Rijksuniversiteit Groningen 20 January 2014 Word count: 14780
The Effect of Alliance Portfolio Diversity on Firm
Performance: Evidence from the Oil and Gas Industry
Author: M. Kwakernaak
First supervisor: P.M.M. de Faria Second supervisor: dr. K.J. McCarthy
University of Groningen Faculty of Economics and Business Master of Business Administration
1
The Effect of Alliance Portfolio Diversity on Firm
Performance: Evidence from the Oil and Gas Industry
Kwakernaak, Marnix
Faculty of Economics and Business, University of Groningen, The Netherlands
Few studies examined the relationship between alliance portfolio diversity and firm performance. Scholars used a variety of dimensions to address the alliance portfolio diversity concept (e.g. partner diversity, geographical diversity, functional diversity, governance diversity). However, mixed results were found of the influence of alliance portfolio diversity on firm performance outcomes. Building on the transaction costs theory and the resource-based view, this study extents prior research on the relationship between alliance portfolio diversity and firm performance, using a multidimensional alliance portfolio diversity approach. I argue that increasing diversity in partner’s industrial, organizational, and national background, has an inverted U-shape influence on firm performance, and that higher alliance portfolio functional diversity increases, and higher alliance portfolio governance diversity decreases firm performance. Hypothesises were tested with a balanced panel dataset consisting of 35 large firms in the oil and gas industry, during a ten-year period from 2001 to 2010. The findings of this show that higher functional diversity positively influence firm performance (p<0.5). No significant results are found for an inverted U-shape relation between the partner diversity constructs and firm performance, neither for a negative influence of increasing governance diversity on firm performance. In line with previous research, this paper confirms the positive influence of a higher functional diverse alliance portfolio on firm performance, suggesting that by means of balancing both exploitation and exploration activities in an alliance portfolio, firms have access to valuable alliance partner’s resources, which can result in increased firm performance.
Keywords: Alliance portfolio diversity, partner diversity, functional diversity, governance diversity,
oil and gas industry
1. INTRODUCTION
2 be used to enter new markets (George et al., 2001). Many firms enter multiple alliances at the same time. A set of simultaneously managed alliances is commonly referred to as an alliance portfolio (Kale & Singh, 2009). Hoffman (2007) argued that researching the firm performance implications of the sum of a firm’s single alliances may give another research result, than examining the whole bundle of alliances that a firm holds simultaneously. For instance, a single alliance partner may not contribute to the net profit of a firm, but can provide valuable knowledge or ideas which can be used in the longer term. Hoffman (2007) argues that alliance portfolio management has become an important strategic issue.
In line with previous literature, this paper defines an alliance portfolio as a collection of multiple alliances with
different partners simultaneously held by a focal firm (Bae & Gargiulo, 2004). Hereby, this study takes an egocentric
view focussing on all the alliances that a focal firm holds on a certain moment in time (Hoffman, 2007).
Different characteristics of an alliance portfolio are investigated by former scholars. For instance, one literature stream looked at the implication of alliance portfolio size (Powel et al., 1996; Vassolo et al., 2004), other scholars argued that not only the size of the alliance portfolio matters, but stressed the importance of the composition of alliance partners in a portfolio, (Darr & Kurtzberg, 2000; Duysters & Lokshin, 2011; Wuyts et al., 2004). Prior research examined the differences and similarities among the alliance partners, regarding their industrial or national background, functional purpose, or the way that the are governed. However, they found mixed results on the way diversity among alliance partners influence firm performance (Goerzen & Beamish, 2005; Lavie et al., 2012; Lavie & Rosenkopf, 2006; Sarkar et al., 2009). Hence, the goal of this paper is to strengthen the alliance portfolio diversity literature by researching the effect of alliance portfolio diversity on a firm performance. This paper use a multidimensional alliance portfolio diversity view that incorporates partner diversity, functional diversity, and governance diversity.
Hence, the research question for this paper is: to what extent higher alliance portfolio diversity influences firm
performance?
The following theoretical and managerial implications are addressed in this study. First, it will extent the alliance portfolio diversity literature, in an attempt to find evidence for the alliance portfolio diversity – firm performance relationship in the oil and gas industry. Former empirical work has been done in the bio technology sector (Bruyaka & Durand, 2012; Baum et al., 2000), Automotive sector (Jiang et al., 2010), and has focussed on technology alliances (Neyens & Faems, 2013), or has taken a multi-sector approach (Sarkar et al., 2009; Duysters et al., 2012; Collins, 2013; Cui & O’Connor, 2012; Goerzen & Beamish, 2005). Due to increasing role strategic alliances have regarding firm performance of oil and gas firms (Green, 2003), the high international character of the oil and gas industry (Levy & Kolk, 2002), and alliance portfolio strategies which tend to focus on operational efficiency and exploration activities (Silvestre & Dalcol, 2010), this paper will examine the alliance portfolio diversity – firm performance relationship in the oil and gas industry.
3 portfolio management, by means of researching to what extent different diversity degrees influence a firm’s financial performance. Reuer & Ragozzino (2006) noted in their study that the alliance portfolio of an average large firm, with over 30 alliances, can count for 6-15 percent of a firm’s value, thereby stressing the managerial importance of alliance portfolios in today’s businesses.
Based on the resource-based view, and the transaction costs theory, this study hypothesized the influence of the different alliance portfolio diversity constructs on firm performance. From a resource-based view, firms are assumed to form strategic alliances in order to extract external resources and capabilities which are valuable, rare, difficult to imitate, and non-substitutable (Das & Teng, 2000). A more diverse set of alliance partners can increase the likelihood that a firm is able to access valuable new resources and capabilities, complementary to a firm’s exciting bundle of resources. This can increase a firm’s competitive advantage and delivers better financial performance (Sivakumar et al., 2011).
From a transaction costs perspective (Williamson, 1996), it is argued that firms choose to form strategic alliances because this governance structure offers stronger incentives and adaptive capabilities, and more administrative control, compare to respectively, integrating external resources or market transactions (Tsang, 2000). This theory mainly focuses on the transaction costs of forming and maintaining alliances. Goerzen & Beamish (2005) suggested that by due to increased diversity in a firm’s alliance portfolio, transaction costs and coordination costs also increase substantially. This may hamper the net financial benefits a firm intent to derive by means of its alliance portfolio.
This study used a panel dataset of 35 oil and gas firms to examine the influence of higher alliance portfolio diversity of firm performance. The findings show that a more functional diverse alliance portfolio can have a substantial positive influence on firm performance. Thereby, stressing the importance of balancing partners with different functional purposes in a firm’s alliance portfolio. I found no significant results concerning the influence of alliance portfolio partner, and governance diversity on firm performance.
