Cross-border M & A’s and Board Nationality Diversity

Cross-border M & A’s and Board Nationality

Diversity

Wesley Kaufmann, LLM MSc

1322710

RijksUniversiteit Groningen Faculty of Economics & Business

Master Thesis IB&M Supervisor: Dr. K. Van Veen Second Supervisor: Dr. C.L.M. Hermes

1

Abstract

2

CONTENTS

1. Introduction...4

2. Regression Analysis...8

2.1 Literature Review...8

2.1.1 Overview of TMT Diversity Studies ...8

2.1.2 Internationalization of the TMT ...10

2.2 Methodology ...12

2.2.1 Dataset and sample firms ...12

2.2.2 Variables...14

2.2.3 Methodological Approach...17

2.2.4 Descriptives...18

2.3 Results Regression Analysis ...22

2.4 Sub-conclusion ...25

3. Case Analysis...26

3.1 Literature Review...26

3.1.1 The role of cross-border M&A’s on Board nationality diversity: existing studies...26

3.1.2 The role of cross-border M&A’s on Board nationality diversity: four scenario’s...28

3.1.2.1 Scenario 1: the cross-border M&A has no discernible effect on Board nationality diversity...28

3.1.2.2 Scenario 2: the cross-border M&A at first results in a larger number of foreigners, but later on the number of foreigners decreases...31

3.1.2.3 Scenario 3: the cross-border M&A results in a lasting larger number of foreigners on the Board, due to inflow from the acquired firm...34

3.1.2.4 Scenario 4: the cross-border M&A is the starting point of further Board nationality diversity...35

3.2 Methodology ...37

3.2.1 Dataset, sample firms, variables, and methodological approach ...37

3.2.2 Descriptives...42

3.3 Results case analyses ...48

3.3.1 Scenario 1 Cases...48

3.3.1.1 Sample firms with nationality homogenous Boards ...48

3.3.1.2 Sample firms with decreased Board nationality diversity...49

3.3.1.3 Sample firms with unrelated increased Board nationality diversity ...50

3

3.3.4.1 Gradual Board nationality diversity ...64

3.3.4.2 Immediate Board nationality diversity...69

Summary ...71

3.4 Sub-Conclusion...71

4. Limitations and potential for future research ...74

5. Conclusion ...76

References...78

Appendix...81

4

1. Introduction

Recently, there has been a growing interest in the composition of Top Management Teams (TMT) of large multinational firms. Most of the studies dealing with the top management team have revolved around the upper echelon perspective which has been put forward by Hambrick and Mason (1984). The main idea behind the upper echelon perspective is that “individual attributes, such as sex, age, tenure, personality traits and values, influence the preferences and attitudes of members, as well as the resulting team dynamics.” (Heijltjes et al. 2003, p. 90). As a result, various studies have dealt specifically with TMT diversity and its effects on certain firm performance measures. Carter et al. (2003) investigate the relationship between Board diversity and firm value for Fortune 1000 firms, and find “important evidence of a positive relationship between firm value and diversity on the board of directors.” (Carter et al. 2003, p. 51). Similarly, Erhardt et al. (2003) investigate the relationship between Board diversity and firm financial performance. In this study the focus is on another form of Board diversity, namely demographic diversity. Their findings are that diverse Boards are found in conjunction with increased firm financial performance. The authors continue by stating that “regardless of whether it is the cause or result of performance, it does appear that firms should seriously consider the potential for the enhanced representation and perspective diversity might create.” (Erhardt et al. 2003, p. 109).

As the abovementioned overview has indicated, there have been numerous studies comparing Board diversity to firm performance; and all of these reviewed studies have found a positive relationship between these two variables. However, one of the questions which has been put forward more recently is: how diverse are Boards in terms of nationality diversity? According to Staples (2007b, p. 4), there are a number of reasons why Board globalization (which is equal to the term Board nationality diversity in my study) is an interesting research subject. Firstly, multinational firms are central to globalization and as a result “it is important for sociologists to study the people and organizations that control resources that often dwarf those of nation-states”(Staples, 2007b, p. 4). Heijltjes et al. (2003, p. 89) confirm this line of reasoning, as they mention that top management team issues have generally been ignored, despite the unprecedented internationalization of business. Therefore, Board nationality diversity can also be placed in the wider context of studies dealing with globalization.

The second reason which Staples (2007b, p. 4) offers as support for investigating Board globalization is sociologists’ interest in the possible emergence of a “Transnational Business Class” (Staples, 2007b, p. 4). Further research into Board globalization could benefit the debate concerning this concept.

Finally, Staples mentions that more empirical research dealing with globalisation is required, as currently “the ratio of theoretical treatises and speculative essays to empirically grounded studies is heavily weighted toward the former.” (Staples, 2007b, p. 4- 5).

5 effect on long-term M&A performance. Krug & Hegarty (1997, p. 674) investigate post-acquisition turnover among US Top Management Teams, and compare domestic and foreign acquisitions: “crossborder acquisitions are associated with higher rates of turnover over time when compared to turnover in purely domestic acquisitions” (Krug & Hegarty 1997, p. 674). Given this line of reasoning, it would appear that Board changes following a cross-border M&A are an important determinant of post-M&A success. A better understanding of these Board dynamics could therefore provide valuable insights for future research dealing with success factors of cross-border M&A’s.

Given the arguments mentioned above, it seems clear that there are a number of good reasons for studying Board nationality diversity. Heijltjes et al. (2003, p. 90) mention however that “nationality is a rarely explored source of diversity in most demographic research.” Heijltjes et al. (2003) attempt to fill this gap in the literature by investigating the internationalization of top management teams in The Netherlands and Sweden. For their study, Heijltjes et al. (2003) focus on Executive Board members. The authors find that for these two sample countries only about 25 per cent of the top companies had foreign Board members in 1999. The percentage of foreign Board members in both countries was about 10-11 percent (Heijltjes et al. 2003, p.93).

Another study dealing explicitly with the internationalization of TMTs is the study performed by Van Veen & Marsman (2007). In this study information concerning over 2000 top managers, from 363 companies, located in fifteen EU countries is retrieved, in order to determine how nationality diversified Executive Boards are. The study indicates that the percentage of foreigners on the Board is rather low (around 15%), but that there is a high variance across countries and across companies. This high variance across countries is confirmed by Van Veen & Elbertsen (2007), who focus specifically on Board nationality diversity in the UK, Germany, and The Netherlands. The authors find inter alia that “countries governance regime has an effect on the structural opportunities for foreigners to enter boards of MNCs within this country”, and that there are significant country differences (Van Veen & Elbertsen 2007, p. 14).

