JHD_DP_2011-002

1 Discussion Paper Series Fifty Years of International Business Theory and Beyond FIFTY YEARS OF INTERNATIONAL BUSINESS THEORYAND BEYOND A.Rugman University of Reading, UK A.Verbeke University of Calgary, Canada Q.Nguyen University of Reading, UK The aim of this series is to disseminate new research of academic distinction in the fields of international business and strategy. Papers are preliminary drafts, circulated to stimulate discussion and critical comment. Publication in the series does not imply that the content of the paper reflects the views of Henley Business School, the John H. Dunning Centre or the University of Reading. John H. Dunning Centre for International BusinessDiscussion Paper No. 2011-001 June 2011 [email protected] www.henley.reading.ac.uk/dunn 2 FIFTY YEARS OF INTERNATIONAL BUSINESS THEORYAND BEYOND Alan M. Rugman (Corresponding author) Professor of International Business Henley Business School, School of Management University of Reading Henley on Thames, Oxon, RG9 3AU, England E-mail: [email protected] Alain Verbeke Professor of Strategy and Global ManagementMcCaig Chair in Management Haskayne School of Business University of Calgary Quyen T.K. Nguyen Henley Business School, School of Management University of Reading Submitted to Global Strategy Journal Acknowledgements:wearepleasedtoacknowledgehelpfulcommentsfromtwo referees,andfromProfessorsPeterBuckley,MarkCasson,Michael-JoergOesterle,and Joachim Wolf. Wealso receivedhelpfulcommentsfromparticipants atseminarsat the University of Leeds, York University and the University of Reading. Version: March 31, 2011 Accepted for publication in Management International Review, volume 51, 2011 3 ABSTRACT As the field of international business has matured, there have been shifts in the core unit of analysis.First, there was analysis at country level, using national statistics on tradeandforeigndirectinvestment(FDI).Next,thefocusshiftedtothe multinationalenterprise(MNE)andtheparentsfirmspecificadvantages(FSAs).Eventually the MNEwasanalysedasanetworkandthesubsidiarybecameaunitof analysis.We untangle the last fifty years of international business theory using a classification by these three units of analysis. This is the country-specific advantage (CSA) and firm-specific advantage (FSA) matrix. Will this integrative framework continue to be useful inthefuture?WedemonstratethatthisislikelyastheCSA/FSAmatrixpermits integrationofpotentiallyusefulalternativeunitsofanalysis,includingthebroad region of the triad. Lookingforward,we developanewframework,visualizedintwomatrices,to show how distance really matters and how FSAs function in international business. Key to thisaretheconceptsofcompoundeddistanceandresourcerecombinationbarriers facing MNEs when operating across national borders. Keywords:multinationalenterprises(MNEs);firmspecificadvantages(FSAs);country specific advantages (CSAs); compounded distance; subsidiary; network; regions; resource recombination barriers; theory 4 INTRODUCTION In this paper, we examine the literature on international business (IB) over the last fifty years. We do this in the first three sections. In the second half of the paper, we develop new frameworks to analyze unresolved issues in the theoretical and empirical literature of IB that will require research in the future. As the field of IB has matured, there have been shifts in the core unit of analysis. In the pre-Hymer(1960)era,internationaleconomistsdominatedthefieldandfocusedon nationalcompetitivenessatthecountrylevel,usingnationalstatisticsontradeand foreign investment.During the 1970s, the focus shifted to foreign direct investment by themultinationalenterprise(MNE)andthetransferacrossbordersofitsfirmspecific advantages (FSAs), both stand-alone competences (such as patented R&D knowledge and brand names) and higher-order capabilities.As of the 1980s, more attention was devoted to the MNE as a differentiated network with the MNE subsidiary as the unit of analysis.Inaddition,insidetheMNE,substantialacademicworkwasperformedfocusingon entrepreneurial growth in specific cultural, economic and institutional contexts. Finally, a number of studies focused more on clusters and networks of independent companies.The IB literature during the past 50 years has used various units of analysis. Mainstream neoclassical international economics builds upon the strong assumption that differences infactorendowmentsacrossborderswillleadtointernationaltransactions,whether transfersofcapitalorgoods.Inotherwords,itisassumedthatthereisnotreallyan organizational challenge to be addressed in creating an efficient system of international exchange.Vernon(1966)andDunning(1958)extendthiswork,butrecognizethe importanceoffirmsintheirpioneeringIBstudies,whichrepresentthefirststageof modernIBanalysis. Vernonsproductlifecycleframework,published in1966,basically arguethattheUnitedStateshasatechnology-related,countryspecificadvantage(CSA) embeddedinUSownedmultinationalenterprises(MNEs).TheseUSparentfirmscreate miniaturereplicas(branchplants)inCanadaandWesternEuropesuchthattechnology would be transferred from the parent firm to these foreign subsidiaries. This leads to an indirecttransferoftechnologytohostcountriesandtootherbenefitsofforeign ownership.IntheUKcontext,Dunning(1958)observesthattheUKsubsidiariesofUS MNEsupgradethemacroeconomicinfrastructureastheymanufacturetechnologically intensiveproductsandservices;theyupgradeUKjobs;theypaytaxes;theyincrease productivityandotherwiseimprovetheCSAsoftheUnitedKingdom.Rugman(1980b) 5 confirmsthisoutcomeinaCanadiancontext,withCanadiansubsidiariesofUSMNEs providing net economic benefits to Canada. Clearly,theunderlyingunitofanalysisintheworkbyVernon(1966),Dunning(1958) andRugman(1980b)isthecountryfactors,eventhoughtheiranalysesreallyfocuson how these country factors interact with MNE activity as a conduit for their absorption at themicro-levelinthehomecountry,andsubsequentdiffusion/exploitation internationally.Hymer(1960),astheintellectualfatherofthesecondstageinmodernIBstudies, pioneeredafundamentalchangeintheunitofanalysisadoptedinIBstudies:he positionstheMNEanditsFSAsatthecoreofhisanalyticalapproach.Hymersgreat insight is his recognition of the MNEs possession of FSAs, required to offset the liability of foreignness (LOF) when operating abroad (Hymer, 1960; Zaheer, 1995).Unfortunately, Hymer (1960) exaggerates the potential of MNEs to exploit their FSAs as monopolists by exertingeffortstoclosemarketsandexercisingexcessivemarketpower.Structural marketimperfectionsasaresultofgovernmentregulationandMNEscreatingentry barriers,ultimatelyattheexpenseofconsumerwelfare,mayexistbutarenowless common (Dunning and Rugman, 1985). The reality today is that few MNEs, even among theworldslargestones,actuallybenefitfromuncontestableanduncontested monopolies.Theabsenceofsuchmarketpowerisdemonstratedbytherelatively uncommonoccurrenceoftrulyglobalfirms,withabalanceddistributionoftheirsales and assets across the triad of North America, Europe and Asia: few firms appear able to emulatetheirhomeregionsuccessinthehostregionsofthetriad(Rugman,2005; Rugman and Verbeke, 2004, 2008a,b,c). Theliabilityofforeignness,meaningtheimpactofvariousformsofdistance(cultural, economic,institutionalandgeographic),explainswhyMNEshavedifficultiesoperating inforeignmarkets,especiallywhenfacingrivalsnothinderedbysuchdistance.The UppsalamodelofinternationalexpansionproposedbyJohansonandVahlne(1977) essentiallyprovidesthemirrorimageofHymersanalysis.TheUppsalamodelsuggests that there are stages of internationalisation, whereby the potential benefits of exploiting FSAsabroadneedtobeweightedagainsttherisksofoperatinginunknownforeign environments and the costs of learning to do business there. Consequently, according to the original 1977 model, firms initially expand in nearby geographic countries that may havesimilarCSAs.AsthefirmlearnstoovercometheLOFitthenexpandsintomore 6 distant country markets, at which stage the unfamiliar cultural, economic, and political environmentwillbeoffsetagainstthefirmsabilityofrecombiningitsFSAswithhost country CSAs. Recent work by Hennart (2009b) rethinks the nature of such recombinations of FSAs and hostcountryCSAs,anddemonstratesthatinmanycasestheboundarybetweenCSAs andFSAsbecomessomewhatblurred.Indeed,ifsomeoftheCSAsleadingto international expansion are actually not freely accessible, but access is controlled by host countryactors(e.g.,closeddistributionnetworkspreventingsalestocustomersorlocal monopoliesonnaturalresourcesownershipandexploitationpreventingpurchasing these resources), then the challenge for the MNE is to develop relationships, i.e., a type of location bound (LB) FSA with powerful local actors to open up access to the desired CSAs. We define LB FSAs later.TheUppsalamodelrepresentsoneintellectualapproachtoexplainMNEentrymode selection. Here the firm can choose to exploit its FSAs abroad, either through exporting, FDIwithwhollyownedsubsidiaries,licensing,orinternationaljointventures.Inthis literature,thecountryremainsasaco-unitofanalysis,alongwiththeMNE.However, the nature of this research places more emphasis on MNE strategic management process issues and the roles of managers in the parent firm. InthethirdstageofcontemporaryworkinIB,theunitofanalysishasbecomethe subsidiaryoftheMNEandthesubsidiarymanager.Theclearestexpressionofthis approach can be found in Birkinshaw (1996, 1997, 2000).Refining work by Rugman and Bennett (1982), and Rugman (1983b) on world product mandates in a Canadian context, Birkinshaw (1996) has developed the concept of subsidiary initiatives, whereby the focus isreallyoninnovativerecombinationsofbothhomeandhostcountryCSAs,andthe FSAs held (or being newly developed) by the MNEs units in these countries. He observes thatsuchinnovativerecombinationscanultimatelygeneratenewtypesofFSAsacross the MNE network and strengthen the MNEs overall competitive advantage. In summary, over the last fifty years, the literature of IB has developed from a somewhat basicfocusonCSAsandFSAsthatareclearlyseparateanddistinguishablefromeach other(Rugman,1981)towardsamorenuancedunderstandingofthelinkagesbetween themandthemannerinwhichMNEmanagersinthehomeandhosteconomieswill interacttodevelopnovelrecombinationsofhomeandhostCSAsandtheFSAsheldby 7 variousMNEunits,dispersedacrossborders.Overall,theunitofanalysishasshifted from the country-level, to the parent MNE, and now increasingly to the subsidiary level, oftenwithafocusofthesubsidiarysroleintheinternalMNEnetwork.Below,we expand on the implications of adopting alternative units of analysis in IB research. FROM COUNTRY LEVEL TO FIRM LEVEL ANALYSISSeveral theoretical approaches have been used to explain the MNEs strategic investment motives,foreignentrymode,ownership,structuredecisionsandperformance.This sectionprovidesareflective reviewandsynthesis oftheliteratureatfirmleveland the interactions between country and firm level.Hymer (1960) explains why a firm engages in international operations by bringing the focus from the country level to the firm level. HymermovestowardsananalysisoftheMNEbaseduponindustrialorganization theoriesbyshowingthattheMNEisaninstitutionforinternationalproductionrather