ارتباط نامتجانس بین اتحاد سرمایه اجتماعی و عملکرد شرکت دارویی جهانی

Some literature suggests that three dimensions of social capital—such as information volume, diversity and richness—have positive association with the firm's performance. However, these constructs and their implications cannot be taken for granted because they may not produce similar results in all contextual settings. The article develops and tests associations between dimensions of social capital and organizational performance in a multidimensional network system. This study provides an analysis on 252 pharmaceutical firms. The analysis shows disparate but insightful results. First, there is a positive association between information volume and the firm's performance. Second, there is a negative association between information diversity and the firm's performance. Third, there is no significant association between information richness and the firm's performance. Therefore, different dimensions of social capital had different links with the firm's performance. The paper provides some explanation and implications of these findings.

Research on inter-firm strategic alliance and organizational networks has profoundly diffused in management and organizational literature in recent decades (Anderson et al., 1994, Contractor and Lorange, 2002a, Gulati, 1998 and Ritter et al., 2004). Like other technical resources, inter-firm alliance social capital is a strategic resource of the firm that can render a unique value to the firm (Nahapiet and Ghoshal, 1998 and Tsai and Ghoshal, 1998). A value rendering resource of the firm need to be rare, valuable, inimitable and non-substitutable (Barney, 2001 and Wernerfelt, 1984). Alliance social capital reflects these dynamics by providing access to stock and flow of information for functional advantages and strategic opportunities recognition (Podolny, 2001). In this sense, alliance social capital is a resource that can differentiate the firm from its competitors. The difference in the resource can lead to different levels of performance of firms in an industry. Naturally, business organizations are seeking network resources to increase their social capita. There is a fair amount of consensus that there is an increasing trend in the inter-organizational alliance formation and related research (Gulati, 1998). Some literature suggests that there is a positive association between organizational alliance social capital and organizational competence (Ritter, 1999). It implies that the firm's social capital and its performance should be positively correlated. The debate emerges at the relational and structural levels in the organizational network system. Some of the empirical literature on alliance social capital claims that current research social network has some limitations (Koka & Prescott, 2002: 812). Some studies limit their focus to a structural dimension of social capital (Burt, 1992, Powell and Koput, 1996 and Walker et al., 1997). Other studies limit their focus to a relational dimension of the alliance social capital (Anderson et al., 1994, Gulati, 1995 and Wilkinson, 2008). Since the structural and relational dimensions constitute the firm's alliance social capital, both need to be included in an analysis (Rowley, Behrens, & Krackhardt, 2000). Because alliance social capital is a multidimensional phenomenon, the relational and structural dimensions may influence its complexity and implications in business enterprises (Ritter, 1999, Wilkinson, 2008 and Young et al., 2009). The multidimensional perspective suggests the antecedents to alliance social capital can lead to different outcomes (Håkansson & Snehota, 1995). However, multidimensional alliance social capital is also limited in scope. First, the models proposed are not universal in their business implications (Kilduff & Tsai, 2003). Second, qualitative and quantitative results may interact but not reflect each other (Anderson et al., 1994). Third, empirical studies from the steel industry (Koka & Prescott, 2002) and engineering sectors capture a cross-country analysis on outsourcing from developed to developing economies (Young et al., 2009). However, this evidence may not be relevant to the emerging technologies and their contexts. Conventional industries such as the steel industry tend to have a low degree of uncertainty because their technologies are well defined. These sectors depend on primary functions more than secondary functions (Anderson et al., 1994). For instance, the steel industry or mechanical and electrical engineering functions (Ritter, 1999) are less dynamic than the biopharmaceutical industry. The latter kind is highly information intensive (Madhok, 2000). Biopharmaceutical industry needs continuous flow of knowledge from the upstream to the downstream in the value chain (Arora and Gambardella, 1994 and Stuart et al., 2007). Related institutions govern this flow of the technology (Bartholomew, 1997 and Casper and Whitley, 2004). Fourth, the alliance social capital is a complex phenomenon. On one hand, it requires diversity of partners. On the other hand, it requires interdependence between partners (Håkansson & Snehota, 1995). Therefore, the nature of complexity resulting from multiple links merits for a multidimensional alliance social capital be developed and used in the emerging knowledge intensive sector. Some studies have used alliance social capital in international contexts (Arregle et al., 2007, Goerzen and Beamish, 2005, Gulati et al., 2009, Leana and Pil, 2006, Lee, 2007, Oxley and Sampson, 2004, Subramaniam and Youndt, 2005 and Wu, 2008). None has explored its implication for the firm's performance in sectors like the biopharmaceutical industry. More recently, Young et al. (2009) find a correlation between inter-organizational alliance and