Although Heijltjes et al. (2003), Van Veen & Marsman (2007) and Van Veen & Elbertsen (2007) specifically investigate how nationality diversified Boards are, the process of nationality diversification itself is largely treated as a “black box”, as for instance few case analyses are provided to clarify Board nationality diversity in practice. Van Veen & Marsman (2007, p. 44) mention that cross-border mergers & acquisitions probably play a larger role than natural selection processes when it comes to Board nationality diversification, but support for this statement is based on anecdotal evidence. Heijltjes et al. (2003) also mention a large number of potential nationality diversity determinants: firm size, the size of the foreign workforce, the degree of foreign involvement in terms of foreign sales, geographic diversification or cultural diversification, the internationalization of ownership, an international orientation of the firm, the type of industry, and country of origin. However, strong empirical support for these determinants appears to be lacking.

6 Staples (2007b) further examines the link between cross-border M&A and Board globalization in a follow-up article. His results are that “a cross-border acquisition almost always results in a more multinational board of directors; that multinational boards are more likely to do cross-border deals; and that once a board becomes multinational it stays that way.” (Staples 2007b, p. 1). Staples (2007b, p. 2) also mentions that multinational boards are concentrated in Europe. Staples’ findings with regard to the role of cross-border M&A’s are interesting, as Heijltjes et al. (2003) did not mention cross-border M&A as a potential determinant of Board nationality diversity. At the same time, Staples (2007a & 2007b) does not appear to specifically have taken into account the determinants put forward by Heijltjes et al. (2003). Furthermore, the studies by Staples (2007a & 2007b) provide only limited case background, and the case analyses which are provided are rather limited in scope, as they are basically geared towards supporting one hypothesis only: cross-border M&A’s will result in more foreigners on the Board. As a result, Staples articles also largely treat Board nationality diversity as a “black box”, albeit with a specific focus on cross-border M&A as the main determinant.

Concluding, a comprehensive two-fold study investigating the link between Board nationality diversity and cross-border M&A deals might benefit our understanding of the issues mentioned earlier, as to the author’s knowledge no such study exists at the moment. In order to perform this comprehensive study, two research questions are identified, which will be analyzed in section 2 and section 3 of this thesis respectively:

First Research Question: Are firms which have been involved in cross-border M&A more Board nationality diversified, compared to firms which have not?

This analysis is meant not only to replicate Staples (2007a & 2007b) results, indicating that cross-border mergers & acquisitions are the primary determinant of Board nationality diversity, but also to improve his analysis by specifically taking into account other potential Board nationality diversity determinants. In order to answer this research question, I will make use of the studies performed by Van Veen & Marsman (2007) and Van Veen & Elbertsen (2007). These studies have resulted in a large dataset containing inter alia information on TMT’s of 363 large European firms. This information is used to create a new sample of 324 large European multinational firms1, consisting of firm-specific information (such as assets, employees, country of incorporation, industry, and number of countries active) as well as country-specific information (such as GDP, and population of the country of incorporation). Furthermore, information regarding cross-border M&A activity is retrieved from the UNCTAD World Investment Reports. These reports provide an overview of cross-border M&A deals involving a value of $1 billion or more, for the sample period of 1995-2005. The UNCTAD border M&A data is cross-referenced with the 324 sample firms, and used in order to create a cross-border M&A dummy. Finally, an Ordinary Least Squares (OLS) regression analysis is performed, with percentage of foreigners on the Board as the dependent variable, and the firm- and country-specific variables as independent variables (including the cross-border M&A dummy). This OLS regression will be analysed in order to investigate if cross-border M&A activity results in a higher percentage of foreigners on the Board, taking into account various other potential determinants of Board nationality diversity.

1

7

Second Research Question: What type of Board nationality diversity-related processes take place following a cross-border M&A?

This analysis is meant to provide a more in-depth and comprehensive study of the impact of cross-border M&A on Board nationality diversity. Whereas the analysis related to the first research question is of a general nature, this second analytical part is based on case analyses, in order to accurately identify the processes occurring in the Board following a cross-border M&A. Four potential scenario’s related to Board nationality diversity changes following a cross-border M&A are identified:

Scenario 1: the cross-border M&A has no discernible effect on Board nationality diversity

Scenario 2: the cross-border M&A at first results in a larger number of foreigners, but later on the number of foreigners decreases

Scenario 3: the cross-border M&A results in a lasting larger number of foreigners on the Board, due to inflow from the acquired firm

Scenario 4: the cross-border M&A is the starting point of further Board nationality diversity In order to ascertain which scenario is relevant in which case, case analyses are performed for 46 sample firms which have been involved in significant cross-border M&A deals during the sample period. For each of these sample firms the composition of the Board is retrieved one year before, one year after, and four years after the cross-border M&A took place. By retrieving the Board composition at three different points in time, it becomes possible to perform a longitudinal case analysis for each sample firm, which enables me to classify each sample firm according to one of the four abovementioned scenario’s. These case analyses will be used to offer an in-depth and comprehensive analysis of the impact of cross-border M&A on Board nationality diversity.

8

2. Regression Analysis

In this section the first research question of this thesis will be discussed: will firms which have been involved in cross-border M&A have a higher percentage of foreigners on the Board, compared to firms which have not? Section 2.1 provides a literature review, section 2.2 the methodology, section 2.3 the results, and section 2.4 the sub-conclusion for this analysis.

2.1 Literature Review

Section 2.1 is concerned with a literature review in which firstly an overview is provided of more general studies dealing with diversity of the top management team in section 2.2.1, followed by a description of studies dealing with internationalization of the top management team in section 2.2.2.

2.1.1 Overview of TMT Diversity Studies

Most of the studies since the mid 1980s related to the top management team have revolved around the upper echelon perspective (Heijltjes et al. 2003, p. 89), which has been put forward in a paper by Hambrick and Mason (1984). In this paper the authors argue for a new emphasis for macro-organizational research: “an emphasis on the dominant coalition of the organization, in particular its top managers” (Hambrick and Mason 1984, p.193). The main idea behind the upper echelon perspective is that “individual attributes, such as sex, age, tenure, personality traits and values, influence the preferences and attitudes of members, as well as the resulting team dynamics.” (Heijltjes et al. 2003, p. 90).

The upper echelon perspective has been a starting point for various other studies related to top management team characteristics. Chaganti & Sambharya (1987) for example study the link between the upper echelon perspective and strategic orientation of the firm.i

Additionally, Hambrick et al. (1996) focus on top management team characteristics (specifically on team heterogeneity) in order to explore the executive origins of firms’ competitive moves. The authors use a large sample of actions and reactions of 32 US airlines over eight years in order to investigate this link. Their results indicate that the net effect on airline performance of TMT’s that are diverse in terms of functional backgrounds, education, and company tenure, is positive. (Hambrick et al. 1996, p. 659).

