Are internal and external networks mutually exclusive? How firm’s organizational structure influences the relationship between alliance portfolio and innovation performance

(1)

Are internal and external networks mutually exclusive? How firm’s organizational

structure influences the relationship between alliance portfolio and innovation

performance

Master Thesis

University of Groningen Faculty of Economics and Business

MSc BA Strategic Innovation Management

25th June 2018

Supervisor: Dr. Florian Noseleit

Co-assessor: Dr. Pedro de Faria

Name: Veneta Fesliyska

Student number: S3504395

E-Mail: v.fesliyska@student.rug.nl

(2)

1

Abstract

(3)

2

Introduction

Due to the technological progress, global market is now an unstable and complex ground for firms which were in habit of innovating internally. Companies needed to modify their business models if they want to sustain their competitive advantage. Their product and process innovations had to be adopted to these fast-changing paces of economies. Innovation in isolation is already impossible, as external knowledge is a necessity for superior innovation performance. One of the key practices used by firms to achieve higher innovation performance was forming an alliance with a reliable partner (Schilling, 2015). This is the reason why prior studies claim that by the turn of the 21st century, the majority of the world’s conglomerates had over 20% of their assets, and over 30% of their annual research expenditures, confined in alliance relationships (Ernst, 2006). Another possible pathway utilized for external knowledge access is broadening organization’s borders worldwide by establishing numbers of subsidiaries. Accordingly, one might recognize the similar role of those two mechanisms used by organizations in their intention to attain excessive innovation performance and, therefore, ask the question how are they interrelated.

Alliance portfolios are seen not only as an opportunity for a firm to access valuable external knowledge from their partners. According to the resource-based view, engaging in a wide bundle of different strategic alliances simultaneously and forming an alliance portfolio provides firms with access to a broad share of complementary resources (George, Zahra, Wheatley, & Khan, 2001), leads to greater opportunities to realize economies of scale and scope (Lavie & Miller, 2008), and enhances the comprehensive learning capabilities of all partners (Kale & Singh, 2007). Looking at its structural configuration, social network theory points at several portfolio characteristics that leverage firms’ innovative performance such as its size and partner’s technological and geographical diversity. Howbeit, there is one particular structural feature that additionally ameliorates the benefit from the dense connectivity between partners, namely the level of centrality in firm’s positioning in its alliance network (Faems, Janssens, & Neyens, 2012; Tsai, 2001). The compelling advantage of the central position within the alliance portfolio is that it facilitates firm’s access to unique network resources by shortening the path lengths between partners (Dong, McCarthy, & Schoenmakers, 2017). Following the presumption made by Schilling (2015) that alliances are essential means of the innovation process, I want to further explore the impact of network centrality by answering the subsequent research question:

(5)

4 In addition to this, researchers also provided an evidence for the importance of firm’s internal network concretizing that subsidiaries have the unique privilege to be exposed to dissimilar cultural, political, technological, societal and legal conditions (O'Donnell, 2000). Subsidiaries are aligned with the organizations, they do not operate completely independent, therefore all their preeminent priorities, requirements, organizational routines and strategic targets are adjusted to the headquarters’ ones (Bouquet & Birkinshaw, 2008). If we recall the main drawbacks of alliance collaborations, subsidiaries provide a pathway to overcome the plausible moral hazard concerns and lowering the risk of opportunistic behavior pertained usually to interfirm partnerships by correspondingly lowering the transaction and contractual costs (Balakrishnan & Koza, 1993; Kogut, Shan, & Walker, 1992; Gulati, 1995b; Powell, Koput, & Smith-Doerr, 1996). Although a growing body of literature is focusing on alliances as mechanism used by firms to acquire new external knowledge (Kim, C. S., & Inkpen, A. C. , 2005), it is not yet considered how their role in the innovative performance could be substituted by the firm’s internal network of subsidiaries. Alliance portfolios are predominantly seen as one of the most prevailing means used by firms to leverage existing knowledge, maintain access to complementary diverse resources and extent of their internal R&D activities (Eisenhardt & Schoonhoven, 1996; Hagedoorn, 1993; Powell, Koput, & Smith-Doerr, 1996). On the other hand, Kogut and Zander (1993) identified subsidiaries as a highly unique environment to develop valuable knowledge because these subunits build an especially efficient channel to overcome barriers in international knowledge flows.

Introducing those two contradicting points, a question appears if the network of subsidiaries can substitute firm’s alliance portfolio partners in their seek to accomplish superior innovative performance. Hereby, the following sub-questions can be formulated as:

Sub-Questions: What is the impact of firm’s number of subsidiaries on its innovation performance? Do subsidiaries have the capability to substitute alliance portfolio partners in their strategic innovative importance?

(6)

5 The findings of this study show that network centrality has a direct positive and significant effect on innovation performance. Moreover, I found the same positive and significant direct impact of firm’s number of subsidiaries on the firm’s number of patent applications for 1997. Nevertheless, this study implies that large number of subsidiaries acts like a substitute for firm’s central position within its alliance portfolio, arguing the superior innovation performance could be achieved by firms without being centrally positioned in the alliance portfolio if they possess an immense internal network of subunits.

(7)

6

Literature review

Alliances and Innovation

In the highly competitive market nowadays, innovation plays compelling part in firm’s outstanding performance (Dong, McCarthy, & Schoenmakers, 2017). Firms are fighting through the battle in this market by actively searching for novel and unique knowledge sources which can support them throughout the innovation process. Resource-based view scholars have maintained that most of the firms do not possess all the fundamental resources and knowledge inputs internally (Dias & Magrico, 2011; Eisenhardt & Schoonhoven, 1996). In order to build up a sustainable competitive advantage, organizations need to access knowledge outside firm’s borders, especially in industries with high levels of technological change and environmental uncertainty (Dias & Magrico, 2011; Eisenhardt & Schoonhoven, 1996; Powell, Koput, & Smith-Doerr, 1996). Therefore, a significant change in organizations’ behavior occurred as a response of the rapidly growing market complexity and competition among players, answering the need to adapt quickly to the abbreviated product and process life cycles (Schilling, 2015). Nowadays, organizations are not able to innovate in isolation (Dong, McCarthy, & Schoenmakers, 2017). Instead of trying to gather relevant knowledge internally through complex processes and routines, firms started looking outside firm’s boundaries to find potential partners with whom to collaborate for the needed innovative capabilities (Deeds & Hill, 1996; Eisenhardt & Schoonhoven, 1996; Mowery, Oxley, & Silverman, 1996; Sakakibara, 2002).

(8)

7 those agreements not only led to the optimization and abbreviation of the innovation process, but also made it less costly in a way that each partner in the relationship does not need to create all essential knowledge internally.