thaninternationalexchange.HedistinguishesbetweenFDIandportfolioinvestmentin terms of the presence of firm-level control in the former and the absence thereof in the latter. Hymer rejects country level portfolio investment theory with its simplifying (and empiricallyincorrect)assumptionsofthemovementsofcapitalasanexplanationfor FDI, Dunning and Rugman (1985). For Hymer, two conditions have to be fulfilled to explain the existence of FDI: (i) foreign firms must possess a countervailing advantage over local firms to make such investment viable,and(ii)themarketforsellingthisadvantagemustbeimperfect.Hymerargues that for firms to own and control value-adding activities, they must possess some kind of monopolistic advantages sufficient to outweigh the liability of foreignness (LOF), arising fromlackofknowledgeaboutlocalcustoms,differencesinlocaltastes,andunfamiliar legalsystems,when competingwithindigenousfirmsinhost countryproduction.The MNEs proprietary FSAs typically include elements such as product differentiation ability, superiormarketinganddistributionskills,trademarksorbrandnames,accesstoraw materials,economiesofscale,accesstocapital,intangibleassetssuchasproprietary technology,patents,managementskills,theabilitytoachieveverticalandhorizontal integration, etc. Hymer focuses on imperfections in final output markets, as expressed in monopolisticadvantagesheldbyindividualMNEsandentrybarriersleadingto reductionsinconsumerwelfare.Hymerspioneeringviewshavebeenrecognizedasan influentialcontributiontothetheoryoftheMNEandFDI.Hewasthefirsttocontrast suchfirm-levelFDIwiththeprevailingorthodoxybyeconomistswhoexplainFDIasa 8 countrylevelfinancial(portfolio)investmentdecisiondeterminedbyinterestrate differentialsacrossnationalborders.HymerrecognizesthatFDIisafirm-levelstrategy decisionratherthanacapital-marketfinancialdecision(DunningandRugman,1985).Hence, FDI will occur mainly in imperfect markets.Thoughalsofocusingonthefirmastheunitofanalysis,internalizationtheoryhasits originsintheworkofvariousscholarsassociatedwiththeReadingSchool:Buckley and Casson (1976, 2009), Rugman (1981), and Hennart (1982). Here, the MNEs existence isnotcausedbymonopolisticadvantagesleadingtoentrybarriersandconsumer exploitation, but by its efficiency properties, i.e., its capacity to reduce transaction costs when replacing an inefficient or non-feasible arms length transaction in the market by aninternaltransaction,insidethefirm,especiallyinthecontextoftransferring intermediate(mostlyknowledge-based)outputsacrossborders(Rugman,1980a,b; Rugman,LecrawandBooth,1985;Grubaugh,1987).TheMNEsactivitiestypically enhanceratherthanreduceconsumerwelfarebecauseefficientlycoordinated transactions substitute for inefficient ones.Internalizationtheoryeconomists(BuckleyandCasson,1976;Rugman,1981)explain whyfirmsbecomeinvolvedininternationalproduction.Here,theemphasisswitches fromtheconventionalactofFDIatthecountrylevel,totheleveloftheinstitution making the investment, i.e. the MNE. The essential argument of internalization theory is thatfirmsaimatmaximizingprofitbyinternalizingtheirintermediatemarkets (typicallythemarketsforintangibleassetssuchastechnology,productionknowhow, brands, etc.,) across national borders in the face of various market imperfections (such as thepublicgoodsexternalityassociatedwithpricinganintermediateproductlike knowledge,thelackoffuturemarkets,informationasymmetriesbetweenbuyersand sellers,governmentinterventionintheformoftradebarriersortheineffective application of the national patent system). InternalizationtheoryextendstotheMNEthecentralideasofCoasiantransactioncost economics theory (Coase, 1937), developed in a domestic context.Rugman (1980b, 1981) indicatesthatwhileHymer(1960),Kindleberger(1969)andCaves(1971,1982)make marketimperfectionsinfinaloutputmarketsthecentreoftheirtheory,noneofthese authors specifically identify internalization of intermediate product markets as the core of a theory to explain FDI and the existence of MNEs, in contrast to Buckley and Casson (1976)andCasson(1979).BuckleyandCasson(1976)showthatwhenmarketsfor 9 intermediateproductsareimperfect,thereisanincentivetobypassthembycreating internal markets. Here, interdependent activities are brought under common ownership and control. The internalization of markets across national boundaries de facto generates an MNE. Rugman(1981)arguesthatinternalizationtheoryisageneraltheoryoftheMNE.He demonstrates that internalization encompasses within itself the reasons for international (aswellas)domesticproduction.HeemphasizestheroleofMNEsinovercoming imperfectionsinvariousexternalmarkets,aswellasthepolicyimplicationsthereof (Hennart,2009a).Rugmanappliesthetheoryofinternalizationtothepublicdebateon foreignownershipinCanadaandhesharplycriticizestheinappropriateregulationof MNEs(Rugman,1980b,1981).Rugmanarguesthattheefficienciesresultingfrom internalizationarenotacknowledgedtotheirfullextent,andinsteadregulatory measuresimposedbygovernmentresultfromtheunfoundedassumption,inlinewith Hymersview,thatMNEsnormallycommandmonopolisticpositionsthattheywill systematically use to exploit the consumer.Eden(2005)suggeststhatRugmansmostimportantcontributiontointernalization theoryrevolvesaroundtwoelements:first,hisroleinbuildingthetheoryof internalizationasageneraltheoryoftheMNEandsecond,hisbridgingofthegap betweeninternalizationtheorywithstrategicmanagementthinking,bydevelopingthe conceptsoflocationbound(LB)andnon-locationbound(NLB)firmspecificadvantages (FSAs) (Rugman and Verbeke, 1992, 2001, 2003), Rugman (1981) emphasizes that each MNE commands an idiosyncratic set of FSAs, which giveitacompetitiveadvantagerelative tootherfirms.TheseFSAsarisewhentheMNE has developed special knowhow or a capability that is unavailable to others and cannot be duplicated by them, except in the long run at high costs. This thinking anticipates the modern resource based view (RBV) of the firm (Prahalad and Hamel, 1990; Barney, 1991) developedafulldecadelater.Inmanycases,suchFSAsarisefromupstreamresearch and development (R&D) expenditures that lead to new products or production processes. Inothercases,innovationoccursatthemoredownstreamlevel,andcanleadto differentiated product lines, thereby generating an FSA in marketing or distribution. The criticalcapabilityoftheMNEcanalsobesomeuniqueelementofitsmanagement structureorcoreroutinesthatconferanFSA(Rugman,1983a,1985;Rugmanand McIlveen, 1985).10 However,Rugman(1980b,1981)notesthatpossessingFSAsisanecessarybutnota sufficientconditionforFDItotakeplace.OneMNEobjectivemaybetoestablish property rights over its FSAs so that these would not be dissipated to other firms. To the extentthatnationalinstitutionalregimes,suchaspatentprotectionsystems,are consideredinsufficienttopreventunwanteddissipation,theninternalmarketsreplace externalones.TheMNEtransfers,deploysandexploitsitsFSAsthroughtheuseof foreignsubsidiariesthatmonitor,meterandregulatetheuseofFSAsabroad.The internalmarketoftheMNEpermitsittomaximizeitsworldwideearningswithout incurringtherisksofFSAdissipationbyexternalactorssuchaslicensingagents, franchisees,etc.(Rugman,1981).ThegreatstrengthoftheMNEisthatitreplaces exogenouscoordinationsystemsprevailinginexternalmarkets(usuallywithpricingat their core) by coordination through a balanced mix of hierarchical control, socialization andinternalprices.Inshort,Rugman(1981)showsthatMNEsdevelopinresponseto imperfections in the goods and factor markets. The CSAs of a nation that provide a basic levelofcomparativeadvantageareaugmentedbyFSAs,internaltotheMNE,and conferring competitive advantage.Hennart(1982)developedaslightlydifferentversionofinternalizationtheoryas comparedtoBuckleyandCasson(1976)andRugman(1981).Heshowsthatfor international expansion to take place, setting up facilities abroad must be more efficient thanexportingtoforeignmarkets(whichentailsdomesticinternalization)andafirm mustfinditdesirabletoowntheforeignfacilities.ThisisthecaseiftheMNEcan organizeinter-dependenciesbetweeneconomicactorslocatedindifferentcountries more efficiently than markets. Three conditions must be satisfied:first, interdependent actors must be located in different countries (otherwise, only domestic economic activity would occur); second, the MNE must be the most efficient governance system to organize theseinterdependencies(otherwise,onlydomesticactorslocatedindifferentcountries wouldbeinvolvedininternationaltransactions,andnotanMNE);third,thecosts incurredbyMNEstoorganizetheseinterdependenciesinthemarket(asinthecaseof licensing)mustbehigherthanthoseoforganizingthemwithinMNEs(seeHennart, 2009a).Managing interdependencies refers to (a) accessing, (b) recombining, and (c) orchestrating the productive usage of various sets of resources that are dispersed geographically. Such resourcesmayinvolveknowhow,rawmaterialsandcomponents,marketingand 11 distributionservices,financialcapital,etc.FDItakesplacewhenfirmsinternalize markets for these resources. For example, an MNE that wants to exploit abroad its firm-specific knowledge will choose to transfer this knowledge internally rather than license ittoforeignproducersifthemarketforthisknowledgeissubjecttohightransaction costs (Hennart, 1982), but the final decision on entry mode choice does not only depend on the MNEs FSAs.It also very much depends on the complementary resources needed bytheMNEfromforeignactorstomaketheexploitationofitsownFSAsfeasibleand potentiallyprofitable(whichexplainswhyIBisalwaysconcernedwithmanaging interdependencies), Hennart (2009b) Theeclecticparadigm,developedandsubsequentlyextendedintofiveversionsby Dunning(1977,1988,1998),integrates several theorystreamson crossborderactivities at the country and firm levels to explain FDI (see Eden and Dai, 2010 for a review of five versionsoftheeclecticparadigm).Dunningproposesthatthreetypesofadvantages influenceFDI:(i)ownership(O)advantages,(ii)location(L)advantagesand(iii) internalization (I) advantages.Ownershipadvantagescanbedividedintoassetadvantages(Oa)andtransactional variables(Ot).Oaincludevarioustangibleandintangibleassetssuchaspatented technology,brandnames,etc.,whereasOtreferstostrengthsincoordinatingand taking advantage of operating a network of geographically dispersed affiliates. Location (L) advantages reflect foreign countries having some country-specific advantages (CSAs)vis--visothercountries,intermsofnaturalresources,factorsofproduction, demand conditions, etc. Location advantages also include elements of the cultural, legal, political and broad institutional environment in which the firm operates, and that make somecountriesmoreattractivethanotherones.Inaddition,Dunning(1977)identifies themarketstructureatthecountrylevelandgovernmentpoliciesasbeingpotential locationadvantages.HealsoarguesthatthedeterminantsofFDImaydifferfromone industry to another.Internalization(I)advantagesrefertobenefitsofcreating,transferring,deploying, recombining and exploiting FSAs internally instead of via contractual arrangements with outsideparties.Here,thecommongovernanceofgeographicallydispersedvalue-added activitieswithinasinglefirmiscomparativelymoreefficientandeffectivethan governancebyindependentmarketactorsorevenbyanequityjointventurewhere 12 more