International ambidexterity and MNE performance

(1)

Department of Economics and Business

MSc. Business Administration: International Management

Final version Master’s thesis: International Management

“International ambidexterity and MNE

performance”

Written by: Alec Paul Jacques Goedert Student Number: 11374098

(2)

Statement of Originality

This document is written by Alec Paul Jacques Goedert who declares to take full responsibility for the contents of this document.

I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.

The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.

(3)

List of figures

Figure 1: Conceptual framework ... 24

Figure 2: Measure of international ambidexterity ... 29

List of tables

Table 1: Categorization scheme ... 28

Table 2: Country distribution ... 31

Table 3: Industry distribution ... 32

Table 4: Means, standard deviations and correlations ... 35

Table 5: Regression results ... 39

(5)

Abstract

Ambidexterity, which is commonly defined as a firm’s ability to simultaneously pursue exploitation and exploration, has been shown to lead to improved performance in the context of single country research. In international business, however, scholars have traditionally emphasized multinational enterprises’ (MNEs’) ability to advance the exploitation of their firm-specific advantages as the primary requisite for international success. In light of the occasionally large differences in institutions, infrastructure, technological development, etc. across countries, more recent theories have highlighted the importance of learning and developing new capabilities to successfully operate in an international setting. In this context, Prange & Verdier (2011) introduce the concept of international ambidexterity, which refers to an MNE’s ability to reconcile exploitation and exploration on its internationalization path. Collecting data of 105 companies from nine European countries, the current study investigates the relationship between international ambidexterity and MNE performance. The findings do not support the notion that more internationally ambidextrous firms outperform their peers. Moreover, the results do not confirm that MNEs’ geographical scope, number of foreign subsidiaries and tangible resource base influence this relationship. In light of these findings, it is argued that international ambidexterity, which is commonly defined in terms of orientation towards exploitation and/or exploration, may be insufficient in the context of international business. Due to the variety of factors that may impinge on the successful implementation of MNEs’ international strategies, international ambidexterity in itself may be futile if MNEs are unable to establish proper mechanisms that ensure successful execution of the underlying orientation.

(6)

1. Introduction

In today’s fast-moving and dynamic environment, successful firms do not only increase the efficiency of their existing operations in current environmental conditions, but are also able to adapt to environmental changes, and renew their capabilities and knowledge base (Raisch & Birkinshaw, 2008). This is as true for domestically oriented firms as it is for internationally operating firms. Indeed, for multinational enterprises (MNEs), it is essential to not only rely on existing resources and capabilities, but also to search for opportunities to occasionally renew their resource base and acquire new competencies (Prange & Verdier, 2011; Makino, et al., 2002). The fact that MNEs have to deal with intense environmental pressures, as a result of operating in different institutional contexts and facing strong international competitors, intensifies this necessity (Hitt, et al., 2016).

As such, although MNEs have to make many important decisions, such as which countries to enter or how standardized their product offering should be, a concern of particular importance is how they can increase alignment with existing conditions and simultaneously stay adaptive and search for new opportunities (Ireland & Webb, 2007; Gibson & Birkinshaw, 2004). In other words, firms have to undertake exploitation and exploration at the same time (March, 1991; Levinthal & March, 1993). The ability to simultaneously pursue exploitative and explorative activities has been labeled ambidexterity (Gibson & Birkinshaw, 2004). More specifically, in the context of international business (IB), international ambidexterity denotes an MNE’s ability to reconcile exploration and exploitation along its internationalization path (Prange & Verdier, 2011; Hsu, et al., 2013). The consensus in the literature is that firms which manage to simultaneously pursue exploitative and explorative activities will depict higher performance (Gupta, et al., 2006). Firms which overly focus on exploitation may be successful in the short run, as they improve existing competencies and increase their alignment with current environmental conditions, but they will struggle once conditions change (Lubatkin, et al., 2006;

(7)

in current environmental conditions can become a liability once conditions change (Tushman & O'Reilly, 1996; Kristal, et al., 2010). Similarly, firms which overly focus on exploration may be able to generate many new ideas and identify new opportunities, but they will fail to develop them to their full potential, as they do not appropriately exploit them (March, 1991). Thus, ambidextrous firms, by avoiding the drawbacks of solely focusing on either exploitation or exploration, are said to outperform their rivals (Gibson & Birkinshaw, 2004; Ireland & Webb, 2007; Lubatkin, et al., 2006; He & Wong, 2004; Cao, et al., 2009; Luo & Rui, 2009).

Traditional theories in IB suggest that firms internationalize in an attempt to exploit existing capabilities and competencies in foreign markets (Hymer, 1976; Makino, et al., 2002) or to increase the efficiency of their operations (Buckley & Casson, 2009). Those are activities that are typically associated with exploitation (Hitt, et al., 2016). However, more recent theories and insights show that firms do not only internationalize to exploit existing competencies, but also to acquire new assets, knowledge, and capabilities outside their home country (Dunning, 2000; Barkema & Drogendijk, 2007; Demirbag & Glaister, 2010). In other words, MNEs increasingly undertake explorative activities to increase their competitiveness and to ensure their viability in the long run. This can be understood as a direct response to the increasing environmental complexity and the intensification of global competition that MNEs have to deal with and demonstrates contemporary MNEs’ need to handle exploitation and exploration simultaneously, i.e., to be internationally ambidextrous (Hsu, et al., 2013).

Prior studies with a single country focus have corroborated the notion that highly ambidextrous firms indeed depict better performance than less ambidextrous firms (Cao, et al., 2009; Gibson & Birkinshaw, 2004; He & Wong, 2004; Lubatkin, et al., 2006). However, in the context of internationalizing firms, empirical evidence is much scarcer. Luo & Rui (2009) and Jonsson & Foss (2011) provide case study evidence for the idea that international ambidexterity allows MNEs to become more competitive and to beat their global competitors. Hsu, et al. (2013)

(8)

exploitative and explorative foreign direct investment (FDI), i.e., MNEs that are more ambidextrous, achieve better performance. However, they solely focus on companies originating from Taiwan, which limits the generalizability of their results to other developing countries, not to mention developed countries. Moreover, little is known about the contextual conditions in which international ambidexterity is most beneficial to MNEs. Although Hsu, et al. (2013) consider the possibility of moderators influencing the relationship between international ambidexterity and MNE performance, their list is not complete. As such, adopting a contingency theory perspective, this paper investigates the moderating role of factors in MNEs’ internal and external environment, which potentially influence the relationship between international ambidexterity and MNE performance. Thus, this paper contributes to the existing literature by extending the research on international ambidexterity to MNEs from developed countries. In addition, the focus on contextual factors which potentially influence the relationship between international ambidexterity and MNE performance will provide a clearer understanding of the effects of international ambidexterity. These thoughts lead to the following research question: What is the effect of international ambidexterity on an MNE’s performance

and are there internal and external factors that moderate the relationship between these two?

