The economic ‘theory of the firm,’ currently dominated by the contracting perspective pioneered by two Nobel laureates, Ronald Coase and Oliver Williamson, needs to be augmented to account for these new factors, particularly the superiority of the firm (i.e. internal organization) over ‘markets’ for the creation, transfer, protection, alignment, and realignment of intangible assets. Complementarities and co-specialization are advanced as two emerging concepts of particular relevance to a new theory of the learning firm, an organization ‘which has the capacity to learn effectively and hence to prosper’ (Easterby-Smith and Lyles, 2003: 2).
Context
As explained above, fundamental changes in the global economy are changing the way firms develop and deploy new knowledge. More open and competitive trading regimes have increased the importance of know-how and other intangible assets. There are significant implications for the theory of the firm, if such a theory is to connect meaningfully with the contemporary economy.
This section begins by introducing some of the theories of the firm that have emerged outside mainstream economics. Subsequent sections use the dynamic capabilities framework to reconsider the ‘problems’ for which firms are the solution, showing the complementarity of the contracting and capabilities perspectives. The final section argues that a more complete theory of the firm will recognize that firms exist in part to compensate for weak or non-existent markets for know-how. For the economic system to work, entrepreneurs and managers are required to orchestrate the resources/competences needed for creating and capturing the value of an innovation. Without managers and management, economic theory cannot explain the evolution and growth of the economy.
One would hope that the theory of the firm would provide some insight into firms as they exist today. Unfortunately, whether one uses the lens of transaction costs (e.g. Coase, 1937; Williamson, 1985), ownership perspectives (e.g. Hart and Moore, 1990), incentive perspectives (e.g. Holmstrom and Milgrom, 1994), or other ‘modern’ theories of the firm, nicely summarized and illustrated by Roberts (2004), the many theories available today still seem to caricature firms, at least those engaged in learning and innovation. Mainstream economics must reconceptualize how markets and market processes relate to learning and innovation and to the theory of the firm if economic theory is to have both relevance and rigorousness.
Furthermore, as Gibbons (2005) has noted, many theories of the firm today can more properly be characterized as theories of the boundaries of the firm. Gibbons further points out, following Cyert and March (1963), that the term ‘theory of the firm’ is more apt for descriptive and prescriptive models of firms’ decision-making processes. Gibbons provides an excellent survey of four theories of the firm—what he calls (1) rent seeking, (2) property rights, (3) incentive systems, and (4) adaptations. He makes oblique reference to the resources/capabilities approach which he ‘expects . . . to play key roles in future formal theories of the firm.’ This section and those that follow are designed to turn some of Gibbons’ perceived potential into actuality. The capabilities approach recognizes values in all four streams and incorporates some ideas from each.
First, however, a little history: to help overcome blatant deficiencies in standard production-function theory of the firm, transaction cost economics arose. Coase and Williamson were the pioneers. Both received Nobel Prizes in recognition of their significance. Williamson himself sees the ‘relation between competence and governance as both rival and complementary—more the latter than the former’ (1999a: 1106). Knowledge-based theories indirectly respond to challenges raised by Winter (1988), Demsetz (1988), and others. Emanating from the field of strategic management (e.g. Wernerfelt, 1984; Teece, 1982, 1986), they show considerable capacity to inform the theory of the modern firm.
Dynamic capabilities, co-specialization, and transaction costs
Ronald Coase in his classic (1937) article on the nature of the firm described firms and markets as alternative modes of governance, with a profit-seeking orientation causing the choice between them to be made so as to minimize transaction costs. The boundaries of the firm are set by bringing transactions into the firms so that at the margin the costs of internal organizing are equilibrated with the costs associated with transacting in the market. The Coasian firm has a simple decision-making calculus that supposedly explains the firm’s boundaries. How resource allocation decisions inside the firm are made is not addressed in Coase’s analysis of the nature of the firm.
A substantial literature has emerged since 1937 on the relative efficiencies of firms and markets. This literature, greatly expanded by Oliver Williamson (1975, 1985) and others, has come to be known as transaction cost economics. It analyzes the relative efficiencies of governance modes: markets and internal organization, as well as intermediate forms or organization such as strategic alliances.
Contractual difficulties associated with asset specificity are at the heart of the relative efficiency calculations in transaction cost economics. When irreversible investments in specific assets are needed to support efficient production, then the preferred organizational mode is internal organization. Internal organization minimizes exposure to the hazards of opportunistic recontracting and allows more flexible adaptation (Williamson, 1975, 1985).