than one firm is the residual claimant. Firms decide to operate in foreign countries byconsideringtheparticularsetofownershipandlocationadvantagestheyface.The entry modes are selected on the basis of internalization advantages (or the lack thereof).Inadditiontoitscontributionasasynthesizingframework,theOLIparadigmallows identifying the key location advantages of four types of international production: natural resourceseeking,marketseeking,efficiencyseeking,andstrategicassetseeking (Dunning,1998).IncontrasttotheHymerKindlebergerCavesapproach,Dunning devotessomeattentiontomanagerialissuesrelatedtotheFDIprocess,especiallyin termsofthecomplextrade-offstobemadewhenweighingalternativemodesof operatinginforeigncountriesandassessing thebenefitsthereoffor theMNEitselfand its various stakeholders in geographically dispersed jurisdictions. Dunningseclecticparadigm,however,strugglestointegratecountryandfirmlevel interactions.Fromthefirmsviewpoint,the(O)and(I)arenotindependentparameters inmanagerialdecisionmakingbutneedtobeconsideredjointly,with(I)beingthe dominant consideration. The existence of the MNE itself, resulting from FDI, implies that(O)neededtobeinternalizedintermsoftheprocessesof(O)creation,transfer, deployment,recombinationandprofitableexploitation(Rugman,1985;2010;Casson, 1987).Inthiscontext,Itaki(1991)hasvoicedthestrongestcriticismoftheeclectic paradigm,andhasclaimedthatan(O)advantagecouldactuallybederivedfroman(I) advantage,inwhichcaseitwouldberedundanttoconsiderthesetwovariablesas separatedeterminants.Itakihasfurtherpointedouttheinseparabilityofthe(O) advantage from the (L) advantage. He argues that the (O) advantage in economic terms is unavoidably influenced by - and inseparable from - location factors. Hence, (L) and (O) are simultaneouslydetermined.Despitetheaboveshortcomings,whichreflectarelative lackoftheoreticalparsimony,Dunningseclecticparadigmundoubtedlyrepresentsthe mostcomprehensiveframeworktoexplainforeignentrymodechoicesandthe economic efficiency implications thereof.Inparallelwiththedevelopmentofinternalizationtheory,agroupofScandinavian researchers,includingscholarsfromUppsalaUniversity,Sweden(JohansonandVahlne, 1977)andtheHelsinkiSchoolofEconomics,Finland(Luostarinnen,1979)have attemptedtoexplain theprocessby whichfirmsfromsmall,domesticmarketssuchas the Scandinavian countries, internationalize their activities.13 DrawingupontheclassicworksofCyertandMarch(1963),andAharoni(1966),the Scandinavianmodelproposesthatinternationalizationisacumulative,path-dependent processwherebyafirmsinternationalexpansionpatternisafunctionofitspast internationalexperienceandknowledgebase(JohansonandWiedersheim,1975; JohansonandVahlne,1977,1990).Internationalizationtheoryarguesthatafirmwith littleornointernationalexperience,typicallyentersaforeignmarketbyexporting.It progresses to establish a sales subsidiary and eventually to invest in production facilities. The driving force of this internationalization process is experiential market knowledge (Johanson and Vahlne, 1990). JohansonandVahlne(1977)alsointroducedtheconceptofpsychicdistance.Psychic distancereferstothedegreetowhichafirmisuncertainofthecharacteristicsofa foreignmarket(JohansonandWiedersheim,1975).Followingthepsychicdistance concept,firmsundertakeinternationalexpansioninanincrementalmanner.Here,the internationalization model postulates that firms will first enter the foreign markets with whichtheyarerelativelyfamiliar(i.e.geographically,culturallyandinstitutionally proximate),andthen,capitalizingontheknowledgeacquiredfromexportingto-or investingin-thosemarkets,successivelyprogresstopsychicallyandculturallymore distantenvironments(JohansonandWiedersheim,1975;JohnsonandVahlne,1977). Several empirical studies have indeed shown that the MNEs level of foreign experience directlyinfluencesitsselectionofamarketentrymode,seeforexample,Loreeand Guisingers (1995) and Li (1994). However,internationalizationtheorycanbebetteralignedwiththeargumentsof internalizationtheory.Rugman(1980a),andFinaandRugman(1996)havepointedout thatanMNEengagesinforeignproductioninordertoavoiddissipationoftherents derivedfromitsFSAsthatwerecreatedatconsiderableeffortandcosts.Therefore, internalizationtheorysuggeststhatafirmconsiderexplicitlytherelativecostsof servicing foreign markets by first, exporting to foreign markets with the FSAs embodied in final products, second, engaging in FDI or third, licensing a foreign producer. This last option becomes attractive especially when the technology licensed is not any longer the technology on which the firms survival and future growth depends.Themodeofentrychangesovertimeastherelativecostsandbenefitsassociatedwith each of these strategies change. The above stages in serving foreign markets are almost 14 the reverse of the internationalization stages, or the Aharoni (1966) approach, which use (1)licensingasthefirststep,(2)exporting,(3)establishmentoflocalwarehouseand directlocalsales,(4)localassemblyandpackaging,(5)formationofjointventure,(6) foreign direct investment (that is, full scale local production and marketing by a wholly ownedsubsidiary).Furthermore,Rugman(2005)alsoquestionsinternationalization theoryinthatitlacksseriousconceptualgroundingandgeneralizability,especiallyin termofwhatexactlyconstitutesgeographicproximityorexperientallearning,andthe mechanisms through which these concepts influence FDI decisions and geographic sales dispersion.Similarly,RuigrokandWagner(2003)alsoquestioninternationalizationtheoryintheir studyofGermanmanufacturingcompanies.Theyarguethataccordingtotheprinciple ofinitialforeignlocationbasedonthepsychicdistancepremise,Germanfirmsare likelytotargetSwitzerlandandAustria(Germanspeakingcountries).However,both countriesareverysmallmarketsandtheyhaveneverbeenabletoattractsubstantial German FDI. Instead, the typical German firm expands early into other European, North AmericanandAsiancountries.Thesenationsarecharacterizedbyhigherpsychic distance.Thus,Germanfirmsappeartohavepursuedhighdistanceexpansion strategies from the outset, driven by the nature of the location advantages of these larger (highdistance)marketsandpossiblybythecomplementaryresourcesofferedbylocal actors in those environments.Theoverallproblemwiththeinternationalizationtheoryapproachisthatitlargely neglects two critical elements. First, the nature of the MNE FSAs, which determines to a largeextentthepotentialnetbenefitsofinternalizationvis--visalternativemodesof operatinginforeignmarkets(e.g.,theimportanceoftacitversusfullycodified knowledge).Second,thepresenceorabsenceofnaturalandgovernment-imposed marketimperfections(e.g.,anineffectivepatentprotectionsystem),whichmaymake the use of external markets a non-starter. For example, exporting usually takes place in theabsenceofgovernment-imposedmarketimperfections,i.e.,whenthereareno barrierstofreetrade,whereasFDIpreciselyoccurswhensuchbarriersexist.Inturn, licensing takes place when foreign markets are fully segmented, the firm no longer has much to lose by sharing its FSAs, and credible licensees can be found with the requisite resources that complement the MNEs FSAs.15 FROM FIRM LEVEL TO SUBSIDIARY LEVEL ANALYSISBirkinshaw and Pedersen (2009) have summarized the research applying FDI theory and theoriesoftheMNEtothesubsidiarylevelandtheinteractionsbetweenMNEhead officeanditssubsidiaries.Today,werecognizethatwhiletherelevantunitofanalysis formostIB theoryis stilltheMNEasawhole, becausemostkeystrategic decisionsare taken at that level, there is often a problem in translating and applying firm-level theory to the subsidiary unit. The subsidiary becomes the key building block of the MNE, which is viewed as a differentiated network rather than a monolithic hierarchy. In other words, noseriousMNEnetworkanalysiscanbeconductedwithoutunderstandingeach subsidiarysidiosyncraticresourcebase,strategy,assignedroleinsidetheMNE,and linkageswithothersubsidiaries.Inthiscontext,Birkinshaw(1997,1998)hasshifted focustothesubsidiarymanagerandthepossibilityofhavingsubsidiaryinitiatives instrumentaltoFSAdevelopment.Theshiftinfocusfromtheparentfirmtothe subsidiary as a unit of analysis has several origins. First,astreamofresearchinCanadaexaminedtheextenttowhichCanadian subsidiaries of foreign MNEs can act autonomously with world product mandates (WPM). Thisispartlyahostcountrylevelinteractionwithsubsidiarylevelmanagement,but vettedbytheMNEsheadofficeinthehomecountry.Inparticular,theCanadian government wanted to see more R&D in the Canadian subsidiaries of US MNEs, Rugman andBennett(1982), PoynterandRugman(1982),Rugman(1983b),RugmanandDouglas (1986), DCruz (1986). In a public policy context, Rugman (1980b, 1981) criticizes the lack ofeffectivenessandefficiencyassociatedwithpolicyeffortstoboostR&Dspendingin Canadiansubsidiaries.HefindsthattheCanadiansubsidiariesofUSMNEsindeeddo onlyhalfoftheR&Dperunitofsalesascomparedtotheirparentfirms.However,the R&DexpendituresofasetofCanadianownedcompaniesofsimilarsizeisalsowell underhalfthoseofUSparentMNEs.Inotherwords,therelativelackofR&Dfoundin Canadian subsidiaries of foreign MNEs is largely due to country factors rather than firm factors. Canada has poor CSAs as a location for R&D, and Canadian owned firms as well asUSsubsidiariesinCanadabothlackFSAsrelatedtoR&Doutputs.Rugmans conclusions on the inefficiency of attempts to boost artificial national R&D expenditures (whichalwaysreflectacost,butnotnecessarilyanybenefittothefirmsandcountry involved)havesubsequentlybeenvalidatedbyMoore(1996),Birkinshaw(1997)and others. 16 Second, research on the strategy and structure of the MNE moved from a focus upon the centralized, hierarchical multidivisional form typology of the 1960s and 1970s (Stopford andWells,1972;Williamson,1981;Egelhoff,1982)towardsanunderstandingofthe linkages between the parent firm and its subsidiaries. This was a parent firm interaction withsubsidiarymanagers.Inparticular,thepopularizationofthePrahaladandDoz (1981,1987)integration-responsivenessframeworkbyBartlettandGhoshal(1989) established theintellectualfoundationforadifferentiatedinternalnetworkperspective astherelevantorganizationalstructure.Thisworkbuildsontheconceptualinsightsof PrahaladandDoz(1981,1987)whoshowthatsubsidiariescandevelopLBFSAs,albeit sometimesassociatedwithnegativeoutcomesfortheMNE,andthereforerequiringre-centralization,seeVerbeke(2009)andalso Mudambiand Navarra(2004)forananalysis ofdysfunctionalities.MuchsubsequentworkonMNEnetworkswasperformedin Scandinavia. Perhaps the best-known network framework is Hedlunds (1986).He argues that the M-form, parent driven MNE would be replaced by the N-form, or network based, MNE. The Scandinavian and Canadian interest in the subsidiary is a useful counterpoint to the earlier US-led focus on centralized and hierarchical MNEs from large economies. ThemostinfluentialexponentoftheviewthatsubsidiarymanagerscandevelopFSAs throughsubsidiaryinitiativesisBirkinshaw(1996,1997,seealsoBirkinshawandHood, 1998,2001;Birkinshaw,HoodandJonsson,1998;Moore2001;MooreandBirkinshaw, 1998).Birkinshawdemonstratesthatthesubsidiary-andinsomecaseseventhe subsidiarymanagerasdriver/facilitatorof subsidiaryinitiatives-may representauseful unitofanalysiswhentryingtounderstandinnovationprocessesinsidetheMNE.Many strategicdecisionscriticaltoinnovationmaybetakenatthesubsidiarylevelandcan lead to new FSA generation. In this context, Rugman and Verbeke (2009a) have suggested that CSAs of host countries may be used in a leveraged way.MNEs make dual use of CSAs from the home and host countries,andsubsidiariesthroughouttheMNEnetworkmaybecriticalinresource recombinationefforts,aviewconsistentwiththedoublediamondframeworkof Rugman and Verbeke (1993).If MNE operations in various countries can be instrumental tonewknowledgegeneration,thisopensthedoorfortwo-wayflowsofFDI, sophisticated forms of parent-subsidiary relationships and complex network functioning inside MNEs (Rugman and Verbeke, 2001; Rugman and DCruz, 2000). 