the manager's perceived functional importance. Thus, the research question seeks answers to whether the alliance social capital of the firm contributes to its performance in the biopharmaceutical industry, and if so, how does it matter from multiple relations and functions. This article uses biopharmaceutical industry as its research context. In a theoretical context, this industry differs from others in two aspects: knowledge and institutions. In the former case, technical pressure induces inter-organizational ties (Podolny and Page, 1998 and Zimmerman and Zeitz, 2002), and in the latter case, institutional pressure induces the need and opportunities for inter-organizational networks (Suchman, 1995). Firms in the biopharmaceutical industry are induced by both types of pressures (Scott, 2001). Thus, there is a theoretical rationale for using this industry to develop and test the implication of a complex alliance social capital of the firm and its actual performance. There is also an empirical reason for using the biopharmaceutical industry for this study. Firstly, biopharmaceutical sector is a dual market structure in the alliance social capital, comprising a large number of SME (small and medium enterprises) and a small number of large pharmaceutical firms. The evidence is drawn on the large focal firm. Secondly, since most inter-firm alliances in this sector began to emerge in the early 1990s, the evidence is drawn from this relatively recent period (1994–2005). Thirdly, unlike established industries such as the steel industry and other conventional sectors that exist globally (Ghauri and Buckley, 2006, Koka and Prescott, 2002 and Young et al., 2009), biopharmaceutical firms are based in North America, Western Europe, and Japan. Finally, this pharmaceutical industry follows stringent institutional rules. The manipulation of the data and information by the firm is fairly limited. Hence, the findings can provide some useful for the future implications. The following section develops theory and hypotheses. The third section describes methods. The fourth section reports the findings. The final section provides discussion, followed by a conclusion and some implications.

In this study, I developed and tested three main antecedents and constructs in the formation of alliance social capital and empirically tested their implication in the organizational performance. I expect that the three dimensions of the alliance can lead to some fundamental insights for a better understanding. The insight comes from the gap between the prior literature and this study. The prior literature has proposed but has not fully empirically analyzed the role of alliance social capital in the organizational performance (Anderson et al., 1994). The literature that has analyzed the implication of alliance social capital on the organizational performance has a limited scope. It follows mainstream industrial sectors (Koka & Prescott, 2002) and multinational corporations (Goerzen, 2007). The current article focuses on the biotechnology sector. This sector has some unique attributes. The industry is highly dynamic with a high level of uncertainty about the future. Competition is increasing in this sector. Continuous innovation is the main source of survival and growth of the firm. For such an industrial environment, an analysis of alliance social capital and its performance implications merit some systemic attention. Although alliance social capital can have many antecedents and eventual constructs, the current study focuses on three relatively well define constructs. They are information volume, diversity, and richness. As indicated by some earlier literatures (Anderson et al., 1994), I used three constructs to measures the role of alliance social capital formation and their eventual implications on the outcome. Thus, this study discusses three dimensions of social capital and their six constitutive components. The findings reveal some insights that have not been reported in the past literature. Partially, these findings support prior literature, and partially they deviate from it. Consistent to the past literature, the findings indicate that information volume has a positive association with the firm's performance (Goerzen, 2007 and Goerzen and Beamish, 2005). In contrast, information diversity has a negative association with the firm's performance. However, information richness shows no significant relationship with the firm's performance. These multivariate analyses are different from an individual analysis. At the individual construct level (Anderson et al., 1994), however, all three antecedents and their constructs show a positive association with the firm's performance. For instance, information diversity and richness alone show positive associations with the firm's performance. Once I introduced the information volume into the regression, the direction of the coefficient changed from positive to negative. It implies that information volume might have mediated the correlations of diversity and richness. Thus, the potential confounding relations of information volume led us to conclude that the number of partners and ties might have reduced the positive association between information diversity and the firm's performance, and between information richness and the firm's performance in the pharmaceutical industry. Some of the prior literature supports these findings and some contradicts. In support, Goerzen and Beamish (2005: 348) find that a small proportion of firms has been successful in managing diverse alliances because diversity tends to reduce the performance of the alliance, especially financial performance. Similarly, they conclude that firms that enter into repeated ties show inferior performance (Goerzen, 2007). However, some earlier studies suggest