Printer Friendly

Who makes you central? Analyzing the influence of international alliance experience on network centrality of start-up firms.


* The importance of network centrality for the success of international start-ups has been frequently emphasized. Research so far has however been mute regarding the question of how firms can increase their network centrality. This paper applies a knowledge-based framework to analyze changes in the structural position of firms within their research network.

* Analyzing longitudinal event history data for the complete German biotech population for 1995 till the end of 2006 our findings show that the prior international alliance history matters for the status within a firms network. Alliances with international partners as well as alliances with partners from a variety of countries enhance the subsequent movement towards a more central position in the research network. However, multi-partner alliances inhibit the subsequent movement towards a central position, pointing to challenges in managing a portfolio of simultaneous alliances.

* The findings emphasize the necessity for a young firm to enlarge its experience with international alliances from a variety of countries but to concentrate its efforts with regards to the number of simultaneous partners.

Keywords: Alliances * Alliance portfolios * Networks * Network position * Country diversity


Firms participate in research networks for a variety of reasons, such as accessing resources and capabilities of partners, reducing risks and sharing costs (Kogut 2000; Tidd et al. 1997). Research networks have shown to be particularly valuable for start-ups in knowledge-intensive industries, such as the biotech industry. Findings by Powell and Brantley (1992) and Liebeskind et al. (1992) show that networks in this industry make an important contribution to knowledge transfer and foster growth and innovative success of participating firms. This and other research is driven by the empirical observation that young firms seldom innovate in isolation, but through research networks with multiple actors, such as suppliers, supporting and service firms, financial institutes, competitors and research institutes (von Hippel 1988; Shah et al. 1994). Moreover, these innovation networks are rarely restricted to a single national context, but exhibit a global reach incorporating actors from multiple geographic locations (Powell et al. 1996; Ritter and Gemunden 2003).

International research networks offer advantages for technology-intensive firms to share costs/risks, to access complementary assets and knowledge of other organizations, to maintain flexibility to respond swiftly to diverse sources of new knowledge and market opportunities, and to hedge on new technology fields to avoid being subject to subsequent exclusion (Wong and He 2003; DeBresson and Amesse 1991; Hagedoorn et al. 2000; OECD 2002). On a more broader term, such networks are especially important for international start-ups (Andersson and Wictor 2003; Chetty and Blankenburg-Holm 2000), who are dependent on relationships with, for example, research institutes, universities or more established firms (Oviatt and McDougall 1999), whereby both the social and the business aspects of networking are important (Johanisson and Monsted 1997).

So far, however, little research has been conducted on the factors that determine a firms' structural position within the international research network, and therefore its potential access to the networks' knowledge resources. Drawing on arguments from the knowledge-based theory of the firm in this paper we analyze factors that influence the structural status of biotech start-ups within the international research network. Specifically, we are interested in determining whether and how a firm's alliance history can influence its position within the research network.

We choose the German biotechnology industry as an appropriate context to examine our research questions. We construct the research network for the complete population of German biotechnology start-ups based on all alliance events since the founding of the industry in the mid 1990's. We formulate five hypotheses linking different types of alliance experiences to a firms' network centrality. All of our hypotheses are confirmed, suggesting that the type of prior alliance partner matters for the position of a firm in the research network. Confirming our first hypothesis, prior experience with research alliances significantly increases the attractiveness of the firm as a network partner. When differentiating the type of alliance experience clearly the nationality of the alliance partner matters most. While collaborating with international partners from different countries significantly increases network centrality and status, there are negative effects from collaborating with private firms such as pharmaceutical firms. Similarly having a history of alliances with multiple partners simultaneously influences the centrality score negatively. These interesting findings points to challenges in managing a portfolio of international alliances.

Importance of Network Position

The importance of network membership for young firms is widely recognized in international management and international entrepreneurship (c.f. Johanson and Vahlne 2003; Chetty and Wilson 2003; Chetty and Blankenburg-Holm 2000; Blomstermo et al. 2004a, 2004b). Various studies have provided strong empirical support for the fact that the internationalization strategy of start-ups benefits from linkages to suppliers, customers, or even competitors who are themselves internationally connected (c.f. Al-Laham and Souitaris 2008). International networks provide firms with information about market opportunities and give them access to local resources, thus fostering their internationalization strategy (i.e. Zahra et al. 2003; Prashantam 2005). Furthermore, firms develop international linkages to draw upon country-specific knowledge (Almeida et al. 2002) that can be used for knowledge exploration or exploitation (Coombs et al. 2006). International networks are, however, not only important for a firms internationalization strategy, but also for a firms innovative success (Ritter and Gemunden 2003; Shan et al. 1994; Hagedoorn and Schakenraad 1994; yon Hippel 1988). This is especially valid for international research networks.

It is widely recognized that the only way for an organization to sustain innovation is to constantly upgrade its knowledge base (Dosi et al. 1988; Acs and Audretsch 1990; Iansiti and Clark 1994; Spender 1996). From a strategic point of view the importance of knowledge for competitive advantage has been emphasized in recent contribution to a knowledge-based theory of the firm (see Spender and Grant 1996). However, since a start-ups initial endowment with valuable knowledge in the form of accumulated capabilities and know-how is limited, accessing external knowledge flows is of crucial importance (see March 1991; DeCarolis and Deeds 1999). Consequently, network membership has been emphasized as a means for young firms to access external knowledge (Hansen 1999; Hansen 2002; Nahapiet and Ghoshal 1998; Tsai 2001; Uzzi and Lancaster 2003; Uzzi 1997).

Some of this research highlights the institutional properties of inter-organizational networks in dealing with complex and uncertain knowledge based transactions (see Grandori and Soda 1995; Alter and Hage 1993). In particular in sectors where knowledge is developing rapidly such as biotechnology, some studies have demonstrated that intense interorganizational network activity promotes innovation (Powell et al. 1996). According to Powell et al. (Powell et al. 1996, 2005; Powell and Brantley 1992) and Liebeskind et al. (1996), research networks in biotechnology make an important contribution to innovation because they serve as sources of new, diverse reliable information and knowledge.

Most of that work has focused upon a research networks' structural properties and the fit between such properties--particularly the strength of ties, and different kinds of knowledge transmission (Shan et al. 1994; Stuart and Podolny 1996; Hansen et al. 1999; Uzzi 1996, 1997, 1999; Ahuja 2000). So far, however, only a few studies have examined the structural position of a firm within its research network and its effect on knowledge transfer. That research has shown that firms with a high centrality within their research network--e.g., core firms--have greater access to and control over resources, which proves particularly valuable in knowledge-based industries (Podolny 1993; Podolny and Stuart 1995; Rowley and Baum 2002). For instance, Tsai (2001) explores that both the most innovative and profitable business units occupy central positions within their research network. Similarly, findings by Gnyawali and Madhaven (2001) show that a superior position in the research network translates into a resource advantage and an increased likelihood of innovative advantages. Central firms can therefore access new information and critical knowledge easier and faster than peripheral firms.

There are several reasons fort his effect (Shipilov 2003, p. 15). Network centrality is directly linked to network status (Rao et al. 2008). Firstly, peripheral firms in the industry are actively seeking involvement with high-status actors. For a lower-status firm, one way of attracting the attention of high-status partners is to offer them access to information or know-how that can be jointly exploited. In this case, high-status firms do not need to conduct industry-wide searches for novel information themselves, but they can rely upon their partners to bring this information to their attention. Second, high qualified scientists in the industry would rather work for high-status, than for the lower-status organizations (Frank 1985). This human capital brings with it not only its own knowledge, but also the information and capabilities embedded in its own personal networks. Thus, higher status firms are likely to have access to wide and diverse knowledge networks of qualified scientists in their industry, which could be used for conducting industry-wide searches when necessary.

To sum up, recent research points to the fact that finns with a central position within a research network e.g. firms with a high network status--have easier access to critical knowledge than peripheral firms, and build up their innovative capabilities faster than peripheral competitors. Therefore, over time the position within the network should translate into success. However, so far studies have been mute with regards to dynamic issues, in particular the question of factors determining a young firm's position within its network. Given the high desirability a central network position entails, we would expect young firms to spend significant efforts in increasing their network centrality, e.g. to gain a more favorable position within their research network. However, we do not know much about the factors that increase centrality within a research network. Given this gap in the literature, our efforts in this paper aim on analyzing strategies for firms to improve their network position.

Hypothesis Development

This paper concentrates on the effects that experience with research alliances will have on the position of a biotech firm within the international research network. As a general hypothesis we state that the history of research alliances a firm has accumulated will influence its desirability as a network partner, and hence increase its network centrality. More specifically, we will argue that the type and the characteristics of alliance partners will moderate this influence in several ways.