17 Rugman and Verbeke (1992, 2001, 2003) have argued that FSAs can be created anywhere intheMNEnetwork,bothintheparentcompanyathomeandintheforeign subsidiaries.FSAscanbelocation-bound(LB)ornon-locationbound(NLB).TheLBFSAs reflect strengths deployable and exploitable in a limited geographic area, such as a single country or a limited set of countries or region, but cannot be profitably exploited outside of this area, whether as an intermediate output (e.g. managerial skills, R&D knowledge) or embodied in final products. LB FSAs may include an excellent local reputation, a well-positioned retail network, privileged relationships with domestic economic actors, etc. In contrast,NLBFSAsrepresentcompanystrengthsthatcaneasilybetransferredacross locationsatlowcost,deployedandprofitablyexploited,withonlylimitedneedfor resourcerecombination.SuchNLBFSAstypicallyincludetheupstreampatented technologicalknowledge,andthedownstreambrandnames.Theactualtransferacross borderscanagainoccurintheformofintermediateproductsorembodiedinfinal outputs. RugmanandVerbeke(2001)haveshownthatsubsidiaryinitiativesmayleadtothe developmentofLBFSAs(aresource-basedexpressionofhostcountrynational responsiveness)butthesecanbetransformedintonon-locationbound(NLB)FSAs, namelywhenbeingaugmentedwithbestpracticeattributesinsidetheMNEnetwork (e.g,asaresultofproductivityincreasesandaddeddifferentiation).Indeed,in synthesizing the literature on subsidiary initiatives, Rugman and Verbeke (2001) find ten generic types of capability development processes inside MNEs, of which Birkinshaw had identified those whereby subsidiary initiatives are critical.This framework incorporates the thinking of Birkinshaw and Pedersen (2009) who align the resource based view (RBV) of the firm with the resources and capabilities developed and held in an MNE (Wernerfelt, 1984; Rumelt, 1984, 1997; Barney, 1991; Mahoney and Pandian,1992;Peteraf,1993;Teece,PisanoandShuen,1997;RugmanandVerbeke, 2002).Some FSAs are likely to be held at MNE parent firm level while others are held at subsidiary level.Birkinshaw and Pedersen (2009) argue that if the subsidiary is a valid unit of analysisin itsownright,itshouldbepossibletounbundleresourcesandcapabilitiesbetweenthe subsidiary and the MNE. Considering basic resources first, most tangible resources (plant, equipment and people) are held primarily at the subsidiary level, while most intangible 18 resources(financial,organizational,andreputational)areheldatthefirmlevel. Capabilities are much harder to unbundle between firm and subsidiary levels of analysis. Some are clearly held at the firm level and shared across subsidiaries, such as a particular organizationalculture.Othersaremorelikelytobespecifictoaparticularsubsidiary, suchasaparticularlyeffectivewayofhandlinglocallabourrelationsorprivileged relationshipswithgovernmentagenciestosecurecommercialcontracts.Most capabilities,however,sitsomewherebetweenthetwolevels.Thecriteriausedto evaluateresourcesintheRBVintermsoftheircontributiontocompetitiveadvantage (valuable,rare,non-imitable,non-substitutable)arenotnecessarilythemostrelevantat thesubsidiarylevel.Thus,BirkinshawandPedersen(2009)suggestthatratherthan simplyanalyzingsubsidiarylevelresourcesintermsoftheirpotentialforcompetitive advantage,itisnecessarytoconsiderrecombiningthemwithotherresources,or leveragingthemonaworldwidebasis.ThisprocesscouldgenerateNLBFSAs,inthe spirit of Rugman and Verbeke (1992). Essentially,Birkinshaw argues that subsidiary managers can develop bothhost country, LB FSAs but also NLB FSAs, through subsidiary initiatives. Subsidiary initiatives that lead to NLB knowledge reflect the subsidiary and its managers moving beyond their assigned charter,andgainingworldproductmandatesorcriticalrolesininternationalvalue addedchainsinsidetheMNEsinternalnetwork,ordevelopingsubsidiaryspecific advantages (SSAs) (Rugman and Verbeke, 2001).Relatedworkexaminingtheroleandfunctionofsubsidiarymanagershasbeen undertakenbyHolmandPedersen(2000),Andersson,ForsgrenandHolm(2002,2007), Forsgren,Holm,andJohanson,(1995),FossandPedersen(2002),HolmandPedersen (2000), Malnight (1996) among others. TheaboveanalysissuggeststhatresearchintheIBfieldhasevolvedoverthelastfifty years. The unit of analysis has shifted from the country level to the firm level and finally tothesubsidiarylevel.MNEsubsidiarystrategyhasreceivedsignificantattention, especiallyinthecontextoftheMNEasadifferentiatednetwork.Nowwemoveonto explore theevolution ofIB theoriesfromanequilibriumorientedtheoreticalfocustoa moredynamicorientedconceptualization.Weshowthatthebasicconceptual foundations of IB theory remain as logical developments of internalization theory and its offshoots.19 THE CLASSIC FRAMEWORK FOR IB THEORY The three basic units of analysis can be analyzed in the classic CSA/FSA matrix of Figure 1,derivedfromRugman(1981).Heretheimpactofcountryfactorsisdepictedonthe verticalaxis,rangingfromweaktostrongCSAimpactonIBtransactions(seeRugman and Verbeke, 2009a fora comprehensive overview on location, competitiveness and the MNEs). Conversely, on the horizontal axis, we depict the importance of the firm factors, i.e., the FSAs, ranging again from a weak to a strong impact. Figure 1 here In cell 1 of Figure 1, only CSAs matter to explain the scope and direction ofIB activities. In cell 1, mainstream international economics explains how comparative advantage will leadthehomecountrytoexportgoodsandserviceswhichbuilduponitsrelatively abundant factor inputs of labour, capital and natural resources. For example, Canada will exportnewsprint,SaudiArabiawillexportoil,andChinawillattractmanufacturing assembly through its abundant cheap labour. Cell 1 also captures cultural stereotypes, as popularized by Hofstede (1983) and the GLOBE (2006) studies by House and others (2004), wherebyculturalcharacteristicsandculturaldistancevis--visothernationsareviewed asinstrumentalordetrimentaltoacountryssuccessinIB.Inaddition,cell1includes situationswherebypoliticalandadministrativerulesgreatlyaffectIBtransactions.For example, host governments may restrict and regulate both imports and exports, as well as inward and outward FDI (Rugman, 1980b; Rugman and Verbeke, 2009b). In such cases, variedandcomplexinteractionsmayoccurbetweenMNEsandhostgovernments (Vernon, 1971, 1991; see Rugman and Verbeke, 2009b for a comprehensive overview) In cell 4, the opposite situation prevails: here,country factors do not matter much, and competitiveadvantageresultssolelyfromFSAsunaffectedbygeography,intermsof locationalimpactsontheirdevelopment,transferacrossborders,deployability, recombination requirements and profitable exploitation. This situation is consistent with theviewespousedbymostmainstreamresourcebasedview(RBV)scholarsinstrategic management.TheseauthorsfocusonFSAsonly(referredto,interalia,ascore competencesandcapabilities),anddonotrecognizetheimportanceoflocationand CSAs. Indeed, authors such as Wernerfelt (1984), Rumelt (1984, 1997) and Barney (1991) developedtheRBVinisolationofcountryeffects.ExamplesofstrongFSAsincell4 20 include the allegedly location-independent brand equity of the firm, and the managerial resources and capabilities of the top management team to grow the firm (Penrose, 1959). Incell3,bothCSAsandFSAsmatter.ThisistheuniquestageforIBtheory.Here,the firmbeingstudiedisanMNE,operatingacrossmultiplecountries,andtryingto coordinatevariousresourcedependenciesacrossborders.Bothhomeandhostcountry CSAs may be important in terms of how the FSAs are managed. CSAs affect the processes of developing, transferring across borders, deploying, recombining with other resources and profitably exploiting FSAs, which are always internalized to some extent. Such FSAs canincludehigherordergovernancecapabilitiesandcoreoperatingroutinesfollowing the firms dominant logic.Cell3situationsalsoallowforcomplexintra-MNEnetworklinkages,sometimeswith sophisticatedvaluechainrelationshipsamongthevarioussubsidiariesinvolved,each holdingspecificsetsofFSAsinparticularvaluechainfunctionsandbenefitingfrom particular CSA bundles (Rugman,Verbeke and Yuan, 2011).In each of these situations it isrecombiningresourcesacrossbordersthatmatters,seeRugmanandVerbeke(2001) and Verbeke (2009).Importantly, the three units of analysis intersect and overlap here. ThemainfocuswhenstudyingresourcerecombinationpatternsmaybetheMNE,but usefulanalysisalsorequiresanunderstandingofwhichCSAsinthehomeandhost countriescanbeleveragedbythefirm,aswellassufficientappreciationforthe geographicaldispersionofMNEFSAsacrossthemultipleunitsandsubsidiariesinits internal network.This CSA/FSA matrix is consistent with Meyer, Estrin, Kumar and Peng (2009), who have advocated a combination of institutional analysis (focusing on a subset of CSAs) and RBV thinking(withtheresourcebasedviewprovidingtoolsforstudyingthenatureand strengthofFSAs).Itisalsoconsistentwiththelatestversionofinternationalization theory (Johanson and Vahlne, 2009), which is now focused on understanding the liability ofoutsidershipratherthantheliabilityofforeignness.Outsidershipisconcernedwith accesstoresourcesorratherthelackofsuchaccess,mainlybecauseofrelational shortcomings. Resources are needed to develop the requisite LB FSAs so as to link these withtheMNEsextantNLBFSAs,andtotakefullbenefitoftheCSAsofthehost countries entered. We now turn to issues in the literature less resolved than discussed so far. These new issues are likely to be the basis for future research. 