A2 a Dynamic Capabilities-based Entrepreneurial

7/25/2019 A2 a Dynamic Capabilities-based Entrepreneurial

1/30

PERSPECTIVE

A dynamic capabilities-based entrepreneurialtheory of the multinational enterprise

David J Teece

Institute for Business Innovation, Haas School of

Business, UC Berkeley, USA

Correspondence:DJ Teece, Berkeley Research Group,

2200 Powell Street, Suite 1200, Emeryville,

CA 94608, USA.

Tel: +1 510 285 3300;

Fax: +1 510 285 3271;

email: [email protected]

Received: 15 December 2011

Revised: 27 August 2013

Accepted: 31 August 2013

Abstract

This paper develops a dynamic capabilities-based theory of the multinationalenterprise (MNE). It first reviews scholarship on the MNE, with a focus onwhat has come to be known as internalization theory. One prong of thistheory develops contractual/transaction cost-informed governance perspec-

tives; and another develops technology transfer and capabilities perspectives.In this paper, it is suggested that the latter has been somewhat neglected.However, if fully integrated as part of a more complete approach, it canbuttress transaction cost/governance issues and expand the range of phe-nomena that can be explained. In this more integrated framework, dynamiccapabilities coupled with good strategy are seen as necessary to sustainsuperior enterprise performance, especially in fast-moving global environ-ments. Entrepreneurial management and transformational leadership areincorporated into a capabilities theory of the MNE. The framework is thenused to explain how strategy and dynamic capabilities together determinefirm-level sustained competitive advantage in global environments. It issuggested that this framework complements contract-based perspectives onthe MNE and can help integrate international management and international

business perspectives.Journal of International Business Studies(2014)45,837. doi:10.1057/jibs.2013.54

Keywords: transaction cost theory; transaction cost economics, or transaction costanalysis; internationalization theories and foreign market entry; competitive advantage;dynamic capabilities and capability development; entrepreneurship business strategy;intellectual capital

The online version of this article is available Open Access

INTRODUCTIONA multinational enterprise (MNE) is a business rm that setsstrategy and manages operations for the development and utiliza-

tion of income-generating assets in more than one country inthe pursuit of prots over time. A robust theory of the businessenterprise ought to be able to provide insight into global scope,network characteristics, and the basis of sustained competitiveadvantage (SCA), if any. Accordingly, the study of internationalbusiness should not be divorced from the study of internationalmanagement, and the theory of the MNE should not be a distantcousin to the theory of the business enterprise more generally.However, incomplete global integration and the existence ofheterogeneous national economies and geographies leave specialissues and considerations that a theory of the MNE must embrace,but hasnt yet done so.

Journal of International Business Studies (2014) 45, 837 2014 Academy of International Business All rights reserved 0047-2506

www.jibs.net


2/30

The theory of the rm has long contendedwith issues such as why rms exist and whatdetermines their boundaries. More specically,a robust theory of the rm should also be able toexplain:

(1) why somerms grow and go global while somerms stay domestic;

(2) the product, as well as geographical identity andscope, of the rms activities1;

(3) market-entry timing and mode; and(4) explain the drivers of foreign direct investment

(FDI) and the role of subsidiaries.

Most critically, an acceptable theory of the MNEshould be able to provide insight into how theenterprise builds and protects SCA.2

This paper endeavors to ll voids and inadequacies

in the theory of the MNE and competitive advantageby drawing on scholarship on organizational cap-abilities,3 business strategy, and entrepreneurship.One goal is to bring greater cohesion to the eld ofinternational business by securing convergencebetween internalization scholars and what I willcall international management scholars, such asBartlett, Ghoshal, and Doz, who have come to eschewinternalization theories in favor of other approaches.Another goal is to bring greater integration with theeld of strategic management, which also claims tohave something to say about the SCA of global rms.A third goal is to respond to the challenge of severalscholars to bring the international business literatureinto better contact with entrepreneurship theory.Mark Casson (1986b: 54) some time ago called for adynamic theory of countries advantages using theeconomic theory of the entrepreneur.4 Jones andWadhwani (2007: 2) likewise recognized the oppor-tunity and the need to employ an entrepreneurialperspective to deepen our understanding of aspects ofthe history of global capitalism.

A nal goal is to integrate economic, organiza-tional, and entrepreneurial theories of the rm bydemonstrating how both governance and entrepre-

neurship/capabilities perspectives are needed toshed light on the nature of the MNE, and thefoundations of SCA.5 I agree with my UC Berkeleycolleague Oliver Williamson that capabilities andgovernance perspectives are both rival and comple-mentary more the latter than the former(Williamson, 1999: 1106). I also submit that thecapabilities view encompasses governance/contrac-tual views and can provide the framework withinwhich governance/transaction cost minimizationdecisions take place.6

These four goals constitute an ambitious, multi-disciplinary agenda. That agenda transcends thedeep theoretical issues addressed by Frank Knight(1921) and by Nobel Laureate Coase (1937).7

The structure of the paper is as follows. It begins

with a review of a number of early approaches to thetheory of the MNE, and then identies variousshortcomings, with attendant hints as to how onemight amend these deciencies. The direction oftravel is toward a capabilities theory, which is embel-lished as the paper evolves. The framework is thenapplied to classic MNE questions, and exploratoryinsights are reviewed. The paper builds upon earlierefforts to bring capabilities into the theory of theMNE (Augier & Teece, 2007, 2008; Pitelis & Teece,2010; Teece, 2006a). Because of the richness inexisting theories in the eld of international busi-

ness and prior efforts to bridge some of the divides(e.g., Rugman & Verbeke, 1992, 2003), there isplenty of good scholarship to draw upon and toincorporate into the capabilities/entrepreneurshipframework, thereby hopefully creating a more robustand integrative theory of the MNE, while simulta-neously blurring the lines between the internationalbusiness and international management literatures.

CONTEMPORARY THEORIES OF THE MNEThe internalization perspective has dominatedmuch of the literature on the MNE over the past 30years (Dunning & Lundan, 2008). This perspectiveattempts to explain the reasons for internationalproduction and the phenomenon of the MNE byappealing to market failure considerations. Suchfailures help explain whyrms internalize transac-tions across national borders. However, the perspec-tive does not address the reasons for differential rmperformance.8

The internalization perspective is arguably morerobust than the earlier HymerKindleberger para-digm. The former is substantially an efciency-basedexplanation of FDI and the MNE; the latter a market

power explanation. While Hymer (1968) did note inone article a specically Coasian justication forinternalization, he was deeply wedded to standardtheories of the rm, and to the MasonBain struc-tureconductperformance paradigm of industrialorganization (Dunning & Pitelis, 2008). Hymersanalysis became impaired when he quickly movedfrom determining that the MNE had special advan-tages to asserting that the exploitation of its mono-poly power and monopolistic advantages was themain reason for its existence and evolution, and was

A capabilities-based theory of the MNE David J Teece

9

Journal of International Business Studies


3/30

therefore something to be regulated or otherwiselimited by government controls (Teece, 1981a).9

The internalization school advanced understand-ing of the MNE beyond where Hymer left it, not leastby emphasizing market failures due to contracting

problems. This led to an efciency-based explana-tion of the MNE. There are two prongs (rationales) tointernalization:

(1) transaction costs/hold-up issues that are avoidedby internalization; and

(2) resource transfer cost savings and learningissues, which are facilitated when technologytransfers occur inside the MNE.

The rst prong was advanced by Buckley andCasson (1976), Dunning (1981), Rugman (1981),Teece (1975, 1976, 1981a), Williamson (1981),

and others. This particular internalization

school

sees contractual issues and associated market fail-ures as the crucial reason for internalization. Thisclass of papers can be thought of as representingthe transaction-cost-based, comparative-governance-based, or exchange-based theory of internaliza-tion.10 Early contributions in this vein (e.g., Casson,1979) explicitly viewed it as a two-way street, notingthat internalized transactions could, when circum-stances warranted, be externalized (outsourced), butthat awareness has generally given way to a narrowerfocus on what rms choose to integrate.

Buckley and Cassons (1976) work was the mostthorough early attempt in this genre to extendCoases (1937) paper into the global context. Theyargued that MNEs minimized transaction costs result-ing from the public goods aspects of some intermedi-ate, mostly intangible, assets via global coordinationand the managerial control of these assets. This prongof the internalization school examined the relativeadvantages associated with different entry modes(e.g., exports, licensing, and FDI). In this same vein,Hennart (1982) explored conditions under whichinternational interdependencies could be dealtwith in a transaction-costs-efcient manner through

employment contracts, rather than arms length mar-ket transactions. Rugman (1981) also highlighted therole of MNEs in overcoming market imperfections ininternational markets.11 This version of the internali-zation paradigm has become so pervasive that MarkCasson could quite correctly claim that by the mid-1980sthe modern theory of the MNE is essentially ageneral theory of contractual relations in interna-tional business(Casson, 1986a: 6).

