Directors crossing borders: The determinants of the dominant direction of director movement within a dyad of countries

1

Directors crossing borders: The determinants of the dominant direction of director

movement within a dyad of countries

June, 2013 University of Groningen

Mike Kruijer BSc Studentnumber: s1884131 Email: m.kruijer.1@student.rug.nl

Master Thesis MSc International Business & Management Supervisor: dr. van Veen

Co-assessor: dr. Ritsema

University of Groningen Faculty of Economic and Business

2

ABSTRACT

3

ACKNOWLEDGEMENT

Writing a master thesis is a large project, one which requires the help and support of other people to complete in a proper fashion. As such, I would like to take this opportunity to express my gratitude to all those who were somehow involved in the process.

First of all, I would like to thank my first supervisor, dr. Kees van Veen for all the support I received during the writing process of this thesis. The unannounced visits to your office, the difficult questions I often posed and the silent moments where we both were just staring somewhere in front of us to think of a solution for a new (methodological) difficulty I ran into, I hope it was all worth it in the end. Second, I would like to thank Sjors van der Heide for his critical remarks and academic eye and Torben Bethke for always trying to push me to higher levels. Your critical remarks and tips were highly appreciated.

Third, I would like to use this opportunity to thank my family and friends for their emotional support and for their constant interest in the process of this thesis. A special thanks goes out to my parents; I couldn’t have done this without your love and (financial) support.

Last but not least, I would like to thank Lissa for always being able to find the right words to motivate me and for always pushing me in the right direction. You’re support has been invaluable.

4

LIST OF FIGURES

Figure 1: Illustration of the unit of analysis ... 9

Figure 2: Graphical depiction of the dependent variable ... 32

Figure 3: SPSS output for OLS ... 67

LIST OF TABLES

Table 2: Overview of hypothesized directions between countries for H1 ... 39

Table 3: Absolute amount of directors moving between countries for H1 ... 40

Table 4: Relative amount of moving directors between countries for H1 ... 41

Table 5: Absolute amount of moving directors for H1, multiplied by the population factor ... 43

Table 6: Absolute amount of moving directors for H1, multiplied by the GDP (PPP) factor ... 43

Table 7: Relative amount of moving directors for H1, multiplied by population factor ... 44

Table 8: Relative amount of moving directors for H1, multiplied by GDP (PPP) factor ... 45

Table 9: Relative numbers of moving directors for H2 ... 49

Table 10: Relative numbers of moving directors for H3a ... 50

Table 11: Relative numbers of moving directors for H3b ... 53

Tabel 12: Relative numbers of moving directors for H4a (Population)... 54

Tabel 13: Relative numbers of moving directors for H4a (GDP adjusted for PPP) ... 55

Table 14: Hypothesized dominant direction based on FDI (H4b) ... 57

Table 15: Relative numbers of moving directors for H4b ... 58

Table 16: Relative numbers of moving directors for H4c ... 60

Table 17: Relative numbers of moving directors for H5 ... 61

5

TABLE OF CONTENT

2.1.1 Incentives for Interlocking ... 11

2.1.2 Consequences of Interlocking ... 13

2.1.3 Direct and Indirect Ties ... 13

2.1.4 The Interorganizational and Intraclass Perspectives ... 14

2.2 Upper Echelons... 15

2.2.1 The TMT ... 16

2.2.2 The Upper Echelons Perspective ... 16

2.3 Internationalization of the TMT ... 17

2.3.1 Internationalization Theory and Board Members ... 17

2.3.2 Diversity of the TMT ... 18 3. HYPOTHESES ... 20 3.1 Varieties of Capitalism ... 20 3.1.1 Supply of Finance ... 21 3.1.2 Governance regimes ... 22 3.1.3 Underlying assumption ... 24 3.2 Cultural Tightness ... 25 3.3 Institutional Distance ... 26

3.4 Country Size, FDI, and the Quantity of Large Firms ... 27

3.5 The Market for Top Managers ... 29

6

5.1.2 Cultural Tightness ... 47

5.1.3 Timing of Entering the European Union ... 48

5.1.4 Quality of Institutions ... 51

5.1.5 Country Size ... 52

5.1.6 Foreign Direct Investment ... 56

5.1.7 Quantity of Large Firms ... 59

5.1.8 The Market for Top Managers ... 62

5.2 Statistical Analysis ... 63

5.2.1 Ordinary Least Squares ... 63

5.2.2 Statistical Input in More Detail ... 63

5.2.3 Preparing the OLS... 65

5.2.4 Performing the Regression ... 66

5.2.5 Robustness Check ... 67

5.2.6 Interpreting the Results ... 68

6. DISCUSSION ... 74

6.1 Discussing the Results ... 74

6.2 Limitations & Recommendations for Future Research ... 78

7. CONCLUSION ... 80

8. BIBLIOGRAPHY ... 82

9. APPENDICES... 93

9.1 Appendix 1: Country characteristics... 93

9.2 Appendix 2: The Institutional Pillar ... 95

9.3 Appendix 3: Factor Scores ... 96

9.4 Appendix 4: FDI Data 1997 - 2007 ... 97

9.5 Appendix 5: Relative FDI data 1997 - 2007 ... 98

9.6 Appendix 6: P-P plot ... 99

9.7 Appendix 7: Scatterplots ... 100

9.8 Appendix 8: Bivariate Correlation Matrix ... 101

7

1. INTRODUCTION

Globalization is an increasingly important phenomenon and has been so since the introduction of steamships, railroads, and the telegraph (Baldwin, 2006). The globalization trend is said to include developments such as increasing international competition, the internationalization of production, international labor migration, new forms of international governance, and the decline and disintegration of the nation-state (Carroll & Fennema, 2002; Therborn, 2000). On the board level, the emerging global economy has led to a boost in the number of foreign board members originating from different markets (Staples, 2007; van Veen & Marsman, 2008; van Veen, Rao Sahib, & Aangeenburg, 2013). Companies are adapting their board recruitment policies to acquire high levels of international knowledge, relations, and experience (Carpenter, Geletkanycz, & Sanders, 2004; Nielsen, 2010; van Veen, Rao Sahib, & Aangeenburg, 2013), which has been found to lead to a greater propensity to enter foreign markets, eventually resulting in higher firm performance (Carpenter, Geletkanycz, & Sanders, 2004; Carter, Simkins, & Simpson, 2003; Masulis, Wang, & Xie, 2012; Miller & Carmen Triana, 2009). Though companies can hire foreign board members directly from the foreign country, another option is to recruit internationally experienced top managers that form interlocking directorates. This practice has been the focus of political and scholarly interest since the early 20th _{century (Burris, 2005) and requires a definition. ‘An interlocking directorate}

occurs when a person affiliated with one organization sits on the board of directors of another organization’ (Mizruchi, 1996, p. 271). The interlocking directorates resulting from this practice create a complex pattern of connections between nominally independent firms (Koenig, Gogel, & Sonquist, 1979).