Main Effect: Consequences of Research Alliances for Network Centrality

In general a research alliance has two partners--usually one of the two partners initiates the alliance by identifying and targeting the other. To theoretically explain the effect of a firm's experience with research alliances on its network position we look at the issue from two perspectives: How do prior alliances help the focal firm to initiate new alliances and build up alliance capabilities? And how do prior alliances (and the knowledge they represent) make the focal firm more attractive as a network partner? We therefore take the perspective of both alliance partners.

First, we argue that research alliances are an important means to build up a stock of strategic relevant knowledge in the biotech-industry, which in turn makes the biotech firm a valuable partner in the international research network. We draw upon the empirical observation that the amount of prior alliance experience a particular biotech firm encounters represents its stock of externally acquired knowledge (e.g. DeCarolis and Deeds 1999; Al-Laham and Souitaris 2008). In the case of a research alliance this knowledge will be of scientific nature, therefore making the focal firm a technically attractive partner for a firm sourcing for scientific knowledge. Established research relationships will make the focal firm a more attractive partner by increasing its direct access to technological knowledge and resources (George et al. 2002; Mian 1997). This is consistent with the observation of Zucker et al. (1998) who argue that scientific knowledge does not transfer via informal spillovers but through direct collaboration. Furthermore, prior alliances increase the visibility of the focal firm as a reliable alliance partner, hence we expect signaling and reputation effects to the network partner sourcing for knowledge.

From a second perspective we argue that by accumulating that stock of external knowledge a biotech firm accumulates valuable capabilities to initiate and manage research alliances (so called alliance capabilities). By providing valuable information (Gulati 1995, 1999) experience with cooperation (Walker et al. 1997) and procedural knowledge of how to manage alliances (Kogut and Zander 1992) prior alliances play an important role in shaping the formation of future alliances. Repeated alliance engagements over time contribute to the build-up of an alliance management capability (Dyer and Singh 1998; Gulati 1999; Rothaermel and Deeds 2006). Additionally firms learn how to manage alliances more formally, such as to draw contracts and to sign agreements (Mayer and Argyris 2004). This means that the more relationships a firm has, the more it knows about how to manage them so the less costly it is to form new relationships (Walker et al. 1997).

Even more so, by engaging in research alliances firms build competency in integrating and recombining the various components of their knowledge stock to develop new scientific knowledge. Henderson and Cockburn (1994) provide considerable support for the importance of "architectural competence" as a source of competitive advantage, a concept similar to what others have called integrative capabilities (Lawrence and Lorsch 1967), combinative capabilities (Kogut and Zander 1992), higher-order capabilities (Sanchez and Heene 1997) or dynamic capabilities (Teece et al. 1997).

Given these findings, we argue that the unique capabilities of a firm to recombine and integrate its own (prior) and the new acquired knowledge develop over time as firms show a history of successful partnering (Anand and Khanna 2000). Given these positive effects, we expect them to be more valuable and sought after network partner than firms without alliance experience. Hence, over time their network centrality should increase with increasing past alliances experience. We therefore state our main hypothesis as follows:

Hypothesis 1: A history of research alliances will increase the network centrality of a German biotech firm.

In the subsequent sections we will refine this basic line of argumentation by specifying the type of partner a biotech firm can cooperate with, and the consequences thereof.

Moderating Effects: Partner Type and Characteristics

Domestic versus International Partners

Our second hypothesis draws upon the observation that roughly 40% of all research alliances in the German biotech industry are formed with international partners. What influence could international ties have on a biotech firms' network position? To develop our hypothesis we draw upon two streams of argumentation. First, we assume that international linkages provide more valuable knowledge and capabilities to the focal firm then national linkages, which over time will make it a more valuable network partner. From a knowledge-based point of view, international linkages provide a fruitful arena for knowledge exchange due to the greater heterogeneity and diversity of partners' knowledge bases. As has been discussed extensively in the literature, the efficiency of learning from alliance partners is dependent upon the relative absorptive capacity of both partners. Following prior research, the amount of knowledge transfer in alliances is largely determined by the relative relationship between knowledge bases of partners (c.f. Lane and Lubatkin 1998). Organizations will have the greatest potential to learn from organizations with similar basic knowledge but heterogeneous specialized knowledge (relative absorptive capacity). Furthermore, research points to the fact that operating in diverse contexts increases the variety of events and ideas to which a firm is exposed, leading to a more extensive knowledge base (Huber 1991; March 1991). Transferring these ideas to an international context, Barkema and Vermeulen (1998) argue accordingly that firms who operate in international markets develop a richer knowledge structure and stronger learning skills, which make them a desirable network partner.

From a second perspective we expect prior alliances of the focal firm to act as quality signals to potential international partners. Gulati's research (1995) has shown that there is greater distrust between partners when forming cross-national alliances. For the international alliance partner, a big concern is the predictability of the local partner's behavior. Faced with a lack of prior experience with a particular domestic partner, the next logical step is to rely on the reputation of that firm, which is a direct consequence of that firms' prior relational behavior (Granovetter 1985). Within a network, reputational considerations play an important role in a firm's potential for future ties, because these social affiliations serve as a source of legitimacy (Uzzi and Lancaster 2003; Gulati and Gargiulo 1999; Uzzi 1996). The signaling properties of prior ties are particularly important for knowledge-based industries (Podolny 1994) and for collaboration across national boundaries (Al-Laham and Amburgey 2005). A history of the national partner with research alliances can signal trustworthiness and reputation, "because the reputational consequences of opportunistic behavior are greater in a domestic context" (Gulati 1995, p. 95).

To sum up, we expect international ties of German biotech firms to moderate the influence of prior alliances on network status. More specifically, we expect international ties to increase the positive influence of alliance experience on network status. We therefore state:

Hypothesis 2: The influence of prior research alliances on the network centrality of German biotech firms will be higher the more international partners the firm has cooperated with.

Diversity of Partners Countries of Origins

In addition to the question whether a biotech firm has formed prior alliances with international partners it is important to consider how many distinct countries these partners represent. With regards to the diversity of knowledge bases of alliance partners we assume that allying with partners from a variety of countries represents a broader and richer forum for knowledge exchange then allying with partners from one specific country only.

As firms are interacting with alliance partners from a variety of international countries they benefit from the market experience of these firms in several ways. First, they 'circulate in different flows of information' (Burt 2000, p. 352), and are exposed to different market knowledge, internal structures, and research routines of these firms (Miller and Chert 1996; Mezias and Glynn 1993). This allows the focal firm to develop breakthrough innovations and novel knowledge by combining and comparing individual pieces of scientific knowledge and research routines from a variety of different countries. Comparing scientific knowledge from different countries helps firms also understand the underlying demand structures common to different markets, which is of crucial importance for their subsequent commercialization efforts. Moreover, through interactions with a broad array of international network partners, firms may "discover links between members' knowledge and create new knowledge than member had previously possessed (p. 583)" (Lewis et al. 2005).

Furthermore, as we have pointed out above, information variety positively affects knowledge transfer and learning in firms (Huber 1991). Firms with connections to partners from a variety of countries develop a broader knowledge base and more sophisticated routines for the development of their own innovative capabilities (Tsai 2001). Erramilli (1991) found that firms that are exposed to limited variation accumulate limited knowledge about the opportunities in a given market. In contrast, firm with a broad array of market exposure are likely to search for more information about customers needs, learn new and different things, and accumulate more higher-order routines (Perkin and Rao 1990). Based on the above, we expect that firms that connect to a variety of markets will be able to access more valuable and innovative knowledge then firms concentrating on a single or few markets. These firms should therefore be more sought after partners in the international network. The diversity of alliance partners countries should thus moderate our main hypothesis as follows:

Hypothesis 3: The influence of prior research alliances on the network centrality of German biotech firms will be higher the more diverse the international partners countries of origins are.

Private versus Public Partners

Our fourth hypothesis differentiates alliances by the type of partner involved with the biotechnology firm: Public research organizations or private firms. We hypothesize that both types of partners show different characteristics, which influence the amount of knowledge the biotech firm can access, and hence influence its desirability as a network partner.

We draw upon the observation that the different types of partners vary substantially in their scientific knowledge base and dominant logics. The scientific knowledge base of biotechnology firms is close to that of public research organizations (e.g. both conduct research in the area of molecular biology). Although biotechnology firms share enough scientific knowledge with public research organizations to effectively learn and transfer knowledge, they show a great dissimilarity with regards to their operational knowledge. This is due to their dissimilar dominant logic. Public organizations engaged in basic research are oriented towards the production of knowledge rather than the conversion of scientific knowledge into commercial products.