The second and relatively neglected prong tointernalization does not see its essence as resulting

from transaction costs saved because hold-up risksare abated. Rather, it emphasizes the common(organizational) culture of an integrated enterpriseand the ease of coordination inside the rm, ascompared with coordination through the market.

Besides easing potential contractual problems, inte-gration opens pathways to learning, and to sharingknow-how and expertise through cross-border tech-nology and know-how transfer within the MNE. Inthis view, the MNE also provides for easy inter-change of personnel across borders, and for betterappropriability and trade secrecy. It thus mitigatesintellectual property concerns, too, since technol-ogy transfer is to wholly owned business units andnot to third parties, purportedly yielding greatercontrol.

In this second prong of the theory, facilitating

opportunity identi

cation, personnel exchanges,learning, integration, and assisting in technologytransfer are likely to be very important, and cannotall be squeezed under the rubric of economizing ontransaction costs. The essence of the MNE in thisprong of the literature is less about saving on trans-action costs and more about being entrepreneurialand effective in the development, transfer, andorchestration of differentiated organizational andtechnological capabilities (Teece, 1981a). Cantwell(1989) developed a variant of this prong and called itthe industrial dynamics and technological accu-mulation perspective, as it moved the focus awayfrom industrial structure toward industrial evolutionin which FDI led to the generation offresh techno-logical advantagesabroad and at home (2).

This second prong has evolved into a knowledge-based approach to the MNE. Somewhat in the spiritof Teece (1976, 1977a, 1981a), Kogut and Zander(1992) saw the MNE as an instrument for generatingand harboring tacit and explicit knowledge, and fortransferring technology and industrial know-howacross borders. In these formulations, the expansionof enterprise boundaries required and facilitated thetransfer of knowledge. Internal knowledge transac-

tions are preferred, not primarily for transaction costreasons, but because of the lower resource costs oftransmitting knowledge internally vs across a market(Tallman, 2003; Teece, 1976, 1977a). In Kogut andZanders model, opportunism is rejected as anongoing factor because rms exist to provide a socialcommunity supporting voluntaristic actions.

Both prongs of internalization provide importantand relevant insights into the MNE. Cantwell (1989)was early to recognize the need to combine con-tractual frameworks with a theory of capability


10



4/30

development. Notwithstanding his early contribu-tions, international business scholarship has left cap-abilities considerations underdeveloped, to itsconsiderable detriment (e.g., Birkinshaw & Hood,1998; Cantwell, 2009; Langlois, 2007). Because of

the shortcomings of therst prong set out in the nextsection, the time is now ripe for the second prong (i.e., capabilities) to be strengthened, augmented withentrepreneurial considerations, and linked to a trans-action-cost-based comparative governance perspec-tive.12 Once this is accomplished, the second prongought to be sufciently robust to serve as the structurewithin which the transaction cost perspective can benested.

SOME SHORTCOMINGS OF NAKEDTRANSACTION-COST-BASED THEORIES OF THE

MNEIn this section, I rst highlight some shortcomings ofthe quasi-neoclassical transaction-cost-comparativegovernance perspective, and then provide hints asto how these can be remedied.13 Subsequent sec-tions aim to provide a combined entrepreneurial/capabilities conceptual perspective within whichtransaction costs can be contained.

Capabilities and Learning UnexploredEarly contributions to internalization, such as Hymer(1976), Buckley and Casson (1976), and, to someextent, Williamson (1981), drew to varying degreeson neoclassical marginal analysis and ignored orunderplayed the importance of dynamics and, inparticular, learning and capability augmentation.Even when the framework was broadened to includeadditional phenomena, capabilities and learning wereneglected. For instance, in explaining the boundariesof the MNE, John Dunning (1995) suggested thatownership and location matter along with internali-zation factors (his OLI model). Buckley and Casson(1998) seemed to accept these elements, too. In sub-sequent work, moreover, Buckley and Casson have

endeavored to address dynamics, innovation, exibil-ity, real options, international entrepreneurship, jointventures and cultural issues. They have not, however,embraced issues of capabilities in a robust manner.Buckley (forthcoming) summarizes this impressivework and explains why behavioral and sociologicalviews are hard to integrate with internalization, asthey do not follow the rational choice axioms. Whilethe challenge is considerable, the goal (theoreticalintegration) is attainable. Earlier work by Teece (1982)shows that transaction-cost-type and capabilities-type

theories can coexist. Furthermore, one possibleinterpretation of the ownership factor in Dunningis that it is a proxy for capabilities (albeit a static one,particularly in early iterations of the OLI model).14 Itis clear that initial steps toward a capabilities

approach have already been taken.However, even if Dunnings ownership factor is

interpreted as embracing rm-specic factors andnational institutions (systems of innovation and pro-duction), and even if the ownership factor is acceptedas a proxy for rm-level capabilities, there is stilla dearth of theoretical structure and content aroundtheir nature, origins, orchestration, replicability/transferability, and imitability. This is becauseneither transaction-cost-based internalization theorynor OLI explains very well the sources of rm-levelasset ownership and capability advantages vis--vis

competitors. While capabilities are obviously builtin large part through learning, the Ofactor in Dunn-ing has little to say about that (Pitelis, 2007).15 Itis important to recognize that learning is a keymechanism by which rm-specic assets develop.16

In more recent writing, Dunning and Lundan usethe path-dependent resources and capabilities of arm and its institutional infrastructure to explaindynamic growth, and highlight the need to link themicrostructure of capabilities to the evolution of the(institutional) macrostructure (Cantwell, Dunning, &Lundan, 2010; Dunning & Lundan, 2008). Althoughthis recent scholarship has been helpful in enhancingour understanding of the dynamics of the internaliza-tion process of rms, large gaps nevertheless exist.The theory of technological accumulation discussedin Cantwell (1989) remains an important mechanismby which rms build technological capabilities. How-ever, given the ever-greater global dispersion of tech-nology, reliance on in-house R&D as the sole basis ofcompetitive advantage is no longer tenable. Technol-ogies from both within and beyond the enterprisemust be orchestrated effectively to achieve timelydelivery of differentiated products and servicesthat customers value (Augier &Teece, 2007; Pitelis,

2004).

Cross-Border Market Creation and Co-CreationIgnoredMarket creation and co-creation are both entrepre-neurial and dynamic concepts that have always beenseminal functions of the MNE. However, marketcreation and co-creation have been largely ignoredin the rst (transaction cost) prong of the internali-zation literature. These activities are very differentfrom market-entry mode selection decisions, upon


11



5/30

which MNE theory has in recent decades put somuch emphasis (e.g., Brouthers, 2013; Hennart,2009; Zahra, Ireland, & Hitt, 2000).

The transaction cost approach to internalizationtheory has focused on entry mode such as procure-

ment/supply contracts, joint ventures, and whollyowned subsidiaries. To explain entry mode, transac-tion cost theory implicitly assumes preexisting mar-kets, which fail under certain conditions (e.g.,where asset specicity or complex know-how trans-fers are involved), necessitating the emergence of theMNE and FDI to address these failures by internaliz-ing (under a management structure) transactionsthat would otherwise likely evolve in an unfavorableway for one of the parties. However, it has long beenrecognized that the market failure assumption ismerely an analytic convenience. Markets only fail

relative to a hypothetical perfect market, whichrarely exists. Infatuation with market failure and thefunctions (or lack thereof) of markets has deectedattention away from more important issues aroundthe very existence of markets. Market creation andco-creation functions are not merely a response to amarket that has somehow failed to perform (relativeto an idealistic standard). Rather, it is often the casethat the market has quite simply failed to emergeand/or needs to be created or co-created (Pitelis &Teece, 2010) by entrepreneurially managed businessenterprises.17

Put differently, even if markets do exist, they maybe very thin or otherwise imperfect. This is particu-larly true for more specialized, idiosyncratic, anduncertain demand-and-supply requirements andopportunities.18 Hence, rather than solving transac-tional difculties by simply internalizing all activity,entrepreneurial MNE managers must often considerwhat is tantamount to creating markets for ideas orfor products, and bolstering the capabilities of sup-pliers in order to have markets they can sell into orfrom which they can source raw materials andcomponents. A market creation and co-creationview of the MNE is obviously rather different from

contractual approaches.19

It follows from the above that the rationale for theMNE is not just to achieve efciencies (relative to someexternal benchmark) from internal transfers of tech-nology and intermediate products, but also to createand manage co-specialization and, if need be, to createnew markets and expand old ones. Indeed, it isrecognized elsewhere in this paper that a prime reasonwhy MNEs exist is that their cross-border pres-ence, entrepreneurial capacities, and organizationalcapabilities are integral to the market creation and co-

creation process, both upstream and downstream, andalso laterally.