the opposite. They conclude a positive correlation between diversity of alliance partners and the organizational performance (Beckman and Haunschild, 2002, Hargadon and Sutton, 1997, Koka and Prescott, 2002 and Rodan and Galunic, 2004). Likewise, some other studies find a positive link between repeated ties and the organizational performance (Gulati, 1995, Gulati et al., 2000, Podolny, 1994 and Walker et al., 1997). Why the findings from this study are different from some of the prior literature? The explanation may lie in technical, structural or relational reasons. Technically, the volume of vertical flow of industrial biotechnology may determine the outcome (Teece, 1977). The more the number of input technologies there are, the more the potential value of the information may be. A typical pharmaceutical firm needs more than 50 partners to identify a target technology (Heller & Eisenberg, 1998). This quantity of ties and partners matter more than diverse sources and repeated partners in primary functions (Anderson et al., 1994). Secondly, biotechnology is a codified phenomenon. Codified knowledge requires low integration in order to reduce cost and operational rigidity (Teece, 2000). Naturally, imposing integrated relationships such as repeated ties and other relational investment can hamper efficient search and access to information from alternative sources (Nahapiet & Ghoshal, 1998). Thus, repeated ties lose advantages. Thirdly, biopharmaceutical firms heavily rely on institutional protection for their proprietary knowledge. Firms in established industries tend to use an integrated governance mechanism to reduce opportunity risk to their proprietary knowledge (Dyer & Singh, 1998). Therefore, repeated ties may be warranted in the latter case but not in the former case. Fourthly, the biopharmaceutical industry is a high-discretion industry. In a high-discretion industry, senior managers have a relatively greater managerial latitude-of-action (Hambrick & Abrahamson, 1995). This implies that managerial strategic choices may influence the outcome (Abrahamson & Hambrick, 1997) and not the network environment alone. Similarly, firms create links for primary and secondary functions. Primary functions provide economies of scales and efficiencies in the production system of the firm, and secondary functions for symbolic reasons. The secondary function can disparately influence the primary function (Anderson et al., 1994). Regarding structural and relational explanation, Anderson et al. (1994) provide some cues to why some contextual construct lead to positive association and others to negative ones. Firstly, dyadic relations are not mere faceless things in the environment. Firm make some specific choices about their network links (Anderson et al., 1994). These links enable firms to share resources with their network partners. Since the secondary links can alter the nature of the relationship, they can also alter the outcome. Naturally, the alteration in links related the diversity and richness might have altered the correlation sign from the positive to the negative one. Indeed some links are attractive more than others in a broader network. Second, not all relational links can be expected to have a positive outcome. Some focal relations can have harmful consequences. Instead of attractive network identity and its positive performance, the antecedent links can have deleterious influence on the firm's social capital and its implications (Anderson et al., 1994). Thirdly, there might be a positive correlation between the firm's social capital and its performance at one time. There might be a negative correlation between the firm's social capital and its performance at another time. Some outcome of the interaction might be quantitative (volume based), and other may be based on qualitative perception. Because some alternative conditions can influence the outcome, the direct constructs of social networks should not be seen as positive or negative predictors from the same links (Anderson et al., 1994). Perhaps for this reason, biotechnology firms tend to avoid forming alliances with the same market actor. They avoid commitment in relationships (Kogut, Weijian, & Walker, 1992). There is also a reverse side of the explanation whereby managers consider organizational performance as the antecedent to social capital. A low-performing firm is likely to increase its social capital for a better performance. However, there may not be a positive link between social capital and the firm's performance. Therefore, the correlation between a dimension of social capital and the organizational performance can be negative or not significant. This alternative referring to perception, cognitive consistency and self serving biases is equally a valid alternative explanation of the phenomenon (Rong & Wilkinson, 2011). Although results reveal some interesting patterns, there are several potential limitations in the current study. Firstly, the data analysis is a cross-section analysis, and not a longitudinal one. Second, the dependent variable is the average of 12 years. The performance is not year-specific. In addition, researchers often capture performance with both financial (accounting and marketing) and subjective variables such as innovation. I used only one of the financial measures. Thus, the performance results can be somewhat biased. I support that some alternative measure of performance of the alliance can be insightful (Goerzen & Beamish, 2005). New product development is one such measure. Thirdly, national diversity may be related to cultural variables (Dow and Karunaratan, 2006, Hofstede, 1997 and Young et al., 2009). A control variable such as culture might have helped improve the results. The future research can take these considerations into account.