Analysis of Dynamic Capabilities and Transaction Cost Theories on Example of R&D Practices Maria Rumyantseva (London, England) Research and development (R&D) is the central source of technical innovation and competitive advantage for firms (Mansfield, 1962; Arrow, 1962; Tushman and Anderson, 1986). Recently, firms have begun to take a fresh look at how to optimize the value of their R&D activities by restructuring these activities (Chesbrough, 2003). This kind of restructuring, in some cases, includes the outright transfer of R&D units. An important question in this context is how firms can make the best use of R&D capabilities within and beyond the boundaries of the firm. There are two important elements in the externalization process – the structuring of the transaction and the subsequent relationship between the firm and its separated R&D unit. The first component is related to a number of theories, the majority of which are rooted in the economics literature.1 Of these, the transaction cost theory (Williamson, 1975; 1985) is the most relevant conceptual framework that addresses the mechanism of such transactions. This theory has been successfully adopted in the management literature and is frequently used in combination with the theory of the firm (Madhok, 2002; Afuah, 2002; Delmas, 1999). The second component, the relationship between the original owner and externalized R&D units, can be viewed through the lens of a range of theories.2 The most comprehensive framework explaining such a relationship is offered by the dynamic capabilities theory (Teece et al., 1990).3 Through its focus on knowledge-intensive processes, which determine the development of the firm and its entities over time, this theory possesses the potential to explain inter-firm relationships and their impact on the firm (Zollo and Winter, 2002). While the joint analysis of the dynamic capabilities and transaction cost theories is pursued in a number of studies (Langlois and Foss, 1999; Nooteboom, 2004; Jacobides, 2005), most of them remain conceptual in nature (Hoetker, 2005). Both theories – transaction cost and dynamic capabilities – in combination appear to be able to provide new insights into the value of externalization. Such an analysis is highly interesting from a theoretical perspective, as the investigation into the R&D externalization process could close conceptual gaps when combining those two partially complementary theories. Particularly, the dynamic capabilities theory favors a flexible organization and disintegration of the value chain4 according to the specialized capabilities of the firm (Collis, 1994; Jacobides and Winter, 2005), but it does not explicitly consider mechanisms enabling such reorganization. It further avoids specification of the firm-market interface, which is essential when discussing the disintegration of the value chain. The transaction cost theory, on the contrary, offers a comprehensive explanation of the firm-market interface (Williamson, 1985; 1991), but avoids acknowledging the existence of capabilities.5 This is due to the theory’s focus on the short-term efficiency of transactions, which runs counter to the long-term perspective necessary for the explanation of capabilities (Langlois and Foss, 1999). These conceptual gaps indicate that both theories provide complementary explanations which, if analyzed jointly, could contribute to their refinement. While R&D externalization offers interesting insights for theory, the study of this process is equally important for managerial practice. Although there is evidence for the transformation from a contract to a relationship between the two parties (Gartner, 2004), and a growing number of companies prioritize this emerging process at a corporate level, its execution remains limited to a standard contract (Craig and Willmott, 2005). The discrepancy between the increased importance of R&D cooperation and the unchanged structuring of externalization transactions could potentially lead to a failure to achieve the objectives of the transformation. A combination of these two eminent streams of research could provide a powerful impetus for the emerging field of studies of innovation processes. ^ Dynamic Capabilities Theory The theory of dynamic capabilities is subject to frequent revisions and refinements, which pose constrains on the attempts to provide a synthesis of its principles. To avoid oversimplification of the explanation of this theory in a too generic form, this paper is structured so as to allow for the various views of dynamic capabilities to be simultaneously presented. ^ Underlying Theories As an understanding of the sources of competitive advantage gradually evolves in management science, the proponents of the dynamic capabilities theory argue that at the current stage of strategic management theory development, the dynamic capabilities perspective is the most advanced concept at explaining the origin of the competitive advantage of a firm (Augier and Teece, 2004). In this perspective, the analysis of the historical development of the interpretations of competitive advantage allows one to follow their evolution and, therefore, to reconstruct the emergence of the dynamic capabilities theory. Chronologically, the competitive advantage was, first, attributed to a firm’s positioning on the market (Porter, 1980, 1985), later, to the conditions enabling sustainability of such advantage (Wernerfelt, 1984; Barney, 1986; 1989; 1991; Dierickx and Cool, 1989; Peteraf, 1993), and, finally, to the organizational capabilities of the firm (Barney, 1992; Teece et al., 1997; Lado and Wilson, 1994). While the dynamic capabilities theory emphasizes the role of capabilities, it also synthesizes principles from the approaches antecedent to it. Therefore, the theory offers “an integrative methodology and perspective in which several theoretical traditions are used as tools for analyzing the dynamics of business organizations” (Augier and Teece, 2004: 20). The major contributions of other theories to the dynamic capabilities concept are discussed in this section. They include (a) the resource-based theory, (b) the innovation studies, (c) the behavioral theory, (d) the evolutionary theory, and (e) the knowledge-based theory. Their joint influence on (f) the dynamic capabilities theory is discussed in the summary of the subchapter. (a) Resource-Based Theory Finding its conceptual roots in the resource-based theory, the dynamic capabilities theory has inherited a number of theoretical assumptions. The key assumption of the resource-based theory, which certainly influenced the formation of the dynamic capabilities concept, is the growth of a firm. Starting with the seminal work of Edith Penrose (1959) the argument of a dominant role of growth in the development of a firm became an indispensable part of the theory of the firm.6 The concept of growth presented by Penrose encompassed several previously overlooked principles. First, in this approach the corporate growth is attributed to the productive resources of a firm. The bundles of such resources constitute a firm and are differentiated by Penrose as physical and human. Although they are considered as the only assets of the firm, the role of organizational functions is also briefly mentioned. For instance, it is indicated that the industrial R&D, particularly if the firm focuses on more generic R&D activities, could assist a firm in finding new business areas (ibid: 116-117). Second, the crucial element of this concept is the assumption that managers could learn and through the learning increase the resource base of a firm. As Penrose commented: “Many of the productive services created through an increase in knowledge that occurs as a result of experience gained in the operation of the firm as time passes will remain unused if the firm fails to expand” (ibid: 