A different picture emerges when we look at the second type of partner. The scientific knowledge of biotechnology firms and pharmaceutical or chemical firms is dissimilar (e.g. molecular biology vs. biochemistry). However, the operational knowledge base of biotechnology firms is closer to that of traditional pharmaceutical firms than to that of public research organizations. This is due to the fact that, as entities seeking economic rents, biotechnology firms and traditional pharmaceutical firms have similar dominant logics with both striving for the conversion of scientific knowledge into commercial products (Bettis and Prahald 1995).

These distinctions are important for the degree of learning that can occur and knowledge that can be transferred within the alliance. Lane and Lubatkin (1998) extending and refining the firm-level concept of absorptive capacity to a dyadic level, argue that the amount of knowledge transfer is largely determined by the relative relationship between knowledge bases of partners. They therefore expect the similarity in knowledge bases of partners to influence the amount of knowledge transfer. With regards to the biotech industry, basic knowledge is reflected in the scientific knowledge bases of the organizations, while specialized knowledge is reflected in the operational knowledge bases of the organizations. Dedicated biotechnology firms therefore share similar basic knowledge with public research organizations (disciplinary foundations) but differ with regard to their specialized knowledge (commercialization versus basic research). Alternatively dedicated biotechnology firms and traditional pharmaceutical or chemical firms share specialized knowledge but differ with regard to their basic knowledge base. We therefore argue that relationships with public organizations should produce higher relative absorptive capacity and thus higher knowledge transfer for biotechnology firms than relationships with pharmaceutical firms. Their attractiveness as a network partner will therefore be increased.

There are however additional considerations indicating positive consequences of public partnering for the focal firm. Prior research has indicated that high status network partners are more likely to choose a new partner if it is directly backed by a credible third party whom they trust (Powell et al. 1996) and who offers a signaling value (Gulati and Higgins 2003; De Carolis and Deeds 1999). Formalized ties to public research institutions can play this signaling role. Relationships with established and reputable research institutions can enhance a biotech firms' legitimacy in the eyes of the potential network partner (Gulati and Higgins 2003).

In summarizing our arguments we hypothesize that a history of alliances with public research institutes will make the focal firm a more attractive network partner than a history of alliances with private firms. We therefore state:

Hypothesis 4: The influence of prior research alliances on the network centrality of German biotech firms will be lower the more private partners the firm has cooperated with.

Multiple Partners

Our final hypothesis differentiates the prior research alliance experience a biotech firm has by the number of different partners. Our general argument stated above refers to the positive effects experience with research alliances has on a firm's status within the international research network. However, we assume that it makes a difference whether a firm has a history of alliances with a single partner--e.g, repeat alliances--or with multiple partners. In doing so we see a downside to a large involvement in alliances. Many of the cooperative arrangements in biotechnology govern some carefully specified and narrow range of products, and confine the parties to well-defined activities. However, the problem is that each alliance partner has rights to claim to the ownership of specific technologies (Shah 1990). Therefore, a large number of cooperative partners in a single relationship may block the exclusive technological advantages a network partner is looking for and decrease the firm's attractiveness. Our latter argument is explicitly validated by the basic arguments established in the knowledge-based view of the firm. According to that view, firms benefit most from inimitable, idiosyncratic and non-transferable capabilities (Spender 1996; Liebeskind 1996). The more partners a biotech firm has cooperated with, the more the valuable knowledge will be spread across the industry, and the less the former requirements will be met.

To sum up, we expect firms having a history of research alliances with multiple partners will be less attractive as a network partner.

Hypothesis 5: The influence of prior research alliances on the network centrality of German biotech firms will be lower the more multiple partners the firm has cooperated with.

Empirical Setting, Data and Methods

Empirical Setting: The German Biotech Industry

Hampered by a hostile regulatory environment for genetic research throughout the 1980s and early 1990s, and facing additional institutional constraints, the German biotechnology industry was de facto not existent prior to the mid 1990s (Casper 2000). In the mid 1990s the German government introduced a series of new technology policies and programs designed to orchestrate the development of innovative technologies and small business start-ups. The most successful of these programs has been the so called "Bio-Regio" Competition, that was designed to create and promote clusters of entrepreneurial start-up firms by funding biotechnology promotion offices in 17 different German regions (see Ernst and Young 1998, 2000, 2002, 2003). Together with other institutional changes this has led to a dramatic increase in growth rates for German biotech start ups. Over the last five years, more than 500 new biotechnology start-ups have been founded in Germany, most of them located in bioregio-clusters around universities and public research institutes (Ernst and Young 2003).

The German biotech industry is also an industry with a high degree of internationalization. More then 40% of all alliances in this young industry are formed with international partners, moreover organizations from the U.S. comprise the majority of selected partners. Table 1 and Fig. 2 give some information on the nationality distribution of international alliance partners. An explanation for this interesting pattern has to take the specifics of the institutional context in the U.S. and Germany into account. Murmann (2003, p. 93) observes that since the end of the 1960s the United States had surpassed Germany in basic and biomedical research in the life sciences due to differing institutional contexts, such as differing funding patterns of the Federal government for basic research in molecular biology. According to Murmann's (2003) analysis, in Germany the new field of recombinant DNA did not nearly develop at the same speed because German law prohibited the manipulation of genetic materials in response to the euthanasia legacy of the Nazi era. As a result Germany was not at the forefront of the new science, and did only start to devote significant resources to the new technological regime in the mid 1990's, as we have described above.

Given the relatively short history of the German biotech industry, it seems plausible that the German population has not yet developed the same level of innovative capabilities as have the U.S. or major European players such as the British or French biotech industries. Evidence from comparable industry settings has shown that under these circumstances firms source critical knowledge internationally (see Ernst 2000; Kristensen and Lund-Vinding 2001).

Turning to the firm level, there are certain unique characteristics that distinguish firms in the biotech industry from other industries. These characteristics include a long and complex product development and approval cycle, heavy reliance upon basic scientific research and a set of very heterogeneous technologies with the potential to transform various application fields. If they are to succeed, biotechnology firms are forced to engage in intensive competition for knowledge and intellectual property. Due to their very specialized business models biotechnology firms periodically have to access knowledge from outside the firm, i.e. form research alliances with various partners and on an international basis. As a consequence, the biotechnology industry has been identified as the industry with the highest international alliance frequency among several industries characterized by high alliance activity (Hagedoorn 2002).


The data used in the study consists of the complete population of 853 German biotechnology start-ups in existence in 1995 or founded thereafter. We used four primary sources to compile the sample. The first were the daily registration and deregistration records of the German Commercial Register ("Bundeszentralregister") in Berlin, the second the "Yearbooks of the German Biotechnology Industry" published yearly by the German company, Biocom AG. The third source were the monthly TRANSCRIPT newsmagazine that reports on the German biotech industry; this periodical is also published by Biocom AG. The final source were the monthly records from the German Patent and Trademark Office in Munich, published by PATHOS GmbH as the primary source for the assignment of patents. The firms were observed from 1995 until the end of 2006.

These data, and other sources, were used to construct an event history for each company. Event histories are data structures that include information on the number, timing and sequence of the events that are being examined. Each firm's history began at the time of its incorporation and ended at the time of an event or at the end of the month, whichever came first.

The organization's second spell began on the following day and ended at the time of an event or at the end of the month. This pattern continued until the firm exited (through failure or acquisition) or until the end of the observation period, in which case spells were coded as "right censored." This procedure allowed timevarying covariates to be updated throughout the firm's history at monthly intervals. In those cases where only the month and year of an event could be determined, the day was set at the midpoint of the month to minimize errors in timing.

The event history data was used to construct a pooled cross section time series data structure. First, the alliances were used to construct the network in existence at the end of every year. In other words, using all of the alliances across firms we constructed the network in existence at the end of 1995, the end of 1996, et cetera. We then used these annual networks to calculate annual centrality scores for each of the firms. We also converted the event history data to annual measures for our alliance based independent variables through a process of aggregation. For example variables such as international partnerships were constructed by summing all of the partnerships in a given year to create an end of the year measure.

Dependent Variable

Bonacich's (1987) eigenvector centrality was used to measure the status of firms in the German biotech industry research network. All research and development alliances in effect during a calendar year were used to construct the research network for that year.

The UCINET program was used to construct the Bonacich (eigenvector) centrality score for each organization in the network. This indicator can formally be defined as:

[S.sub.t] (a, B) = [[infinity].summation over (k = 0) a[B.sup.k][R.sup.k+1.sub.t] 1.