The significance of director interlocks for firms cannot be located at the level of the specific interfirm dyad, but must be interpreted as a general resource that facilitates the flow of communication, monitoring of events, or projection of influence across the larger corporate network (Burris, 2005; Mintz & Schwartz, 1981; Mizruchi & Stearns, 1988). Research on interlocking directorates has been greatly enriched by studies that shift the focus of attention from dyads of interconnected firms to the structural position of firms within the overall interlock network (Burris, 2005) and are most usefully analyzed at the aggregate level (Mintz & Schwartz, 1981). Exploring interlocking directorates allows researchers to map out the interorganizational network of corporate governance (Peng, 2001) and may work as useful diagnostics for the existence of ongoing and potential relationships among major corporations (Mintz & Schwartz, 1981).

8 has been explained in different ways (Mizruchi, 1996; Nollert, 2005; van Veen & Kratzer, 2011). However, most of these studies have used old data, often focusing on one specific country, and omitted making any comparisons (van Veen & Kratzer, 2011). Furthermore, complementing studies on interlocking directorates, studies on nationality diversity within upper echelons have highlighted the importance of nationality diversity on boards and its significant impact on firm performance and stockholder returns, but do not help in understanding where new international board members come from in the first place (van Veen, Rao Sahib, & Aangeenburg, 2013). The studies of van Veen, Rao Sahib, and Aangeenburg (2013) and van Veen and Kratzer (2011) are helpful in understanding the relative numbers of international board members within countries. However, they only present the total amount of director movement between two selected countries and do not explore the actual direction of the movement of directors. These authors pointed out a gap within the current literature: analyzing the actual direction of the movement of directors on an aggregate level.

1.1 Literature Gap

This research aims to fill this gap by analyzing the dominant direction of director movement in order to test whether mobility patterns between two countries are reciprocal or dominated by one in the dyadic relationship between two countries.

Analyzing the actual direction of the movement of directors can be interesting for a variety of reasons. First and foremost, it may enhance understanding of how boards balance the necessity to internationalize and increase the nationality diversity on the one hand, with the risks of becoming too diverse on the other (van Veen, Rao Sahib, & Aangeenburg, 2013). While a number of studies showed that high nationality diversity in upper echelons leads to a greater propensity to enter foreign markets, resulting in higher performance (Carpenter, Geletkanycz, & Sanders, 2004; Masulis, Wang, & Xie, 2012; Miller & Carmen Triana, 2009; van Veen, Rao Sahib, & Aangeenburg, 2013), increasing diversity is not free of risks and can be seen as a two-edged sword (van Veen, Rao Sahib, & Aangeenburg, 2013); it may lead to the formation of subgroups within teams (Earley & Mosakowski, 2000), foster conflict and turnover, and diminish morale, cohesion and group performance (Barkema & Shvyrkov, 2007; Harrison & Klein, 2007). Thus, too much diversity can lead to a negative effect on the performance of companies (van Veen, Rao Sahib, & Aangeenburg, 2013). As such, compan/ies may recruit directors from abroad, but will try to minimize differences at the same time (van Veen, Rao Sahib, & Aangeenburg, 2013).Hence, analyzing the direction of interlocking directorates enables descriptive indicators of how companies deal with this two-edged sword.

9 the development of a single European market (Heemskerk, 2011) and may explain why certain countries form the foundation of this development. This will add to the research stream on the presence of a transnational business community (Carroll & Fennema, 2002, 2004; Kentor & Jang, 2004).

Finally, analyzing the direction of interlocking directorates has a descriptive contribution since this provides the opportunity to empirically present and systematically compare director networks between countries.

1.2 Research Question

When incorporating the expected theoretical contributions of analyzing the direction of director movement, the following research question can be derived:

What factors are determinants of the dominant direction of director movement between two countries?

Figure 1: Illustration of the unit of analysis

In order to simplify the aim of this research question, it is illustrated in Figure 1. Thus, with representing the amount of outgoing directors from country A to country B and representing the reciprocal flow, what factors determine why ? With this research question, the unit of analysis of this relationship becomes the balance of the amount of interlocking directorates between two countries.

11

2. THEORETICAL FRAMEWORK

The lack of a cohesive conceptual framework for the study of moving directors has given rise to a body of literature that is inconsistent and confusing (Schoorman, Bazerman, & Atkin, 1981). As such, it is necessary to integrate several contributions from former studies. The following theoretical framework will present underlying theories and streams of research related to director movement such as interlocking directorates, upper echelons, and board diversity. They form the foundation of this empirical work and serves as critical background information for the remainder of this paper.

2.1 Interlocking directorates

Besides the hiring practices in which companies hire foreign directors to serve on their board, boards may also enhance the nationality diversity on their boards by means of interlocking directorates. As stated earlier, ‘an interlocking directorate occurs when a person affiliated with one organization sits on the board of directors of another organization’ (Mizruchi, 1996, p. 271). Following this definition, the coming paragraphs analyze the incentives for interlocking in the first place and examine the results of interlocking. Moreover, information is provided on different types of interlocking directorates and the two prevailing perspectives on interlocking are introduced: the interorganizational and intraclass approaches.

2.1.1 Incentives for Interlocking

The increasing European framework of interlocking directorates (Carroll & Fennema, 2002, 2004; Kentor & Jang, 2004) increases the need of understanding why interlocks are formed in the first place. Literature over the years has pinpointed a variety of incentives for the formation of interlocking directorates.

12 interlocks, so far the literature is uncertain on whether interlocking directorates actually restrict competition.

A second interpretation of interlocking is that it reflects attempts by organizations to co-opt sources of environmental uncertainty (Allen, 1974; Burt, 1980; Mizruchi, 1996; Palmer, Friedland, & Singh, 1986; Pennings, 1980; Pfeffer & Salancik, 1978; Schoorman, Bazerman, & Atkin, 1981; Zajac, 1988). Typically, organizations are assumed to be somewhat dependent upon one another, with the degree of interdependence between two organizations being a function of the extent to which they exchange resources (Allen, 1974; Burt, 1980). As a result, companies will follow strategies that permit them to reduce or control important sources of uncertainty in their environments that enables them to limit or anticipate potentially disruptive actions of other organizations (Allen, 1974). Interlocking directorates may be one of those strategies, as they provide their partners with the opportunity to exchange specific information about their operations or general information about the industrial sectors in which they are located (Palmer, 1983). Thus, interlocking directorates may be formed as a result of a desire to reduce environmental uncertainty for a firm.

