Innovation through Alliance Portfolios: The Role of Centrality, International
Alliances, and R&D Alliances
Master Thesis
University of Groningen Faculty of Economics and Business MSc BA Strategic Innovation Management
17. January 2019
Hendrik Opitz S2712598 h.opitz@student.rug.nl
Supervisor: Dr. Pedro de Faria Co-assessor: Dr. Florian Noseleit
Abstract
Table of Contents
1.Introduction ...4
2. Literature Review ...6
2.1 Strategic Alliances and Innovation ... 6
2.2 Alliance Portfolio Characteristics ... 7
2.3 Firm Alliance Portfolio Centrality and Innovation ... 10
2.4 The Moderating Effect of International Alliances ... 11
2.5 The Moderating Effect of R&D Alliances... 16
3. Methodology ...19 3.1 Data Collection ... 19 3.2 Measures ... 20 3.3 Analytical Method ... 22 4. Results...23 4. 1 Descriptive Statistics... 23 4.2 Hypotheses Testing ... 25 4.3 Robustness Check ... 27 5. Discussion ...29 5.1 Research Implications... 29 5.2 Managerial Implications ... 32
5.3 Limitations and Future Research ... 32
6. Conclusion ...33
1.Introduction
The innovative capabilities of a company are widely accepted to be one of the key drivers in achieving business success (Rogers, 2004). Innovative capabilities refer to the specific competencies and expertise of a firm regarding the development and the introduction of new products and processes (Haagedoorn & Duysters, 2002). Organizations deploy a considerable amount of resources to establish an innovation culture and introduce new products to the market. Inputs for new ideas that a company aims to further develop can have several sources. Not only formal research or the creativity of the own workforce can lead to innovations, but also other external sources can be used to establish a steady flow of ideas (Rogers, 2004). Strategic alliances are one of the prior mentioned sources and have become popular organizational instruments in the 1990s (Haagedoorn & Duysters, 2002). Not only do alliances ensure access to new knowledge and capabilities, but they also help organizations to enter into new markets and increase their market power (Haagedoorn & Duysters, 2002).
The relational view further focuses on the processes and routines between alliance partners to gain a deeper understanding of the sources of competitive advantages (Dyer & Signh, 1998). Thereby, it argues that critical resources of an organization may not be derived internally but are embedded in routines and processes between a pair of firms. Each alliance partner possesses unique, complementary resource endowments. By leveraging the complementary assets of the partner in combination with its own assets, the firms are able to develop new products or achieve superior performance (Dyer & Signh, 1998). Since the firms are dependent on the assets of their partners, a single firm in isolation could not produce the same outputs. Thereby, a strategic alliance can be the cause of achieving a sustained competitive advantage that others cannot imitate (Dyer & Signh, 1998).
A major criticism of a substantial number of former studies is that they focus on alliances as a single event, rather than emphasizing on the synergies and tradeoffs that a whole portfolio of alliances offers (George, et al., 2001). Characteristics of the portfolio of alliances can influence the benefits that a firm enjoys, concerning both financial and innovation performance (Nooteboom, 1999). Social network theory aims to solve this issue by focusing on the whole network of alliances instead of examining single alliances in isolation (Lavie, 2007).
whether it has a central or a less central position in the portfolio (Lavie, 2007; Macauly, et al., 2018). Occupying a central position grants the firm with many direct ties to other members of the network, which will make it a valuable partner to ally with since it can gain access to a variety of knowledge, skills, and resources in order to enhance its own innovative performance (Gnyawali & Madhavan, 2001; Tsai, 2001; Macauly, et al., 2018). Former research paid a significant amount of attention towards the effects of alliance portfolio centrality on firm performance and innovativeness. This research adds to the former literature by further testing the impact of alliance network centrality on the innovation performance. Specifically, this study explores whether a positive or negative coherence exists between alliance portfolio centrality and innovation performance. Thus, the following research question is addressed:
Does a higher level of centrality within an alliance portfolio enhance the innovation performance?
Moreover, this research extends the above-mentioned literature by introducing alliance portfolio characteristics into the alliance portfolio centrality literature. Past research was concerned with the direct effects of alliance portfolio characteristics such as international or R&D alliances on the innovative performance of the firm (Wassmer, 2010; Lavie & Miller, 2008; Sampson, 2004). This study further investigates how the relationship between alliance portfolio centrality and the innovation performance is influenced by the share of international alliances and R&D alliances respectively, to test the following two sub-questions:
1. How does a high share of international alliances influence the relationship between alliance portfolio centrality and innovation performance?
2. How does a high share of R&D alliances influence the relationship between alliance portfolio centrality and innovation performance?
years after the formation of a partnership, making it a count data dependent variable. Hence, I performed a negative binomial regression to test the hypotheses.
The results of this study offer new insights into the centrality literature and the composition of the alliance portfolio. A firm that occupies a central position in its alliance portfolio indeed achieves a better innovation performance than a firm that holds a less central position. Having a high share of international alliances in the alliance portfolio, however, negatively influences the focal relationship. The hypothesized positive effect of having a high share of R&D alliances in the alliance portfolio on the same relationship is not supported by the data. Eventually, this research concludes by giving managerial implications, limitations of this study and possible directions for future research.
2. Literature Review
2.1 Strategic Alliances and Innovation
In order to survive in the fast-paced, globalized marketplace, companies face the need to be innovative, as well as respond quickly to the innovations of competitors (Narula & Haagedoorn, 1999). But even the largest companies tend to struggle in sustaining high levels of productivity over a longer time span and strive to access external knowledge to further develop innovations (Haeussler & Higgins, 2014). A popular vehicle to obtain knowledge that resides outside of the boundaries of the firm is a cooperative relationship like strategic alliances (Grant & Baden-Fuller, 2003).
A strategic alliance consists of at least two legally independent organizations that exchange or share resources and collaborate in the development of technologies, services or products (Todeva & Knoke, 2005; Gulati, 1999). Activities that are performed by a strategic alliance can take diverse forms, e.g., technical collaboration, shared new product development, shared manufacturing arrangements, supplier-buyer partnerships, cross-selling arrangements, and franchising (Grant & Baden-Fuller, 2004).
possessed by the alliance partner. These resources can be either concrete, as financial resources and specific skills (Hamel, et al., 1989), or more abstract, such as market power and legitimacy (Baum & Oliver, 1991). By leveraging the recombination potential of the resources of the alliance partners, firms are more likely to improve their strategic position and overcome the vulnerability they are facing (Eisenhardt & Schoonhoven, 1996). If as a result of the collaboration, the vulnerability of the organizations’ decreases, a shift in the perceived complementarity of resources and the associated payoffs to the alliance partners is likely to occur (Hamilton & Singh, 1991). Hence, partners assess the value of each other in the alliance. If the value of one partner is significantly reduced, the alliance is likely to be either terminated or redefined.