9 team itself. The link between Board diversity and improved firm performance is taken into account in a number of other studies however.

Carter et al. (2003) for example investigate the relationship between Board diversity and firm value for Fortune 1000 firms.iii The main conclusion of the authors is that, after having controlled for size, industry and other corporate governance measures, there appears to be a statistically significant positive relationship between the presence of women or minorities on the Board and firm value. This is seen as “important evidence of a positive relationship between firm value and diversity on the board of directors” (Carter et al. 2003, p. 51).

Similarly, Erhardt et al. (2003) deal with Board diversity and firm financial performance. In this study the focus is on demographic diversity. According to the authors: “as board functioning is highly related to organisational performance….. the question becomes whether increased demographic diversity on boards affects overall company performance.” (Erhardt et al. 2003, p. 104). ivErhardt et al. (2003) conclude that diverse Boards are found in conjunction with increased firm financial performance. The authors continue by stating that firms should seriously consider the potential of diversity (Erhardt et al. 2003, p. 109).

Furthermore, Carpenter (2002) re-examines the link between TMT heterogeneity and firm performance. The results of this study are two-fold. Firstly, “the positive relationships between TMT educational, functional, and tenure heterogeneity and performance are contingent on complexity, as indicated by a firm’s international strategy.” (Carpenter 2002, p. 275). Secondly, “such relationships are clearly stronger in short-tenured top management teams” (Carpenter 2002, p. 275).

The relationship between Board diversity and firm performance was investigated in yet another context by Siciliano (1996). The author investigated the relationship between Board diversity and organizational performance for 240 non-profit YMCA organizations. v Results of the study were mixed, but generally in favour of diversity: greater occupational diversity revealed higher levels of social performance and fundraising results, gender diversity compared favourably to the organization's level of social performance but showed a negative association for level of funds raised, while the diversity in board member age groupings was linked to higher levels of donations (Siciliano 1996, p. 1313).

Finally, Elron (1997) specifically looks at the functioning of TMTs, and finds that “there is general support for the importance of cultural heterogeneity for the functioning of the TMTs and their international subsidiaries”, providing further support for the hypothesis that Board diversity is linked to better firm performance. vi

10

2.1.2 Internationalization of the TMT

The overview of studies dealing with Board diversity provided in section 2.1.1 has shown that, although there is a wide variety of diversity studies which find a positive relationship between Board diversity and firm performance, few studies have explicitly dealt with Board nationality diversity. According to one of the few studies dealing with this issue (Staples 2007a, p.311), it has become widely accepted that Boards are increasingly taking on more foreign Directors. Furthermore, this alleged increase in the number of foreign Board members is often accompanied by a normative judgement that the increase of foreign Board members is a positive development:

“indeed, one might say that having a multinational board is becoming the mark of the truly global corporation – or at least corporations with global aspirations.” (Staples 2007a, p. 311)

Similarly, Mandl (2003) states that Boards should go global if they are to maximize their effectiveness. But actual empirical support in favour of these claims is largely lacking. Despite this lack of empirical support, internationalization of the TMT is likely to be important for a number of reasons though, as mentioned earlier in the introduction of this thesis.

Firstly, multinational firms are central to globalization and as a result “it is important for sociologists to study the people and organizations that control resources that often dwarf those of nation-states” (Staples, 2007b, p. 4). The second argument in favour of investigating Board globalization is sociologists’ interest in the possible emergence of a Transnational Business Class (Staples, 2007b, p. 4). According to Staples (2007b, p. 4-5), “while researchers have yet to provide convincing evidence that such elite transnational networks have formed or are forming, some have suggested that an increase in TNC board globalization might be a harbinger of things to come.” Further research into Board globalization could benefit the debate concerning this concept.

Furthermore, Staples mentions that more empirical research dealing with globalization is required, as currently there are more “theoretical treatises and speculative essays” than empirical studies related to globalization. (Staples, 2007b, p. 4- 5).

Finally, there is another reason why Board nationality diversity studies would be of interest to us, namely the link between Board nationality diversity and post-cross-border M&A firm performance. Recall that according to Van Os (2005, p. 1), only a minor percentage of mergers can be considered successful, and this percentage is likely to be even lower for cross-border M&A’s. There is no consensus however which factors ultimately determine M&A success, although a number of studies have looked at the link between acquired firm TMT turnover and M&A success; i.e. Cannella & Hambrick (1993), and Krug & Hegarty (1997). Also recall from the discussion of these studies in the introduction of this thesis that acquired firm TMT turnover is associated with worse post-M&A performance, and that this is particularly the case for cross-border M&A’s.

11 Given the arguments mentioned above, it seems clear that there are a number of valid reasons for studying Board nationality diversity in general, and the role of cross-border M&A’s in particular. Heijltjes et al. (2003, p. 90) mention however that “nationality is a rarely explored source of diversity in most demographic research.” One of the reasons for this lack of academic interest is the underlying assumption that the number of foreign Board members was expected to be low:

“A cursory overview of some of the world’s largest manufacturing firms with foreign sales of over 50 per cent of turnover, reveals that only a few large firms have foreign nationals on their boards…..In all, it seems that the vast majority of companies are uninationally managed, despite the fact that a lot of them are characterized by a substantial degree of international involvement.” (Heijltjes et al. 2003, p. 90).

These results are based on the situation in the mid 1990s however, and as evidenced by the previously discussed articles by Heijltjes et al. (2003), Staples (2007a & 2007b), Van Veen & Marsman (2007) and Van Veen & Elbertsen (2007)2 this assumption no longer holds true today. These studies have indicated that, although Board nationality diversity is not very deep, it has become more widespread. The studies by Staples (2007a & 2007b) are particularly noteworthy in this regard, as he finds that “the most common way for a company to end up with a multinational board was as the result of a cross-border M&A.” (Staples 2007a, p. 316).

Staples’ findings with regard to the role of cross-border M&A’s are particularly noteworthy for two distinct reasons. Firstly, to my knowledge the articles by Staples are the only studies to date which provide an explanatory (as opposed to mainly descriptive) approach to Board nationality diversity. Secondly, Staples’ findings are particularly interesting, as for instance Heijltjes et al. (2003) did not mention cross-border M&A as a potential determinant of Board nationality diversity in their own study.

At the same time, Staples (2007a) does not appear to specifically have taken into account the potential determinants of Board nationality diversity put forward by Heijltjes et al. (2003).3 As the actual role of cross-border M&A’s on Board nationality diversity is therefore not unambiguously determined yet, this contradiction leads me to my first research question:

First Research Question: Are firms which have been involved in cross-border M&A more Board nationality diversified, compared to firms which have not?