Some consideration of market creation is alreadypresent in Cassons important work on entrepre-neurship (Casson, 1982, 1997, 2005). However,

market-making in his theory is rather neoclassical,overly focused on individual action, and notlinked very well to the MNE. In particular, Cassonsapproach does not seem to recognize the importanceof the capabilities of the enterprise and its manage-ment in shaping markets, inuencing trends, shap-ing demand, and assembling the complementsneeded for new markets to be viable.

The reality that needs to be reected in the theoryis that entrepreneurial MNEs can help build the newecosystems within which global rms operate. MNEsfacilitate investment in complements and other

infrastructure needed for new products to belaunched successfully. By investing in complements,MNEs can enhance the vitality of particular businessecosystems. Ecosystems are thus partly endogenous,as they are often co-created by (global) companies.20

This is in contrast to Porter (1980, 1985) and the basicindustrial economics model, which has the industryas the domain of analysis, and market structuredetermined exogenously. In this paper, the conceptof ecosystems (not the industry) is advanced as theappropriate domain for competitive analysis.

Entrepreneurship Suppressed, EquilibriumAssumed, Management Muted, LeadershipIgnoredThe Coasian view of the rm has resources allocatedand decisions made by a manager who internalizestransactions until an indifference point is reachedwhere the marginal cost of internalization is equal tothe marginal cost of relying on the market. Theseperspectives employ neoclassical tools (both marginalanalysis and equilibrium concepts) to explain man-agement behavior and the nature of the enterprise.Moreover, factors of production and technology aregiven, and prices are (implicitly) known. The eco-

nomic problem (or business model choice) becomesone of whether to engage in market exchange or in(vertical) integration. In a cross-border context, thelatter implies FDI.21 In the Coasian rm, there is atbest a modest role for the manager, no room for theentrepreneur, and no need for the leader.

Williamson (1981, 1985) extended Coase by dee-pening the contractual underpinnings of the trans-action cost framework. In so doing, he created apredictive model ofrm boundaries by featuring therole of asset specicity. In Williamsons contractual


12



6/30

schema, bounded rationality coupled with uncer-tainty leads to an inordinate number of potentialcontingencies, which in turn render complete con-tingent-claims contracting impossible. Contracts aretherefore necessarily incomplete. This in turn leads

to recontracting hazards. These are mitigated, if noteliminated, by internalization. Location issues arenot addressed, except indirectly (i.e., inasmuch asthey impact transaction cost).22

In short, in endeavoring to build a theory of therm, neither Coase nor Williamson focused on theimportant role the business enterprise plays in search-ing for and/or developing new opportunities, eitherat home or abroad. Nor did they feature learning orleadership. Rather, the evolution of the enterprise(including the MNE) is due, in these frameworks, tomanagements desire to minimize transaction costs,

and in particular to guard against opportunism.Opportunity, on the other hand, is almost completelyneglected.

Organizational change is also missing. What lim-ited change there is Williamsons fundamentaltransformation being perhaps the main exemplar is not due to entrepreneurship or innovation. Rather,it arises from past investments that are held up (i.e.,rent is extracted) because of extortionate recontract-ing by opportunistic contractual partners who takeadvantage of changes in bargaining positions onceidiosyncratic investments have been made.

Put differently, internalization theories, to theextent to which they rely on Coase and Williamson,posit ubiquitous contractual problems, which inturn lead to market failures. Internalization over-comes these problems primarily by changing gov-ernance structures. Entrepreneurial and managerialfunctions such as opportunity discovery, learning,and knowledge creation play almost no role in theiranalyses. Neither entrepreneurship nor leadership isneeded or featured.

In reality, however, entrepreneurs and entrepre-neurial managers working in an organizational con-text discover and create new knowledge and help

commercialize new technologies at home andabroad. They learn about new opportunities, andsometimes help create them and transfer technolo-gies as needed. Because the market for information/knowledge about new opportunities (Gans & Stern,2010; Teece, 1981b) isnt well developed, entrepre-neurs and managers must build organizationalcapabilities inside businessesrms to assist in knowl-edge creation and knowledge capture (Teece, 1986b,2006b). To be effective, such rms often need to beglobal in scope. The neglect of these entrepreneurial/

managerial functions in the theory of the rm wouldappear to be a serious omission.23 The neglect of therole of the leader, particularly in organizationaltransformation, is equally serious.

Control

Follows Ownership of (Foreign)

Subsidiaries; Inter-Firm Relationships EnigmaticMost governance-based theories of FDI implicitlyassume, or explicitly state (e.g., Hennart, 2010), thatstrategic control comes through ownership andresides with the parent, that getting incentives rightis not only necessary but sufcient to achieve align-ment of goals and economic efciency, and thatsubsidiaries are just that. In isolation, and puttingnancial constraints to one side, this view of theMNE tends to see the wholly owned subsidiaries asthe preferred organizational form, because protect-

ing speci

c assets from recontracting hazards is themain purpose of the MNE.24

In reality, the common ownership of businessunits doesnt eliminate incentive problems, nor doesit necessarily achieve control. This is particularlytrue in the MNE context, where the identication ofhost countriesemployees with the MNE shareholderis less than perfect. This is well recognized by interna-tional business scholars; but the internalization the-ory of the MNE struggles to nd an elegant way totake this into account. If asset ownership is neithernecessary nor sufcient for incentive alignment, asHelper and Sako (2012) suggest, then internalizationtheory must somehow be modied if it is to capturethe essence of the MNE. Certainly, evidence thatMNEs choose to internalize in order to guard againstrecontracting hazards is weak at best.25

The business historian Alfred Chandler was one ofthe rst to identify the structural virtues of a fullyintegrated rm that was global in scope. In his para-digm, (vertical) integration was necessary to achievecoordination (Chandler, 1977). However, cross-bordercommunication technology has improved dramati-cally in recent decades, and capabilities are moreglobally dispersed than ever. As a result, coordination

in the supply chain seems to be less dependent oninternalization than it once was. The implications forinternalization theories are considerable.

Consider Apple, which is known for the ingenuityof its designs, yet it owns none of its own manufac-turing. It has tight supply relations with many com-panies some pure contractors (e.g., Foxconn,headquartered in Taiwan with factories in China)and at least one competitor (Samsung, headquarteredin Korea). Apple helps provide nancing to some ofits suppliers, and may obtain exclusive purchase


13



7/30

arrangements from them for short periods (e.g., threeyears). Contractual arrangements appear to sufce forApple to achieve the necessary coordination, whileretaining the exibility needed to respond to marketforces. Powerful examples of heavily outsourced com-

panies, such as Apple, serve to remind us that stan-dard contractual approaches to internalization needto be combined with, and arguably embedded into,entrepreneurial, capability, and networkparadigmsof the rm.

These outsourcing arrangements have been stu-died under several names, including internationalproduction networks(e.g., Ernst & Guerrieri, 1998)and global value chains (e.g., Geref, Humphrey,& Sturgeon, 2005). Buckley introduced the expres-sion global factory to characterize a network inwhich Brand owners will control design, engineer-

ing and marketing while outsourcing large areas ofproduction to parts suppliers and they may wellcontract out nal assembly (Buckley, 2007: 115).Buckley claims that his global factory network isheld together by control of key assets and ows ofknowledge and intermediate products (Buckley,2009: 230). The global (virtual) factory conceptappears to be the structure selected by Buckley toaccommodate notions of dynamic capabilities.

Competitive Advantage NeglectedAs Geoffrey Jones notes, the recognition that multi-nationals are profoundly heterogeneous is one ofthe most important lessons from history (Jones,2005: 289). The theory of the multinational rm ineconomics and transaction cost theory is noexception does not deal effectively with thisheterogeneity, and hence cannot address issuesrelating to competitive advantage, that is, the foun-dation for enterprise- or business-level nancialperformance that is both superior (supernormal)and durable. While different governance forms areseen as suitable for certain transaction types, theframework is silent as to how competitive advantageis built and preserved for particular rms. It is true

that the invention of new and superior governancemodes can be a source of temporary competitiveadvantage, but, as a general matter, there is no easyway to protect innovations in governance fromrapid imitation. Hence, governance advantages willerode, sometimes rather quickly.

Moreover, thanks to the very development oftransaction cost economics and its wide dissemina-tion, knowledge of the tools of good governance isavailable in the public domain. Adoption of goodgovernance protocols may of course be impaired by

factors inside the enterprise. If so, new governancemodes may serve to create differentiation. But trans-action cost analysis does not contribute signicantinsights into when and where this is the case.Accordingly, it is of limited relevance to scholars

and practitioners seeking an understanding of com-petitive advantage.

Neglect of rm-level heterogeneity and the parti-culars of competitive advantage is perhaps the pri-mary reason for the schism between the elds ofinternational business and international (strategic)management. The former largely ignores it; the latterembraces it. Until this chasm is bridged, interna-tional business scholars will have little to say tomanagers, and international management scholarswill have little to contribute to public policy or theunderstanding of national competitiveness.