54). Penrose saw a capacity in management to learn and reconfigure internal resources as the main source of the firm’s growth. However, if a firm does not realize its potential to expand, it accumulates a slack of resources. The idle resources stemming from managerial learning remain “in the concealed form of unused abilities” (ibid: 54) rather than staying in a form of tangible assets. In this perspective the firm is seen as a fluid organization which constantly learns, creates new resource combinations, and keeps some internal space for resource maneuvering. Penrose defined the latter organizational quality as “fungibility” and saw the firm as possessing an option to search for the best internal use of available corporate resources. As managers learn, they initiate the change in resource combinations: “The possibilities of using services change with changes in knowledge … there is a close connection between the type of knowledge possessed by the personnel in the firm and the services obtainable from its material resources” (ibid: 76). The assumption of resource fungibility and the role of learning in the organization have allowed further theorizing of the capabilities of a firm, which, ultimately, has led to a conceptualization of organizational capabilities (Amit and Schoemaker, 1993; Lado and Wilson, 1994; Teece and Pisano, 1994). However, while focusing on a firm’s internal processes, Penrose did not elaborate on the external constrains of growth and left an impression that the firm could expand eternally. Similarly, later studies have underemphasized the role of constrains on a firm’s growth. However, it was not so much the second statement that managers learn and develop unused capacities (the Penrosian effect), but merely the first concept of the firm as a pool of resources which has received most attention in subsequent theoretical developments. On the basis of this concept, the theory has introduced a fundamental assumption that in order to be successful a firm has to grow. This implies an internalization of necessary and excess resources and means that the growth is a primary goal of a firm where other strategies, not pursuing it, are neglected. As Brian Silverman explains: “Her (Penrosian) framework is unidirectional – firms grow, but never shrink; firms acquire, but never divest” (Silverman, 1999: 1112). In spite of this misbalance the approach is widely accepted. Studies of different organizational forms of a firm, such as vertically integrated companies and conglomerates (Rumelt, 1984; Williamson 1975; Monteverde and Teece, 1982), have implicitly understood growth as a success and created a somewhat biased perception of the development of a firm. (b) Schumpeterian Tradition and Innovation Studies The ideas of Joseph Schumpeter contributed to the initiation of a number of developments in economics and management theory. Among others they contributed to evolutionary economics, organizational theory, and theory of technology and innovation, which have directly influenced the formation of the dynamic capabilities theory (Augier and Teece, 2004). The name of Schumpeter (1942) is strongly associated with the literature on innovation, which has been guided by his influential assumption of the fundamental role of innovation for the economic growth and corporate development. The key hypothesis of his concept that a firm’s size influences its innovative output, has since led many researchers to investigate forms of organization by contrasting incumbent firms and new entrants. A basic assumption usually made in such analyzes is that firms remain statically entrenched in their category, at least for the period of study. Similarly, studies providing a more complex picture of the innovation process, such as technology life cycle studies, often profoundly investigate organizational forms that are efficient at supporting innovation during particular phases of the technology cycle, but also tend to neglect their potentially dynamic nature (Abernathy and Utterback, 1978; Nooteboom, 1999). This view circumvents the role of incumbent firms as inactive entities passively accepting their destiny and not attempting to improve their competitive position. Such a perspective was broadly adopted in the theoretical literature (Arrow, 1962; Tushman and Anderson, 1986; Tripsas, 1997). A persistent failure of established firms in the face of radical innovation has been also documented in empirical studies (Cooper and Schendel, 1976; Majumdar, 1982; Tushman and Anderson, 1986; Henderson and Clark, 1990; Christensen, 1993). In summary, Schumpeter speaks about innovations done in a form of novel combinations by new industry entrants. Consequently, it remains somewhat unclear how large incumbent firms could change their innovation structures in order to adapt to a new environment, and whether they are capable of transforming internal structures to an ultimately different, more economic state. (c) Behavioral Theory of the Firm The behavioral theory of the firm was not intended as a theory of strategy (Simon, 1956; 1987). However, several insights from this perspective are important for the construction of the dynamic capabilities theory (Teece et al., 2002). In terms of relevance to strategy, the most basic contribution of the behavioral theory of a firm is in its appreciation of the heterogeneity of firms (Teece et al., 2002). The behavioral theory is built around a political conception of organizational goals, a bounded rationality of expectations, an adaptive nature of rules, and an understanding of how the interactions among these factors affect decisions in a firm (Cyert and March, 1963). Cyert and March (1963) emphasized the uniqueness of a firm. According to their approach, organizations and organizational actors differ in terms of their intentions, their knowledge, and their decisions. The ideas of behavioral theory on adaptation and dynamic character of expectations and goals have been borrowed by the dynamic capabilities theory (Augier and Teece, 2004). Although dynamic capabilities involve much more than “coordination” and “adaptation”, also emphasizing proactive search, selection, and subsequent implementation of particular strategies, the notions of coordination and adaptation serve as an important background for its development. (d) Evolutionary Theory of the Firm The proponents of this theory argue that as firms face strategic decisions on the basis of their past history, it is natural to view questions related to the development of strategy and competencies in an evolutionary setting (Simon, 1993; Winter, 2000). The evolutionary theory of the firm inherits Alfred Marshall’s understanding of a firm’s size, which he explains as: “Firms rise and fall, but the representative firm remains always of the same size” (1925: 367). According to this concept, the equilibrium in the industry is a combination of both a disequilibrium of the population of firms and an equilibrium of the supply and demand in the industry. The “representative firm” is thought to bridge the dynamic analysis at the level of a firm and the static one at the industry level. Many ideas significant for the development of the evolutionary theory were introduced by Schumpeter. Schumpeter was early to recognize that bounded rationality is necessary for the theory of innovation and dynamics (Schumpeter, 1934). In what was first intended to be entitled a “Neo-Schumpeterian theory of the firm” (Winter, 1968), Nelson and Winter (1982) integrated insights from Schumpeter with ideas of Alchain (1950), Hayek (1945), and Cyert and March (1963). The firm in this view is seen as a profit-seeking entity whose primary activities are to build and exploit valuable knowledge assets. Firms come with “routines” and “competences”, which are recurrent patterns of action and may be changed through search