The structure of this paper will be as follows. The first chapter will provide the reader with a theoretical background on ambidexterity and the findings of prior empirical studies. The next chapter will introduce the conceptual framework including the hypotheses. The methodology chapter will elaborate on the sample compilation, the data collection procedure and the operationalization of the variables. In the following chapter, the results of the statistical analysis will be presented. The paper concludes with a discussion of the most important findings, the limitations of the study and directions for future research.

(9)

2. Literature review

2.1. Exploitation and exploration

At the core of the ambidexterity literature is the premise that firms generally need to divide their time and resources between two sorts of activities, namely exploitation and exploration (He & Wong, 2004; March, 1991). Exploitation denotes the (re-)utilization and improvement of already established knowledge, routines and capabilities (Kristal, et al., 2010; Raisch & Birkinshaw, 2008; Barkema & Drogendijk, 2007). As such, exploitative activities typically involve the recombination, adjustment and reconfiguration of what has already been learnt, and lead to incremental enhancements of existing technologies and market operations. Firms pursue exploitation to foster their alignment with existing external circumstances and to ensure continuous fulfillment of the contemporary expectations of their current customer base (Lubatkin, et al., 2006). As exploitation mostly involves the application and refinement of what is already known, the financial outcomes of exploitative activities are less uncertain, closer in time, less variable, generally positive and subject to higher predictability and reliability (March, 1991; Im & Rai, 2008; He & Wong, 2004; Auh & Menguc, 2005). Gupta, et al. (2006) note that there is some disagreement in the literature as to whether exploitation involves learning or not. One stream of the literature interprets exploitation as the use of an existing knowledge base and capabilities without making any improvements or refinements. From this point of view, any type of learning, incremental or not, falls under the label of exploration. However, this view of exploitation is clearly at odds with March’s (1991) initial conceptualization of the term. Hence, exploitative activities do indeed involve learning, but it is a different kind of learning, namely more incremental (Gupta, et al., 2006).

Exploration, in contrast to exploitation, refers to firm activities concerned with the acquisition of new capabilities, the gathering of new knowledge, and the search for new opportunities (Kristal, et al., 2010; Levinthal & March, 1993; Floyd & Lane, 2000). Thus, explorative

(10)

activities involve the departure from established routines and lead to the development of new technological trajectories and the opening up of new markets. As such, whereas exploitation is more concerned with fulfilling preferences of existing customers and increasing the firm’s alignment with its current environment, exploration is oriented towards the fulfilment of the (latent) demands of new customers in new markets and being prepared to adapt to new environmental circumstances (Lubatkin, et al., 2006; Gibson & Birkinshaw, 2004). However, building on Barkema & Drogendijk’s (2007) discussion of cultural knowledge, it should be noted that explorative activities should not take a firm too far away from its existing knowledge base. If the new technologies or opportunities are too unrelated with regards to a firm’s existing competencies, it is unlikely that it will be able to make commercial use of ‘the new’. Thus, explorative activities should be somewhat related to existing knowledge and capabilities. In contrast to the outcomes of exploitation, exploration outcomes are subject to higher uncertainty and variability (Im & Rai, 2008; He & Wong, 2004). Furthermore, the outcomes of explorative actions take longer to manifest themselves and are more often negative than positive. The reason for this is that the results of extending what is already known and learnt are always less risky and closer in time than the results of operations that drive a firm away from its existing abilities and knowledge base (March, 1991; Barkema & Drogendijk, 2007).

With regards to the compatibility of exploration and exploitation, two different opinions have emerged in the literature (Gupta, et al., 2006; Kristal, et al., 2010). The first view sees exploitation and exploration as fundamentally at odds with each other (March, 1991) and defines them as “two ends of a continuum” (Gupta, et al., 2006, p. 693). First, exploitation and exploration contend for a firm’s limited resources with each other, such that money and other inputs spent on exploitation cannot be used for explorative purposes anymore, and vice versa (March, 1991; Patel, et al., 2013). Moreover, attempting to foster the alignment with current environmental conditions will inevitably weigh on a firm’s ability to innovate, adapt to changes in its external environment, and seize upcoming opportunities (Patel, et al., 2013). In a similar

(11)

vein, exploring new possibilities will put a burden on how fast a firm can enhance existing capabilities (He & Wong, 2004). Finally, exploitative and explorative activities require different organizational routines, processes and attitudes (Im & Rai, 2008; Raisch & Birkinshaw, 2008). On the one hand, exploitation is more associated with a top-down approach, where upper management defines a relatively rigid framework that allows for continuous refinement of existing knowledge and capabilities. On the other hand, exploration is associated with a bottom-up approach that gives lower management and employees more autonomy to identify new opportunities and develop a new range of capabilities (Lubatkin, et al., 2006). Thus, decentralization of at least some of a company’s operations is necessary for successful exploration, whereas exploitation is better achieved through centralized decision-making processes. As such, according to this view, there is necessarily a trade-off between exploitation and exploration (Cao, et al., 2009; March, 1991).

However, the second view that has emerged in the literature does not consider exploitation and exploration as two extremes of a continuum, but rather sees them as orthogonal to one another (Gupta, et al., 2006). From this perspective, there is no trade-off between exploration and exploitation and firms can choose to maximize both types of activities. In other words, a firm’s decision to invest time and money in explorative activities has no implications on how much resources it can subsequently spend on exploitative activities, and vice versa (Cao, et al., 2009). The rationale behind the idea that exploitation and exploration are not incommensurate with one another, is that the basic assumption of resource scarcity, which is presented in March’s (1991) seminal paper, is flawed or at least incomplete. Authors in favor of the orthogonality perspective generally argue that although most resources are indeed subject to constraints, others are not. For instance, although it may be true that production input factors, money, time, etc. are not unlimitedly available to a firm, this is typically not true for the supply of knowledge and information. Moreover, firms have the ability to form alliances with each other, a circumstance which involves at least a partial abolition of their resource constraints (Gupta, et

(12)

al., 2006). Thus, from this perspective, exploration and exploitation are not subject to a trade-off (Cao, et al., 2009), but are two complementary elements of a whole that fuel each other (He & Wong, 2004). The distinction between these two views is important, as it has a bearing on the definition of (international) ambidexterity, which will be discussed in the next section of this paper.

Although the previous discussion made it clear that exploitation and exploration are by no means independent of each other, it should be noted that their interdependency goes beyond the incompatibility/complementarity debate. Rothaermel & Deeds (2004) note that explorative activities, by detecting new possibilities and technologies, provide the basis for future exploitation. Indeed, the fundamental logic behind exploitation is that firms improve existing capabilities and refine existing knowledge (March, 1991; Levinthal & March, 1993), but a necessary condition for this to take place is that firms undertake some form of explorative actions first. In order to be able to exploit, a firm has to explore at least once and only if this exploration has been successful can further exploitation take place (Rothaermel & Deeds, 2004).