In order to answer this research question, a regression analysis is performed. This analysis is meant not only to replicate Staples (2007a & 2007b) results, indicating that cross-border mergers & acquisitions are the primary determinant of Board nationality diversity, but also to improve his analysis by specifically taking into account other potential Board nationality diversity determinants. The methodology of this analysis is provided in section 2.2, while the results are shown in section 2.3. Finally, section 2.4 contains a sub-conclusion with regard to the regression analysis of section 2.

2

Specifics of these studies have been discussed in detail in the Introduction of this thesis, and will therefore not be repeated here

3

Recall from the introduction of this thesis that these potential determinants include: firm size, the size of the foreign workforce, the degree of foreign involvement in terms of foreign sales, geographic diversification or cultural

2.2 Methodology

In this section the methodology of the regression analysis is provided. Section 2.2.1 deals with the dataset and sample firms, section 2.2.2 with the variables, section 2.2.3 with the methodological approach, and finally section 2.2.4 provides some descriptive statistics of this analysis. The results of the regression analysis itself are provided in section 2.3 of this thesis.

2.2.1 Dataset and sample firms

The dataset for the regression analysis consists of 324 sample firms, an overview of which is provided in table 1, which is located in the Appendix of this thesis.

---- insert table 1 about here ----

This dataset is based on the dataset created by Van Veen & Marsman (2007) and Van Veen & Elbertsen (2007). Their dataset contains information regarding TMT composition, as well as firm- and country specific variables for 363 large multinational firms from the earliest 15 Members of the EU.4 The reason for selecting these firms is two-fold. Firstly, these companies are often blue-chip companies, which implies that they have a relatively long corporate history. Furthermore, normally they are financially sound, and can be considered to be stable by definition. Additionally, these companies are required by law to publish company data in their annual report, which means that the required data will be accessible for these sample firms. Secondly, as will be discussed in more detail in section 2.2.2, cross-border M&A data is taken from the UNCTAD World Investment Reports. These reports only deal with cross-border M&A’s involving a value of over $1 billion. It stands to reason that only the largest European firms can be expected to be able to afford cross-border M&A’s with such values.

Finally, Staples (2007b, page 2) has concluded that multinational Boards are concentrated in Europe, which means that if there is a significant relationship between cross-border M&A’s and Board nationality diversity (as stated by Staples 2007a & 2007b), this relationship is likely to be strongest for European sample firms.

For each of the 324 sample firms data was retrieved on a number of variables for the year 2005, which are discussed in detail in section 2.2.2. An overview of the used variables is provided in table 2. Recall that most of this information has been retrieved by Elbertsen & Marsman for their respective master theses (as opposed to the data used for the case analyses in section 3, which is entirely my own).

Table 2: Overview of regression analysis variables 5 Dependent variable

- Percentage of foreigners on the Board

*

4

The smaller sample size of my own study (324 instead of 363) is due to the fact that I have used Datastream and Worldscope data for certain variables, whereas Elbertsen & Marsman have used information retrieved from corporate annual reports and websites. Some of these variables were not available for all of the original 363 sample firms, hence the smaller sample size.

5

Independent variables - Assets

- Employees - Sales

- Number of countries active

*

- Population of country of origin

*

*

*

- Industry

- Board structure

- Cross-border M&A dummy

Information on most of these independent variables was obtained from the DataStream and Worldscope databases, namely: assets, employees, sales, country of origin, and industry. The information related to the macro-economic variables (population of country of origin, GDP of country of origin, and CME / LME), as well as the firm-specific variables number of countries active and board structure have been obtained through information on the Internet or company annual reports. The cross-border M&A dummy was created by making use of UNCTAD World Investment Reports, which will be discussed in more detail in section 2.2.2.

Information regarding the dependent variable (percentage of foreigners on the Board) was more difficult to retrieve. The approach which has been taken by Elbertsen & Marsman (and which I will copy in section 3) is as follows. The first step has been to retrieve the 2005 annual reports for each sample firm, and to copy the names of the Board members at that time. The focus of this regression analysis is on Executive Board members (for 2-tier firms) and Executive Directors (for 1-tier firms), as these Board members are expected to be comparable for purposes of this analysis. With this list of names in place, the next step was to ascertain which Board member could be considered as a foreign national. A foreign national is defined as a Board member with a nationality which is different from the country of incorporation of the sample firm. Determining the nationality of Board members is not a straightforward process, as mentioned by Staples (2007a, p. 314):

“The most daunting tasks facing anyone who studies board globalisation is obtaining valid and reliable data on the citizenship of corporate board members. Such information is not consistently provided by corporations nor are corporations required to provide it to most regulating agencies.”

member’s surname would be looked at, and a determination made for which country this surname was most common. That Board member would then be assigned a nationality based on the general connotation of his or her surname. Undoubtedly, some Board members have been assigned a wrong nationality during this process. However, given the restraints which one faces when attempting to assign nationalities to Board members, I believe that the sequenced approach taken here by Elbertsen & Marsman is the best one possible.

2.2.2 Variables

Having discussed the dataset and information retrieval process above, it is now time to further clarify the choice of variables for this thesis. The dependent variable is discussed in section 2.2.2.1, while section 2.2.2.2 deals with the independent variables of the regression analysis.

2.2.2.1 Dependent Variable

between foreign Board members, the most obvious approach would be to use the number of foreigners on the Board, and divide this by Board size. This provides me with the following operationalization of Board nationality diversity, which is in line with the approaches taken by Heijltjes et al. (2003) and Van Veen & Marsman (2007):

Percentage of Foreigners on the Board = (Number of Foreigners on the Board / Board Size) * 100%

2.2.2.2 Independent variables

Although studies dealing with Board nationality diversity have so far been limited, a number of potential determinants have been identified in the literature, which will be included in the regression analysis as independent variables.

Firstly, Van Veen & Marsman (2007) conclude that differences in percentage of foreigners on the Board appear to be largely the result of differences between countries. Therefore, country dummies are included to capture these differences. In fact, these country differences could be so significant, that they might “overshadow” the other variables of the regression analysis. In order to prevent this from happening, two regression analyses will be performed; one analysis without, and another analysis including the country dummies.

Similarly, Heijltjes et al. (2003, p. 94) state that differences in Board internationalization could be due to the industry of the incumbent firm:

“As market structure, consumer needs and production logistics vary, industries themselves vary in their degree and nature of internationalization.”

Consequently, industry dummies are included in the regression analysis for the sample firms as well.

Another potential determinant of Board nationality diversity is firm size. Heijltjes et al. (2003, p. 93) mention that the number of foreigners on the Board appears related to size. Three common proxies for firm size are therefore used in this regression analysis: assets, employees, and sales. In their study Heijltjes et al. (2003) furthermore state that

“Many companies have become multinational in terms of the number of national markets they serve, the number of countries where they own or control production facilities, the foreign workforce they employ, and even in ownership.”