Moreover, as a third incentive, inter-corporate relations and directorship interlocks in particular constitute a governance structure that is an alternative to market-hierarchy forms of economic organizations (Aguilera, 1998; Williamson, 1985). In order to minimize transaction costs, firms may use network structures such as sharing directors as alternatives to arms-length relationships and in-house ownership. Interlocking directorates are an important source of information and constitute a web of communication through which general business information and opinions can be transmitted, based on personal trust and individual integrity, hereby lowering transaction costs between firms (Aguilera, 1998; Carroll & Fennema, 2002; Mintz & Schwartz, 1981; Nollert, 2005; Nooteboom, 1999; Pfeffer & Salancik, 1978). Thus, depending on the nature of the organization’s environment, the management of an organization’s linkages to financial institutions, suppliers, and customers may be just as crucial to the effectiveness of the total organization as is internal management, with interlocking directorates representing a possible strategy for managing these linkages (Schoorman, Bazerman, & Atkin, 1981). In conclusion, corporate interlocking is a cooperative strategy for regulating the relationships between organizations which are to some extent dependent upon another (Allen, 1974) and has the potential to reduce transaction costs between firms.

13 interlocking directorates, since the concept of legitimacy is difficult to test empirically. As such, this incentive for the formation of interlocking directorates has not been confirmed through empirical means.

For the fifth and final incentive, it has to be recognized that interlocks occur between organizations, but are created by individuals (Mizruchi, 1996). As such, interlocks may also be formed following personal motives. Individuals may join boards for financial remuneration, prestige, and contacts that may prove useful in securing subsequent employment opportunities (Zajac, 1988). In addition, outside directors may also be chosen as individuals because they will add prestige to their organization and enhance the legitimacy of a firm (Mace, 1971). Thus, interlocking directorates may also be formed as the result of personal incentives of a director or because of certain characteristics or skills that a director might posses.

2.1.2 Consequences of Interlocking

‘If interlocks are to be worth studying, it is essential that they be shown to have consequences for the behavior of firms’ (Mizruchi, 1996, p. 280). Most of the analyses of the determinants of interlocks have implied various consequences for individual firms (Mizruchi, 1996), closely related to the incentives for interlocking.

Mizruchi (1996) identified four consequences of interlocking. First, interlocks are argued to enhance communication amongst competitors and function as collusive mechanism. Second, interlocks are assumed to pacify the resource provider’s management and function as mechanism of cooptation. That is, large organizations generally recognize the dependence and need for joint action with other firms (Mintz & Schwartz, 1981; Nollert, 2005). They are continuously aware of their interdependency relationships with other firms and the collective advantage of pursuing a qualified joint profit maximizing relationship (Williamson, 1965), which is why interorganizational relationships such as interlocking directorates are established that ensure the supply of scarce resources within the industry (Mintz & Schwartz, 1981). Third, interlocks are assumed to provide the monitoring firm with information on the receiving firm’s operations as well as potential influence on its operation and hence functions as monitoring mechanism. Finally, interlocks may serve as reflection of social cohesion, since they are assumed to facilitate the political unity necessary for effective political action (Burris, 2005; Mizruchi, 1992; Nollert, 2005).

2.1.3 Direct and Indirect Ties

14 company sits on the board of directors of another company (Mizruchi, 1996) – characterizes what is typically referred to as a direct interlock (Schoorman, Bazerman, & Atkin, 1981). However, firms may also be interlocked through a more complex pattern in which directors from two separate organizations are both in the board of a third organization, referred to as an indirect interlock (Schoorman, Bazerman, & Atkin, 1981). Both direct and indirect interlocks can yield benefits, although when an organization is requiring an individual for his/her expertise in a certain area or for enhancing its reputation, only direct interlocks would achieve these ends (Schoorman, Bazerman, & Atkin, 1981). For the purpose of this research, only direct interlocks will be used, since they are easier to trace and are expected to result in more reliable results compared to indirect interlocks.

2.1.4 The Interorganizational and Intraclass Perspectives

Interlocks have been studied from two different but compatible perspectives, which are called the interorganizational and the intraclass approaches (Palmer, 1983; Palmer, Friedland, & Singh, 1986). The interorganizational perspective proclaims that corporations create interlocks in response to their need for resources controlled by other organization in their environment (Aldrich, 1979; Burt, 1983; Mintz & Schwartz, 1981; Palmer, 1983; Palmer, Friedland, & Singh, 1986; Pfeffer & Salancik, 1978; Zajac, 1988), whereas the intraclass perspective asserts that corporate directors create interlocks in response to their need to organize themselves as a social class (Palmer, 1983; Palmer, Friedland, & Singh, 1986; Useem, 1997).

The interorganizational approach sees organizations as entities that possess interests (Palmer, 1983), which function as incentives to form relationships with other organizations. In pursuit of these interests, they form interlocking directorates which provides them with the opportunity to exchange specific information and to formulate policies that are more sensitive to their environment (Allen, 1974; Dooley, 1969; Palmer, 1983). In addition, the formation of interlocking ties between companies allows partners to influence one another’s board-level policies (Palmer, 1983) by appointing one of their own representatives on the board of another company (Palmer, Friedland, & Singh, 1986). The concept of the intraorganizational approach can best be summarized the following: ‘Interorganizational linkages helps firms manage dependence (…), through three kinds of linkages: information exchange, informal coordination (e.g., tacit price setting agreements), and formal coordination (e.g., long term contracts, joint ventures, interfirm stockholding)’ (Palmer, Friedland, & Singh, 1986, p. 783).

15 directorates. In this way, the resulting interlocking directorates can be perceived as a result of the relationships between the top managers and the other directors of the boards on which they sit (Palmer, 1983), leading to a complete assemblage which may be called a social network (Koenig & Gogel, 1981). Through this social network, ‘’interlocks facilitate direct interaction and the communication of techniques, values, and beliefs between directors (…), hereby guiding managerial behavior, socializing new directors into this culture, and socially controlling deviant managerial behavior’ (Palmer, 1983, p. 42). In addition, top managers can use their social contacts to get better paid jobs or executive compensations, which normally cannot be justified by market rules (Nollert, 2005). Finally, directors pursue the formation of multiple interlocks because their position in the boards social network is determined by his/her position in other social networks (Palmer, 1983; Useem, 1997). All in all, it can be postulated that there are copious personal incentives for top managers to form interlocking directorates.

However, rather than being two separate explanations for the phenomenon of interlocking directorates, the study of Palmer, Friedland, and Singh (1986) on the reconstitution of broken ties showed that the interorganizational and intraclass perspectives provide compatible and possibly complementary explanations of interlocking. Both the interorganizational and intraclass pose incentives for firms or the actors of firms to form interlocking ties and are both anteceding factors of the interlocking directorate network, as summarized in Table 1.

Incentives for interlocking Intraclass/Interorganizational Restricting Competition Interorganizational

Reducing Uncertainty Interorganizational

Lowering transaction costs Interorganizational

Organizational Prestige Interorganizational

Personal motives Intraclass

Table 1: Categorization of the incentives for interlocking As such, both perspectives will be used in the formation of the hypothesis.