2.2. International ambidexterity

2.2.1. Definition

In general terms, ambidexterity denotes a firm’s capability to perform two dissimilar activities at the same time (e.g., standardization versus customization of products and services) (Gibson & Birkinshaw, 2004; Han, 2007). More specifically, ambidexterity refers to a firm’s ability to simultaneously pursue exploitative and explorative activities (Raisch & Birkinshaw, 2008; Lubatkin, et al., 2006; Gupta, et al., 2006; Hsu, et al., 2013). At the core of the ambidexterity literature is the idea that firms which manage to engage in both types of activities will outperform those that only focus on either exploitation or exploration (March, 1991; Levinthal & March, 1993; Tushman & O'Reilly, 1996). Prange & Verdier (2011) extend the notion of

(13)

ambidexterity to the IB context and introduce the concept of international ambidexterity. They define it as an internationalizing firm’s ability to reconcile the conflicting goals of explorative and exploitative international expansions. This suggests that ambidexterity is as relevant to MNEs as it is to domestically oriented companies. One might argue that ambidexterity is even more important to MNEs than it is to firms with a single country focus. Internationally operating firms generally have to deal with rapidly changing environments and technological trajectories (Prange & Verdier, 2011). In addition, these firms often face fierce international competitors and have to deal with the complexity of operating in different countries and institutional contexts at the same time (Hitt, et al., 2016), suggesting that there is less room for mistakes. Today’s competitive landscape requires firms, multinational or not, to handle exploitation and exploration simultaneously if they are to stay sustainably competitive (Hsu, et al., 2013; Ireland & Webb, 2007; Prange & Verdier, 2011), and thus requires them to consider the short run and the long run (Luo & Rui, 2009). The rationale is that only focusing on one type of activity to the exclusion of the other, will leave a company with important disadvantages and will endanger its viability (March, 1991; Prange & Verdier, 2011; Hsu, et al., 2013; Floyd & Lane, 2000; Jansen, et al., 2012).

On the one hand, firms that solely focus on exploration to the detriment of exploitation run the risk of being trapped in an ongoing cycle of unprofitable change and renewal (Raisch & Birkinshaw, 2008). An inappropriately strong focus on exploration leads to a situation, where firms have to bear the full costs of trying out new things and exploring new possibilities, without ever being able to reap the advantages of it (March, 1991; Cao, et al., 2009). Firms that overly focus on exploration may have an initial advantage compared to their peers, when it comes to the discovery of new markets, capabilities or technologies (Lubatkin, et al., 2006), but they will ultimately underperform, as they subsequently fail to devote enough resources to the exploitation of those discoveries (He & Wong, 2004). As such, these firms are likely to end up with a range of underdeveloped capabilities, opportunities and ideas without ever developing

(14)

them to their full potential (March, 1991). Furthermore, as already mentioned, exploration outcomes are more uncertain, distant in time and mostly negative. As such, a purely explorative focus is also unfavorable, because it increases a firm’s vulnerability compared to competing firms that also focus on achieving efficiency and not only renewal. Whereas a solely explorative firm may endlessly wait for the next breakthrough and makes losses in the process, others drive it out of business (Lubatkin, et al., 2006). Levinthal & March (1993) note that the mechanisms of exploration are self-reinforcing in the sense that exploration leads to failure, which leads firms to explore even more, fail again, explore again, etc. and it may be very difficult to break out of this cycle. They call this the “failure trap” (p. 105).

On the other hand, firms that solely focus on exploitation run the risk that their knowledge and capabilities become obsolete over time (March, 1991; Casillas & Moreno-Menéndez, 2014). Although it is true that a strong focus on exploitation and thereby attempting to bring existing technologies and competencies to perfection may be associated with high performance in the short run, such a behavior makes firms extremely vulnerable to more radical changes in the environment and technological trajectories (Raisch & Birkinshaw, 2008; Cao, et al., 2009). Tushman & O’Reilly (1996) argue that firms have to be willing to cannibalize their existing business every now and then if they want to stay sustainably competitive. Purely exploitation oriented firms will continuously improve their alignment with current environmental conditions, but once these conditions change, they will be at a serious disadvantage compared to companies with a more explorative mindset (Lubatkin, et al., 2006). As such, the problem with overly focusing on exploitation is not one of being unable to create a competitive advantage. The problem rather lies in the missing sustainability of the latter (He & Wong, 2004; Auh & Menguc, 2005). Similar to the self-reinforcing mechanisms of engaging in too much exploration, exploitation is also subject to a self-reinforcing cycle. Companies have a natural tendency to engage in activities that they are particularly good at. As they refine and improve the skills and knowledge required for these activities, they become better and better in their

(15)

particular sphere of competence and as a result engage even more in those activities. This is called the “success trap” (Levinthal & March, 1993, p. 106). Although potentially efficient in the short run (Casillas & Moreno-Menéndez, 2014; Tushman & O'Reilly, 1996), such a behavior is ultimately self-destructive in the future (March, 1991).

2.2.2. Two views in the literature

As has already been mentioned in a previous section of this paper, there are two different views regarding the (in)compatibility of exploitation and exploration. Accordingly, two different views on what an ambidextrous firm actually looks like have emerged. Although both views agree on the notion that ambidextrous firms somehow pursue exploitation and exploration at the same time - the reasons for this have been explained in the previous section - they fundamentally differ with regards to the rivalry between the two (Gupta, et al., 2006; He & Wong, 2004; Cao, et al., 2009).

On the one hand, scholars who view exploration and exploitation as two extremes of a continuum define ambidexterity as a firm’s ability to achieve an effective balance between exploitative and explorative activities (Raisch & Birkinshaw, 2008; Hsu, et al., 2013; He & Wong, 2004; Cao, et al., 2009; Auh & Menguc, 2005). Thus, the relative extent of exploitation and exploration is at the center of interest and not the combined magnitude of the two (Cao, et al., 2009). As such, the smaller the degree of imbalance between exploitative and explorative activities, the more ambidextrous a firm is (He & Wong, 2004). Although exploitative and explorative activities are said to contend for a firm’s scarce resources, firms must balance the two in order to stay competitive in the short run and the long run (March, 1991). Building on the incompatibility argument, Levinthal & March (1993) assert that a firm needs to “engage in sufficient exploitation to ensure its current viability and, at the same time, […] devote enough energy to exploration to ensure its future viability” (p. 105). As such, although trade-offs between exploitation and exploration are inevitable, these trade-offs are seen as a mechanism

(16)

(Cao, et al., 2009; Han, 2007). On the other hand, scholars who question March’s (1991) assertion of scarce resources and thus view exploitation and exploration as orthogonal to each other, define ambidexterity as a firm’s ability to boost the joint magnitude of exploitative and explorative activities (Raisch & Birkinshaw, 2008; He & Wong, 2004). As such, the relative

magnitude of either exploitation or exploration is not of interest, neither is the degree of

imbalance between the two (Cao, et al., 2009). From this perspective, exploitative and explorative activities are not subject to trade-offs, but are seen as complementary and supportive of each other (Raisch & Birkinshaw, 2008). Hence, the higher the joint magnitude of exploitation and exploration, the more ambidextrous a firm is (He & Wong, 2004).