Given this globalisation of business activity, it therefore stands to reason that the number of countries active, as a proxy for firm multinationality, will also have a significant impact on the percentage of foreigners on the Board. The number of countries active variable is based on such information being provided by the sample firms in their annual reports, or corporate websites.

of non-executives, it is likely that the differences between 1- and 2-tier Board structures are likely to result in other factors influencing Board nationality diversity. Based on the abovementioned argument by Staples (2007a, p. 320), we would expect that in general 1-tier Boards are more nationality diversified compared to 2-tier Boards. This statement is corroborated by for instance the case analysis of French firm Suez (which will be discussed at length in section 3 of this thesis), where following a number of M&A deals the firm decided to change from a nationality homogeneous 2-tier Board to a nationality diversified 1-tier Board. As a result, when creating the Board structure dummy the 1-tier firms are given a 1, and the 2-tier firms a 0.

In order to specifically exclude some of the macro-economic factors which could influence the percentage of foreigners on the Board, three additional variables are added: population of country of incorporation, GDP of the country of incorporation, and a CME / LME variable. While the first two of these variables are basically country size proxies, and therefore require little explanation, the CME / LME variable merits some further attention. According to the Varieties of Capitalism (VOC) dichotomy introduced by Hall & Soskice (2001), there is a difference between liberal and coordinated market economies. Firms in liberal market economies normally coordinate their activities primarily through (internal) hierarchies and competitive relationships, while firms in coordinated market economies are more dependant on network relationships outside the market. According to Van Veen & Marsman (2007, p. 38), this “social capital” which exists in CME countries makes it more difficult for outsiders to develop a career; hence LME firms are expected to be more open to foreigners entering the Board. Given the fact that most LME firms have a 1-tier Board, and most CME firms a 2-tier Board, this approach is consistent with the earlier mentioned Board structure dummy. However, as most countries do not fit nicely into the LME / CME dichotomy, each country is assigned a numerical value between 0 and 1, based on the degree to which a country fits the CME model (closer to 0), or the LME model (closer to 1).

The final element of the regression analysis (which is arguably the most important element) is the cross-border M&A dummy. The cross-border M&A dummy was constructed by making use of the UNCTAD World Investment Reports 1996 – 2006. These reports contain information on cross-border M&A’s entailing a value of $ 1 billion or more, during the years 1995 – 2005. As the author was unable to find an interactive database containing this information, all the information contained in these eleven Reports was copied manually into an Excel file. Ultimately, this data retrieval process resulted in 977 cross-border M&A deals over the ten-year period of interest. This dataset was then cross-referenced with the sample firm list containing the names of the 324 sample firms, as mentioned at the beginning of this section. Sample firms which had been involved in cross-border M&A activity during the sample period according to the UNCTAD data were given a numerical value of 1, while sample firms which did not participate in cross-border M&A were given a numerical value of 0.

A summary of all the variables used in the regression analysis is provided in table 3.

Table 3: Definition of variables used in regression analysis

Dependent Variable

Independent Variables: Firm-Specific

Assets: represent the total assets of the company converted to U.S. dollars using the fiscal year end exchange rate (in millions $).

Board structure dummy: created manually based on annual reports and corporate websites. Sample firm is given a 1 if it has a 1-tier Board structure, and a 0 if it has a 2-tier Board structure.

Country dummies: created manually based on Incorporation Code-Foreign (Country Incorporation Code). Sample firm is given a 1 if it belongs to the relevant country, and a 0 if this is not the case.

Cross-border M&A dummy: created manually to indicate whether a sample firm has been involved in a cross-border M&A deal involving $ 1 billion or more during the sample period; the sample firms is assigned a value of 1 if such an M&A deal has taken place, and a value of 0 if this is not the case.

D1 – D8 Industry dummies: created manually by making use of the first digit of the Primary SIC code. Hence, d1 captures all the firms that have a SIC code starting with 1, etc. Definition of Primary SIC code: The Industry Classification Code is a four-digit system of classification that identifies a company’s primary operations. Standard & Poor’s Compustat assigns these codes by analyzing the sales breakdown from a company’s 10K (annual report required by the Securities and Exchange Commission each year) and annual report. The assigned classification is reviewed each year when the company is updated by analyzing the product-line breakout in the 10K or annual report.

Employees: represent the number of both full and part time employees of the company. It excludes: (1) Seasonal employees (2) Emergency employees

Number of countries active: represents the number of countries the sample firm is active in for the year 2005. This information is based on annual reports and corporate websites.

Sales: represents the net sales or revenues of the company converted to U.S. dollars using the fiscal year end exchange rate (in millions $).

Independent Variables: Macro-level

CME – LME: created manually based on the Varieties of Capitalism framework by Hall & Soskice (2001) for each sample country, with a value closer to 0 if the sample country better fits the CME model, and a value closer to 1 if the sample country better first the LME model.

GDP: represents the GDP level of each sample country for the year 2005.

Population: represents the number of inhabitants of each sample country for the year 2005.

2.2.3 Methodological Approach

2.2.4 Descriptives

In this section the descriptives of all the variables used in the regression analysis are provided. Whereas some of the variables can take a multitude of values (such as assets, employees, GDP, etc.) other variables are limited in this regard: the dummy variables can either take the value of 0 or 1, while the CME / LME variable consists of values between 0 and 1. For these latter variables frequency tables are provided, while for the other variables more descriptives are given in the relevant table. As most of the independent variables are used as control variables in this regression analysis, not all of their descriptives are discussed in this section.

Table 4: Descriptives of dependent and some independent variables

Variable N Minimum Maximum Mean Std. Dev.

Percentage foreigners 324 0 1 0,16 0,2 Assets 324 26,3 1499302,3 86828,2 234779,3 Employees 324 34,0 460800,0 49117,3 74354,9 Sales 324 23,4 292862,1 18819,2 32429,7 Countriesactive 324 1 220 43,3 47,8 Population 324 455,0 82500,9 31516,9 27731,5 GDP 324 17401 1800822 676326,4 630469,2

Table 4 shows that the average percentage of foreigners on the Board for all the sample firms is 16%. This number appears rather low, but is in line with the studies by for instance Heijltjes et al. (2003) and Van Veen & Marsman (2007). The percentage of foreigners is not identical to that of Van Veen & Marsman (2007), as I have used a slightly smaller sub-set, as described in section 2.2.1.

Another interesting finding is the fact that the average number of countries active is 43,3 countries for the entire sample, which indicates that the sample firms can be considered to be very internationally oriented (on average). The next tables are concerned with the CME / LME variable and dummy variables.