Alliance Portfolios/Networks

Strategic management literature has broadly covered most topics in order to explain the incentives of each firm to form a strategic alliance as well as the forthcoming effect from those alliances on firms’ innovation performance. The consequential impact of alliances on firm’s performance could not be addressed without having in consideration that firms usually participate in more than one interfirm collaboration. There is a significant interdependence not only between the two partners forming this collaboration, but also between the indirect ties that affect obliquely firm’s performance (Ahuja, 2000; Baum, Calabrese, & Silverman, 2000; George, Zahra, Wheatley, & Khan, 2001; Lavie & Rosenkopf, 2006). Discussing the effect of alliance formation on innovation performance, researchers contemplate to the network approach, claiming that every partner within the network can play a compelling role in achieving greater innovative outcome (Lavie & Rosenkopf, 2006). This is how the portfolio approach was adopted to the strategic management literature, launching the observation of cross-network collaborations, indirect ties and triadic relationships.

Initially, following the social network perspective, Doz and Hamel (1998) presented the alliance portfolio as the set of mutual alliances maintained by a focal firm, which was later altered to the set of alliances in which a firm is involved, all direct ties with partner firms (Bae & Insead, 2004; Baum, Calabrese, & Silverman, 2000; Rowley, Behrens, & Krackhardt, 2000). Others scholars approach the alliance portfolio aspect from a learning perspective, defining it as a focal firm’s accumulated alliance experience (Anand & Khanna, 2000; Hoang & Rothaermel, 2005; Kale, Dyer, & Singh, 2002; Reuer, Park, & Zollo, 2002; Simonin, 1997). Exploring the alliance portfolio configuration, researchers in this field used theories grounded in the social network literature, the resource-based view, organizational learning or a combination of the three (Ahuja, 2000; Baum, Calabrese, & Silverman, 2000; George, Zahra, Wheatley, & Khan, 2001; Lavie & Rosenkopf, 2006; Stuart, Hoang, & Hybels, 19999). Overall, the most common view within the management literature uses ‘firm’s collection of direct alliances with partners’ (Lavie, 2007) as a characterization of the term. Although dyadic relationships are the fundamental building blocks of the alliance system, the structure of the portfolio by its very nature is a convoluted network of interconnected alliances (Cranmer, Desmarais, & Menninga, 2012).

(9)

8 reliable partner by adding precise criteria to facilitate the collaboration success. This allows organizations to strengthen their position in obtaining and, moreover, sustaining innovation (Faems, Van Looy, & Debackere, 2005; Katila & Ahuja, 2002; Laursen & Salter, 2006). Faems et al (2005) found in their study that, in terms of management decisions, firms should contemplate the idea of a “portfolio approach with interorganizational arrangements”. Rather than focusing on simple dyadic collaborations when realizing their innovation strategies, firms need to look over the broader perspective of the alliance network. However, there are different dimensions discussed in the literature which represent those portfolio characteristics that influence the innovation performance, namely the portfolio size (depth and breadth), diversity and central positioning of a firm (Wassmer, 2010).

Alliance Portfolios Composition

The structural characteristics of the alliance network mentioned above are from a significant importance for every firm’s innovation success. By organizing them precisely to boost firm’s innovative performance, the overall network structure could also enhance the possibility for higher innovation rates, profitable distribution of governance practices, knowledge spillovers, and other advantageous possibilities (Ahuja, 2000; Gilsing, Nooteboom, Vanhaverbeke, Duysters, & van den Oord, 2008; Gulati & Higgins, 2003; Robinson & Stuart, 2007; Rosenkopf & Almeida, 2003; Schilling & Phelps, 2007). One of the reasons why firms need exceedingly disseminated portfolio with greater number of partners is that it contributes to greater innovation outcomes because such firms are normally exposed to a broader and more diverse external knowledge base (Ahuja, 2000; Baum, Calabrese, & Silverman, 2000; Rothaermel, 2001; Rowley, Behrens, & Krackhardt, 2000; Stuart, Hoang, & Hybels, 19999). This external knowledge base depends not only on the accessible resources outside the firm’s borders, but also on the density of firms clustered in a given geographical area, sector of activity, social ties, nature of the knowledge and partners involved within an alliance collaboration (Jaffe, Trajtenberg, & Henderson, 1993; Tecce, 1986; Saxenian, 1994; Lavie, 2006). The composition of a focal firm’s alliance portfolio typically not only signals for quality, quantity, and diversity of information and resources to which the focal company has access, but also influences the efficiency of access to network resources, and the flexibility or stability of the focal company’s position in the interorganizational field (Hoffman, 2007).

Alliance Portfolio Size

(10)

9 facilitating collaborations as well as constraining interdependencies across all embedded alliances (Faems, Janssens, & Neyens, 2012). However, another point of view is that there are diminishing returns from each additional alliance added to an alliance portfolio once their alliance portfolio has gone above a certain size (Wassmer, 2010). Hence, due to the inconclusive results in the literature regarding the effect of the size of the portfolio on innovation performance, it is often separated into two different dimensions - breadth and depth. While discussing the breadth of the portfolio, we normally address wide range of partners that differ in their field of specificity and, therefore, provide knowledge base that is rich in various market niches (Belderbos, Carree, & Lokshin, 2006; Laursen & Salter, 2006; Lavie, 2006; Nieto & Santamaria, 2007; Beers & Zand, 2014). The depth of the portfolio is often characterized with partners with similar areas of development that specialize in one specific marker niche and tend to accomplish communal innovation outcome (Belderbos, Carree, & Lokshin, 2006; Laursen & Salter, 2006; Lavie, 2006; Nieto & Santamaria, 2007; Beers & Zand, 2014).

Alliance Portfolio Diversity

Researchers also strongly emphasis on the diversified portfolio partners in terms of geographical and technological knowledge. Alliance portfolio diversity have been defined as the degree of technological variance in partners, the position of firms’ partnerships across the innovation value chain and the geographic scope of their partnerships (Jiang, Tao, & Santoro, 2010). There is no doubt that in order to acquire critical external knowledge that is divergent to their own knowledge base, firms need the capability to learn from their partners in the portfolio. This two-sided process can be valuable for all participants in the collaboration if they have quite diversified niches of specialization (Lavie & Miller, 2008). Heterogeneity of partners with various sizes, structures and market fields can increase firm’s breadth of search, learning capabilities, resource access and, moreover, reduce the threat of the possible core rigidities (Jiang, Tao, & Santoro, 2010). Partners could vary in terms of geographical proximity, technological specifications, market niches, structural and relational configuration. For example, compared to collaborations with domestic partners, international alliance portfolios provide greater flexibility, responsiveness, greater extend of accessible knowledge base scope, adaptability to global market conditions, and reduction of risk and uncertainty (Eisenhardt & Schoonhoven, 1996; Hagedoorn, 1993; Harrigan, 1985; Kogut & Kulatilaka, 1994; Powell, Koput, & Smith-Doerr, 1996; Teece, 1992; Levinthal & March, 1993). Diversified partners can offer network resources that extricate firms from their own competency traps and stimulate innovations, new solutions, and new skills (Lavie & Miller, 2008).