The distinction between these two views has fundamental theoretical and empirical implications (Cao, et al., 2009). Consider the following example: firm A has an exploitation score of 20 and an exploration score of 10, whereas firm B has an exploitation and exploration score both equal to 10. Scholars in favor of the continuum view would argue that firm B is more ambidextrous than firm A (i.e., exploitation and exploration are in balance) and should, therefore, depict higher performance. Scholars in favor of the orthogonality view would argue that firm A is more ambidextrous than firm B, because its joint magnitude of exploitation and exploration - as measured by either the product or sum of the two scores - is higher than for firm B. Hence, depending on the conceptualization, different conclusions and theoretical predictions emerge with regards to which firm is more ambidextrous and thus should perform best (Gupta, et al., 2006; Cao, et al., 2009; He & Wong, 2004). Gupta, et al. (2006) come to the conclusion that neither view is superior to the other and that the validity of the continuum or orthogonality view essentially depends on the existence of resource constraints. As such, if resource constraints are in place, defining ambidexterity as the ability to achieve a balance between exploitation and exploration is more appropriate than defining it as the ability to boost the joint magnitude of the two, and vice versa.

(17)

Although the argument that firms which simultaneously pursue exploitation and exploration will outperform those firms that only focus on one end of the spectrum has been brought forward by many authors (e.g., March, 1991; Levinthal & March, 1993; Ireland &Webb, 2007), empirical tests of this assertion have been much scarcer (Raisch & Birkinshaw, 2008). He & Wong (2004) test the influence of ambidexterity on firm performance in the context of technological innovation strategies at the firm level. Recognizing that two different interpretations of ambidexterity exist in the literature, they show that a high imbalance between exploitation and exploration negatively influences sales growth, whereas the interaction between the two positively influences sales growth. Moreover, their findings suggest that pushing exploitation and exploration to extreme limits has negative performance implications and that achieving a balance between exploitative and explorative activities is not enough to sustain high performance if both fall under a certain threshold. Turning their attention to business units, Gibson & Birkinshaw (2004) find support for their hypothesis that the ability of a business unit to simultaneously pursue alignment (i.e., exploitation) and adaptability (i.e., exploration) is positively associated with firm performance. In addition, they find that there doesn’t seem to be a trade-off between exploitation and exploration. Similarly, Lubatkin, et al. (2006) find support for their hypothesis that ambidexterity is positively related to the performance of small- and medium-sized firms. Building on He & Wong’s (2004) insight that two different views on ambidexterity can be found in the literature, Cao, et al. (2009) are the first to conceptually distinguish between these two interpretations. They introduce a ‘balance dimension’ of ambidexterity and a ‘combined dimension’ of ambidexterity, suggesting that the two views of ambidexterity capture different aspects of the phenomenon. Their findings are relatively in line with those of He & Wong (2004). Moreover, although their results show that there seems to be a positive interaction effect between the balance and combined dimension of ambidexterity, a high score on the combined dimension cannot compensate for a severe imbalance between exploitation and exploration. Finally, Cao, et al.’s (2009) findings

(18)

corroborate Gupta, et al.’s (2006) assertion that achieving a balance between exploitation and exploration is more conducive to performance sustainability under resource constraints, whereas firms with greater availability of resources are in less need of achieving such a balance. Thus, greater resource availability may allow firms to overcome the trade-off between exploration and exploitation (Cao, et al., 2009). To sum up, empirical results provide support for both the balance and joint magnitude view of ambidexterity. Indeed, former findings suggest that achieving a balance between and boosting the joint magnitude of exploitation and exploration both positively influence firm performance. However, a firm’s availability of resources seems to determine which type of ambidexterity is more desirable.

In addition to the studies that analyze the performance implications of ambidexterity on the firm or business unit level, some scholars investigate the effects of ambidexterity in inter-organizational relationships. Drawing a sample from the biotechnology industry, Rothaermel & Deeds (2004) find support for their hypothesis that small biotechnology firms engage sequentially in explorative and exploitative alliances to first generate ideas for new products and then introduce them to the market, suggesting that both types of alliances are necessary for successful product developments. However, the authors do not investigate the performance implications of engaging in both types of alliances nor do they test whether explorative and exploitative alliances should be in balance or not. Im & Rai (2008) find that there are synergistic effects between exploitative (i.e., exchange of information aimed at gains in the short run) and explorative (i.e., exchange of information aimed at gains in the long run) knowledge sharing and that high levels of both lead to better performance in long-term inter-organizational relationships. Studies that deal with ambidexterity in the context of IB are very scarce. The study of Hsu, et al. (2013) is the only one that empirically tests the ambidexterity hypothesis in the context of internationalization. Drawing a sample of 207 Taiwanese multinational companies operating in the high-technology industry, the authors find support for their hypothesis that MNEs with a higher degree of imbalance between exploitative and explorative

(19)

FDI (i.e., a lower level of ambidexterity) underperform MNEs with a lower degree of imbalance (i.e., a higher level of ambidexterity). Additional case study evidence for the notion that ambidextrous firms depict better performance comes from Jonsson & Foss (2011) and Luo & Rui (2009). Although not directly building on the exploitation-exploration dichotomy, Jonsson & Foss (2011) show how IKEA has managed to become more competitive internationally over the years by balancing the conflicting demands of being locally responsive and cost-efficient. Like March (1991), the authors argue that there is a trade-off between these two objectives, but a balance is nonetheless necessary. Luo & Rui (2009) describe how multinational companies from developing countries have been able to overcome critical situations and outrival their global competitors owing to their ambidextrous nature. However, it should be noted that, contrary to other scholars, they define ambidexterity as a multidimensional construct of which the simultaneous pursuit of exploitation and exploration is only one dimension.

In summary, although empirical evidence suggests that ambidexterity - as measured by either the degree of balance between or joint magnitude of exploitation and exploration - positively influences firm performance, most of these studies have a single country focus (e.g., Cao, et al., 2009; Lubatkin, et al., 2006). Furthermore, the few studies that deal with international ambidexterity in the context of internationally operating firms either focus exclusively on emerging countries (Hsu, et al., 2013), adopt a qualitative research design (Jonsson & Foss, 2011) or both (Luo & Rui, 2009). Moreover, the only quantitative study dealing with ambidexterity in an IB context, conducted by Hsu, et al. (2013), solely focuses on firms from Taiwan, which impedes the generalizability of the results to firms from other emerging countries, not to mention firms from developed countries. Thus, existing research seems to have completely neglected the effect of international ambidexterity in the light of MNEs from developed countries. Even though the logic behind achieving ambidexterity can easily be extended to internationally operating firms, MNEs face a much greater complexity than domestically-focused firms (Hitt, et al., 2016). They do not only have to deal with greater

(20)

competition, but also have to coordinate a web of foreign operations and subsidiaries (Lu & Beamish, 2004). Thus, studying the performance implications of ambidexterity in the context of MNEs is important, as these implications might not be the same as in a single country context. Moreover, although Hsu, et al. (2013) consider the existence of moderating variables influencing the effect of international ambidexterity on MNE performance, it would be bold to suggest that their list is complete. As such, still little is known about the contextual factors that moderate the relationship between international ambidexterity and MNE performance. Thus, an interesting avenue of research is to extend the study of international ambidexterity to developed countries and to explicitly consider the potentially moderating role of factors in MNEs’ external and internal environment, which might be of relevance. These thoughts lead to the following research question: What is the effect of international ambidexterity on an MNE’s

performance and are there internal and external factors that moderate the relationship between these two?