The paper is structured as follows. First, a review is given of the existing literature on alliance portfolio diversity, and by means of a resource based view and transaction costs perspective, hypotheses are developed on the relationship between partner diversity, functional diversity, governance diversity and firm performance. Second, the research methodology is discussed. Subsequently, the analyses and results are presented. The final section will discusses the results, point out the main theoretical and managerial implications of the findings, addresses the limitations of this study, and suggests directions for future research.
2. THEORY AND HYPOTHESES
4 In turn, the value of the firm’s strategic alliances, seen from a portfolio view, may differ compared to the value of the sum of the set of individual alliances (Vassolo et al. 2004). This underlines the basic assumption and motivation of this study. According to Darr & Kurtzberg (2000), an important determinant of a successful alliance portfolio lies in the similarities and differences between the alliance partners. This study defines alliance portfolio diversity as the
heterogeneity of resources, capabilities, and knowledge present across the portfolio of partnership of the focal firm
(Collins, 2013).
One of the main motivations for firms to build alliance portfolios is to access, integrate, and leverage valuable partner resources, which are not available in their current resource stock (Ahuja, 2000; Gulati, 2007; Hoffmann, 2007; Lavie, 2006). At the same time, alliance portfolios are used to mitigate risks and uncertainty created by a firm’s external environment (George et al., 2001; Hoffmann, 2007).
Previous research found mixed results on the influence of differences and similarities among the alliance partners on the performance of a firm. On the one hand, for instance Baum et al. (2000) found that an alliance portfolio that consist of more diverse alliance partners, increases firm performance due to an increased variety of resources available to the a firm. This was mainly due to increasing innovation performance. The innovation performance implication of increasing alliance portfolio diversity was confirmed by the study of Faems et al. (2005). Mouri et al. (2012) found that a high functional diverse horizontal alliance portfolio positively affects the initial public offering in financial markets. Terjesen et al. (2011) researched the effect of manufacturing capabilities on venture performance, and concluded that by means of alliance portfolio partner diversity and geographical diversity the utilization of manufacturing skills in young high-technology firms is enhanced. Moreover, Jiang et al. (2010) found that firms learn to leverage partner’s complimentary, and supplementary resources and capabilities via higher alliance portfolio diversity. Additionally, diverse alliance portfolios can provide advantages in terms of endorsement and legitimacy, through alliance partners with good public reputations or a specific public status (Gargiulo & Benassi, 2000; Koka & Prescott, 2002).
5 partner’s culture, alliance experience, challenges of knowledge appropriation, and the need for synergy and integration. In other words, the influence of alliance portfolio diversity on firm performance seems to be ambiguous.
2.1. Resource-based view and Transaction cost theory
Traditionally, the resources-based view (Barney, 1991; Penrose, 1959; Peteraf, 1993; Wernerfelt, 1984) argued that internal resources and capabilities determine the value creation of a firm. In particular, if the resources and capabilities are valuable, rare, difficult to imitate, and non-substitutable, they form the basis of the firm’s competitive advantage (Barney, 1991). According to Das & Teng (2000) the accumulated resources are critical for a firm’s competitive strategy. Building on the resource-based view, this study argues that via the set of alliances, a firm potentially gain access to valuable resources, which would not be available to a firm without these specific alliance partners (Das & Teng, 2000). These alliance partner’s resources may encompass financial assets, human resources, R&D investments, marketing efforts, reputation, skills, or specific knowledge (Lavie et al., 2007, 2008). By means of combining the complementary resources available through a diverse set of alliance partners in the right way, a firm can increase its firm performance and enhance its sustainable competitive advantage (Sivakumar et al., 2011).
6
2.2. Partner diversity
A larger diversity of alliance partners can offer a firm a wider variety of resources and knowledge to a firm (Cui & O’Connor, 2012). This study defines alliance portfolio partner diversity as the degree to which the focal firm’s alliance
portfolio partners are different in industrial background, country of location, and organizational structure.
Former studies showed mixed results regarding the influence of alliance partner diversity on firm performance. Some scholars argued that firms with different types of alliance partners provide access to distinct skills, unique information and knowledge, capital, and some alliances can provide legitimacy (Dodgson, 1992; George et al., 2002). A diverse set of alliance partners can have a positive influence on a firm’s innovativeness, because partners can offer the opportunity to aquire new knowledge and solutions, which can be used for developing new technologies and products (Swaminathan & Moorman 2009; Wuyts et al., 2004). Besides, access to diverse resources increases a firm’s flexibility to coop with a changing external environment (Sampson, 2007). By virtue of combining new resources and knowledge a firm can increase its learning capabilities and become better in utilizing external knowledge, which in turn, can increase a firm’s financial performance (Baum et al., 2000).
However, previous research also suggested that high partner diversity can have negative outcomes for a firm. If the available alliance portfolio knowledge is too diverse, it can impede routine building and hamper knowledge utilization (Vasudeva & Anand, 2011). Goerzen & Beamish (2005) found that high degrees of national and industry diversity substantially increases transaction and coordination cost. This had a negative effect on the firm’s corporate performance. Faems et al. (2010) confirmed the cost implications, and observed a substantial firm level cost increase as a result of increasing technological alliance portfolio diversity.
As a consequence of the ambiguous findings regarding the effect of higher alliance portfolio partner diversity on firm performance, this study expects a curvilinear relationship. In line with the previous discussion, this study argues that increasing partner diversity can provide access to a valuable pool of external resources, which positively influences firm’s financial returns. However, after some threshold, it is expected that higher levels of partner diversity result in increasing transaction and coordination costs, which outweigh the financial benefits drawn from a diverse set of alliance partners. This study takes a multidimensional partner diversity approach, and examines partner diversity along the following dimension; industry diversity, national diversity, and organizational diversity.
2.2.1. Industry diversity
A firm’s alliance portfolio can consist of partners from different industries. This study defines alliance partner industry diversity as the degree to which the focal firm’s alliance portfolio partners differ in industrial backgrounds.
7 enhance firm performance, either via growing productivity (Oum et al., 2004), or an increased ability to value and apply alliance partner’s knowledge (George et al., 2001).
Alliances formed between competing firms, supposedly have large overlap concerning backgrounds, experiences, knowledge, and technological bases. This enables a firm to learn new skills, acquire the appropriate knowledge, and potentially imitate the partners operating processes or output more easily (Cohen & Levinthal, 1990). On the contrary, too much overlap in knowledge bases will inhibit learning opportunities, because the alliances partner does not possess new valuable knowledge (Grant, 1996). Besides, such such alliances are likely to miss a sufficient level of complementarities (Nakamura et al., 1996), which may result in a lack of synergy, or conflict of interest between partners due to the same business goals (Parise & Casher, 2003). Mowery et al. (1996) found in the biotechnology industry, that partners from non-competing industries experienced significant higher levels of knowledge transfer, which can induces potential firm benefits (Dussauge et al., 2004; Dyer & Hatch, 2006; Khanna et al., 1998; von Hippel, 1988). So, on the one hand, allying with competitors can increase the relative easy to which firms can learn industry specific organizational processes and skills. On the other hand, alliance partners form non-competing industry can provide new knowledge and learning opportunities. Hence, from an alliance portfolio point of view, balancing the degree of competing and non-competing supposedly can be enhance benefits for a firm.