(11)

10 works that investigate alliance portfolio diversity found that it has a positive correlation with innovativeness (Ahuja, 2000; Katila & Ahuja, 2002; Baum, Calabrese, & Silverman, 2000). Firms use diversified external knowledge sources to face technological change and the increasing complexity of knowledge processes (Noseleit & de Faria, 2013). Some scholars also look at the focal firm’s value chain (Baum, Calabrese, & Silverman, 2000; Powell, Koput, & Smith-Doerr, 1996) making a distinction between types of partners as upstream, downstream, and horizontal alliance partners in the portfolio. While upstream partners (universities and research institutes) are often used for explorative collaborative activities as fundamental research, experimentation, and improvisation, downstream partners (customers and suppliers) are suitable for exploitative activities, such as fine-tuning, upscaling and adjusting existing products (Faems, Van Looy, & Debackere, 2005; Lavie & Rosenkopf, 2006). Finally, horizontal partners, presented by competitors, allow the focal firm to benefit from the partner industry-specific knowledge and build platforms and industry standards simultaneously (Kim & Higgins, 2007).

Nevertheless, immensely diversified alliance portfolio does not only have beneficial effect on firm’s innovative performance (Faems, Van Looy, & Debackere, 2005). Hereby, the literature declared also the drawbacks that firms can experience by adapting the portfolio approach if it is not adjusted to firm’s capabilities and own resource base. A major dissimilarity between alliance partners would lead to higher complexity in the management of the collaborations, which in turn, according to transaction cost theory approach, will increase the management and transaction costs (Wassmer, 2010). Moreover, the possibility that the partner will encounter difficulties in the knowledge sharing process is higher when there is a significant distance between the two or multiple technological domains, which will negatively influence the benefits of such partnerships (Lane & Lubatkin, 1998). Furthermore, an increase in the diversity can lead to more conflicts due to nationality and cultural differences (Jiang, Tao, & Santoro, 2010; Lavie & Miller, 2008). From the perspective of organizational learning theory, the knowledge sharing capabilities within a highly diversified but dense portfolio can also alter a focal firm’s innovative outcome negatively by stimulating a vicious circle different complex contacts, reducing the flexibility for collaboration with new partners and diminishing the responsiveness to new market trends (Capaldo, 2007; Wassmer, 2010).

(12)

11

Hypothesis Development

Network Centrality

In the former research works, centrality was mainly discussed in the network framework as an emblematic aspect of the network structure (Wassmer, 2010). Howbeit, previous studies also pointed on the importance of the central position in alliance portfolio for focal firm’s superior innovative outcome (Powell, Koput, & Smith-Doerr, 1996; Dong, McCarthy, & Schoenmakers, 2017). The term “network centrality” is either defined as the extent to which an actor is central in a network (Freeman, Roeder, & Mulholland, 1979), or the level of exposure of the unit to valuable external knowledge (Tsai, 2001). Combining both definitions together, it could be summarized that the more central position each firm holds in the portfolio, the more likely it is to access valuable information (Gulati, 1999; Powell, Koput, & Smith-Doerr, 1996). Another advantage of the network centrality that was formerly discussed in the literature is the reduction of the search costs for alliance partners (Chung , Singh, & Lee, 2000). Companies that occupy a central position within the alliance network favor more from their access to valuable and unique resources than the ones that are not that centrally located (Ahuja, 2000; Powell, Koput, & Smith-Doerr, 1996). Seen as a valuable information source in the industry network, the central location improves the access to relevant information about the existence and the reliability of potential future partners that each firm needs in its strategic planning (Gulati, 1999).

Social network theory mainly focuses on the geographical proximity when discussing this particular structural characteristic. Howbeit, network centrality is not only an indicator of a particular positioning of the firm in its alliance portfolio, it also signals for the level of connectivity between the partners. This particular positioning of a focal firm within a network not only enhances the likelihood of receiving valuable information (Gulati, 1999; Powell, Koput, & Smith-Doerr, 1996), but also leads to substantial reduction of the search costs for alliance partners (Chung , Singh, & Lee, 2000). Various positions in the alliance network usually characterize with different opportunities for the organization, for example the ease of access to external knowledge, which is an essential capability in new products’ development (Tsai, 2001). The central position does not only embellish the ability to access desired strategic resources, but also characterize with better opportunities to gain advantage from all partners in the network (Coleman, 1990; Tsai & Ghoshal, 1998).

(13)

12 unique access to all other partner’s knowledge, information and practices, which helps them to adapt the external knowledge and practices to their market needs, respond the emerging market trends and deal with competitive challenges (Tsai, 2001). Overall, the structural position in every interfirm network is proven to be from an essential strategic importance, acting as a facilitator and inhibitor of innovation activities (Dong, McCarthy, & Schoenmakers, 2017). The shorter the average path length between units in an alliance network, the more dense the network connectivity is, which alleviates the ease of knowledge transfer and, therefore, results on superior innovation performance of all members in the chain (Fleming, King lll, & Juda, 2007). Stronger ties in the alliance network intensifies the process of transferring tacit and complex knowledge and helps units to overcome the problems related to geographic distance, technological differences, and competition between units (Hansen, 1999; Tsai, 2001; Hansen & Løvås, 2004).

As a result, looking at the network centrality as a measure that implies not only the position of the firm within the network but also captures the size and the density of the network in which a firm is embedded (Schilling, 2015), it is expected to have a positive impact on firm’s innovation performance. Therefore, my first hypothesis will be formulated as:

Hypothesis 1: Companies that occupy a central position in their alliance network achieve on

(14)

13

Network of subsidiaries and alliance network centrality

When it comes to international competitiveness, not only the alliance partners play a substantial role as a source of strategic resources (Birkinshaw, 1996). Another potentially important source of competitive advantage, which was immensely discussed by the strategy and management scholars, is firm’s networks of subsidiaries (Ghoshal & Bartlett, 1990; Brikinshaw & Hood, 1998; Rugman & Verbeke, 2001). One of the reasons for this could be found in the fact that often subsidiaries are exposed to different cultural, political, technological, societal and legal environments, which enables them to encounter different markets, competitors and management practices (O'Donnell, 2000). Large multinational companies that usually establish broad network of subsidiaries which are globally spread across the market, benefit from greater flexibility, responsiveness, adaptability to global market conditions, and reduction of risk and uncertainty (O'Donnell, 2000). According to Tsai (2001), those interunit links and networks presented by subsidiaries play a crucial role in the internal learning process from which organizational units determine new opportunities and gather novel knowledge. Hereby, not only the firm position within the alliance network and alliance partners grant firm the opportunity to access essential resources and external knowledge, the subsidiaries of the firm are often seen as a substantial source of reliable capabilities that are fundamental for innovative performance (Tsai, 2001; Gulati, 1999; Powell, Koput, & Smith-Doerr, 1996).

