出口和外国直接投资作为替代战略外文翻译.docx

出口和外国直接投资作为替代战略外文翻译.docx

《出口和外国直接投资作为替代战略外文翻译.docx》由会员分享，可在线阅读，更多相关《出口和外国直接投资作为替代战略外文翻译.docx（17页珍藏版）》请在冰豆网上搜索。

出口和外国直接投资作为替代战略外文翻译

外文翻译

原文

EXPORTINGANDFDIASALTERNATIVESTRATEGIES

MaterialSource：

OXFORDREVIEWOFECONOMICPOLICY，VOL.20，

Author：

KeithHead，JohnRies

Exportsandoverseasproductionarealternativemodesforservingforeigncustomers.Empiricalstudiesusuallyfindthatforeignmarketsareservedthroughbothmodesandthatcountriesreceivinghighlevelsofexportsalsohostlargeamountsofforeigndirectinvestment（FDI）.ThispaperevaluatesseveralpossiblewaystoreconcilethefactsaboutFDIandexportswiththestandardtheoryofmultinationalcorporations.WearguethatcoexistenceandcorrelationofFDIandexportsareconsistentwithmodelswherethetwomodesaresubstitutes.Thissubstitutiverelationshipfindscollaborativeevidenceintheresultsofseveralpapers.Nevertheless，asignificantbodyofevidencesuggeststhatFDIsometimescomplementsexportsthroughthemechanismofstimulatingexportsofintermediategoodsforusebyoverseasaffiliates.

I.INTRODUCTION

Multinationalcorporations（MNCs）recurrentlyfacethesamequestion：

shouldwesupplyproductXtocustomersincountryjusingourexistingfactoryincountryiorshouldweinvestinafacilitythatcanmanufacturelocallyincountryj?

Togiveconcreteexamples，bothauthorsofthispaperdrivecarsmanufacturedbyJapanesefirms.OnecarwasproducedatSubaru’smainplantinGunma，Japan，whiletheotherwasproducedinGeorgetown，Kentucky.Theseexamplesmakeitclearthat，atthelevelofanindividualproduct，exportingandforeigndirectinvestment（FDI）arealternativestrategiesforreachinganoverseascustomer.Muchempiricalworkhasstruggledtofindstatisticalevidencetosupportthisproposition.

ThispaperoutlinesalternativetheoriesoftheMNCtoidentifytheeconomicmechanismslinkingFDIandexports.WethenreviewthelargeliteratureonwhetherexportingandFDIaresubstitutesorcomplements.Withfewexceptions，moststudiesfindapositiverelationshipbetweenexportsandFDI.Someauthorsinterpretthisasevidenceof‘complementarity’betweenthetwovariables.Othersquestiontheeconometrics，arguingthatthepositiverelationshipcouldbespurious.Wearguethattheempiricalevidencedoesnotcontradictthetheoreticalpredictionofsubstitutionattheproductlevel.Indeed，substitutionisfoundinstudiesthatfocusonnarrowproductlines.However，manufacturingadownstreamproductinaforeigncountrymayinducetradeinupstreamproductsandtherebyleadtopositivelycorrelatedFDIandexports.TheempiricalliteratureprovidesdirectandindirectsupportforthisformofeconomiccomplementaritybetweenFDIandexports.

WhyshouldwecarewhetherexportsandFDIaresubstitutesorcomplements?

Theredonotappeartobeanyissueswheretheappropriatepolicychoicedependsdirectlyontheanswer.However，governmentsformulatingtradeandforeigninvestmentpoliciesshouldbenefitfromamorecompleteunderstandingofhowMNCsmakeinternationalproductiondecisionsandtheramificationsofthesedecisionsfortheirhomeoperations（includingexporting）.Forexample，wouldalessfavourabletaxtreatmentforincomeearnedbyoverseasaffiliatesstimulatethemanufacturingsectorathome?

TotheextentthatFDIandexportsaresubstitutes，raisingthetaxcostofFDImightinducemoreproductionathome.If，instead，overseasactivitytendstocomplementhomeactivity，itmaynotactuallybeindomesticworkers’intereststodiscourageoverseasinvestment.Thepaperproceedsasfollows.SectionIIbeginswithasimpleexpositionofthestandardmodeloftheFDIversusexportdecisionandidentifiesthekeyfactorsinfluencingthechoice.ThepredictionofthemodelthatfirmswillreachagivenforeignmarketeitherthroughFDIorexportingappearstobeatoddswithcountry-andfirm-levelevidenceillustratedbyfiguresinsectionIII.Inthetwosectionsthatfollow，wediscussextensionsofthestandardmodelofMNCsthatcanexplainthecoexistenceandpositivecorrelationofFDIandexports.InsectionVI，wesurveytheregressionevidence，describingthemethodsusedtoinvestigatetheexport–FDIrelationship.SectionVIIprovidesourconclusionsonwhatlessonsthetheoryoftheMNCshoulddrawfromtheempiricalliterature.