Relating strategic alliances to the performance of a firm, Dyer and Singh (1998) argue that an alliance is likely to enhance overall firm performance. Partners contributions to the partnership can take the form of idiosyncratic knowledge, assets, and resources or capabilities the partners aim to exchange, combine, or invest in. Through the joint idiosyncratic resource contribution of the partners, unique specialized assets are developed. Furthermore, knowledge-sharing routines help the partners to exploit new sources of ideas and information by drawing on the specialized knowledge of each other (Dyer & Signh, 1998). Hence, the formation of an alliance helps the partners to achieve a sustained competitive advantage. As a result, the partners generate supernormal profits, that cannot be produced by a firm in isolation. George et al. (2001) agree with that view arguing that through the formation of an alliance, firms can gain access to new resources, like knowledge or assets that have been developed external to the firm, which spurs new product development and innovation. Ultimately, this enhances the profitability of the company, as well as its growth.
2.2 Alliance Portfolio Characteristics
size of the portfolio (Lahiri & Narayanan, 2013) or its characteristics such as the diversity of partners within the portfolio (Oerlemans, et al., 2013).
Strategic alliances offer opportunities to access complementary assets that reside outside of the boundaries of the firm (Teece, 1982). By leveraging these assets, organizations can fully exploit their firm-specific capabilities and develop new ones, thereby improving its innovative capabilities. Following this logic, with increasing size of the alliance portfolio, also the number of external assets available to the focal firm will increase (Lahiri & Narayanan, 2013). By combining external assets such as knowledge, with internal resources, firms are likely to enhance their innovation performance (Ahuja, 2000).
With the increasing size of the alliance portfolio, the organization is also prone to challenges. First of all, the search, identification, and transfer of knowledge will be aggravated (Duysters, et al., 1999). Secondly, prior literature pointed out that shared routines between the partners facilitate the transfer of knowledge (Phene, et al., 2005). These routines develop during the time of the partnership. When utilizing the resources across a portfolio of alliances, organizations have to adapt their routines to several partners simultaneously to develop products and processes. With the increasing size of the alliance portfolio, firms are unable to do so and impede the transfer of knowledge and organizational learning, ultimately inhibiting innovation performance.
Another stream of research within the alliance portfolio literature is concerned with the degree of diversity of the firm’s network of alliances. Diversity in the alliance portfolio is characterized by the degree of variance in partners, the governance structures of the alliance, and the functional purposes (Jiang, et al., 2010). Having a more diverse set of alliance partners grants the firm with access to more varied information and capabilities and hence provides more opportunities for learning while reducing the risk of intra-alliance rivalry (Baum, et al., 2000). Partners can be for example universities and research institutions, suppliers, buyers, or even competitors and hence possess different types of knowledge (Oerlemans, et al., 2013).
performance. Thus, low levels of diversity keep the innovation activities of the firm rather low. As the level of diversity increases to a moderate level, the firm is able to access a wider variety of complementary assets. The company profits the most from a moderate amount of diverse knowledge inflow. At moderate levels, it is able to make use of varied external knowledge to develop new ideas and innovations, while it is still able to deal with diversity issues of the portfolio (Oerlemans, et al., 2013). Hence, a moderate alliance portfolio diversity is beneficial for innovation. If the degree of diversity is further increasing, the organization is exposed to too many ideas. Managers might be overwhelmed by the diverse inflow of ideas that they do not know which ideas to choose and how to manage them (de Leeuw, et al., 2014). Due to the high number of ideas, only a few are taken seriously and get the attention that is needed to be successfully developed and implemented. Moreover, ideas or resources could reach the firm at the wrong time or place to be fully exploited (Koput, 1997). Lastly, due to the diverse range of partners and differing governance structures, the costs of communication, coordination, and monitoring as well as the likelihood of opportunism, which can result in unintended knowledge spillovers increase (Oerlemans, et al., 2013). These reasons ultimately impede the innovation performance of the firm.
The concepts of alliance portfolio size and diversity help to get a general understanding of the importance of the composition of an alliance portfolio and are meaningful contributions to the former literature. However, they are already well developed. Another important stream of literature is the social network theory and particularly the position of a firm in its network. Social network theory focuses on the enduring patterns of relationships between interacting organizations and thereby overcomes the alliance independence assumption (Lavie, 2007). Thus, it centers the network of alliances instead of single alliances in isolation. Actors of a network are interdependent, and their membership grants them with opportunities such as social capital (Lavie, 2007). Social capital refers to certain benefits that derive from closed network structures through either direct or indirect ties to other members (Portes, 1998). Direct ties provide direct links to other parties of the network, whereas indirect ties link firms that are part of the same network, but not directly connected (Coleman, 1988).
2.3 Firm Alliance Portfolio Centrality and Innovation
Studies about the structural embeddedness took the position of the firm in its alliance portfolio into consideration, specifically the centrality (Lavie, 2007). Alliance portfolio centrality depends on two conditions. First of all, it relies on the number of linkages that the firm possesses to its partners (Freeman, 1978). Additionally, also the quality of their partner firms is an important measure. Freeman (1987) defines the quality of the linkages as how many ties the potential allies have to other firms. Even though quality might also mean other things, this study follows Freeman’s definition. Accordingly, a high number of linkages equals a high quality of the potential partner. Hence, if a firm has many linkages to partner firms that also have many ties to their respective partners, it is perceived to be more central than a firm that has either a lower amount of links or many links but of lower quality (Macauly, et al., 2018). Differently speaking, a centrally positioned firm has many ties to its alliance partners, who are themselves connected to many other firms.
Organizations actively seek connections to other firms to profit from their knowledge (Smith-Doerr & Powell, 2003). Knowledge bases, practices and capabilities differ across organizations. In environments, where knowledge is advancing rapidly, expertise’s are widely dispersed, and uncertainty about the best practices to solve current problems is high. Occupying a central position in the alliance portfolio links the focal firm to several partners. Due to that, it can draw on a wide range of external assets such as management skills, technologies, and money from its partners (Gnyawali & Madhavan, 2001). These assets facilitate the generation of new ideas and best practices and are shared among the partners in order to enhance the cost efficiency and advance the firm’s innovation activities (Tsai, 2001). The focal firm is also more likely to overcome problems in the design and manufacturing phase by applying the external knowledge that it can access due to its ties to other companies (Dougherty & Hardy, 1996).
than across networks, the firm should engage in ties to organizations that are themselves not connected to the firm’s existing network (Gilsing, et al., 2008). Such a link will provide the firm with access to entirely new information and entrepreneurial opportunities and guarantees that both partners have access to different information flows.