In some ways, but not in others, the dynamic capabilities approach is consistent with a Coasian perspective. It conceptualizes the firm and markets as alternative modes of governance. However, the selection of what to organize (manage) internally versus through alliances or the market depends on a number of factors somewhat outside the realm of traditional transaction costs, including the availability and the non-tradability of intangible assets, capabilities, and to some extent on what Langlois (1992) has termed ‘dynamic transaction costs.’18
The notion of ‘non-tradability’ advanced earlier and elsewhere (Teece, 1980) does not precisely match Coasian or Williamsonian concepts of ‘transaction costs.’ There is nevertheless a strong relationship between specific assets and non-traded or thinly traded assets. However, there are reasons why assets are not traded (or are thinly traded) that do not relate to asset specificity and transaction costs as such. For example, there may be no viable business model for licensing certain types of know-how in disembodied form.
Indeed, many companies will simply not license what they consider ‘strategic’ technological assets, especially not to direct competitors. The reason, at one level, is because a contract cannot be written that would compensate the licensor for the likely loss of customers if the licensee uses the licensor’s technology to compete against the licensor.19 Theoretically, a licensor ought to be indifferent between own sales and the sales of a licensee if the royalty rate is set to enable royalties to equalize with lost profits. However, such arrangements are rarely, if ever, seen, in part because there is likely to be ambiguity with respect to which customers and what sales are actually lost to the licensee (a standard transaction cost issue). Accordingly, it is uncommon in the actual world to see exclusive licenses (to direct competitors) when the licensor is able to sell in the same territory. At another level, it may simply be because there are differences in expectations with respect to the profit potential associated with the use of the technology. There are also likely concerns with respect to whether the licensor or the licensee will capture the ‘learning by using’ know-how associated with exploiting the technology. Negotiating and contractually specifying and monitoring sharing arrangements are also likely to be, as Williamson’s framework suggests, very difficult.20
In short, the business model that firms use to capture value from innovation is usually one that involves selling products that contain new knowledge embedded in their design and manufacture. It is rare that firms will rely entirely on an unbundled business model in which patent/trade secret licensing is used as a mechanism to capture value from know-how. Rambus, Inc, and Dolby Labs are amongst the exceptions. Their success with this business model has not been unambiguously positive, especially for Rambus.21
In capabilities-based theories of the firm, the concept of co-specialization is particularly important (Teece, 1986, 2007a). Assets that are co-specialized to each other need to be employed in conjunction with each other, usually inside the firm (Teece, 1980). Co-specialization and the organizational challenges associated with achieving scope economies and seizing new opportunities isn’t the emphasis in the path-breaking scholarship of Ronald Coase, Armen Alchian, Harold Demsetz, or Oliver Williamson. However, it is a phenomenon that requires (theoretical) attention. Some is provided below.
Cospecialized assets are the building blocks of firms. Building and assembling co-specialized assets inside the firm (rather than accessing them through a skein of contracts) is not done primarily to guard against opportunism and recontracting hazards, although in some cases that may be important. Instead, because effective coordination and alignment of assets/resources/competences is important, but difficult to achieve through the price system, special value can accrue to achieving good alignment. This is more easily done inside the firm. Achieving such alignment through internalization goes beyond what Barnard (1938) has suggested as the functions of the executive—which he sees in achieving cooperative adaptation.
The imperative for internalization is not just a matter of minimizing Williamsonian transaction costs. Rather, at least in the dynamic capabilities framework, the distinctive role of the (entrepreneurial) manager is to ‘orchestrate’ co-specialized assets. Performed astutely and proactively, such orchestration can: (1) keep co-specialized assets in value-creating alignment, (2) identify new co-specialized assets to be developed through the investment process, and (3) divest or run down co-specialized assets that no longer yield special value. These goals cannot be readily achieved through contracting mechanisms in part because of dynamic transaction costs (the costs of negotiating, etc.) but also because there may not be a competent entity to build or ‘supply’ the assets that are needed. In short, capabilities must often be built, they cannot be bought, and there is limited utility in labeling this conundrum as a transactions cost problem.