and learning. However, the rationalization of the industry as a static system as suggested by the evolutionary theory limits the possibilities of learning and change from the industry to the level of the firm. (e) Knowledge-Based Theory The knowledge-based theory initially arose at the interface of innovation and knowledge management research. It focuses on the development of a firm and considers it as a repository of distinct technological and organizational knowledge, which facilitate the processes of learning and growth (Dosi and Marengo, 1992; Nonaka, 1994). The principal role of such resources for the competitive advantage of the firm is distinguished and further explored by a number of conceptual studies (Conner, 1991; Kogut and Zander, 1992; Nonaka, 1994; Leonard-Barton, 1995; Foss, 1996). One of the contributors to this approach is Leonard-Barton, who grounds a discussion in the literature on “core competence” and focuses on “the whole system of knowledge management”, which is seen as an integral element of the competitive advantage, or “core technological capability” (Leonard-Barton, 1995). The proponents of this theory generally emphasize the openness of the firm to the acquisition of external knowledge (Leonard-Barton, 1995; Nonaka, 1994).7 Another important concept of the knowledge-based theory is referred to as “combinative capabilities” discussed by Kogut and Zander (1992). Here the emphasis is put on the firm’s ability to handle change by transforming old capabilities into new ones. Two points regarding the nature of this transformation are emphasized. First, firms produce new capabilities by recombining existing capabilities and other knowledge. Second, the ability of a firm to do this is affected by the organizing principles guiding its operations. These principles include matters of formal structure but, more importantly, they also account for the internal social relations, which are shaped in part by differences in the knowledge bases of individuals and groups within a firm (Dosi et al., 2000). The theoretical and empirical studies of this approach argue that knowledge-based theory explanations are most applicable to situations of uncertainty, where capabilities are divergent and communication is impossible or very costly (Langlois, 1992; Foss, 1996). However, knowledge-based theory does not expand this argument and provides limited explanation of the interactions between a firm and the market. (f) Dynamic Capabilities Theory Each of the theories reviewed in this section provides a distinct contribution to the dynamic capabilities theory. If the resource-based theory emphasizes the growth of the firm achieved through its resources and managerial learning, the innovation studies provide the link between the size of a firm and its innovativeness. The behavioral theory accentuates the processes of intra-firm coordination and adaptation, which are also elaborated in evolutionary theory, focusing on change at the level of the firm. The knowledge-based theory advances the concept of knowledge-intensive assets and specifies that such assets are particularly vulnerable in environments with a high degree of uncertainty. In combination, these principles comprise the major assumptions underlying the dynamic capabilities theory. The dynamic capability perspective follows Hayek (1945), the behavioral, and the evolutionary theorists in emphasizing that coordination as an economic problem only occurs because of change. The change factor imposes certain characteristics on the managerial behavior which represent a mix of Schumpeterian (the manager introduces novelty and seeks new combinations) and evolutionary (the manager endeavors to promote and shape learning) principles. Taking into account evidence from the resource-based and knowledge-based theories, the role of internal resources of a firm is essential for creating a competitive advantage. However, even unique, valuable, immobile resources (Barney, 1991) could ensure only one type of rents, the Ricardian rents. Such rents, also defined as quasi rents, could be competed away over time. This happens because resources alone, no matter how critical for the current competitive advantage, could not sustain such an advantage over time as this will need additional capability to adjust them to the changing environment. It is the capabilities to recombine resources in a more efficient way which provide an innovative potential and allow for alternative Schumpeterian rents. The Schumpeterian rents, in their turn, can be sustained only if capabilities underlying them are skillfully maintained over time.^ Classification of Firm’s Capabilities Although a number of authors have argued that there is no consensus on the definition of capabilities (Foss, 2003; Baldwin and Clark, 1994; Henderson and Cockburn, 1994), existing concepts could be classified into three categories (Winter, 2003). While the first category of (a) zero-level capabilities is limited to the management of existing corporate routines, the second category of (b) first-order capabilities facilitates incremental change in the firm. The third category of (c) second-order and “meta” capabilities initiates a change at the level of a firm. Each of the categories described is subsumed to a larger set of a firm’s capabilities and resources, the majority of which is recognized as sources of durable firm heterogeneity. The three categories of capabilities and an alternative mechanism of (d) ad hoc problem-solving are analyzed in this section. (a) Zero-Level Capabilities: Management of Routines The first category of capabilities can be referred to as those which manage the operations and, thus, the routines. They are well recognizable in situations where the firm remains in a relatively stable position as the market experiences no rapid or continuous unpredictable change. These capabilities enable the firm to keep manufacturing and selling the same products, on the same scale, and to the same customers over time. The capabilities exercised in such a stationary process determine a firm’s efficiency as it carries on its current business activities. They are defined as productive or zero-level capabilities representing the “how to earn a living now” capabilities (Winter, 2003: 992).(b) First-Order Capabilities: Management of Incremental Change This category of capabilities creates a prototype of dynamic capabilities. There is a consensus in the literature that “dynamic capabilities” contrast with ordinary productive capabilities by being concerned with change (Collis, 1994). The capabilities in this category are, however, limited to facilitation of an incremental change and could well be explained through the development of a new product. In many respects, such capabilities remain highly patterned and “routine”. They facilitate change in the existing routines and productive capabilities, and because of this the resource base of the firm is gradually improved (Eisenhardt and Martin, 2000).