Another essential factor of alliance portfolio centrality is that a company that engages in many ties is also considered to be a prestigious partner to form an alliance with (Phelps, et al., 2012). Information about a central firm tends to be available and shows a positive sign of its quality to others. Additionally, due to the number of ties, central firms are thought to possess valuable resources that other organizations desire (Macauly, et al., 2018). This good reputation grants a central firm more power and status and can also foreclose partnering opportunities of competing firms (Gnyawali & Madhavan, 2001). Combining this with the early access to valuable information and developments, the overall bargaining power of the central firm increases (Gilsing, et al., 2008).
Innovation is the result of new ideas that have been developed and implemented (Björk & Magnusson, 2009). As mentioned above, recent literature identifies two mechanisms that enable firms that occupy a central position in their alliance portfolio to achieve a higher innovation performance. First of all, by having many ties to the alliance partners, the focal firm can make use of a wider variety of knowledge and therefore ensure a steady flow of ideas. Secondly, information about a central firm is more accessible, making it a more reliable partner and hence increase the reputation of the focal firm. As a result, a firm that has a central position in its alliance portfolio is expected to have a better innovation performance: H1: There is a positive relationship between network centrality and innovation performance.
2.4 The Moderating Effect of International Alliances
Having international alliances in the alliance portfolio can provide the firm with many benefits such as access to resources, assets, capabilities, and information that cannot be accessed from local partners (Wassmer, 2010). Organizations assimilate knowledge that is related to its knowledge base, thus an overlap with its partner’s knowledge base eases the sharing and transfer of knowledge (Cohen & Levinthal, 1989). However, having many international partners in the portfolio can also increase the complexity of the alliance portfolio. In such cases, the firm must invest more to deal with differences that are rooted in the national background, in order to effectively share knowledge and develop idiosyncratic procedures to cooperate with a high number of foreign partners (Lavie & Miller, 2008). Investigating the effects of different degrees of portfolio internationalization on firm performance, Lavie and Miller (2008) found a sigmoid relationship, arguing that at low levels knowledge is not distinctive enough, while at high levels coordination costs are rising. These coordination costs tend to outweigh the benefits of having foreign partners and might end up in mistrust, conflict, lack of commitment and ineffective interactions.
An aspect that the recent literature neglects is how a high share of international alliances will affect the relationship between alliance portfolio centrality and the innovative performance of a firm. Specifically, this study investigates how the benefits of being centrally located in the network are influenced by having a high share of international alliances. Occupying a central position in the alliance portfolio enhances the access to external knowledge as well as the reputation. These two mechanisms help a central organization to achieve a better innovation performance than firms that do not occupy a central position. However, both mechanisms are likely to be affected by having a high share of international alliances, as the following paragraphs discuss.
Hence, more time has to be spent on communication, development of common managerial approaches, and the design of compatible work routines.
Sirmon and Lane (2004) pose that cultures constitute at the national, organizational, and professional level. The national culture is overarching, thus influencing the organizational as well as the professional culture (Sirmon & Lane, 2004). A national culture is constituted by deeply set values that are common to the whole nation (Sirmon & Lane, 2004). It is learned early in life and provides the individuals with a way of thinking of ‘how things ought to be’ and ‘how they ought to be done.’ These values and norms of national culture trickle down and influence the professional and organizational cultures (Sirmon & Lane, 2004).
Next to the national culture, also organizational cultures are likely to influence the collaboration (Pothukuchi, et al., 2002). Differences in the shared values and beliefs between two organizations can lead to a lack of common understanding and hence impede the interpretation of each other’s strategic intent, which is critical in global markets and alliances (Sirmon & Lane, 2004). As a result, effective sharing, combining and leveraging of resources such as knowledge will be inhibited ultimately leading to a decrease in learning and satisfaction with the alliance (Pothukuchi, et al., 2002). Since access to new information is dependent on partners sharing their knowledge, differences in the organizational cultures are likely to impede a steady flow of information reaching the firm.
Also, the professional culture of a company is likely to have an effect on knowledge sharing between companies that are embedded in different national environments. A professional culture is constituted by again shared values, beliefs, and norms that are shared between people in a functionally similar job (Sirmon & Lane, 2004). They are rooted in the occupational training and education that individuals have received and lead to a broader understanding of how their job should be performed (Jordan, 1990). Since educational systems differ per country, also the professional cultures are distinct (Sirmon & Lane, 2004). When in an international alliance two different professional cultures have to collaborate, a lack of shared basic knowledge that is reasoned in their different occupational socialization and the lack of experience to collaborate with individuals from outside their professional culture can severely inhibit the communication (Sirmon & Lane, 2004). Again, this hampers the transfer of knowledge, which is crucial for firms to gain access to new information.
innovation performance. Cultural differences lead to misunderstandings and inefficient collaboration, thereby inhibiting the flow of information. Organizations that occupy a central position in its network of partnerships benefit from access to information since they are well connected in their network due to their many direct ties to other organizations who serve as a source of information (Gnyawali & Madhavan, 2001). Firms that do not have a central position are less well connected and have fewer direct links to their partners. Therefore, they benefit less from access to information than central firms. A high share of international alliances is likely to have a negative effect on the access to information from other organizations. Centrally positioned firms can reach far more information than non-central companies and hence, are more affected by the negative effects of a high share of international alliances. As a result, alliance portfolio centrality and the share of international alliances interact in a way that the positive effects of having better access to information of centrally located firms will be weakened by having a high share of international alliances in the portfolio.
The second mechanism that is expected to be affected by international alliances is the reputation that resides from a central position in the alliance portfolio. When operating across borders, trust between alliance partners plays an important role. According to Gulati (1995) trust can be described as a type of expectation that eases the fear that one of their partners will act opportunistically. Especially for firms facing greater uncertainty surrounding future events and the response of the partner to those futures events, trust plays a crucial role. When initiating an alliance, there will always be some vulnerability. Sources of this vulnerability may be the loss of items that are of value for the focal firm, as personnel, markets, technology, or know-how (Parkhe, 1998). If there is a low degree of trust between alliance partners, the partner is more likely to misappropriate the assets of the focal firm, and as a result, the vulnerability is likely to increase. Usually, more and better information is available about domestic partners, leading to a higher degree of trust when forming a domestic alliance. For international partners it is harder to gather information about their opportunistic behavior, leading to a higher degree of mistrust (Gulati, 1995). Hence, initiating an alliance with a foreign partner can increase the risk of misappropriation of knowledge.
firm does not have a direct connection to a third organization, but is part of the same network, the third organization is likely to be informed about any opportunistic behavior of the focal firm due to the information flowing around in the network. Hence, even if there is no firsthand experience between the partners, information about opportunism become available through social ties (Kwon, et al., 2016). This information can be transferred on an inter-organizational level throughout the network. When a firm hence has a high share of international partners, it may be perceived as a higher risk, since the international partners may not be trustworthy and misappropriate the knowledge of other members of the network. Having many international partners, therefore, could lead to the focal firm being less perceived as a prestigious and reliable partner, but more as a risk of unintended knowledge spillovers. Hence, firms might be less willing to initiate a new alliance with the focal firm due to its negative reputation of being a risk of unintended knowledge spillovers.