21 THE FUTURE OF DISTANCE IN INTRA-FIRM AND INTER-FIRM NETWORKS ThekeyscholarlyandmanagerialchallengeinIB,irrespectiveoftheunitofanalysis selected,istounderstandproperlyhowdistanceaffectsthetransferability, deployability,recombinationandprofitableexploitationacrossbordersof(quasi-) proprietary know-how, whether in the form of stand-alone competences or higher-order capabilities.The main weakness of many scholarly analyses is the incorrect assessment ofwhatdistancereallymeanswhenperformingIBtransactions.Thisleadsto overestimates of the non-location boundedness of extant FSAs and underestimates of the need for melding investments in host environments, so as to create new, LB FSAs, often in concert with other economic actors. The outcome of such incorrect assessment in the IB literature is structurally flawed theories of how MNEs really operate across borders.Basically, there are three types of FSAs: stand alone FSAs (such as patented knowledge or abrandname),routines(i.e.,thewaythingsaredoneinsidethefirm),and recombinationcapabilities(i.e.,thecapacitytoaugmentinaproductivefashionthe MNEsexistingresourcebasewithnewlyaccessibleresources)(Verbeke,2009). Traditional thinking is that each of these builds upon home country CSAs. Drawing upon homecountryCSAstypicallyleadstoLBFSAs,tiedtothehomecountry.Aportionof thesestand-aloneFSAs,routinesandrecombinationcapabilitiescanbecome internationallytransferrable,deployableandprofitablyexploitable,i.e.thisportion becomesnon-locationbound(NLB).FSAsintheNLBcategorytypicallyincludeR&D knowledge,systemintegrationcapabilities,managerialcapabilities,easyaccessto capital, and sometimes brand names, to the extent that foreign consumers confer value to these. However, upstream FSAs are usually much more NLB than downstream ones. NLBFSAscanbetransferredinternationallythroughtheMNEintra-firmnetwork.However, in most cases, these must be complemented with investments in new LB FSAs in the host countries where they are to be exploited. The MNE may need to draw upon complementary resources held by external actors in the host country.It is only through these complementary resources, and the ensuing LB FSAs that the MNE is able to access hostcountryCSAs(e.g.,accesstoalargeconsumermarket).Iftheseresourcescanbe freelypurchasedonthemarket,theMNEwilldevelopthenewLBFSAsonitsown.In contrast,iftheseresourcescannotbepurchased,jointventuresandothertypesof alliancesmayresult,especiallyiftheMNEwishestokeepsomedirectcontroloverits 22 ownNLBFSAstransferredabroad.Inthecasewhentherearestrongimperfectionsin themarketsforresourcesandnosatisfactoryjointventurepartnerscanbefound,the only option is a merger or acquisition. Here the MNE is forced to purchase not only the resources required for accessing the host country market, but a local firm in its entirety, includingpossiblesubstantialunwantedassetsandcapabilities.Obviously,iftheMNE can simply augment its internationally transferrable NLB FSAs with self-created LB FSAs, fully controlled by the MNE through its internal network, then FSA dissipation risks can be avoided.However,hostcountrysubsidiariesmaydomuchmorethansimplyaugmenthome-based,NLBFSAswithrequisitelocalstrengths.Theymaydevelopsubsidiaryspecific advantages (SSAs) as a result of their autonomous initiatives. Rugman and Verbeke (2001) defineSSAsasidiosyncraticstrengthsdevelopedbyhostcountrysubsidiarymanagers buildinguponhostcountryCSAs.However,SSAsmayrepresentnotjustLBknowledge butnew,NLBstrengthsinthesenseofhavingprofitpotentialabroad.Thechallenge associatedwithSSAsisthattheactualcapability(ormorespecificallythecapability carrier,suchasthemanagementteammasteringtherelevantknowledge)cannotbe simplytransferredacrossborders.Inotherwords,theunderlyingknowledgeis embedded in the subsidiary and its linkages with local actors.As Rugman and Verbeke (2001)pointout,SSAsareoftenviewedwithgreatskepticismbythecorporatehead office,astheydonotconstitutenetworkknowledgebydefinition.Anunresolved question is whether SSAs can somehow be upgraded (e.g., through codifying knowledge and sending expatriates to the subsidiary to learn about the new capability), so that they would become relevant to the entire MNE subsidiary network. FutureresearchongeographicdistanceandSSAsmightbeusefullylinkedtorecent empiricalandtheoreticalworkanalyzingtheregionalnatureofMNEs(Rugmanand Verbeke,2004;Rugman,2005).Giventhattheempiricaldatashowthattheworlds largest500firmsoperatemainlywithintheirhomeregionofthetriad,itisapparent thatSSAs,aswellasmoreconventionalNLBFSAs,aredifficulttoexploitprofitably outside of their home region. The problem may not reside solely in technical difficulties associatedwithinternationalFSAtransfer,butalsoinchallengesof(a)effective deployment in a host environment, depending upon the recipients absorptive capacity; (b) appropriate recombination of the NLB FSAs with newly created or newly accessed LB 23 FSAs;and(c)managerialeffectivenessinprofitablyexploitingthenewlycreatedFSA bundles.Thesechallengesbecomecompoundedasdistanceincreases,whethereconomic, cultural,institutional,ormerelygeographic.ItbecomesmoredifficultforseniorMNE managers at the head office to understand critical success factors and to act upon related challenges(aboundedrationalityproblem).Italsobecomesmoredifficulttoengagein propermonitoringandcorrectionofhumanbehavior,especiallywheneffortsare diverted from engagement towards achieving company goals; this is a bounded reliability problem (Verbeke and Greidanus, 2009).Regional boundaries such as in NAFTA and the EU, as well the boundaries separating Asia from the rest of the world, represent a useful first cut at separating lower distance, intra-regional environments from higher distance, inter-regionalones.Inotherwords,thelocationboundednessofFSAsisoftenmainly intra-regional in nature:FSAs can be relatively easily transferred, deployed, recombined and profitably exploited throughout the home region compared to between regions. The comparatively easier resource recombination challenge, as compared to what prevails in host regions, is especially critical. It remains somewhat of a puzzle as to why inter-regional distance remains so important, whereas the world economy otherwise appears to be moving towards greater integration withsupportingpoliticalandinstitutionalstandardizationbeingdrivenbyimproved informationtechnologyandinformationaccessingcapabilities.Inourviewitisthe compounded distance, more specifically the need to manage various distance dimensions simultaneously,thatreallyexplainstheregionalnatureoftheworldeconomy.The compoundeddistanceisinsufficientlycapturedineffortstosimplyaddorotherwise aggregatevarioustypesofdistance.Ghemawat(2001,2007)hasanintroductory discussionofthis,thoughstilladoptingalargelyadditiveapproach.Futureresearch shouldengageinfurthercarefulreflectionandtestingofthecompoundeddistance concept. ThekeyinsighthereisthatthevariousdistancedimensionstypicallymeasuredinIB studies are not independent of each other.For example, regional economic integration fostersinstitutionalcoordination,andmaycontributetoloweringactualcultural distancebyincreasingmobilityoflabourandmanagerialbestpractices.Atthesame time,improvingthecommontransportinfrastructure,addingtransportconnections 24 (e.g.,inairtravel)intermsoffrequencyandqualitymayreducetheimpactof geographical distance.In addition, one specific parameter related to distance dimension hasbeenneglectedsystematicallyinlarge-scalestudiesanalyzingthegeographicscope ofinternationalexpansionpatterns.Thisisthepresenceandthestrengthofthe relationshipsbetweentheMNEandrelevantactorsoperatinginthehostenvironment.Asnotedearlier,JohansonandVahlne(2009)havearguedintheirextensionof internationalizationtheorythatbeinganoutsiderinforeignnetworksmayexplainthe absence of competitive success.The liability of host region outsidership, meaning simply that most MNEs have few if any powerful relationships with actors in host regions, may goalongwayinexplainingtheabsenceofinternationalsuccessinthesehostregions. Relationships may well be the key missing factor (as a LB FSA) to overcome compounded distance and to allow access to coveted CSAs in host regions. At this time, it remains unclear as to the extent to which forces of global integration can actuallyovercomethecurrentlytriadregionbarriersobservedbyRugman(2005). Researchers will need to continue to examine carefully the extent to which region-bound FSAs, including SSAs held by MNE subsidiaries, can be transformed from theirapparent currentregionalexploitabilityintoNLBFSAswithatrulyglobalreach,meaningthat theycaneasilybetransferred,deployed,andrecombinedwithotherresourcestothe extent necessary, and profitably exploited around the world.The above analysis can be related to the earlier Figure 1. The most interesting aspect of Figure1liesincell3,wherebytheMNEsuccessfullyaccessesandleveragesCSAsin home and host environments, and is able to put together bundles of LB and NLB FSAs to achieve success in a variety of host environments. The requirements of this approach go far beyond merely adapting homegrown FSAs to the host country environment, as in the national responsiveness strategy of Bartlett and Ghoshal (1989). For example, in the case whereaBirkinshawtypeSSAheldbyanMNEsubsidiaryincell3iscriticaltoMNE internationalgrowthandprofitability,thechallengefortheMNEheadofficeisto providesufficientresourcestothesubsidiarytoallowglobaldeploymentand exploitation(byallowingaworldproductmandateorprovidingthestatusofglobal centreofexcellence).Alternatively,investmentsmustbemadetoturntheSSAintoa strengththatcanbetransferredinsidetheMNEnetwork(e.g.,throughcodifying knowledge,orbysendingsubsidiaryexpertstootherunitstoprovidetraining,orby 25 allowingdirectinteractionandknowledgetransfersamongsubsidiaries,etc.)andthe subsequent, dispersed exploitation out of many units. Essentiallythissectionhasdiscussedpotentialcombinationsofhomeandhostcountry CSAs and associated FSAs. All of this provides sophisticated substance to cell 3 of Figure 1. The future for research in IB needs to expand upon this type of deep analysis of cell 3 inFigure1.Inparticular,effortsarestillintheirinfancytounderstandthenatureand extentofrecombinationefforts,wherebyextantFSAbundlesaremeldedwith complementary resources in host environments, so as to access host country CSAs. Acriticalfeatureofinternalizationtheoryisitsfocusongovernance.Inotherwords, disciplinedexecutionoftheFSA-CSArecombinationprocessesdescribedaboverequires muchmorethansuggestedbytherathersimplisticcase-basedanalysisofBartlettand Ghoshal (1989) and their broad integration responsiveness framework.This framework provides little guidance in terms of proper governance beyond some general suggestions on how to improve normative integration, i.e., socialization, Rugman, Verbeke and Yuan (2011).Similarly,limitedcaseanalysisontheso-calledmetanationalbyDoz,Santosand Williamson (2001), in terms of building globally deployable NLB FSAs based on the CSAs ofperipheralhostcountryenvironments,isnotparticularlyuseful.Dozandhisco-authorsplacethefutureoftheMNEinthehandsofsmallsensingandmagnetunits, almosttheequivalentofUSArmyDeltaForceunits.Theseunitsaresupposedtowork outsideoftherealmoftheMNEsoperatingdivisions.Sensorsaresupposedtoidentify andseekoutgeographicallydispersed,unique,andhithertofullyignoredbodiesof valuable knowledge. Magnets are then supposed to follow up the sensors activities, and torecombinethevariousuniquepiecesofknowledgeidentifiedbythesensorsinto commerciallyviableproducts.Inafinalstage,the MNEsoperatingdivisions must then implement the new solutions put forward by the sensors and magnets.The operating divisions role is simply to engage in scaling up and ensuring the commercial viability of the products pushed onto the MNE network by the sensors and magnets.VerbekeandKenworthy(2008)haveargued,however,thatthemetanationalformof governance is unlikely to displace the conventional, hierarchical, M-form organizational structure,especiallyintermsofhownovelresourcerecombinationsareturnedinto actualnon-locationboundFSAs.Theyhavefocusedontheneedtoviewresource 26 recombination,i.e.,theinnovationprocess,initsentirety.ThismeansthatR&Dor othereffortstodevelop/accessnewknowledgecannotbestructurallydivorcedfrom