TOWARD A CAPABILITIES THEORY OF THE MNE

Antecedents: Resources and Early CapabilitiesPerspectivesInternalization theories have yielded importantinsights into the MNE. This is true with respect toboth prongs of the approach. However, as the inter-national business eld has matured, and as thetheory of the rm in economics has evolved, rela-tively neglected (capabilities-related) factors nowseem more salient. My early efforts to understandinternational technology transfer, which were madeunder the internalization rubric (e.g., Teece, 1976,1977a), used elements of a capabilities story thattranscended transaction costs. However, in theseearly treatments, capabilities considerations weresoon swamped by contractual concerns (Teece,1985, 1986a). This needs to be readdressed in myown work, as well as in the literature at large.

A capability is the capacity to utilize resources toperform a task or an activity, against the opposition ofcircumstance. Essentially, capabilities ow from theastute bundling or orchestration of resources. Theorganizational and managerial technology of the

rm and its ability to transfer technology (embeddedin routines and resources) across distances and bor-ders are very much implicated in the rms nationaland global capabilities.

The (dynamic) capabilities framework is an entre-preneurial approach that emphasizes the importanceof (signature) business processes, both inside therm and also in linking the rm to external partners.It also recognizes the importance of critical resourcesand good strategy. It is not animated primarily bytransaction cost or contractual concerns. Rather, it


14



8/30

builds on the resource-based approach. It is focusedmore on opportunity than on opportunism, and onthe efcient and effective transfer of technologybetween and among the various organizational unitsof the rm.

Teece (1980, 1982) explored Penroses ideas ofresource fungibility by assessing how the nature of aresource, and in particular itstradability(or lack ofit), affected diversication. A rm with excess factorservices (i.e., above those needed for its current andprojected production program) may nd it moreprotable to monetize those services via a new userather than through a market transaction. While thisresearch focused on product diversication, it wasalso applicable to international diversication.

What has been missing is systematic attention tohow (entrepreneurial) management can deploy or

redeploy the nontradable assets and resources at itsdisposal. Wernerfelt (1984) and Barney (1991) beganthe task of lling this gap, building on Penrosesview that making better use of resources was impor-tant to enterprise growth and development. Penroseherself viewed entrepreneurship as one of theresources of the rm, stating, [W]e include entre-preneurs among the resources of the rm and therange of ideas of entrepreneurs among the servicesrendered (Penrose, 1959: 86). In this regard, she wasperhaps describing a dynamic capability, at least inthe sense referred to here.

Despite Penroses abiding interest in the interna-tional rm, she did not pay particular attention tothe application of her theory to the case of the MNE(Pitelis, 2007). Moreover, although Penrose did recog-nize the importance of entrepreneurship, she did notfully address the roles of entrepreneurs in designingbusiness models or organizations (Augier & Teece,2007), or in building competitive advantage. Rather itwas Wernerfelt (1984) and Barney (1986, 1991) whoarticulated the relationships among rm resourcesand competitive advantage. They focused on thepossession of the right resources as the main mechan-ism for the generation of economic rent. The

resources to worry about were dened by Barney(1991) as those meeting his criteria of valuable, rare,inimitable, and non-substitutable (VRIN). Implicitly,Barney was inviting strategists to focus on intellectualcapital (Teece, 2000), since this is the class of assetsthat most frequently meets the VRIN criteria.

As already noted, Teece (1976, 1977a) andCantwell (1989) early on had threads of a resourceand capabilities approach under development. Thisinitially manifested itself as a knowledge-basedapproach to the MNE with Teece (1981a,1981b) and

Kogut and Zander (1992, 1995). However, knowl-edge-based approaches which emphasize how thegeneration and transfer of knowledge, along withtransaction cost problems in the market for know-how, dictate integration and/or foreign ownership

are insufciently robust to capture many relevantentrepreneurial and capabilities features. Interest-ingly, recent work in international business (someof it generated by advocates of the transaction-cost-based or exchange-based paradigm) has alreadybegun to focus on an entrepreneurial/capabilitiesapproach (e.g., Buckley, 2009; Casson, 2000, 2005;Dunning & Lundan, 2010;26 Pitelis, 2004; Pitelis &Teece, 2010; Rugman & Verbeke, 2003). It is nowtime to carry these efforts forward on a comprehen-sive basis into a (dynamic capabilities-based entre-preneurial) theory of the MNE.27

A robust theory of the MNE that explains its scope,its boundaries, and the role of subsidiaries, while alsoproviding insights into the competitive advantage ofparticular MNEs, requires that the entrepreneurship,resources, and capabilities concepts be somehowamalgamated. The (dynamic) capabilities approach,advanced in the eld of strategic management andapplied below in the context of the MNE, endeavorsto do so, and is the subject matter of most of theremainder of the paper. The goal is to help shape amore robust theory of the MNE that highlights howrm-level sustainable (durable) competitive advan-tage is both built and maintained.

Understanding sustainable competitive advantage(and not just the boundaries of the rm) is a broadermandate for a theory of the MNE than what inter-nalization scholars typically accept. Coming up witha better theory of the MNE that does double duty(i.e., explains boundaries and competitive advan-tage) requires a comprehensive, multidisciplinaryapproach to understanding managerial decision-making and business organization in a contextwhere intangible assets are important, and wherethere are rapid changes, frequent discontinuities,and great complexity, engendered in part by the fact

that the worlds needs and desires have not (yet)become irrevocably homogenized.28 Labor is notfully mobile, and many national institutions remaindistinct. The world is, and is likely to remain, onlysemiglobalized (Ghemawat, 2003). It is this world ofrm-level heterogeneity and semiglobalization andnot the hypothetical worlds of perfect competitionor oligopoly that animates the inquiry here.

The capabilities framework resonates well withCantwells (1989) work.29 Cantwell recognized, cor-rectly, that a theory of the MNE based on an exchange


15



9/30

(or transaction cost/governance) framework masksany active role for management (Cantwell, 1989:215). He also argued that ownership advantages areendogenous and developed through innovation andstrategy, and showed how MNEs extend their capabil-

ities and their overall innovation potential usingglobal networks.

The capabilities perspective can also be seen asconsistent with other international business litera-ture, including recent thinking on knowledge man-agement in a networked MNE (e.g., Rugman &DCruz, 2000; Rugman & Verbeke, 2001, 2002,2003; Vahlne & Johanson, 2013). As noted earlier,some knowledge-based theories of the MNE can bethought of as special cases of a resources/capabilitiestheory of the MNE. An essential characteristic oforganizations/rms is that they can generate and

embody knowledge, which cant be easily boughtand sold. Sometimes the only way to capitalize on

knowledge is to start a rm and build the necessarycomplementary assets (Teece, 1986b, 2006b). Tofully capitalize on opportunities, such rms mustoften be global from the beginning. In short, cap-abilities generally have to be built, as they cannot bebought.

The dynamic capabilities framework developedbelow goes beyond the knowledge and technologicalelements highlighted in earlier research to moreexplicitly include managerial and organizationalcapabilities as determinants of competitive advan-tage. It contends that the active development andastute orchestration of tangible and intangible assetsby both parent and subsidiaries lie at the heart of therationale for the MNE and, together with strategy,determine its longer-run success.30

Denitions and Core Building BlocksThe original denition of dynamic capabilities byTeece, Pisano, and Shuen (1997: 516) referred to theability of an organization and its management tointegrate, build, and recongure internal and exter-nal competencies to address rapidly changing envir-

onments.31 Eisenhardt and Martin (2000) extendedthis to also embrace what I refer to as shaping theenvironment.

Teece et al. (1997) identied the core buildingblocks of dynamic capabilities under the tripartiterubrics of processes, positions, and paths. This wassupplemented in Teece (2007) by a more appliedfocus organized around sensing, seizing, and trans-forming. In what follows, I relate the two taxonomies,and then show how strategy ts in. Important clarify-ing distinctions between ordinary and dynamic

capabilities are made. Application to the MNEfollows.

ProcessesTeece et al. (1997) identied three classes of processes/

managerial functions that are relevant to dynamiccapabilities under the following rubrics: coordina-tion/integration; guided learning; and recongura-tion/transformation. Organizational processes embedthe strategy and business model of the business intothe day-to-day routines of employees. The effective-ness of organizational routines is buttressed bystrong and consistent organizational values.

Dynamic capabilities thus reside, at least in part, inthe managerial, entrepreneurial, and leadershipskills of the rms top management, and in manage-ments ability to design, develop, implement, and

modify these routines. Either way,

rms with super-ior dynamic capabilities have learned to adjust tochanging environments, and also to shape the (busi-ness) environment.