Centrally positioned companies will be more affected by a high share of international partnerships than firms that do not occupy a central position. One of their main benefits is the positive reputation that makes them a desired partner with valuable knowledge. Among the many ties to other organizations, information about the focal firm is easily transferable and retrievable by other organizations (Phelps, et al., 2012). Due to that, the negative effects of a high share of international alliances influences firms in a central position. Organizations in a non-central position have less access to external knowledge and are thus perceived to possess fewer valuable knowledge (Macauly, et al., 2018). They have fewer direct ties which further decreases the amount of information that is available about the company. Hence, the negative effect of having many international alliances is less relevant for non-central firms, since less information is transferred via their ties to other companies. As a result, the second mechanism that enables a central firm to enhance its innovation performance, the reputation, is weakened when the firm has a high share of international alliances in its portfolio.
H2: There is a positive relationship between network centrality and innovation
performance and this relationship is negatively moderated by the share of international alliances.
2.5 The Moderating Effect of R&D Alliances
R&D alliances have become a popular vehicle to acquire and leverage technological capabilities (Oxley & Sampson, 2004). R&D refers to the standard research and development activity aimed to increase technical or scientific knowledge as well as applying the gained knowledge to create improved or new processes or products (Haagedoorn, 2002). An R&D alliance can be defined as a formal arrangement between at least two otherwise independent organizations in which R&D is at least part of the cooperation (Faems, 2006). This study tests how the relationship between alliance portfolio centrality and innovation performance is the share of R&D alliances in the alliance portfolio.
In response to competitive pressures, firms increasingly search for knowledge that has been developed external to the firm (Oxley & Sampson, 2004). Making use of inter-firm R&D alliances has become very common in order to access new knowledge, shorten the development time of products, spreading risks and costs of the development, and realize economies of scale and scope in the R&D activities of the firm (Sampson, 2004). Further, Lin et al. (2012) argue that a higher proposition of R&D alliances in the alliance portfolio of a firm signals a higher knowledge acquisition capacity. Due to the mutual sharing of knowledge that comes with an R&D alliance, the firm is able to acquire the knowledge and boost its innovative performance.
This study investigates the indirect effect of having a high share of R&D alliances on the relationship between alliance portfolio centrality and innovation performance. Two mechanisms, namely the flow of information and the firm’s reputation, have been pointed out to enhance the innovation performance of companies that occupy a central position in its alliance portfolio.
while downstream value chain activities take care of distribution, marketing, and customer services (Das, et al., 1998). Thus, firms that aim to learn from their partners are likely to engage in upstream value chain activities since they involve the production and sharing of knowledge. Specifically, firms that participate in an R&D alliance need to interact on a frequent basis and mutually share the knowledge that each partner possesses. These frequent interactions facilitate the learning from each other and allow the partners to observe and also imitate the best practices of each other (Lin, et al., 2012). As a result, there is an increased flow of information between the companies, and the gathered experience and learnings may go beyond the lifetime of the R&D alliance. Oxley and Sampson (2004) agree with that view arguing that it requires many points of contact to bring a joint R&D project through commercialization. On the several points of contact, information will be exchanged which facilitates the flow of information across organizational boundaries.
Firms that take a central position in their network benefit from an enhanced access to information due to their direct ties. Having a high share of R&D alliances further enhances the access since the organization has even more points of contact where information is exchanged. Hence, a centrally positioned firm benefits from high shares of R&D alliances due to an enhanced access to information from their many partners. Opposed to that, firms that occupy a less central position have fewer direct links to their partners. Due to that, their access to information is already limited, also limiting the benefits of having a high share of R&D alliances (Gnyawali & Madhavan, 2001). Thus, they are less affected by having a high share of R&D alliances than central firms. Therefore, a high share of R&D alliances in the portfolio enhances the first mechanism in a way that centrally positioned firms have even better access to information, helping them to achieve a better innovation performance.
can indicate that the firm has already developed valuable knowledge, which makes it more compelling to ally with (Qi, et al., 2015). Consequently, the firm will be perceived as a valuable partner to collaborate with, ultimately, leading to a better reputation. Therefore, having a higher share of R&D alliances in the alliance portfolio will serve as an indicator for valuable knowledge that also other firms aim to possess, thereby enhancing the second mechanism, the reputation of a firm.
Information about the reputation can be transferred through the links that the company has to its partners. These partners also have links to their allies, which thereby become informed about the focal firm even without having a direct tie (Kwon, et al., 2016). Companies that occupy a central position in their network have many direct ties to their partners who broadcast information about the focal firm to their partners. Hence, information about the positive reputation of the focal firm spreads around. Opposed to that, a firm in a less central position has fewer ties to its partners. Hence, it has fewer links that transfer the positive reputation. Thereby, the positive effects of having a high share of R&D alliances in the portfolio are limited and less relevant for non-central companies. Eventually, centrally positioned firms benefit from a spreading reputation that is enhanced by having a high share of R&D alliances, helping them to achieve a better innovation performance.
Summarizing, a high share of R&D alliances is expected to have a positive effect on both mechanisms that have been pointed out above. R&D alliances require a frequent interaction between the partners. This close interaction involves an intended and unintended sharing of knowledge. Due to that, the flow of information will be increased. Furthermore, a higher share of R&D alliances reflects that the organization possesses valuable assets. These assets are also desired by other companies leading to an increase in the reputation of the focal firm as a precious partner that can offer lots of useful assets. Hence, it is expected that a high share of R&D alliances in the overall portfolio has a positive impact on the relationship between alliance portfolio centrality and innovation performance:
H3: There is a positive relationship between network centrality and innovation
Figure 1: Conceptual Model
H2: -
H1: +
H3: +
3. Methodology
This study aims to test the relationship of alliance portfolio centrality on the innovation performance. Thereafter, the moderating effects of a high share of international alliances in the alliance portfolio, as well as a high share of R&D alliances in the alliance portfolio, on the prior mentioned relationship are tested. In order to do so, this section provides an overview of the data collection process, the variables that are used for the analysis, and the analytical method that is performed to test the hypothesis.
3.1 Data Collection
The original dataset that I extended with additional information for this study was created by Schilling (2015) with the aim to explore the effect of technology shocks on the collaboration of firms, as well as the separate influence of both on the innovation performance. The dataset consists of 449 North American firms that were part of the global technology collaboration network between the years 1990 and 2005 and were publicly held for at least three years during this time period. Thereafter, 86 additional firms were added that met the same condition of being publicly held for at least three years between 1990 and 2005 but were not part of the global technology collaboration network. Hence, the sample
yields a total amount of 535 firms. The firms used in the dataset are part of multiple industries, namely transportation equipment, air and space, construction and materials, food and textiles, pharmaceutical, biotechnological and medical, IT, machines and instruments, chemical, and lastly plastics and oil. The dataset provides sufficient information to test the first hypothesis.
