（1999）WhichRegional hequality？TheEvolutionofRural‐UrbanandInland‐

CoastallnequalityinChinahoml983tol995血”"α/q／

Ｃｏｍｐａ７α"veEco"omjcs,27,686-701

RaviKanburandXiaoboZhang（2001）FiftyYearsof RegionallnequalityｉｎＣｈｍａ：AJoumeythrough

Notes

lltcouldbeshowedbythefbllowingequatlon：

γ(0)=血(#)Ｙｈ(,)-X(r)].-'〃c(o）

V(O)：presentvalueofexpectednetincomegapbetween

ruralandurban

P(t):possibilityofmjgrationlaborfindingajobinlabom YU(t),Ｙｒ(t):averageincomeinurbanandruralarea C(O):migrationcost(transportationfeeetc.）

ZWearegratefUltoProfMarukawaofUnivDfTokyofbr hiskindnessofsharmgthedistantdata

Refbrence

AmartyaSen（1973）Ｏ〃ＥＣＯ"ｏｍ/che9"αﾉﾉﾘOxfbrd UniversityPress

AyingLiu，ShUjieYaoandZongyiZhang（２００１）

ConvergenceofChina，sRegionallncome：1952-97.

Ｃ/zmeseECo"ｏｍｊｃｓＲｅＷｅｗ，12,243-358

BobHamiltonandJohnWhalley（1984）EHiciencyand DistributionalImplicationsofGlobalRestrictionson LabourMobility・ん"r"α/q/、eye/qpﾉ?ze"ｒＥｍ"ｏｍｊｍ,１４，

６１－７５

ChinaStateStatisticsBureau，variouseditions，Ｃｈｊ"α 肋"srjcα/形α７６００ﾉﾋBeijing:ChmaStatisticalPress ChrisBramallandMarionEJones（1993）Rurallncome

人間社会環境研究第15号２００８．３

1４０

（、Chinese）

XinzhengShi(2002)EmpiricalResearchonUrban-Rural

hcomeDifferentials:TheCaseofChina

UnpublishedManuscript,CCERPekmgUniversity XinZhengShi，TbrrySicularandYaohuiZhao（2002）

AnalyzingUrban‐RurallncomehequalitymChma PaperpresentedatthelntemationalSymposiumon EqualityandSocialJusticeinTiPansitionalChinaBeqing，

Ｊｕｌｙｌｌ‐１２

YaohuiZhao（1999）LabourMigrationandEaming

Differences：ＴｈｅＣａｓｅｏｆＲｕｒａｌＣｈｍａＥｃｏ"ｏｍｊｃ

比veﾉOpme"/α"`α〃だＣ/zα"ge,47(4),767‐782 YaohuiZhao（1999)LeavingtheCountryside:Rural-to‐

UrbanMigrationDecisionsmChinaAmerjcα〃ＥＣＯ"ｏｍＪｃ ＲｅＷｅｗ,８９，２８１－２８６

Revolution,RefbnnandOpenness・WOrkingPaper2001- O4，DepartmentofEconomicsandManagement，Comell University

Sandraponcet（Z006DProvmcialmigrationdynamicsin China:Borders,costsandeconomicmotivations,Regjo"αノ Sbje"ｃｅｍａＵ>もα〃ＥＣＯ"ｏｍｊｃｓ３６,385-398

ScottRozelle（1994)RuralIndustrializationandlncreasing lnequality：EmergingPattemsinChina，sRefbnning EconomybJD”"αノｑ/・ＣＯﾉﾘＷｒａｍ′ｅ比o"omjcs，１９，

３６２－３９１

ThomasRLyons(1991)InterprovincialDisparitiesinChina：

OutputandConsumption，1952-1987．ＥＣＯ"ｏｍｊｃ 比veﾉOpmc"/Ｃ"〃α/Ｃｈα"ge,39,471-506

XiaoyuFanandShufenQie（2002）TheSituationand FeaturesofRuralLabourMigrationC/ｚｊ"αbSiZJﾉﾉS"Ｃｓ，７