Positions (resources)32

As noted earlier, the asset positioning of a companymatters. I am referring not just to balance sheetassets (plant and equipment and the like) but alsoto human capital and knowledge assets. Teeceet al. (1997) identied technological assets, comple-mentary assets (technological or otherwise), nan-cial assets, reputational assets, market structureassets, and institutional assets. It is obvious that aroad construction company will need access toheavy equipment (e.g., bulldozers and dump trucks),and a home-building company will need access toarchitectural services, as well as construction toolsand skilled and unskilled labor. A bank will neednancial assets, and the talent to build and runsystems for loan origination and underwriting, etc.

The rms position, as dened by its resources, isenhanced if the resources meet Barneys VRIN cri-teria. As I have noted elsewhere (Teece, 2000), theclass of assets most likely to satisfy VRIN criteria is

intellectual capital, particularly technology andknow-how. Intellectual capital readily meets mostof the VRIN criteria because it tends to be tacit andidiosyncratic, and has fuzzy edges. In essence, thecriteria distinguish between ubiquitous resourcesavailable to all at competitive prices and those thatare more specic or unique. Furthermore, the VRINcriteria recognize that a unique asset is not valuablefor its own sake. It delivers value to the rm and itsstakeholders only if it supports a point of differencethat is appealing to the customer, and which,


16



10/30

furthermore, cannot easily be replicated by otherswith different assets.

Needless to say, it should be immediately apparentthat, in fast-paced competitive environments, posi-tions/resources alone are generally ofeeting value.

The way assets need to be deployed (usually inclusters or combinations) is likely to be dynamicand involve astute and entrepreneurial orchestra-tionactivity by management. As noted in Teece etal. (1997: 515):

The global competitive battles in high-technology indus-

tries have demonstrated the need for an expanded para-

digm to understand how competitive advantage is achieved.

Well-known companies have followed a resource-based

strategyof accumulating valuable technology assetsHow-

ever, this strategy is often not enough to support a signicant

competitive advantage.

Clearly, the manner in which assets and otherresources are coordinated and orchestrated is at leastas important to competitive success as the identity ofthe assets themselves. This is where asset orchestra-tion and market creation (or co-creation) come intoplay (Pitelis & Teece, 2010). Whereas neoclassicaland transaction cost economics assume that marketsexist, even if they dont function well, the capabil-ities approach makes no such assumption. Marketsmay have to be created, in the sense that newproducts and services are introduced for whichafter-sales support and product training, forinstance, may be lacking and may have to be built.This is what Singer did globally to allow marketdevelopment of the sewing machine. In India, Gill-ette has likewise been promoting the benets ofremoving beards in order to broaden the market forits safety razors. The need for such creation activitiesto expand markets is assumed away in transaction-based approaches, where there is almost always aparty (or customer) to transact with. This reduces theproblem to one of contracts, when in fact thefundamental problem may be one of market exis-tence or market expansion.

The problem stems in part from what was referred

to earlier as the equilibrium assumption. In a per-fect world of markets (spot, term, future, etc.), therm has full information about competitors, aboutcomplementors in investment decisions, and aboutwhat consumers really want. But, in reality, much ofthis information is proprietary, tacit, or diffuse, andthus inaccessible. The decision to invest depends onsensing an opportunity and also on sensing howpotential competitors and complementors willrespond. This is not a capability required in aneoclassical world of perfect competition.

The focus of the dynamic capabilities framework ison how rms can create, extend, integrate, modify,and deploy their resources and/or specic assets whilesimultaneously managing competitive threats andeffectuating necessary transformations.33 Whereas

other approaches emphasize tangible asset/resourceownership and protection, the dynamic capabilitiesperspective emphasizes intangible assets and resourceaugmentation, and also asset orchestration.

By embedding (managerial) asset orchestrationinto the theory of the MNE, the eld of internationalbusiness can build bridges to topics in internationalmanagement.

Paths (strategy)It is important to recognize that strategy must gohand in hand with processes, resources (positions),

and capabilities. Strategy, when developed success-fully, involves deploying the rms scarce assets tosupport market needs and gain advantage overrivals, while recognizing market and technologicalopportunities and any constraints imposed by therms historical path of evolution.

Put differently, the managerial orchestration thatis core to enhancing processes and exploiting posi-tions must be guided and informed by strategy andvice versa.34 Strategy needs to be consistent, coher-ent, and embrace innovation. While it is necessarilyshaped by the legacy of the past, it also shapes thepath ahead. Strategy will determine which productsto make, which customers to target, how to deploythe rms resources, what the optimal timing will be,and how to keep competitors at bay.

A strategy can be dened as a coherent set ofanalyses, concepts, policies, arguments, and actionsthat respond to a high-stakes challenge (Rumelt,2011: 6). A good strategy has:

(1) prescient diagnoses;(2) a guiding policy; and(3) coherent action.

These three functions constitute what Rumelt

(2011) calls the kernel of strategy.35 A good strat-egy will often not appear fully formed, but insteademerge after a period of trial and error (provided thebusiness environment is sufciently forgiving toallow experimentation). While the actions dictatedby the strategy may be visible to rivals, and freelyimitable, rivals may not perceive it in their interestto do so until it is too late, because the underlyingdiagnosis and policy can be kept secret.

In the framework advanced here, dynamiccapabilities and business strategies codetermine


17



11/30

performance.36 Firms with weaker capabilities willrequire different strategies from rms with strongercapabilities. Strong dynamic capabilities requirerms to sense, seize, and transform in conjunctionwith a sound strategy. A sound strategy must in turn

have a strong kernel.For purposes of operationalizing the framework,

dynamic capabilities can usefully be disaggregatedinto three clusters of processes and entrepreneurial/managerial orchestration activities conducted insiderms (Teece, 2007):

(1) identication and assessment of opportunities athome and abroad (sensing);

(2) mobilization of resources globally to addressopportunities, and to capture value from doingso (seizing); and

(3) continued renewal (transforming).

These activities are required of the rms manage-ment if the rm is to sustain itself as markets andtechnologies change. In a global context, the MNEsmanagement must not only be entrepreneurial, butalso cosmopolitan, or what Perlmutter (1969) calledgeocentric.

It is important to emphasize that the frameworkadvanced here sees the effectiveness of dynamiccapabilities as being compromised by poor strategy.Strategy and dynamic capabilities can be seen asanalytically distinct concepts, although they are

in practice interrelated (Table 1). For instance, sen-sing is important to dynamic capabilities but alsocontains a strong element of diagnosis, which isimportant to strategy; seizing needs to be connectedto both a guiding policy and coherent action; andtransforming that is value protecting and enhancingrequires a guiding policy and coherent action. Thenature of the managerial tasks for various elementsof strategy is outlined in Table 1. Entrepreneurialmanagement is especially relevant to the rmsability to be prescient and sense opportunities andthreats (both market- and technology-related).

Replicability and Imitability: Ordinary vs DynamicCapabilitiesIn the dynamic capabilities framework, considerableemphasis is placed on the replicability and

imitability of organizational processes and positions(Teece, Pisano, & Shuen, 1997). Clearly, if one isinterested in sustainable competitive advantage, oneneeds to take imitability into account. That which iseasily replicated by the rm is scalable, possibly

globally.37

However, that which is easily imitatedby others will clearly not be able to support superiornancial returns. When examining competitiveadvantage, it is therefore critical to distinguishbetween ordinary (and easily replicable) capabil-ities and dynamic capabilities, which by their verynature are hard to replicate. As explained below,ordinary capabilities support technical tness, whiledynamic capabilities support evolutionary tness.The former is about the enterprise doing thingsright; the latter has more to do with doing theright things.

Ordinary capabilities: FoundationsIt is perhaps easier to understand what dynamiccapabilities are as a class by juxtaposing them againstordinary capabilities.38 Ordinary capabilities can bebroken into operational, administrative, and govern-ance capabilities (Teece, forthcoming). Here I empha-size that ordinary capabilities are about producingand selling a dened (and static) set of products andservices. The degree of prociency, however obtained,indicates the strength of the ordinary capability, forwhich practice often makes perfect.

Ordinary capabilities simply allow an existingproduct or service to be made, sold, and serviced.They will not necessarily permit the MNE to growexcept in environments with low competition, notechnological disruptions, and very limited globali-zation.39 When local capabilities in jurisdictionswhere MNEs operate are weak relative to those theMNE can transfer to an afliate, ordinary capabilitiesmay nevertheless allow an MNE to possess competi-tive advantages for indenite periods.

Ordinary capabilities and their diffusion matter tothe MNE.40 They undergird the MNEs technicaltness. Technical tness41 supports static efcien-cies; but unless competition is very weak, anddemand is strong,42 ordinary capabilities are unli-kely to support durable competitive advantage. Suchcapabilities allow an organization to keep earning

Table 1 The interrelation of dynamic capabilities and strategy

Strategy kernel Diagnosis Guiding policy Coherent action

Related dynamic capabilities clusters Sensing Seizing/Transformation Seizing/Transformation

Nature of managerial orchestration Entrepreneurial Administrative Leadership


18



12/30

its living by producing and selling the same product,on the same scale and to the same customer popula-tion over time(Winter, 2003: 992).

