Growth, population and industrialization, and urba

JournalofUrbanEconomics63(2008)96–115∗,JikunHuanga,ScottRozelleb,EmiUchidacaCenterforChineseAgriculturalPolicy,InstituteofGeographicalSciencesandNaturalResourcesResearch,ChineseAcademyofSciences,Beijing100101,ChinabFreemanSpogliInstituteforInternationalStudies,StanfordUniversity,EncinaHallEast,E301Stanford,CA94305-6055,USAcDepartmentofEnvironmentalandNaturalResourceEconomics,UniversityofRhodeIsland,Kingston,RI02881,USAReceived1February2006;revised18December2006Availableonline30January2007AbstractChinaisexperiencingurbanizationatanunprecedentedrateoverthelasttwodecades.TheoverallgoalofthispaperistounderstandtheextentofandthefactorsdrivingurbanexpansioninChinafromthelate1980sto2000.Weuseauniquethree-periodpaneldatasetofhigh-resolutionsatelliteimagerydataandsocioeconomicdataforentireareaofcoterminousChina.Consistentwithanumberofthekeyhypothesesgeneratedbythemonocentricmodel,ourresultsdemonstratethepowerfulrolethatthegrowthofincomehasplayedinChina’surbanexpansion.Insomeempiricalmodels,theotherkeyvariablesinthemonocentricmodel—population,thevalueofagriculturallandandtransportationcosts—alsomatter.Adaptingthebasicempiricalmodeltoaccountfortheenvironmentindevelopingcountries,wealsofindthatindustrializationandtheriseoftheservicesectorappeartohaveaffectedthegrowthoftheurbancore,buttheirrolewasrelativelysmallwhencomparedtothedirecteffectsofeconomicgrowth.Wealsomakeamethodologicalcontribution,demonstratingthepotentialimportanceofaccountingforunobservedfixedeffects.©2007ElsevierInc.Allrightsreserved.Keywords:Urbanization;Spatialscale;Monocentricurbanmodel;Industrialization;Remotesensing;Econometricanalyses;Decompositionanalyses;China*Correspondingauthor.Fax:[phone].E-mailaddresses:[e-mail](X.Deng),[e-mail](J.Huang),[e-mail](S.Rozelle),[e-mail](E.Uchida).0094-1190/$–seefrontmatter©2007ElsevierInc.Allrightsreserved.doi:10.1016/j.jue.2006.12.006X.Dengetal./JournalofUrbanEconomics63(2008)96–11597Economistsinurbaneconomicshavelongbeeninterestedinunderstandingtheprocessofurbanization.Inoneoftheearliesttheoreticalefforts,researchersdefinedandidentifiedtheoreti-callyanumberofthecausesandconsequencesoftheexpansionofurbanland[1,20,23].Amongotherthings,themonocentricurbanmodelgenerateshypotheseslinkingthechangesinurbanlandareatosomeofthefundamentalbuildingblocksofeconomies:income,population,agricul-turallandvaluesandtransportationcosts.Insubsequentyears,empiricaleconomistshavetestedthesehypotheseswithincreasinglylargedatasetsandincreasinglyprecisemeasures.Manyofthem—fromBruecknerandFansler[2]toMcGrath[18]—findsupportforthepredictionofthemonocentriccitymodel.Interestingly,althoughperhapsunsurprisingly,almostalllarge,systematicstudiesofthede-terminantsofurbanareahavebeendoneindevelopedcountries.Thisisinspiteofanargumentthatanunderstandingofurbanizationisevenmoreimportantinthedevelopingworldsincecitiesaregrowingfasterindevelopingcountries,problemsrelatedtotheconversionoflandintobuilt-upareaaremoresevereandthescopeforintervening(aswellasthepotentialgainsfromgoodpolicies)maybegreater.Thatisnottosaythattherehavenotbeeneffortstostudytheprocessofurbanizationindevelopingcountriesanditscauses[10,27,29,30].But,mostlikelyduetodatalimitations,large,nationwideeconometricstudiesusinghighqualitydataarerare.Inresponse,theoverallgoalofthisstudyistoprovideempirical-basedevidencethatwillhelpanswerseveralkeyquestionsontheextentandcausesofurbanlandexpansionsinthedevelop-ingworld—focusinginourcaseonChina.Specifically,thisstudywillbeginwithaconceptualframeworkfromtheurbaneconomicsliteratureandtestaseriesofhypothesesfromthemono-centriccitymodelusingaChina-widedatasetontheexpansionofurbanareainmorethan2000“metropolitanregions”overtime.Inthepursuitofthisobjective,wewillseektoanswerseveralfundamentalquestionsaboutChina:Duringthereformera,whathasbeentheextentofurbanlandexpansionofChina?Whatarethemaindeterminantsthatexplainurbanlandexpansiondur-ingthereformera?Whichfactorshavebeenthemostimportantintermsoftheirimpactontheexpansionofurbanland?Takentogether,wehopethatwecanprovideinformationtopolicymak-ersabouttheextentandmaincausesofurbanizationinChinasothaturbanlandusepolicywillhaveamoreempiricalbasis.Moregenerally,wealsowillshowthatwhenurbaneconomistshaveaccesstodataovertimethechoiceofmethodologythatcancontrolformanyoftheunobservedfactorswhichaffecttheexpansionofaregion’surbanareawillaffectthefindings.Therestofthispaperisorganizedasthefollows.Thenextsectionreviewstheurbaneconomicliterature—boththeoreticalandempirical—andmotivatesthehypothesestobetested.ThethirdandfourthsectionsintroducethedatausedinthisstudyandreviewtherecordofurbanlandexpansionacrossChinaandhowitcorrelateswithsomeofthekeyvariablesthatweareinterestedin.Thefifthandsixthsectionspresenttheempiricalmodelanddescribethefindingsofthetwomainpartsofouranalysis:thedeterminantsofurbanexpansion(intheeconometricanalysis)andmeasurementoftheimportanceofthed