(c) Second-Order and “Meta” Capabilities: Management of Firm-Level Change The capabilities from the third category could facilitate more radical change than simply the development of a new product. These capabilities are primarily targeted at creating new capabilities. One of the most cited definitions of such capabilities is, “the ability of an organization to learn, adapt, change and renew over time” (Teece et al., 1994: 20). A common theme of dynamic improvement to the activities of a firm appears in all definitions of this category (Collis, 1994). Many of them also share appreciation of the role of an entrepreneurial function. It is observed that in order to be efficient, these dynamic capabilities have to be constantly practiced and reshaped (Winter, 2003). Further elaborations on this observation emphasize that there is a need to balance the costs of the capability and the use that is actually made of this capability. As capabilities are vulnerable to a “rustiness” problem, their successful maintenance requires frequent exercise and, therefore, is costly. Obviously, dynamic capabilities can not be substituted by ad hoc problem solving. If the firm does not invest in making the response to exogenous change a routine, or, if it only invests in making familiar types of change a routine, it could be disadvantaged relative to more flexible players investing in higher-order capabilities. Collis (1994) makes an interesting related suggestion that there is a historical tendency for the locus of competitive action to rise in the capabilities hierarchy: “Valuable capability would concern the rate at which the organizational structures that produce rapid innovations were innovated. But an even better source would be the ability to innovate the structures that produce better product innovation. And so on…” (Collis, 1994: 144). Theoretically, there is no logical end to the search for the capability which is the ultimate source of sustainable competitive advantage. In practice, the cost of developing and maintaining capabilities, as well as external market change, could at any time rebalance internal capabilities in favor of lower-order capabilities supplemented by ad hoc problem solving. (d) Ad Hoc Mechanism The high cost of a continuous maintenance of capabilities raises value of alternative mechanisms of organizational change. For instance, ad hoc problem solving sometime plays the role of a successful substitute for the dynamic capabilities (Winter, 2003). When change occurs due to the unforeseen alterations in the external environment, the adjustment in a firm behavior will not often depend on the dynamic capabilities. Such adjustments are better defined as ad hoc problem-solving, implying behavior that is largely non-repetitive, fairly or intently rational, and merely reactive or passive (Winter, 2003). Therefore, they could not be considered as routines. The ad hoc problem solving and the exercise of dynamic capabilities are two different mechanisms of organizational change. More important, the irregular nature of the former implies much lower costs for the firm exercising it. In contrast, repetitive exercising of dynamic capabilities requires substantially higher expenditures for their maintenance. (e) Joint Analysis of the Capabilities of the Firm Generally, it is difficult to make sharp distinctions among the three categories of capabilities, since they all describe the ability of the firm to perform an activity more efficiently than competitors (which could be static or dynamic, routinized or creative, incremental or full of foresight). The difference between them is quite relative.8 Furthermore, as capabilities tend to change over time, the difference between them will also fluctuate. A number of authors have argued that a capability is able to evolve, “just as products have development paths that follow recognizable patterns, known as product lifecycles, so do capabilities” (Helfat and Peteraf, 2003: 998). And, although, non-dynamic capabilities are changed under the influence of dynamic capabilities, all capabilities have the potential to accommodate change (Raff, 2000). A lower-order capability could gradually evolve into a higher-order capability in an evolutionary way and every “capability may pass through multiple stages of transformation before it faces an ultimate decline” (Helfat and Peteraf, 2003: 998). Similar ideas of capabilities development referring to a larger domain of a firm’s learning were presented by Helfat and Raubitschek (2000) and Zollo and Winter (2002). Zott has expressed them as follows: “The mechanisms that constitute dynamic capability can be conceptualized as routines for variation (including search through imitation and experimentation), selection, and retention, which are ingredients to a system of evolutionary learning” (2003: 120). However, presenting capabilities in a form of a firm’s internal capability lifecycle (Helfat and Peteraf, 2003), or as a part of an evolutionary development of a firm (Zollo and Winter, 2002), any classification of capabilities tends to be subjective and temporal. Despite its remarkable complexity and highly situational nature, the concept of capabilities continues to attract increasing attention of scholars. This is due to the fact that capabilities, particularly higher-order capabilities, are a key to the advancement of the theory of the firm (Augier and Teece, 2004). As Zott explains: “Dynamic capabilities are more than a simple addition to RBV since they manipulate the resources that directly engender rents” (2003: 120).9 In brief, the value of organizational capabilities is context dependent. Capabilities can be valuable, but depending on their application, they are not always the source of “sustainable” competitive advantage. Because of this quality scholars in strategy express some doubts whether it is possible to find the ultimate source of sustainable competitive advantage (Helfat and Peteraf, 2003; Collis, 1994). However, it seems that the confusion surrounding the concept of capability stems from the desire to find too generic formulas for sustaining competitive advantage (Winter, 2003) and, due to this, scholars do not account for all aspects of the capabilities. A closer look at the recent attempts to further deconstruct the capabilities into components (i.e. to make a sharper distinction between productive and dynamic capabilities) indicates where further refinements of this theory are being pursued.^ Distinction between Productive and Dynamic Capabilities The initial definition of selected capabilities of the firm as dynamic was made by a few influential contributors (Teece et al., 1990; Teece and Pisano, 1994; Collis, 1994). In these conceptual papers only the learning routines were identified as contributing to the dynamic capabilities, where other routines, such as productive routines, were separated in a different category. The category of non-dynamic productive routines includes all basic functions of the firm that constitute operations management. These routines are usually well understood and partially codified. The easy diffusion of knowledge in such functions means that they can be equally well positioned inside and outside corporate boundaries. Therefore, under the assumption of a static environment, the decision to let the function be performed by the third party could be adequately supported by a standard contractual framework. The relationship between the various functions within the firm is a different matter. The function of system integration guiding this relationship is still considered to be highly entrepreneurial. The identification and implementation of such new combinations is the crucial essence of dynamic capabilities.10 The dynamic capabilities are distinguished from other capabilities of the firm due, primarily, to their ability to act and produce change in the functions of a firm. Such action is often initiated by the economic change that triggers and substantially enhances the dynamism of corporate coordination. Therefore, the identification of the dynamic capabilities is easier in firms placed in the Schumpeterian environment of rapid technological change.11 The role of dynamic capabilities is to coordinate the change processes, which occurs in parallel inside and outside the firm. The challenge which dynamic capabilities face is to synchronize the corporate change with the market change and to make a firm’s internal functions operate in unison within the adjusted corporate order. The capabilities discussed in the previous sections can be generally considered as either dynamic or productive. While the dynamic capabilities are able to distinguish the firm among competitors and produce