In this expression, a is a scaling coefficient, B is a weighting parameter that can range between zero and the absolute value of the inverse of the value of the maximum eigenvalue of the sociomatrix [R.sub.t], 1 is a column vector where each element has the value "1," and s is also a column vector where element [S.sub.i,t] denotes the status of biotech firm i. Given this specification, a biotech firms position (status) is a function of the number and the position of the firms with which it forms cooperative research agreements. In turn, the position of these partners is the function of the number and the position of their partners, and so on. The B parameter is set equal to the reciprocal of the maximum eigenvalue. The firm's eigenvector centrality at the end of each calendar year was used as our independent variable.

There are several measures of network centrality that can be used for different purposes. 'Degree centrality' is a count of the number of partners with which a focal firm is allied (i.e. another term for network size). 'Betweenness centrality' is a calculation of the probability that a firm lies between two other firms on the shortest path between them and indicates potential indirect access to information or control that the firm employs. 'Closeness centrality' is a measure of the shortest path between a firm and all others. We used Bonacich's (1987) eigenvector centrality which has the major advantage of measuring the centrality of the firm in the network not only by counting the number of ties, but also by calculating the position of the firm's partners and including them in the measure. Therefore, a biotech firm's centrality is a function of the number and the centrality of the firms with which it forms cooperative research agreements. In turn, the centrality of these partners is the function of the number and the centrality of their partners, and so on. For example take two biotech firms A and B each of which has 5 ties. They will have identical degree centrality. However if firm A has ties to partners that themselves have many ties and firm B has ties to partners with few ties then firm A will have a higher eigenvector centrality than firm B. Betweenness and closeness centrality were conceptually less appropriate than the eigenvector centrality in our case, because they treat all ties equally, whereas eigenvector centrality weights partners by their own centrality (Soh et al. 2004). Since eigenvector centrality is widely used in network analysis as a measure of status or prestige we decided to use it as opposed to other centrality measures. Our measure is therefore a measure of the status of a firm in the network (see Wassermann and Faust 1994, p. 205).

The relevance of using a status measure becomes evident when we visualize the German network at a given point in time. In Fig. 1 we have included a graphical representation of the German biotech-network for the Year 2005. All research alliances in place in 2005 were used to construct the network, utilizing the spring embedded mode in the software packages UCInet and NetDraw for visualization.

The graphical analysis in Fig. l a unveils that the network consists of a large and dense core (all nodes are connected by at least one path) and a periphery of unconnected smaller components. Figure l b depicts an isolated view on these peripheral components. Turning to our dependent construct--the eigenvector centrality--we see that a firm can have a lot of ties (i.e. a high degree centrality) but still remain in the periphery of the network (as visualized in Fig. 1b). The eigenvector centrality score however reflects the position of the firm in the network core in a more accurate way.


Independent Variables

Our primary independent variables are prior research alliances of the firm with domestic or foreign partners, prior research alliances of the firm with other firms or with public organizations, prior research alliances of the firm with multiple partners and the diversity of countries within which the biotech firms have formed alliances. Prior research alliances established by a firm is a simple count of all the R&D agreements established by a German biotech firm. Prior international alliances is a count of alliances where the partner is based outside of Germany (e.g. the German biotech firm forms an alliance with a partner based in the United States). Table 1 and Fig. 2 provide a descriptive overview of the distribution of international alliance partners of German biotech firms according to their country of origin.

Prior private alliances is a count of alliances where the partner is a firm (such as another biotech company or a pharmaceutical company) rather than a public organization (such as a university or non-profit research institute). Prior multiple partner alliances is a count of alliances where there is more than one partner involved in the same alliance.

Country diversity represents the number of distinct countries where a biotech company has formed alliance partnerships. For example let's observe a firm that has formed partnerships with a U.S firm, then another U.S. firm then a Canadian firm. After the first alliance the country diversity index is 1. After the second alliance the country diversity index is still 1 since it has already formed an alliance with a U.S. firm. After the third alliance, the country diversity index is 2 because there are 2 countries that the German firm has formed alliances within. As we see from Table 1 and Fig. 2 German biotech firms form alliances with partners from a variety of countries, although U.S. firms dominate by numbers.

In order to test our hypotheses H2-H5 about moderating effects we constructed four multiplicative interaction terms. The first is the number of prior alliances multiplied by the number of prior international alliances. The second is the number of prior alliances multiplied by the number of prior private alliances. The third is the number of prior alliances multiplied by the number of multiple partner alliances. The last is the number of prior alliances multiplied by country diversity (the number of distinct countries in prior alliances). As with the dependent variable, the values of all of the primary independent variables were taken at the end of each calendar year. We therefore condensed the more detailed information contained in our event history structure into series of annual observations.

Control Variables

We included as controls a number of variables at the firm and industry level known or expected to affect the network position but not included in our hypotheses. One firm-level control was age, measured as the number of days since the founding or qualification of the firm. The second control variable was size, measured by the number of employees the firm reported employing. We used two variables to measure the absorptive capacity of a firm to control for her ability to recognize the value of new, external knowledge, assimilate it, and apply it to commercial innovations (Cohen and Levinthal 1990).


The first of these was the number of research domains of the firm as an indicator of the breadth of the knowledge base of the firm. We used the self-reports of firms compiled in the Yearbook records to classify each firm e.g. genetic diagnostics, polymer protein coating, tissue engineering, among others. The number of research domains in which firms were active was a simple count. The second control for absorptive capacity was technological sophistication of the firm as an indicator of the complexity of her technological knowledge base. To measure the sophistication of technological capabilities we coded the laboratory types firms reported to utilize. We therefore follow Casper (2000) and argue that the sophistication of a biotech firm's technological knowledge is to a great part reflected in the technological sophistication and complexity of the laboratories and research facilities it uses. A total of 8 laboratory types was used to classify each firm: Chemical lab, chemical-biological lab, L1, L2 and L3-Lab, S1, S2 and S3-Lab. These Lab-Typs are classified according to the requirements of the German Ministry for Education and Research. Among these classifications, L3 and S3 laboratories reflect the highest technological complexity and security standards. We therefore constructed a dummy indicating whether the firm was utilizing a L3 or S3 laboratory. The annual number of corporate patents granted in genetic engineering was used to measure cumulative patent activity.

On the industry level we included density measures of the regional clusters the biotech firms are located in. To measure the density of the local clusters we coded for the number of Core-Biotechnology firms, universities, private or public research institutes, laboratory equipment and materials suppliers, consulting firms such as IT-Services and management consultants, and financial institutes located in the same 2-digit postal code area as the firm. Clustering at the 2-digit level represented a compromise between a smaller geographic region such as the 5-digit postal code such as the city and a larger region such as the "Bundesland" (state). To measure the variety of the local supporting cluster we constructed the Herfindahl index as follows: 1 - [summation][p.sup.2.sub.i] where the [p.sub.i] represents the proportion organizational category i in the 2 digit postal region. The values of all of the control variables were also taken at the end of each calendar year.


We used a cross sectional time series regression with fixed effects:

y(i,t) = a + B* X(i,t) + u(i) + e(i,t)

where y(i, t) is the Eigenvector centrality-score for firm i in year t. The fixed effect specification includes a firm specific component u(i) in the error term and a random error term e(i,t) which is not specific to the firm. We choose the fixed effect specification for our initial modeling to allow for unobserved heterogeneity among firms. The F-Test was used to evaluate the goodness of fit of the models, and t-tests were used to evaluate the statistical significance of individual parameters.

We also estimated all of the models with a random effects specification to evaluate the sensitivity of our analysis to model specification. The specification of estimation model had no substantial impact on our results. Although there was some variation in the size of the coefficients the signs and statistical significance are the same regardless of which estimation technique is used. However, our results show that there is significant correlation between the error terms and the independent variables. Despite any drawbacks the fixed effects model addresses this issue through the inclusion of the firm specific component of the error term but the random effects specification does not. We therefore report the results of the fixed effect model only.


Table 2 provides means and standard deviations for the variables in our models as well as a correlation matrix. Table 2 indicates relatively moderate inter-correlations among most of the independent variables. This is not surprising since all of the alliance based variables tend to correlate and the multiplicative interaction terms will as well. Given the large number of observations (2501 firm-years) in the data multi-co linearity is however less likely to be a problem. Nonetheless, our results should be viewed with the usual caution.

Table 3 provides the results of our regression models. Model 1 provides parameter estimates for only the control variables, models 2, 3, 4, 5 and 6 add subsequently the interactions of our main independent variables, and model 7 is the complete model with all our variables. The results of model 2, 3, 4, 5 and 6 are only reported for information purposes, only model 7 is used to evaluate our hypotheses.