However, firms should be careful with increasing industry diversity in their alliance portfolios. Costs may increase, due to difficulties in aligning partner’s strategic goals, and the additional complexity of coordinating scare resources among multiple partners (Hoang & Rothaermel, 2005). Alliance partners with different industrial backgrounds can also experience difficulties in sharing knowledge, because organizational processes and routines are to some extent different (Jiang et al., 2010) Moreover, high industry diversity increases the complexity of managing the alliance portfolio (Hoffman, 2005). McGill & Santoro (2009) confirmed that dispersion of knowledge between alliance partners makes it hard to acquire partner’s knowledge, build knowledge sharing routines, and manage the complexity of a diverse alliance portfolio.
So consequently, firms will probably benefit from an industry diverse alliance portfolio, because it increases the likelihood that diverse partners provide complementary, industry specific resources and skills via competing alliance partners, and simultaneously provide novel knowledge via non-competing alliance partners. However, if the industry diversity increases, also the alliance portfolio difficulties in goal alignment, building knowledge sharing routines, and coordination increases. This assumable causes increasing transaction costs which at some point, may outweigh benefits of an increased industry diverse alliance portfolio. Therefore, this study hypothesizes:
8
2.2.2. National diversity
Nowadays, due to an increasing globalization, more firms are looking for business opportunities across national boundaries. This movement is also reflected in strategic alliance formation. A vast amount of literature examines the impact of firm’s international expansion and cross-border alliances. However, the impact of increasing internationalization of alliance portfolios is still insufficiently highlighted (Lavie & Miller, 2008).
This study defines national diversity in alliance portfolios as the degree to which a firm’s alliance portfolio
partners originate from different national backgrounds. National diversity in alliance portfolios is reflected in, for
instance, geographical distance, institutional differences, cultural differences, or in differences regarding economic development (Ghemawat, 2001).
A national diverse alliance portfolio can provide several benefits to a firm. International strategic alliances are a pre-eminent way for firms to facilitate business expansion into new markets (Duysters & Lokshin, 2011). An essential argument of the resource-based view, suggest that there are no natural limits for a firm to leverage and exploit its recourses across new geographical areas, as long as the strategic assets of the firm remain value, rarity, inimitability, and non-substitutable (Barney, 1991; Goerzen & Beamish, 2003). In particular, upstream and downstream alliances with foreign partners offer access to new sources of technologies, extend the exposure and market reach of current owned resources, and can enhance firm’s economies of scale and scope (Contractor & Lorange, 1988; Eisenhardt & Schoonhoven, 1996; Tallman & Li, 1996). Terjesen et al. (2011) found that geographical diverse alliance portfolios enhance the effectiveness of the managerial capabilities, which can enhanced venture performance. Additionally, international partnering provides the opportunity to access distinct foreign resources, information, and capabilities (Wassmer, 2010). Moreover, an national diverse alliance portfolio increases the opportunity to leverage a firm’s current strategic assets and increases the scope of new external resources. This can result in reduced risk and uncertainty, as well as greater opportunities to adapt and respond to changing external market conditions (Hagedoorn, 1993; Kogut & Kulatilaka, 1993; Powell et al., 1996). Lavie & Miller (2008) argued that foreign alliance experience can help to overcome national differences and increase the firm’s capacity to manage international alliances. As a consequence, the firm will be better able to utilize the international resources, which can positively influence firm performance. Besides, evidence was found that international acquired knowledge increase a firm’s capability of successfully introducing new products in foreign markets (Subramaniam & Venkatraman, 2001).
9 effect on a firm’s financial performance (Denis et al., 2002). Hence, a national diverse alliance portfolio may result in increasing alliance portfolio costs. Negative outcomes are confirmed by studies which found that a too diverse national alliance portfolios causes a decrease in innovation performance (Sivakumar et al., 2011). In similar vein, it may also negatively influence firm’s economic performance, because of increased managerial and organization complexity (Denis et al., 2002; Goerzen & Beamish, 2005). Lavie & Miller (2008) performed a research in the U.S. software industry, and found that increased alliance portfolio internationalization creates a lack of absorptive capacity and coordination problems which decreases financial performance.
Consequently, increasing national alliance portfolio diversity enables a firm to access potential new valuable resources and knowledge as well as offers the opportunity to exploit the current owned resources to foreign markets. However, if national diversity becomes too high, social and cultural differences may impede knowledge sharing, and create communication and coordination problems, resulting in increasing transaction costs. In turn, decreasing access to external knowledge and increasing managerial costs of holding a large national diverse alliance portfolio are expected to outweigh the financial benefits after a certain threshold of national diversity is reached. Hence, this study hypothesizes:
Hypothesis 1b: Alliance portfolio partner national diversity has an inverted U-shaped effect on firm performance, where, as alliance portfolio partner national diversity increases, firm performance first increases, and then decreases.
2.2.3. Organizational diversity
Organizations differ among each other as a result of different management styles, diverging reactions to a changing external environment, and different routines to coop with internal processes and tasks (Lavie et al., 2012). In addition, one firm may differ from another due to a distinct corporate culture, and divergent strategic goals (Parkhe, 1993).
Organizational diversity in alliance portfolios may emanate from alliances formed between parent corporations and subsidiary, or public and private organization (Jiang et al., 2010). Prior literature recognized the heterogeneity between private and public sector (Rainey & Bozeman, 2000). Saz-Carranza & Longo (2012) argued that private actors are mainly focused on efficiency, compared to governments, which basic goals are to provide legitimacy. Universities are primarily focused on teaching and scientific research, where transfer of knowledge and technologies is still lagging behind (Etzkowitz, 2003).
The study defines alliance partner’s organizational diversity as the degree to which alliance portfolio partners
differ in organizational ownership and organizational level.
10 alliance partners. Universities can provide new knowledge through sector related basic science; this can lower R&D expenditures, and give access to critical human resources (Lavie, 2007; Oliver & Liebeskind, 1998). Government partners can accommodate legitimacy, new business opportunities, and sometimes provide financing contacts (Pangarkar & Wu, 2013). Moreover, governments might be resource rich and powerful alliance partners, through which firms acquire financial support in order to mitigate organizational costs (Bae & Gargiulo, 2004). Hence, this study expects that an alliance portfolio with higher diversity in partner’s organizational backgrounds, provides more access to distinct, organizational specific resources and processes, and learning opportunities, which will enhance the firm performance. This was confirmed by the study of Jiang et al. (2010), who found that a wider exposure to partner’s organizational diverse backgrounds, has a positive influence on firm performance.
