外商直接投资能长期促进经济增长吗？【外文翻译】.doc

1、0本科毕业论文外文翻译外文题目DOSEFOREIGNDIRECTINVESTMENTALWAYSENHANCEECONOMICGROWTH出处KYKLOS,VOL562003FASC4,491508作者JOZEMENCINGER原文DOESFOREIGNDIRECTINVESTMENTALWAYSENHANCEECONOMICGROWTHJOZEMENCINGERINTRODUCTIONFORINTERNATIONALFINANCIALINSTITUTIONS,POLITICIANS,ANDTHEVASTMAJORITYOFECONOMISTSFOREIGNDIRECTINVESTMENTFD

2、IAPPEARSTOBEASORTOFPANACEAFOREVERYECONOMICPROBLEMINTHEEMERGINGMARKETECONOMIESITSPOSITIVEIMPACTONECONOMICGROWTHHASACQUIREDTHESTATUSOFCONVENTIONALFACTECONOMICTHEORYNAMELYSUGGESTSTHATUNFETTEREDINTERNATIONALCAPITALFLOWSFOSTEREFFICIENTALLOCATIONOFRESOURCES,WHICHBYITSELFSHOULDPROMOTEGROWTHTHEECONOMICBENEF

3、ITSOFFDIARECONSIDEREDTOBETWOFOLDFIRST,FDICANHELPCOUNTRIESIFDOMESTICSAVINGSAREINSUFFICIENTTOFINANCEECONOMICEXPANSIONSECONDLY,AFOREIGNCORPORATEPRESENCEISASSOCIATEDWITHPOSITIVEEXTERNALITIESTHEALMOSTDESPERATEEFFORTSOFMANYCOUNTRIESTOATTRACTASMUCHFDIASPOSSIBLEINDIRECTLYSUPPORTTHETHEORYHOWEVER,SUBSTANTIALG

4、AINSOFINWARDFDIFORTHEHOSTCOUNTRIESHAVEBEENMUCHMOREASSERTEDTHANCONFIRMEDBYEMPIRICALEVIDENCETHERESULTSOFARAPIDLYGROWINGNUMBEROFEMPIRICALSTUDIESONTHERELATIONBETWEENFDIANDECONOMICGROWTHDIFFER,ALTHOUGHMOSTSTUDIESSTARTWITHESSENTIALLYTHESAMEBENCHMARKCROSSCOUNTRYGROWTHMODELTHEDIFFERENCESINTHESETSOFTHECOUNTR

5、IESINCLUDED,SAMPLEPERIODS,DATA,ANDESTIMATIONTECHNIQUESHAMPERCOMPARISONSACROSSTHESTUDIESEDISONETAL2002INMANYSTUDIESDEALINGWITHSUBSETSOFTHECOUNTRIES,FDIORFDIINCOMBINATIONWITHSOMEOTHERFACTORORFACTORS,ISPOSITIVELYRELATEDTOGROWTH,WHILESEVERALSTUDIESRODRIK1998,GRILLIANDMILESIFERRETTI1995,KRAAY11998HAVEFOU

6、NDNOSIGNIFICANTRELATIONSHIPBETWEENFDIANDGROWTHMOSTSTUDIESHAVESTRESSEDTHEDIFFERENCESAMONGTHESETSOFCOUNTRIESINCLUDEDREGARDINGTHEIRTRADEPOLICIES,INSTITUTIONALCHARACTERISTICS,ORLEVELOFDEVELOPMENTMORETHANTWODECADESAGOBHAGWATI1978SUGGESTEDTHATTHEIMPACTOFFDIONGROWTHDEPENDSONTHETRADEPOLICYOFAHOSTCOUNTRYINEX

7、PORTPROMOTINGCOUNTRIESFDIWOULDINCREASEGROWTH,WHILEITWOULDHAVENOIMPACTINACOUNTRYWITHANIMPORTSUBSTITUTIONTRADEPOLICYHISHYPOTHESISWASTESTEDBYBALASUBRAMANAYAM,SALISUANDSAPSFORD1996THEOUTCOMEWASMIXEDBLOMSTRM,LIPSEYANDZEJAN1994FOUNDTHATFDIONLYPROMOTESGROWTHINHIGHERINCOMEDEVELOPINGCOUNTRIESEMPIRICALSTUDIES