2.2 Upper Echelons

16 the organization, functional background, education, socio-economic roots, and a manager’s financial position (Hambrick & Mason, 1984). Because of the idiosyncrasies of the decision makers, the UE perspective and the concept of directors crossing borders can be argued to be interrelated, since the characteristics and origin of an executive may explain aspects of the strategy of the firm.

2.2.1 The TMT

However, if one follows the approach of Hambrick and Mason (1984), it is imperative that the top management team (TMT) of a company is clearly defined and demarcated. After all, when testing the impact of top management characteristics on aspects of a firm’s strategy or performance, it is important to know whose characteristics to test. Since countries differ strongly in their corporate governance system (Aguilera & Jackson, 2010), there is a range of different structures for the top layers of firms. This raises the question; who actually constitutes the top management team?

When investigating the top management team, most studies use the construct postulated in the early 1960’s by Cyert and March (1963) in that one should look at the ‘dominant coalition’ within an organization (Carpenter, Geletkanycz, & Sanders, 2004). The dominant coalition within an organization comprises those individuals most able to influence what the organization actually does (Connolly, Conlon, & Deutsch, 1980) and are also known as the ‘major decision makers’(Price, 1972) or the ‘executive core’ (Zald, 1963). This group of senior executives should be of interest because the group and its members provide an interface between the firm and its environment and are relatively powerful. Therefore, their choices are likely to have an impact on the organization (Carpenter, Geletkanycz, & Sanders, 2004). Often, these individuals can be identified through their title or position which, especially for listed companies, can be found in the annual reports. Following this line of reasoning, the board of directors of a company is seen as the top management team within this study.

2.2.2 The Upper Echelons Perspective

17 The central tenet of the UE perspective has been corroborated in a variety of studies. For example, when investigating the extent to which executive characteristics explain the influence of the top management team (TMT) on innovation, Hoffman and Hegarty (1993) showed that different executive characteristics explain influence on different types of innovation. Other research found evidence for linking different TMT characteristics to different strategic actions of firms (Bantel & Jackson, 1989; Finkelstein & Hambrick, 1990; Wiersema & Bantel, 1992) and outlined how TMT characteristics may have an influence on firm performance (D'Aveni, 1990; Haleblian & Finkelstein, 1993; Michel & Hambrick, 1992; Norburn & Birley, 1998). Furthermore, TMT characteristics are not only relevant for the focal firm, but have also been shown to influence partner selection. For example, as Eisenhardt and Schoonhoven (1996) show, the greater the number of previous employers and senior ranks earlier held by the TMT, the greater the rate of alliance formation. All in all, when synthesizing the studies on the UE perspective, Carpenter, Geletkanycz, and Sanders (2004, p. 20) state the following: ‘Recent studies have greatly enhanced the generalizability of the UE model. For example, researchers have demonstrated that executive effects exist in not only domestic but also international settings, mature firms as well as newly founded firms, and for-profit enterprises and public agencies. They have provided critical evidence to suggest that TMT demography impacts strategy in its various forms – not only business and corporate strategy profiles, but also international agendas, alliance formation, and acquisitions.’

In conclusion, following the UE perspective, there is a relation between the increasing movement of directors across borders and the demographics of these directors with firm strategies and performance.

2.3 Internationalization of the TMT

2.3.1 Internationalization Theory and Board Members

18 The increasing pressure on expanding beyond borders for firms gave rise to an international market for executive labor in Europe (Ruigrok & Greve, 2008). The recruitment of foreign executives can reduce the impact of the liability of foreignness – a concept that stipulates that firms face additional economic and social costs when doing business abroad (Johanson & Vahlne, 2009) – due to the familiarity of the executive within the targeted foreign environment (Ruigrok & Greve, 2008). Thus, ‘the emerging global economy has led to an increase in the number of foreign board members originating from different markets’ (van Veen, Rao Sahib, & Aangeenburg, 2013, p. 2).

More factors have contributed to the increased international mobility of executives, commencing within the 1990s when large European multinationals made first steps to recruit foreign TMT members (Heijltjes, Olie, & Glunk, 2003; Ruigrok & Greve, 2008). Examples of such factors are ‘a wave of cross-border mergers, the increasing use of English as a lingua franca in large multinationals, a convergence of educational formats, and increasing international capital markets as well as markets for goods and services’ (Ruigrok & Greve, 2008, p. 2). As a result, nationality now has become a main factor in the search and selection process of executive and non-executive directors in many listed companies in European countries (Oxelheim & Randoy, 2003; Ruigrok & Greve, 2008).

2.3.2 Diversity of the TMT

However, an increased level of nationality diversity is not a good thing per se for companies. Companies have to carefully balance the up- and downsides of recruiting diverse boards : they have to balance the necessity to internationalize and increase the nationality diversity on the one hand, with the risks of becoming too diverse on the other (van Veen, Rao Sahib, & Aangeenburg, 2013). A broad range of researchers have investigated the positive sides of increased diversity of the top management team. For example, Bantel and Jackson (1989) have shown in their research on the TMTs within banks that highly educated teams with a diverse range of functional backgrounds leads to more innovative banks. In another research, Wiersema and Bantel (1992) found a relation between strategic change and different education backgrounds of members of the TMT. Furthermore, in their investigation of the TMT within 32 large US airlines, Hambrick, Cho, and Chen (1996) showed that a diverse team in terms of company tenure, education, and functional backgrounds exhibited a greater propensity to take strategic actions and respond to environmental changes. Overall, researchers have shown that cognitive diversity as a result of different experiences in terms of tenure, age, education, and other demographic components may promote constructive debates within the TMT and lead to strategic innovation (Barkema & Shvyrkov, 2007).

19 Over the years, researchers also found a range of negative effects of a diverse group. Diversity may lead to the formation of subgroups within teams (Earley & Mosakowski, 2000), foster conflict and turnover, and diminish morale, cohesion, and group performance (Barkema & Shvyrkov, 2007; Harrison & Klein, 2007). In addition, diversity within groups also reduced interpersonal liking, psychological commitment, and intergroup communication (Tsui, Egan, & O'Reilly III, 1992; Tsui & O'Reilly III, 1989). The formation of subgroups within teams is especially likely in case of so-called strong faultline settings (Lau & Murnighan, 2005). Strong faultline settings occur if the ‘subgroups differ in terms of several demographic characteristics at the same time’ (Barkema & Shvyrkov, 2007, p. 664). An example of a strong faultline setting within a group may be one in which half of the group is male, nearing their pension, and having an engineering background, whereas the other half of the group is young, female, and of Hispanic origin. In this case, these faultlines settings may hamper constructive debate within the group and negatively affect the capacity of the TMT to enter new foreign markets (Barkema & Shvyrkov, 2007; van Veen, Rao Sahib, & Aangeenburg, 2013). ‘Thus, nationality diversity can also lead to separation dynamics on a board which has a negative effect on performance indicators of companies’ (van Veen, Rao Sahib, & Aangeenburg, 2013, p. 3). However, although there may be substantial negative effects of too much diversity within a group, ‘recent research suggests that, over time, the negative effects of demographic diversity may diminish, allowing diversity’s positive effects to be approached if not achieved’ (Lau & Murnighan, 2005, p. 645).