Table 5: Frequency table for CME / LME variable

Table 6: Frequency table for Country dummies Country Frequency Austria 20 (6%) Belgium 17 (5%) Denmark 26 (8%) Germany 28 (9%) Finland 21 (6%) France 39 (12%) Greece 12 (4%) Ireland 17 (5%) Italy 22 (7%) Luxembourg 2 (<1%) Netherlands 23 (7%) Portugal 19 (6%) Spain 25 (8%) Sweden 25 (8%) United Kingdom 28 (9%) TOTAL 324

Table 7: Frequency table for Industry dummies

Industry Frequency

D1 (Agriculture, Construction, Mining) 16 (5%) D2 (Manufacturing) 63 (19%) D3 (Primary Materials) 69 (21%) D4 (Transportation, Communication) 59 (18%) D5 (Wholesale Trade, Retail Trade) 28 (9%) D6 (Finance, Insurance, Real Estate) 70 (22%)

D7 (Services) 17 (5%)

D8 (Other Services) 2 (<1%)

TOTAL 324

Table 8: Frequency table for Board Structure dummy

Value Frequency

2-tier 86 (26,5%) 1-tier 238 (73,5%)

TOTAL 324

al. (2003) claim that the Dutch and Swedish corporate governance systems are both 2-tier, I believe (based on an analysis of the Swedish sample firms) that the Swedish system should be classified as a 1-tier system. However, arguments can be made in favour of both positions, which might further explain the bias in favour of 1-tier Board firms in my sample.

A final interesting finding, as evidenced by table 9, is the fact that the frequency table for the cross-border M&A dummy shows that of the 324 sample firms, 109 firms have been involved in significant cross-border M&A deals (as acquiring firms) during the sample period, which equals about one third of the sample.

Table 9: Frequency table for cross-border M&A dummy

Value Frequency

0 215 (66,4%) 1 109 (33,6%)

TOTAL 324

At this stage it might be worthwhile to provide some further information related to these 109 sample firms which have been involved in cross-border M&A’s.

Firstly, the distribution of these 109 firms over industries and countries is shown in tables 10 and 11.

Table 10: Industry overview of sample firms engaged in cross-border M&A6

Industry Frequency

D1 (Agriculture, Construction, Mining) 2 (2%)

D2 (Manufacturing) 30 (28%)

D3 (Primary Materials) 17 (16%) D4 (Transportation, Communication) 23 (21%) D5 (Wholesale Trade, Retail Trade) 9 (8%) D6 (Finance, Insurance, Real Estate) 23 (20%)

D7 (Services) 5 (5%)

TOTAL 109

The distribution of sample firms engaged in cross-border M&A over the various industries clearly shows that, at least for this sub-sample of large European firms, most cross-border M&A activity is concentrated in the manufacturing, transportation / communication, finance, and primary materials industries; which combined account for more than 85% of the 109 cross-border M&A sample firms.

Table 11: Country overview of sample firms engaged in cross-border M&A7

Country Frequency Austria 3 (2,8%) Belgium 6 (5,5%) Germany 17 (15,6%) 6

No sample firms from industry d8 are part of this sub-sample, which is unsurprising as in the complete sample only 2 firms belonged to this industry.

7

Denmark 3 (2,8%) Spain 4 (3,7%) Finland 7 (6,4%) France 20 (18,3%) United Kingdom 19 (17,4%) Ireland 1 (<1%) Italy 4 (3,7%) Luxembourg 1 (<1%) Netherlands 17 (15,6%) Portugal 1 (<1%) Sweden 6 (5,5%) TOTAL 109

2.3 Results Regression Analysis

Section 2.2 has provided a description of the regression analysis which has been performed in order to investigate the role of cross-border M&A deals on Board nationality diversity, taking into account various other potential determinants of such diversity. In this section the results of two separate regression analyses are provided; one without country dummies, and one including country dummies. A summary of these regression analyses is provided in table 12 and table 13.

Table 12: Results of regression analysis without country dummies8 Dependent variable: percentage of foreigners on the Board

N R Square Adjusted R Square Std Error of the Estimate

323 0,17 0,13 0,22

Unstandardized Coef. Standardized Coef.

Independent Variable B Std. Error Beta T Sig.

Constant ,52 ,19 2,73 ,00 Assets -1,97E-09 ,00 -,00 -,03 ,98 Employees 2,67E-07 ,00 ,09 1,14 ,25 Sales 3,11E-07 ,00 ,04 ,58 ,56 Countriesactive ,00 ,00 ,13 1,96 ,05 Population -8,85E-07 ,00 -,11 -1,38 ,17 GDP -3,58E-08 ,00 -,01 -1,15 ,25 CME-LME -,18 ,16 -,07 -1,15 ,25 D1 -,30 ,17 -,28 -1,82 ,07 D2 -,24 ,16 -,41 -1,52 ,13 D3 -,28 ,16 -,49 -1,76 ,08 D4 -,30 ,16 -,50 -1,91 ,06 D5 -,29 ,16 -,36 -1,83 ,07 D6 -,28 ,16 -,50 -1,78 ,08 D7 -,25 ,16 -,24 -1,53 ,13 Board Structure -,03 ,03 -,06 -,92 ,36 M&A dummy ,14 ,03 ,28 4,48 ,00 8

Table 13: Results of regression analysis including country dummies9

Dependent variable: percentage of foreigners on the Board

N R Square Adjusted R Square Std Error of the Estimate

323 0,41 0,35 0,19

Unstandardized Coef. Standardized Coef.

Independent Variable B Std. Error Beta T Sig.

(Constant) ,02 ,39 ,06 ,96 Assets -1,68E-08 ,00 -,02 -,26 ,79 Employees 9,55E-08 ,00 ,03 ,46 ,65 Sales 1,64E-07 ,00 ,02 ,35 ,73 Countriesactive ,00 ,00 ,09 1,49 ,14 Population -8,50E-06 ,00 -1,01 -3,54 ,00 CME-LME ,45 ,52 ,18 ,87 ,39 D1 -,19 ,15 -,18 -1,31 ,19 D2 -,13 ,14 -,21 -,91 ,37 D3 -,16 ,14 -,28 -1,14 ,26 D4 -,17 ,14 -,27 -1,19 ,23 D5 -,18 ,14 -,21 -1,24 ,22 D6 -,17 ,14 -,29 -1,19 ,24 D7 -,15 ,14 -,14 -1,04 ,30 DAustria -,00 ,08 -,00 -,03 ,97 DBelgium ,02 ,06 ,02 ,27 ,79 DGermany ,68 ,17 ,82 3,90 ,00 DDenmark -,18 ,06 -,21 -3,02 ,00 Dspain ,18 ,09 ,21 2,00 ,05 DFinland -,12 ,06 -,13 -2,09 ,04 DFrance ,36 ,13 ,51 2,81 ,01 DUnited Kingdom ,61 ,15 ,73 4,04 ,00 DGreece -,02 ,08 -,01 -,19 ,85 DIreland -,10 ,10 -,10 -,98 ,33 DItaly ,33 ,13 ,36 2,59 ,01 DLuxembourg ,61 ,16 ,21 3,75 ,00 DNetherlands ,33 ,07 ,37 4,73 ,00 DPortugal -,08 ,06 -,09 -1,40 ,16 Board Structure ,08 ,04 ,15 1,75 ,08 M&A dummy ,06 ,03 ,12 2,15 ,03 9