The parameter estimates in Model I indicate that only two of the control variables have a significant impact on the centrality of firms: The lagged dependent variable and size. Model 2 adds a parameter estimate for prior research alliances. For the control variables the lagged dependent variable remains positive and significant but size is no longer significant. Research breadth becomes significant with a negative effect. More importantly the effect for prior research alliance is quite strong and has a significant positive effect.

Model 3 adds a multiplicative interaction term for prior international alliances. For the control variables the lagged dependent variable remains positive and significant but research breadth is no longer significant. Instead, the number of prior patents becomes significant with a negative effect. The prior international interaction term has (as predicted) a significant positive effect. However the main effect for prior research alliances is no longer significant. Model 4 adds a multiplicative term for prior country diversity. For the control variables, the lagged dependent variable is again positive and significant. The effect of prior research alliances is again positive and significant as is the international alliance interaction term. However, the country diversity interaction term is significant but negative.

Model 5 adds a multiplicative term for private partner alliances. For the control variables, the lagged dependent variable is again positive and significant. Similarly, the number of prior patents remains negative and significant. The effect of prior research alliances is again positive and significant as is the international alliance interaction term. The country diversity interaction term is again significant but negative. The private partner interaction term is also negative and significant. Model 6 adds a multiplicative term for prior multiple partner alliances. The interaction term is negative and significant, all other variables remain in the same direction and significance.

Model 7 is the final model with all of our control variables, all of our main variables, and all of the interaction terms. This model is used to discuss our results. We confirm all of the five hypotheses. The effect of prior research alliances is positive and significant, confirming our first Hypothesis (H1). The international alliance interaction term is positive and significant, confirming our second Hypothesis (H2). The country-diversity interaction term is positive and significant, thus confirming our Hypothesis 3. As hypothesized, the private partner interaction term and the multiple partner interaction term are significant and negative, thus confirming Hypotheses 4 and 5. The examination of the F-Values for all seven models indicates that each of the models provides a progressively improvement in fit.

Discussion and Conclusion

In this paper we examined factors influencing a biotechnology firm's position in the international research network from a knowledge-based perspective. Our research question was motivated by recent findings pointing to the fact that firms with a central position within a research network e.g. firms with a high network centrality--have easier access to critical knowledge than peripheral firms, and build up their innovative capabilities faster than peripheral competitors.

Examining the research conducted so far, we noted that studies on networks have been largely mute with regards to factors determining a firm's position within the network. Given the high desirability a central network position entails, our efforts in this paper therefore aim on analyzing strategies for firms to increase their network centrality. Specifically, we hypothesize that alliance experience should influence the desirability of a biotech firm as a research partner, and hence its position within the research network. We choose the German biotechnology industry as an appropriate context to examine our research questions. That industry is a salient example of a very young and knowledge intensive setting in which cooperation to generate new knowledge is the most important component of firms business model. More so, the German biotech industry is also an example for an industry characterized by a high degree of international embeddedness due to the prevalence of large multinational pharmaceutical companies, international suppliers and due to the global character of the scientific discovery process. Despite the short history of the industry, more than 40% of all research ties in the German population are international ties, characterizing the network therefore as truly international in nature.

In constructing the research network of that industry and examining all alliance events in the complete population of German biotechnology firms since 1995, we formulate five hypotheses linking different characteristics of alliance partners to a firm's network centrality score. All of our hypotheses are supported by our data.

We anchor all of our hypotheses in the knowledge based view (Spender 1996), which suggests that firms can access knowledge by means of strategic co operations. We formulate a main hypothesis and four hypotheses reflecting moderating effects. As our main hypothesis we expect a firm's accumulated experience with research alliances to increase it's network status. Following the literature (i.e. DeCarolis and Deeds 1999; Anand and Khanna 2000) we argue that the amounts of prior alliance experience particular biotech firm encounters represents its stock of externally acquired knowledge. This stock in turn makes the firm an attractive network partner in subsequent alliances. And second, we argue that by accumulating that stock of external knowledge a biotech firm accumulates valuable capabilities that foster its innovativeness, and thereby increase its attractiveness as a network partner for firm's sourcing for knowledge. Our findings clearly confirm our main hypothesis. The eigenvector centrality of the German biotech firm as a measure of its network status significantly increases the more research alliances the firm has formed in the past, indicating that the firm is moving towards the central regions of the international research network. The eigenvector centrality is not simply a function of the number of alliances of the focal firm, but a function of the status of its partners--thus the focal firm gains in network status. We explain this finding with the value of the knowledge and capabilities the firm has accumulated in its alliance history, and with signaling effects regarding its reliability as a research partner.

Our second hypothesis addresses the moderating effect of the nationality of prior partners a biotech firm has allied with. Given the specific context of the young German biotech industry, we assume that domestic firms cooperate with international partners--i.e, from the U.S.--to build up their innovative capabilities and to access complementary knowledge (see Murmann 2003). From a knowledge-based point of view, these international linkages provide a fruitful learning arena due to the greater heterogeneity and diversity of partners knowledge bases. We therefore expect firms with international partners to have better accessed the valuable capabilities then firms with purely domestic partners, and hence to become a more attractive network partner in subsequent partnerships. They should therefore occupy more central positions within the network. Our findings confirm our hypothesis, showing that a history of international alliance partners increases a biotech firm's position within the network. Beside the purely knowledge-based explanation for this finding, we additionally assume status benefits for international connected firms in that young industry: Those firms might be considered a more reliable alliance partner, or a more prospective alliance partner regarding future outcomes. International ties thus act as "endowments" for young German firms, a phenomenon that has been confirmed by prior research for a purely U.S. context (see Stuart 2000; Stuart et al. 1999).

Our findings also confirm our third hypothesis. Cooperating with a variety of partners from different countries does make the firm more important in the network, and increases its centrality over time. We believe that these findings point to the benefits of building up international diversity with regards to research alliances. The benefits of diversity seem to be especially important for the young start-ups in the German biotech industry. The more variety the firm encounters in its alliance portfolio, the more knowledge it can access, an the more valuable it becomes as an international network partner.

Our data also confirm our fourth hypothesis. We assumed that it matters whether a biotech firm cooperates with a public research organizations or a private firm. In laying the foundation for our hypothesis, we draw upon the knowledge-based view and the important concept of relative absorptive capacity. Following that concept, organizations will have the greatest potential to access knowledge from organizations with similar basic knowledge but different specialized knowledge (see Lane and Lubatkin 1998). According to our observation, for a biotech firm the amount of accessible knowledge should be higher when cooperating with a public research institute then cooperating with a private firm (i.e. pharmaceutical company). We therefore expected these biotech firms to be a more desirable partner within the research network, and we expected a history of cooperations with private firms to decrease their network centrality more then cooperations with universities. The results of our analysis confirm our hypothesis. The dummy variable for private firm significantly depresses the influence of alliance experience on network status, indicating a negative effect for private firms, and a reverse effect for public organizations. It therefore matters whether a biotech firm partners with a private or a public firm; a history of partnering with public research institutes and universities sends a strong positive signal to potential network partners.

Finally, our data confirm our fifth hypothesis, pointing to the fact that a history of research alliances with multiple partners deteriorates a firm's network centrality. We argued that a large number of simultaneous cooperative partners may block the exclusive technological advantages a network partner is looking for and decrease the firm's attractiveness. Given our findings, we are convinced that a history of alliances with multiple partners increases the risk of disentangling the focal firm's valuable knowledge; spreading the knowledge across the industry and increasing the danger of being imitated by rivals. The technologically valuable capabilities the biotech firm might have acquired are therefore no longer a distinguishing feature for its network status. We also believe that the results point to the necessity to concentrate on a few alliances with few partners simultaneously, instead of spreading the efforts across partners.

We see several major contributions of our findings. First, we enhance the literature in international management in several ways. The "network school" in international management has so far concentrated its effort on analyzing the consequences of network membership for the timing, speed and sequences of firms internationalization strategy (see Johanson and Vahlne 2003; Blomstermo et al. 2004b). Within the emerging field of international entrepreneurship, the network approach has mainly analyzed the effects of network membership for firms survival, growth and innovativeness (see Zahra and George 2005; Zahra et al. 2003; Oviatt and McDougall 1999). In both approaches, however, research has been mute with questions regarding the position of firms within their networks, and more specific factors that determine changes in that position. For managers it seems of utmost importance to gain knowledge on dynamic network processes, such as changes in network size, cohesion, teachability or coreness. Knowing what factors determine network position is a central building block for a strategic network management. We thereby close an important gap in both streams of research.