Firm’s private entities function in an intra-firm network, including the headquarters and associated subsidiaries (Ghoshal & Bartlett, 1990), which makes this already established chain a unique environment for cultivating valuable knowledge. Мultiunit firms are seen as especially effective channels to stun barriers in the international knowledge flows (Kogut & Zander, 1993). Those interunit collaborations possess relevant knowledge which is usually not produced inside each entity's borders, so all of them have the unique ability to benefit from knowledge produced in the headquarters or other subsidiaries (Sofka, Seshu, & de Faria, 2014). Colakoglu, Yamao and Lepak (2014) claim that each subsidiary has the ability to acquire, assimilate, and exploit knowledge from its headquarters and its peer subsidiaries which is an exceptional advantage from the inclusion in such network.

(15)

14 expand the subsidiary’s capacity to develop new knowledge and reduces the transaction risk and costs of the alliance collaborations (Andersson, Forsgren, & Holm, 2007; Gulati, 1999; Malmberg & Maskell, 2002).

Considering alliance network centrality as another structural dimension discussed above, it is likewise observed as a mean to access valuable knowledge outside the firm boundaries. However, the central positioning within an interfirm knowledge network could act as a “double-edged sword” for the innovation output (Dong, McCarthy, & Schoenmakers, 2017). The central position combined with dense knowledge ties to all the units in the corporate network may cause the so-called information overload, block profitable search for innovation and blur the innovation focus of the focal firm (Dong, McCarthy, & Schoenmakers, 2017). Adopting the transaction costs approach, alliances are often seen as complex organizational arrangements that require substantial efforts to identify the right partners for collaborative activities, detailed assessments for ratifying contracts, which in turn increases the transaction costs and requires considerable amount of management attention to sustain the relationship (Ring & Van de Ven, 1994; Gulati, Khanna, & Nohria, 1994; Doz, 1996).

Thus, if those supplementary costs and efforts used to govern alliance collaborations are devoted to the network of subsidiaries with the aim to embellish stronger internal knowledge exchange, there is an attainable possibility that such novel knowledge will be generated inside the firm’s boundaries (Tsai, 2001). Innovativeness is a process that is closely tied to firm’s absorptive capacity, defined as the ability of a firm to recognize the value of new, external information, assimilate it, and apply it to commercial ends (Cohen & Levinthan, 2000). Therefore, firms can also use their internal collaborations to enhance this capability by learning from the skills and competencies of their internal corporate network (Cohen & Levinthan, 2000; Lane & Lubatkin, 1998). If the organization’s focal target is on coordinating the learning processes and stimulating knowledge transfer between subsidiaries (Foss, 1997; Schleimer & Pedersen, 2013), the internal corporate network can improve its the innovation output (Ciabuschi, Dellestrand, & Martin, 2011), increase the absorptive capacity of each entity (Schleimer & Pedersen, 2013) and the strengthen the general progress of all subsidiaries (Bouquet & Birkinshaw, 2008).

(16)

15 certain knowledge transfer barriers, combining them within one innovation strategy could lead to loss in the main advantage gained from such network collaborations (Ciabuschi, Dellestrand, & Holm, 2012; Tsang, 2000; Jandik & Kali, 2009; Oxley, 1997). Focusing on one of the two channels for knowledge generation increases the feasibility to achieve innovation output. These two internal and external network aspects act more as a substitutable in terms of innovation performance. By establishing an extensive network of subsidiaries, firms could gather valuable external knowledge for innovative activities from its own subunits without being centrally positioned in their alliance network. Thus, the positive correlation between network centrality and innovation performance is expected to diminish when the focal firm already possess a large network of subsidiaries from which to gather valuable external knowledge. Following this reasoning, I hypothesize:

Hypothesis 2: The number of subsidiaries of the focal firm influences negatively the relationship

between an alliance network centrality and the innovation performance.

(17)

16

Methodology

This section aims to explain the data collection process and to clarify the measurements used for the different variables in the model. To investigate the effect of network centrality on innovation performance, moderated by the size of subsidiaries’ network, a theory-testing approach has been used. Van Aken, Berends & Van der Bij (2012) pointed on the importance of using a theory testing approach if the literature streams are already elaborated and not scattered, but there is still a literature gap in the theoretical explanations.

The dataset used for the aim of this research was primary structured by Schilling (2015), and used in the paper “Technology Shocks, Technological Collaboration, and Innovation Outcomes”. It was collected with the aim to conduct exploratory study on alliance activity and the effect of technology shock on alliance behavior, how these technological changes lead to a change in the global technology collaboration network and their link to innovative outcomes. Therefore, she used the SDC database to extract a prior set of 13,906 firms that participated in the global technology collaboration network between 1990 and 2005. After that she reduced this current set by only identifying North American firms that were publicly held for at least three years of the study period (1990 - 2005) and that applied for at least one subsequently granted patent during that time. These additional criteria led to a total number of 535 North American firms (Schilling, 2015).

Data Collection

(18)

17 information regarding the lack of alliance partners. Later in this section a brief explanation of all the variables will be provided.

The following step was additional data collection regarding the total number of subsidiaries (foreign and local ones) of each firm for the years 1997 and 2017. When the information was not present for 1997, the latest available was gathered from the period 1990 to 2005. For the former information about 1997, SEC (Securities and Exchange Commission) information platform was used as it contains 99% of the important facts for the US public companies and their owners. From the available annual reports from 1997, I searched particularly for “Exhibit 21” of each report which is the part that consists the list of all subsidiaries of each registrant. Although not all the firms had a separated exhibit for the list of subsidiaries, most of them had a particular section in the report that clarified the total number of subsidiaries. However, when the company’s information was not available in the platform mentioned above, the search proceeds by typing “annual report” and company’s name because most of the annual public materials are available in their private website or other information websites. Later on, Orbis was the additional platform used for the data regarding 2017 as it contains information for firm’s current working status, its subsidiaries, the headquarter location and financial information. Only the information about subsidiaries was gathered through this platform. To check for possible bias in the data, all the information for 2017 was additionally compared to the one available in the open platform “Google”. The up-to-date information was latterly used as a robustness check for the base model and provided interesting insights regarding the time lag importance of the main concept of this research.