20

3. HYPOTHESES

This chapter outlines the hypotheses related to the research question and grounds them in the literature. It sequentially analyzes the concepts of the varieties of capitalism, cultural tightness, institutional differences, country and company size, and the market for top managers.

3.1 Varieties of Capitalism

The economic behavior of firms is embedded within a wider set of institutions which affect economic outcomes on a macro level (van Veen & Kratzer, 2011). Institutions can be defined as the rules of the game in a society or, more formally, the humanly devised constraints that shape human interaction (North, 1990). Relating institutions to the concept of interlocking directorates, Stokman and Wasseur (1985) showed that different institutional systems lead to structural differences in the characteristics of the interlocking directorate networks. This finding which was supported by Heinze (2004), who concluded that the explanation of differences between the German and British situations was rooted in the different kinds of capitalism in the two countries. This shows that there are reasons to assume that interlocking directorate networks differ strongly by country as a result of institutional differences (van Veen & Kratzer, 2011). The following paragraph analyzes these institutional differences in-depth by means of the varieties of capitalism approach (Hall & Soskice, 2001).

21 profits. On the other end of the spectrum, firms in CMEs depend heavily on non-market relationships to coordinate their endeavors with other actors and to construct their core competencies, generally entailing more extensive relational or incomplete contracting, network monitoring, and a high reliance on collaborative relationships to build the competencies of the firm (Hall & Soskice, 2001). In addition, CME countries are characterized by a long-term orientation. Based on the attributes of the typical CME and LME, one might expect to find much denser interlocking directorate networks within CME-oriented countries than in LME-oriented countries (van Veen & Kratzer, 2011).

Within the broad range of factors included in the varieties of capitalism approach, multiple reasons can be found that support this premise. Within the next sections, the main focus is on the supply of capital and the governance systems currently in place. After this, an underlying assumption is presented which is used to form a hypothesis.

3.1.1 Supply of Finance

22 & Mizruchi, 1999; Heemskerk, 2011). Thus, it can be concluded that historically speaking, CMEs have dense networks of interlock directorates, although the network seems to be diminishing due to the divestment of shareholdings in large commercial banks.

When looking at the financial system within LME countries, firms in LMEs tend to rely more heavily on bond and equity markets for external finance than do those in CMEs (Aguilera & Jackson, 2003; Hall & Soskice, 2001). Although bank lending also occurs in LME countries, banks do not assign officers responsible for monitoring the lending company, since doing so exposes the bank to problems of moral hazard where officers can take advantage of fluid labor markets to move elsewhere (Hall & Soskice, 2001) and to use the knowledge they gained for the benefit of competing firms. Thus, LME economies generally lack close-knit corporate networks providing investors with inside information about the progress of companies (Hall & Soskice, 2001) and attach less value to an interlocking directorate than CME economies. All in all, it can be concluded that in terms of supply of finance, CME countries are more likely to have high levels of interlocking directorates than LME countries.

3.1.2 Governance regimes

Governance regimes within countries can vary substantially (Aguilera & Jackson, 2003; van Veen & Elbertsen, 2008). Within the varieties of capitalism, structural features can be found that may affect the accessibility of boards for foreigners, hereby indicating the extent to which interlocking directorates are possible in the first place. Per country, the rights of being represented in the board for both employees and shareholders vary greatly in their strength and scope (Locke, Kochan, & Piore, 1995), ranging from rights to information, consultation, and codetermination (Aguilera & Jackson, 2003). In addition, the ownership structure of companies differs around the world (La Porta, Lopez-de-Silanes, & Shleifer, 1999), which can also be argued to have an influence on the direction of interlocks. In order to enable a comparison between the two principal varieties of capitalism and outline the influence on the direction of interlocks, it is necessary to describe the current systems in place.

23 a large proportion of the shares within a company (Aguilera & Jackson, 2003; van Veen & Elbertsen, 2008).

On the other side of the varieties of capitalism spectrum, representing a typical CME country, Germany provides formal internal channels to give labor a voice in the firm’s decision making by providing legal rights to information, consultation, and codetermination in key decisions (Aguilera & Jackson, 2003). The German board system is a two-tier system with two separate segments, an executive board and a supervisory board (van Veen & Elbertsen, 2008), hereby formally separating the management and control functions. The recruitment of executives within German boards is based on the principle of codetermination, implying that at least one-third of the supervisory board seats are reserved for employee representatives who are elected by the domestic workforce of the company (van Veen & Elbertsen, 2008). The remaining seats of the supervisory board are reserved for shareholder representatives. Furthermore, La Porta, Lopez-de-Silanes, and Shleifer (1999) show that the shares of large German firms are not likely to be widely held. Ownership by large blockholders, such as families, states, or financial institutions, are likely to dominate and reduce the possibility for outsiders to develop powerful positions since large blockholders will demand a strong position within the board (van Veen & Elbertsen, 2008).

24 hands of a single group, the greater is the loss of control incurred with the addition of a single outside group on the firm’s board (Burt, Christman, & Kilburn, 1980). Where control of a corporation is already dispersed such that each interest group has little control over decisions, the addition of yet another voice on the board entails little loss of control for each group currently on the board (Burt, Christman, & Kilburn, 1980). In conclusion, a dominant kinship or management group in a directorate would be expected to oppose extensive interlocking with other firms (Aguilera, 1998; Burt, 1980; Dooley, 1969; Pfeffer, 1972; Zald, 1969).

The inference of the different board recruitment practices between CMEs and LMEs is that since CME board recruitment is based on the concept of codetermination of employees, the potential percentage of directors interlocking in CMEs is smaller than in LMEs. Furthermore, the dominance of blockholders in CME countries was shown to have a negative effect on the amount of interlocking directorates. Thus, both board recruitment practices and company ownership would suggest that LMEs have more interlocking directorates than CMEs. However, the arguments for this inference may be offset by the difference between the two varieties of capitalism in board size. The German boards are particularly large, with a mean of 24.6 seats, which is roughly twice the size of the boards in the UK (van Veen & Elbertsen, 2008). So even though at least one-third of the German board should consist of employee representatives, on an absolute level, there are more directors remaining who may form an interlock. Thus, relatively speaking, CME countries may have a lower percentage of directors eligible for interlocking positions, but this is offset by the larger size of their boards.