On the contrary, due to the divergent objectives and conflict of interests between partners, it becomes more difficult to draw alliance contracts, align partner goals, develop joint decision making, communication, and knowledge transferring routines (Piva & Rossi-Lamastra, 2013). As a result of organizational diversity, alliance partners develop different attitudes towards a changing external environment (Freeman, 1984). For instance, a private actor wants to seize a new business opportunity; however, the public actor may be cautious due to its public role. Tension and conflict of interest may arise, which hinder the collaboration (Lavie et al., 2011). Besides, the characteristics of the embedded knowledge may vary across different types of organizations. For example, scholars found in alliances between private firms and universities, partners experience difficulties to build knowledge-sharing routines due to distinct attributes of their internal knowledge (Piva & Rossi-Lamastra, 2013). Organizational differences are known to hamper the transfer of resources between strategic alliance partners (Huxham & Beech, 2003).
Thus, this study expects that an alliance portfolio with different types of organizations increases the firm performance, due to the increases likelihood that a firm acquires access to valuable resources, novel knowledge, and new business opportunities to exploit the current resource base. However, if organizational diversity in an alliance portfolio becomes too high, transaction costs of aligning organizational goals, processes, and knowledge sharing may outweigh the financial returns. Therefore, this study hypothesizes:
Hypothesis 1c: Alliance portfolio partner organizational diversity has an inverted U-shaped effect on firm performance, where, as alliance portfolio partner organizational diversity increases, firm performance first increases, and then decreases.
2.3. Functional diversity
11 alliances may represent the same or different functional purposes (Cui & O’Connor, 2012). This paper defines alliance portfolio functional diversity as the degree to which a firm’s portfolio alliances differ in functional purposes.
In literature a common distinction is made between alliance functional purposes based on exploration or exploitation activities. R&D alliances are used to enlarge firm’s technological capabilities, and to acquire new knowledge, therefore they are commonly exploratory in nature. Exploration alliances involve activities intend to discover new business opportunities, get access to specific technological capabilities, and to acquire new knowledge to build unique capabilities and enhance innovation (Koza & Lewin, 1998; Rothaermel, 2001). These alliances intend to supply innovative business ideas in order to secure future financial returns, therefore enhances a firm’s long-term competitive advantages (March, 1991). Manufacturing, distribution, and marketing alliances are generally used to utilize existing resources, and therefore exploitative in nature (March, 1991; Santoro & McGill, 2005). Exploitation activities are aimed at leveraging, integrating, and implementing existing resources and knowledge in order to capitalize the current strategic assets of a firm (Koza & Lewin 1998; Lavie & Rosenkopf, 2006; Rothaermel, 2001; Rothaermel & Deeds, 2004). By doing so, a firm wants to commercialize its current set of resources in order to generate short-term financial returns (Koza & Lewin, 1998; Rothaermel, 2001; Rothaermel & Deeds, 2004).
Hence, it is argued that exploration and exploitation activities are both essential for organizations (March, 1991). When an alliance portfolio contains excessive exploration activities, a firm bares the risk of investing too much resources in new business opportunities, and simultaneously lacking the capacity to increase financial performance (Uotila et al., 2009). Focussing too much on exploitation alliance activities creates the risk of ending up without a competitive advantageous in the long run (March, 1991). By virtue of balancing alliance partners with different functional purposes in an alliance portfolio, a firm will be able to simultaneously exploit its current internal resources, and explore new knowledge and unique external resources. For instance, Mouri et al. (2012) found that increased alliance portfolio functional diversity has a positive influence on the shareholder value of a firm.
However, increasing the diversity may also result in increasing communication difficulties between partners with different functions (Oxley & Sampson, 2004). Higher functional diversity lowers the degree of knowledge overlap between the partners, and can potentially limit the degree of knowledge transfer across alliances (Lane & Lubatkin, 1998). Although beyond the scope of this study, Cui & O’Connor (2012) argued that to overcome conflicts of interest, divergent organizational routines, and coordination problems in functional diverse alliance portfolios, firms should also take into account former partner experience. This can help to lower initial alliance costs and increase the firm’s financial returns from a balanced alliance portfolio (Lavie et al., 2011). It is expected that firms learn to mitigate potential increasing transaction costs of a more diverse functional alliance portfolio and generate increasing net financial returns.
12
Hypothesis 2: Increasing alliance portfolio functional diversity is positively associated with firm performance.
2.4. Governance diversity
Different governance forms can be applied to the alliances in a portfolio. Non-equity based alliances are solely managed through contracts, while in equity-based alliances one firm employs a certain percentage of equity ownership into another firm (Zollo et al., 2002). In line with this study this involves a minority ownership stake of less than 50 percent, otherwise it is perceived as an acquisition. The partnering firms can also create a separate entity, which is called a joint venture. The joint venture is subject to joint control of the parent firms (Shenkar & Zeira, 1987). A specific alliance governance structure has consequences for the functioning of an alliance and the final returns for the partner firms (Gulati & Singh, 1998). Governance specific capabilities applied in a distinct governance structure may vary regarding communication routines, control enforcing mechanisms, knowledge sharing processes, and degree of commitment (Kogut, 1988; Sarkar et al., 2009). Different alliance governance structures can be applied simultaneously in a firm alliance portfolio, and each affects the functioning of an alliance on a different way (Gulati & Singh, 1998; Oxley, 1997; Pisano, 1989). This paper defines governance diversity as the degree to which
alliance governance types are heterogeneously present in a firm’s alliance portfolio.
Alliance portfolio diversity in governance structures can both have positive and negative effects on a firm. On the one hand, engaging in new alliance governance structures enables a firm to gain new experience in managing different alliance structures. Former researchers found that alliance experience (e.g. governance structure experience) increases a firm’s ability to control a specific type of alliance, creates learning opportunity via standardized knowledge sharing routines (Levinthal & March, 1993; Dekker & Van den Abbeele, 2010). Increasing the diversity of governance structures applied in an alliance portfolio, can therefore induce new learning opportunities, and enhance alliance managements capabilities.
However, due to the lack of initial governance specific experience, costs of learning and managing a new type of governance structure may increase, which can negatively influence a firm’s financial performance (Madhok, 2002; Sampson, 2005). At the same time, when firms apply new governance structures, they have to apply unfamiliar coordination mechanisms and safeguarding structures. Especially, when firms make alliance specific investments, the increased uncertainty that their resources will become obsolete, increases the costs of contracting, coordination, and monitor the alliance (Subramanian et al., 2006). Hence, it is expected that applying more diverse governance structures in an alliance portfolio will increase a firm’s transaction costs.
13 exploit a partner’s resources via institutionalized knowledge sharing routines and applying governance specific managerial capabilities, which more likely results in increasing financial returns for the focal firm (Sampson, 2005). Although the potential learning opportunities, of a more diverse governance alliance portfolio, this study assumes that via increasing safeguarding and managerial costs, higher alliance portfolio governance diversity will have a negative influence on firm performance. Therefore, this study hypothesizes:
Hypothesis 3: Increasing alliance portfolio governance diversity is negatively associated with firm performance.
