Economic Growth Foreign Direct Investment and Economic Openness An Empirical Study of the US.docx

Economic Growth Foreign Direct Investment and Economic Openness An Empirical Study of the US.docx

《Economic Growth Foreign Direct Investment and Economic Openness An Empirical Study of the US.docx》由会员分享，可在线阅读，更多相关《Economic Growth Foreign Direct Investment and Economic Openness An Empirical Study of the US.docx（14页珍藏版）》请在冰豆网上搜索。

EconomicGrowthForeignDirectInvestmentandEconomicOpennessAnEmpiricalStudyoftheUS

EconomicGrowth,ForeignDirectInvestmentandEconomicOpenness:

AnEmpiricalStudyoftheU.S.

JeffreyBreece

SchoolofFinance,RenminUniversityofChina,

Beijing,P.R.China

ABSTRACT:

Thispaperanalyzesthecausalrelationshipsbetweeneconomicgrowth,foreigndirectinvestment,andeconomicopennessbyrelyingonU.S.datafrom1960to2010.ThesethreefactorshaveandcontinuetoplayavitalroleinthedevelopmentoftheU.S,mandatinganindepthanalysisoftheirinter-workings.WeestablishaVARframeworktotestforGrangerCausality,inordertoexplorebeyondmerecorrelation.Wefindabilateralcausalrelationshipexistsbetweeneconomicgrowthandeconomicopennessaswellasonefromforeigndirectinvestmenttoeconomicopenness.Weconcludewithseveralpolicyimplicationsthatcanbedrawnfromourfindings.

I.Introduction

TheUnitedStateshastheworld’slargesteconomy,isthelargestrecipientofforeigndirectinvestmentandoneofthemosteconomicallyopencountriesintheworld,accordingtotheHeritageFoundation’sEconomicFreedomoftheWorldIndex.Manyacademiccircleshavenotedthatthisconcurrencecanbynomeansbeacoincidenceandtheremustexistsomeunderlyingconnections.TheseconnectionsmayholdtheanswertohowtheU.S.becamesuchadominantpowerinsuchashortamountoftimeandhowitcanagainenterintoaperiodofprosperity.

Overthelatterhalfofthe20thcenturyandcontinuinguntilthepresentday,thesethreeaspectshaveprogressedinunisonwithoneanother,ascanbeseeninFIGURE1.Especiallyevidentinthedataisthesimilarresponsesofeconomicgrowth（RGDP）,foreigndirectinvestment（FDI）andeconomicopenness（measuredastotaltradetorealgrossdomesticproduct）（EO）torecenteconomicshocks.Allthreesufferedsynchronizedperiodsofnegativegrowthfromthebustoftheboom,asinthe3rdquarterof2001whenRealGDPfell0.3%,economicopennessfell0.033%andFDIfell0.88%.Inaddition,themorerecentrecessioncausedbythesub-primemortgagecrisisisalsovisiblethroughallthreefactors,particularlyFDIwhichappearstobemoresensitivethaneconomicgrowthandeconomicopenness.

FIGURE1

Thetandemgrowthofthesethreefactorssuggestsanobviouscorrelationoverthepast50years,butwhataboutcausation?

Howdotheinterdependentworkingsofeconomicgrowth,foreigndirectinvestmentandeconomicopennessallfunctiontogether?

AndwhatrolewillthisplayinthefuturedevelopmentoftheU.S.?

ThispaperexaminesthepreviousquestionsthroughuseofaVectorAuto-regressionframeworkwithhopesofunderstandinghowthesevariablesarerelatedintheU.S.Theremainedofthepaperisorganizedasfollows.AreviewofrelevantliteratureisputforthinSection2.ThedatausedinthisstudyispresentedinSection3.Section4illustratesthemodelsanddiscussestheempiricalresults.Section5concludes.

II.LiteratureReview

Alargeamountofresearchhasbeendedicatedtowardstherespectivefieldsofeconomicgrowth,foreigndirectinvestmentandeconomicopenness.However,thebulkofthisworkisconcernedwithonlyapair-wiseanalysisofthefactors,andeventhen,itispredominantlyconductedonaseriesofdevelopingcountries.ThespecificcaseoftheU.S.isnormallynotconsidered.Nevertheless,manyrecentfindingsareapplicabletotheU.S.overthepast50years.

Therelationshipbetweenthegrowthrateandinflowofforeigndirectinvestmenthasbeenexploredinamannerofdifferentways.BerthelemyandDemuger（2000）,whileinvestigating24Chineseprovincesfrom1985to1996,useendogenoussimultaneousequationstorevealforeigndirectinvestments’largepositiveinfluenceoneconomicgrowth.Similarly,Alfaroetal.（2000）arriveatasimilarconclusionwhentheymodelasimpleeconomy,findingthatforeigndirectinvestmentplaysanimportantroleineconomicgrowth,butisdependentonthematurityoffinancialmarkets.Otherstudieshavebeenconductedthatsupportasimilarhypothesis;onefindsthatforeigndirectinvestmentcanleadtohighergrowthrates,especiallyincountriesthataremoreeconomicallyopen（Nair-ReichertandWeinhold2000）.AnotherstudyusingGrangerCausalityfindsthatforeigndirectinvestmentcausesgrowththroughknowledgetransfersandtheadoptionofnewtechnology（HansenandRand（2006）.However,therehavebeenpaperswhichadvocatethecounterargument.CarkovicandLevine（2002）,aftercorrectingforbiasesfoundinpreviousstudies,concludethatforeigndirectinvestmentdoesnothaveasignificanteffectongrowth.