In order to test the second and third hypotheses, I collected additional data. Both hypotheses are either concerned about the influence of a high share of international alliances, or R&D alliances on the focal relationship. For that reason, I collected the total number of alliances between the years 1992 and 2005 for every single firm and added it to the dataset. The data was retrieved using the SDC database. I searched for each firm that is covered by the dataset in the column “Participant_Ultimate Parent Name” of the SDC database. This column covers the names of all companies that collaborated together. If alliances for a company were found, I applied the option “Split Alliances” to split all the alliance partners into different rows. Based on that, I counted all the partners that a firm had in every single year. This procedure was repeated for every year of the dataset.
In addition to that, I collected the number of international alliances, as well as the number of R&D partners of the firms to examine the share of international alliances and the share of R&D alliances of the total portfolio respectively. Again, the SDC database provided sufficient information on the number of international and R&D partners. If the column “Cross_Border_Participants” was marked with a “Y”, the alliances were accounted as an international alliance. The column “Research_And_Development_Flag” gave disclosure whether it was an R&D collaboration or not. Eventually, I calculated both shares by dividing the total number of either international or R&D partners by the total number of alliance partners and adding it to the dataset.
3.2 Measures Dependent Variable
that the norms and systems of getting a patent granted differ across regions. In this study, however, only North American firms have been used, eliminating the prior mentioned problems. Past research found that there are time lags between the investments in R&D and the granted patents (Sampson, 2007). Hence, any outcome of collaborative activity might not be visible in the year that the alliance is formed. Since some alliances were also formed at the end of a year, during this analysis, this research assesses the innovation performance of the firm by the number of successful patent applications two years after the alliance was initiated.
Independent Variable
Alliance Portfolio Centrality: Alliance portfolio centrality is measured using the distance-weighted reach variable from the dataset created by Schilling (2015). This variable describes the sum of the reciprocal distances to every organization that can be reached from the focal company. A firm has a distance weighted reach of 2 if it is directly connected to two other organizations. If the organization is only connected to 1 firm that is directly connected to a third firm, the distance weighted reach will equal 1.5. The advantage of this approach is that it provides a convincing measure of the size as well as the connectivity of the network of a firm. Thereby, it considers that the overall network has multiple components and might even change over time. The higher the distance weighted reach of a firm is, the more central it is positioned in its network (Schilling, 2015). A lower number for that variable indicates that the network lacks density.
Moderating variables
Share of International Alliances: This study uses data that I retrieved from the SDC database to calculate the share of international alliances. For every single firm for the time period 1992 to 2005, I first assessed the total number of alliance partners and thereafter the total number of foreign partners. Based on these two numbers, I calculated the share of international alliances for every year.
Control Variables
I included different control variables to control for some effects that might influence the dependent variable. Rogers (2004) argues that larger firms usually have more financial resources to fund their innovation activities. In order to control for the firm size, yearly sales data was included. Moreover, the control variables sum of patents and patent stock have been included to control for past patents on which basis new patents have been developed (Coad & Rao, 2008). The stock of patents includes all patents of the last three years, including the year of observation. The total sum of patents are all the patents the firm has been granted during the time of observation from 1990 until 2005. Lastly, I also include dummy variables signaling the different industries in order to control for sectoral differences in the propensity to patent (Schilling, 2005)
3.3 Analytical Method
In order to formulate the hypothesis, this research draws on existing literature that is already well developed. Hence, a theory testing approach is followed in this study. In order to test the hypotheses, it is important to determine which statistical regression is appropriate to be used. Greene (1994) argues that if the dependent variable has discrete, nonnegative counts of numbers, the use of either the Poisson model or a Negative Binomial Regression is appropriate. This study considers patents that have been granted two years after the formation of an alliance as the dependent variable. Hence, patents apply to the prior mentioned criteria by being a non-negative, count variable taking values from 0 to 4390.
4. Results
This section, first of all, discusses the descriptive statistics. Thereafter, I interpret the results of the analysis in order to check whether the hypotheses can be confirmed or rejected.
4. 1 Descriptive Statistics
The descriptive statistics, as well as the correlations of the variables, are presented in table 1. Two years after the formation of the alliances, the dependent variable has a mean count of almost 90 patents. In total 361,773 patents have been granted. For the moderating variable share of international alliances, there is a mean count of 20 percent, while the second variable share of R&D alliances has a mean count of 23 percent. The sum of patents from the year 1990 to 2005 has a mean count of 1307 patents, while two years after the formation of an alliance the mean count of the stock of patents at that time equals 250. With regard to the firm size, the mean sales of each firm, given in thousands, in each whole year accounted for 4534,87 US-Dollar.
The second part of table 1 considers the correlations between the variables. Correlations are useful to determine whether a relationship between two variables exists and how strong it is (Taylor, 1990). Correlation coefficients can range from -1 to 1, while a correlation coefficient that equals zero indicates that there is no relationship between the variables. Grewal et al. (2004) argue that if the correlation coefficient is higher than 0.6 or -0.6, there is an increased risk of multicollinearity. Multicollinearity increases the risk to obtain inaccurate estimates of the standard errors and coefficients as well as inference errors (Mason & Perreault Jr., 1991). In the below presented data, one relationship has a high correlation coefficient and hence might suffer from multicollinearity. Namely, the variables sum of
patents and patent stock have a correlation coefficient of 0.8989. A reason for that can be that
the variable patent stock is also included in the sum of patents. The variance inflation factor (VIF) is a common method to test further whether the data is prone to multicollinearity problems. If the VIF is higher than 10, multicollinearity is considered to be a problem (Tu, et al., 2005).
Table 1: Descriptive Statistics
4.2 Hypotheses Testing
Table 3 provides a summary of the negative binomial regression. In total, four different models are tested. The first model includes only the control variables and the dependent variable. Model 2 further adds the independent variable to check the first hypothesis. After that, also the direct effect of the moderating variables on the dependent variables is tested. Lastly, in the fourth model, all variables and interaction effects are included in order to test the second and third hypothesis. All models are statistically significant since the chi-squared test equals 0.000.
In the first model, all control variables are added to find out whether they have a significant effect on the dependent variable patents t+2. The control variables are significant at the p<0.01 level. These same results can also be seen in the remaining models, just with slightly different coefficients.
The second model includes not only the control variables but also the independent variable distance weighted reach to test the first hypothesis, namely if a positive effect of a centrally positioned firm on its innovation performance exists. The independent variable has a positive coefficient (ß=0.0004) which is significant at the p<0.01 level. Thus, the first hypothesis is confirmed. However, since the coefficient is rather small, also the positive effects of being more central positioned are limited. But overall, the model shows support for hypothesis 1.