Second, this paper contributes to research in strategic management, specifically to research on alliances and networks. By addressing the consequences of a firm's alliance experience for its network position we address the broader issue of alliance capabilities in an international setting. Considering prior research pointing to performance consequences of network centrality (Stuart 2000; Uzzi 1996; Ahuja 2000; Anand and Khanna 2000) we furthermore shed light on the underlying causality chain between international alliance experience and performance.

Finally, only sparse research has examined knowledge-based factors that determine a firm's position within its research network (Powell et al. 2005). We therefore apply and extend the knowledge-based theory of the firm to explain firm's cooperative efforts in a knowledge intensive, dynamic context such as the biotechnology industry.

DOI 10.1007/s11575-010-0035-2

Published online: 18.05.2010


Acs, Z. J., & Audretsch, D. B. (1990). innovation and small firms. Cambridge: The MIT Press.

Ahuja, G. (2000). Collaboration networks, structural holes and innovation: A longitudinal study. Administrative Science Quarterly, 45(3), 425-455.

Al-Laham, A., & Amburgey, T. L. (2005). Knowledge sourcing in foreign direct investments: An empirical examination of target profiles. Management International Review, 45(3), 1-29.

Al-Laham, A., & Souitaris, V. (2008). Network embeddedness and new venture internationalization. Analyzing international linkages in the German biotech industry. Journal of Business Venturing, 23(5), 567-586.

Almeida, P., Song, J., & Grant, R.M. (2002). Are firms superior to alliances and markets? An empirical test of cross-board alliances, Organization Science, 13(2), 147.

Alter, C., & Hage, J. (1993) Organizations working together. Newbury, PA: Sage. Anand, B. T., & Khanna, T. (2000). Do firms learn how to create value? The case of alliance. Strategic Management Journal, (Special Issue March), 295 315.

Andersson, S., & Wictor, I. (2003). Innovative internationalisation in new firms: Born globals--the Swedish Case. Journal of international Entrepreneurship, 1(3), 249-275.

Barkema, H. G., & Vermeulen, F. (1998). International expansion through start-up or acquisition: A learning perspective. Academy of Management Journal, 41(1), 7-26.

Bettis, R. A., & Prahalad C. K. (1995). The dominant logic: Retrospective and extension. Strategic Management Journal, 16(1), 5-14.

Blomstermo, A., Eriksson, K., & Deo Sharma, D. (2004a). Domestic activity and knowledge development in the internationalization process of firms. Journal of International Entrepreneurship 2(3), 239-258.

Blomstermo, A., Eriksson, K., Lindstrand, A., & Deo Sharma, D. (2004b). The perceived usefulness of network experiential knowledge in the internationalizing firm. Journal of International Management, 10(3), 355-373.

Bonacich, P. (1987). Power and centrality: The family of measures. American Journal of Sociology 92(5), 1170-1183.

Butt, R. S. (2000). The network structure of social capital. Research in Organizational Behavior, 22, (edited by Robert I. Sutton and Barry M. Staw. Elsevier Science).

Casper, S. (2000). Institutional adaptiveness, technology policy, and the diffusion of new business models: The case of German biotechnology. Organization Studies, 21(5), 887-914.

Chetty, S., & Blankenburg-Holm, D. (2000). Internationalization of small to medium-sized manufacturing firms: A network approach. International Business Review, 9(1), 77-93.

Chetty, S. K., & Wilson, H. I. M. (2003). Collaborating with competitors to acquire resources. International Business Review, 12(1), 61-81.

Cohen, W. M., & Levinthal D. A. (1990). Absorptive capacity: A new perspective on learning and innovation. Administrative Science Quarterly, 35(1), 128-152.

Coombs, J. E., Mudambi, R., & Deeds, D .L. (2006). An examination of the investments in U.S. biotechnology firms by foreign and domestic partners. Journal of Business Venturing, 21(4), 405-428.

DeBresson, C., & Amesse, F. (1991). Networks of innovators: A review and introduction to the issue. Research Policy, 20(5), 363 379.

DeCarolis, D. M., & Deeds, D. L. (1999). The impact of stocks and flows of organizational knowledge on firm performance: An empirical investigation of the biotechnology industry. Strategic Management Journal, 20(10), 953-969.

Dosi, G., Freeman, C., Nelson, R., Silverberg, G., & Soete, L. (1988). Technical change and economic theory. London: Continuum International Publishing.

Dyer, J. H., & Singh, H. (1998). The relational view: Cooperative strategy and sources of interorganizational competitive advantage. Academy of Management Review, 23(4), 660-679.

Erramilli, M. K. (1991). The experience factor in foreign market entry behaviour of service firms. Journal of International Business Studies, 22(3), 479-501.

Ernst, D. (2000). Inter-organizational knowledge outsourcing: What permits small Taiwanese firms to compete in the computer industry. Asia Pacific Journal of Management, 17(2), 223-255.

Ernst, & Young. (Ed.). (1998). Aufbruchstimmung. Erster Deutscher Biotechnologie Report, 1998.

Ernst, & Young. (Ed.). (2000). Granderzeit. Zweiter Deutscher Biotechnologie Report, 2000.

Ernst, & Young. (Ed.). (2002). Neue Chancen. Deutscher Biotechnologie Report, 2002.

Ernst, & Young. (Ed.). (2003). Zeit der Bewahrung. Deutscher Biotechnologie Report, 2003.

Frank, R. H. (1985). Choosing the right pond. Human behavior and the quest for status. Oxford: Oxford University Press.

George, G., Zahra, S., & Robley W. D. (2002). The effects of business-university alliances on innovative output and financial performance: A study of publicly traded biotechnology companies. Journal of Business Venturing, 17(6), 577-609.

Gnyawali, D. R., & Madhaven, R. (2001). Cooperative networks and competitive dynamics: A structural embeddedness perspective. Academy of Management Review, 26(3), 421-445.

Grandori, A., & Soda, G. (1995). Inter-firm networks: Antecedents, mechanisms and forms. Organization Studies, 16(2), 184-214.

Granovetter M. (1985). Economic action and social structure: The problem of embeddedness. American Journal of Sociology, 91(3), 481-510.

Gulati, R. (1995). Social structure and alliance formation patterns: a longitudinal analysis. Administrative Science Quarterly, 40(4), 619-652.

Gulati, R. (1999). Network location and learning: The influence of network resources and firm capabilities on alliance formation. Strategic Management Journal, 20(5), 397-420.

Gulati, R., & Gargiulo, M. (1999). Where do interorganizational networks come from? American Journal of Sociology, 104(5), 1439-1493.

Gulati, R., & Higgins, M. C. (2003). Which ties matter when? The contingent effects of interorganizational partnerships on IPO success. Strategic Management Journal, 24(2), 127-144.

Hagedoorn, J. (2002). Inter-firm R&D partnerships: An overview of major trends and patterns since 1960. Reseach Policy, 31(4), 477-492.

Hagedoorn, J., & Schakenraad, J. (1994). The effect of strategic technology alliances on company performance. SMJ, 15(4), 291-310.

Hagedoorn, J., Link, A. N., & Vonortas, N. S. (2000). Research partnership. Research Policy, 29(4/5), 567-586.

Hansen, M. T. (1999). The search-transfer problem: The role of weak ties in sharing knowledge across organization subteams. Administrative Science Quarterly, 44(1), 82-111.

Hansen, M. T. (2002). Knowledge networks: Explaining effective knowledge sharing in multiunit companies. Organization Science, 13(3), 232-248.

Hansen, M. T., Nohria, N., & Tierney, T. (1999). What's your strategy for managing knowledge? Harvard Business Review, 77(2), 106-116.

Henderson, R., & Cockburn, I. (1994). Measuring competence? Exploring firm effects in pharmaceutical research. Strategic Management Journal, 15(S1), 63-84.

Huber, G. P. (1991). Organizational learning: The contributing processes and literatures. Organization Sciences, 2(Special Issue), 88-115.

Iansiti, M., & Clark, K. B. (1994). Integration and dynamic capability: Evidence from product development in automobiles and mainframe computers. Industrial and Corporate Change, 3(3), 557-605.

Johannisson, B., & Monsted, M. (1997). Contextualizing entrepreneurial networking--the case of Scandinavia. International Studies of Management and Organization, 27(4), 297-312.

Johanson, J., & Vahlne, J.-E. (2003). Business relationship learning and commitment in the internationalization process. Journal of International Entrepreneurship, 1(1), 83-101.

Kogut, B., & Zander, U. (1992). Knowledge of the firm, combinative capabilities and the replication of technology. Organization Science, 3(3), 383-397.

Kogut, B. (2000). The network as knowledge: Generative rules and the emergence of structure. Strategic Management Journal, 21(Special Issue March), 405-425.