2.5. Oil and gas industry
This paper attempts to find evidence for the alliance portfolio diversity – firm performance relationship in the oil and gas industry. Strategic alliances are known to be commonly used in the biotechnology industry, telecommunications, medical devices, and software industry (Baker et al., 2008). Also firms from the oil and gas sector are using strategic alliances more prevalent. Because of increasing operation costs and falling oil and gas prices, these companies are more often starting to collaborate and form strategic alliances, thereby reducing costs by eliminating redundant practices and increasing the value of operation assets (Green, 2003). Ernst & Steinhubl (1997) found in a survey among senior managers in large oil and gas companies, that a vast majority of the managers expected strategic alliances to become the main source of increasing firm performance. Hence, this study will use this particular industry setting to research the impact of alliance portfolio diversity on firm performance. This study will extent the alliance portfolio literature in an attempt to find evidence for the findings of jiang et al. (2010), who also research the alliance portfolio diversity – firm performance relationship, using the similar multidimensional alliance portfolio diversity approach. Jiang et al. (2010) examined the automotive industry, and in order to extent the generalizability of their findings, this study will investigate the oil and gas industry. The following section will explain why the oil and gas industry is significantly different compared to the automotive industry, and why it is relevant as a research setting to examine the influence of alliance portfolio diversity on firm performance.
The main similarity between the automotive industry and oil and gas industry, is that both are matured industries. Moreover, both industries have an oligopolistic structure (Alper & Mumco, 2007; Levy & Kolk, 2002), which implies, that a few large players in the industry dominate the market (Alper & Mumco, 2007). However, several industry differences are addressed in the next section, explaining why the oil and gas industry distinct itself from the automotive industry.
14 together with competitors. Therefore, it is expected that due to the different structural design of alliance formations between the industries, the degree of functional diversity and partner diversity in alliance portfolios may differ.
Second, the institutional environment of the oil and gas sector is highly international in scope. The tendency to internationalize seems essential for oil and gas firms, in order to coop with the increasing competition (Silvestre & Dalcon, 2010). Stable relations with foreign partners, which are based on trust and efficient knowledge exchange, contribute to the firm’s learning process. This allows them to acquire the necessary technological capabilities, facilitate access to new markets, handle market challenges, and carry out innovations, which is essential for sustainable competitive advantages of oil and gas firms (Bell & Albu, 1999; Bell & Giuliani, 2007). Grant & Cibin (1996) pointed out that oil and gas companies tend to adopt a global, rather than multi-domestic alliance strategy in their production and refining operations. This is mainly due to a uniform international oil and gas price (Levy & Kolk, 2002). On the contrary, the automotive firms tend to focus on local and regional price competition, and marketing strategies, which determine the success of automotive companies (Löffler & Decker, 2012). Additionally, in relation to firm’s market entry opportunities, the oil and gas companies tend to focus their alliance strategy on political strategies due to the (inter)national ownership of specific oil and gas fields and government owned oil and gas companies (Levy & Kolk, 2002), whereas the automotive industry focus on national or regional competition and growth opportunities (Kumar & Waheed, 2007). Consequently, oil and gas firms are more likely to collaborate with international institutional partners, compared to the automotive industry, where firms are more likely have to coop with national or regional institutions. Hence, it is expected this will be reflected in the alliance diversity degree related to the type and national backgrounds of alliance partners.
Third, in both industries, firms benefits significantly from increased operational productivity. However, contrary to the oil and gas industry, firms from the automotive industry are increasingly focussing on product innovation over the last years (Requena-Silvente & Walker, 2009). It is expected that this tendency is reflected in the alliance portfolio strategies of car manufacturers, moving towards more marketing collaborations, instead of partnering with competitors, who may imitate product innovation outcomes. At the same time, alliance networks of oil and gas companies are more focussing on joint exploration and exploitation activities with competitors. Particular concentrating on technological innovations, and exploring new oil and gas fields to develop (Ernst & Steinhulb, 1997; Silvestre & Dalcon, 2010). In order to stay competitive in the automotive industry, it is important to avoid innovations to be imitated by competitors (Dyer & Nobeoka, 2000). Hence, car manufactures more likely focus on acquiring valuable R&D knowledge and marketing resources through their alliance portfolios. Oil and gas companies are expected to concentrate their alliance portfolio strategy on acquiring technical knowledge, complementary operational assets, new oil and gas fields, using competitors and governments.
15 that the oil and gas sector is an interesting industry setting for alliance portfolio diversity studies, and potentially give new insights, compared to the study of Jiang et al. (2010).
3. RESEARCH METHODOLOGY
Wassmer (2010) points out that in alliance portfolio literature scholars often neglect to define clear boundaries of their research. This attempts to avoid an incomplete or biased alliance portfolio view, hence, the dataset contains both bilateral and multilateral alliances, as well as domestic and international alliances. A strategic alliance is defined as a voluntary arrangement among independent firms to exchange or share resources and engage in the co-development or provision of products, services, or technologies (Gulati, 1998). This includes (international) joint ventures.
3.1. Data collection
This study used a quantitative panel data set, company data was extracted from the Securities Data Corporation (SDC) Platinum Database. This database provided detailed information on the alliance portfolios of firms, whose primary Standard Industrial Classification (SIC) codes were in oil and gas extraction (division mining), and petroleum refining (division manufacturing) (SIC codes: 1311, 1321, 1381-1389, 2911). Firm performance data and firm size data was obtained from the Orbis. Orbis is a global company database which contains financial data on companies worldwide in an approximately 5 to 10-year window.
3.2. Sample
First, the 50 biggest oil and gas companies worldwide in 2010, were identified. The reason why I chose the largest organizations, is that it is commonly assumed that the largest companies in an industry give an appropriate view of the total industry. Besides, taking into account the oligopolistic character of the oil and gas industry (Levy and Kolk, 2002), it is assumed that the sample of the 50 biggest oil and gas companies give a good representation of the total industry. The reason for choosing the worldwide biggest oil and gas companies in the specific year 2010, is because I wanted to cover the relatively recent time spam; 2001-2010, thereby researching up-to-data industrial conditions. Moreover, the data sample represented companies from countries around the world, thereby increasing the external validity of the research.
16 Another eight companies were excluded from the dataset. According to the SDC Platinum Data, these firms did not hold any alliance between the years 2001-2005 or 2006-2010. Brahma (2009) argued that in order to increase the validity, firms lacking essential data should be left out the research, otherwise this could influence the research findings. The next step was to identify mergers and acquisitions conducted by the focal firms in the last 5 years period prior to the measure points; 2005 and 2010 (e.g. company websites and Wikipedia). In case the focal firm acquired more than 50% of the total assets of another firm between 2001-2005 or 2006-2010, and if the firm did not continue as an distinct subsidiary (determined by the financial data provided by Orbis), then the alliances of the acquired firm were added to the existing alliance portfolio of the focal firm. In case two companies merged, both firm’s alliances portfolios were pulled together resulting in a combined alliance portfolio. In the final stage, six more companies were dropped from the sample, due to lack of financial performance data and total assets data in Orbis, leaving a total sample size of 35 oil and gas companies and 1071 alliances.