Measures

Dependent Variable

(19)

18 the count of all successful patents applied for in the three years (leading up to and including the observation year 1997) as a control variable in the model.

Independent variable

Network centrality: In measuring the central position of a focal firm within a network of alliances, I use the distance-weighted reach, presented by Schilling (2015). Network centrality is used initially in the alliance literature to identify firm’s ability to access more valuable resources and take advantage from its central position (Tsai, 2001; Ahuja, 2000; Powell, Koput, & Smith-Doerr, 1996; Schilling, 2015). The variable is calculated by the sum of the reciprocal distance to every organization that is reachable from a focal firm and is being defined as the minimum geodesic distance from this firm to a chosen partner (Schilling, 2015). Clarifying the definition given above, we can say that a firm is occupying a central position within the alliance network when the path lengths that relate it to the other partners in the network are shorter. This additionally illustrates that all partners in the alliance portfolio are connected to the same central “hub”. Therefore, the higher the distance-weighted reach, the more centrally positioned is the focal firm within the network. Low value of this variable identifies a low density in the network. By counting only the path length between connected pairs of nodes, the measurement avoids the problem with limitless path length which normally could occur in disconnected networks. It also provides more reliable measure than the simple average path length between connected pairs because it factors the size of connected components. Overall, it is mainly based on the focal firm’s position in the global technology collaboration network. Because this distance-weighted reach (such as firm-level alliance counts) is positively skewed, the measure was log-transformed to improve its normality (Schilling, 2015).

Moderator

(20)

19 (Histogram 1, Appendix A), it has to be log-transformed for the empirical analysis in order to display better distributional properties. After this conversion, the level of skewness was lowered to a certain extent although it was still not normally distributed (Histogram 2, Appendix A). However, the comparison between the use of these two variables in the final model showed that there is better model fit with the log-transformed count. This additionally supported the decision to use the log variable in the final analysis. Therefore, for the following analytic method, the variable counting the number of subsidiaries for each focal firm is utilized as a continuous variable rather than a count one.

Control Variables

(21)

20

Analytical method

Alliance portfolio stream is a broadly investigated topic throughout the strategic management literature. Guided from all the theory that is available in line with this research, a theory - testing approach is followed. As cross-sectional data was available regarding firm’s number of subsidiaries, the research main focus is only one particular year of examination. Therefore, the decision to choose one year for investigation, namely 1997, was initiated with the intense to provide better understanding of the effect from collaboration between units operating in different industry sectors. As not all the firms in the database have an information available for all the years from the time period 1990 – 2005, I checked which is the year that is most commonly met within the sample. Seeing that for the year 1997 I can extract a greater sub-sample from the whole database let to the decision of using this particular time period for the research. Furthermore, this analytical method provides a better opportunity to differentiate between the two different network perspectives and make a comparison between units instead of within a single one. As the dependent variable is a count variable, a linear regression is not a relevant statistical method. This variable cannot take negative values so Poisson or Negative Binomial Regression are the two possible regressions conforming the analysis (Hausman, Hall, & Griliches, 1984).

(22)

21

Results

Descriptive Statistics

The results from the descriptive statistics and correlations between variables are presented in Тable 1. The mean count of patent application in 1997 is on average 97 (Mean = 97.04) while the sum of all the patents applied between 1990 and 2005, excluding the year of observation, reaches 1211 (Mean = 1210.73). This means that the number of patent applications for 1997 is relatively high when compared with the count for all years included in the database. Moreover, we notice that the standard deviations of all count variables which measure patents are exceedingly high. This indicates that firms in the sample vary substantially in their level of innovativeness. The variable distance-weighted reach also shows slightly higher standard deviation (SD=0.9; Mean=0.77). This means that there is a minor variance in firm’s position in their alliance portfolio and, on average, they operate in a quite dense portfolio network (distance-weighted reach < 1). All measures that have been log-transformed for the empirical analysis tend to exhibit better distributional properties.

(23)

22 Table 1. Descriptive Statistics

Hypotheses testing

Table 2 presents a summary of the results from the negative binomial regression used for the hypotheses testing. The table consists of 4 models. Model 1 displays only the effect of all control variables on the dependent variable. Model 2 includes control variables and the independent variables. In Model 3, I include only the moderator without the independent variable to allow comparison between their effects on dependent variable. Model 4 presents the interaction effect between the independent and the moderating variable on the dependent variable, including all control variables.

(24)

23 in Model 3 and 4, the level of significance of firm size lowers until it completely disappears in the last model. As both measures were initially used in the strategic management literature as an indicator of firm size (Lavie & Miller, 2008; Tsai, 2001; Roth & O'Donnell, 1996), I replaced the main control for firm size with number of subsidiaries in Model 1. The analysis indicated that the fit between all variables in the model diminished from this change. This additional examination indicates that in a study observing the alliance portfolio and its effects on firm performance, the number of subsidiaries as a measure of firm size is not so reliable control for patents as the yearly sales data although it presents an influential moderating effect in the following models. However, a measure regarding the organizational structure and size should not be underestimated as control for potential analytical bias in such observation focused on the alliance portfolio steam.

In model 2, I add to the analysis the independent variable distance-weighted reach, which indicates the level of centrality in firm’s alliance portfolio. There is no significant change in the coefficients of the control variables in this model. The independent variable shows a significant and positive interaction with the dependent variable. In line with hypothesis 1, we indicate that for a one unit change in the distance-weighted reach, firm’s patent applications in 1997 increased by almost 14% (β= 0.135). This positive and significant relationship between centrality in the alliance portfolio and patent applications implies that firms could benefit for their innovation performance from the level of centrality in their alliance portfolio positioning. Therefore, hypothesis 1 is strongly supported.

Although it was not hypothesized, Model 3 displays only the interaction between firm’s network of subsidiaries and its innovation performance. By excluding the variable measuring the level of centrality in alliance portfolio, we are able to investigate the individual effect of the count of subsidiaries on the number of firm’s patent applications. In this model for the first time we observe the change in the level of significance of the control variable firm size. It does not disappear entirely, but its diminishes while the coefficient decreases from β= 0.062** to β= 0.04*. In spite of this slight change, the results concerning the effect of the internal network size on innovation performance clarify that there is a positive and significant correlation between firm’s number of subsidiaries and its number of patent applications. The numbers show that for one unit change in the number of subsidiaries, we estimate an 8% increase the number of patent applications for 1997, other things being equal (β= 0.077). This can lead to the conclusion that firms could use their network of subsidiaries to improve their overall innovative performance by internalizing their innovative activities and using the benefits that its own network of subsidiaries provides to outperform their competitors.