3.1.3 Underlying assumption

The literature on interlocking directorates so far is ambiguous on whether country characteristics have a pulling or a pushing influence on the direction of director movement. Thus, if a country’s institutions are characterized by a high amount of coordination between different parties, does that mean they will attract more foreign directors or that they will send out more directors? Answering this question is essential in the further analysis of the hypothesis.

25 another firm to its board in an attempt to persuade the director to influence his or her firm of principal affiliation in the interests of the receiving firm´ (Palmer, 1983, p. 46). The same logic applies on a country level, where the dominant country which scores highest on coordination is likely to appoint directors from other firms so the principal firm can be influenced. Moreover, when internationalizing, the firm will profit more from hiring foreign board executives on its board than if it would send a manager of its own to learn about the foreign business environment. As such, for the varieties of capitalism approach, the underlying assumption is that when a country demonstrates characteristics that are likely to support the formation of interlocking directorates, that country will attract more interlocks than it sends.

In summary, it can be argued that CME countries are more likely to attract foreign directors than LME countries. This is due to the high dependence on non-market relationships to coordinate firm activities, the long-term perspective taken, the supply of finance that is not entirely dependent on publicly-available data, and the large board sizes which offset the negative influence of the board recruitment practices and ownership structure in CME countries. This leads to the following hypothesis:

H1: In a two-country dyad, the more LME-oriented country will send more directors to the other country than the other way around.

Although many of the developed nations can be classified as liberal or coordinated market economies, the point here is to enable a comparison between many kinds of different economies with attention to the ways in which firms coordinate their endeavors and the institutions supporting this (Hall & Soskice, 2001). It is important to note that there may be variation between the two ideal types of the LME/CME continuum, with different institutional structures underpinning the strategic coordination (Hall & Soskice, 2001).

3.2 Cultural Tightness

An emerging stream of research has started to investigate the construct of cultural tightness/looseness, which describes the extent to which social norms constrain individuals in different societies (Crossland & Hambrick, 2011; Gelfand, Nishii, & Raver, 2007).

26 within-societal variation, depending on the strength of social norms and degree of sanctioning within societies when the norms are not adhered to (Gelfand, Nishii, & Raver, 2007; Rohner, 1984). In order to capture the void of this within-societal variation within the literature on culture, the concept of cultural tightness-looseness was introduced (Boldt, 1978; Gelfand, Nishii, & Raver, 2007; Pelto, 1968). Defined as ‘the strength of social norms and the degree of sanctioning within societies’ (Gelfand, Nishii, & Raver, 2007, p. 1226), cultural tightness reflects the extent to which norms are widely shared within a society and the extent to which transgressions will lead to repercussions (Crossland & Hambrick, 2011).

Societies characterized by tight cultures and strong norm enforcement will provide clear expectations of how executives should act in particular situations (Crossland & Hambrick, 2011). Hence, individuals in tight societies engage in more frequent monitoring of their behavior and have more intense negative self-reactions when their behavior differs from social standards, as compared to individuals in loose societies (Gelfand, Nishii, & Raver, 2007). Thus, individuals must have an extremely reliable mental compass to successfully perform in a tight society (Tetlock, 2002). This strict self-monitoring behavior is a useful tool in blending into a foreign society, which makes it easier for individuals to integrate from a tight society in a loose society than the other way around.

Furthermore, norm transgression will be recognized and stringently sanctioned in culturally tight societies, leading to greater constraints on corporate leaders (Gelfand, Nishii, & Raver, 2007). These constraints may be difficult to deal with for an executive originating from a culturally loose society, where the executive generally has higher latitudes of actions (Crossland & Hambrick, 2011).

All in all, it is possible to argue that it is easier for workers from tight societies to blend into a loose society than the other way around. The same logic can also be applied on the job market for directors, leading to the following hypothesis:

H2: In a two-country dyad, the tighter country will send more directors to the other country than the other way around.

3.3 Institutional Distance

27 is a better match. However, the results of this study do not show the influence of institutional factors on the direction of director movements.

An example of a formal institution that is likely to influence the direction of director movement is the European Union. One of the basic principles of the European Union is the freedom of movement of factors of production and, in particular, of workers (Belot & Ederveen, 2005). As such, being a member in the European Union can be argued to facilitate the transfer of directors between countries, both outwards and inwards. However, countries that entered the European Union have done so in several distinctive waves, at different points in time. As a result, some countries are more likely to and have more time to establish cross-border interlocking directorates than others have (van Veen, Rao Sahib, & Aangeenburg, 2013). Companies from countries in the earlier waves of the European Union will be more acquainted with different cultures and be more receptive and accepting of cultural differences. As such, it is likely that countries that enter the EU in an earlier stage will have more international directors in their boards, leading to the following hypothesis:

H3a: In a two-country dyad, the country entering the European Union at a later time will send more directors to the other country than the other way around.

Furthermore, regarding institutional distance, it can be argued that it is easier for directors to overcome the institutional barriers in a country with well-developed formal institutions than when directors go to a country with lesser formal institutions. Moreover, countries with well-developed institutions are likely to have higher levels of education, which in turn enhances the director selection process, making it more likely for a firm to identify and select foreign directors. This implies that there will be more international directorates heading to well-developed countries with developed institutions than the other way around. As such, the following hypothesis can be constructed:

H3b: In a two-country dyad, the country with the lower institutional quality will send more directors to the other country than the other way around.

3.4 Country Size, FDI, and the Quantity of Large Firms

It can be argued that there is a direct relationship between a country’s size and the dominant direction of director movement. The following paragraph describes the background of this link and hypothesizes a relationship.

28 growth process since they can rely on the home market for a longer time (Ruigrok & Greve, 2008), which delays their need for foreign information and thus, for hiring foreign directors on their board to overcome the liability of foreignness. In smaller markets on the other hand, firms may quickly capture the home market after which internationalization is necessary to ensure further growth (Fennema & Schijf, 1978).

Furthermore, firms with large home markets will have a larger and more specialized pool of managerial talent to draw upon, reducing the need to search for new executives in the international sphere (Ruigrok & Greve, 2008). Small countries are often limited to homegrown managerial talent, which may put pressure on them to look outside of their regular environment for capable executive talent. Thus, companies from small countries are more likely to attract foreign executives, because the required executive with a certain skill-set is not available in the home country and the need to internationalize is larger in an earlier stage within small countries. This leads to the following hypothesis:

H4a: In a two-country dyad, the larger country will send more directors to the other country than the other way around.

29 with a higher level of FDI relative to the other country will send more directors relative to the other country. As such, the following hypothesis follows:

H4b: In a two-country dyad, the country investing most in the other country will send more directors to the other country than the other way around.