3.3. Study
This paper uses a two-period longitudinal study to examine the effect of increasing alliance portfolio diversity on firm performance. The two periods 2001-2005 and 2006-2010 were chosen in order to analyse the current tendency concerning alliance portfolio diversity and firm performance in the oil and gas industry. By means of repeated observations of the same variables over time, the longitudinal study approach offers the opportunity to examine the potential curvilinear effects of increasing alliance partner portfolio diversity on firm performance. I looked at the effect of the independent variables on the dependent variable for each firm specific. The longitudinal study enables to research the effect of increasing alliance portfolio diversity degrees on firm performance over time.
3.4. Measures
3.4.1. Dependent variable. In line with previous studies (Faems et al., 2010; Jiang et al., 2010), this study use net
17 facing similar market conditions (Jacobson, 1987). Due to the international price of oil and gas, as well as the particular industry dataset, it is assumed that net profit margin is an appropriate measure in this study.
3.4.2 Independent variables. In this study alliances portfolio diversity is measured along three dimensions;
partner, functional, and governance diversity. In line with previous research (Jiang et al., 2010), partner diversity is divided into three different measures; industry, national, and organizational diversity.
To determine the diversity degrees, this study coded the diversity constructs as follow. Industry diversity is measured by means of categorizing the alliance partners, based on the amount of comparable primary SIC codes it shares with the focal firm: a ‘4’ is given to an alliance partner with the same four-digit SIC code, subsequently a ‘3’ for the same first three-digit SIC code, a ‘2’ for the same first two-digit SIC code, a ‘1’ for the same first one-digit SIC code, and a ‘0’ if the partners firms shares no SIC code digit. Organizational diversity is measured by categorizing alliance partners as follow: a ‘1’ for the same ownership (public-public, private-private) or organizational level (subsidiary-subsidiary, joint venture-joint venture), a ‘0’ is giving for alliance partners with different ownership (public-private, public-government, and private-government) or different organizational level (subsidiary-joint venture). Functional diversity is measured by coding the alliance partners into four categories, based on their functional purpose; ‘1’ for R&D, ‘2’ for manufacturing, ‘3’ for marketing, and ‘4’ for alliance partners with another functional purpose. Governance diversity is measured by coding a ‘1’ for a joint venture alliance and a ‘0’ for a non-joint venture alliance.
After coding the four above diversity constructs, the next step is to translate the results of these codifications into diversity measures. This paper will use the Blau diversity index (Blau, 1977) to translate the codifications into measures. The Blau diversity index measures the heterogeneity of the categorical variables, and is often used in diversity literature (Jiang et al., 2010). The Blau index quantifies the diversity of a group and is statistically defined as:
D = 1 - ∑ p
i2Where,
D = represents degree of diversity
p= represents the proportion belonging to a given category i = is the number of different categories.
18 independent variables are: industry diversity = 0.8, organizational diversity = 0.5, functional diversity = 0.75, and
governance diversity = 0.5. Due to the variation in the amount of alliance portfolio partners, ranging from 0 to 87, I
conducted the absolute diversity scores, corrected by the total amount of alliance portfolio partners.
National diversity is measured by coding the alliance portfolios with a continuous variable; a ‘0’ for an alliance
portfolio with no foreign partners, a ‘1’ for an alliance portfolio with one foreign partner, a ‘2’ for an alliance portfolio with two foreign partners, and so forth.
3.4.3. Control variables. Two control variables are included in this study to control for possible confounding effects. The first one is firm size, which is measured by the three year average of total assets of the focal firm, over respectively the years; 2005-2007 and 2010-2013. Former literature argued that firm size is correlated with the tendency to form alliances (Harrigan, 1988; Belderbos et al., 2006). Larger firms may find it easier to manage a larger alliance portfolio, and assumable have more resources available to invest in multiple alliances (Duysters et al., 2012). Besides, firm size may influences firm performance (Ahuja et al., 2008). The second control variable is alliance
portfolio size, and is measured by the focal firm’s total amount of alliances in the years 2001-2005 and 2006-2010.
Alliance portfolio size may influence the firm performance via economies of scale and scope, increased alliance management experience, or learning opportunities (Kale & Singh, 2007; Lavie & Miller, 2008). Therefore, this study includes alliance portfolio size as a control variable. Due to the proposed U-shaped effects of the partner hypothesises on firm performance; Portfolio size squared was added to measure the potential hypothesized curvilinear effects of increasing partner diversity on firm performance. Because off the single industry approach, no control variables were included controlling for industrial differences among the sample firms.
4. ANALYSES AND RESULTS
4.1. Descriptive statistics
19 As well as the functional diversity mean (0.27 < 0.60). As table 1 show, only functional diversity is strongly correlated with firm’s net profit margin (p < 0.001). Organizational diversity and governance diversity are significant under a p-value less than 0.1. The different independent variables show high correlations among each other. This is largely similar to the correlations coefficients found in the automotive sector (Jiang et al., 2010).
Due to the high correlations between the independent variables, I tested for multicollinearity, in order to see to what extend the distinct independent variables give valid results. I measured the variance inflation factors (VIF) of the independent variables, which resulted in a VIF of 1.39 (= < 10), a tolerance of 0.7257, and a conditional number of 8.6308, indicating no sign of multicollinearity problems (Mason & Perreault, 1991).
In order to determine whether to use a fixed or random effects model in this empirical study, I performed a Hausman test (Hausman, 1978). The Hausman test is used to check whether a fixed or random effects model is more efficient to use, resulting in more consisting results. If the P-value of the null hypothesis (coefficients estimated by the random effects model estimators are equal to the coefficients of the fixed effects model) is significant (P-value < 0.05), then a fixed effects model is more appropriate to use than a random effects model. Due to the fact that the rank of the differenced variance matrix did not equal the number of coefficients being tested in model 3, I rescaled my variables so that the coefficients are on a similar scale. The Hausman test showed a significant result (model 2 X2, 8d.f. = 39.07, p = 0.00 and model 3 X2, 11d.f. =206.17, p = 0.00), hence the more efficient fixed effects model is applied in this study.