Rather than stressing opportunism (although opportunism surely exists and must be guarded against), the emphasis in dynamic capabilities is on building specialized assets (that cannot be bought) and on change processes (to keep the enterprise aligned with its business environment). These processes include research and development, remolding the business architecture, asset selection, and asset orchestration. In dynamic capabilities, ‘small numbers’ bargaining is at the core, as in Williamson (1975). Importantly, the emphasis in dynamic capabilities is not just on protecting value from recontracting hazards; it’s also on creating the assets that in transaction cost economics become the object of rent appropriation.
The basic unit of analysis for dynamic capabilities is not the transaction (as in transaction cost economies) but the firm and the (largely intangible) specific assets it creates and controls. To the extent that the emphasis in dynamic capabilities is on deals and contracts (explicit or implicit) it is less concerned with avoiding opportunism and more concerned with embracing opportunity. However, there is also considerable emphasis on ‘production,’ learning, and innovation. These considerations are largely absent from alternative theories of competitive advantage and from alternative theories of the firm.
The fundamental economic ‘problems’ to be solved by the (learning) firm
As earlier sections made clear, the fundamental problems solved by the learning firm are not just coordination to overcome high transaction costs (and other issues flowing from incomplete contracts) but also the design of opportunity and value capture mechanisms. These mechanisms can help solve the appropriability problems and help create the new organizational capabilities needed to address new opportunities as they arise. These theoretical challenges require the joining of transaction cost economics and capabilities theory. The problems associated with creating and capturing value are as important as coordination and incentive design in defining the nature of the (learning) firm.
Likewise, the economic problem being addressed here has little to do with incentive design and principal–agent problems. Managing expert talent (literati and numerati) has less to do with metering and monitoring to detect and punish opportunism than it has to do with detecting, monitoring, and metering opportunity.
Alchian, Demsetz, and Williamson have all emphasized opportunistic free riding as one organizing principle. This clearly is an important issue. Williamson assumes, correctly so, that human actors are boundedly rational, self-interest seeking, and opportunistic. The dynamic capabilities framework emphasizes other (arguably less ubiquitous and unevenly distributed but nevertheless more salient) traits of human nature: (1) entrepreneurship and pursuit of high-risk/high-reward opportunities, and (2) foresight and acumen. Williamson (1999a) appears to recognize that skills and foresight are not uniformly distributed. He quotes businessman Rudolf Spreckels—‘Whenever I see something badly done, or not done at all, I see an opportunity to make a fortune.’ Williamson comments: ‘Those instincts, if widely operative, will influence the practice and ought to influence the theory of economic organization’ (1999a: 1089). This statement invites a capabilities-based theory of the firm.
There are other differences between transaction cost and capabilities perspectives. Williamson makes the transaction the unit of analysis, with (the degree of) asset specificity a key explanatory variable in organizational design. In dynamic capabilities, complementary assets and the degree of their co-specialization are important explanatory variables. The firm is the focus if not the unit of analysis.
The utility of transaction cost economics and related frameworks for make-buy-ally and related governance decisions are not in dispute. But transaction cost economics leaves us without an understanding of the distinctive role of the manager. Executives must not only choose governance modes (between market arrangements, alliances, and internal organization), they must also understand how to design and implement different governance structures, to coordinate investment activities, and to design and implement business models, and choose appropriability strategies.
A dynamic knowledge-based theory of the firm is not at odds with Coase, Williamson, Hart, Moore, and others. In the dynamic capabilities framework, opportunism is not held in abeyance, nor are principal–agent and incentive issues ignored. But the essence of the learning firm lies in the generation, configuration, and leveraging of knowledge assets and organizational capabilities to allow the owners (shareholders) to create and capture value.
While the understanding of the existence and growth of the firm can be assisted by transaction cost theory, the advantages of organizing economic activity inside the firm go well beyond savings in transaction costs, however these are manifested. Advantages also flow from the ability of entrepreneurial managers to combine idiosyncratic co-specialized assets not just to achieve ‘scope economies,’ but to create and capture value by offering distinctive services (solutions) to customers while solving the firm’s appropriability problems.
Each of these ‘problems’ has a transaction cost dimension; but, in some cases, over-reliance on the transaction cost economics apparatus will add unnecessary baggage. If, for instance, one wants to understand issues surrounding creating value, not simply protecting value created, transaction costs can only go part of the way. The firm’s routines for sensing, seizing, and transforming can provide a basis for profitability well beyond the avoidance of contracting costs and hazards.