Schumpeterian rents (second-order and “meta” capabilities), or possess the potential for further development into second-order capabilities (first-order capabilities), productive capabilities are targeted at maintaining the operational functions of the firm (zero-level capabilities). Related studies argue that the dynamic capabilities have to be continuously synchronized and coordinated inside the firm (Zott, 2003). Following the definition of Schumpeter, such coordination could be described as “new combinations”. These can be completely novel configurations of existing capabilities and assets, or standard configurations of new capabilities and assets, or hybrids involving both. While the conceptualization of resources deployable into multiple product areas was first advanced by Penrose (1959), she did not elaborate on this assumption further and provided no explanation of how the development takes place. However, in firms of significant size this could be a non-trivial task. Particularly, in large hierarchies achieving functional “integration” is a major organizational accomplishment, which demands consistent and very complex managerial efforts. Such “integration” is described by Porter (1994) in a concept of “strategic fit”. Similarly, Miles and Snow highlight the same concept, although their focus is made on the relationship between strategy, structure, and processes. They note that, “fit is both a state and a progress – is best conceptualized as a journey rather than a destination” (Miles and Snow, 1978: 11). In many ways the traditional literature on fit is consistent with dynamic capabilities principles; however, it does not provide the explanation of a continuous organizational adaptation and synchronization essential for the dynamic capabilities. Such a constant change in a firm’s capabilities logically implies that the redeployment process should involve interactions with the external environment. Indeed, the acquisition of resources from external sources is well studied in research that is related to dynamic capabilities. For instance, Henderson and Cockburn (1994) stress that a firm’s ability to integrate knowledge from external resources, the “architectural competence”, is positively associated with research productivity as measured by patent counts. Similarly, Iansiti and Clark (1994) found that “integration capability” is associated with the positive performance of a firm. Notably, the process of asset placement outside a firm’s boundaries has received much less attention in the related research. It seems, that in order to increase its explanation power, the dynamic capabilities theory should offer conceptualizations of both the likely internal demand for various assets, a firm’s ability to integrate desired assets, and its ability to divest undesired assets. Only if a firm possesses the three conceptualizations can dynamic capabilities be further separated from operational management, as the former deal with designing and implementing difficult to imitate “change” routines, achieving efficient asset selection, and performing a “fit” function. In summary, the dynamic capabilities theory provides a coherent framework for studying a firm under conditions of uncertainty, rapid technological change, and ambiguity. Emphasizing a firm’s internal processes, the theory sees the source of a firm’s competitive advantage in terms of its unique capabilities. However, such capabilities enable a sustainable competitive advantage only if they are continuously practiced and purposefully developed. Over the time, as a firm learns and improves its resource and capability base, some of its resources will turn redundant or obsolete. Eventually, new combinations will employ a portion of current resources, but part of them will inevitably turn into ballast. The redundancies accumulated over time are difficult to detect and evaluate using only a firm’s internal perspective. Nevertheless, the dynamic capabilities theory contains no adequate description of a firm-market interface and does not explain how a firm can eliminate its redundant resources. Therefore, in spite of the fact that the concept of how a firm externalizes redundant or non-core resources appears to be equally important with the process of investment decisions (Chandler, 1990; Helfat, 1988), it is ambiguous as far the dynamic capabilities theory is concerned. This imbalance is quite contradictory if one considers that the dynamic capabilities theory does not favor a hierarchical mode of governance and tends to employ organizational forms that are flatter and easily responding to environmental change. Maintaining a flat organizational form seems to be unrealistic if a firm constantly accumulates resources but does not possess an efficient mechanism to eliminate redundant and non-core assets. A further empirical validation of the theory could potentially provide new evidence with respect to the dynamic capabilities development.^ Transaction Cost Theory The transaction cost theory accounts for a longer history and a more established concept than the dynamic capabilities theory. Nevertheless, this theory has strongly been criticized for its logic. Such criticism, among others, addresses the lack of dynamism. This paper, first, discusses the intellectual roots of the transaction cost theory. The analysis of the theory is continued by discussing its major elements and selected principles, which are exposed to a particularly strong criticism and, as a result, are continuously refined. ^ Underlying and Related Theories Over the past 30 years, the predominant theoretical explanation of the choice of the boundaries between the market and hierarchy was developed by a transaction cost theory (Williamson, 1975; 1991; 1996; Madhok, 2002), which to a large extent modified and refined the previously formulated principles (Williamson, 1985; 1998). Interestingly, while the transaction cost theory is associated with the industrial economics stream of research, the major intellectual antecedents of this theory comprise conceptualizations provided by prominent contributors to the theory of the firm (Williamson, 1999). The most influential conceptualization, which predefined the emergence of the transaction cost theory, is the classical work of Ronald Coase, “The nature of the firm” (1937). Coase’s unconventional approach was to define the firm, not as an economic production function mechanism, but as a part of a dichotomy of the firm and market, where the choice between these two alternative forms of governance is dominated by the difference in the transaction cost. In his later articles, Coase (1960; 1992) elaborated on the notion of the zero transaction cost but still emphasized that choices always have to be made between feasible alternatives. This emphasis means that the analysis of the institutional actions has to be made considering existing transaction costs. Similarly, John Commons has stressed the importance of analysis of the economics of the organization and defined a transaction as a unit of such an analysis: “The ultimate unit of activity … must contain in itself the three principles of conflict, mutuality, and order. This unit is a transaction” (1932: 4). Chester Barnard has similarly insisted on the importance of organization and its underestimation in economic theory. In his book, “The functions of the executive” (1938), Barnard argued that cooperative adaptation is a central problem of economic organization. He further described such adaptation as, “conscious, deliberate, purposeful” (ibid: 4), which could be achieved by the administration of a firm. While Barnard’s ideas were fundamental to the development of the organization theory (Simon, 1947; 1957; Cyert and March, 1963), they also influenced the formation of the transaction cost theory (Williamson, 1999). Another contributor to the development