Kristensen, P. S., & Lund V. A. (2001). Importance of collaboration partners in product development. In A. Plunket, C. Voisin, & B. Bellon (Eds.), The dynamics of industrial collaboration. Cheltenham: Edward Elgar.

Lane, P. J., & Lubatkin, M. (1998). Relative absorptive capacity and interorganizational learning. Strategic Management Journal, 19(5), 461-477.

Lawrence, P. R., & Lorsch, J. W. (1967). Organization and environment: Managing differentiation and integration. Irwin: Homewood IL.

Lewis, K., Lange, D., & Gillis, L. (2005). Transactive memory systems, learning, and learning transfer. Organization Science, 16(6), 581-598.

Liebeskind, J. P. (1996). Knowledge, strategy and the theory of the firm. Strategic Management Journal, 17(Special Issue Winter), 93-107.

Liebeskind, J., Wiersema, M., & Hansen, G. (1992). LBOs, corporate restructuring, and the incentive-intensity hypothesis, financial management. Financial Management Association, 21 (1/Spring), 73-88.

Liebeskind, J. P., Oliver, A., Zucker, L., & Brewer, M. (1996). Social networks, learning and flexibility: Sourcing scientific knowledge in new biotechnology firms. Organization Science, 7(4), 428-443.

March, J. G. (1991). Exploration and exploitation in organizational learning. Organization Science, 2(1), 71-87.

Mayer, K. J., & Argyres, N. (2004). Learning to contract: Evidence from the personal computer industry. Organization Science, 15(4), 394-410.

Mezias, S. J., & Glynn, M. A. (1993). The three faces of corporate renewal: Institution, revolution, and evolution. Strategic Management Journal, 14(2), 77-101.

Mian, S. A. (1997). Assessing and managing the university technology business incubator: An integrative framework. Journal of Business Venturing, 12(4), 251-285.

Miller, D., & Chert, M.-J. (1996). The simplicity of competitive repertoires: An empirical analysis. Strategic Management Journal, 17(6), 419-439.

Murmann, J. R (2003). The coevolution of industries and national institutions: Theory and evidence. Paper presented at the Annual Academy of Management Meeting, Seattle 2003.

Nahapiet J., & Ghoshal S. (1998). Social capital, intellectual capital, and organizational advantage. Academy. Management Review, 23(2), 242-266.

OECD. (2002). Dynamising National Innovation Systems. Paris: OECD.

Oviatt, B. M., & McDougall, P. P. (1999). Accelerated internationalization: Why are new and small ventures internationalizing in greater numbers and with increasing speed? In R. Wright (Eds.), Global Strategic Management. Stamford, CT: JAI Press.

Perkins, S. W., & Rao C. R. (1990). The role of experience in information use and decision making by marketing managers. Journal of Marketing Research, 27(1), 1-10.

Podolny, J. (1993). A status-based model of market competition. American Journal of Sociology, 98(4), 829-872.

Podolny, J. (1994). Market uncertainty and the social character of economic exchange. Administrative Science Quarterly, 39(3), 458-470.

Podolny, J. M., & Stuart, T. E. (1995). A role-based ecology of technical change. American Journal of Sociology, 100(5), 1224-1260.

Powell, W. W., & Brantley, P. (1992). Competitive cooperation in biotechnology: Learning through networks? In N. Nohria, & R. Eccles (Eds.), Networks and organizations. Boston: Harvard University Press.

Powell, W. W., Koput, K. W., & Smith-Doerr, L. (1996). Interorganizational collaboration and the locus of innovation: networks of learning in biotechnology. Administrative Science Quarterly, 41(1), 116-145.

Powell, W. W., White, D. R., Koput, K. W., & Owen-Smith, J. (2005). Network dynamics and field evolution: The growth of interorganizational collaboration in the life sciences. American Journal of Sociology, 110(4), 901-975.

Prashantam, S. (2005). Towards a knowledge-based conceptualization of internationalization. Journal of International Entrepreneurship, 3(1), 37-52.

Rao, R. S., Chandy, R. K., & Prabhu J. C. (2008). The fruits of legitimacy: Why some new ventures gain more from innovation than others. Journal of Marketing, 72(4), 58-75.

Ritter, T., & Gemunden, H. G. (2003). Network competence: Its impact on innovation success and its antecedents. Journal of Business Research, 56(9), 2003, pp. 745 755.

Rothaermel, F. T., & Deeds, D. L. (2006). Alliance type, alliance experience and alliance management capability in high technology ventures. Journal of Business Venturing, 21(4), 429-460.

Rowley, T. J., & Baum, J. A. C. (2002). The dynamics of network moves and network strategies. Paper presented at the Academy of Management Meeting, Denver CO August.

Sanchez, R., & Heene, A. (Eds.). (1997). Strategic learning and knowledge management. Chichester: Wiley.

Shan, W. (1990). An empirical analysis of organizational strategies by entrepreneurial high-technology firms. Strategic Management Journal, 11(2), 129-139.

Shan, W., Walker, G., & Kogut, B. (1994). Interfirm cooperation and startup innovation in the biotechnology industry. Strategic Management Journal, 15(5), 387-394.

Shipilov, A. (2003). Should you bank on your network? Relational and positional embededdness in the making of financial capital. Working Paper, Joseph L. Rotman School of Management University of Toronto.

Soh, R H., Mahmood, I. R, & Mitchell, W. (2004). Dynamic inducements in R&D investment: Market signals and network locations. Academy of Management Journal, 47(6), 907-917.

Spender, J. C., & Grant, R. M. (1996). Knowledge and the firm: Overview. Strategic Management Journal, 17(Special issue Winter), 5-9.

Spender, J. C. (1996). Making knowledge the basis of a dynamic theory of the firm. Strategic Management Journal, 17(Special Issue Winter), 45-63.

Stuart, T., & Podolny J. (1996). Local search and the evolution of technological capabilities. Strategic Management Journal, 17(S1), 21-38.

Stuart, T. E. (2000). lnterorganizational alliances and the performance of firms: A study of growth and innovation rates in high-technology industry. Strategic Management Journal, 21(8), 791-811.

Stuart, T. E., Hoang, H., & Hybels, R. C. (1999). Interorganizational endorsements and the performance of entrepreneurial ventures. Administrative Science Quarterly, 44(2), 315-349.

Teece, D. J., Pisano, G., & Shuen, A. (1997). Dynamic capabilities and strategic management. Strategic Management Journal, 18(7), 509-533.

Tidd, J., Bessant, J., & Pavitt, K. (1997). Managing innovation: Integrating technological, market and organizational change. Chichester: Wiley.

Tsai, W. (2001). Knowledge transfer in intraorganizational networks: Effects of network position and absorptive capacity on business unit innovation and performance. Academy of Management Journal, 44(5), 996-1004.

Uzzi, B. (1996). The sources and consequences of embeddedness for economic performance of organizations. American Sociological Review, 61(4), 674-698.

Uzzi, B. (1997). Social structure and competition in interfirm networks: The paradox of embeddedness. Administrative Science Quarterly, 42(1), 35-67.

Uzzi, B. (1999) Social relations and networks in the making of financial capital. American Sociological Review, 64(4), 481-505.

Uzzi, B., & Lancaster, R. (2003). The role of relationships in interfirm knowledge transfer and learning, the case of corporate debt markets. Management Science, 49(4), 383-399. von Hippel, E. (1988). The sources of innovation. New York: Oxford University Press.

Walker, G., Kogut, B., & Shan, W. (1997). Social capital, structural holes and the formation of an industry network. Organization Science, 8(2), 109-125.

Wassermann, S., & Faust, K. (1994). Social network analysis. Methods and applications. Cambridge: University Press.

Wong, P. K., & He, Z. L. (2003). Local embeddedness, global networking, and the innovation performance of firms. Paper presented at the Annual Academy of Management Meeting, Seattle.

Zahra, S., Matherne, B., & Carleton, J. M. (2003). Technological resource leveraging and the internationalization of new ventures. Journal of International Entrepreneurship, 1(2), 163-186.

Zahra, S., & George, G. (2005). International entrepreneurship: The current status of the field and future research agenda. In M. Hitt, D. Ireland, M. Camp, & D. Sexton (Eds.), Strategic entrepreneurship. (S. 255-288). Oxford: Wiley.

Zucker, L., Darby, M., & Brewer, M. (1998). Intellectual human capital and the birth of U.S. biotechnology enterprises. American Economic Review, 88(1), 290-306.