production and marketing/sales.This is broadly consistent with the thinking of Buckley (2009) on the global factory.TherearegoodreasonswhymostlargeMNEsareorganizedindivisions(typically geographicandproductdivisions)andintonationalunitswithspecificmandatesin tightlycoordinated,internationalvaluechains:suchgovernanceiscomparativelymore efficient and effective than alternative forms of organization. The modern M-form allows economizingonboundedrationalityandboundedreliability,andimplementing innovation processes in their entirety, from technological FSA development to after sales service to customers, a view consistent with Wolf and Egelhoffs (2010) critical analysis of variousformsofnetworkorganizationsproposedbyIBscholarswhoneglectthe limitations of networks as compared to, for example, modern matrix structures. Using an informationprocessingapproach,WolfandEgelhoff(2011)convincinglyarguethat decentralizednetworkorganizationssimplydonotworkascomparedtoM-form structuressuchasmatrixorganizations,whentheMNEoperatesinmultiplehigh distanceenvironments,needstofocusatleastpartlyontheefficiencyofexisting operations, and must balance the exploitation of NLB FSAs with developing LB FSAs.Inotherwords,muchofthetraditionalinternationalmanagementstrategyliterature stillmakessimplisticassumptionsthatglobalstrategiesarefeasible,andcanbe implementedwithoutmuchattentiondevotedtogovernance.Infact,globalstrategies and truly global NLB FSA bundles are likely to be extremely rare, as demonstrated,inter alia, by the regional nature of MNEs in terms of their (limited) sales and asset dispersion, and the regional elements in their governance structures (Rugman and Verbeke, 2008a). Future research will likely build upon the theoretical and empirical insights of Rugman andVerbeke(2004)andhopefullyrecognizeregionalstrategyandstructureasan efficientalternativetonon-feasibleglobalapproaches,especiallyinthecontextof internalgovernance.Globalgovernancesometimesdoesappearpossibleinthecontext ofglobalalliancessuchasintheairlineindustry,whereindividualairlinesperform largelyregionalroles,andselectparticular,narrowvaluechainactivitieswhere resourcescanbepooledandrecombinedtocreateNLBalliancespecificadvantages (ASAs), which unfortunately are lost to an airline when it decides to leave the alliance.27 OTHER UNITS OF ANALYSIS AND THE ROLE OF DISTANCE IN IB Infocused,contemporaryIBstudies,nineotherunitsofanalysis havebeenadopted,in additiontothebigthreediscussedinearliersections.Thesesetsinclude:first,the entrepreneur; the MNE head office and the top management team; and the value chain; Second,theexpatriate;thecentreofexcellence;andthemodelfactory.Third,the outsourcing agreement; the joint venture/strategic alliance and the cluster. The first three additional units of analysis are really subsets of the firm as the main unit of analysis, but with a focus on, respectively, who established the firm, what governance mechanism leads the firm and how the firms value added functions are dispersed across space. The second set of three additional units is concerned mainly with what occurs at the subsidiary level, but with a focus on, respectively, who is sent by the head office to performanumberoftasksinthesubsidiary,whatuniquerolethesubsidiarycan performinsidetheMNEnetworkandhowitcanbeamodelforexceptionalvalue creation.Thethirdsetexplicitlylooksatthecomplementaryresourcesprovidedby external actors that can be productively recombined with extant FSAs. Here, the critical point to remember is that any serious IB analysis always considers the CSAs of home and host countries, as well as bundles of LB and NL FSAs. The challenge is alwaystoexplain,predictorguidehowcompetitivesuccesscanbeachievedoutsideof the home nation, or how such success is hampered by a variety of internal and external constraints.Forexample,thestrategicchallengesfacinganentrepreneurwith international expansion ambitions are similar to those facing a large MNE, especially in terms of the bounded rationality and bounded reliability constraints to be overcome, and the transaction costs related to these constraints (Casson, 1982).Theabovealsoholdsfortheanalysisofnetworks.Theproblemwithsomeofthe strategyandIBliteratureisanexaggeratedfocusonthebenefitsofnetworks,buta relative neglect of its costs. Network theory needs to be applied to IB with caution. When there are intra-firm alignments the network approach is fully consistent with analysis at the parent and subsidiary level, as discussed above. However, inter-firm linkages, which includejointventuresandstrategicalliancesdonotrelateeasilytobasicIBtheory, especiallyinternalizationtheory.Here,thenetworkliteraturelargelyignoresthe ownershipissue,inthesenseofrequiredFSAprotectionagainstdissipationandthe 28 appropriatedistributionofthe cooperationbenefitsand costsamong thepartners, who are all residual claimants (Narula and Hagedoorn, 1999; Gulati, Nohria and Zaheer, 2006). Inmoregeneralterms,themainchallengeassociatedwiththemyriadofalternative units of analysis adopted in theIB literature is that the authors sometimes lose track of the continued relevance of distance and its various dimensions. One example is the study ofso-calledbornglobalswherebyauthorsoftensimplyassumethatcell3ofFigure1 opportunitiesareavailabletothemacrosstheworldwithoutunderstandingthe challengesofinternationalexpansion,especiallyoutsideofthehomeregion.Inmany ways, a simplistic use of cell 3 in Figure 1 thinking represents a born global illusion for international entrepreneurship, Rugman and Almodovar (2011).Studiesonanotherunitofanalysis,namelythevaluechain,areoftenalsoassociated with a rather shallow understanding of the compounded distance concept. A value chain dispersed across borders equates to a firm making strategic decisions to access particular home and host country CSAs by deploying idiosyncratic FSA bundles in each country and thencoordinatingthesebundles.Inthiscontext,theoftenexpressedviewthatMNEs pursueglobalsourcingandareengagedinaglobalracefortalent,ratherthan pursuing more focused and selective initiatives in these areas, such as regional sourcing orlocalsourcing,oftenleadstoabsurdconclusions,(Rugman,LiandOh,2009).More specifically,mostoftheliteratureonoffshoringandoutsourcingislargelyacell1in Figure1phenomenon.WesternfirmshaveoffshoredmanufacturingtoChinadueto ChinasCSAincheaplabour.Similarly,IThasbeenoffshoredtoIndiaduetoIndias relatively cheap, skilled and educated labour. The relevant insight is that offshoring is difficult. The distance to countries such as China andIndiamaybelargeforWesternMNEs,andthereforeoffshoringeffortsarenot deployedgloballybutareverytargeted,oftentowardslargecosmopolitancentres. Withinthese centres,widelyavailableknowledgeofbusinessbestpractices,competent legalandfinancialcounselling,andapoolofexperiencedmanagersmayalleviatethe boundedrationalityandboundedreliabilitychallengestobeovercome.Inmostcases, substantialmeldinginvestmentsmustbeperformedfortheMNEtobesuccessfulin recombining the MNEs extant NLB FSAs with the FSAs of local resource providers. Here, theadditionaldistancechallengeisthatoutsidesuppliersofoffshoringservicesmay wanttoworkupthesmilingcurve,i.e,movetowardsbothmoreupstreamand 29 downstream value chain activities as compared to typical component supply or basic ICT services,therebycreatingmorevalueaddedperunitdelivered,Mudambi(2008). However,thereisadangerfortheMNEinthatitmayultimatelyloseitscompetitive edgeinsomeoftheseactivitiesvis--visoffshoringservicesproviders,especiallyin higher distance, peripheral markets. Fortunately, much of the work on the global factory, which looks at the production side of the value chain is consistent with the complexities of cell 3 in Figure 1 thinking, see Buckley and Ghauri (2004), Buckley and Hashai (2005), Buckley (2009).In examining the changinglocationandownershipstrategiesofMNEs,namely,thehubandspokeand theglobalfactory,BuckleyandGhauri(2004)showtheincreasinglysophisticated decisionmakingofmanagersinMNEstoslicemorefinelythe valueaddedactivitiesof firms. In finding optimum locations for each closely defined activity, they are deepening the international division of labour. Ownership strategies are also becoming increasingly complex, leading many MNEs to apply a control matrix, whereby operating strategies are decided upon location by location, and can range from wholly owned subsidiary units via FDItomarketrelationshipssuchassubcontracting,withjointventuresasoptionson subsequentdecisionsinadynamicpattern,LiandRugman(2008).However,fully understandingcell3complexities,probablymeansthathereagaintheso-calledglobal factory may be an illusion, and that the reality may be more one of regional factories. Figure 2 here The implication of the above is not only that distance still matters, but should never be underestimatedinIBstudies(especiallynotthecompoundeddistance),assuggestedby Figure 2. The vertical axis of Figure 2 lists the various possible units of analysis, whereby we include here, for illustrative purposes only, four units of analysis, consistent with cell 3 in Figure 1: the entrepreneur, the MNE, the MNE subsidiary with intra-firm networks, andinter-firmarrangements.ThehorizontalaxisdichotomizesthereachofMNE FSAs/capabilities into global NLB FSAs versus limited NLB FSAs.The relevance of Figure 2 is that we can position a substantial number of contemporary IB concepts on the left hand side. Cell 1 represents the so-called born global firm with anambitiousentrepreneuratitscorewhohasworldwidegrowthambitions.Cell2 representstheglobalMNEwithsalesandassetsdispersedinabalancedfashionacross theglobe,ofwhichthereareonlynineexamples,Rugman(2005),orthemetanational withitssensorsandmagnetsgatheringknowledgefromthefarcornersoftheplanet. 