of the transaction cost theory is Kenneth Arrow (1962). He suggested that the market failure is a relative category which could be more accurately described by employing the category of transaction cost as they are able to explain the impediment or blocking of market formation. In this way, he has located the concept in economic theory. Alongside with rationalizing the existence of transaction costs, these and other related concepts have proposed that there are multiple sources and aspects of such costs. For instance, Coase (1937, 1988) emphasized the ‘frictional’ costs, such as those of identifying a potential supplier, negotiating, drafting a contract, and monitoring it. Later, Williamson (1975, 1985) transformed the subject by shifting attention to the costs of transactional hazards and of governance arrangements to limit such hazards. His focus is on the tendency of transactions to attract difficulties for reasons associated with bounded rationality and opportunism. In this concept, the frictional costs differ from the transaction costs due to their quality to be present even when things go well.12 Despite the prominent antecedents, the concept of transaction cost remained “vague and elastic” for a long time (Williamson, 1999: 1088). Its excessive degrees of freedom allowed almost any outcome to be rationalized by one of the specifications of the transaction cost (Fischer, 1977). As late as in 1970s, the refinement and full-speed operationalization of the concept has begun. Since then transaction cost theory has gone through informal, preformal, semiformal, and fully formal modes of analysis (Williamson, 1996: 18-20). Further advancement of the transaction cost theory is enhanced by joint analysis with related theories of the new institutional economics. In particular, empirical studies combining principles of transaction cost theory with those of property rights (Grossman and Hart, 1986; Hart, 1990; 1991) and agency theories (Jensen and Meckling, 1976; Harris and Raviv, 1978) have contributed to a better calibration of the instruments of the transaction cost theory. For instance, in his research on contractual governance of R&D projects, Ulset (1996) found that the combined approach of transaction cost theory and property rights theory provides a more precise description of governance tools (e.g. such as control and property rights). A wider range of control mechanisms is made available by the agency theory which offers an in-depth analysis of transaction-based agency relationships between independent market players (Jensen and Meckling, 1976). The costs of relationships, or agency costs, are comprised of expenses for monitoring and bonding mechanisms designed to reduce opportunism and unavoidable conflicts between principals and agents. Employed by the agency theory governance mechanisms include various forms of contracts, ranging in character from formal to informal, explicit to implicit, and objective to subjective (Barney and Ouchi, 1986). The strong focus of this theory on the interaction process between agents well complements a broader conceptualization of the choice of governance analyzed in the transaction cost theory. A brief summary of major contributions to the development of the transaction cost theory are outlined in table 1. Table 1: Principles Contributing to Transaction Cost Theory Relevant Principles Selected Contributors Firm - market dichotomyDifferences in transaction costs “Frictional” costs Ronald Coase (1937) Importance of economics of organization Transaction as a unit of activity John Commons (1932) Economic theory underestimates organizationFunctions of executive Cooperative adaptation Chester Barnard (1938) Market failure Transaction costs explain impediments and blockings of market formation Kenneth Arrow (1967) Transaction cost stems from difficulty to measure and monitor performance Alchian and Demsetz (1972) Transaction cost stems from inability to specify goods and services needed Jacobides and Croson (2001) Relevant Principles Selected Contributors Property Rights Theory Control and property rights as governance tools Grossman and Hart (1986); Hart (1990); (1991) Agency Theory Control mechanisms Various forms of contracts Jensen and Meckling (1976); Harris and Raviv (1978) Source: Own analysis^ Conceptualization of Transaction Cost In this section the key elements of the transaction cost theory are analyzed and their interpretation of the specific assets, the issue central for the present study, is discussed. First, (a) the general logic of the theory regarding the governance of specific assets is outlined. Then (b) the dual concept of commodity goods (for which the price mechanism is efficient) and specific assets (for which the price mechanism is considered to be inappropriate) is presented. While specific assets are prescribed to be kept inside a firm, the theory does not specify the corporate mechanism for dealing with such assets and limits (c) the role of managers to passive decisions of allocating existing assets. Finally, the advancements of the transaction cost theory indicate that despite the role of the firm as (d) a value protecting mechanism, (e) the hybrid organizational forms could compromise the duality of the market vs. hierarchy mechanisms and exemplify an efficient mode for dealing with specific assets. Furthermore, the static nature of the transaction cost decisions is questioned with (f) a hypothesis of dynamic transaction costs. (a) Governance of Specific Assets In his work, Coase (1932) distinguished between governance in a hierarchy and market organizational modes. The choice between them is made so as to minimize the transaction cost which defines all organizational modes ranging from the market to internal coordination in the firm. In the latter, transaction costs form the boundaries of the firm, so that at the margin, the internal costs of organizing are equilibrated with the costs associated with transacting in the market. The boundary choices are driven largely by the specificity of assets involved in the exchange (Shelanski and Klein, 1995), which may trigger a threat of opportunistic behavior of companies involved in the exchange and require costly contractual safeguards, use of property rights, and sufficient incentives to deter participants from opportunism. According to the transaction cost theory, a vertical integration in the presence of exchange-specific assets, is a preferred governance solution (Hennart, 1991; Osborn and Baughn, 1990). It is accepted that it is useful to think of a firm and market as alternative modes of governance. Therefore, whether to use an internal mode of governance or a market mode of governance is dependent on the tradability of the assets in question. (b) Price Mechanism and the Role of the Firm The price mechanism means that a large portion of allocation of resources can take place via the market in a quick and efficient way.13 This is true for commodity markets, but a large portion of goods and most services are difficult to exchange in open, organized, and well-developed markets. Mutually beneficial transactions frequently do not happen if the object of the transaction is associated with property rights that are poorly defined, contain assets that are difficult to transfer, and have a value that is hard to measure. The bounded rationality of involved actors and the incomplete information they possess contribute to an inadequacy of a market valuation of specific assets. Therefore, when specific assets are needed to support efficient production, the preferred organizational mode is a hierarchy. For the situations of thin or non-existent markets, transaction cost theory clearly favors allocation of assets inside the firm. Furthermore, the allocation of resources inside a firm not only substitutes but also complements the allocation of