The main purpose of the two regression analyses has been to answer the research question: do firms which have been involved in cross-border M&A deals have a higher percentage of foreigners on the Board than other firms which have not? In order to answer this question a cross-border M&A dummy was created. The results for both regression analyses clearly indicate that the cross-border M&A dummy has a significant impact on the percentage of foreigners on the Board, as both these dummies are significant at the 5% significance level (with a p-value of 0,000 for the M&A dummy in the regression analysis without the country dummies, and a p-value of 0,032 for the M&A dummy including country dummies). This result indicates that, as hypothesized, and in line with for instance Staples (2007a) results, cross-border M&A’s have a significant effect on Board nationality diversity. The next question then becomes; is this significant effect positive or negative? In order to answer this question I need to look at the signs of the beta coefficients for the cross-border M&A dummies. As both beta coefficients are positive (at 0,276 and 0,122 respectively), the conclusion is that cross-border M&A’s have a significantly positive effect on the percentage of foreigners on the Board.

The interpretation of the cross-border M&A dummies has already provided the most important results for this section of my thesis. However, a number of other interesting issues also arise from the two regression analyses. Firstly, according to my analyses, firm size is not directly significantly related to the percentage of foreigners on the Board, as none of the firm size proxies (assets, employees and sales) are even close to the 5% significance level in either of the regression analyses. In fact, the same holds for most of the other independent variables, as the regression analysis without country dummies only has one significant variable; the cross-border M&A dummy. Although some of the industry dummies have a p-value which is lower than 0,1, and despite the fact that the number of countries active variable is almost significant at a p-value of 0,052, it is quite surprising to find that none of the other independent variables have been found to be significant in the regression analysis without country dummies. There is however an intuitive explanation for this, which is confirmed by the regression analysis including the country dummies: one of the most important determinants (if not the most important) of Board nationality diversity is differences between countries. Of the 14 country dummies included in the second regression analysis, 9 are found to be significant at the 5% significance level. Additionally, including the country dummies in the analysis boosts the R-square from about 17% (in the analysis without country dummies) to about 41%. These findings are exactly in line with the results by Van Veen & Marsman (2007), who found that with regard to the percentage of foreigners on the Board, differences between countries are very large. Including the country dummies has not significantly altered the results obtained from the regression analysis without country dummies though; the M&A dummy is still significant, and the only variable which turned from insignificant in the analysis without county dummies, to significant in the analysis including country dummies, is the macro-level variable of population of the acquiring country.

2.4 Sub-conclusion

Section 2 of the thesis has shown that a great number of studies have looked at various aspects of Board diversity (i.e. Chaganti & Sambharya 1987, Hambrick et al. 1996, Carter et al. 2003, etc.). However, few studies have dealt explicitly with the issue of Board nationality diversity. Exceptions are Heijltjes et al. (2003), Van Veen & Marsman (2007), Van Veen & Elbertsen (2007) and Staples (2007a & 2007b). Most of these studies have treated Board nationality diversity as a “black box”, although Staples (2007a & 2007b) specifically looks at cross-border M&A’s as a determinant of Board nationality diversity. His studies fail to take into account other potential determinants however, which offers room for a more comprehensive analysis taking into account the role of cross-border M&A’s on Board nationality diversity. In order to determine if firms which have been involved in cross-border M&A are more Board nationality diversified compared to firms which have not, a regression analysis is performed, based on 324 large European sample firms. Unlike the analyses performed by Staples, this regression analysis also takes into account a number of firm specific and macro economic variables. In order to test the impact of cross-border M&A activity, a cross-border M&A dummy is created. Descriptives of the regression analysis indicate a number of interesting things. Firstly, the average percentage of foreigners on the Board for the entire sample is relatively low at 16%. Furthermore, the sample firms are (on average) truly multinational, as the average number of countries active is just over 43 countries for the entire sample. Of the 324 sample firms, 109 have engaged in significant cross-border M&A deals during the sample period. These sample firms are primarily located in the manufacturing, transportation / communication, finance, and primary materials industries. Quite surprising is the fact that most of these sample firms (two thirds) are located in just four European countries: France, Germany, UK, and The Netherlands. Given its small size, the large number of cross-border M&A deals by Dutch firms is rather unexpected, and supports the contention that firms in certain countries are more prone to engage in cross-border M&A’s than others. An interesting follow-up question would then be; are firms from certain countries also more prone to allow foreigners on their Boards? These issues will be taken up in more detail in section 3 of this thesis.

3. Case Analysis

In section 2 of this thesis a regression analysis was provided which indicated that the number of foreigners on the Board is positively related to cross-border M&A activity, taking into account a number of firm and country specific variables. However, this regression analysis has only indicated that on average firms which have been involved in cross-border M&A deals will have more foreigners on the Board, compared to other firms which have not. Board nationality diversity itself is still largely treated as a black box in this regression analysis. In order to give a more accurate description of the Board dynamics following a cross-border M&A deal, a more in-depth, case-based longitudinal approach is required. Therefore, section 3 provides case analyses of 46 sample firms, in order to obtain important firm-level information related to cross-border M&A’s and Board nationality diversity. Section 3.1 provides the literature review, section 3.2 the methodology, section 3.3 the results of the case analyses, and section 3.4 contains a sub-conclusion.

3.1 Literature Review

As mentioned in section 2 of this thesis, studies dealing with Board nationality diversity are limited. Studies dealing explicitly with the link between Board nationality and cross-border M&A’s are even rarer; the only studies (to my knowledge) which deal explicitly with this link are the ones by Staples (2007a & 2007b). These studies by Staples, although clearly relevant to my own study, provide little firm-level information regarding the Board dynamics following a cross-border M&A however. The aim of the analysis provided in section 3 of this thesis is to further add to our current understanding of the influence of cross-border M&A’s on Board nationality diversity by providing a large number of case analyses. These case analyses are based on four potential scenario’s. Section 3.1.1 will firstly provide an overview of the two articles by Staples, as well as the existing research gap, while section 3.1.2 is used to introduce the four potential scenario’s of cross-border M&A influence on Board nationality diversity.