30 Cell3representstheMNEsubsidiarywithaglobalmandate.Finally,cell4represents global outsourcing networks and global alliances.Unfortunately, the real world cases that can actually be placed in cells 1 to 4 are few and farbetweenduetothetyrannyofcompoundeddistance.Mostbornglobalsarereally bornregionalsincell5.MostlargeMNEshaveahomeregionsalesandasset concentration,andoperatewithgeographicdivisionsincell6,RugmanandVerbeke (2004, 2008a,b,c). Most MNE subsidiaries perform a regional mandate role in the internal MNEnetworkincell7.Finally,mostoutsourcingnetworksandalliances,eventhose designedbythelargestMNEsintheworld,areagainveryselectiveintermsof geographiccoverage,placingthemincell8.Globaloutsourcing/offshoringincell4is largely a non-starter, because of the transaction costs associated with it. In short, in the future, much more research needs to be done on the right hand side of Figure 2; theIB fieldneedstomovebeyondtheleftfieldofFigure2.Weneedtofocusinsteadonthe limits to globalization, resulting from two sets of parameters, as shown in Figure 3. Figure 3 here Theessenceofexplainingtheliteratureonglobalversusregionalstrategiescanbe divinedwithreferencetoFigure3.Asexplainedabove,muchoftheliteratureinIB ignores the complexity of geographic distance. Therefore, we place the complexity of the cultural,institutional,economic,andgeographic componentsofcompoundeddistance on the vertical axis. Essentially, each MNE needs to identify the relevant weighting of the componentsofthecompoundeddistanceitfaces.Itthenneedstobuildonitsextant FSAs base to further develop new FSAs which will overcome the resource recombination barriersitfaces.Therefore,onthehorizontalaxisofFigure3,weplacetheresource recombination barriers to the growth of the MNE.The presence of compounded distance means that it is not sufficient to add the isolated effectsofcultural,institutional,economicandgeographicdistancecomponents.There maybemultiplicativeeffectsarisingfromthesedistancedimensions,especiallyin interregional as opposed to intraregional settings. In other words, rather than looking at CSAsassourcesofnetbenefitstotheMNE,assuggestedbyFigure1,thefocusshould shift instead to the challenge of accessing these CSAs.Theexistenceofresourcerecombinationbarriersdemonstratesthedifficultiesof creatingthe rightmix ofFSAsthat wouldguarantee competitiveness in,forexample,a host region market.In other words, it is important not simply to assume the presence of 31 FSAsintheMNEthatwouldalmostautomaticallyconfercompetitiveadvantage,as suggestedbyFigure1,butinsteadtorecognizethat,typically,NLBFSAsneedtobe meldedwithLBFSAs,aprocessthatmayencountersevereimplementationproblems, interalia,wheneachsegmentofthefirmsvaluechainrequiresidiosyncraticresource recombination efforts, Rugman, Verbeke and Yuan (2011). A similar approach to finding, and testing, several types of LB FSAs appears in Lo, Mahoney and Tan (2010). Incell1,thereisamixofhighcompoundeddistanceandlowresourcerecombination barriers: this is the assumption of a complex world, combined with the MNEs ability to createtherightFSAmixwithouttoomuchdifficulty.Thisisthemainassumption adoptedin,forexamplethe transnationalandmetanationalillusions.Incell2,alow compoundeddistanceandlowresourcerecombinationbarrierssuggestasimpleworld where all is global (global firms, born globals, global mandates, etc.). Cell 3 in turn shows theregionalsolution,wherebyfirmsfacehighcompoundeddistance,andatthesame times struggle with important resource recombination barriers when entering into host regions. Finally, cell 4 describes the often-observed position of smaller firms with mainly LB FSAs because of high resource recombination barriers. Such firms have great difficulty overcoming even limited compounded distance (across the national borders of the home region).These firms operate mostly on a local level.Inconclusion,Figure3helpsustobuilduponthetensionbetweentheliteratureon simplisticaspectsofglobalizationandglobalstrategies,ascomparedtotherealitiesof morecomplexdecisionmakingrequiredbyMNEsdealingwiththerealitiesof compoundeddistanceandresourcerecombinationbarriers.Incell1ofFigure3,the world is complex, in terms of compounded distance but the literature on metanationals andtransnationalssimplyassumesthatfirmscanovercomethis.Inreality,thereare complexitiesfacingthefirmindevelopingnewFSAstoovercomeresource recombination barriers. In the future, IB research needs to engage with cell 3 rather than continuing to be bogged down in cell 1. In a similar manner, much of the literature over thelastfiftyyearshasdealtwithcell2ofFigure3.Thishasledtosimplisticworkon globalfirms,globalmandatesandbornglobalfirmswhichbasicallyassumethatthe worldissimpleasbothcompoundeddistanceandresourcerecombinationbarriersare ignored. In reality, even if compounded distance is not a critical challenge, there remain firmsthatoperateincell4,developingmainlyLBFSAs,andoperatingdomestically. Thesefirmslackthemotivationtoexpandinternationally,forexample,becauseofan 32 uncontesteddomesticmarket.Insummary,futureresearchinIBneedstodistinguish muchmorecarefullybetweenthefourcellsofFigure3.Intheconcludingsection,we suggest that such future research may well benefit from a focus upon the regional nature of IB. CONCLUSIONS REGIONAL VERSUS GLOBAL STRATEGIES We have shown that the three key units of analysis in IB theory over the past fifty years have been the country, the firm (MNE) and the subsidiary.The most promising area for furtherIBtheorydevelopmentisinessencethestudyoftheinteractionsamongthese three parameters, with the subsidiary as the key building block.In particular, we have shown that the subsidiary has emerged as a new unit of analysis due to the importance ofnetworkthinkingfromstrategicmanagement.Here,theSSAconceptrepresentsthe culmination of fifty years of IB analysis. An SSA results from (a) recombining knowledge transferred from the network with newly created knowledge; (b) autonomously assumed (extended) subsidiary roles; and (c) subsidiary knowledge embedded in idiosyncratic host country locations. Paradoxically, the SSA concept is conceptually almost the opposite to conventionalnetworkthinkinginmainstreamstrategy,withitsexaggeratedfocuson knowledge sharing. Looking ahead, it is apparent that, irrespective of the unit of analysis chosen, the region asanexpressionofhowdistancemattersinIB-hasemergedasanimportant parameter. Here, theoretical work is in its infancy as the focus of recent research on the nature and extent of regional versus global activity by MNEs has been largely empirical. Thisresearchonlybecamepossibleduetochangesinaccountingstandardsinthelate 1990s, which required firms to identify the broad regions of the triad in which their sales and assets take place, Rugman (2000, 2005).There is confirmation of the lack of global firms and the importance of the region as a unit of analysis in Fisch and Oesterle (2003); Asmussen(2008);CollinsonandRugman(2008);RugmanandOh(2008);Hejazi(2007); Kolk(2010);Sethi(2009);YinandChoi(2005);RugmanandVerbeke(2008a,b,c);Grosse (2005). Thereneedstobemoreconceptuallydriveninquirytosolvetheempiricalpuzzlethat most of even the largest MNEs appear unable to move beyond their home region and go global. A transaction cost economics (TCE) based analysis may be particularly useful here, Rugman and Verbeke (2005). Such analysis would recognize the barriers to recombining 33 extantNLBFSAsheldbytheMNE,withLBFSAstobeprovidedbythirdparties(inthe absenceofwellfunctioningexternalmarketsfortheresourcesunderlyingtheseLB FSAs). This resource recombination process, in turn, may be a precondition to accessing the coveted CSAs in host markets, such as a large customer base at the downstream end or a highly skilled and productive workforce at the upstream end of the value chain, but requires overcoming resource recombination barriers. Currentlythereappearstobealiabilityofinterregionalforeignnessfacingeventhe largestMNEs,thoughthisconcepthasonlybeensketchedbyRugmanandVerbeke (2004, 2007).If the great majority of the worlds 500 largest firms grow their sales and assetsmainlywithintheirhomeregion,thissuggeststhattheirbusinessmodelsand strategiesarehomeregionbased.Theremustbeprohibitiveresourcerecombination barriersassociatedwithadaptingorchangingtheirbusinessmodelstoachievetruly global sales in the other two regions of the triad. Thus, IB theory must redirect thinking at the country level on the liability of inter-regional foreignness caused by compounded distanceandmovethistowardsafocusuponregionalstrategy.Inotherwords,the country level of analysis may need to move towards a regional level when assessing the impacts of differences in culture, political regulation, economic and financial institutions etc.Newresearchonthemultinationalityandperformancerelationshipneedsto incorporate this regional dimension, Rugman and Oh (2007, 2010), Lee (2010); Oh (2010); Wolf et al. (2008); Li and Li (2007), while avoiding the pitfalls of the past research on this topic (Verbeke and Brugman, 2009) Inrelatedwork,FischandOesterle(2003)pointoutthatthetermglobalizationhas become overused, and has apparently replaced the term internationalization. There are bothconceptualresearchgapsconcerningthedifferencebetweenglobalizationand internationalization,andalackofempiricalknowledgeabouttheactuallevelof globalization of MNEs. In the spirit of Rugman (2000), Fisch and Oesterle (2003) present a newquantitativetool,whichcombinesgeographicspreadandculturaldiversity measures,soastointegratemultipledimensionsofinternationalizationintoa globalization index instead of a simple internationalization measure. Fisch and Oesterle (2003)applythismeasuretoassesstheglobalizationofthemostinternationalized German MNEs, among the top 100 non -financial MNEs from developed economies. The resultssuggest thattheseMNEsareneitherglobalizednor showa straightforwardpath towardsglobalizationinthepastdecade.Thisoutcomecontradictsthecommon 34 assumptionofglobalMNEs.Thisnewconstructcanbeoperationalizedtomeasurethe degree of internationalization (Glaum and Oesterle, 2007). More research like this needs to be undertaken on the metrics of internationalization, not globalization. Intermsofinternationalhumanresourcedevelopmentandcross-culturalstudies,a regionalfocuswillrequirefundamentalrethinkingofthetraditionalcountryfocus.If MNEsoperateregionally,thenwhatistheuseofHofstede(1983),Houseetal.(2004), GLOBE(2006)andtheKogutandSingh(1980)metrics?Thesetraditionalempirical shortcutsforcross-culturalanalysismayneedtoberevisedataregionallevel.For example,TungandVerbeke(2010)havesuggestedrecentlythatanalysisofexpatriates andthetrainingofseniorexecutivesforChinesefirmsmaybesubjecttoaninverse resonanceandthattheremaybeaneedtoadoptaregionalfocus,e.g.anAsian-Asian perspective rather than the traditional Western-Asian perspective. In many ways, the ten theoreticalproblemsincross-culturalresearchidentifiedbyShenkar(2001)remain unresolved, mainly because research in this area is poorly embedded in IB theory. In this context, Wolf, Dunemann and Egelhoff (2008) examine the economic, psychological, and sociologicalreasonsforhome-regionorientedMNEs(RugmanandVerbeke,2004; Rugman, 2005).In terms of political integration, the experiment of the EU offers many insights into the benefitsofeconomicintegrationonaregionallevelwithcommoninstitutionsand regulations.ButdoesthisapplytoNAFTA,inwhichtheextentofsocialandpolitical integrationismuchlessthanintheEU?OrtoAsia,whereatbestaNAFTAtypesetof loose integrationismorelikely toarise thanthe tightly regulatory structureof theEU?WhyisitthatNorthAmericanandAsianMNEsarenearlyashomeregionbasedas European ones, without the same degree of political and social integration? Afinalexampleoftheneedforbettertheoreticalanalysis,buildingupontheregional phenomenon,istheworkoninternationalentrepreneurshipandso-calledbornglobal firms.Closescrutinyofthebornglobalphenomenon,wherebycompanies internationalizenearorattheirfounding(KnightandCavusgil,1996),suggeststhat thesefirmsactuallyachievetheirexportssaleslargelywithinthehomeregion.For example,Knight,MadsenandServais(2004)investigatethe292bornglobalfirmsin Denmark and the United States. Their findings are that Danish firms generate on average of 71 percent of their total sales from abroad, mainly in Europe, compared to 47 percent 35 for the American sample. This reflects the small size of the Danish domestic market and Denmarks proximity to numerous neighbouring countries in Europe, of which Germany accountsfor50percentoftheirsales.TheAmericanfirmsgenerate53percentoftheir totalsalesintheUnitedStates.Tobemoreaccurate,thesefirmscanbedescribedas born regional, not born global, (Rugman, 2000; Fisch and Oesterle, 2003).There appears to be no robust empirical evidence of any born global firms other than a fewITfirmsfromIndiaandpossiblysomesmallfirmsfromIsrael,Rugmanand Almodvar,(2011).Instead,therearebornregionals(Lee,2010;Lopezetal.,2009). Further,whenasmallfirmmakesopportunisticforeignsales,theseareusuallyinthe formofexports,notFDI.Yet,theclassicarticleoninternationalnewventure(born globals)byOviattandMcDougall(1994)usesDunningsOLIframeworkofMNEsto explainbornglobalactivity,insteadofmaintainingamoreappropriatefocusupon exportsandinternationalmarketing.Again,thisareaofresearchisstillveryweakly embedded in basic IB theory. In conclusion, future theoretical work inIB must continue to study the linkages among thekeyunitsofanalysisadoptedinthepreviousfiftyyears,withafocusonthe subsidiary as the key building block, taking into account the reality of regional strategy andstructureformostMNEs.Weneedtoavoidbeingsidetrackedbyill-conceived concepts such as the metanational and empirically illiterate subfields of research such as the borngloballiterature,whichignoresthe costsof distance. Weshouldfocusinstead on the limits to globalization, resulting from severe frictions, i.e. resource recombination barriers,occurringwhenattemptsaremadetoconductbusinessbeyondthehome region,wherecompoundeddistancemaybehigh.Itisonlybylinkingbravenew empiricalworkonemergingissues,suchastheregionalsolution,withgoodbasic theory,i.e.sophisticated,comparativeinstitutionalanalysis,asprovidedby internalization theory, that the field of international business will continue to prosper in the future. 