4.2. Results
To examine the impact of alliance portfolio diversity on firm performance, I conducted multiple OLS (ordinary least squares) regression analyses. Due the fact that this study use a small dataset (N=35), the probability value of p<0.1 is included in the results. Although not significant, Fisher (1925) argued that a p-value between 0.05 and 0.1 provide Table 1. Descriptive statistics
Variables Mean S.D. 1 2 3 4 5 6 7 Net profit margin 20.29 13.94
Firm size 786.75 905.99 -0.08 Portfolio size 15.29 14.53 -0.18 0.75*** Industry diversity 0.44 0.21 0.09 0.30* 0.39*** National diversity 17.49 18.65 -0.14 0.73*** 0.93*** 0.36** Organizational diversity 0.38 0.15 -0.22† 0.29* 0.35** 0.34** 0.30* Functional diversity 0.27 0.21 -0.43*** 0.14 0.36** 0.21† 0.26* 0.39*** Governance diversity 0.32 0.19 -0.21† 0.25* 0.37** 0.08 0.36** 0.47*** 0.39*** N=70, † p<0.10, ∗ p<0.05, ∗∗ p<0.01, ∗∗∗ p<0.001.
20 some reason to suspect that the relation between the independent variable and dependent variable is not totally random, and that it shows a trend towards significance.
Model 1 (table 2) includes all the control variables. In contrast to previous research (Goerzen & Beamish, 2005; Jiang et al., 2010), this model shows that firm size has a small, but significant negative effect on firm performance (β = -0.01, p<0.05). Portfolio size and portfolio size squared are not significant.
Model 2 added the portfolio diversity variables. The model show no significant relation between partner industry diversity, national diversity, or organizational diversity and firm performance. Alliance portfolio functional diversity has a significant effect on firm performance (β = 17.46, p<0.05), offering support for hypothesis 2. The relation between governance diversity and firm performance was negative but not significant (β = -13.03, p<0.01), thus providing no support for hypothesis 3.
Model 3 includes the squared terms of the industry diversity, national diversity, and organizational diversity in order to test for curvilinear relations with firm performance. This model shows no significant results for the squared terms of industry diversity and organizational diversity in relation to firm performance. Since the root terms and squared terms of industry diversity, national diversity, and organizational diversity are not significant, hypotheses 1a, 1 b, and 1c are not supported. The control variables retain the same relationship as in model 1 and 2.
Fixed effects models measure the within effect, looking at a change in a variable over time within a firm, and its effect on firm performance. Model 2 and 3 shows an explanatory value of respectively 43% and 50% (model 2 R2 within = 0.43, model 3 R2 within = 0.50), By adding the squared terms of the partner diversity dimension the explanatory value of the model increased (R2 within chance=7%). It appears that the data moderately fits the models. Table 2. Fixed-effects OLS regression results
Independent variables Model 1 Model 2 Model 3
Firm size -0.01* -0.01** -0.01* Portfolio size -0.09 0.16 0.82 Portfolio size squared 0.00 0.01 -0.01 Industry diversity -7.54 -3.86 National diversity 0.13 -0.61 Organizational diversity -8.90 -30.28 Functional diversity 17.46* 15.40† Governance diversity -13.03† -12.80† Industry diversity -1.68 squared National diversity 0.01† squared Organizational diversity 51.52 squared Model F 9.73 9.11 9.22 R2 within 0.22 0.43 0.50 R2 between 0.00 0.08 0.08 R2 overall 0.00 0.02 0.02 1. Regression coefficients (β) are shown.
2. † p < 0.1, *p < 0.05, **p < 0.01.
21
5. DISCUSSION AND CONCLUSION
Increasing attention is being paid to the role that alliance portfolios play in relation to firm performance. Previous research suggested that diversity in alliance portfolios significantly influences firm performance (Cui & O’Connor, 2012; Duysters & Lokshin, 2011; Faems et al, 2010; Goerzen & Beamish, 2005; Jiang et al., 2010). On the one hand, increased diversity can provide more input of novel knowledge, resources, skills (McGill & Santoro, 2009), but may also increase the uncertainty and managerial complexity of an alliance portfolio (Duysters & Lokshin, 2011). However, scholars found mixed results of the impact of diverse alliance portfolios on firm performance. From a resource-based, and transaction costs perspective, this study attempted to find empirical prove for the alliance portfolio diversity – performance relation by using a dataset containing 1071 strategic alliances formed by 35 large oil and gas companies. Alliance portfolio diversity was measured by means of three commonly, but not simultaneously used diversity dimension; partner diversity, functional diversity, and governance diversity.
5.1. Partner diversity and firm performance
This study drew on recombinatory literature research and hypothesized that alliance portfolio partner diversity; divided in industry diversity, national diversity, and organizational diversity, would have an inverted U-shape relation with firm performance. No significant results were found for any of these hypotheses. The results suggest that increasing alliance portfolio partner diversity does not influence the firm performance. The findings partly contradict prior research. Goerzen & Beamish (2005) found a positive relation between national diversity and firm economic performance, and a negative relation between industry diversity and firm economic performance. The study of Jiang et al. (2010) confirmed the negative influence of industry diverse alliance portfolios. Regarding increasing organizational diversity, they found a strong positive effect on firm performance. Similar to their results, this study found no significant result of the effect of increasing alliance portfolio national diversity on firm performance.
A reason for the non-significant result regarding national diversity may be due to the international character of the oil and gas firms. Due to an international diverse internal resource set, and international alliance experience of the focal firm, the potential benefits and costs of a national diverse alliance portfolio might be mitigated (Grant & Cibin, 1996). This can decrease the effect of foreign diversity in alliance portfolios (Jiang et al., 2010). However, this was not confirmed by this study, and could be a direction for future research.
22 firm performance (Stuart, 2000). For instance, prior research found that knowledge plays an important role in the biotechnical industry in increasing firm performance via enhanced innovation performance (Baum et al., 2000). In the oil and gas industry operational productivity, exploitation of strategic assets on large scale, and technical innovation are suggested to determine to a large extent the firm performance (Silvestre & Dalcol, 2010). Specific alliance partner’s attributes can provide access to new oil and gas field, and provide complementary resources to exploit new operation business areas. Oil and gas firms potentially focus on specific complementary resources, regardless of industrial background, country of origin, or organizational type. So, in this research setting, alliance partner’s characteristics may play a significant role. Hence, further research could examine the role of specific alliance partner’s attributes in the oil and gas industry.
A potential explanation for the non-significant findings stems from prior studies, suggesting that alliance portfolio management capabilities and partner experience can moderate the relation between alliance partner diversity and firm performance (Cui & O’Connor, 2012; Duysters & Lokshin, 2011; Zollo et al., 2002). Especially, in this research setting containing a dataset of large multinational oil and gas firm, which have experienced a long history of doing business (over 50 years), partner experience may play an important role. Although beyond the scope of this research, future research could use the multidimensional partner diversity construct to examine the role of alliance portfolio management capabilities and prior partner experience, in the relation between alliance portfolio partner diversity and firm performance.