In the third model, only the moderating variables, as well as the control variables, are included to test if a direct effect of the moderators on the innovation performance exists. The results show that both moderators, a high share of international alliances (ß=0.429) and a high
share of R&D alliances (ß=0.191) have a positive and significant (p<0.01) effect on the
innovation performance of a firm. Especially the findings for international collaborations are quite interesting as they contradict the coordination problems that prior literature has found in coherence with a high share of international alliances in the alliance portfolio (Lavie & Miller, 2008).
formation of the alliance, still exists. However, a high share of R&D alliances turns insignificant (p>0.1). Having now a closer look at the second and third hypothesis, first of all, we can conclude that there is a negative interaction effect between distance weighted reach and a
high share of international alliances (ß=-0.0008), which again is significant at the p<0.01 level.
This means that firms that are centrally positioned in their alliance portfolio while having a high share of international alliances in their alliance portfolio impair their innovation performance. These findings are in line with the second hypothesis, which thus can be confirmed.
For the third hypothesis, the interaction between alliance portfolio centrality and a high share of R&D alliances, the model also shows a negative relationship (ß=-0.0002) which is however insignificant (p>0.1). Hence, the model does not provide support for the third hypothesis, which cannot be confirmed. So overall, the results of the negative binomial regression show support for hypotheses one and two, while the third hypothesis cannot be verified.
4.3 Robustness Check
In order to test whether the results of the model are significant and reliable even when tested in another way, I perform the same negative binomial regression with patents t+1 as a dependent variable. Table 4 presents the results of the negative binomial regression for patents that have been granted one year after the formation of an alliance.
Again, the chi-squared test of all models equals 0.000. Hence, all models are statistically significant. Taking a closer look at the different models, no significant differences can be noticed compared to the prior performed analysis. The coefficients and significances of the control variables stay almost the same throughout all four models.
The first hypothesis can be confirmed by the second model showing a positive and significant coefficient (ß=0.0004; p<0.01). Also, the positive direct effect of the moderating variables in the third model stays the same, only a high share of R&D alliances has a lower significance (p<0.05 instead of p<0.01) than in the prior tested model.
The fourth model further confirms the second hypothesis, showing a negative and significant interaction effect (ß=-0.0007; p<0.02) between alliance portfolio centrality and a
high share of international alliances. Additionally, the model did not provide support for the
third hypothesis since the interaction effect between alliance portfolio centrality and a high
share of R&D alliances is insignificant (ß=-0.0003; p>0.1).
Furthermore, I performed another negative binomial regression for patents t+2 to serve as a second robustness check. I tested two different models neglecting either the patent
stock or the sum of patents since these variables have a high correlation factor and hence are
potentially multicollinear. The results of the second robustness check are displayed in table 5. Both models are in line with the originally tested model, further confirming the positive and significant effect of distance weighted reach on the innovation performance while possessing a high share of international alliances has a negative and significant influence on the focal relationship. Again, no significant results were found for the second hypothesis, the effect of having a high share of R&D alliances.
Table 5: Robustness Check, Negative Binomial Regression neglecting correlated variables
5. Discussion
The following section discusses all the findings of the empirical analysis and links it to the relevant literature. After that, it provides managerial implications and eventually limitations of this study as well as possible directions for future research.
5.1 Research Implications
research aims to extend the literature by further discussing the position of a firm in its portfolio and the consequences for innovation performance. Moreover, it tests how the relationship between alliance portfolio centrality and innovation performance is influenced by the share of international alliances and R&D alliances.
In order to do so, first of all, I assessed the direct effect of occupying a central position in the portfolio on the firms’ innovation performance. Recent literature argues that alliance portfolio centrality enhances the innovation performance of a firm (Gnyawali & Madhavan, 2001; Tsai, 2001; Macauly, et al., 2018). This research confirms this relationship by showing that there is a significant and positive effect of alliance portfolio centrality on the innovation performance of the firm. Occupying a central position in the alliance portfolio offers access to knowledge and best practices that help the firm to come up with innovative ideas and thus enhance the innovation performance (Gnyawali & Madhavan, 2001; Tsai, 2001). Furthermore, by engaging in many ties to network partners, firms’ signal that they possess valuable resources and knowledge (Phelps, et al., 2012). When allying, other organizations search for partners that offer valuable resources, hence the focal firm is a desired partner with a good reputation granting it access to new resources and knowledge that ultimately lead to an enhanced innovation performance (Macauly, et al., 2018; Gnyawali & Madhavan, 2001).
In addition, I further extended the relationship by investigating the effect of having a high share of international alliances in the portfolio of alliances on the above-discussed relationship. The results indicate that the relationship between alliance portfolio centrality and innovation performance is negatively influenced by having a high share of international alliances.
Moreover, by having many direct ties to other organizations, a centrally positioned firm signals that it possesses knowledge and skills that other companies value (Phelps, et al., 2012). Due to the many ties, information can be transferred to other members of the network (Kwon, et al., 2016). Thus, information about the firm is easily available. Having a high share of international alliances in the portfolio, however, can lead to mistrust and the fear of knowledge misappropriation, since there is less information available about the reliability of foreign partners (Gulati, 1995). This leads to the focal firm being perceived as a risk of unintended knowledge spillovers by others (Parkhe, 1998; Gulati, 1995). This information is shared among the links to other firms. As a result, the positive reputation that helps a centrally positioned firm to enhance its innovation performance is weakened by having a high share of international alliances in the portfolio.
Overall, there are two mechanisms that enable a centrally positioned firm to achieve a better innovation performance than a non-central company. It has better access to information paired with an increased reputation. As described above, however, having a high share of international alliances in the portfolio of partnerships negatively influences both mechanisms. Thus, both the relationship between alliance portfolio centrality and innovation performance is weakened.
Further, former literature argues that if a firm’s alliance portfolio is characterized by a high share of international alliances, it is likely to be exposed to coordination costs. These coordination costs become overwhelming, thereby limiting the innovativeness of the firm (Lavie & Miller, 2008). Surprisingly, the results show a positive direct effect of a high share of international alliances on the innovation performance of a firm. The access to knowledge, skills, and resources that cannot be retrieved from local partners is one of the main benefits of initiating alliances with foreign partners (Wassmer, 2010). These benefits seem to outweigh the coordination costs, leading to a positive effect of international alliances on the innovation performance.
the links to other companies (Macauly, et al., 2018). Through those links, firms are able to access knowledge that has been developed by other firms (Smith-Doerr & Powell, 2003). Hence, alliance portfolio centrality considers the access to knowledge, while R&D alliances facilitate the joint development of new knowledge. Rather, it could be that the not the type of alliance is of matter, but with whom the alliance is formed. Noseleit and de Faria (2013) found that the R&D efforts of a firm are more productive when participating in an alliance within the same industry or with related industries. Alliances with unrelated industries, however, hamper the R&D efforts. Thus, it could be the case that not the type of alliance is relevant but whether the partners are based in the same or related industry, or in different industries.