Conversely,apaperputforthbyChowdhuryandMavrotas（2006）findthatinthecaseofChile,adevelopingcountryandinmanywayscomparabletotheU.S.inthelatterhalfofthe20thcentury,itisGDPthatcausesforeigndirectinvestment.Theyalsofindevidenceforbi-directionalcausalitybetweenthetwovariablesforothercountries.

Whiletherelationshipbetweeneconomicgrowthandforeigndirectinvestmenthasundergonemuchscrutinyrecently,fordecadesacademicshavebeeninvestigatingeconomicopennessinasimilarcontext.Harrison（1994）,foundapositiveconnectionbetweengrowthandopenness,usingavariedassortmentofopennessmeasures,andfindsthatcausalityrunsinbothdirections.Thisisconducivewiththecommoneconomictheorythathighgrowthprovidesanimportantmomentumfortheopeningupofmarketsandfallinggrowthratesfacilitateariseinprotectionism.InconjunctionwithHarrison,Edwards（1997）andSunetal.（2003）bothreportthatopencountriesregularlyhavehighergrowthrates.However,theymakenoclaimconcerningcausalityforthesefactors.

Theassociationbetweenforeigndirectinvestmentandeconomicopennesshasreceivedlittleattentionuntilrecently,butisbecominganincreasinglyimportantissue.ThisisespeciallytrueinthetimesfollowingtherecentfinancialcrisisandthedevelopingtraderelationswithChina,whereregulationsandpoliciesconcerningeconomicopennessareontheforefrontofpoliticalagendas.Twostudies,conductedbyQuazi（2007）andNajarzadehandShahri（2008）bothfindthatindicesofeconomicfreedom,mostnotablytheHeritageFoundation’sEconomicFreedomoftheWorldIndex,areasignificantandrobustdeterminantofforeigndirectinvestment.Similarly,BengoaandSanchez-Robles（2003）determinethateconomicfreedomhasapositivecausationonthelevelofinflowingforeigndirectinvestment.

Thegoalofthispaperistobuildonpreviousliteraturebyenlargingtheinvestigationofeconomicgrowth,foreigndirectinvestmentandeconomicfreedomfrombi-lateraltosimultaneous.Thisisdoneinwiththeambitionofreinforcingordiscountingpaststudiesonthesubject.Todothiswewillusethedataandempiricalmethodsmentionedbelow.

III.Data

ThedatausedinthisstudyisquarterlydatafromtheU.S.during1960Q1to2010Q3andistakenfromtheFederalReserveBankofSt.Louis.Wetakeeconomicgrowthasrealgrossdomesticproduct（RGDP）inbillionsofchained2005dollars.Foreigndirectinvestment（IFDIAV2）isatwoperiodmovingaverageofthenetinflowofforeigndirectinvestmentandisalsoinbillionsofchained2005dollars.Amovingaverageisusedforforeigndirectinvestmentinordersosmoothoutthedata.Economicopenness（EO）isthesumofnetimportsandexportsdividedbytherealgrossdomesticproducttogetaproportionofthepercentageoftradetoGDP.AlldataisseasonallyadjustedandissummarizedinTABLE1.

TABLE1

Variables

Observations

Mean

S.D.

Minimum

Maximum

RGDP

202

744.969

3241.007

2802.616

13363.49

IFDIAV2

202

87.502

132.090

-70.132

729.494

EO

202

0.156

0.0683

0.071

0.292

WethenconductanAugmented-DickeyFullerTesttotestifRGDP,IFDIandEOareallstationaryserieswiththenullhypothesisthatthevariableshaveaunitrootandarestationary.TheresultsaredisplayedinTABLE2.

TABLE2

Variables

ADF-statistic

10%

5%

1%

P-value

Judgment

RGDP

0.883

-2.574

-2.876

-3.463

0.995

AcceptNull

IFDIAV2

-1.230

-2.575

-2.876

-3.464

0.6612

AcceptNull

EO

1.149

-2.574

-2.876

-3.46

0.998

AcceptNull

FromTABLE2wecanconcludethatRGDP,IFDIAV2andEOareallnotstationary.Therefore,weconductthesameAugmented-DickeyFullerTestontheirgrowthrates（firstorderconditions）,whichareRGDPGR,IFDIAV2GRandEOGRrespectively.AgraphofthefirstorderconditionscanbeseeninFIGURE2andtheoutcomeoftheADFTestareinTABLE3.

FIGURE2

TABLE3

Variables

ADF-statistic

10%

5%

1%

P-value

Judgment

RGDPGR

-6.797

-2.57

-2.876

-3.463

0.000

RejectNull

IFDIAV2GR

-8.542

-2.575

-2.876

-3.464

0.000

RejectNull

EOGR

-20.301

-2.574

-2.876

-3.463

0.000

RejectNull

WecanseefromFIGURE2andTABLE3thatRGDPGR,IFDIAV2GRandEOGRallhavep-valuesof0.000,meaningthattheyarecanrejectthenullhypothesiswithabsoluteconfidence.Nowthatthevariablesarestationarywecanusetheminourmodels.

IV.EmpiricalMethodology

A.）BestLagOrderSelection

InordertogenerateaVARmodelthatwillprovidereliableresultsweneedtochoosethelagorderthatensuresthattheerrortermsinthemodelwillsatisfythevectorwhitenoiseprocess.Todeterminethelagorderwegiveamaximumlagorderof8（2years）andchoosethelagorderindicatedbythemajorityofcriterions.TheresultscanbeseeninTABLE4.

TABLE4

 Lag

LR

FPE

AIC

SC

HQ

0

NA 

 1.386131

 8.840147

 8.890863

 8.860685

1

 95.45931

 0.918257

 8.428336

  8.631197*

 8.510488

2

 33.90770

 0.840099

 8.339301

 8.69