Ordinary capabilities enable the rm to performdenable tasks. They rest on (1) non-VRIN resources

and (2) practices, even best practices. The level ofordinary capabilities can therefore be measuredagainst a particular task or standard. Best practice,in a sense, does precisely that.43 Best managementpractices, for example, can be thought of as thosethat continuously collect and analyze performanceinformation, that set challenging and interlinkedshort- and long-run targets, and that reward highperformers and retrain/re low performers(Bloom,Genakos, Sadun, & Van Reenen, 2012).44 Many bestpractices, however, diffuse rather quickly in a worldwhere everyone has access to similar benchmarks.

Bob Lutz (2011), the former vice chairman at Gen-eral Motors, illustrates this point for the automotiveindustry:

The operations portion of the automobile business has been

thoroughly optimized over many decades, doesnt vary much

from one automobile company to another, and can be

managed with a focus on repetitive process. It is thehard

part of the car business and requires little in the way of

creativity, vision or imagination. Almost all car companies do

this very well, and there is little or no competitive advantage

to be gained by trying even harder in procurement,

manufacturing or wholesale.

This statement is revealing, as it indicates how bestpractices, hence ordinary capabilities, are widelydistributed, at least in the global automotive indus-try.45 If so, they can no longer be the foundation ofcompetitive advantage, as elaborated below.

Ordinary capabilities: Replication and transfer.What undermines the power of ordinary capabilitiesto serve as the foundation of competitive advantagefor a particular MNE is that such capabilities can beimitated much more easily today than in earliertimes. A good deal of know-how, which used to be

tacit and proprietary just two or three decades ago, isnow explicit and in the public domain availablefrom consultants, schools of engineering, and thepublic literature.46 Explicit (codied) knowledge tra-vels easily, and the Internet, by allowing low-costaccess to information, has helped enable this. Theimplication is that the barriers to the transfer ofordinary capabilities have been dramatically reducedin recent decades.

Indeed, many basic business services (e.g.,accounting, sales, human resource management) can

today be readily outsourced to computing resourcesresident in the cloud.47 These developments enabled by Internet protocols, the general march ofcomputer-processing power, and the growth of fatclients greatly facilitate starting up, as well as

running, a business. Many routine operational andadministrative capabilities can be supported remotelyby independent providers. Hence, they are no longeras critical to competitive advantage. For example, theimplications ofcloud computingfor the MNE areprofound. In short, the Internet facilitates the avail-ability of ordinary capabilities not just because oflow cost and easy access to the ow of information,as Richard Nelson has emphasized,48 but because oflow cost and easy access to the computing, softwareresources, and data storage needed to support basic,yet high-quality, business functions.

Knowledge transfer within an organization pre-sents a host of difculties (Szulanski, 1996). Andreplicability does not always imply imitability.Knowledge may remain difcult for external organi-zations to replicate to the extent that it is embeddedin interactions among people, tasks, and tools(Argote & Ingram, 2000).

Notably, MNEs investing abroad appear to adoptgood management practices in almost every countryin which they operate (Bloom et al., 2012: 14).Indeed, Bloom et al. found that foreign multina-tionals are generally better managed than host-country rms. MNEs may thus succeed for a whilewith strong ordinary capabilities, because ordinarycapabilities developed at home may temporarily bedistinctive abroad.49

Some less-developed economies still lack domesticrms performing what, from a developed-countryperspective, would be thought of as mundanetasks. Yum! brands success in China, for example,appears to be due in large part to its ability to trans-fer and adapt ordinary capabilities (Starvish,2011). This adaptation is itself partially a dynamiccapability.

Another barrierto imitation is the simple failure

of rivals to implement publicly available best prac-tices (Knott, 2003). Bloom et al. (2012: 13) found intheir study that there is a wide dispersion withrespect to good management practices within everycountry and across countries, as shown in Figure 1.In a survey of more than 10,000 organizations across20 countries, they also found that foreign MNEswere generally better managed (i.e., they had betterordinary capabilities) than domestic rms (Bloomet al., 2012: 23). Brazil and India had a large tail ofvery badly managed rms.


19



13/30

Nevertheless, competition and imitation will, overtime, lead to the erosion of any advantage fromordinary capabilities. This may occur slowly, butcan be rapid in contexts where the absorptivecapacity of external organizations is high. AnMNE subsidiary relying solely on strong ordinarycapabilities in a particular host country willnd, ifthe ordinary capabilities are imitable (e.g., viaknowledge spillovers through employee turnover)and competitors can enter, that its advantage willsteadily diminish. In short, ordinary capabilitieswill not support long-run competitive advantageunless competition is suppressed by governmen-tally or privately imposed entry barriers, or by weakphysical and social infrastructure that preventsordinary capabilities from quickly diffusingthroughout the economy.

DYNAMIC CAPABILITIES: ASSESSMENTSince the late 1970s, local differentiation, globalintegration, and innovation have characterized suc-cessful rms operating globally. In the global econ-omy today, the competitive advantage of thebusiness rm appears to rest on the developmentand deployment of intangible assets,50 relationships,and human capital. These developments haveplaced a premium on the ability of companies tobecome entrepreneurial and agile at home and

abroad, requiring in turn that management operatewith less authority, and organize to allow and pro-mote exibility, responsiveness, and learning. Thisrequires dynamic capabilities.

As already noted, dynamic capabilities are under-girded by processes (routines) and resources (posi-tions). Dynamic capabilities rely not just on bestpractices but on signature practices; not just onany resources but on VRIN resources. They alsorequire astute managerial orchestration guided bywhat Rumelt (2011) has called good strategy. Table2 illustrates this, and contrasts it with ordinarycapabilities.

Signature processes and signature business modelsare beyond industry best practices. Such processesembody a companys history, experience, culture,and creativity (Gratton & Ghoshal, 2005). Because of

their deep roots, they are not so easily replicated byothers who do not share this history, and may havedifferent values, too. Over longer periods of time,such processes and business models may becomesomewhat imitable by others. As Gratton andGhoshal point out, such a transformation occurredwith Toyotas lean manufacturing model, the ToyotaSystem of Production.

Whether signature processes and business modelsare goodmay take some time to become apparent.Eventually, it should show up in key performance

0.8

0.6

0.4

0.2

0.0

0.8

0.6

0.4

0.2

0.0

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

Fractionoffirms

Firm management scores, from 1 (worst practice) to 5 (best practice)

U.S. Brazil China

U.K. India Greece and Portugal

Bars are the histogramof firms in each country

Line is the smoothedU.S. density, shownfor comparison tothe U.S.

Figure 1 Best-practice diffusion.

Source: Bloom et al. (2012).


20



14/30

indicators. However, the replicability of a process orbusiness model is often confounded, particularlyexternally, by what Lippman and Rumelt (1982) calluncertain imitability. This, along with a high tacitcomponent to the underlying knowledge, may keepa signature process effectively proprietary.

There is an obvious opportunity for all businessenterprises to learn, and to embed that learning in

new signature processes and business models.Hence the MNE competing in diverse contexts hasthe opportunity to develop distinct signature pro-cesses and models in different geographies. Accord-ingly, the MNE as such may have an advantage inthe development of new products and signatureprocesses and models, as it can more readily runmultiple, simultaneous experiments than can apure domestic enterprise. Moreover, adaptationand adoption of new processes inside the MNE arelikely to be easier than they would be across unaf-liated enterprises. Certainly, top managementcan endeavor to drive such adoption inside thecompany.

A corollary of the fact that VRIN resources andsignature processes and business models are pro-ducts of the rms heritage and past managerialdecisions is that dynamic capabilities tend to getbuilt, are difcult to imitate, and cannot generally bebought. For example, Tim Cook, a long-time execu-tive at Apple and its current CEO, said in February2013: Apple has the ability to innovate in all threeof these spheres and create magicThis isnt some-thing you can just write a check for. This is some-thing you build over decades (AFP, 2013). This is

the reason for the stickiness of dynamic capabil-ities that is, they dont tend to travel well, they arecomplex, and they are hard to gure out and toimplement.

Once again, Bob Lutz (2011) of General Motors putit most succinctly:

Where the real work of making a car company successful

suddenly turns complex, and where the winners are sepa-

rated from the losers, is in the long-cycle product develop-

ment process, where short-term day-to-day metrics and the

tabulation of results are meaningless.

Dynamic capabilities also help characterize howan enterprise obtains strengths, extends thesestrengths (for instance by developing new businessmodels), synchronizes business processes and mod-els with the business environment, and/or shapes thebusiness environment in its favor (Teece et al.,1997). They are higher-order, difcult-to-replicatecapabilities. Asset orchestration is implicated, and

dynamic capabilities support the

rm beyondmerely achieving superior coordination. They arebased on processes that are beyond best practice, andon resources that meet the VRIN criteria.