The larger a firm is, the more impact its actions have on other organizations and accordingly the more the firm needs representatives who can integrate and legitimize the firm in its external environment (Burt, 1980). This idea has been tested extensively, with the main finding that firm size in terms of assets or annual sales is positively correlated with the number of different corporations represented on the firm’s board of directors (Burt, Christman, & Kilburn, 1980; Dooley, 1969; Pfeffer, 1972). Thus, it is likely that countries with a high quantity of large firms will attract more foreign directors to serve on the boards. As such, the following hypothesis follows:

H4c: In a two-country dyad, the country with the lowest quantity of large firms will send more directors to the other country than the other way around.

3.5 The Market for Top Managers

In order to understand the dimensions of the market for top managers, it is necessary to take a look at theories of international migration. The macro theory of international migration is perhaps the oldest and best-known theory, originally developed to explain labor migration in the process of economic development (Massey et al., 1993). This theory argues that international migration is caused by geographic differences in supply and demand for labor (Harris & Todaro, 1970; Lewis, 1954). This means that countries with a high amount of labor relative to the levels of capital will have low wages, whereas countries with limited supply of labor are characterized by high wages. As a result of the wage differences, workers from low wage countries will move to high wage countries. This emigration pattern lowers supply of labor in the low wage country, hereby increasing the level of wages, whereas an opposite effect occurs in the high wage country, eventually leading to an equilibrium between the two countries that reflects only the costs of international movement (Massey et al., 1993). The relevant conclusion that can be used from this theory is the fact that the international migration of workers is caused by differences in wage rates between countries.

30 be discouraged from leaving the well-known for a less know and riskier environment because of language and cultural barriers. Moreover, executives may rely highly on their current network which may be of limited value in a new environment. Furthermore, executives’ spouses and children may prefer to settle down in one place for a longer period rather than move around from one country to the other (Ruigrok & Greve, 2008). Finally, Ruigrok and Greve (2008) argue that the pension systems of many countries effectively discourages international executive mobility. All in all, to offset all these barriers, executives moving abroad and taking up an uncertain challenge will typically expect a generous compensation package (Ruigrok & Greve, 2008). Because of the high switching costs for executives to move to another country, countries that have higher wages are more likely to attract foreign executives. As such, the following can be hypothesized:

31

4. METHODOLOGY

The following section outlines the origin of the data used and will explain how it has been collected. Furthermore, the hypotheses are operationalized into measurable variables that enable an analysis. Finally, this section will illustrate the methods used and defend the use of the methods where necessary.

4.1 Data

In order to test the hypotheses, this research uses a stratified sample that covers companies from 15 European countries in the year 2007. The dataset includes data on several variables of board memberships from approximately 5500 top management positions in 347 companies. The data was used and collected by van Veen and Marsman (2008) in 2005 and was updated in a later version for 2006 and 2007 by van Veen, Rao Sahib, and Aangeenburg (2013). The dataset comprises information from the 15 countries first to access the European Union: Austria, Belgium, Denmark, Germany, Finland, France, Greece, Ireland, Italy, the Netherlands, Portugal, Spain, Sweden, and the United Kingdom. The difficulty of obtaining data from later entrants to the EU and the intensive data collection exercise accompanying it made the mentioned authors decide to stick with these sample countries. Furthermore, with top management primarily moving to other top companies, the authors chose to focus on approximately 30 of the biggest enlisted companies within a country. This lead to a sample with predominantly multinational corporations (MNCs). In general, ‘they have long histories, are financially strong, are comparatively stable and are considered to be the most important companies in these countries, when looking at their market capitalization’(van Veen & Marsman, 2008, p. 191). The usage of MNCs as sample companies enabled collecting the data through annual reports of these firms. When information was lacking, this was supplemented by using internet sources such as Google Finance (Finance, 2007), ZoomInfo (Zoominfo, 2007), and ‘Top Management’ (TopManagement, 2007).

4.2 Operationalization of Variables

32 The dependent variable is operationalized as the amount of outgoing internationalizing directors from country I to country J ( relative to the amount of outgoing internationalizing directors from country J to country I ( . This becomes:

Where is the relative balance of outgoing directors between the two countries. This is also depicted graphically in Figure 2:

Figure 2: Graphical depiction of the dependent variable

4.2.2 Independent Variables

4.2.2.1 Varieties of Capitalism

33 Due to the fact that the dependent variable is the relative amount of director movement within a dyad, the independent variables have to be altered to represent information for a dyad as well. As such, the single score of a country has to be divided by the score of the other country in the dyad. All in all, the operationalization of this variable becomes:

(1)

In which represents the relative coordination scores for the dyad and and represent the sores on the coordination index for respectively country i and country j. Both Greece and Luxembourg have not been included in the sample of Hall and Gingerich (2004).

4.2.2.2 Cultural Tightness

For the operationalization of the cultural tightness hypothesis, the country scores of Gelfand et al. (2011) are used. In this work, it is argued that tightness-looseness of a country is manifested in ecological, historical, and institutional contexts, but also in everyday situations. The overall strength of social norms and tolerance for deviation was measured on a six-item Likert scale that assessed the degree to which norms are pervasive, clearly defined, and reliably imposed within nations (Gelfand et al., 2011) within a group of 6823 respondents across 33 nations. For five of the sample countries, Gelfand et al. (2011) did not collect data on the cultural looseness/tightness scores, being Denmark, Finland, Ireland, Luxembourg, and Sweden. As such, these countries have been omitted for the further analysis of this specific hypothesis. Finally, a special note has to be made for Germany: Gelfand et al. (2011) measured the cultural looseness/tightness scores in two distinct areas within Germany, hereby splitting up the results into Germany (Former East) and Germany (Former West). For the purpose of this analysis, the average of both scores is used to function as a tightness score for Germany as a whole.

Again, the scores for this variable have to be adjusted so they have the same unit of analysis as the dependent variable. As such, the variable is operationalized as:

(2)

In this case, represents the relative scores of the cultural tightness index for the dyad with country and country . and are the cultural tightness scores for respectively country and country .

4.2.2.3 Entrance of EU

34 various countries. The required data is gathered through the official EU website (EU, 2013). With all the sample countries being member of the EU, all of them are included in the further analysis of H3a. Finally, similar to the previous independent variable, the independent variable should represent the same unit of analysis as the dependent variable. In order to do so, the year of entry was deducted from the year of the director data. For example, a country which entered in 1981 would get an EU score of 26 (2007-1981=26). With this input, the mathematical description of this variable becomes:

(3)

In this case, represents the relative scores for the dyad with country and country . and are the scores for respectively country I and country j.