(25)

24 variable is positive and significant, equal to β= 0,288, when firm’s count of subsidiaries is equal to 0. Moreover, the same positive and significant direct effect of the number of subsidiaries could be observed in the table, other things being equal (β= 0.158). These positive and significant results confirm that they are both from a strategic importance for firms that aim to achieve a superior performance. However, we can see that their interaction effect is negative and significant (𝛽= - 0.054). This implies that the high level of centrality in an alliance portfolio combined with large number of subsidiaries can result to a lower number of patent applications. Although individually both structural characteristics are beneficial for the focal firm, this result signifies that large number of subsidiaries lowers the importance of network centrality in leveraging firm’s innovative performance. Their combination in the model presents a substitution effect between the two different network perspectives. Therefore, hypothesis 2 is supported.

(26)

25 Table 2. Negative Binomial regression: Interaction effects for 1997

Robustness check

(27)

26 The second robustness check provided more interesting insights in terms of time variation. As the base model introduces the moderating effect of number of subsidiaries for the year 1997, I included the data from 2017 that was gathered during the data collection process. I extracted the count for 1997 from the one for 2017 and assembled a new variable which aimed to indicate the change in the network of subsidiaries in this 20 years’ time period. Additionally, I changed the dependent variable to the log-transformed measure available in the panel data that illustrates the effects of patent application for the year X with a year delay. This means that if the patent was applied in 1997, this variable will indicate the effect of the independent variable on patenting in the following year 1998. To include all years available in the data set in this robust check, I removed the if-clause from the regression and added the years from 1990 to 2005 to the model. Considering that in this model the dependent variable is log-transformed, the Negative Binomial Regression was replaced by xtreg command which is an appropriate analytical method for longitudinal observation of panel data set. Moreover, I used randomized effects to include the variation of the effects between entities. Table 3 presents the final results from the robust check with a comparison of the base model.

(28)

27 Table 3. Robustness check indicating the time lag significance

(29)

28

Discussion

In the following section, the results from the statistical tests will be discussed and possible reasoning why those results occurred will be given. Moreover, the research implications of the results as well as the one for managers in practice will be provided. In the last part of this section, the limitations of this research are presented and directions for future research are given.

Research in the field of alliance portfolios is predominantly focused on the structural configuration of the portfolio and its direct impact on the innovative performance of the focal firm (Wassmer, 2010). Organizations substantially differ in terms of size and structural composition which influences their innovative capabilities as well as their collaborative activities (Gulati, 1999). Therefore, the internal organizational scope and the overall internal structure was embodied in the alliance portfolio stream only as control measure indicating dissimilarities between firm’s size. However, the direct effect of firm’s internal organizational structure on the overall benefits which it could obtains from its portfolio partners was still underdeveloped. Following this line of reasoning, the aim of this study is to investigate whether а firm’s position in an alliance portfolio influences innovation performance and how this relationship is moderated by its number of subsidiaries. In their incentive to generate superior innovation performance, firms need to devote their capacity in maintaining one of the interfirm or intra-firm network collaborations which provides them with greater opportunities to stay compatible on the market.

(30)

29 helps organizations to overcome the problem with transferring tacit and complex knowledge to each other (Hansen, 1999; Tsai, 2001; Hansen & Løvås, 2004). Therefore, the central position in the alliance portfolio is not only a strategic advantage for firms in their innovation growth, it is a crucial feature in the external network configuration.

This study also sheds light on the strategic importance that the individual effect of organizational structure has on firm’s innovation output. Introduced by the impact that firm’s number of subsidiaries has on the relationship between alliance portfolio centrality and innovation performance, this research provides deeper understanding of the substitutional effect which the internal and external network have for firms’ innovation output. The results above indicated that both network centrality and network of subsidiaries have a highly significant positive effect on innovation performance. Both internal and external unit networks were previously seen as equally influential factor on firm’s innovation performance. However, as they were only separately examined in the research area, the literature stayed inconsistent about their interdependence. Although the central position in the alliance portfolio enhances firm’s overall innovation performance, companies own organizational structure contributes to the same superior capabilities such as the ease access to valuable external knowledge. In this contribution, the importance of the central positioning within an alliance portfolio could be preempt from its own internal network as it accelerates similar knowledge acquisition potential. Hereby, the analysis in this study confirmed their substitution effect suggested in hypothesis 2 by showing a negative and significant influence of the count of subsidiaries on the relationship between network centrality and innovation performance (𝛽 = - 0,054). Albeit significant, the substitution effect isn’t that large. The logic behind the given result could be encompassed to the sensitivity of the variables as they are both a propitious mechanism used from companies to improve their innovative performance.

(31)

30 to access the local market or in their need for essential regional knowledge and enhancing their own internal capabilities (Forsgren, Pedersen, & Foss, 1999).

Although firm’s central position in an alliance network position and the strength of ties play an important role in the organizational learning process (Gulati, 1999; Powell, Koput, & Smith-Doerr, 1996; Tsai, 2001), a successful innovative outcome can be achieved also inside the organizational structure by the network of subsidiaries. Subsidiaries are normally classified within the literature as one of the most advantageous sources from which multinational companies can obtain valuable external knowledge. A multiunit cross-border knowledge exchange is a reliable method used from companies in order to enhance their capability to gain and sustain competitive advantage (Hedlund, 1986; Hymer, 1976; Kogut & Zander, 1993). In MNCs, the valuable knowledge is geographically dispersed between all entities without the accompanying increase in alliance sharing costs (Caves & Caves, 1996; Foss & Pedersen, 2002; Gupta & Govindarajan, 2000). Moreover, the collaboration between subsidiaries allows each entity to learn from the others and benefit from knowledge developed outside its own borders but which have been transferred by usual organizational routines (Tsai, 2001).

(32)

31

Research Implications

This research is initiated deriving from the aim to bridge the gap between firm’s organizational structure and its external network role in firm’s ability to attain superior innovative performance. The study contributes to prior literature by adding the internal structural approach to the research field geared towards alliance portfolio concept. Organization’s size and structural composition have been mostly used as a control for possible dissimilarities in firm’s capabilities to achieve innovative performance. Therefore, this paper provides an indication that the network of subsidiaries can actually influence negatively the relationship between alliance portfolio centrality and its innovation performance. According to the resource-based view, organizations simultaneously engage in different alliance collaborations in order to obtain a vast amount of complementary resources and as external network provides better opportunity for them to realize economies of scale and scope. Hereby, this study adds the internal network approach as another reliable pathway for a firm to achieve a superior innovation performance as it indicated that the central position in an alliance network becomes less advantageous if the firm also possess a large network of subsidiaries. Moreover, relying on the internal collaborations could further reduce the transaction costs, as these costs, according to TCE, presents one of the substantial drawback alliance collaborative activities. Keeping in mind that most researches are focused either on one of these two perspective in their investigations, this research could be seen as a starting point in the adoption of the internal structural approach to the overall alliance stream and vice-a-versa.