Prof. A. Al-Laham ([mail]) Chair for Strategic and International Management, Management Area, University of Mannheim, Mannheim, Germany e-mail:

Prof. T. L. Amburgey Department of Strategic Management, Joseph L. Rotman School of Management, University of Toronto, Toronto, Canada
Table 1: Distribution of
international alliance partners
of German biotech
firms according to country
of origin (cumulative
count, 1995-2006)


United States 75
United Kingdom 22
Switzerland 20
Japan 8
France 6
Canada 6
Denmark 6
Sweden 6
Belgium 6
Netherlands 4
Austria 4
Italy 4
Israel 2
Ireland 2
Spain 2
Poland 2
Czechoslovakia 2
South Korea 1
Argentina 1
Norway 1
South Africa 1
Total 181

Table 2: Descriptive statistics for variables of analysis

Variables Mean S.D. Lag. 1.

Lag. dep 0.5719 5.4709 1.00

1. Age 1568.131 1323.33 0.0249 1.00

2. Size/No of 28.874 69.2090 0.1834 0.1764 *

3. Research 1.6989 2.9685 0.0385 0.1503 *

4. Techn. 0.03438 0.1822 -0.0197 0.0524 *

5. Prior Patents 10.904 13.2726 0.0097 0.3136 *

6. Regional 47.9288 25.2851 0.0961 * 0.0564 *

7. Regional 0.3261 0.0816 -0.0210 * -0.0437

8. Prior Dom. 0.5158 1.7539 0.4761 * 0.1076 *

9. Prior Int. 0.2111 1.1129 0.5454 * 0.1379 *

10. Prior Priv. 0.5457 2.560 0.6630 * 0.1282 *

11. Prior Pub. 0.1711 0.6433 0.3331 * 0.1214 *

12. Prior Mult 0.0899 0.4023 0.2880 * 0.0860 *

Variables 2. 3. 4. 5.

Lag. dep

1. Age

2. Size/No of 1.00

3. Research 0.1195 * 1.00

4. Techn. 0.0116 0.0507 * 1.00

5. Prior Patents 0.0369 0.0338 0.1841 1.00

6. Regional 0.0500 * 0.0443 0.0104 -0.0446 *

7. Regional -0.0767 * -0.0411 0.0567 * 0.2042 *

8. Prior Dom. 0.2206 * 0.1332 * -0.0372 0.0451 *

9. Prior Int. 0.3125 * 0.1287 * -0.0059 0.0471 *

10. Prior Priv. 0.3174 * 0.1226 * -0.0214 0.0458 *

11. Prior Pub. 0.2013 * 0.1116 * -0.0078 0.0579 *

12. Prior Mult 0.1644 * 0.1100 * -0.0204 0.0281

Variables 6. 7. 8. 9.

Lag. dep

1. Age

2. Size/No of

3. Research

4. Techn.

5. Prior Patents

6. Regional 1.00

7. Regional -0.3941 * 1.00

8. Prior Dom. 0.1723 * -0.0579 * 1.00

9. Prior Int. 0.1103 * -0.0586 * 0.6104 * 1.00

10. Prior Priv. 0.1448 * -0.0472 * 0.6699 * 0.8794 *

11. Prior Pub. 0.1449 * -0.0578 * 0.7632 * 0.5451 *

12. Prior Mult 0.1041 * -0.0655 * 0.8008 * 0.4755 *

Variables 10. 11. 12.

Lag. dep

1. Age

2. Size/No of

3. Research

4. Techn.

5. Prior Patents

6. Regional

7. Regional

8. Prior Dom.

9. Prior Int.

10. Prior Priv. 1.00

11. Prior Pub. 0.4959 * 1.00

12. Prior Mult 0.4236 * 0.6743 * 1.00

* = significant at p < 0.01 based on 2,501 observations

Table 3: Coefficients of pooled cross-sectional time series analysis

Variables Model 1 Model 2 Model 3

Lag dep. 0.3277 *** 0.2082 *** 0.1617 ***
 (0.0190) (0.0204) (0.0202)
Age 0.00004 -0.0002 -0.0001
 (0.00017) (0.0001) (0.0002)
Size (No. of employees) 0.0071 *** 0.0022 -0.0004
 (0.0019) (0.0019) (0.0018)
Research breadth -0.0166 -0.1089 * 0.0053
 (0.0541) (0.0522) (0.0516)
Technological -0.1268 0.2666 -0.0694
 sophistication (1.2994) (1.242) (1.203)
Number of prior patents 0.1329 -0.0980 -0.1181 **
 (0.0539) (0.0545) (0.0528)
Regional density -0.0012 -0.0042 -0.0005
 (0.0084) (0.0081) (0.0078)
Regional Herfindahl 1.3093 1.147 1.737
 (2.0739) (1.982) (1.919)
Hl Prior R&D Alliances 1.252 *** 0.2076
 (count) (0.0960) (0.1326)
H2 Prior R&D Alliances 0.0949 ***
 X intern. Partner (0.0086)
H3 Prior R&D All. X
 country div.
H4 Prior R&D Alliances
 X private partner
H5 Prior R&D All. X
 multiple partner
Prior intern. partners
Prior multiple partners
Prior private partners
Country diversity
Number of observations 2,501 2,501 2,501
Number of firms 717 717 717
F-Value 42.67 60.47 70.23
 (8.1776) (9.1775) (10.1774)
P Value p<0.00001 p<0.00001 p<0.00001

Variables Model 4 Model 5 Model 6

Lag dep. 0.1740 *** 0.1982 *** 0.1302 ***
 (0.0196) (0.0177) (0.0153)
Age -0.0001 -0.0001 -0.0001
 (0.0001) (0.0001) (0.0001)
Size (No. of employees) -0.0018 -0.0021 -0.0019
 (0.0017) (0.0016) (0.0014)
Research breadth 0.0345 0.0218 -0.0663
 (0.05043) (0.0452) (0.0386)
Technological -0.0279 -0.2515 -0.2746
 sophistication (1.171) (1.052) (0.8960)
Number of prior patents -0.0828 -0.1180 ** -0.0572
 (0.0515) (0.0463) (0.0395)
Regional density 0.0006 0.0078 0.0046
 (0.0076) (0.0068) (0.0058)
Regional Herfindahl 1.523 2.156 3.330 **
 (1.869) (1.679) (1.431)
Hl Prior R&D Alliances 1.080 *** 1.0905 *** 2.039 ***
 (count) (0.1569) (0.1409) (0.1254)
H2 Prior R&D Alliances 0.1735 *** 0.4702 *** 0.5214 ***
 X intern. Partner (0.1569) (0.0177) (0.0153)
H3 Prior R&D All. X -0.4496 *** -0.2534 *** -0.2550 ***
 country div. (0.0459) (0.0423) (0.0360)
H4 Prior R&D Alliances -0.1624 *** -0.1612 ***
 X private partner (0.0078) (0.0007)
H5 Prior R&D All. X -1.158 ***
 multiple partner (0.0446)
Prior intern. partners
Prior multiple partners
Prior private partners
Country diversity
Number of observations 2,501 2,501 2,501
Number of firms 717 717 717
F-Value 75.98 121.71 206.91
 (11.1773) (12.1772) (13.1771)
P Value p<0.00001 p<0.00001 p<0.00001

Variables Model 7

Lag dep. 0.10352 ***
Age -0.00018
Size (No. of employees) -0.0014023
Research breadth -0.04270
Technological -0.03799
 sophistication (0.85980)
Number of prior patents 0.00197
Regional density 0.00369
Regional Herfindahl 2.4834
Hl Prior R&D Alliances 2.115 ***
 (count) (0.18838)
H2 Prior R&D Alliances 0.4532 ***
 X intern. Partner (0.02375)
H3 Prior R&D All. X 0.1063 *
 country div. (0.0520)
H4 Prior R&D Alliances -0.1782 ***
 X private partner (0.01141)
H5 Prior R&D All. X -1.414 ***
 multiple partner (0.05648)
Prior intern. partners 1.755 ***
 (count) (0.42038)
Prior multiple partners 3.161 ***
 (count) (0.46028)
Prior private partners -0.0001
 (count) (0.1849)
Country diversity -5.023 ***
 (count) (0.53974)
Number of observations 2,501
Number of firms 717
F-Value 181.38
P Value p<0.00001

Note: Significant at ([dagger]) p < 0.10; * p < 0.05;
** p < 0.01; *** p< 0.001
COPYRIGHT 2010 Gabler Verlag
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2010 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Al-Laham, Andreas; Amburgey, Terry L.
Publication:Management International Review
Date:May 1, 2010
Previous Article:Dynamics of experience, environment and MNE ownership strategy.
Next Article:Is market orientation affected by the size and diversity of customer networks?

Terms of use | Privacy policy | Copyright © 2021 Farlex, Inc. | Feedback | For webmasters |