36 Figure 1: The CSA/ FSA Matrix. Firm-Specific Advantages (FSAs) Weak Strong Country-Specific Advantages (CSAs) Strong 13 Weak 24 Source: Adapted from Chapter 8 in Rugman, Inside the Multinationals, New York: Columbia University Press, 1981; 25th anniversary edition, Basingstoke, UK and New York: Palgrave Macmillan, 2006. 37 Figure 2: Distance and the Nature of International Business Studies Reach of Firm-Specific Advantages (FSAs) Global NLB FSAs Limited NLB FSAs ENTREPRENEUR 1 BORN GLOBAL 5 BORN REGIONAL Unit of analysis MNE 2 GLOBAL MNEMETANATIONAL 6REGIONAL MNE GEOGRAPHICDIVISIONS

MNE SUBSIDIARY INTRA-FIRM NETWORK 3 GLOBAL MANDATE 7 REGIONAL MANDATE INTER-FIRM NETWORK 4 GLOBAL NETWORK 8 REGIONAL NETWORK 38 Figure 3: MNE Strategies for Growth in a Complex WorldResource Recombination Barriers to Growth Low High Compounded Distance in International ExpansionHigh 1 METANATIONAL TRANSNATIONAL 3 REGIONAL Low 2 GLOBAL 4 LOCAL 39 REFERENCES Aharoni,Y.(1966).TheForeignInvestmentDecisionProcess.Boston:DivisionofResearch, Graduate School of Business Administration, Harvard University. Asmussen,C.G.(2009).Local,Regional,orGlobal?QuantifyingMNEGeographicScope. Journal of International Business Studies, 40, 1192 1250. Andersson,U.,Forsgren,M.&Holm,U.(2002).TheStrategicImpactofExternal Networks:SubsidiaryPerformanceandCompetenceDevelopmentintheMultinational Corporation. Strategic Management Journal, 23(11), 979 96. Andersson,U.,Forsgren,M.&Holm,U.(2007).BalancingSubsidiaryInfluenceinthe FederativeMNC:ABusinessNetworkView.JournalofInternationalBusinessStudies,38(5), 802 18. Bartlett, C. & Ghoshal, S. (1989). Managing Across Borders The Transnational Solution. Boston, Massachusetts: Harvard Business School Press.Barney,J.B.,(1991).FirmResourcesandSustainedCompetitiveAdvantage.Journalof Management, 17(1), 99120.Birkinshaw,J.M.(1996).HowMultinationalSubsidiaryMandatesAreGainedandLost. Journal of International Business Studies, 27(3), 467 496. Birkinshaw,J.M.(1997).EntrepreneurshipinMultinationalCorporations:The Characteristics of Subsidiary Initiatives. Strategic Management Journal, 18(3), 207 229. Birkinshaw,J.M.(2000).TheDeterminantsandConsequencesofSubsidiaryInitiativein Multinational Corporations. Entrepreneurship Theory and Practise , 24(1), 9-35. Birkinshaw,J.M&Hood,N.(1998)MultinationalSubsidiaryEvolution:Capabilityand CharterChangeinForeignOwnedSubsidiaryCompanies.Academy of Management Review, 23(4), 773-795. Birkinshaw,J.M&Hood,N.(2001).UnleashInnovationinForeignSubsidiaries.Harvard Business Review, 79(3), 131-137. 40 Birkinshaw,J.&Pedersen,T.(2009).StrategyandManagementinMNESubsidiaries.In A.M.Rugman,(Ed),OxfordHandbookofInternationalBusiness,2ndEdition(pp.367-388). Oxford: Oxford University Press. Birkinshaw,J.M,Hood&N.,Jonsson,S.(1998).BuildingFirmSpecificAdvantagesin Multinational Corporation: The Role of Subsidiary Initiative. Strategic Management Journal, 19(3), 221-241. Buckley,P.J(2009).InternalizationThinkingFromtheMultinationalEnterprisetothe Global Factory. International Business Review, 18(3), 224 235. Buckley,P.&Casson,M.,(1976).The Futureof the Multinational Enterprise.Basingstokeand London: Macmillan. Buckley,P.&Casson,M.,(2009).TheInternalisationTheoryofTheMultinational EnterpriseAReviewofTheProgressofAResearchAgendaAfter30Years.Journalof International Business Studies, 40(9), 1563-1580 Buckley, P.J. & Ghauri, P.N. (2004). Globalization, Economic Geography, and the Strategy of the Multinational Enterprises. Journal of International Business Studies, 35, 81 98. Buckley,P.J.&Hashai,N.(2005).FirmConfigurationandInternationalization:AModel. International Business Review, 14(6), 655 675. Casson, M. (1979). Alternatives to the Multinational Enterprise. New York: Holmes & Meier. Casson, M. (1982) The Entrepreneur. Oxford: Martin Robertson Casson, M. (1987). The Firm and the Market. Oxford: Basil Blackwell. Caves,R.E.,(1971),InternationalCorporations:TheIndustrialEconomicsofForeign Investment, Economica, 38, 1-27Caves,R.E.(1982,1996).MultinationalEnterprisesandEconomicAnalysis.1stand2ndEdns. Cambridge: Cambridge University Press. Coase, R.H. (1937), The Nature of the Firm, Economica, NS4, 386 405.Collinson,S.&Rugman,A.M.(2008).TheRegionalNatureofJapaneseMultinational Business. Journal of International Business Studies, 39(2), 215-230. 41 Cyert, R.M. & March, J.G. (1963). A Behavioural Theory of the Firm. New Jersey: Prentice Hall. DCruz, J. (1986). Managing The Multinational Subsidiaries. In H. Etemad and L.S. Dulude (Eds.), Managing the Multinational Subsidiary (pp 75-89). London: Croom Helm. Doz,Y., Santos,J.& Williamson,P.(2001).From Global to Metanational: How Companies Win in the Knowledge Economy? Harvard: Harvard Business School Press.Dunning,J.H.(1958).American Investment in British Manufacturing Industry.London:Allen& Unwin. Dunning, J.H. (1977). Trade, Location of Economic Activity and the MNEs: A Search for an Eclectic Paradigm. In P. Wijikman, (Ed.), The International Allocation of Economic Activity. (pp. 395-418). London: Macmillan. Dunning,J.H.(1988).TheEclecticParadigmofInternationalProduction:ARestatement and Some Possible Extension. Journal of International Business Studies, 19(1), 1-31. Dunning,J.H.(1998).LocationandtheMultinationalEnterprise:ANeglectedFactor? Journal of International Business Studies, 29(1), 45-66. Dunning,J.H.&Rugman,A.M.(1985).TheInfluenceofHymersDissertationonthe TheoryofForeignDirectInvestment.AmericanEconomicReview:PapersandProceedings, 75(2), 228 260. Eden,L.(2005).Cave!HicDragones!AlanM.RugmansContributiontotheFieldof International Business. In A. Verbeke (Ed), Internalization, International Diversification and the MultinationalEnterprise,EssaysinHonourofAlanM.Rugman(ResearchinGlobalStrategic Management), Volume 11 (pp 9-27). Oxford: Elsevier.Eden,L.&Dai,L.(2010).RethinkingtheOinDunningsOLIParadigm.Multinational Business Review, 18(2), 13- 34. Egelhoff,W.(1982).Strategyand Structurein MultinationalCorporations:A Revisionof the Stopford and Wells Model. Administrative Science Quarterly, 27(3), 435 458. Fina, E. & Rugman, A.M. (1996). A Test of Internalization Theory and Internationalization Theory: The Upjohn Company. Management International Review, 36(3), 199 213. 42 Fisch,J.H&Oesterle,M-J(2003).Exploringthe GlobalizationofGermanMNCs withthe Complex Spread and Diversity Measure. Schmalenbach Business Review, 55(1), 2 21. Forsgren, M., Holm., U., & Johanson, J. (1995). Division Headquarters Go Abroad: A Step in the Internationalization and the Multinational Corporation. Journal of Management Studies, 32(4), 475 91. Foss, N.J., & Pedersen, T. (2002). Transferring Knowledge in MNCs: the Role of Sources of SubsidiaryKnowledgeandOrganizationalContext.JournalofInternationalManagement, 8(1), 1-19. Glaum, M. & Oesterle, M-J. (2007). 40 Years of Research on Internationalization and Firm Performance:MoreQuestionsthanAnswers?ManagementInternationalReview,47(3), Focused Issue, 307-317. Ghemawat,P.(2001).DistanceStillMatters:TheHardRealityofGlobalExpansion. Harvard Business Review, 79(8): 137 147. Ghemawat,P.(2007).Redefining Global Strategy: Crossing Borders in a World Where Differences Still Matter. Harvard: Harvard Business School Press. Grosse,R.(2005).AretheLargestFinancialInstitutionsReallyGlobal?Management International Review, 45(1), 129-145. Grubaugh, S.J. (1987), Determinants of Direct Foreign Investment,Review of Economics and Statistics, 69(1), 149-52 Gulati, R., Nohria, N. & Zaheer, A. ( 2006). Strategic networks. Strategic Management Journal, 11, 293 309. Hedlund,G.(1986).TheHypermodernMNC:AHeterarchy?HumanResourceManagement, 25(1), 9-35. Hennart, J.F. (1982), A Theory of Multinational Enterprise, Ann Arbor, University of Michigan Press. Hennart, J. F. (2009a). The Theories of The Multinational Enterprise. In A.M. Rugman (Ed). TheOxfordHandbookofInternationalBusiness,2ndEdition(pp125-145),Oxford:Oxford University Press. 43 Hennart, J.F. (2009b). Down with MNE Centric Theories! Market Entry and Expansion as theBundlingofMNEandLocalAssets.JournalofInternationalBusinessStudies,40,1432-1454. Hejazi,W.(2007).ReconsideringtheConcentrationofUSMNEActivity:IsitGlobal, Regional, or National? Management International Review, 47(1), 5-27. Hofstede,G.(1983).TheCulturalRelativityofOrganizationalPracticesandTheories. Journal of International Business Studies, 14(2), 75-89. Hofstede,G.(2006).WhatDidGLOBEReallyMeasure?ResearchersMindsVersus Respondents Minds. Journal of International Business Studies, 37(6), 882 896. Holm, U. & Pedersen, T. (2000). The Emergence and Impact of Centres of Excellence: A Subsidiary Perspectives. London: McMillan. House, R.J, Hanges, P.J., Javidan, M. Dorfman, P.W., & Gupta, V. (2004). Culture, Leadership and Organizations: The GLOBE study of 62 Societies. Thousand Oaks, CA: Sage.Hymer,S.H.,(1976),TheInternationalOperationsofNationalFirms:AStudyofDirectForeign Investment. Cambridge, Mass.: MIT Press.Itaki,M.(1991).ACriticalAssessmentoftheEclecticTheoryoftheMultinational Enterprise. Journal of International Business Studies, 22, 445 460. Johanson, J. & Wiedersheim, P.F. (1975). The Internationalization of the Firm the Four Swedish Cases. Journal of Management Studies, 12(3), 305 22. Johanson, J., & Vahlne, J.-E. (1977). The Internationalization Process of the Firm: A Model ofKnowledgeDevelopmentandIncreasingForeignMarketCommitments,Journalof International Business Studies, 8(1), 23-32. Johanson,J.&Vahlne,J.-E(1990).TheMechanismofInternationalization.International Marketing Review, 7(4), 11-24. Johanson,J.&Vahlne,J.-E(2009).TheUppsal