5.2 Managerial Implications
This research aims to give organizations clarity how their innovation performance is affected by their position in their alliance portfolio as well as its composition. Managers need to be aware that overall the innovativeness of a firm is determined by an interplay of different factors. Even though the centrality of a firm in its network contributes to an enhanced innovation performance, also other factors that seem to be unrelated to the firms’ position determine the strength of this relationship. This research provides support that international alliances are one of these factors that influence the focal relationship. So overall, it is important for managers to acknowledge that the innovation performance is not determined by a single variable, but by an interplay of different factors.
5.3 Limitations and Future Research
Moreover, this study measures the innovativeness of a firm as the number of patents a firm got granted in a specific year. However, organizations might have developed innovations without applying for a patent and also the economic impact of the patents may differ greatly (Acs & Audretsch, 1988). Therefore, future research can test whether the hypotheses hold with alternative measures of innovativeness, e.g. the R&D expenditures of a firm. When performed for other areas than North America, this measure might also overcome the limitation that the norms and systems of getting a patent granted differ across regions.
Lastly, the effect of a high share of R&D alliances on the relationship between alliance portfolio centrality and innovation performance could not be clarified in this study. Therefore, it seems that the type of alliance does not have an effect on the focal relationship. However, the industry in which the companies are based might have an influence. Future research could further investigate whether this assumption holds or not.
6. Conclusion
This study aims to include alliance portfolio characteristics into the alliance portfolio centrality literature. By researching the effects of the share of international and R&D alliances respectively, this research shows that not only the position of a firm in a network is important but also its composition.
7. Bibliography
Acs, Z. J., Anselin, L. & Varga, A., 2002. Patents and innovation counts as measures of regional production of new knowledge. Research Policy, 31(7), pp. 1069-1085. Acs, Z. J. & Audretsch, D. B., 1988. Innovation in Large and Small Firms: An Empirical Analysis. The American Economic Review, 78(4), pp. 678-690.
Ahuja, G., 2000. Collaboration networks, structural holes, and innovation: A longitudinal study. Administrative Science Quarterly, 45(3), pp. 425-455.
Baum , J. A., Calabrese, J. & Silverman, B. S., 2000. Don't go it alone: Alliance network composition and startups' performance in Canadian biotechnology. Strategic Management
Journal, 21(3), pp. 267-294.
Baum, J. A. & Oliver, C., 1991. Institutional Linkages and Organizational Mortality.
Administrative Science Quarterly, pp. 187-218.
Björk, J. & Magnusson, M., 2009. Where Do Good Innovation Ideas Come From? Exploring the Influence of Network Connectivity on Innovation Idea Quality. Journal of Product
Innovation Management, 26(6), pp. 662-670.
Chung, S., Singh, H. & Lee, K., 2000. Complementarity, Status Similarity and Social Capital as Drivers of Alliance Formation. Strategic Management Journal, 21(1), pp. 1-22.
Coad , A. & Rao, R., 2008. Innovation and firm growth in high-tech sectors: A quantile regression approach. Research Policy, 37(4), pp. 633-648.
Cohen, W. M. & Levinthal, D. A., 1989. Innovation and learning: the two faces of R & D. The
Economic Journal , 99(397), pp. 569-596.
Coleman, J. S., 1988. Social capital in the creation of human capital. American Journal of
Sociology, Volume 94, pp. 95-120.
Das, S., Sen, P. K. & Sengupta, S., 1998. Impact of Strategic Alliances on Firm Valuation. The
Academy of Management Journal , 41(1), pp. 27-41.
de Leeuw, T., Lokshin, B. & Duysters, G., 2014. Returns to alliance portfolio diversity: The relative effects of partner diversity on firm's innovative performance and productivity.
Journal of Business Research, 67(9), pp. 1839-1849.
Dougherty, D. & Hardy, C., 1996. Sustained product innovation in large, mature
organizations: Overcoming innovation-to-organization problems. Academy of Managment
Journal, 39(5), pp. 1120-1153.
Dyer, J. H. & Signh, H., 1998. The Relational View: Cooperative Strategy and Sources of Interorganizational Competitive Advantage. Academy of Management Review, 23(4), pp. 660-679.
Eisenhardt, K. M. & Schoonhoven, C. B., 1996. Resource-based View of Strategic Alliance Formation: Strategic and Social Effects in Entrepreneurial Firms. Organization Science, 7(2), pp. 136-150.
Faems, D., 2006. Collaboration for Innovation: Processes of Governance and Learning in R&D
Alliances, s.l.: s.n.
Freeman, L. C., 1978. Centrality in social networks conceptual clarification. Social Networks, 1(3), pp. 215-239.
George, G., Zahra, S. A., Weathley, K. K. & Khan, R., 2001. The Effects of Alliance Portfolio Characteristics and Absorptive Capacity on Performance: A Study of Biotechnology Firms.
The Journal of High Technology Management Research, 12(2), pp. 205-226.
Gilsing, V. et al., 2008. Network embeddedness and the exploration of novel technologies: Technological distance, betweenness centrality and density. Research Policy, Volume 37, pp. 1717-1731.
Gnyawali, D. R. & Madhavan, R., 2001. Cooperative Networks and Competitive Dynamics: A Structural Embeddedness Perspective. Academy of Management Review, 26(3), pp. 431-445. Goerzen, A. & Beamish, P. W., 2005. The effect of alliance network diversity on multinational enterprise performance. Strategic Management Journal, 26(4), pp. 333-354.
Grant, R. M. & Baden-Fuller, C., 2003. A Knowledge Accessing Theory of Strategic Alliances.
Journal of Management Studies, 41(1), pp. 61-84.
Greene, W. H., 1994. Accounting for excess zeros and sample selection in Poisson and negative binomial regression models.
Grewal, R., Cote, J. A. & Baumgartner, H., 2004. Multicollinearity and Measurement Error in Structural Equation Models: Implications for Theory Testing. Marketing Science, 23(4), pp. 519-529.
Haagedoorn, J., 2002. Inter-firm R&D partnerships: an overview of major trends and patterns since 1960. Research Policy, 31(4), pp. 477-492.
Haagedoorn, J. & Duysters, G., 2002. External sources of innovative capabilities: the preferences for strategic alliances or mergers and acquisitions. Journal of Management
Studies , 39(2), pp. 167-188.
Haeussler, C. & Higgins, M. J., 2014. Strategic Alliances: Trading Ownership for Capabilities.
Journal of Economics & Management Strategy , 23(1), pp. 178-203.
Hamel, G., Dos, Y. L. & Prahalad, C. K., 1989. Collaborate with your Competitors and Win.
Harvard Business Review, 67(1), pp. 133-139.