II.SINGLE-PRODUCTMNCSANDTHECASEFORSUBSTITUTION

WebeginwiththesimplestpossiblemodelforconsideringtheformationofanMNC.Thefirmproducesoneproductthatitsellstoconsumersinboththehomecountry（home，H）andtheforeigncountry（foreign，F）.Figure1illustratesthethreeoptionsavailabletothefirm.Thedefaultpositionofafirmasitfirstservesaforeignmarketis‘homecentralization’.Asecondoptionistoopenaplantintheforeigncountrytoservethatmarketwhilecontinuingtoservethehomemarketwiththeoriginalplant.Wecallthis‘replication’becauseitinvolvescreatingareplicaofthehomeplantintheforeigncountry.TheuseofthereplicationformeliminatestheexportsassociatedwithhomecentralizationsincetheMNCnowserveseachmarketlocally.

Athirdoptionistoshutdownthehomeplantanduseanewfactoryintheforeigncountrytomanufactureforbothmarkets.This‘foreigncentralization’forminvolvesimportingbackintothehomemarket.Presumablysomehead-officeactivities（control）remainathomeorelsewewouldjustseethisashomecentralizationfromtheperspectiveoftheforeigncountry.WewillexplicitlyconsideratwostageMNClater，butfornowwewanttoretainthefocusonasinglebusinessunit.

Wecanusealittlebitofalgebratomakeouranalysismoreprecise.Wetakethesizeofeachmarketforthefirm’sproductasgiven，withtherebeingMHcustomersathomeandMFcustomersintheforeigncountry.Wespecifythelevelofdemandineachcountryexogenouslytoavoidhavingtoconsiderpricingdecisions.Theunderlyingassumptionswouldbethateachconsumerhasdemandforoneunitforapricelessthanorequaltohisorhervaluationandzerounitsforanypriceaboveit.Thereforethefirmchargesapriceequaltoeachconsumer’swillingnesstopay.Thismeansthataggregaterevenuesdonotdependonthestrategythefirmselectsandwecanfocusonidentifyingthecost-minimizingstrategy.Thisapproachisfineforoutliningthebasiccaseforsubstitution.However，whenweconsiderthecaseforcomplementarity，itwillproveimportanttoallowfordownward-slopingdemandcurves.

Figure1

ThreeStrategiesforaSingle-productFirm

Weassumethatthereisacompositevariablefactorthatwewillrefertoaslabouranddenotethecompositevariablefactorpricesas‘wages’WHathomeandWFabroad.TheindividualfirmpossessesauniquetechnologyallowingittransformvariableinputsintofinalgoodswithproductivityAthatdoesnotdependonthecountryinwhichitproduces.

Thetotalcostofhomecentralization，CH，isgivenas

CH=（wH/A）MH+（wH/A+TF）MF+K

（1）

WherewH/Aisthemarginalcostofhomeproduction，TFrepresentsthetradecostsincurredinexportingtotheforeignmarket，MFisthesizeoftheforeignmarket，andKrepresentsthefixedcostsofthecapital（land，buildings，equipment）deployedatthehomefactory.Forsimplicity，assumethatthiscostdoesnotdependonwhichcountryafactoryisbuiltin.Ifthefirmweretoopenanotherplant，itwouldhavetoincurthecapitalcoststwice.However，itwouldbeabletoavoidtradecostsbyservingmarketslocally.Thus，thecostsofreplicationaregivenby

CR=（wH/A）MH+（wF/A）MF+2K.

（2）

ShuttingdownthehomeplantandrelocatingittotheforeigncountrycanlowerfixedcostsbacktoK，assumingthatallthehomeinvestmentKisreversible（i.e.thatitcanbesoldatfullvalueorrelocatedatzerocost）.Underforeigncentralization，tradebecomesnecessaryagain.Thecostsofimportingforeign-madeproductsintohome，aredenotedasTH.Thereforethecostsofforeigncentralizationaregivenby

CF=（wF/A）MF+（wF/A+TH）MH+K.（3）

MultinationalbusinessstrategyinthisexamplejustrequiresustocompareCH，CR，andCFandselecttheformthatyieldsthelowestcost.Three-waycomparisonscanbecomplicated，sowewillsetasideforeigncentralizationandconsidertherelativemeritsofhomecentralization（exporting）versusreplication.Homecentralizationispreferredwhen

CR–CH=K+（wF/A）MF–（wH/A+TF）MF>0.（4）

Notethatthecostsofproducingforthehomemarkethavecancelledeachotheroutsincebothformsinvolveusingthehomefactoryforthatmarket.Dividingbythesizeoftheforeignmarket，MF，wecanseethatexportingispreferabletoreplicationwhen

wF/A+K/MF>wH/A+TF.（5）

Figure2graphstheleft-andright-handsidesoftheinequality.Itshowsthatreplicatingoverseasinvestmentcanonlybejustifiedwhentheforeignmarketislargeenough.

Wecansolveforthecriticalmarketsizerequiredtojustifyreplicatinginvestment.ThisistheMF*thatsetsCH=CR：

MF*=K/[TF–（wF–wH）/A].（6）

FirmscentralizeproductionforbothmarketsathomeandexporttotheforeignmarketwhenMF0）.Thelargerthehomecomparativeadvantage，thebiggertheforeignmarketwillhavetobetoju