ساختار بازار دوگانه و احتمال روابط تکرارشده - شواهد از بیوتکنولوژی دارویی

This paper analyses the role of repeated ties in the high-tech pharmaceutical biotechnology industry, a sector that is characterized by a strong dual market structure. Our most important finding is that previous ties in pairs of large pharmaceutical companies and small biotechnology firms have a negative effect on their subsequent partnering. An explanation for this result is found in the context of understanding the specifics of large-small coalitions in a high-tech dual market structure. Unlike what is known about repeated ties in many other industries, this high-tech dual market structure indicates that R&D partnerships between a small number of very large companies and a large group of dependent, small firms are not characterized by mutual dependence, similarity, or equality.

Repeated ties in pairs of cooperating companies play an important role in the current understanding of inter-firm partnership formation (see amongst others, Chung et al., 2000, Ciborra, 1991, Dyer and Singh, 1998, Gulati, 1995, Hagedoorn, 1993, Lundvall, 1993, Lyles, 1994 and Mowery et al., 1998). This research indicates that prior partnerships in specific pairs of companies can influence the likelihood that these pairs will continue to form partnerships. In order to contribute to this body of literature, our study will attempt to deepen the understanding of repeated ties and partnership formation by considering the industrial context of pairs of companies in a high-tech setting. More specifically, this paper analyzes the role of repeated ties in a high-tech industry with a strong dual market structure, i.e. the international, pharmaceutical biotechnology industry. Such a dual market structure is largely determined by on the one hand a group of large, integrated, international, and established companies and on the other hand a group of relatively small, specialized firms. The high-tech dual market structure in the pharmaceutical biotechnology industry is apparent in the role played by a small group of very large pharmaceutical companies and a large group of relatively small biotechnology firms (Powell, 1996, Powell et al., 2005 and Saviotti, 1998). Previous research has already established that inter-firm partnering in the pharmaceutical biotechnology industry is rather specific in terms of the concentration of inter-firm R&D partnering within these two groups of companies (Kenney, 1986, Powell, 1996 and Rothaermel, 2000). To the best of our knowledge, the current literature on the effect of previous ties on subsequent inter-firm partnering has paid no attention to the specific setting of a dual market structure in a high-tech industry. Early, seminal contributions to the analysis of dual market structures, however, did reveal the impact of such market structures on innovative performance, productivity, earnings, financial issues, labor relations, and marketing practices (see Averitt, 1968, Bowring, 1986 and Sutton, 1992, 1998). It is the objective of this paper to investigate whether in this sort of industry the effect of previous ties on subsequent inter-firm partnering might also be different from what is known from many studies on other industries with more evenly distributed populations of companies. In the following section we discuss our hypothesis, derived from current theory, that stipulates the expected effect of previous ties on subsequent partnering. In order to test our hypothesis, we construct a panel dataset that contains information about the partnering behavior of pairs of companies. In the methodological section we discuss the data, sample, and the specific variables we include in our panel logit models. This is followed by sections that present and discuss the results and the implications of our research.