4.2.2.4 Institutions

‘Only if institutions can be measured with a minimum degree of confidence are empirical statements such as ‘institutions matter for y’ credible’ (Voigt, 2009, p. 4). Thus, it is imperative that a method is chosen that covers the various institutional components of a country in a proper manner. For purpose of this study, the scores of the Global Competitiveness Report from the World Economic Forum (WEF) are used. This report offers a structured, systematic, and comprehensive approach to identifying and measuring the drivers of economic performance of more than 140 countries (WEF, 2013). The competitiveness scores for the countries are captured by including a weighted average of many different components, each measuring a different aspect of competitiveness (Schwab, 2013). These components are grouped into 12 pillars of competitiveness: (1) institutions, (2) infrastructure, (3) macroeconomic environment, (4) health and primary education, (5) higher education and training, (6) goods market efficiency, (7) labor market efficiency, (8) financial market development, (9) technological readiness, (10) market size, (11) business sophistication, and (12) innovation (Schwab, 2013). The WEF uses qualitative data from their own Executive Opinion Survey to measure these pillars (Lall, 2001), the results of which are supplemented by data from international organizations or national sources to complete the scores on the pillars (Schwab, 2013).

35 2007-2008 (Schwab, 2008) have been used for the analysis, since the data used on the movement of directors comprises the same time period.

With the adjustment for the unit of analysis in mind, this variable is operationalized as: (4)

Where represents the relative scores of the two countries within the dyad and and the scores for country and country .

4.2.2.5 Country Size

In order to test H4a – the influence of country size on interlocking directorates – two separate variables were used. Since country size can be measured in different ways with potentially different outcomes, both the total population and the levels of GDP where collected from the CIA Factbook data of 2007 (CIA-Factbook, 2007). The national GDP levels are adjusted conform Purchasing Power Parity (PPP), which excludes the influence of different currency values on the total level of a nation’s GDP.

All in all, country size will be measured through two operationalized variables: (5)

(6)

Here and represent the relative scores of the two countries within the dyad for both population and GDP, whereas and represent the data for country . Finally, and represent the population and GDP input from country .

4.2.2.6 Foreign Direct Investment

H4b hypothesizes that there is a link between de degree of Foreign Direct Investment (FDI) and the direction of interlocking directorates. In order to facilitate measurement of this hypothesis, the data available from the OECD on bilateral FDI was used (OECD, 2013).

36 Peculiarly enough, there is a difference in data between what country A invests in country B and what country B receives from country A. To clarify this, this is equal to when person A hands over 10 Euros to person B, but person B somehow only receives 8 Euros. Whether this is a result of different accounting systems or that some money is actually lost on the way is unclear and is the topic of debate within the field of economics. This difference impedes the operationalization of this variable and leaves two options. Either I average the inflow in country A with the outflow from country B, or I choose one of the two and assume that is representative of the actual FDI levels. Due to time constrictions and convenience, I choose the latter. Moreover, because FDI levels often fluctuate to a large extent, I decided to use a longer time-span of 10 years to ensure more reliable data.

In summary, I used the inflowing FDI for all country pairs from 1998 to 2007, leading to approximately 2100 entries of inflowing FDI. Unfortunately, the OECD restricts access to some FDI data because of confidentiality of the investments or only started gathering data at a later point in time. For example, the data on inflowing FDI of Belgium only started after 2002. Although this reduces the reliability of FDI data of Belgium to a certain extent, a time span of 6 years can still be perceived as representing the overall flow of investment towards Belgium and is used accordingly. Finally, FDI flows generally show positive numbers, in which investments within a country exceed disinvestments. However, this is not always the case and some dyads show a negative inflow of FDI, indicating that some of the components of the inwards FDI are negative and not offset by positive amounts of other investments (UNCTAD, 2013).

All in all, FDI is operationalized as:

(7)

Here, represents the relative amount of foreign direct investment for the dyad , and and the scores of country and respectively.

4.2.2.7 Quantity of Large Companies

Following H4c, there may be a link between the amount of large companies and the direction of interlocking directorates. In order to test this hypothesis, the companies of the Forbes Global 2000 list of publically traded companies were tested on their origin. The Forbes Global 2000 list comprises a composite ranking from four metrics: sales, profits, assets, and market value (Forbes, 2007). The total amounts of Forbes Global 2000 companies of each country were aggregated and function as basis for testing this hypothesis.

37 (8)

Within this variable, represents the relative amount of large companies for the country dyad, whereas and represents the absolute amount of large companies within countries and

4.2.2.8 The Market for Top Managers

Top management pay is an elusive concept within the literature and data is very difficult to acquire. This is mainly due to the confidentiality of top management pay levels and the fact that rewards for top management often consist of both a fixed and a variable component. As such, there is no data or construct available that completely captures the notion of top management pay levels.

The concept of Gross Domestic Product per capita can be used as a close proxy in this case. This will show how much on average person earns within a country and can be used to compare it to the GDP per capita for other countries. To ensure that the variable is not influenced by differences in the value of different currencies, the GDP per capita data are all adjusted for price levels. The country level data is collected from the CIA Factbook historical data section, so the data of the year 2007 can be used.

This final variable is operationalized the following:

(9)

Here, represents the relative GDP per capita of the dyad , whereas and represent the GDP per capita of country and country .

4.2.3 Control Variables

Although Europe is expected to have high levels of educational quality, there are still differences between the countries. These differences may have an influence on the dominant directions of director movement because there are simply relatively more eligible candidates present in a country with a high quality of the educational system compared to a lower quality of education. As such, I will control for the educational quality of a country when testing the influence of the operationalized variables on the direction of directors.

In order to include this variable, I will look at the scores of the fifth pillar of the Global Competitiveness Report (Schwab, 2008) which concerns higher education and training. ‘This pillar measures secondary and tertiary enrollment rates as well as the quality of education as assessed by the business community’ (Schwab, 2008, p. 5).’

38 (10)

39

5. RESULTS

5.1 One-Dimensional Analysis

The coming sub-chapters analyze the hypotheses and operationalized variables in a one-dimensional fashion. Every single variable is tested for its influence on the dependent variable as to check whether one variable on its own is an accurate prediction of the dominant direction of cross-border director movements. How this is done will be explained throughout the first variable which is based on the varieties of capitalism approach.

5.1.1 Varieties of Capitalism

The higher the scores on the coordination index, the more CME-oriented a country is, with a country having a 1.0 score fulfilling the ideal description of a CME country. Following H1, it is argued that in a two-country relationship, there will be fewer directors going from the country with a higher coordination index score to the other country than the other way around (see 2.3 for the assumption on direction). When this reasoning is applied on all sample countries, this leads to the following figutablere (Table 2) of hypothesized directions between countries (see Appendix 1 for a list of country abbreviations and the coordination scores):

Receiving UK IE ES NL FR SE DK FI PT BE IT DE AT Sending UK + + + + + + + + + + + + IE - + + + + + + + + + + + ES - - + + + + + + + + + + NL - - - + + + + + + + + + FR - - - - 0 + + + + + + + SE - - - - 0 + + + + + + + DK - - - + + + + + + FI - - - 0 + + + + PT - - - 0 + + + + BE - - - + + + IT - - - + + DE - - - + AT - - - Table 2: Overview of hypothesized directions between countries for H1