|
Article Excerpt
1. Introduction
Organizations innovate, in part, through combinations of existing and new knowledge (Kogut and Zander 1992). They must, therefore, often turn to external sources to gain new ideas, insights, and expertise. This ability to acquire this knowledge from external entities is, however, limited by an organization's own experience and expertise (Nelson and Winter 1982). Scholars have come to agree that a firm's search for new knowledge is technologically and geographically bounded (Jaffe et al. 1993, Stuart and Podolny 1996).
Recent research emphasizes the importance of knowledge recombinations across technologies (Rosenkopf and Nerkar 2001, Fleming 2001) and geographic regions (Song et al. 2001). But if knowledge flows between technologically and geographically proximate firms, how can companies overcome localization, and reach out for distant and, perhaps, unique knowledge? This paper suggests that two mechanisms--the mobility of inventors and the formation of strategic alliances--can enable firms to overcome geographic and technological constraints.
We explore how a cohort of semiconductor firms uses these mechanisms to draw on the knowledge stocks of other firms. The study employs patent citation data to observe this phenomenon. We pay particular attention to whether contextual factors (geographic location and technological expertise) influence the efficacy of the mechanisms. Thus, a key contribution of this paper rests in its ability to explore both proximal and distant influences on interfirm knowledge flows systematically and simultaneously.
2. Contexts, Mechanisms, and Knowledge Flows
Studies in evolutionary economics highlight the importance of path dependence in the innovative process (Nelson and Winter 1982, Dosi 1988). These studies suggest that the results of past searches for knowledge become the natural starting points for new searches, as firms rely on their own experience and established knowledge bases to determine what is important and useful. Research in organizational learning (March and Simon 1958, Cyert and March 1963) also makes a similar point regarding the search for new knowledge. This literature suggests that boundedly rational decision makers rely on established organizational practices to drive the search for knowledge. Routines or "socially constructed programs of action" are relatively stable and greatly influenced by the experience and history of the firm and the individuals therein (Nelson and Winter 1982, Baum et al. 2000). Firms, thus, recognize and absorb external knowledge close to their existing knowledge base (Cohen and Levinthal 1990). Hence, even as firms seek to expand their knowledge stocks by looking externally, the resultant search processes are restricted to familiar and proximate areas. Thus, the search for new knowledge is often restricted to a firm's technological or geographic context.
2.1. Contextual Localization of Knowledge
We use the term context to refer to technological or geographic landscape in which a firm operates. Technological and geographical contexts derive from each firm's accumulated decisions regarding technological and locational choices.
Technological Similarity. Studies of innovation and technological development suggest that the tendency toward technologically local search is pervasive. For instance, Helfat's (1994) study of petroleum firms shows that the allocation of R&D spending among various lines of technology varies little across time. Martin and Mitchell (1998) show that new product introductions are heavily influenced by the designs of existing products. While technologically local search is attributed primarily to internal organizational constraints, a variety of interfirm relational mechanisms reinforce this tendency. For example, firms and their employees participate in cooperative technical organizations such as technical committees (Rosenkopf and Tushman 1998, Rosenkopf et al. 2001), jointly author technical papers (Liebeskind et al. 1996), and form alliances (Mowery et al. 1996). Thus, they interact most frequently with other firms and individuals with similar technological expertise. Social networks emerge between professiona ls with common technological interests (von Hippel 1987). and these external relationships reinforce the internal organizational emphasis toward local search.
Empirical studies of patent data suggest a relationship between technological similarity and knowledge flows without testing directly for this relationship. For example, Stuart and Podolny (1996) construct a technological landscape using common patent citations, and show that every major Japanese semiconductor firm, save one, maintained a similar position on this landscape across a 10-year period. This implies that firms continue to draw upon the knowledge stocks of firms most technologically similar to them.
HYPOTHESIS 1. Technological similarity increases the likelihood that a focal firm will draw upon the knowledge stock of another firm.