resources by markets as the latter possess no efficient mechanisms for trading thin assets. While transaction cost theory has been used in empirical analysis of production and supply activities (Monteverde and Teece, 1982; Stuckey, 1983; Masten, 1984; Walker and Weber, 1984; Balakrishnan and Wernerfelt, 1986), as well as in marketing (Anderson and Schmittlein, 1984), its application to the research of specific assets (such as R&D) in the definition of a firm’s boundaries has been limited (Pisano, 1990). The rationale behind this lack of studies on specific assets is the accepted assumption that an internal organization possesses an ability to defeat opportunism and make specific assets work for the purposes of a firm. However, this assumption does not explain why it occurs. This treatment excludes any insights on process design, internal resource allocation, and asset alignment, as a set of factors explaining the performance of a firm. In other words, it leaves an explanation of how specific assets are handled beyond the scope of traditional transaction cost theory. The theory is, however, very efficient at explaining the treatment of commodity goods and low-specificity assets. It indicates that the market enables rapid adaptation with respect to assets which are actively traded using a price mechanism. If activities are done outside a firm, the emphasis is put on designing contracts that are efficient, in the sense that “hold up” or re-contracting costs are minimized or avoided. Therefore, the main emphasis in the transaction cost theory is put not on the selection of activities that are needed to be done or on the explanation of how these activities have to be done, but rather on the determination of where they are to be done. The mechanism provided by the transaction cost theory solely serves to avoid the negative consequences of opportunism in the case of a transaction. If the activity is done outside the firm, the emphasis is put on a contractual mechanism that helps to avoid “hold up” problems. (c) Role of Management and Concept of “Fit” The transaction cost theory leaves the role of management somewhat indistinctive. It limits managerial functions to making the choice of governance mode between market arrangement and internal organization. The managers do not necessarily have to understand what drives or should drive asset selection, configuration, or investment decisions. The theory conceptualizes managers only as cautious executives of the asset allocation decisions, and in this description the term “managerial discretion” emphasizes the passive role of a manager. However, in later refinements, the transaction cost theory revises this view and makes a claim that the role of management is “significant” in transaction cost theory (Williamson, 1999: 1001). Nevertheless, this only means that the role of a manager is limited to exercising organizational “fit” in a form of adaptive process of organization and reorganization of a firm’s existing resources. (d) Firm as Value Protection Mechanism The transaction cost theory focuses on the economizing on transaction costs with the emphasis put on how this mechanism is organized via the governance framework. The governance framework explains asset and value protection, but leaves unexplained the process of value creation. Therefore, the transaction cost theory reflects the static situation of choices between internalization and externalization. This is supported by the assumption that non-traded and thinly traded assets have to be organized internally. The transaction cost theory implies that both an evaluation and a choice of governance structure are situational decisions with a short-term orientation at best. Furthermore, though there is much utility and exploratory power in the transaction cost framework, the contractual scheme upon which it is built deals with existing resources and does not examine how new resources are discovered, how they are accumulated, and how a firm learns. Therefore, the structure and behavior of a modern firm cannot be fully explained by simply appealing to transaction costs. This is an important element of management, but it is not the main concern of management scholars or practitioners. (e) Role of Hybrid Organizational Forms Any advancement of the traditional transaction cost logic to justify the hybrid governance form of specific assets (such as R&D) is constrained by the previous assumptions of this theory. For instance, the transaction cost theory is quite pervasive at explaining why there should be no outsourcing of R&D. The reasoning is that the high degrees of complexity and uncertainty associated with such a transaction make contractual relationships difficult or impossible to negotiate. Under conditions of bounded rationality imposed by the complexity of the transaction, the bargainers are unable to define complete contractual agreements because neither all alternative features nor their corresponding prices can be determined (Williamson, 1975). Similarly, Ouchi and Bolton note in relation to R&D projects: “When the task in hand is complex or uncertain, contractual agreements will be difficult to specify and monitor, and post-contractual opportunism will be common” (1988: 12). Because of these reasons, the transaction-cost model rationalizes that R&D is impossible to trade in the market-mediated system, and it rather should be carried out within the firm. This assumption could be relaxed by the fact that not all R&D has high degrees of uncertainty and complexity attached to it (Freeman and Soete, 1997). There are some kinds of R&D that have low levels of complexity (i.e. minor technological improvements and product differentiation) where market-mediated contracts can theoretically function.14 On the basis of this advancement of the transaction cost theory, several attempts were made to use the theory for explanation of a shift from the in-house provision of research toward a more market mediated approach. However, although useful as a general guide for such changes, transaction cost theory is difficult to apply for specific organizational contexts or more aggregated empirical surveys (Foss, 1996; Englander, 1988). This happened due to what Robins has termed “serious logical and empirical flows” (1987: 68) of the theory. However, there was also positive evidence of operationalization such as Brockhoff’s (1992) application of the transaction cost logic for the analysis of the R&D cooperation. The recent developments in terms of the transaction cost theory further elaborate on hybrid forms of governance and explain them as able to cope with greater complexity in contractual and organizational forms, than the traditional market and hierarchical organizations (Williamson, 1996). (f) Dynamic Transaction Cost The proponents of the dynamic approach to the transaction cost theory argue that although the context in which the transaction cost decisions are made is fixed in the short term, it is subject to change in the medium term as firms make incremental adaptations and explore new directions in their search for profit. The slow evolving factors of dominating contracting norms, firm reputations, and transactions technology, as well as the on-going process of learning to contract determine the environment and its change (Mayer and Argyres, 2004). The environment is further altered by the change in the existing transactional “interfaces” that can support market exchange to a greater or lesser extent (Baldwin and Clark, 2003). Therefore, in the intermediate and longer terms a firm could influence the shape of the transaction cost context and the way the transactions are made evolves (Jacobides and Winter, 2005). In summary, the recent efforts to extend the transaction cost theory, also including the explanation of placement specific assets outside the boundaries