Firms with strong dynamic capabilities exhibittechnological and market agility. To achieve this,they use less hierarchy. Agility, coupled with theability to sense new opportunities and threats, sup-ports evolutionary tness.51 This inevitably requiresthat rms constantly create new technologies, differ-entiated and superior processes, and better businessmodels to stay ahead of the competition, stay intune with the market, and even shape the market ifnecessary.52 The rm must be able to simultaneouslycope with changes in the external environmentand with changes caused by processes internal tothe rm (Greiner, 1998).53 It will be aided if it hassufcient resources and superior information, talent,and capital, including relationship capital. However,absent the required ability to orchestrate resources,and to create and execute a quality strategy, suchresources are likely to be of little value.

As noted, strong dynamic capabilities will helporganizations to stay relevant to marketplace needsand technological opportunities. Organizations

must change their capabilities to reect anticipatedchanges in markets, technologies, and the businessenvironment more generally. However, as Winterexplains, change can be reactive rms can easilyget into a re ghting mode, which he describesas high paced, contingent, opportunistic and per-haps creative search for satisfactory alternativebehaviors. Winter (2003: 993) called this ad hocproblem-solving. This is in contrast to routine-directed problem-solving. In Winters terminology,the latter is a capability. He correctly recognized

Table 2 Elements of the capabilities framework

Core building blocks Weak ordinary capabilities Strong ordinary capabilities Strong dynamic capabilities

Processes (routines) Sub-par practices Best practices Signature practices and business models

Positions (resources) Few ordinary resources Munificent ordinary resources VRIN resources

Paths (strategy) Doing things poorly Doing things right Doing the right things (good strategy)


21



15/30

that it is possible that on close examination evenre ghting approaches to problem-solving mayhave micro-routines embedded within. Certainly,skills are implicated.

The individual and organizational skills at issue

with dynamic capabilities are much more orientedto creating unique problem-solving methodologiesand signature processes. Problem-solving is verymuch a dynamic capability.54 There is much dis-tance between the purely routinized and that whichis purelyad hoc. The middle ground also constitutesa (dynamic) capability. Indeed, most invention isntfully directed. The innovation process is neithercompletely routinized norad hoc.

The capabilities approach, expanded upon below,sees MNE activity as driven by the opportunity toleverage capabilities and create and capture value

from innovation on a global scale. Entrepreneurialmanagers are not just resource allocators; they alsosense, shape, and exploit opportunities. A theory ofthe (multinational) rm that doesnt recognize thislogic and these phenomena, and their associatedlocational dimensions, will be unable to explain theMNE's sustainable competitive advantage.55

To create and exploit opportunities globally, entre-preneurial activity must be linked up with capitaland other complementary assets, because propertyrights over discoveries and inventions are incom-plete. Some ownership and control over comple-mentary assets is likely to be needed to assist the

MNE in the appropriation of value needed to sup-port continued investment (Teece, 1986b, 2006b).As explained in Teece (1980, 1982, 1986b), man-agers, entrepreneurs, and innovators cannot justleave it up to the market to line up specic assets

and develop new ones, and integrate them into awell-functioning global invention, production, andmarketing system that provides the theoretical rai-son dtreand management for the MNE. They arethemselves the instruments that make markets workwell. Even if Coasian transaction costs were zero,learning, co-creation, and orchestration functionswould still need to be carried out. The entrepreneu-rially managed MNE is a vehicle designed to do so.

The rm is indeed, as Coase (1937: 388) noted, anisland of conscious power but it is unsatisfactory toframe managerial capacity primarily in transaction

cost-minimizing terms, as Coase asserted. Rather,the functions of management can be framed interms of assisting in the building and/or securingand deploying of VRIN resources and signatureprocesses not typically available for sale (or, if avail-able, not routinely priced in a liquid market). Thebusiness rm is an island of (non-market) resourceallocation orchestrated to enhance learning, valuecreation, know-how transfer, and value capture.These factors help explain why it is necessary totransform internalization theory into an entrepre-neurial/capabilities theory of the MNE. The eco-nomic logic of the framework is reected in Figure 2.

Capabilities (Dynamic)

Organizational Heritage;Managerial Decisions

(Orchestration Processes:

Sensing, Seizing,Transforming)

Resources (VRIN)

Strategy,(Prescient Diagnoses

Guiding Policy;

Coherent Action)

Competitive Advantage

Build

Build

Buy?

Buy?

(Ordinary)

Capabilities

(Generic)Resources

Figure 2 Logical structure of the dynamic capabilities paradigm.


22



16/30

CAPABILITIES AND MNE PERFORMANCEWhile ordinary capabilities are insufcient for long-term survival and growth, dynamic capabilitiesenable the rm to have a better chance of establish-ing and maintaining competitive advantage (and

concomitant superior performance) in economieswhere change is rapid, and intangible assets arecritical to competitive differentiation.56 However, asindicated in Figure 2, the rm also needs goodstrategy.

Dynamic capabilities are hard to develop, anddifcult to transfer across borders, in part becausethey are tacit, in part because they are oftenembedded in a unique set of relationships andhistories, and in part because of uncertain imitabil-ity. In short, dynamic capabilities undergird thefuture of any MNE, because, along with strategy,

they undergird competitive advantage in fast-moving, knowledge-based economies. They oftenlie at the heart of both short- and long-cycle productdevelopment processes.

Bartlett and Ghoshal (2002: 14) noted that asmajor global competitors achieve parity in the scaleof their operations and their international marketpositions, the ability to link and leverage knowledgeis increasingly the factor that differentiates thewinners from the losers and survivors. Theseauthors were tilting toward elements of a dynamiccapabilities framework, because such a framework isalso about linking and leveraging know-how. Goodstrategy, strong ordinary capabilities, scale (in somecircumstances), and strong dynamic capabilities areall needed for long-term growth and survival in theframework advanced here.

As noted, ordinary capabilities are about doingthings right, whereas dynamic capabilities are aboutdoing the right things, at the right time, based onunique processes, organizational culture, and a pres-cient assessment of the business environment andtechnological opportunities.57 By the right things, Irefer to investment in new products, processes, andbusiness models that are in tune with the rms

business environments at home and abroad, andwith its strategy. The late Steve Jobs made a strongstatement with respect to the importance of spend-ing money on the right things:

Innovation has nothing to do with how many R&D dollars

you have. When Apple came up with the Mac, IBM was

spending at least one hundred times more on R&D. Its

about how much you get it. (cited in Kirkpatrick, 1998)

Getting it requires strong dynamic capabilities.The dynamic capabilities perspective goes beyond

organizational t,58 and also beyond a nancial-statement view of enterprise strength, to emphasizerecognition of the most promising opportunitiesand the managerial orchestration needed to create,accommodate, and fashion resources both inside

and outside the rm, at home and abroad. Includedare the external linkages and alliances that arecommon in the global economy, and well documen-ted and analyzed in the international businessliterature.

At a quite general level, dynamic capabilities areabout how an enterprise seizes the future and devel-ops the products, processes, and business models tomeet (and shape) ever-changing markets. Dynamiccapabilities result from superior top managementorchestration skills. They are hard to teach, in partbecause there is a large tacit component (Teece et al.,

1997).The greater the diversity and rate of change inbusiness environments, and the greater the impor-tance of intangible (including relationship) assets,the more critical good strategy and strong dynamiccapabilities become for the MNEs growth and nan-cial performance. To maintain competitiveness, theMNE must develop and maintain asset alignmentboth internally and with collaborating rms. TheMNE and its partner rms must develop and deliverjoint solutions that are in tune with customerneeds in multiple environments. It is not just amatter of selecting the right organizational bound-aries to achieve t, although that is clearly oneelement.59 Strong dynamic capabilities include theprocesses, business models, and leadership skillsneeded to effectuate high-performance sensing, seiz-ing, and transforming. Strong dynamic capabilitieshelp ensure evolutionary tness; ordinary capabil-ities are more attuned to the requirements for tech-nicaltness.60

THE CAPABILITIES APPROACH JUXTAPOSEDAGAINST TRADITIONAL MNE THEORY

Prior sections of this paper have treated the rm in ageneral way, consistent with the need to encompassmultinationality. This section explores particularissues squarely within the established domain ofinternational business studies. It is, after all, thepotential for synergistic interaction between head-quarters and foreign locations (Cantwell, 2009),with their distinct institutional contexts and cap-abilities proles, that distinguishes the MNE casefrom the general theory of the rm. Efforts are madeto explain how a capabilities approach leads to an


23



17/30

all-encompassing perspective that can incorporateboth prongs of the internalization paradigm.

Leveraging Capabilities Through HorizontalExpansionThe specication of market-entry strategies is a keyattribute of various internalization/governance the-ories of the MNE. One way to observe the differentinsights from the two main approaches to internali-zation is to examine how they impact perspectiveson MNE expansion.