3. Conceptual framework

3.1. International ambidexterity

Due to the fast-changing, unstable and complex nature of today’s business environment, MNEs cannot solely rely on the possession and deployment of their existing resources and capabilities (Luo, 2000), but they also have to create new competences if they want to be successful (Hitt, et al., 2016). Initially, international expansion requires distinct and inimitable resources, capabilities or knowledge which set MNEs apart from their competitors. These distinct competences allow internationalizing firms to offset the hazardous effects of the liability of foreignness and give them a competitive edge in their foreign markets (Luo, 2000; Prange & Verdier, 2011). By means of FDI, these MNE competences are typically transferred from the parent companies to the foreign affiliates. The transfer of existing capabilities and knowledge across borders is characteristic of exploitation (Hsu, et al., 2013). However, faced with

(21)

ever-shortening product life cycles and the rapid obsolescence of once valuable knowledge and technologies, the creation and acquisition of new knowledge and capabilities have become indispensable (Luo, 2000; Prange & Verdier, 2011). Therefore, acquiring new knowledge and capabilities has emerged as a strong motive for FDI (Hitt, et al., 1997; Hsu, et al., 2013; Makino, et al., 2002). Put simply, relying on existing resources and capabilities does not offer an MNE a sustainable competitive advantage anymore. This means that internationalization requires regular renewal of an MNE’s existing resource and knowledge base. However, international expansion does not only require renewal, it also offers opportunities to accomplish it. Through exposure to diverse cultural attitudes, institutions, ideas and local knowledge, internationalization can serve as a stimulus for the creation of new competences (Luo, 2000). The acquisition of new knowledge and development of new capabilities through FDI is characteristic of exploration (Hsu, et al., 2013). Unlike exploitative FDI, which involves the transfer of competences from the parent firm to its subsidiaries, explorative FDI - if successful - typically leads to knowledge and capability flows from a focal subsidiary to the rest of the MNE network (Luo, 2000).

Only the MNEs which find a way to pursue both exploitative and explorative activities along their internationalization path can sustain continuous growth and remain competitive. The ability to achieve this dual orientation is denoted international ambidexterity (Prange & Verdier, 2011). Although the strategic intent behind exploitative and explorative activities radically differs, pursuing both types of activities allows MNEs to boost their competitive position and to optimally respond to the contemporary competitive challenges. In terms of internationalization, this means that MNEs should sustain explorative and exploitative FDI at the same time (Hsu, et al., 2013). As has already been discussed in the previous chapter, the rationale behind this assertion is that solely focusing on either exploitation or exploration leads to competitive disadvantages (Prange & Verdier, 2011). One the one hand, an excessively strong focus on exploitation may initially allow MNEs to overcome the liability of foreignness

(22)

in foreign markets (Luo, 2000). Over time, however, as competitive conditions change and once valuable capabilities become obsolete, these MNEs lose their competitive edge. On the other hand, an overly strong explorative focus leads to unrewarding renewal of knowledge and capabilities without reaping the benefits of this renewal (March, 1991). As such, MNEs’ short and long run success requires a balance between exploitation and exploration, i.e., they have to be internationally ambidextrous (Prange & Verdier, 2011). International ambidexterity allows firms to simultaneously overcome the liability of foreignness and prepare for environmental changes which open up opportunities for profitable renewal. Thus, it can be expected that more internationally ambidextrous firms depict higher performance than firms with a high imbalance between exploitation and exploration (Hsu, et al., 2013). Finally, given that investing in geographical expansions, and FDI in particular, is a resource-intensive undertaking, the continuum/balance view of ambidexterity - as opposed to the orthogonality/joint magnitude view - is most appropriate in the context of internationalization. Indeed, it cannot be expected that internationalizing MNEs are exempted from resource constraints and can endlessly invest in exploitative and explorative undertakings (Gupta, et al., 2006). These thoughts lead to the following hypothesis.

Hypothesis 1: There is a positive relationship between international ambidexterity and an

MNE’s performance.

As has already been illustrated in the previous chapter of this paper, the idea of a causal relationship between (international) ambidexterity and firm performance outcomes has been empirically supported by a number of studies (e.g., Hsu, et al., 2013; Cao et al., 2009). Adopting a contingency theory perspective, the main contribution of this paper lies in investigating the existence of endogenous and exogenous factors that possibly influence this relationship. The expectation is that international ambidexterity is not equally beneficial to every MNE and that the nature of the relationship between international ambidexterity and MNE performance may

(23)

studies investigating the relationship between internationalization and MNE performance (Hennart, 2007), many authors argue that the investigation of possible contingencies is necessary to account for MNEs’ internal and external environment in the context of internationalization (e.g., Banalieva & Sarathy, 2011; Vermeulen & Barkema, 2002). Complexities in both environments possibly affect the relationship between firm-level characteristics and MNE performance (Vermeulen & Barkema, 2002). As such, given that international ambidexterity is – in the context of this paper - defined as achieving a balance between exploitative and explorative FDI, i.e., exploitative and explorative internationalization (Hsu, et al., 2013), it can indeed be expected that there are critical contingencies influencing the relationship between international ambidexterity and MNE performance. In order to thrive, MNEs must be cognizant of these contingencies, as they impact the strategic requirements for being successful (Child, et al., 2003). Specifically, this study examines the possibly moderating role of an MNE’s geographical scope, number of foreign subsidiaries and tangible resource base.

