CH0112Reversal of Fortune(激励与组织-北京大学,周黎安)

1激励与组织专题报告ReversalofFortune:GeographyandInstitutionsintheMakingoftheModernWorldIncomeDistributionAcemoglu,JohnsonandRobinson,2002QuarterlyJournalofEconomics罗凯2005年6月9日2Background3TwocompetitivehypothesesGeographyhypothesisThepersistenceprediction:totheextentthatgeographicfactorsdonotchangeoverperiodsof500yearsormore,countriesthatwererelativelyrich500yearsagoshouldberelativelyrichtoday.Ifclimate,transportcosts,thetimingoftheNeolithicrevolutionandtheimpactofdiseasematterforincometoday,theyshouldhavematteredatleastasmuchbeforeEuropeanscameintocontactwiththeinhabitantsofthecolonies.4TwocompetitivehypothesesInstitutionshypothesissocietieswithasocialorganizationthatprovidesencouragementforinvestmentwillprosper.Theorganizationofsocietyandinstitutionsalsopersist.(AccordingtoAcemoglu,JohnsonandRobinson,2001).However,amajorshockliketheinstitutionalreversal—Europeancolonialismledtotheestablishmentof,orcontinuationofalreadyexisting,extractiveinstitutionsinpreviouslyprosperousareasandtothedevelopmentofinstitutionsofprivatepropertyinpreviouslypoorareas—candisrupttheorganizationofagroupofsocieties,andpredictsareversalinrelativeincomesamongthesecountries.5TechnologyOLS2SLSNaturalExperiment6Proxiesforeconomicprosperityin1500UrbanizationKuznets(1968,p.1)Bairoch(1988)FigureA1Table2,Columns1-6Figure3Table2,Columns7-10;paneldata;urbanizationandincomepercapitaspanningover200years7FigureA18Proxiesforeconomicprosperityin1500UrbanizationKuznets(1968,p.1)Bairoch(1988)FigureA1Table2,Columns1-6Figure3Table2,Columns7-10;paneldata;urbanizationandincomepercapitaspanningover200years9Table2,Columns1-610Proxiesforeconomicprosperityin1500UrbanizationKuznets(1968,p.1)Bairoch(1988)FigureA1Table2,Columns1-6Figure3Table2,Columns7-10;paneldata;urbanizationandincomepercapitaspanningover200years11Figure312Proxiesforeconomicprosperityin1500UrbanizationKuznets(1968,p.1)Bairoch(1988)FigureA1Table2,Columns1-6Figure3Table2,Columns7-10;paneldata;urbanizationandincomepercapitaspanningover200years13Table2,Columns7-10(panel)14Proxiesforeconomicprosperityin1500PopulationdensityBairoch,1988,chapter1Malthus--transitorydifferences:persistencyThemoredenselysettledareasin1500werealsomoredenselysettledin1000or1AD.FigureA2SomereservationsregardingtheproxyNosimilarrelationshipinrecentdataDemographictransition15FigureA216Proxiesforeconomicprosperityin1500PopulationdensityBairoch,1988,chapter1Malthus--transitorydifferences:persistencyThemoredenselysettledareasin1500werealsomoredenselysettledin1000or1AD.FigureA2SomereservationsregardingtheproxyNosimilarrelationshipinrecentdataDemographictransition17Proxiesforeconomicprosperityin1500PopulationdensityMoreextensive:ausefulcheckonourresultsusingurbanizationdataHighlycorrelatedwithurbanizationImportantnotonlyindocumentingthereversal,butalsoinexplainingit18DataDependentvariableLogGDPpercapita(PPP)in1995LogGDPpercapitain1900and1950IndustrialOutputpercapitaTotalUKindustrialoutput19DataKeyvariables:UrbanizationUrbanizationin1960and1995Urbanizationin1000,1500,and1700PopulationdensityLogPopulationDensityin1AD,1000and1500(alsologpopulationin1500andlogarablelandin1500)20DataControlvariables:Europeansettlementsin1800and1900AverageProtectionagainstExpropriationRisk,1985-95ConstraintsonExecutivein1900,1970,1990andfirstyearofindependence.PercentofEuropeandescentin1975ReligionVariables21DataControlvariables:ColonialDummiesTemperatureVariablesHumidityVariablesSoilQualityNaturalResourcesDummyforLandlockedLatitude22DataIVLogMortality23ResultsDependentVariableislogGDPpercapita(PPPbasis)in1995.Basesampleisallformercoloniesforwhichwehavedata.Urbanizationin1500ispercentofthepopulationlivingintownswith5,000ormorepeople.TheregressionthatincludescontinentdummiesomitsOceania.Theneo-EuropesaretheUSA,Canada,AustraliaandNewZealand.Intheothergeographyregressionweinclude5measuresoftemperature,4measuresofhumidity,7measuresofsoilquality,and5measuresofnaturalresources.TheregressionthatcontrolsforcolonialoriginincludesdummiesforformerFrenchcolony,Spanishcolony,Portuguesecolony,Belgiancolony,Italiancolony,Germancolony,andDutchcolony.Britishcoloniesareomitted.ThereligionvariablesarepercentofthepopulationwhoareMuslim,Catholic,andother;percentProtestantisomitted.FordetailedsourcesanddescriptionsseeAppendixTableA1.24Results25Results26FurtherDiscussionMeasurementError27Results28FurtherDiscussionTemperateDriftTimingofTheReversal&IndustrializationGeography&Industrialization29Results30Results31Results32ResultsPanelregressions;countryandperioddummiesareincludedasindicatedatthefootofeachcolumn.DependentvariableinPanelAisindustrialoutputpercapita,withupto10observationspercountry,1750-1980,fromBairoch(1982).DependentvariableinPanelBislogGDPpercapitafromMaddison(1995).TheindependentvariableofinterestistotalUKindustrialoutputinteractedwithconstraintsontheexecutiveineachcountryfromthePolityIVdataset.ThemaineffectofinstitutionsisevaluatedatthemeanvalueofUKindustrialization.PolityIVp