The impact of the transaction-cost-based interna-lization approach is already evident in the literatureon the MNE (e.g., Hennart, 2009; Zahra et al., 2000).However, this stream of work tells us very little aboutwhich markets the MNE should create and enter;implicitly, one can perhaps read into internalization

that the right markets are those in which the servicesofrm-specic assets generate value. In other words,transaction-cost-based internalization theories helpspecify entry mode, but not the best direction andtiming for expansion. These are important decisions,and a robust theory of the MNE should be able tohelp explain them.

Clearly horizontal market-entry strategies of theMNE are not just about guring out the right con-tractual mode. Firm-specic capabilities will need tobe assessed, both as to relevance abroad and as totransfer costs. Modications and adaptations may

sometimes be required. Intellectual property issueswill need to be analyzed. Replication of capabilities ina different context may be difcult (Teece, 1976,1977a). In the main, the problems likely to beencountered are not contractual ones; they relatemore to technology (and capability) transfer costs,and the assessment of market opportunity.

In effectuating such transfers, it is important forthe MNE to minimize the liability of foreignness(Hymer, 1976; Zaheer, 1995) while simultaneouslyexploiting home-country benets. As Helfat andLieberman note:

Established rms enter markets where they have pre-entry

resources and capabilities that are similar to the resource

requirements of the market of entry. The choice of geo-

graphic markets is most strongly inuenced by specialized

resources and capabilities, including knowledge of the local

market and tacit technological skills. (Helfat & Lieberman,

2002: 738)61

Thus rm-level capabilities act simultaneouslyboth as a constraint on and as an enabler of whatrms can do with respect to foreign market entry.International expansion will be facilitated when the

rm capabilities align with market needs abroad, andmanagement in the parent or subsidiary can keep itthat way. The unevenness in the global business andeconomic landscape navigated by MNEs createsopportunities to both transfer and deploy existing

capabilities, and create new ones, thereby fuelingcross-border expansion.62

Hence, while the boundaries of the MNE may bepartially determined by transaction costs, in largermeasure they are likely to be determined by capabil-ities and the need for, and difculty associated with,replication and the associated transfer of technolo-gies and capabilities. This aspect of the argumentechoes the theory of technological accumulationoutlined by Cantwell (1989, 1995, 1999). It alsoimplies that MNEs will invest abroad to augmenttheir existing capabilities, as their geographically

dispersed networks facilitate the accumulation oftechnological assets over time.Accordingly, the boundaries of the MNE can be

seen as resulting from entrepreneurial managementdeveloping and assembling the particular constella-tion of specic assets that the rms activities requirein each location where it elects to operate. The MNEbecomes the locus for creating and leveraging pro-ducts and capabilities, and for capturing value fromthis process globally.63 In the language of economictheory, value is achieved not just by minimizingtransaction costs but also by exploiting (throughmanagement actions) the implicit bidask spreadsassociated with the transferof intangible assets.

The MNEs country of origin is a key contextualfactor. Regional and national systems of innovation,for instance, shape rm experiences, knowledge, andcapabilities. Moreover, good management is not uni-formly distributed either. Entrepreneurial manage-ment is a scarce resource that is itself geographicallyconcentrated. Hence, as Madhok and Osegowitsch(2000: 326) note:Home country characteristics havea strong bearing on the evolution of the domesticindustry and the international competitiveness ofindividual rms. Firms are in part products of the

environments in which they were born; by goingglobal, they can tap into regional and national sys-tems of innovation outside the home country. Thecapabilities of the MNE stem in part from the diverseenvironments in which they operate and compete.

However, most country advantages may be opento all that choose to invest in a particular hostcountry. Low-cost labor, for example, is generallyfungible across all foreign entrants. Hence, as dis-cussed below, country-specic advantages are not socompelling in the (dynamic) capabilities-based


24



18/30

approach to competitive advantage. They may helpexplain entry, but they are at best one factor behindcompetitive advantage at the enterprise or businesslevel. In short, the essence of the MNE is that itaccepts, adapts to, and capitalizes on institutional,

cultural, and market heterogeneity while simulta-neously trying to capture economies associated withsome kind of (scalable) advantage in certain assets orprocesses it owns or is currently developing(Teece,2006a: 125).

How and When Do MNEs Enter New GeographicMarkets?The mode of entry into a foreign market is the topicon which internalization theories have beenthought to have their rmest footing. However, asdiscussed above, one cannot fully understand

choices with respect to global expansion mode bylooking at transaction cost/governance issuesalone. At least two other factors are at work. First,the presence of pre-entry capabilities, includingslack resources, matters considerably. An MNE will(and should) be reluctant to enter a foreign market(or even a proximate domestic market) if it doesnthave (or cannot readily access) at least strongordinary capabilities and enough slack to replicatethem without hitting internal resource con-straints. The slack resources at issue might even benancial. Indeed, in Teece (1986b: 296), cash washighlighted as an important factor in explainingthe mode of market entry. As Madhok (1997)notes, the rm boundary issues are largely capabil-ity-related. When a rm has strong ordinary anddynamic capabilities, the incremental costs of FDIare likely to be low.

Conversely, when timing is of the essence andcertain capabilities are absent, joint ventures are likelyto be preferred by the enterprise endeavoring to goglobal. When an MNE enters a foreign market, it willneed to replicate some of the capabilities (processes,skills, etc.) employed in the home market. Adjust-ments may be necessary because the skills and know-

how the MNE possesses in one context might notquite work in a different geographic context. Gettingthis t right requires dynamic capabilities.

Timecost tradeoffs have been analyzed andempirically estimated for technology transfer pro-cesses (Teece, 1977b, 1980, 1986b). If the timecosttradeoff is too steep, managers should associatewith (joint) venture partners who can help attenthem.64 Joint ventures and collaboration not onlyreduce nancial outlays; they often also enhance theMNEs ability to access local capabilities. Thus mode

of entry will depend not just on contractual factorsbut also on who owns and controls the requiredcapabilities, the time it takes to transfer them, andthe timing imperatives of market entry.

Scholars have begun to study why and how some

rms internationalize very early in life (e.g., Rennie,1993; Oviatt & McDougall, 1994). The born globalphenomenon is consistent with dynamic capabil-ities. Small companies can have strong dynamiccapabilities, and may be able to access abroad theordinary capabilities necessary to make their foreignmarket-entry strategies viable. Small entrepreneurialrms can quickly create and (with local partners) co-create new markets abroad.

Recent evidence (Arregle, Miller, Hitt, & Beamish,2013) indicates that prior investment in a regionmade up of multiple countries impacts future deci-

sions to invest in these countries: that is, capabilitiescan be redeployed within regions more readily thanbetween them. This nding, which goes beyondprevious country-level analyses, is consistent withcapability transfer being easier with geographicproximity, and with institutional and language simi-larity. The nding is more consistent with capabil-ities theories than with transaction cost/contracttheories, although the two approaches reinforceeach other, because contracting is also easier wheninstitutions are more alike.

Dynamic capabilities themselves (involving asthey do sensing, seizing, and, ultimately, transform-ing) can in most cases be sequenced over time andacross different geographic markets. It is more chal-lenging if the rm has to perform all three simulta-neously in each of its businesses, and in all of itsmarkets. However, such simultaneity is sometimesrequired.65 For example, Yum! Brands (the owner offast-food brands KFC, Taco Bell, and Pizza Hut) hassimultaneously engaged in rapid expansion inChina, and in retrenchment and transformation inone of its established markets, the United Kingdom.

The Role of Headquarters and the Subsidiary inDynamic Capabilities TheoryThe governance (transaction cost) theory of interna-lization has little to say about the role of headquartersand foreign subsidiaries.66 Capabilities perspectives,on the other hand, provide insights into the respec-tive roles of headquarters and subsidiaries.

The headquarters function is where certain cap-abilities reside. The M-form (multidivisional form)of organization facilitates decomposition of theMNEs architecture by allowing considerable auton-omy to regional/country and divisional managers.


25



19/30

Headquarters can enhance the rms capabilities byallowing and facilitating technology transfers amongthe divisions, and by encouraging and supporting theexploitation of complementarities.67

Top management at headquarters performs a most

important global asset orchestration function in thedynamic capabilities framework. They allocate thenancial resources needed for the MNE to createmarkets outside the home jurisdiction while leavingoperational matters to lower levels of the organiza-tion. As Peter Buckley (2009) notes in his (recent)work on the global factory, the management stylethat network congurations require is vastly differ-ent from conventional command and controlmethods. Indeed, he refers to headquarters as acontrolling intelligence or orchestrator of activ-ities (233). In essence, Buckley is advancing a

dynamic capabilities perspective on the MNE.Subsidiaries nevertheless play a vital role in therms dynamic capabilities. They can generate know-how and capabilities fr

Documents

A2 a Dynamic Capabilities-based Entrepreneurial