Hamilton, W. F. & Singh, H., 1991. Strategic Alliances in Technological Innovation:
Cooperation in Biotechnology. Journal of High Technology Management, 2(2), pp. 211-221. Hoffmann, W. H., 2007. Strategies for Managing a Portfolio of Alliances. Strategic
Management Journal , 28(8), pp. 827-856.
Jiang, R. J., Qingjiu, T. T. & Santoro, M. D., 2010. Alliance Portfolio Diversity and Firm Performance. Strategic Management Journal, 31(10), pp. 1136-1144.
Jordan, A. T., 1990. Organizational culture and culture change: a case study. Studies in Third
World Societies, Volume 42, pp. 209-226.
Kim, C.-S. & Inkpen, A. C., 2005. Cross-border R&D alliances, absorptive capacity and technology learning. Journal of International Management, Volume 11, pp. 313-329. Koput, K. W., 1997. A chaotic model of innovative search: some answers, many questions.
Organization Science, 8(5), pp. 528-542.
Kwon, S.-W., Haleblian, J. & Hagedoorn, J., 2016. In country we trust? National trust and the governance of international R&D alliances. Journal of International Business Studies, Volume 47, pp. 807-829.
Lahiri, N. & Narayanan, S., 2013. Vertical Integration, Innovation, and Alliance Portfolio Size: Implications for Firm Performance. Strategic Management Journal, 34(9), pp. 1042-1064. Lane, H. & Beamish, P. W., 1990. Cross-cultural cooperative behavior in joint ventures in LDCs. Management International Review, Volume 30, p. 87.
Lavie, D., 2007. Alliance Portfolio and Firm Performance: A Study of Value Creation and Apporpriation in the U.S. Software Industry. Strategic Management Journal, 28(12), pp. 1187-1212.
Lavie, D. & Miller, S. R., 2008. Alliance Portfolio Internationalization and Firm Performance.
Lin, C. et al., 2012. The alliance innovation performance of R&D alliances-the absorptive capacity perspective. Technovation, Volume 32, pp. 282-292.
Linden, A. & Mäntyniemi, S., 2011. Using the negative binomial distribution to model overdispersion in ecological count data. Ecology, 92(7), pp. 1414-1421.
Macauly, C. D., Richard, O. C., Peng, M. W. & Hasenhuttl, M., 2018. Alliance Network
Centrality, Board Composition, and Corporate Social Performance. Journal of Business Ethics, 151(4), pp. 997-1008.
Mason, C. H. & Perreault Jr., W. D., 1991. Collinearity, power, and interpretation of multiple regression analysis. Journal of Marketing Research, pp. 268-280.
Narula , R. & Haagedoorn, J., 1999. Innovating through strategic alliances: moving towards international partnerships and contractual agreements. Technovation, 19(5), pp. 283-294. Nooteboom, B., 1999. Innovation, Learning, and Industrial Organisation. Cambridge Journal
of Economics , 23(2), pp. 127-150.
Noseleit, F. & de Faria, P., 2013. Complementarities of internal R&D and alliances with different partner types. Journal of Business Research, 66(10), pp. 2000-2006.
Oerlemans, L. A. G., Knoben, J. & Pretorius, M. W., 2013. Alliance portfolio diversity, radical and incremental innovation: The moderating role of technology management. Technovation, 33(6-7), pp. 234-246.
Osborn, R. N. et al., 1998. Embedded Patterns of International Alliance Formation.
Organization Studies, 19(4), pp. 617-638.
Oxley, J. E. & Sampson, R. C., 2004. The Scope and Governance of International R&D Alliances. Strategic Management Journal, Volume 25, pp. 723-749.
Parkhe, A., 1998. Understanding Trust in International Alliances. Journal of World Business, 33(3), pp. 219-240.
Phelps, C., Heidl, R. & Wadhwa, A., 2012. Knowledge, Networks, and Knowledge Networks: A Review and Research Agenda. Journal of Management, 38(4), pp. 1115-1166.
Phene, A., Madhok , A. & Liu, K., 2005. Knowledge transfer within the multinational firm: What drives the speed of transfer?. Management International Review, pp. 53-74. Portes, A., 1998. Social capital: Its origins and applications in modern sociology. Annual
Review of Sociology , 24(1), pp. 1-24.
Pothukuchi, V., Choi, J., Chen, C. C. & Park, S. H., 2002. National and Organizational Culture Differences and International Joint Venture Performance. Journal of International Business
Qi, J., Sutton, N. K. & Zheng, Q., 2015. The Value of Strategic Alliances in Acquisitions and IPOs. Financial Management, 44(2), pp. 387-430.
Rogers, M., 2004. Networks, firm size and innovation. Small Business Economics, 2(2), pp. 141-153.
Sampson, R. C., 2004. Organizational Choice in R&D Alliances: Knowledge-Based and Transaction Cost Perspectives. Managerial and Decision Economics , Volume 25, pp. 421-436.
Sampson, R. C., 2007. R&D Alliances and Firm Performance: The Impact of Technological Diversity and Alliance Organization on Innovation. Academy of Management Journal , 50(2), pp. 364-386.
Schilling, M. A., 2015. Technology Shocks, Technological Collaboration, and Innovation Outcomes. Organization Science, 26(3), pp. 668-686.
Sirmon, D. G. & Lane, P. J., 2004. A Model of Cultural Differences and International Alliance Performance. Journal of International Business Studies , 35(4), pp. 306-319.
Smircich, L., 1983. Concepts of Culture and Organizational Analysis. Administrative Science
Quarterly, pp. 339-358.
Smith-Doerr, L. & Powell, W. W., 2003. Networks and Economic Life. The Handbook of
Economic Sociology .
Taylor, R., 1990. Interpretation of the correlation coefficient: a basic review. Journal of
Diagnostic Medical Sonography, 6(1), pp. 35-39.
Teece, D. J., 1982. Towards an economic theory of the multiproduct firm. Journal of
Economic Behavior & Organization, 3(1), pp. 39-63.
Todeva, E. & Knoke, D., 2005. Strategic Alliances and Models of Collaboration. Management
Decision, 43(1), pp. 123-148.
Tsai, W., 2001. Knowledge Transfer in Intraorganizational Networks: Effects of Network Position and Absorptive Capacity on Business Unit Innovation and Performance. The
Academy of Management Journal, 44(5), pp. 996-1004.
Tu, Y.-K., Gilthorpe, M., Clerehugh, V. & Kellett, M., 2005. Problems of correlations between explanatory variables in multiple regression analyses in the dental literature. British Dental
Journal, 199(7), pp. 457-461.
Uzzi, B., 1996. The Sources and Consequences of Embeddedness for the Economic Performance of Organizations. American Sociological Review, pp. 674-698.
Wassmer, U., 2010. Alliance portfolios: A review and research agenda.. Journal of