Geographic Similarity. Other studies of innovation and technology diffusion point to the geographic localization of knowledge. Jaffe et al. (1993) analyzed patent citation data to demonstrate that firms and universities acquire knowledge from others in geographically proximate locations. A key reason for geographically localized knowledge flows, research suggests, is the establishment of interfirm linkages between firms in the region (Saxenian 1990). These relational linkages may be formalized, such as alliances and supply relationships (von Hippel 1988) or informal, such as regional social networks (Rogers and Larsen 1984) and mobility of engineers (Almeida and Kogut 1999). Case studies document extensive information flows through regional clusters in Italy (Piore and Sabel 1984), Germany (Herrigel 1993), and Silicon Valley (Saxenian 1990). These studies suggest that geographic proximity reduces the cost and increases the frequency of personal contacts that build social relations between players in a network , thereby facilitating the flow of knowledge. Firms exploit these regional relationships to access knowledge from other local firms. Organizations reinforce the tendency for geographically local search through the establishment of boundary spanners and gatekeepers with local experience (Allen 1983), as well as by hiring regional experts from neighboring firms (Almeida and Kogut 1997). Thus, as in the case of technology, the underlying reason for geographic local search is both organizational and relational in nature.
HYPOTHESIS 2. Geographic proximity increases the likelihood that a focal firm will draw upon the knowledge stock of another firm.
Given the tendency for localized search, the organizational characteristics and external relationships are self-reinforcing and hard to change. Not surprisingly, Sorenson and Stuart's (2000) analysis of semiconductor patent citations suggests greater localization of the knowledge search process across time. These studies suggest a second point--given the localness of search, a firm's technological area of expertise or technological context may be relatively stable and difficult to change. Similarly, geographic context is relatively stable, because it is not mere location that matters to interfirm learning, but the more challenging process of getting embedded through the establishment of relationships in regional networks (Saxenian 1994). Thus, firms to a large extent are bound to and limited by the technological and geographic contexts in which they find themselves.
Of course, local, contextually bounded search has some advantages. It restricts the breadth and, therefore, the cost of the search process. Geographically and technologically proximate search also results in the acquisition of knowledge that can be more easily recognized and managed by the organization's existing routines and members. Despite these advantages, recent studies in the area of strategic management suggest that given technological change and the dynamic nature of competition, firms must move beyond local search to successfully compete across time. For example, Porter (1990) points to the emergence of geographically dispersed but specialized regions, in various technologies and industries, emphasizing the need for geographically distant search. Kim and Kogut (1996) show that the dynamic of competition has encouraged semiconductor firms to diversify across technological subfields to maintain their competitive edge. Rosenkopf and Nerkar (2001) demonstrate that external exploration in distant technolo gical domains yields innovations with more impact on a broader set of technological areas. Studies like these support March's (1991) suggestion that firms balance local search (exploitation) with more distant search (exploration).
2.2. Mechanisms Facilitating Knowledge Flows
How do firms, embedded in relatively stable geographic and technological contexts, achieve the balance suggested above? We suggest that alliances and mobility are two distinct mechanisms that firms may employ to access contextually distant knowledge. To begin, we develop hypotheses regarding the two interfirm mechanisms independently from the context in which firms are embedded.
Alliances. A central idea in the literature on alliances is that they are useful mechanisms for knowledge acquisition and learning (Hamel et al. 1989). Empirical studies of affiance formation assume that search for new capabilities and strategic interdependencies within limited social contexts drive partner selection (e.g., Gulati 1995, Eisenhardt and Schoonhoven 1996). Powell et al. (1996) postulate the existence of "networks of learning," and suggest that participation in networks of R&D alliances facilitates the growth of new biotechnology firms, because these networks create access to knowledge.
In-depth case studies provide us with a rich illustration of learning between alliance or network partners at the expense of demonstrating overall knowledge flows across networks. Doz (1996) explores how alliances may be construed as learning processes, where learning occurs in multiple dimensions--environment, task, process, skill, and partner goals--and the amount of learning is facilitated or constrained by initial conditions. Dyer (1997) suggests that the breadth and intensity of the relationship between affiance partners will grow across time. Though there is considerable literature relating learning and alliances, few studies explicitly measure interfirm knowledge flows associated with alliances.
Once again, studies rooted in patent data have begun to explore this issue. In Stuart and Podolny's (1996) study of the major Japanese semiconductor firms, the authors suggest that Matsushita accomplished a technological transition through the strategic use of alliances. In two studies of alliances between multinational firms in 1985-1986, spanning a variety of industries, Mowery et al. (1996, 1998) demonstrate that certain alliances are followed by rises in the cross-citation and common-citation patterns between the firms, suggesting some transfer of knowledge.
HYPOTHESIS 3....
|