There is empirical evidence that outsourcing decisions do not depend on transaction cost (asset specificity) considerations alone. Studies show that ‘system effects’ such as interdependencies and complementarities (Monteverde and Teece, 1982)22 and capability advantages (Argyres, 1996) impact economic organization in a statistically significant manner. These studies seem to indicate that boundary placement influences production learning and impacts research and development efficiency (Armour and Teece, 1980), resulting in lower costs and superior innovation potential.
What then is the role of managers in the theory of the firm? They are not primarily micromanaging creative people to stamp out opportunistic behavior. Nor are they merely engaged in adaptive sequential decision making. Rather, they are helping the organization to create and implement the systems and structures that enable the firm to sense opportunities, act on them, and transform as the environment changes, which inevitably it will.
Opportunism is controlled not just through metrics and internal organization monitoring, but also through high commitment cultures/values. Innovative firms in which a great deal of learning must take place to commercialize new ideas typically need strong values because it’s difficult in the loosely structured internal environments that innovation requires to define and measure performance and implement rigid controls. Incentive issues are powerful as well; creative and entrepreneurial activities need to be encouraged and rewarded.
The complementarity between capabilities-based views and contractual/transaction costs/property rights views is hopefully apparent. It has been remarked on by this author elsewhere, as well as by others (e.g. Foss, 1996).23
The transaction cost economics perspective clearly needs dynamic capabilities, and vice versa. Transaction cost economics assumes what might be referred to as capabilities neutrality. In transaction cost economics, so called ‘production costs’—which might be thought of as a proxy for the firm’s level of (operational) capability—are assumed to be the same across organizational types so that the choice between market and non-market arrangements swings entirely on transaction/governance costs. This assumption is a natural connection point to capabilities theory, which clearly indicates that the levels of capabilities are themselves a function of managerial activity/excellence (or lack thereof). Differences in capabilities can lead to wide disparities in ‘production’ costs within an industry. The very essence of the field of strategic management is built on the recognition that firms are different—not just in governance, but with respect to other features too (Rumelt, Schendel, and Teece, 1991)—and that this drives performance differences.
The (dynamic) capabilities framework, which posits that knowledge assets and their (dynamic) management have become central to profit maximization in an era of globalized commerce and information, suggests a new theory of the firm, one that is consistent with the observation of Alfred Marshall that:
‘capital consists in a great part in knowledge and organization: and of this some part is private property and the other part is not. Knowledge is our most powerful engine of production—organization aids knowledge.’
(Marshall, 1898: 213)
The proposed new capabilities-based theory opens up the black box of the firm and injects, into economic theory, new considerations which are generally not central to the theory of the firm as commonly presented.
Recapping complementarities, co-specialization, and the scope of the (learning) firm
The theory of the (learning) firm has benefited, and can benefit further, from a more rigorous exploration of the concepts of complementarities and co-specialization. The earliest use of the idea of complementarities in economics can be traced to Edgeworth (1881). Early applications to the economic development literature include Hirschman (1958) and to the innovation literature can be found in Rosenberg (1979; 1982) and Teece (1986). Work on complementarities in a strategic context includes Teece (1980), Milgrom and Roberts (1990a, 1990b), and Miller (1988).
Rosenberg notes: ‘Time and again in the history of American technology it has happened that the productivity of a given invention has turned on the availability of complementary technologies . . . these linkages are both numerous and of varying degrees of importance’ (1979: 26–27). Furthermore,
the growing productivity of industrial economies is the complex outcome of large numbers of interlocking, mutually reinforcing technologies, the individual components of which are of very limited economic consequences by themselves. The smallest relevant unit of observation, therefore, is seldom a single innovation but, more typically, an interrelated clustering of innovations.
(Rosenberg, 1979: 28–29)
Complementarities exist when various activities reinforce each other in such a manner that performing multiple activities together lowers/(raises) cost, increases economies/(diseconomies) of scope, or otherwise improves/(depresses) payoffs.24 More technically, complementarities exist when the mixed partial derivatives of a cost function or a payoff function provide positive returns at the margin associated with one variable increasing as the levels of other variables increase too. Doing more of one activity increases the returns from doing more of another. The aggregate economic value achieved by combining two or more complementary factors therefore exceeds the value that would be achieved by applying these factors in isolation.