3.2. Internationalization

In the IB literature, many studies have investigated the impact of internationalization on MNE performance (e.g., Lu & Beamish, 2004; Geringer, et al., 1989; Borda, et al., 2017). In this context, Vermeulen & Barkema (2002) argue that the concept of internationalization can be partitioned into two distinct, but related dimensions. The scope dimension of internationalization refers to the total number of countries an MNE has entered, whereas the scale dimension refers to the total number of foreign affiliates that have been established or acquired by an MNE. Large scale does not necessarily mean large scope, as MNEs may establish a large number of foreign subsidiaries in a narrow geographical area. Thus, building on the insight that scale and scope relate to different aspects of the internationalization phenomenon and that they may have different effects (Vermeulen & Barkema, 2002), they are

(24)

3.2.1. Geographical scope

The more countries an MNE enters on its internationalization path, i.e., the larger its geographical scope, the larger the diversity of informal and formal institutions it is exposed to (Hitt, et al., 1997). With a large geographical scope, MNEs have to deal with a variety of different cultural values, customs, norms, religious beliefs and languages, making effective coordination of foreign operations challenging (Vermeulen & Barkema, 2002). Differing law systems, regulations, trade regimes and possible exchange risk effects additionally hinder the task of effectively coordinating a network of foreign subsidiaries (Hitt, et al., 1997). Indeed, with increasing geographical scope of operations, the amount of information that has to be processed becomes increasingly difficult to handle (Lu & Beamish, 2004). Moreover, operating in a large number of countries with high institutional diversity makes it more challenging for an MNE to digest the knowledge and experience it gathers (Vermeulen & Barkema, 2002). MNEs have to become familiar with the different context-specific requirements in each host country and understand the national specificities of doing business (Zaheer, 1995). Aside from dealing with the institutional diversity, this also involves learning about different customer demands and competitive behaviors, and establishing new relationships with distributors, government officials and suppliers. The larger an MNE’s geographical scope, the higher the attention and time needed to accomplish these things (Vermeulen & Barkema, 2002). Indeed, maneuvering through an ever-increasing web of different institutional, cultural and competitive requirements is problematic (Wu, 2013). In other words, larger diversity in host country contexts does not only lead to higher complexity in coordinating foreign activities, but it also makes learning increasingly time-consuming and inefficient. This suggests that fully reaping the benefits of maintaining an effective balance between exploitative and explorative expansions becomes increasingly challenging, the larger a firm’s geographical scope.

(25)

particularly because each host country context requires very specific and often unique adaptations to firm capabilities, knowledge and resources (Hitt, et al., 2016). Thus, given these increased impediments to an effective cross-border transfer of knowledge and capabilities, fully reaping the advantages of international ambidexterity becomes increasingly problematic. With increasing geographical scope, exploitative as well as explorative FDI become less effective. In contrast, MNEs with a narrower geographical scope may find it easier to effectively transfer knowledge and capabilities across borders (Geringer, et al., 1989). In sum, a larger geographical scope and the accompanying challenges of more difficult coordination, learning and knowledge transfer, have a bearing on how much benefits an MNE can reap from being internationally ambidextrous. International ambidexterity is expected to be most beneficial, the lower the complexity and uncertainty surrounding an MNE’s operations (Hsu, et al., 2013). These thoughts lead to the following hypothesis.

Hypothesis 2: An MNE’s geographical scope negatively moderates the relationship between

international ambidexterity and MNE performance.

3.2.2. Number of foreign subsidiaries

The second aspect of international expansion is the scale dimension or the number of foreign subsidiaries established and acquired by an MNE. On the one hand, a firm with a high number of foreign affiliates has to adapt its structure and processes to account for the particularities of doing business on an international level. This is no cost-free undertaking and does not only require adaptation, but also unlearning of established routines and beliefs of how to do things (Vermeulen & Barkema, 2002). With a limited number of foreign affiliates, however, such an adaptation may not be necessary, as the domestic market remains an MNE’s main priority. Furthermore, with a growing number of foreign subsidiaries, it becomes increasingly difficult to effectively coordinate and align the activities of different foreign subsidiaries (Hitt, et al., 1997; Lu & Beamish, 2004), regardless of whether their purpose is exploitative or explorative.

(26)

On the other hand, with a growing number of foreign subsidiaries, MNEs are able to spread the costs of manufacturing, marketing and R&D across a larger number of affiliates and achieve economies of scale (Hitt, et al., 1997). This gives MNEs the possibility to increase the efficiency and profitability of their exploitative and explorative operations (Vermeulen & Barkema, 2002). In addition, with many foreign subsidiaries, MNEs can maximize the benefits of exploitative and explorative investments. A higher number of exploitative subsidiaries allows MNEs to make better use of their existing knowledge and capabilities, and to optimally reap the benefits of their superior assets in foreign markets (Lu & Beamish, 2004). Similarly, with a higher number of explorative subsidiaries, MNEs can accumulate, develop and transfer more knowledge, capabilities and innovative ideas. This particularly concerns knowledge developed within individual subsidiaries and subsequently disseminated across the MNE network. Some subsidiaries may even become specialized knowledge and capability creation centers (Vermeulen & Barkema, 2002). Finally, as has already been discussed, a number of scholars do not see ambidexterity as the capability of achieving a balance between exploitation and exploration, but rather as “a firm’s effort to increase the combined magnitude of both” (Cao, et al., 2009, p. 782). The rationale is that exploitation and exploration are complementary activities which are supportive of each other. Empirical evidence suggests that both balancing exploitation and exploration, as well as increasing the combined magnitude of them are associated with higher firm performance (Gibson & Birkinshaw, 2004; He & Wong, 2004; Cao, et al., 2009). Combining the balance and joint magnitude views of ambidexterity, this suggests that achieving a balance between exploitative and explorative operations among many foreign subsidiaries should prove to be more beneficial to an MNE than achieving a balance among fewer foreign subsidiaries. In sum, this means that the extra costs of having many foreign affiliates are expected to be offset by the benefits of economies of scale and higher joint magnitude of exploitative and explorative learning. The positive effect of international

(27)

ambidexterity is thus expected to be higher, the larger the number of foreign subsidiaries. This reasoning leads to the following hypothesis.

Hypothesis 3: The number of an MNE’s foreign subsidiaries positively moderates the

relationship between international ambidexterity and MNE performance.

3.3. Size of the tangible resource base

Broadly speaking, the larger an MNE’s tangible resource base, the larger its room for mistakes. An MNE’s financial power is usually an important indicator of the size of its tangible resources (Chang & Rhee, 2011). In this context, Chang & Rhee (2011) make use of the notion of slack resources. These resources give a company higher strategic flexibility and can be used for purposes that are not directly necessary for company survival. Slack or tangible resources serve as a hedge against environmental uncertainty and risk (Cao, et al., 2009), and allow MNEs to pursue a greater diversity of strategic alternatives. Firms which possess a low amount of slack resources typically shy away from risk-taking behavior, as it is more difficult for them to cope with potential setbacks. On the contrary, firms with higher availability of slack resources are more likely to take risks, as emerging challenges can be more effectively addressed and potential setbacks can be more easily absorbed (Chang & Rhee, 2011). It can be expected that the higher the level of tangible resources an MNE has at its disposition, the less important it is to achieve a balance between exploration and exploitation (Cao, et al., 2009). Indeed, the main logic behind achieving a balance between exploration and exploitation is to avoid the adverse performance effects of either capability and knowledge obsolescence (too much exploitation) or unrewarding renewal (too much exploration) (March, 1991). With higher availability of tangible resources, the negative performance effects of an imbalance between exploitation and exploration are less severe, as they can be more easily absorbed by the focal MNE (Chang & Rhee, 2011). Thus, although achieving a balance between exploitation and exploration is a necessary capability for all MNEs, international ambidexterity is presumably of particular

(28)

importance to MNEs with fewer slack resources, as their room for mistakes is smaller. This reasoning is also in line with Gupta, et al. (2006) who argue that achieving a balance between exploitation and exploration is most important to firms which are subject to tight resource constraints. With higher availability of resources, the competition between exploitative and explorative activities for these resources is alleviated and the trade-off between the two is of lesser significance. These thoughts lead to the following hypothesis.