Managerial Implications

(33)

32

drawbacks in firm’s innovative strategy. If firms combine these two means in their organizational planning, it may result in aggravating the overall innovation outcomes.

Limitations and Future Research

There are several limitations that could be found in this research. First, as it uses as a baseline Schilling’s recent study, which investigates the technological shock influence on the alliance activities spike after 1995, this study changes the longitudinal prior research focus to one that examines the effect within only one year. Using cross-sectional analytical method, this paper presents only one-year time frame of examination which can be seen as a drawback in terms of strategic implications. Hereby, a longitudinal study concerning the time period from 1990 to 2005 for the same relationship might contribute to broader conclusions. Due to lack of time and available resources, the year of observation 1997 may suffer from particular imperfections regarding up-to-date market strategies. Nowadays, the innovative markets frequently undergo some rapid changes because of all diversified market opportunities and various consumer needs. Hereby, additional data collection for patent applications and innovation activities for 2017 used as a dependent variable in future researches will somehow explain the outcomes from this particular research with reliable argumentation for the current market situation.

Since patenting norms and systems vary across regions, another limitation of this study can be found in the fact that data was only used for North American firms (Schilling, 2015). Future research might aim their focus on another innovative region to see if the results will stay robust when you adapt them to different regions with dissimilar regulation norms. For example, such investigation concerning the innovative performance of organizations operating on the Chinese market might provide different conclusions because of all geographical and cultural dissimilarities.

Another possible limitation could be the lack of industry control in the model. Excluding the industry variation from the analysis was done so that this research can provide a broader conclusions regarding the final results. Hereby, the focus of only one industry sector would narrow the overall research sample size. However, as industries differentiate in their innovative capabilities as well as in norms and regulations with regard to patent applications, a study assigned only to a particular industry can come up with dissimilar results for this interaction between firm’s internal and external networks. Future studies can focus on industries that characterize with high levels of innovativeness, for example pharmaceutical industry, to clarify if the main outcomes of this study differentiate between industries.

(34)

33 initial research objective of this study, providing interesting new insights crucial for the alliance portfolio stream. Additionally, differentiating between level of innovativeness of the alliance portfolio partners and subsidiaries can also provide an interesting conclusion from a managerial perspective. Such a comparison will ameliorate the resource division of each organization between these two networks.

Conclusion

This study aims to shed light upon the inclusion of internal organizational structure in the widely discussed alliance portfolio stream. The results show that, in general, firm’s subsidiaries are seen as an alternative channel from which organizations can obtain valuable external knowledge. Albeit this research does not aim to undermine the importance of network centrality, it provides an evidence that superior performance could be attained by firm regardless their position in their alliance portfolio. However, the positive outcome could be accomplished only when the organization already possesses an immense number of subsidiaries.

(35)

34

References

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

Aken, J. V., Berends, H., & Bij, H. v. (2012). Problem Solving in Organizations: A Methodological Handbook for Business and Management Students.

Anand, B. N., & Khanna, T. (2000). Do firms learn to create value? The case of alliances. Strategic management journal, 295-315.

Andersson, U., Forsgren, M., & Holm, U. (2007). Balancing subsidiary influence in the federative MNC: A business network view. Journal of International Business Studies, 38(5), 802-818.

Bae, J., & Insead, M. G. (2004). Partner substitutability, alliance network structure, and firm profitability in the telecommunications industry. Academy of Management Journal, 47(6), 843-859.

Balakrishnan , S., & Koza, M. P. (1993). Information asymmetry, adverse selection and joint-ventures: Theory and evidence. Journal of economic behavior & organization, 20(1), 99-117.

Baum, J. A., Calabrese, T., & Silverman, B. S. (2000). Don't go it alone: Alliance network composition and startups' performance in Canadian biotechnology. Strategic management journal, 267-294.

Beers, C., & Zand, F. (2014). R&D cooperation, partner diversity, and innovation performance: an empirical analysis. Journal of Product Innovation Management, 31(2), 292-312. Belderbos, R., Carree, M., & Lokshin, B. (2006). Complementarity in R&D cooperation

strategies. Review of Industrial Organization, 28(4), 401-426.

Birkinshaw, J. (1996). How multinational subsidiary mandates are gained and lost. Journal of International Business Studies, 27(3), 467-495.

Bouquet, C., & Birkinshaw, J. (2008). Weight versus voice: How foreign subsidiaries gain attention from corporate headquarters. Academy of Management journal, 51(3), 577-601.

Brikinshaw, J., & Hood, N. (1998). Multinational subsidiary evolution: Capability and charter change in foreign-owned subsidiary companies. Academy of management review, 23(4), 773-795.

Capaldo, A. (2007). Network structure and innovation: The leveraging of a dual network as a distinctive relational capability. Strategic management journal, 28(6), 585-608.

Caves, R. E., & Caves, R. E. (1996). Multinational enterprise and economic analysis. Cambridge university press, 1817.

Chung , S., Singh, H., & Lee, K. (2000). Complementarity, status similarity and social capital as drivers of alliance formation. Strategic management journal, 1-22.

Ciabuschi, F., Dellestrand, H., & Holm, U. (2012). The role of headquarters in the contemporary MNC. Journal of International Management, 18(3), 213-223.

(36)

35 Cohen, W. M., & Klepper, S. (1996). A reprise of size and R & D. The Economic Journal,

925-951.

Cohen, W. M., & Levinthan, D. A. (2000). Absorptive capacity: A new perspective on learning and innovation. Strategic Learning in a Knowledge economy, 39-67.

Colakoglu, S., Yamao, S., & Lepak, D. P. (2014). Knowledge creation capability in MNC subsidiaries: Examining the roles of global and local knowledge inflows and subsidiary knowledge stocks. International Business Review, 23(1), 91-101.

Coleman, J. S. (1990). Commentary: Social institutions and social theory. American Sociological Review, 55(3),, 333-339.

Colombo, M. G., Grilli, L., & Piva, E. (2006). In search of complementary assets: The

determinants of alliance formation of high-tech start-ups. Research policy, 35(8), 1166-1199.

Cowan, R., & Jonard, N. (2004). Network structure and the diffusion of knowledge. Journal of economic Dynamics and Control, 28(8), 1557-1575.

Cranmer, S. L., Desmarais, B. A., & Menninga, E. J. (2012). Complex dependencies in the alliance network. Conflict Management and Peace Science, 29(3), 279-313.

Das, T. K., & Teng, B. S. (2000). A resource-based theory of strategic alliances. Journal of management, 26(1), 31-61.

Deeds, D. L., & Hill, C. W. (1996). Strategic alliances and the rate of new product development: An empirical study of entrepreneurial biotechnology firms. Journal of business venturing, 11(1), 41-55.