of the firm, are still in the process of development. Meanwhile, the transaction cost theory has the reputation of being a theory which best explains the mechanisms for allocating non-specific assets. The role of management in the transaction cost theory is somewhat limited. The very nature of managerial activity expressed in the coordination and adaptation of non-traded or thinly traded assets (Cyert and March, 1963) is not elaborated in this theory. Because the question usually asked by managers in the transaction cost framework is “where” to allocate resources, the issues of “how” to exploit resources and, in particular, how to treat them under the condition of hierarchy, are not developed in this concept. While the transaction cost theory adequately explains the managerial role in arrangement of market contracts, it does not provide an equivalent explanation of any managerial role for the internal coordination of resources. This contradicts the concepts of the strategic management literature, which suggest that asset selection, development, and configuration, as well as further investment decisions, comprise the essence of managerial activities in the creation of strategic “fit” of a firm (Teece et al., 1997). Beside that, the static decision of boundary choice does not contain space for an option to learn and there is no managerial creativity left in the transaction cost theory. Due to these conceptual weaknesses the opportunistic basis of the transaction cost theory has come up against wide-spread criticism. An accepted criticism is that the link between asset specificity and boundary choice has little to do with opportunistic behavior and failed markets (Foss and Foss, 2004a). The counter argument is that the increase in the specificity of activities does not trigger market failure but enhances the efficiency with which such activities are coordinated (Grant, 1996; Kogut and Zander, 1992; Coff, 2003). The predominant concern of transaction cost theory with allocation decisions does not provide space for taking into account the value of creation. For instance, the focus of the transaction cost theory on finding the remedy from opportunism underemphasizes the fact that there could be a value in the co-specialization. Similarly, the fear of opportunism also extinguishes numerous opportunities available on the market. Other theories show that decisions regarding asset selection are not just driven by transaction costs and governance considerations but by a degree of value that has yet to be discovered (Langlois and Foss, 1999; Conner, 1991). The meaning of the assertion “price is what you pay, value is what you get” has yet to be recognized by the transaction cost theory.^ Comparative Analysis of Dynamic Capabilities and Transaction Cost Theories The significant variance of dynamic capabilities and transaction cost theories in the interpretation of the role of the firm can serve as a basis for their joint interpretation. Here I present such an attempt and elaborate on the criticism of both theories in more detail by analyzing existing studies which combine the two approaches and identifying conceptual gaps that still exist in each of them.^ Differences and Complementarities of Two Theories The transaction cost theory rationalizes the firm as a solution for problems associated with transactions involving specific assets. Thinly and non-traded assets that are valuable and scarce could trigger opportunistic behavior that is why the theory prescribes their integration inside the firm. Advocating the benefits of a hierarchical form of corporate governance, the transaction cost theory focuses on the analysis of make vs. buy decisions that define the internal assets of the firm. This analysis concentrates on the market interface providing a spectrum of available alternatives influencing a firm’s transaction decisions. The make vs. buy decision is considered to be efficient if there is no equally good or better alternative available on the market. The efficiency level of individual decisions accumulated over time, according to this theory, forms the efficiency level of the entire firm. However, considering that the character of each decision is highly situational and depends on particular circumstances under which this decision is made, the principles underlying it will be short-term. The perception of a firm through its accumulated transactions (make vs. buy decisions), which are featured as highly situational, implies that the firm is itself a short-term oriented market player. In contrast, the dynamic capabilities theory rationalizes the firm as a bundle of capabilities that differentiates it from competitors. Neither efficient transactions nor firm-specific resources distinguish the firm on a continuous basis. It is only capabilities nurtured in the firm that are able to create a bottleneck in the industry value chain and ensure sustainable competitive advantage. However, such a bottleneck will not last long unless the firm continually learns and, on the basis of the newly gained knowledge improves its capabilities. Therefore, the efficiency of a firm could be mainly achieved through its constant learning and innovation processes. Obviously, this approach implies a long-term perspective on a corporate development with current decisions on capability development pre-conditioning future results, and to a large extent, determining whether the firm sustains its competitive advantage. A problematic characteristic of the dynamic capabilities theory, which becomes more vivid if the theory is compared with the transaction cost theory, is that it does not directly elaborate on the existence of a firm-market interface. It not only broadly neglects transaction costs, but also leaves unexplained how non-core, redundant or irrelevant corporate resources are handled. In other words, if a firm is a learning organization it creates new resources and capabilities and combines them in new ways, so that some resources will inevitably become obsolete or redundant. This implies that a firm, in order to stay flexible, will need to eliminate its redundant resources and capabilities. The proponents of the dynamic capabilities theory accept this argument and confirm that the theory favors disintegrated organizational structures (Winter, 2003). However, dynamic capabilities theory neither discusses the process supporting disintegration of irrelevant resources nor provides any insights on a firm’s capability to maintain and improve such a process. Transaction cost theory, on the contrary, offers a profound analysis of the firm-market interface. However, due to its focus on the efficiency of individual transactions, the short-term orientation of this theory (defined in some works as static), (Foss and Foss, 2004a) does not provide space for the rationalization of capabilities. Therefore, the transaction cost theory does not account for the existence of capabilities in a firm (Williamson, 1999). It seems that these criticisms could be partially resolved through a combination of the principles of dynamic capabilities and transaction cost theories (Table 2). Table 2: Comparative Analysis of Two Underlying Theories Dynamic Capabilities Theory Transaction Cost Theory Unit of analysis firm specific capability, characterized by astute initial asset selection and their continuous reconfiguration transaction, characterized by its frequency, uncertainty, and involvement of specific assets Human actors bounded rational capability to learnfavor trustsee benefits in fair relationshipsable to foresight this ability is uniformly distributedpossess intra- and entrepreneurial skills which are also uniformly distributed bounded rational and, therefore, produce incomplete contractsopportunistic guided by self-interestable to foresight rather than being myopic Explanation of the firm firm i