8、ONTHESUBJECTHAVETHEREFORENOTREFUTEDTHESTATEMENTTHATINGENERAL,THERESULTSOFTHESESTUDIESINDICATETHATTHESIZEOFINWARDFDISTOCKSORFLOWS,RELATIVETOGDP,ISNOTRELATEDINANYCONSISTENTWAYTORATESOFGROWTHLIPSEY2002,P55ASIZABLELITERATUREONFDIINTRANSITIONECONOMIESCANBELOOSELYDIVIDEDINTOTHESTUDIESDEALINGWITHTHEDETERMI

9、NANTSOFFDIANDTHOSEDEALINGWITHTHEIMPACTSOFFDIONECONOMICPERFORMANCEMOSTOFTHESTUDIESUSEDMICROECONOMICDATAANDDEALTWITHMICROECONOMICISSUESBARRELANDHOLLAND2000,BEVANANDESTRI2000,KONINGS2001,DAMIJANETAL2001ANDONLYAFEWWITHIMPACTSOFFDIONMACROECONOMICPERFORMANCEOFTRANSITIONCOUNTRIESCAMPOSANDKINOSHITA2002,KRKO

10、SKA2001,LIPSCHITZETAL2002WHICHARECONSIDEREDHEREINDEED,WEONLYWANTTOANSWERTWOQUESTIONSFIRSTDIDFOREIGNTAKEOVERS,APREDOMINANTFORMOFFDIINDEVELOPEDTRANSITIONCOUNTRIES,ENHANCETHEIRECONOMICGROWTHDURINGTHEPOSTTRANSITIONPERIODSECONDLYIFNOT,WHATWERETHEREASONSDATAWESHALLCONCENTRATEONAVERYNARROWSAMPLEBOTHREGARDI

11、NGTHESETOFTHECOUNTRIESANDTHEPERIODINCLUDEDTHESAMPLECONSISTSOFEIGHTEUCANDIDATESCZECHREPUBLIC,ESTONIA,HUNGARY,LATVIA,LITHUANIA,POLAND,SLOVAKIA,ANDSLOVENIAINTHEPOSTTRANSITIONPERIOD,WHICHISALSOTHEPERIODOFTHEIRGRADUALACCESSIONTOEUANDHIGHRELIANCEONFDITHENARROWINGOFTHESAMPLETOONLYEIGHTCOUNTRIESANDTOTHEPOST

12、TRANSITIONPERIODONACCOUNTOFTHENUMBEROFOBSERVATIONSISDELIBERATEITISBASEDONEXPERIENCESWITHTRANSITIONPROCESSESINDIFFERENTCOUNTRYGROUPSFIRST,WETRYTOOBSERVEASETOFCOUNTRIESWITHCOMMONINSTITUTIONALANDCULTURALCHARACTERISTICS,APARTICULARGROWTHEXPERIENCE,ANDATAPARTICULARLEVELOFDEVELOPMENTONLYTHESEEIGHTOUTOF25T

13、RANSITIONCOUNTRIESWERE2ACKNOWLEDGEDASFUNCTIONINGMARKETECONOMIESBYEUSTANDARDSANDBECAMECANDIDATESFORENTRYIN20041SECONDLY,TRANSITIONBROUGHTAFUNDAMENTALBREAKINTHEWAYINWHICHECONOMIESFUNCTIONANDTHETRANSITIONPERIODINTHENARROWSENSESHOULDNOTBEINCLUDEDINTHESAMPLE2WITHBOTHRESTRICTIONSWEGETAMUCHNARROWERSETOFTRA

14、NSITIONCOUNTRIESTHANTHEONEUSEDBYCAMPOSANDKINOSHITA2002LETUSBEGINBYEXPLORINGTIMESERIESDATAFOREACHCOUNTRY,ANDCROSSSECTIONDATAFOREACHYEARBYOBSERVINGSIMPLERELATIONSHIPSBETWEENTHEGROWTHOFGROSSDOMESTICPRODUCTRGDP,ANDTHERATIOOFFOREIGNDIRECTINVESTMENTANDGROSSDOMESTICPRODUCTFDI,THELATTERREPRESENTINGALSOAKIND