Dias, J., & Magrico, V. M. (2011). The impact of resource conditions and environmental uncertainty on inter-firm alliance strategies. Applied Economics, 43(6), 757-765. Dong, J., McCarthy, K. J., & Schoenmakers, W. W. (2017). How central is too central?

Organizing interorganizational collaboration networks for breakthrough innovation. Journal of Product Innovation Management, 34(4), 526-542.

Doz, Y. L. (1996). The evolution of cooperation in strategic alliances: Initial conditions or learning processes?. Strategic management journal, 17(S1), 55-83.

Doz, Y. L., & Hamel, G. (1998). Alliance advantage: The art of creating value through partnering. Harvard Business Press.

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), 136-150.

Ernst, E. (2006). Herbal remedies for anxiety–a systematic review of controlled clinical trials. Phytomedicine, 13(3),, 205-208.

Faems, D., Janssens, M., & Neyens, I. (2012). Alliance portfolios and innovation performance: connecting structural and managerial perspectives. Group & Organization Management, 37(2),, 241-268.

Faems, D., Van Looy, B., & Debackere, K. (2005). Interorganizational collaboration and innovation: Toward a portfolio approach. Journal of product innovation management, 22(3), 238-250.

(37)

36 Fleming, L., King lll, C., & Juda, A. I. (2007). Small worlds and regional innovation. Organization

Science, 18(6), 938-954.

Forsgren, M., Pedersen, T., & Foss, N. J. (1999). Accounting for the strengths of MNC subsidiaries: the case of foreign-owned firms in Denmark. International Business Review, 8(2), 181-196.

Foss, N. J. (1997). Resources, firms, and strategies: a reader in the resource-based perspective. Oxford: Oxford University Press on Demand.

Foss, N. J., & Pedersen, T. (2002). Transferring knowledge in MNCs: The role of sources of subsidiary knowledge and organizational context. Journal of International Management, 8(1), 49-67.

Freeman, L. C., Roeder, D., & Mulholland, R. R. (1979). Centrality in social networks: II. Experimental results. Social networks, 2(2), 119-141.

Gardner, W., Mulvey, E. P., & Shaw, E. C. (1995). Regression analyses of counts and rates: Poisson, overdispersed Poisson, and negative binomial models. Psychological bulletin, 118(3), 392.

George, G., Zahra, S. A., Wheatley, 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), 205-226.

Ghoshal, S., & Bartlett, C. A. (1990). The multinational corporation as an interorganizational network. Academy of management review, 15(4), 603-626.

Gilsing, V., Nooteboom, B., Vanhaverbeke, W., Duysters, G., & van den Oord, A. (2008). Network embeddedness and the exploration of novel technologies: Technological distance, betweenness centrality and density. Research policy, 37(10), 1717-1731. Goerzen, A. (2005). Managing alliance networks: Emerging practices of multinational

corporations. The Academy of Management Executive, 19(2), 94-107.

Griliches, Z. (1991). The search for R&D spillovers (No. w3768). National Bureau of Economic Research.

Gulati, R. (1995b). Social structure and alliance formation patterns: A longitudinal analysis. Administrative science quarterly, 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., & Higgins, M. C. (2003). Which ties matter when? The contingent effects of

interorganizational partnerships on IPO success. Strategic Management Journal, 24(2), 127-144.

Gulati, R., Khanna, T., & Nohria, N. (1994). Unilateral commitments and the importance of process in alliances. Sloan management review, 35(3), 61.

Gupta, A. K., & Govindarajan, V. (2000). Knowledge flows within multinational corporations. Strategic management journal, 473-496.

Hagedoorn, J. (1993). Understanding the rationale of strategic technology partnering: Nterorganizational modes of cooperation and sectoral differences. . Strategic management journal, 14(5), , 371-385.

(38)

37 Hansen, M. T., & Løvås, B. (2004). How do multinational companies leverage technological

competencies? Moving from single to interdependent explanations. Strategic Management Journal, 25(8‐9), 801-822.

Hargadon, A. (2003). How breakthroughs happen: The surprising truth about how companies innovate. Harvard Business Press.

Harrigan, K. R. (1985). Vertical integration and corporate strategy. Academy of Management journal, 28(2), 397-425.

Hausman, J. A., Hall, B. H., & Griliches, Z. (1984). Econometric models for count data with an application to the patents-R&D relationship.

Hedlund, G. (1986). The hypermodern MNC—a heterarchy? Human resource management, 25(1), 9-35.

Hoang, H., & Rothaermel, F. T. (2005). The effect of general and partner-specific alliance experience on joint R&D project performance. Academy of Management Journal, 48(2), 332-345.

Hoffman, K. (. (2007). Banach spaces of analytic functions. . Mineola, New York: Courier Corporation.

Hymer, S. H. (1976). International operations of national firms. MIT press.

Jaffe, A. B., Trajtenberg, M., & Henderson, R. (1993). Geographic localization of knowledge spillovers as evidenced by patent citations. Tthe Quarterly journal of Economics, 108(3), 577-598.

Jandik, T., & Kali, R. (2009). Legal systems, information asymmetry, and firm boundaries: Cross-border choices to diversify through mergers, joint ventures, or strategic alliances. Journal of International Business Studies, 40(4), 578-599.

Jiang, R. J., Tao, Q. T., & Santoro, M. D. (2010). Alliance portfolio diversity and firm performance. Strategic management journal, 31(10), 1136-1144.

Kale, P., & Singh, H. (2007). Building firm capabilities through learning: the role of the alliance learning process in alliance capability and firm‐level alliance success. Strategic

management journal, 28(10),, 981-1000.

Kale, P., Dyer, J. H., & Singh, H. (2002). Alliance capability, stock market response, and long‐ term alliance success: the role of the alliance function. Strategic management journal, 23(8), 747-767.

Katila, R. (2002). New product search over time: past ideas in their prime?. Academy of Management Journal, 45(5), 995-1010.

Katila, R., & Ahuja, G. (2002). Something old, something new: A longitudinal study of search behavior and new product introduction. Academy of management journal, 45(6), 1183-1194.

Kim, C. S., & Inkpen, A. C. . (2005). Cross-border R&D alliances, absorptive capacity and technology learning. Journal of International Management, 11(3), 313-329.

Kim, J. W., & Higgins, M. C. (2007). Where do alliances come from?: The effects of upper echelons on alliance formation. Research Policy, 36(4), 499-514.

Kogut, B., & Kulatilaka, N. (1994). Options thinking and platform investments: Investing in opportunity. California Management Review, 36(2), 52-71.

Are internal and external networks mutually exclusive? How firm’s organizational structure influences the relationship between alliance portfolio and innovation performance