Of course, as pointed out by Teece (1980), this in and of itself has no direct implication for the theory of the (boundaries of the) firm, although it has powerful implications for economic organization more generally. The existence of positive complementarities indicates the advantage of having separate activities occur together. However, without more structure to the concept, one cannot predict where the individual firm boundaries should lie, because contractual arrangements exist that, in theory, can enable joint activities to take place without common ownership of the parts.
While the importance of complementarities is now being recognized, the approach still needs additional specificity (with respect to causal relationships amongst key constructs) to allow it to morph fully into a falsifiable theory. Put differently, a robust theory of complementarities that provides economic insight is yet to emerge. While there is little doubt that complementary relationships exist among heterogeneous factors inside the firm (and that these can impact firm performance), the contexts in which such interactions occur is yet to be adequately specified. However, some evidence has been assembled. Monteverde and Teece (1982), while testing for the importance of asset specificity in predicting outsourcing decisions for General Motors and Ford, also found that a ‘systems effect’—defined as ‘the degree to which any given component’s design affects the performance or [system-level integration] of other components’ (1982: 210)—was statistically significant in explaining both firms’ outsourcing decisions. The longstanding notion of strategic ‘fit’ is obviously consistent with notions of complementarity.
It should be noted that the notion of complementarity can be applied at a high level of aggregation, as with the Toyota System of production, or at a high level of specificity, such as the complementarity between the (integrated) design and manufacture of automobile components, e.g. an exterior grill and headlamp assemblies (Monteverde and Teece, 1982). Parmigiani and Mitchell (2009) use the example of automobile dashboards, which they note typically consist of multiple, interrelated, complementary components. Both levels of aggregation seem to provide insights, suggesting the power and generality of the complementarity insights.
Complementarities expressed through their mathematical corollary (supermodularity) break from classical economics. Most classical economics models of production recognize only traditional ‘factors of production’ like labor and capital and assume homogeneity with respect to the distribution of these factors amongst firms. The standard production function sees no benefit from the use of particular inputs—in the sense that, apart from diminishing returns related to fixed factors, there is no special significance to the identity of particular factors of production (Teece and Winter, 1984). Moreover, everything is infinitely divisible—indeed, twice differentiable—and firms maximize some objective function subject to constraints. Complementarity does not require divisibility; changes in one variable may require discrete (non-incremental) changes in another.
With production functions of the standard kind, decision makers need only equate marginal revenues to marginal cost and they will deliver global maxima in output. There are serious issues with this theory surrounding the search for, and the discovery of, a global maximum, if one exists. Complementarity modeled as supermodularity enables some departures from this extreme caricature by at least recognizing local maxima. It also accepts that payoff functions may be discontinuous. Design choices are recognized as being discrete and not necessarily continuous. These perspectives have received endorsement from organizational ecologists and strategic management scholars including Levinthal (1997), Porter and Siggelkow (2008), and Teece (2007a).
However, capabilities theory at present runs the risk of providing more ex post rationalization than ex ante guidance with respect to the particulars of the requirements—with Teece (1986, 2006) being the possible exception since these papers are quite explicit about the contexts in which complementary assets are important for capturing value from a specific innovation. These papers are also able to specify when complementary assets should be included inside the boundaries of the enterprise.
Toward a unified theory of the firm
Knowledge-based theories of the firm see business organizations as accumulating capabilities in path-dependent ways. Recognizing, creating, and exploiting complementarities is very much at the core of what firms do. Sustained ‘abnormal’ or ‘supernormal’ profitability occurs because factor markets for certain types of assets (particularly intangibles and idiosyncratic physical and human assets) are not fully efficient. To take full advantage and earn superior profits, firms need to sense, seize, and transform in ways that exploit inefficient factor markets. Indentifying and securing combinations and permutations of assets which enable the enterprise to address customer needs is key.
As firms build the microfoundations needed to sense, seize, and transform, all the while exploiting complementarities, they lay the foundations for sustained above-average profitability. There is nothing in Ronald Coase’s or Oliver Williamson’s work to explain how firms identify and exploit complementarities and develop competitive advantage. This raises the question of how the Coase/Williamson conceptualizations of the firm relate to dynamic capabilities.
As stated earlier, the knowledge and contracting perspectives are complementary theories/frameworks. No theory of the firm can ignore contractual issues. But neither Coase nor Williamson see a firm as a pure nexus of contracts. Nor do they see firms as merely ‘social communities in which individual and social expertise is transformed into economically useful products and services by the application of a set of higher-order organizing principles’ (Kogut and Zander, 1992).