15、OFREVEALEDLIBERALIZATIONOFCAPITALACCOUNTTHEAVERAGEFDIINTHE19942001PERIODWAS414PERCENTRANGINGBETWEEN605PERCENTINESTONIAAND132PERCENTINSLOVENIA,WHILETHEGROWTHOFGDPBY365PERCENTAYEARWASRANGINGBETWEEN487PERCENTINPOLANDAND197PERCENTINCZECHREPUBLICGROWTHWASMOSTSTABLEINSLOVENIAANDMOSTUNSTABLEINLITHUANIACORR

16、ELATIONCOEFFICIENTSBETWEENGROWTHANDCORRESPONDINGFDIARENEGATIVEINSEVENOUTOFTHEEIGHTCOUNTRIESANDPOSITIVEINLITHUANIAIFDATAAREPOOLED,THESIMPLECORRELATIONCOEFFICIENTREMAINSNEGATIVEASIMILARNEGATIVERESULTISOBTAINEDBYOBSERVINGCROSSSECTIONDATAFOREACHYEARINTHEOBSERVEDPERIODTHEREISANEGATIVECORRELATIONBETWEENGR

17、OWTHANDFDIINSIXOUTOFEIGHTYEARS,POSITIVECORRELATIONIN1997,ANDINSIGNIFICANTIN1998TABLE1ALSOINDICATESTHATSLOVENIA,WITHBYFARTHESMALLESTSHAREOFFOREIGNDIRECTINVESTMENTINGDP,HADANECONOMICGROWTHABOVETHEAVERAGEANDALSOVERYSTABLETHISCANBEEXPLAINEDBYTHEMUCHBETTERINITIALCONDITIONSOFTHECOUNTRYATTHEBEGINNINGOFTRAN

18、SITIONONECOULDTHUSCLAIMTHATECONOMICGROWTHANDCONVERGENCEOFTHECOUNTRYTOTHEEUWOULDBEEVENFASTERWITHMOREFOREIGNDIRECTINVESTMENTTABLE1FOREIGNDIRECTINVESTMENTANDGDPGROWTHCOUNTRYTIMESERIES,19942001COUNTRYAVERAGEFDISTANDARDDEVIATIONAVERAGERGDPSTANDARDDEVIATIONCORRELATIONCOEFFICIENTCZECH5803611972900183REPUBL

19、ICESTONIA603296420397029HUNGARY437241355150069LATVIA605202386347007LITHUANIA343247203583035POLAND273158487189061SLOVAKIA344368444187075SLOVENIA132075426081064ALLCOUNTRIES414295365316018TABLE2FOREIGNDIRECTINVESTMENTANDGDPGROWTHANNUALCROSSSECTIONDATAYEARAVERAGEFDISTANDARDDEVIATIONAVERAGERGDPSTANDARDDE

20、VIATIONCORRELATIONCOEFFICIENT1994303343116510021199537532640426705719962872194231590601997326259594341031199850434036724300119994613051443060482000538333454165042200147923941919901319942001414295365316018THECOUNTRYTIMESERIESANDANNUALCROSSSECTIONDATASERIESEACHCONTAINJUSTEIGHTOBSERVATIONSWHICHDOESNOTA

21、LLOWAPROPERSTATISTICALANALYSIS,MUCHLESSTHEESTABLISHMENTOFACAUSALRELATIONSHIPFORTHATREASONTHERELATIONSHIPBETWEENFOREIGNDIRECTINVESTMENTANDECONOMICGROWTHWASEXAMINEDBYPANELDATAEIGHTCANDIDATESFORTHEPERIOD19942001WHICHPROVIDES64OBSERVATIONSTWOSURPRISINGRESULTSEMERGEDFIRST,ANEGATIVECORRELATIONBETWEENGROWT

22、HANDFDISECOND,THESIGNIFICANCEOFTHENEGATIVESIGNSTRENGTHENEDIFFDIWASLAGGEDSEEFIGURE1IMPLYINGTHATFDIHAMPEREDGROWTHANDREALCONVERGENCEOFTHECANDIDATECOUNTRIESWITHEUTHISCONTRADICTSTHEORYANDINTUITION,ANDHASNOTBEENSUPPORTEDBYEMPIRICALSTUDIESCAUTIOUSNESSTHEREFORESEEMSAPPROPRIATEBEFOREANYSUBSTANTIALINFERENCEIS