5.2. Functional diversity and firm performance
The study of Cui & O’Conner (2012) found a direct negative effect of increased functional diversity on firm innovation. And, although researchers argued that increased functional diversity may hamper inter-partner communication (Lane & Lubatkin, 1998; Oxley & Sampson, 2004), which can increase managerial challenges and increases transaction costs (Faems et al., 2010), this study found a significant positive effect of higher alliance portfolio functional diversity on firm performance, implying that the financial returns outweigh the costs of holding such a diverse alliance portfolio. This result is in line with the findings of Jiang et al. (2010), and suggests that on the one hand, firms with functional diverse alliance portfolios are able to discover new business opportunities, acquire new knowledge and technological capabilities, and enhance innovation capabilities via exploratory orientated alliances, in order to enhance their long-term competitive advantages. On the other hand, firms can capitalize their existing resources, knowledge, and capabilities in order to increase short-term profits (Koza & Lewin 1998; March, 1991; Rothaermel, 2001; Uotila et al., 2009).
23
5.3. Governance diversity and firm performance
No significant result was found on the hypothesis that increasing governance diversity influences firm performance, which suggested that holding an alliance portfolio with a diverse set of governance structures, does not influence firm performance. This raises questions about the importance of management experience, and the costs implication of learning to manage diverse alliance governance structures (Madhok, 2002; Sampson, 2005). Although not significant, the result showed a trend towards significance (P < 0.1) (Fisher, 1925), suggesting that these finding could potentially align with the findings of Jiang et al. (2010) who found a negative influence of higher governance diversity on firm performance. Increasing the sample size would potentially increase the significance of the findings. Moreover, a potential reason for the non-significant result could be rooted in way the construct was measured. This study categorized governance diversity in joint ventures and other types of governance structures. However, other governance structure measures could be applied to research alliance portfolio governance diversity. Anand & Khanna (2000) made a distinction between the functional purposes of a joint venture (e.g. production, marketing, and R&D joint ventures). Besides, a commonly distinction is made between alliance governance structures which involve solely non-equity investments, and certain degrees of equity based investments (Jiang et al., 2010). Additionally, scholars focused on relational governance mechanisms like trust (Das & Teng, 1998). Future research could examine other forms of governance structures and their role in the alliance portfolio diversity – firm performance relationship.
5.4. Control variables
This study included firm size and portfolio size as control variables. The results show a small negative effect on firm size on firm performance, which could suggest that larger firms experience higher transaction costs. Although the effect is small, this contradict a stream of common arguing, that larger firms commonly have lower transaction costs due to higher levels of economies of scale, scope, experience and learning (Nooteboom, 1993). Future research could include firm size as moderator, to examine the implication for the relation between alliance portfolio diversity and firm performance.
24
5.5. Managerial implications
The results in this study suggest that managers should mainly focus on balancing their alliance portfolios, based on the functional purposes of the alliance partners. Benefits appear to emerge equally from R&D, manufacturing, and marketing activities. So, by increasing the functional diversity in a firm’s alliance portfolio, managers are likely to increase the firm’s financial performance. The findings in this study also suggest that it seems that specific partner difference, like industrial, national, organizational backgrounds, do not have a substantial influence on the firm performance of large firms. Besides, applying varying governance structures to alliances also appear to have no direct cost increasing effect. Hence, this paper advises managers to take into account the functional purpose of the alliances, and to aim for a higher spread of R&D, manufacturing, marketing and other alliances in their strategic alliance portfolio strategies.
5.6. Limitations
This empirical study is not without limitations. First, the sample size used in this study in rather limited. The panel dataset consisted of only 35 firms, which decreases the reliability of this study.
Additionally, the sample consisted of only large companies; hence these findings may not be generalizable to small companies. Further research should test the findings, including a larger sample size and a wider dispersion of large, medium, and small size firms.
Second, the measures of the diversity constructs, and firm performance could be more extensive in order to increase the constructs validity. For instance, further research could measure governance diversity by means of a broader distinction between different governance structures (e.g. non-equity, equity) or types of governance mechanisms (formal, relational). Furthermore, firm performance was only measured by means of net profit margin, more financial measures or other alternative firm goals could be included, like market share, return on equity, revenue growth, or new business opportunities, learning objectives, and reputation enhancement. Including more specific measures of the variables could give a richer explanation about the role of alliance portfolio diversity in firms.
Third, the single industry design limits the generalizability of this study’s findings to other industries. Hence, it is suggested that future research on alliance portfolio diversity examines, for example, the telecommunications and electronics industry where alliances are also commonly used (Jiang et al., 2010), or more extensively across industries.
25
6. CONCLUSIONS
This study complements existing alliance portfolio literature by showing that increased alliance portfolio functional diversity positively influences firm performance. I hope my findings will help managers to improve their alliance portfolio strategy. At the same time, I want to stimulate other scholars and master students to further examine the implications of the alliance portfolio diversity constructs for firms. For instance, they can take into account potential moderating and mediating effects like partner alliance partner experience or managerial capabilities. I similar vein, I want to stress the increasing importance of strategic alliances in the oil and gas sector, and encourage further research on the role of strategic alliance in this particular industry.
Acknowledgements
I thank Pedro de Faria for his pleasant supervision and constructive feedback. Also a special thanks to Brenda Bos, who helped me to perform the empirical analyses.
REFERENCES
Ahmad, S., & Schroeder, R. G. (2003). The impact of human resource management practices on operational performance: recognizing country and industry differences. Journal of Operations Management, 21 (1), 19-44. Ahuja, G., Lampert, C. M., Tandon, V. (2008). Moving beyond Schumpeter: Management Research on the
Determinants of Technological Innovation. Academy of Management Annals, 2 (1), 1-98.
Ahuja, G. (2000). The duality of collaboration: Inducements and opportunities in the formation of interfirm linkages.
Strategic Management Journal, 21 (3), 317-343.
Alper, C. E., & Mumcu, A. (2007). Interaction between price, quality and country of origin when estimating automobile demand: the case of Turkey. Applied Economics, 39 (14), 1789-1796.
Anand, B. N., & Khanna, T. (2000). Do firms learn to create value? The case of alliances. Strategic Management
Journal, 21 (3), 295-315.
Bae, J., & Gargiulo, M. (2004). Partner substitutability, alliance network structure, and firm profitability in the telecommunications industry. Academy of Management Journal, 47 (6), 843-859.
Baker, G. P., Gibbons, R., Murphy, K. J. (2008). Strategic alliances: bridges between islands of conscious power.
Journal of the Japanese and International Economies, 22 (2), 146-163.
Barney, J.B. (1991). Firm resources and sustained competitive advantage. Journal of Management 17 (1), 99-120. Baum, J. A. C., Calabrese, T., Silverman, B. S. (2000). Don't go it alone: alliance network composition and startups'
performance in Canadian biotechnology. Strategic Management Journal, 21 (3), 267-294.
Belderbos, R., Carree, M., Lokshin, B. (2006). Complementarity in R&D cooperation strategies. Review of Industrial