Hypothesis 4: The size of an MNE’s tangible resource base negatively moderates the

relationship between international ambidexterity and MNE performance.

The investigated relationships and hypothesized effects are summarized in the framework presented in Figure 1.

Figure 1: Conceptual framework

International ambidexterity MNE performance

Size of tangible resource base

H1 (+)

H4 (-)

Geographical scope Number of foreign subsidiaries

(29)

4. Methodology

4.1. Sample and data collection

The final sample consists of 105 European companies operating in ‘Computer and Electronic Product Manufacturing’ according to the 2012 North American Industry Classification System (NAICS 2012). Firms in high-technology industries typically operate in unstable and dynamic environments, characterized by short-lived competitive advantages and frequent technological change (Wu, 2007). Given this dynamism, high-technology industries provide an appropriate context for this study, as ambidexterity is expected to be of lesser importance in more stable environments (Gibson & Birkinshaw, 2004; Raisch & Birkinshaw, 2008). The focus on high-technology industries is also consistent with former research on the implications of ambidexterity (Hsu, et al., 2013; Patel, et al., 2013; Cao, et al., 2009). Bureau van Dijk’s Orbis database is used to compile an appropriate sample of MNEs. This database “is one of the most comprehensive and inter-temporal pan-European databases containing detailed information about many public and private companies in virtually all European countries” (De Jong & Van Houten, 2014, p. 317). It allows to filter companies according to their industry classification and stock exchange listing. Given that the data for non-listed companies is often incomplete, only publicly listed companies with more than one active foreign subsidiary are considered. In addition, the Orbis database contains a detailed list of each MNE’s directly-owned subsidiaries and their respective locations. Finally, only entities in which the ownership stake is at least 10% are considered. Data for the measurement of the variables was collected for a single year, i.e., 2015.

4.2. Variables and measures

Dependent variable. MNE performance is measured as the average return on assets (ROA) for

(30)

international business research (e.g., Vermeulen & Barkema, 2002; Chang & Rhee, 2011; Lu & Beamish, 2004). Instead of focusing on the ROA of an individual year, the average ROA is used to reduce potential bias effects of single-year outliers (Georgakakis & Ruigrok, 2017). Moreover, given that the effects of explorative activities are expected to manifest themselves only after a longer period of time (March, 1991), focusing on average MNE performance is more appropriate.

Independent variable. International ambidexterity is measured as the inverse value of the ratio

of the absolute difference between an MNE’s foreign exploitative subsidiaries and foreign explorative subsidiaries (Hsu, et al., 2013), and the total number of foreign subsidiaries (see Figure 2). As such, each MNE’s foreign subsidiaries are classified as either exploitative or explorative. Given that the Orbis database provides only incomplete information on indirectly-owned subsidiaries, i.e., subsidiaries of subsidiaries, only directly-indirectly-owned subsidiaries can be considered in this paper. Building on Hsu, et al. (2013), manufacturing subsidiaries are classified as exploitative, whereas R&D and marketing subsidiaries are classified as explorative. However, it should be noted that this classification scheme is incomplete. Indeed, some subsidiaries perform functions (e.g., sales), which are not considered by Hsu, et al. (2013). Therefore, subsidiaries which are responsible for administrative functions, distribution, retail or sales activities are also considered as exploitative. Other subsidiaries perform exploitative and explorative functions (e.g., R&D and manufacturing). These subsidiaries are considered as half exploitation oriented and half exploration oriented, i.e., 0.5 is added to each count. In this instance, it is also important to mention that this paper distinguishes between sales and marketing activities. Only those subsidiaries for which there is explicit information that they undertake marketing activities are considered as explorative. The vast majority of marketing subsidiaries also perform sales activities and are thus almost exclusively considered as half exploitative and half explorative. Subsidiaries for which there is no explicit information that their purpose goes beyond that of pure sales entities are considered as purely exploitative.

(31)

Finally, individual subsidiaries that have not been set up by the MNE itself, but have been added to the company group by means of an acquisition, are classified according to a different categorization scheme. Subsidiaries that have been acquired to get access to the target company’s knowledge, patents, technologies, expertise, innovative products/services or R&D capabilities are considered as purely explorative. Although these acquisitions naturally also involved the addition of manufacturing capacity or sales volume, categorizing these subsidiaries as half exploitative and half explorative, would distort their true purpose. Similarly, acquisitions that have been undertaken for the sole purpose of geographical expansion and extending the firm’s sales or distribution network are considered as purely exploitative. Thus, for individually acquired subsidiaries, the purpose of the acquisition - instead of the subsidiaries’ function after the completion of the respective acquisition - is decisive for their categorization. However, if an acquisition led to the addition of more than one subsidiary to an MNE’s group structure, each subsidiary is categorized according to its specific function, making use of the categorization rule already specified above. Subsidiaries which are identified as holding companies are not considered in this paper, as they do not directly perform a value-adding activity. The information on subsidiaries’ specific tasks was collected through annual reports, the Orbis database, company websites, diverse national company registries and other online sources such as newspaper articles. The complete categorization scheme is summarized in the following table.

(32)

Table 1: Categorization scheme

After the categorization step, the absolute difference between exploitative and explorative subsidiaries is taken as a measure of imbalance between exploitation and exploration. This approach is consistent with prior research (Cao, et al., 2009; He & Wong, 2004; Hsu, et al., 2013). The absolute difference is divided by the total number of foreign subsidiaries to relate the degree of imbalance between exploitation and exploration to the size of an MNE’s foreign operations. For instance, an MNE with 25 exploitative subsidiaries and 20 explorative subsidiaries has the same absolute imbalance as an MNE with 5 exploitative subsidiaries and no explorative subsidiaries. Intuitively, however, it appears clear that the first MNE has achieved a more effective balance between exploitation and exploration than the latter. Putting the absolute difference between exploitative and explorative foreign subsidiaries in relation to the total number of subsidiaries thus allows for a measure of relative imbalance, which more

Exploitative subsidiaries Explorative subsidiaries

Manufacturing Marketing

Sales/Wholesales Research and development (R&D)

Distribution/Retail

Individual acquisitions (Purpose: access to knowledge, patents, technologies, expertise,

new products and R&D capabilities)

Administrative tasks

Individual acquisitions (Purpose: geographical expansion and extension of

(33)

accurately reflects an MNE’s ambidextrous orientation. As a final step, to allow an easier interpretation of the results, this measure is reversed, so that higher numbers indicate a higher level of international ambidexterity (Hsu, et al., 2013). The complete formula to measure an MNE’s degree of international ambidexterity is depicted in Figure 2. IAi stands for the degree

of international ambidexterity of firm i. Exploiti and Explori stand for the number of exploitative

and explorative subsidiaries established abroad respectively. ForeignSubsi is the total number

of subsidiaries established abroad by firm i, i.e., the sum of exploitative and explorative subsidiaries.