23、MADERESULTSSHOULDBESTATISTICALLYTHOROUGHLYTESTEDTHEPOOLEDDATAALLOWTHEAPPLICATIONOFTHESTANDARDGRANGERCAUSALITYTESTSSEEHOLTZEAKINETAL1985THERESULTSFORONEANDTWOLAGSAREINTABLE3FIGURE1GROWTHANDFOREIGNDIRECTINVESTMENTINEIGHTCANDIDATECOUNTRIES,199420014USINGTHERESULTSINTABLE3,AZEROHYPOTHESISTHATFDIDOESNOTI

24、NFLUENCEECONOMICGROWTHCANBEREJECTEDWHICHIMPLIESADOPTIONOFTHEOPPOSITEHYPOTHESIS,IE,THATFDIAFFECTSECONOMICGROWTHSIMILARLY,ANONREJECTIONOFTHEZEROHYPOTHESISTHATECONOMICGROWTHDOESNOTINFLUENCEFDIIMPLIESTHATWECANNOTADOPTTHEOPPOSITEHYPOTHESISTHATECONOMICGROWTHAFFECTSFDIINSHORT,THEASSUMPTIONTHATFDIAFFECTSECO

25、NOMICGROWTHISCONFIRMED,REVERSECAUSALITY,IETHATGROWTHATTRACTSFDI,ISREJECTEDTABLE3GRANGERCAUSALITYTESTSSAMPLE164LAGS1NULLHYPOTHESISOBSERVATIONSFSTATISTICSPROBABILITYGDPDOESNOTGRANGERCAUSEFDI/GDP56065838042067FDI/GDPDOESNOTGRANGERCAUSEGDP496610003011SAMPLE164LAGS2NULLHYPOTHESISOBSERVATIONSFSTATISTICSPR

26、OBABILITYGDPDOESNOTGRANGERCAUSEFDI/GDP48115481032470FDI/GDPDOESNOTGRANGERCAUSEGDP48864001226DESPITESIGNIFICANTSTATISTICSOFTHESTANDARDGRANGERCAUSALITYTESTSTHEREARETWOMAJORDRAWBACKSFIRST,POOLINGIMPLIESTHATTHEUNDERLYINGCAUSALSTRUCTUREISTHESAMEFOREACHCROSSSECTIONALUNITOREACHCANDIDATECOUNTRYIFTHISWERENOT

27、THECASE,WEWOULDGETINCONSISTENTESTIMATESSECOND,THESTANDARDGRANGERCAUSALITYTESTDOESNOTPROVIDEFORTHESIGNOFTHERELATIONSHIPSIMSVERSIONOFTHECAUSALITYTESTWHICHALLOWSFORESTIMATINGINDIVIDUALEFFECTSINTHEFORMOFACOUNTRYSPECIFICINTERCEPTORSLOPEISTHEREFOREAPPLIEDBYESTIMATINGTHEFOLLOWINGTWOEQUATIONSRGDPA0A1FDI1A2F

28、DIA3FDI1A4GDP0AJDUMJ1FDIB0B1RGDP1B2RGDPB3RGDP1B4GDP0BJDUMJ2WHERERGDPISTHEGROWTHRATEOFGROSSDOMESTICPRODUCT,FDIISTHESHAREOFFOREIGNDIRECTINVESTMENTINGDP,GDP0ISTHEINITIALGDPPERCAPITAINPERCENTAGEOFEUAVERAGE,AND5DUMJARECOUNTRYDUMMIESTHERESULTSAREINTABLE4COUNTRYSPECIFICSWERELARGELYABSORBEDBYGDP0UTTERINGTHE

29、LEVELOFDEVELOPMENTTHECOUNTRYDUMMIESWERETHEREFOREINSIGNIFICANTEXCEPTFORESTONIAANDPOLANDINTHEFIRSTEQUATION,ANDCZECHREPUBLICANDPOLANDINTHESECONDEQUATIONINSHORT,THERESULTSCONFIRMTHECAUSALITYLINKFOUNDBYTHESTANDARDGRANGERTESTTHETVALUEOFTHECOEFFICIENTOFTHELAGGEDFDIRATIOISSIGNIFICANTATTHE1PERCENTLEVELANDHIG