This paper studies the particular relation between previous ties in specific pairs of cooperating companies and the likelihood that these pairs continue to build on their collaborative relationship through a range of repeated ties. The current, well-established literature on the formation of new partnerships pays extensive attention to the particular circumstances under which pairs of cooperating companies are likely to continue their partnerships through repeated ties (see Chung et al., 2000, Ciborra, 1991, Dyer and Singh, 1998, Gulati, 1995, Hagedoorn, 1993, Lundvall, 1993, Lyles, 1994, Mowery et al., 1998 and Mytelka, 1991). Specifically, many of these important contributions have argued that mutual dependence, similarity, and equality in pairs of companies are important prerequisites without which a long-term repeated collaborative relationship between these companies is not likely to occur. Conform our expectations the current research clearly shows that prior R&D partnerships between pairs of large pharmaceutical companies and small biotechnology firms have a negative effect on the probability of subsequent partnerships of these pairs. This result is in congruence with the findings of Arora et al. (2001) who argue that specific industry structures facilitate the development of markets for technology (see also Audretsch, 2003). In such technology markets, inter-firm relationships based on quasi-integration, such as repeated ties and equity-based joint ventures, are not likely to occur. An interpretation of the negative role for previous ties in new R&D partnership formation can thus be found in the context of understanding the specifics of large-small coalitions in a high-tech dual market structure such as found in the international, pharmaceutical biotechnology industry (see also Averitt, 1968 and Sutton, 1998). As was previously mentioned, the vast majority of all inter-firm R&D partnerships in the high-tech pharmaceutical biotechnology industry are formed between groups of large pharmaceutical companies and small biotechnology firms (see amongst others, De Rond, 2003, Hagedoorn and Roijakkers, 2002, Powell, 1998 and Rothaermel, 2000). In our sample more than 70% of these pairs are large-small coalitions. These large pharmaceutical companies typically provide their numerous, smaller R&D partners with financial support and regulatory know-how, in return for which large companies acquire access to the research skills of these small biotechnology firms (Arora and Gambardella, 1990, Barley et al., 1992, Pisano, 1991, Powell, 1996, Powell et al., 2005 and Shan et al., 1994). At this point, it is also important to note that, in this particular high-tech industry, large pharmaceutical companies usually have a wide array of changing R&D partnerships, whereas most small biotechnology firms have a limited number of partnerships. In our sample, the average large pharmaceutical company had more than 11 partnerships during the period 1991–1998; the average small biotechnology firm had fewer than four of these partnerships. The actual number of R&D partnerships of small firms is even substantially lower, due to the necessarily strict selection criteria for our panel data sample. As discussed in the above, we had to exclude small biotechnology firms entering the industry later during the period of analysis and small firms that went bankrupt or that were taken over. These small firms had even fewer R&D partnerships than the small firms that were actually included in the sample. So, small biotechnology firms are often highly dependent on their limited number of financial ties to large pharmaceutical companies3 through which they can ensure their very survival. Large companies frequently enter into numerous R&D partnerships with small firms in order to learn from many different sources of technology and to stay current with respect to the latest scientific and technological developments (Arora and Gambardella, 1990, Hagedoorn, 1996, Kenney, 1986, Powell et al., 1996, Powell et al., 2005 and Saviotti, 1998). Once the large pharmaceutical company has absorbed critical knowledge from an individual, smaller research partner, there is no need to continue its partnership with the smaller biotechnological firm through repeated ties. As the smaller firm is concentrating its research efforts in a specific therapeutic area (Averitt, 1968 and Bowring, 1986), it may take years before this firm has managed to switch to a different field. Alternatively, it may take years before its novel research in the current area has reached a phase where it is once again commercially viable for the large company to enter into a new R&D partnership with its former partner.4 The clear lack of incentives to enter into long-term repeated collaborative relationships for groups of large pharmaceutical companies and small biotechnology firms can thus be very well explained within the particular context of a high-tech dual market structure. In this particular market structure, research partners are very dissimilar with respect to their interdependence, size, capabilities, and both groups of companies play a completely different role in R&D partnerships. A large part of the existing alliance literature, discussed in the above, suggests that previous ties in pairs of cooperating companies are an important stimulus for subsequent partnership formation. Although the current contribution does not discard the accepted understanding of inter-firm partnering and the role of repeated ties, it does point at the critical importance of taking the specific industrial context into consideration when studying repeated ties. Particularly, an explanation of our findings within the context of the high-tech pharmaceutical biotechnology industry could be exemplary for an understanding of the importance of repeated ties in high-tech industries. Therefore, it seems interesting for future work to focus on the effect of previous ties on inter-firm partnering within the context of other high-tech industries with diverging dual market structures and specific groups of cooperating companies. It also seems important for future studies of the pharmaceutical biotechnology industry to examine repeated ties between large pharmaceutical companies and small biotechnology firms in light of the increasing availability of venture capital in recent years.