Figure 2: Measure of international ambidexterity

Moderator variables. Geographical scope is measured as the total number of countries in which

an MNE has established directly-owned subsidiaries. The number of foreign subsidiaries is measured as the sum of all the directly-owned subsidiaries a firm has established abroad (Vermeulen & Barkema, 2002). Building on Chang & Rhee (2011), the size of an MNE’s tangible resource base is measured as the ratio of an MNE’s total debt and total assets. The lower this ratio, the higher an MNE’s flexibility in deploying its resources, i.e., the larger the size of its intangible resource base. A high ratio is indicative of financial restraints which limit an MNE’s capability to further debt finance its operations (Chang & Rhee, 2011). To allow a more convenient interpretation of the results, this measure is reversed, so that higher values indicate a larger tangible resource base. The reversing procedure is analog to the procedure for the measurement of international ambidexterity (see Figure 2).

Control Variables. A number of control variables are considered, due to their potential effect

|

Exploit

i

- Explor

i

|

ForeignSubs

i

(34)

economies of scale and is operationalized as a firm’s total assets (Vermeulen & Barkema, 2002). In addition to the effects of economies of scale, larger MNEs may also be able to set up more efficient transportation and communication systems across borders. Finally, larger firm size is typically associated with an MNE’s ability to exercise market power over its stakeholders (Oviatt & McDougall, 2005). Second, firm age is considered, as more experienced MNEs might depict higher performance than less experienced MNEs (Hsu, et al., 2013) and is measured as the number of years passed since the firm’s establishment (Chang & Rhee, 2011; Cao, et al., 2009). With more (international) experience, an MNE may be better able to approach the challenges of fierce domestic and foreign competition, and is better prepared to deal with the diversity of different contexts across countries. With more experience, which is typically the result of a process of incremental learning and gradual development of skills and knowledge (Ogasavara & Hoshino, 2008), MNEs can attenuate the adverse effects of competitive and institutional pressures (Li, et al., 2014) and may thus depict higher performance. Third, a firm’s R&D intensity is considered and is measured as the ratio of an MNE’s R&D expenditures to its sales revenue (Hsu, et al., 2013; Lu & Beamish, 2004). In the past, many studies have shown that more innovative firms do indeed depict higher performance (Rubera & Kirca, 2012). Fourth, subsectors of ‘Computer and Electronics Product Manufacturing’ are considered to account for possible sub-industry effects. Finally, MNEs’ home base is considered to account for possible home country effects. All the data for the measurement of the control variables was retrieved from the Orbis database.

5. Results

5.1. Descriptive statistics

Table 2 shows the country distribution of the 105 companies in the final sample. As can be seen, most companies originate from Germany (28.6%), followed by France (18.1%), Sweden (16.2%), the Netherlands (13.3%) and Finland (10%). Belgium, Switzerland, Norway and

(35)

Austria jointly only make up 14.3%. Table 3 depicts the distribution of the firms for the subsectors of the ‘Computer and Electronic Product Manufacturing’ industry. Most MNEs in the data sample operate in ‘Semiconductor and Other Electronic Component Manufacturing (3344)’ (41.9%), followed by ‘Manufacturing and Reproducing Magnetic and Optical Media (3346)’ (23.8%), ‘Navigational, Measuring, Electromedical, and Control Instruments Manufacturing (3345)’ (15.2%), ‘Communications Equipment Manufacturing (3342)’ (11.4%) and ‘Computer and Peripheral Equipment Manufacturing (3341)’ (7.6%).

Table 2: Country distribution

Country Percentage (%) Frequency

Germany 28.6 30 France 18.1 19 Sweden 16.2 17 The Netherlands 13.3 14 Finland 9.5 10 Belgium 3.8 4 Switzerland 3.8 4 Norway 3.8 4 Austria 2.9 3 Total 100.0 105

(36)

Table 3: Industry distribution

Industry subsector Percentage (%) Frequency

Semiconductor and Other Electronic Component

Manufacturing (3344)

41.9 44

Manufacturing and Reproducing Magnetic and

Optical Media (3346)

23.8 25

Navigational, Measuring, Electromedical, and Control

Instruments Manufacturing (3345)

15.2 16

Communications Equipment

Manufacturing (3342) 11.4 12

Computer and Peripheral Equipment Manufacturing

(3341)

7.6 8

Total 100.0 105

Table 4 contains the descriptive statistics and correlation measures for the dependent variable, the independent variable and the remaining control variables. On average, MNEs in the data sample operate in 6.92 foreign countries with 8.82 foreign subsidiaries and depict a negative performance over the period 2013-2015. Unsurprisingly, firm size has a high positive correlation with the number of foreign subsidiaries and geographical scope, respectively.

(37)

internationalized, as small firm size prevents internationalization (Oviatt & McDougall, 2005). Moreover, MNEs’ number of foreign subsidiaries, geographical scope, firm age and size of the tangible resource base depict a positive correlation with MNE performance respectively. Furthermore, Table 4 suggests that MNEs with a higher degree of international ambidexterity tend to be older, have a higher R&D intensity, operate in more countries and have established more foreign subsidiaries. As can be seen, the correlation between MNEs’ geographical scope and number of foreign subsidiaries is 0.959, suggesting that these two variables suffer from multicollinearity. The values of the variance inflation factors (VIFs) and tolerance levels for these two variables confirm this, as they are above and below the critical values of 10 and 0.2 respectively (Field, 2013). The VIF values and tolerance levels for the other two predictor variables are both around 1.1 and 0.9 respectively. Although a relatively high correlation between the number of foreign subsidiaries and geographical scope could be anticipated, a correlation above 0.9 comes somewhat unexpected. The reason why an MNE’s geographical scope and number of foreign subsidiaries are treated separately in this study is that past research has found evidence for the differing effects of the two (Vermeulen & Barkema, 2002).

The very high correlation can be explained by three main factors. First, some of the companies in the dataset have more than one holding company, which are very often incorporated in a single country and/or in countries where MNEs have already established other subsidiaries. Overall, their exclusion has led to a higher correlation between MNEs’ geographical scope and number of foreign subsidiaries than is the case in reality. Indeed, an analysis of the data shows that their exclusion has led to a small increase in the correlation between MNEs’ number of foreign subsidiaries and their geographical scope from 0.953 to 0.959. Second, due to data restrictions, indirectly-owned subsidiaries cannot be considered. This naturally leads to an underestimation of MNEs’ number of foreign subsidiaries and, to a lesser degree, their geographical scope. Overall, this may again lead to an overestimation of the correlation between these two variables, although this remains naturally speculative at this point. Third, a large