30、HERTHANTHETVALUEOFTHECOEFFICIENTSFORLEADANDCURRENTFDI,ANDTHEOPPOSITEISTRUEFORTHECOEFFICIENTSFORLAGANDLEADOFRGDPTABLE4CAUSALITYTESTSRGDPFFDI,XJAITAFDIGRGDP,XJBITBCONSTANT1015539CONSTANT13199723FDI10447310RGDP10183154FDI0014009RGDP0133120FDI10236161RGDP10437295GDP00075237GDP00136217CZECHREPUBLIC273619

31、1ESTONIA1366116POLAND2023163POLAND3112254R2ANDDW03477135R2ANDDW0433202TOESTIMATETHEROBUSTNESSOFTHECAUSALITYTESTSANDOFTHECOEFFICIENTSINTHESIMPLEREGRESSIONMODELWITHGROWTHBEINGTHEDEPENDENTANDFDITHEINDEPENDENTVARIABLE,EIGHTSUBSAMPLESINEACHONECOUNTRYWASOMITTEDWEREFORMEDTHERESULTSAREINTABLE5TABLE5GRANGERC

32、AUSALITYTESTSANDSIMPLEREGRESSIONONNARROWERSETS6NULLHYPOTHESISRGDPDOESNOTCAUSEFDIRGDPABFDIRGDPABFDI1FDIDOESNOTCAUSERGDPLAG1LAG2CONSTSLOPECONSTSLOPEEIGHTCOUNTRIES05949111147144366019145019702726WITHOUTCZECHR007753014605448630151054010703130WITHOUTESTONIA28433833518344364020144739102423WITHOUTHUNGARY05

33、441111937943659017125038902823WITHOUTLATVIA009661046684447660221650210103029WITHOUTLITHUANIA42111627327052092030284818902019WITHOUTPOLAND03933009343941155015104858502321WITHOUTSLOVAKIA01345206143941453014095048502723WITHOUTSLOVENIA04652712940544053018125338503126DESPITETHEROBUSTNESSOFTHENEGATIVESIGN

34、FORFDIANDTHESOLIDREJECTIONOFTHEREVERSEDCAUSALITYTHERESULTSCOULDBEDISPUTEDBYTHEARGUMENTSOFFDIENDOGENEITYANDABIASDUETOOMITTEDVARIABLESTHESEISSUESAREDEALTWITHINTHEFRAMEWORKOFASTANDARDNEOCLASSICALGROWTHMODELUSEDBYMOSTRESEARCHERSINTHEFIELDINTHESOLOWTYPEGROWTHMODELS,FDIISCONCEIVEDASANADDITIONTOTHECAPITALS

35、TOCKOFTHEHOSTECONOMYFDIISTHUSCONSIDEREDASUBSTITUTEFORDOMESTICCAPITALORAPOSITIVECOMPLEMENTASTANDARDPRODUCTIONFUNCTIONDERIVEDFROMTHEAUGMENTEDSOLOWTYPEMODELANDUSEDBYMANKIW,ROMERANDWEIL1992ANDEASTERLY2001OFTHEFOLLOWINGFORMRGDPFGDP0,RINV,REMP,FDI,REU,DUMMYVARIABLES3WITHRINVREPRESENTINGGROWTHOFGROSSFIXEDI

36、NVESTMENTREMPGROWTHOFEMPLOYMENT,ANDREUGROWTHOFGDPINEUCOUNTRIESPROVIDEDTHERESULTSWHICHAREINTABLE6TABLE6THEAUGMENTEDPRODUCTIONFUNCTION,199420017123456789CONSTANT650503572445694374649455606470665340708463808365GDP000512370037175005123700542140044208005521100552460064241RINV01164300121465011541801314830

37、109392012946101184320132475REMP02411800250194024317902622040319261026220102751930329237FDI102742850298322202772840263272023921702962920267239FDI021822800530490079072REU015703400020010065013DUMMYFORESTONIA18702306139215514121541452160DUMMYFORLITHUANIA129512312631181466136RGDP1008007170143133R20527057

38、4052705870478059005320605DW157151156157157158NOTETSTATISTICSAREINPARENTHESISTHENEGATIVECOEFFICIENTFORFDI1REMAINSSIGNIFICANTLYNEGATIVEINALLSPECIFICATIONSOFTHEPRODUCTIONFUNCTIONANDALLOTHERCOEFFICIENTSAREINACCORDANCEWITHTHETHEORETICALEXPECTATIONSIETHENEGATIVESIGNOFINITIALGDP/CAPITAISINACCORDANCEWITHTHE

