经济学论文

IZADPNo.2995IstheChineseGrowthMiracleBuilttoLast?EswarS.PrasadDISCUSSIONPAPERSERIESForschungsinstitutzurZukunftderArbeitInstitutefortheStudyofLaborAugust2007IstheChineseGrowthMiracleBuilttoLast?EswarS.PrasadCornellUniversityandIZADiscussionPaperNo.2995August2007IZAP.O.Box724053072BonnGermanyPhone:+49-228-3894-0Fax:+49-228-3894-180E-mail:iza@iza.orgThispapercanbedownloadedwithoutchargeat:http://ssrn.com/abstract=1012561AnindextoIZADiscussionPapersislocatedat:http://www.iza.org/publications/dps/Anyopinionsexpressedherearethoseoftheauthor(s)andnotthoseoftheinstitute.ResearchdisseminatedbyIZAmayincludeviewsonpolicy,buttheinstituteitselftakesnoinstitutionalpolicypositions.TheInstitutefortheStudyofLabor(IZA)inBonnisalocalandvirtualinternationalresearchcenterandaplaceofcommunicationbetweenscience,politicsandbusiness.IZAisanindependentnonprofitcompanysupportedbyDeutschePostWorldNet.ThecenterisassociatedwiththeUniversityofBonnandoffersastimulatingresearchenvironmentthroughitsresearchnetworks,researchsupport,andvisitorsanddoctoralprograms.IZAengagesin(i)originalandinternationallycompetitiveresearchinallfieldsoflaboreconomics,(ii)developmentofpolicyconcepts,and(iii)disseminationofresearchresultsandconceptstotheinterestedpublic.IZADiscussionPapersoftenrepresentpreliminaryworkandarecirculatedtoencouragediscussion.Citationofsuchapapershouldaccountforitsprovisionalcharacter.Arevisedversionmaybeavailabledirectlyfromtheauthor.IZADiscussionPaperNo.2995August2007ABSTRACTIstheChineseGrowthMiracleBuilttoLast?*IstheChinesegrowthmiracle–aremarkablyhighgrowthratesustainedforovertwodecades–likelytopersistoraretheseedsofitseventualdemisecontainedinthepoliciesthathaveboostedgrowth?Forallitspresumedflaws,theparticularapproachtomacroeconomicandstructuralpoliciesthathasbeenadoptedbytheChinesegovernmenthashelpedtodeliverhighproductivityandoutputgrowth,alongwithareasonabledegreeofmacroeconomicstability.Intandemwithabenigninternationalenvironment,thisapproachmakesitunlikelythattheeconomywillfaceacollapseingrowth.Buttherecomesapointwhenthepolicydistortionsneededtomaintainthisapproachcouldgenerateimbalances,imposepotentiallylargewelfarecosts,andthemselvesbecomeasourceofinstability.Thetraditionalrisksfacedbyemergingmarketeconomies,especiallythoserelatedtohavinganopencapitalaccount,donotloomlargeinthecaseofChina.Intheprocessofsecuringprotectionagainstexternalrisks,however,Chinesepolicymakersmayhaveincreasedtherisksofinternalinstability.Thereareanumberoffactorsthatcouldtriggerunfavorableeconomicdynamicsthat,eveniftheydon’trisetothelevelofacrisis,couldhaveseriousadverserepercussionsongrowthandwelfare.Theflexibilityandpotencyofmacroeconomictoolstodealwithsuchnegativeshocksisconstrainedbythepanoplyofpoliciesthathassupportedgrowthsofar.JELClassification:F3,E5,O1Keywords:macroeconomicpolicies,exchangerateflexibility,capitalaccountliberalization,financialsectorreformsCorrespondingauthor:EswarS.PrasadDepartmentofAppliedEconomicsandManagementCornellUniversity440WarrenHallIthaca,NY14853USAE-mail:eswar.prasad@cornell.edu*EarlierversionsofthispaperwerepresentedattheAmericanEnterpriseInstitute,theChinaBankingRegulatoryCommission,thePeople’sBankofChina,theChinaCenterforEconomicResearchatPekingUniversity,andtheEBRD.IthankPhilipLevy,JustinLin,LuoPing,LiRuogu,AndrewShengandnumerousothercolleaguesforusefuldiscussionsandcomments.IamgratefultoSunTaoforhelpinobtainingsomeofthedatausedinthispaper.RahulAnandandWellianWirantoprovidedableresearchassistance.1I.OverviewTheChineseeconomyhasbeenhummingalongforovertwodecadesnow,withnosignofslowingdownanytimesoon.TheeconomyheldtogetherwellevenduringthetumultuousperiodoftheAsiancrisisinthelate1990s.Sincethen,annualGDPgrowthhasaveragedabout10percentand,afterabriefflirtationwithdeflationduring2002-03,inflationhasbeguncreepingupbutremainsmoderate,withtherecentspikesinCPIinflationlargelyattributabletofoodpriceincreases(Figure1).ThisremarkablemacroeconomicsuccesshasgivenrisetotwoopposingviewsaboutprospectsfortheChineseeconomy.Oneisthatalliswell—indeed,insomewaysthingsmaybeevenbetterthantheylook—and,withsomeminorpolicyadjustments,thereisnoreasontheparty(theeconomicone,thatis)cannotgoonindefinitely.Theotherviewisthattheeconomicboomisbuiltonahouseofcardsthatcouldcomecrashingdownanyminute.Whereshouldweplaceourbets?Istheeconomyfundamentallystrongorjustpumpeduponsteroids?Evenacasualreaderofeconomichistoryknowsthatapparentlysurethings,bornealongonthebeliefthatthistimethingsarereallydifferent,oftencomeunhingedwithlittlewarning.Butperhapsthingsreallyaredifferentthistime.WhiletheChinesegrowthmiraclehasborrowedpartofitstemplatefromothersuccessstories,therearemanyaspectstotheChineseexperiencethatdomakeitsuigeneris.Forinstance,theinvestmentboomthathasunderpinnedgrowthhasbeenlargelyfinancedoutofdomesticsavings,thestockofforeignexchangereservesiswellbeyondanyreasonableprecautionarynormsand--despitethehugecapitalinflowsduring2001-05,boomingtradesurplusesinthelasttwoyears,andtherapidrateofexpansionofmonetaryandcreditaggregates--therearefewsignsofmajorinflationarypressuresincorepriceindicators.Inthispaper,IwillarguethatthingsareindeeddifferentinthesensethatChina’spresentsituationmakesagrowthcollapseunlikely.However,thisisnottosaythatalliswell.TherearedeepstructuralproblemsintheeconomythatChinesepolicymakersneedtocometogripswith.Ahighrateofgrowthisofcourseoneofthebesttonicsforsuchproblemsastheydiminishtheirscaleinrelativetermsandalsocreateresourcestodealwiththem.Perhaps,therefore,themeanstoachievinghighgrowthmaybelessrelevantthantheendresults.Butrapidgrowthcanhide,orinsomecasesevenexacerbate,anumberofproblemsthatcomehometoroostwhenthegoinggetsrough,asithasahabitofdoing.Forinstance,thefinancialsectorisinpoorshapeandhasdistorteddomesticdemand;thepatternsofinvestmentfinancingcouldleadtoaresurgenceofnonperformingloans(NPLs)inthefutureand,byfuelingabuildupofexcesscapacityinsomesectors,couldgeneratedeflationaryrisksinthemediumterm.Meanwhile,intheshortterm,someofthepressuresarebecomingevidentinotherformssuchasassetpricebooms(intheequitymarkets,inparticular).2Inmyview,however,thesustainabilityofgrowth--whileanimportantconcern--isnotthekeyproblem.Thereareindirectandsubtlecoststothecurrentgrowthmodelthatdeservefarmoreattention.Forinstance,ChinahasheldtheexchangerateoftherenminbirelativetotheU.S.dollarwithinanarrowrangedespiteenormouspressuresforasubstantialappreciationinrecentyears.Andthishasbeenaccomplishedwithoutthetypicalrisingquasi-fiscalcostsofsterilizingtheliquiditygeneratedbylargecapitalinflows.Butsterilizationhasbeenfacilitatedbyfinancialrepressionandarelativelyclosedcapitalaccount.Thishas,amongotherthings,meantverylowrealratesofreturnforhouseholds,whosavealotandhavefewinvestmentopportunitiesotherthandomesticbankdeposits.Thesepolicieshavealsocurtailedfinancialsectordevelopment,leadingtoinefficientintermediationofdomesticcapital.Thereareclearlylargewelfarecostsassociatedwiththeseconstraints.Thesecondissue,whichloomslargerinmymind,isthatthegrowthstrategyhasinvolvedanumberofpolicydistortionsandconstraintsthathavegreatlyreducedtheroomforpolicymaneuverincaseanybigshocksshouldhit.Itisinevitablethat,astheeconomybecomesmorecomplexandintegratedwiththeworldtradeandfinancialsystems,itwillbeexposedtomoreshocks.Suchshockscouldcomefrominternalsources—e.g.,lossofconfidenceinthebankingsystem,socialinstabilitygeneratedbyrisinginequality—orexternalsources—e.g.,internationalcapitalmarketcrises,acollapseofexternaldemand,U.S.tradesanctions,flaring-upoftensionsoverTaiwanetc.Ofcourse,shockstendtomutateintoformsthatonecannotpredictandoftencomefromunexpecteddirections.Chroniclesofcrisesforetoldarefaroutnumberedbycrisesthatwerelargelyunanticipateduntilthingsbegantounravel.Thus,amorerelevantquestionthantheoneaboutsustainabilityonthebasisofpurelyinternaldynamicsofthesystemiswhethertheeconomyhasenoughflexibilitytowithstandandrecoverfromlargeshocks,eitherinternalorexternal.Heretheanswerisfarlessclear.Monetarypolicyistypicallythefirstlineofdefenseagainstsuchshocksbut,withmonetarypolicyconstrainedbytheobjectiveofmaintainingatightlymanagedexchangerate,itcanatbestplayaverylimitedrole.Thereappearstoberoomforfiscalmaneuversincetheexplicitlevelsofthefiscaldeficitandgovernmentdebtarequitelow(Figure2),butthismaybedeceptiveastherearelargecontingentliabilitiesinthestate-ownedbankingsystemandhugeunfundedpensionliabilities.Thefinancialsystemisstilldysfunctionalinmanywaysandmaynotbedeeporrobustenoughtowithstandasignificantshock.SowhatshouldChinadotoprepareitselftodealwithshocks,andmakegrowthmorebalancedandsustainable?Therearesomeanswersthatfewwoulddisagreewith.Thebankingsystemshouldbemademorerobustanddrivenbymarketprinciples,andthefinancialsystemshouldbebroadenedtocreatealternativesourcesoffundingforfirmsandalternativeinvestmentopportunitiesforhouseholdsandfirms.Thestate-ownedenterprisesectorneedstobefurthercorporatizedbyhardeningbudgetconstraints.Thereisaneedforabettersocialsafetynetandabettersystemfordeliveryofsocialservices.3Whileitiseasytocreateanevenlongerlaundrylistofreforms,akeypointisthatmanyofthesereformsareinter-relatedandChinahasreachedapointinitsevolutiontowardsamajormarket-orientedeconomythattryingtoimplementthesereformsinisolationisnotaneffectivewaytoproceed.Forinstance,stablemacroeconomicpoliciesandawell-developedandefficientfinancialsectorareessentialingredientsforbalancedandsustainablegrowth.Inturn,thesetwointermediateobjectiveswouldbehelpedbyeffectivemonetarypolicyandfurthercapitalaccountliberalization.Andaflexibleexchangerateisaprerequisiteforbothofthese.Ignoringtheselinkages—forinstance,tryingtopushforwardwithbankingreformswhileholdingmonetarypolicyhostagetoanexchangerateobjective—makesanalreadydifficultreformprocessevenharder.Similarly,financialrepressionhaskepttherealpriceofcapitalcheapand,alongwithsubsidizedenergyandlandprices,shiftedproductiontowardscapital-intensivemethods.Thisworksatcross-purposeswiththeauthorities’goalofboostingemploymentgrowthandfacilitatingthetransitionofruralunemployedandunderemployedtoemploymentinmanufacturingandservices.Beforeonegetssweptawaybydoomandgloomscenarios,however,itisalsoimportanttorecognizesomeoftheintrinsicstrengthsofthesystem.Whilethestatesectorremainsdominant,thereisavibrantandrapidlyexpandingprivatesectorthat,despiteallofitsownflaws,isprobablycontributingevenmoretooutputandemploymentgrowththansuggestedbyofficialstatistics.Therehasalsobeensignificantreformofthestate-ownedenterpriseandbankingsectors.Thereisalongwaytogoonreformofthesesectorsbut,giventhelegacyproblemsthattheyfacedtobeginwithandtheapparentintractablenessoftherootproblems,suchprogresscannotbedismissedasbeinginconsequential.Thereareotherstructuralfeaturesoftheeconomythatalsoconstitutekeystrengths—alabormarketthatisflexibleinmanydimensions,aphysicalinfrastructurethatistheenvyofmanycountriesatasimilarstageofdevelopment,aneconomythatisveryopentotrade,andahighlycompetitiveexportsector.Indeed,allofthesefactorscometogetherindeliveringtheapparentlyhighrateofproductivitygrowthandsubduedinflation.Ultimately,theessenceofthepolicydebatecanbeframedintermsofthepaceandsequencingofreformsrequiredtoturnthesestrengthsintoforcesthatallowthegrowthmiracletobesustainedandtoreducetherisksofitsbeingderailedbyshocks.Thereis,asalwaysadelicatebalancetobestruckintermsofthereformstrategy;economictheorybyitselfisoflittlepracticalvaluehereasitpointsoutthefirstbestbutnottheleastriskywaytogetthere.Onekeyprinciple,asnotedabove,istorecognizethatthereareinherentlimitstotheincrementalreformstrategythathasworkedwellinthepast.Atacertainlevelofdevelopmentandcomplexityofaneconomy,theconnectionsamongdifferentreformsbecomedifficulttoignore.Thereisalsothepressingneedtotakeadvantageoffavorableconditionsthatnowexist--highgrowth,lowinflation,abenigninternationalenvironment--topushforwardwithreformsthatcouldbemuchmessierandmoredifficulttoimplementunderduresswhentimesarenotasgood.4II.ParsingtheGrowthStoryOnedimensionoftheChinesegrowthstorythatisofparticularrelevanceinthecontextoftheargumentsinthispaperisthecompositionofgrowth.Untilrecently,dataonthecompositionofGDPfromtheexpendituresidewereavailableonlyonanominalbasis.Toprovideahistoricalcomparison,Ipresentthosedatahere.Figure3showsthatinvestmenthasbeenamajorcontributortogrowthduringthisdecade,insomeyearsaccountingfornearlytwo-thirdsofnominalGDPgrowth.Itisalsoworthnotingthatnetexportsbythemselveswerenotamajorcontributortogrowthformostofthisdecade.EventhoughChineseexportshavebeengrowingbyleapsandbounds,thedirectgrowthcontributionofthetradebalancehadbeenrelativelysmalluntil2005-6.Foradevelopingeconomythatistypicallylabor-richbutcapital-poor,highratesofinvestmentoughttobegoodforfastercapitalaccumulationandhighergrowthduringtheconvergencetothesteadystaterateoflong-rungrowth.ButasubstantialfractionofthisinvestmentinChinahasbeenfinancedbycreditprovidedbystate-ownedbanksatlowinterestrates.Indeed,cheapcapitalhasplayedabigpartinskewingthecapital-laborratioandholdingdownemploymentgrowth(Aziz,2006).Inaddition,localgovernmentsprovidesubsidizedlandinordertoencourageinvestment.Andenergypricescontinuetobeadministeredandmadeavailabletoenterprisesatpricesbelowinternationallevels.Hence,thepricesofthefactorsofproductionthatserveascomplementaryinputstophysicalcapital—landandenergy—arealsocheap.Consequently,thegovernmenthasnotbeenabletocooloffinvestmentgrowthinrecentyearsdespitefearsofoverheatingandtheinvestmenttoGDPratiohasremainedabove40percent(Figure4).Oneconsequenceoftheinvestment-heavyexpansionhasbeentherelativelyslowemploymentgrowth.Itisnowherenearapacesufficienttokeepupwiththegrowthinthelaborforce,absorbworkerslaidofffromstateenterprisesthatareretrenching,andalsoabsorbexcessrurallabor(whichsomepolicymakershavespeculatedcouldbeontheorderof150millionpersons,nearlyhalfoftheruralworkforce).2Indeed,duringtheperiod2000-05,growthoftotalnon-agriculturalemploymentaveragedlessthan3percentperannum(seethelastcolumninTable1),comparedtoaveragenon-agriculturalGDPgrowthofabout9.5percent(Table2).Therearetwofeaturesofbanklendingthatmayberelevantforthepriceofcapital—thefirstisthelendingrate.Recentincreasesinthebaselendingratehavebeenfartoosmalltoraisetherealpriceofcapitaltoameaningfullevelforaneconomythatisexperiencingannualrealgrowthofover10percent(Figure5).Thesecondfactoristheabsenceofobviouspenaltiesfornonrepaymentofloansbystate-ownedenterprises,whichhaseffectivelyactedasafurthersubsidyforcapital.Thisfactorhasbecomelessrelevantovertimeasstateenterprisesarebeingsubjecttoincreasinglyhardbudgetconstraintsandbanks’NPLpositionsarebeingcloselymonitored.2SeethechapterbyBrooksinPrasad(ed.,2004)forsomedetailedcalculations.5Theceilingonlendingrateshadactedasadirectinstrumentforsubsidizingcapital.Followinggradualincreasesinthelendingrate,inOctober2004theceilingonlendingrateswasscrappedaltogether(exceptforurbanandruralcreditcooperatives).3Thesubsequentwideningofthegapbetweenthebaselendingrateandtheactuallendingrate(aweightedaveragebasedonloanvolumes)indicatedthatbankswerebeginningtousethismargin(Figure6).However,thismarginhasshrunkrecentlyandthereislittleevidencethatthelargestatecommercialbanks(SCBs),inparticular,areusingthisflexibilitytosubstantiallyredirectlendingtotheprivatesectorathigherinterestrates.ResultsofsurveysoflendinginstitutionsreportedinthePeople’sBankofChina’squarterlyMonetaryPolicyReportsrevealthatbanksstillpriceamajorityoftheirloansaroundthebaselendingrate(whichhasnowbecomeafloor).4Thiscouldinpartreflectconcernsbankshaveabouttheirownrisk-assessmentcapabilities,especiallyinanenvironmentwherethereisstillstrongpressuretoavoidaccumulationofnewNPLs.Alessbenignexplanationisthatbanksarerespondingtoaninformalincentivestructurethatremainsunchanged—loansmadetostateenterprisesarestillregardedaslessriskyintermsofreputationalcoststobankmanagersandloanofficers,whileloansmadetoprivatesectorenterprisesthatbecomenonperformingcouldentailchargesofincompetenceorcorruption.5Deficienciesinthelegalframeworkmayalsoplayarole.Weaklegalprotectionmeansthatcollateralprovisionsaredifficulttoenforce,solendingtotheprivatesectorcarriesadditionalrisks.Alotoftherecentinvestmenthasalsobeenfinancedthroughretainedearnings;suchinvestmentissurelyreasonableasitisbeingundertakenbyprofitablefirms.However,evenherethepictureisfarfromclear.Profitablestateenterpriseswerenot,untilveryrecently,requiredtopaydividendstothestate.Thissuggeststhatsuchinvestmentmayhavebeenspurredbytheminimalratesofreturnonbankdepositswhichmadeevenmarginalinvestmentprojectsseeminthemoney.Therisk,ofcourse,isthatsuchhighratesofinvestmentinindustrieswithfavorabledemandconditionsmaybeleadingtoabuildupofexcesscapacityinthoseveryindustries;thiscouldbecomeevidentifthereweretobeadversedemandshocksinthefuture(GoldsteinandLardy,2006).3DunawayandPrasad(2004)notethepotentialbenefitsofthispolicyshift.4State-ownedcommercialbanks,joint-stockcommercialbanksandpolicybanksaccountedforaboutthree-quartersofallRMBloansmadebydomesticbanksin2006.Inthefourthquarterof2006,about98percentoftheloansmadebythesethreegroupsofbankswereataninterestratenohigherthan1.3timesthebaselendingrate—i.e.,below9percentevenforlong-termloans(seeTables1and2inthePBCMonetaryPolicyReport,2006Q4).5Anecdotalevidencesuggeststhatbanklendinghasbecomeincreasinglyskewedtowardslargeenterprises,whichareofcoursemostlystate-owned.Podpiera(2006)examineslendinggrowth,creditpricing,andregionalpatternsinlendingtolookforevidenceofchangesinthebehaviorofSCBsfollowingrecentreformsandstrengtheningoftheirbalancesheets.HeconcludesthatthepricingofcreditriskremainsratherundifferentiatedandthatSCBsdonotappeartotakeenterpriseprofitabilityintoaccountwhenmakinglendingdecisions.DobsonandKashyap(2006)discusssomeofthechallengesofbankingreformsinChina.6Whileinvestmenthasbeenhigh,nationalsavingshavebeenevenhigher,withbothhouseholdandcorporatesavingsrisinginrecentyears.Theuncertaintiesengenderedbythetransitiontoamarketeconomy,thelimitedavailabilityofinstrumentstoborrowagainstfutureincometofinancepurchases(majordurablegoods,housingetc.),andthelackofinternationalportfoliodiversificationopportunitieshaveallcontributedtohighhouseholdsavings(ChamonandPrasad,2007).Financialsystemrepressionhasmeantthattherearefewalternativestofunnelingthesesavingsintodepositsinthestate-ownedbankingsystem(Figure7).Householdswillinglyholdbankdepositsdespitetheweaknessesofthebankingsystembecauseofimplicitdepositinsuranceprovidedbythegovernment.Thisprovidesabundantliquidityforbankstoexpandcreditwhich,becauseofthedistortedincentivesfacedbylenders,largelyfinancesinvestmentbystateenterprises.Stateenterprisesthatdomakeprofitsarenotrequiredtopaydividends,encouragingthemtoplowretainedearnings(whicharecountedasenterprisesavings)backintoinvestment.Thus,theinvestmentboominrecentyearshasbeenfueledbycheapcreditandoveroptimisticexpectationsoffuturedemandgrowthinsectorsthataredoingwellatpresent.Asformanufacturingexports,Chinesepolicymakersclearlyseethemasanintegralpartofthegrowthstrategy.Policymeasurestoencourageexportsinclude(untilveryrecently)substantialtaxincentivesforforeigndirectinvestmentintoChina—mostofthisFDIhasgoneintotheexportingsector—andthesetting-upofspecialeconomiczonesforexportingfirms.ExportgrowthskyrocketedafterChina’sWTOaccessionin2001,whichopenedupindustrialcountrymarketsthathadpreviouslyrestrictedimportsfromChina.ItisestimatedthatabouthalfofChina’sexportsareaccountedforbyprocessingtrade,whichaccountsforonlylimitedvalueaddedtofinalproductswithinChina.However,farmoreimportantthanthedirectcontributionofthetradebalancetoGDPmayhavebeenthecatalyticeffectofgrossexports,especiallythetechnologicalandefficiencygainsthathavecomewithgreateroutwardorientationoftheeconomy.III.PolicyChoicesandTheirConsequencesInowturntoadiscussionofeconomicpolicychoicesthat,unwittinglyorotherwise,havegeneratedthepatternsofgrowthdescribedabove.China’scurrencypolicyhasofcoursereceivedthemostattentionoflate.Whetherthemaintenanceofafixedexchangerateispartofadeliberatemercantiliststrategytopromoteexport-ledgrowthhasbeenthesubjectofintensedebate.Onthe