There is clearly a way for knowledge-based theories and transaction cost perspectives to be brought together. Arrow (1974) provided a commanding and potentially unifying insight. He observed that the reason firms exist is not simply due to high transaction costs; rather, markets in some situations simply do not work and there is market ‘failure.’25 One can perform a thought experiment and conclude that if the transactions were forced into a market, transaction costs in such circumstances would be very high; but it’s perhaps simpler to just recognize that there are many circumstances where internal organization is clearly a necessary and superior way to organize, and it is desirable for innovative activity to take place inside a firm orchestrated by entrepreneurial managers and surrounded by some kind of management structure.
For purposes of building an applied theory of the learning firm, it is important to specify the contexts in which these market failures are prevalent. The most important (and also the most under-researched) domain within which organization inside the firm is likely to be necessary is the creation, transfer, protection (appropriability), and orchestration (so as to exploit complementarities) of know-how and other intangibles.
As the author noted three decades ago:
unassisted markets are seriously faulted as institutional devices for facilitating trading in many kinds of technological and managerial know-how. The imperfections in the market for know-how for the most part can be traced to the nature of the commodity in question.
(Teece, 1981a: 84)
The market is also imperfect as a tool to create know-how. One can ‘buy in’ technology more easily than one can have it created through a contractual agreement and then transfer it in. ‘Creation’ must frequently be done internally, even though external sourcing is usually a necessary complement to own development. ‘Co-creation’ is perhaps the more helpful concept.
It is only after industrially relevant know-how is first created that it can be traded (via licensing arrangements). Even once it is created, mutually beneficial trades frequently don’t happen because the property rights covering know-how may be poorly defined (fuzzy),26 the asset difficult to transfer, or its use difficult to meter. Internal resource allocation within the firm (a managerially directed activity) is the only viable alternative.
Moreover, because of complementarities and co-specialization, many intangible assets may be more valuable when they can co-evolve in a coordinated way with other assets. The ability to assemble unique configurations of co-specialized assets, as in the case of systemic innovation (Teece, 2000), can, therefore, enhance value. Rosenberg (1979) seems to go further and argues that such coordination and clustering is necessary for value to be created.
In short, managers often create great value by assembling particular constellations of complementary and co-specialized assets, especially knowledge assets, inside the enterprise to produce highly differentiated and innovative goods and services that customers want. This process of identifying, assembling, and orchestrating constellations of complementary and co-specialized assets is a fundamental function of management—and points to the fundamental nature of the modern firm. It’s different from the Coasian firm.
In a globalized, knowledge-based economy, firms can secure short-term advantage from the coordination of bundles of difficult-to-trade assets and competencies, at least when such assets are scarce and difficult to imitate. Advantage that is sustainable over a longer term, however, can only flow from unique abilities possessed by business enterprises to continuously shape, reshape, and orchestrate those assets to create new technology, to respond to competition, achieve critical market mass, exploit complementarities, and serve changing customer needs. The particular (non-imitable) orchestration capacity of a business enterprise—its dynamic capabilities—is the irreducible core of the learning firm. It cannot be reproduced simply by assembling a constellation of contracts.
Fundamentally, business firms know how to do things. Most figure out how to adapt and possibly even shape their environment to some (small) degree. As noted earlier, even Harold Demsetz was willing to see the firm as a repository of knowledge.
However, it is not clear that many economists are willing as yet to recognize the implications of firms being repositories of knowledge and instruments for learning. One exception is Sidney Winter who correctly notes, ‘it is the firms, not the people who work for the firms, that know how to make gasoline, automobiles, and computers’ (1982: 76).
Organizational capabilities are what explain why an enterprise is more than the sum of its parts. They also help explain why the profits of the enterprise cannot be completely competed away in factor markets. Employees can come and go to a certain extent and the organization can continue without interruption.
Mainstream theory too often takes production functions and production sets as given, ignores complementarities and co-specialization, and fails to explain capabilities and heterogeneity amongst firms even in the same industry. Mainstream theory also completely sidesteps the problem of how firms actually perform the tasks of storing the knowledge that underlies productive competence, transferring it internally (or externally), augmenting it in value-enhancing ways, and identifying and exploiting complementarities.