39、THEORYOFCONVERGENCECOUNTRYSPECIFICEFFECTSEXPRESSEDBYDUMMYVARIABLESAREINSIGNIFICANTEXCEPTFORESTONIAANDLITHUANIARESULTHOWCANONEFINDAREASONABLEEXPLANATIONOFTHISRESULTITISSTATISTICALLYRATHERROBUSTBUTCONFLICTSWITHTAKENFORGRANTEDWISDOMACCORDINGTOWHICHFDIENHANCESECONOMICGROWTHBECAUSEITSUBSTITUTESDOMESTICSA

40、VINGSANDINVESTMENT,ANDBECAUSEITBRINGSPRODUCTIVITYSPILLOVERFROMFOREIGNOWNEDTOLOCALFIRMS未完8译文外商直接投资能长期促进经济增长吗JOZEMENCINGER引言对国际金融机构，政治家和大多数经济学家而言，外商直接投资似乎是新兴市场经济中每个经济问题的灵丹妙药；其对经济增长的积极影响已得到了传统事实的认可。经济理论表明，不受约束的国际资本流动可以促进资源的有效配置，这本身就是促进经济的增长。外商直接的经济效益被认为是双重的。首先，外商直接投资可以帮助那些国内储蓄不足的国家来筹集经济扩张资金；其次，国外企业的存在是

41、具有积极的外部效应。许多国家不顾一切的努力去吸引尽可能多的外商直接投资间接的支持了这一理论。然而，对东道国来说，实质上外商直接投资的声称的实质收益要比实证研究证实的多。尽管大多数研究都是基于相同的跨国增长模型，但基于FDI与经济增长的大量快速增长的实证研究的验证结果有所不同，国家设置的不同包括，样品周期，数据以及估计技术阻碍比较（爱迪生等，2002）。在很多研究中发现一些国家的FDI或FDI相关的一些因素与经济增长呈正相关，然而一些研究（克里克1988，GRILLI和MILESIFERRETTI1995,KRAAY1988）发现FDI与经济增长无明显关系。大多数研究强调的是各个国家间的差异包括

42、贸易政策，体制特征或是发展水平。二十多年前巴格瓦蒂就指出FDI对经济增长的影响取决于东道国的贸易政策；在出口导向的国家FDI会促进经济的增长，而在实行进口替代贸易政策的国家FDI对经济增长没有影响。他的假设由BALASUBRAMANAYAM,SALISU和SAPSFORD验证，而这个结果是多样的。BLOMSTRM,LIPSEY和ZEJAN发现FDI促进经济增长仅适用于高收入的发展中国家。关于这个问题的实证研究没有反驳声明一般来说，相对于国民生产总值，实证研究的结果表明FDI股票的回笼或流动，与经济增长率的任何一贯的方式无关（利普西2005，第5页）。对FDI在经济转型期的研究的相当多的著作可以

43、粗略地分为对FDI的决定因素的研究和FDI对经济表现的影响的研究。大多数研究使用的是微观数据和处理微观问题的方法（BARRELANDHOLLAND2000,BEVANANDESTRI2000,KONINGS2001,DAMIJANETAL2001，在这里只有少数研究FDI对转型国家宏观经济表现的影响CAMPOSANDKINOSHITA2002,KRKOSKA2001,LIPSCHITZETAL2002。事实上，我们只要回答两个问题9第一，是否外国接管是成熟的转型国家在后转型期用于提高经济增长的主要方式第二，如果不是，那原因是什么数据我们应该集中精力于包括关于国家的装置和时期的一个极小的样本上。

44、将一系列的国家和一个周期集中在一个小的样本上。这个样本包括8个欧盟的候选国捷克共和国，爱沙尼亚，匈牙利，拉托维亚，立陶宛，波兰，斯洛伐克和斯洛文尼亚，在过度后时期，同时也是他们逐渐入职欧盟和高度依赖FDI的时期。因为对数量的观察，样本缩小到8个国家和过度后时期，这是经过深思熟虑的。它的基础是不同国家集团在过度后时期的经验。首先，我们试着去观察有着共同体制和文化特征，一个特殊增长经历和处于一个特殊发展水平一系列国家。只有这8个国家从25个转型国家脱颖而出被欧盟标准公认为市场经济并且成为20041项中的候选人。其次，转型带来了根本的突破，从狭义上来说，经济运作和的转型期都不应该包括样本2上。由于这

45、两方面的限制，我们得到了比CAMPOSANDKINOSHITA2002还要狭窄的转型国家。我们通过探究每个国家的时间序列数据，和通过观察国内生产总值增长和外商直接投资率和国内生产总值的比率的简单关系探索各个国家每年的截面数据，后者代表的是一种相对自由的资本账户。在19942001年外商平均直接投资在爱沙尼亚是414到605，在斯洛文尼亚是132，而每年的GDP增长率在波兰是365到487，在捷克共和国是197。经济增长最稳定在斯洛文尼亚，最不稳定的在立陶宛。经济增长和FDI的相关系数在8个国家中有7个国家为负，只有立陶宛为正的；如果数据汇集，简单的相关系数仍然为负。通过观察每年的截面数据获得的

46、结果是负的，在8年里有6年经济增长和FDI的呈负相关，在1997年呈现的是正相关，而1998年又是微不足道的。表1表明斯洛文尼亚外商直接投资占GDP的比重最小，可它超过了平均经济增长并且很稳定。这可以被解释为在转型之初国家的初始条件很好，有人因此宣称有更多的外商直接投资会使经济更快增长和更快加入欧盟。表1外商直接投资与经济增长（国家时间序列，19942001）国家平均FDI标准差平均RGDP标准差相关系数捷克共和国580361197290018爱沙尼亚603296420397029匈牙利437241355150069拉托维亚60520238634700710立陶宛343247203583035

47、波兰273158487189061斯洛伐克344368444187075斯洛文尼亚132075426081064所有国家414295365316018表2外商直接投资与GDP增长（截面数据）年份平均FDI标准差平均RGDP标准差相关系数1994303343116510021199537532640426705719962872194231590601997326259594341031199850434036724300119994613051443060482000538333454165042200147923941919901319942001414295365316018国家时间序列和

48、截面数据系列每个都仅包含了8个国家的观测值，没有适当的统计分析，更不用说因果关系的建立。由于这个原因，外商直接投资与经济增长可以用面板数据来说明（8个候选国在19942001）提供了64个观察值。出现了两个惊人的结果第一，经济增长和FDI是负相关；第二，如果FDI滞后显示FDI会阻碍其经济增长以及妨碍这些国家加入到欧盟当中，那么负相关的影响便会加强。这种理论与直观认识相违背，并且得不到实证研究的支持。因此之前看起来谨慎的任何实质性的结论都应该经过统计测试。集中的数据可以用格兰杰因果关系检验（霍尔茨埃金等人，1985）滞后量一、二的结果在表3中。图18个国家经济增长和外商直接投资，1994200

49、1表3的结果表明，假设零FDI不会影响经济增长，会被其相反的假设所拒绝，11即意味着FDI会影响经济增长。同样，相反的零假设经济增长不影响FDI，我们也不能采取相反的假设即经济增长影响FDI。总之，FDI对经济增长影响的假设是被证实的，而经济增长会吸引FDI会被反驳。表3格兰杰因果检验样本164滞后量1NULLHYPOTHESISOBSERVATIONSFSTATISTICSPROBABILITYGDPDOESNOTGRANGERCAUSEFDI/GDP56065838042067FDI/GDPDOESNOTGRANGERCAUSEGDP496610003011样本164滞后量2NULLHYPOTHESISOBSERVATIONSFSTATISTICSPROBABILITYGDPDOESNOTGRANGERCAUSEFDI/GDP48115481032470FDI/GDPDOESNOTGRANGERCAUSEGDP48864001226尽管格兰杰因果关系检验的标准统计还是有其两大缺点。首先，把数据汇集起来意味着对每个截面单位或每个国家有着相同的结构，但如果不在这种情况下，我们会得出不一致的估计。其次，格兰杰因果检验不能提供这种关系的迹象。西姆斯的因果关次检验适用于估计“个别效应”在一个国家具体的截取或截留数据的形式下估算，通过下列两个方程RG