3.1.1 The role of cross-border M&A’s on Board nationality diversity:

existing studies

As mentioned before, the only studies to date dealing explicitly with the influence of cross-border M&A on Board nationality diversity are the ones by Staples (2007a & 2007b), which will now both be discussed. Staples (2007a, p. 316) mentions that cross-border M&A’s are the most important determinant of Board globalization, as he “would estimate that at least 50 per cent of the 116 non-national board appointments were either directly or indirectly made in connection with a cross-border M&A.” Staples (2007a) does not mention specifically what he defines as “directly” or “indirectly” connected appointments. Staples (2007a) provides one case analysis in support of his findings: steel giant Arcelor. According to Staples (2007a, p. 316), French steel firm Usinor Sacilor had no foreign Directors on the Board. But:

According to Staples (2007a), the Arcelor case is “paradigmatic”, implying that the characteristics of this case can be seen as representative for larger samples. No evidence in favour of this statement is provided however.

The follow-up article (Staples 2007b), which deals specifically with the role of cross-border M&A, provides an analysis of the BHP Billiton case. The author’s main conclusion with regard to BHP Billiton is that in 2005 the firm has a globalized Board (with 58,3% foreigners), although all Directors have an Anglo-Saxon nationality (Staples 2007b, p.17). Nonetheless, the Board nationality diversity is largely the result of cross-border M&A activity (mainly due to the BHP (Australia) and Billiton (UK) merger), according to Staples. Unlike the Arcelor case mentioned earlier, the BHP Billiton case is not presented as being paradigmatic:

“We do not expect that each and every cross-border acquisition will result in a more globalized board, though we do expect this to happen more often than not. And even when it does happen, we do not expect it to happen in precisely the same way. And so, the BHP-Billiton case represents just one of the paths through which cross-border acquisitions can lead to more globalized boards; we expect the process to vary based on the histories of the companies involved, as well as a host of other factors probably too numerous to specify.” (Staples 2007b, p. 19)

After having provided the BHP Billiton case analysis, Staples (2007b) continues his analysis by providing descriptives of a sample of 43 multinational companies. Consequently, there is no further specific mentioning of other paths through which cross-border M&A can lead to more globalized Boards.

3.1.2 The role of cross-border M&A’s on Board nationality diversity: four

scenario’s

In this section four scenario’s are introduced which should capture a wide range of Board dynamics which could occur as a result of cross-border M&A deals. The four scenario’s are as follows:

Scenario 1: the cross-border M&A has no discernible effect on Board nationality diversity Scenario 2: the cross-border M&A at first results in a larger number of foreigners, but later on the number of foreigners decreases

Scenario 3: the cross-border M&A results in a lasting larger number of foreigners on the Board, due to inflow from the acquired firm

Scenario 4: the cross-border M&A is the starting point of further Board nationality diversity The reasoning behind these scenario’s will now be discussed in more detail.

3.1.2.1 Scenario 1: the cross-border M&A has no discernible effect on Board

nationality diversity

According to Staples (2007a, p. 311), some authors claim that a multinational board is becoming the mark of a truly multinational company. Similarly, Mandl (2003) states that Boards should go global if they are to maximize their effectiveness. But are most Boards truly nationality diversified in practice? The few studies which have dealt explicitly with this question have reached the conclusion that this is not the case. Heijltjes et al. (2003) find that the percentage of foreign Board members in both Sweden and The Netherlands was about 10-11 percent. Van Veen & Marsman (2007) come up with a number of about 15% regarding Board nationality diversity in a large sample of European multinationals. Finally, Staples (2007a) calls Board globalization a trend that has become more widespread, but is not yet very deep. Hence, although the hypothesized scenario’s of the next sub-sections will provide a multitude of reasons why we expect cross-border M&A’s to have a significant effect on Board nationality diversity, the low number of foreigners on Boards is in itself an indicator that there will not always be such an effect. Staples (2007b, p. 19) mentions in this regard that “we do not expect that each and every cross-border acquisition will result in a more globalized board.” The main question then becomes: what arguments can be put forward in favour of the hypothesis that cross-border M&A deals will have no discernible effect on Board nationality diversity?

Firstly, the relative size of the acquired firm compared to the acquiring firm is likely to be important. Recall from section 2 that of the 324 sample firms there were 109 firms which had been involved in significant cross-border M&A deals during the sample period. Many of these 109 sample firms had been involved in more than one M&A deal during the sample period; in total these 109 firms were responsible for 246 cross-border M&A deals of $1 billion or more. As these M&A deals all involved a value of at least $1 billion, there is already a strong bias towards larger border M&A’s in this sample. Nonetheless, when the value of the cross-border M&A is expressed as a percentage of the market value of the acquiring firm10 at the

10

time of the deal, the vast majority of the cross-border M&A’s (75%) accounts for less than 31% of the acquiring firm’s market value.11 In other words; in most cases the acquiring firm is much larger than the acquired firm. This is shown graphically in table 14.

Table 14: Classification of cross-border M&A’s as percentage of firm market value

M&A as % of acquiring firm market value

Number of M&A’s Cumulative Percentage

0-10 112 50 11-20 38 67 21-30 19 75 31-40 25 87 41-50 11 92 51-60 4 93 61-70 2 94 71-80 2 95 81-90 2 96 91-100 2 97 100+ 7 100 TOTAL 224 100%

We would expect that cross-border M&A’s which are relatively insignificant for an acquiring firm will not have an impact on Board nationality diversity. As table 14 has shown, many of the UNCTAD cross-border M&A deals can be classified as insignificant, since they account for 10% or less of the acquiring firm’s market value. In order to exclude these insignificant cross-border M&A deals, the longitudinal analysis which will be performed later on will only include cross-border M&A’s which accounted for at least 10% of the acquiring firm’s market value at the time (this approach is discussed in detail in section 3.2). Nonetheless, it can still be expected that if the acquired firm is much smaller than the acquiring firm, the influence of the cross-border M&A on Board nationality diversity will be negligible.

Secondly, it is quite clear, both from the studies by Van Veen & Marsman (2007), Van Veen & Elbertsen (2007), and those by Staples (2007a & 2007b), that differences in Board nationality diversity for a great deal depend on country differences. Hence, firms from certain countries might be more inclined to allow foreign members on their Boards, than firms from other countries. In this light, Van Veen & Marsman (2007, p. 38) mention the distinction between CME and LME countries, as introduced by Hall & Soskice (2001). According to the authors, LME countries are expected to be more open towards allowing foreign members on their Boards than CME countries.

But apart from varieties of capitalism, there are also political issues which might prevent foreigners from being allowed to sit on the Board. Staples (2007b, p. 29) for instance mentions: “who believes that the U.S. Government would be willing or able to stand by and allow the boards of ExxonMobil, General Electric, or IBM to be controlled by Chinese, Pakistani, Russian, or Libyan citizens?” Hence, even if a cross-border M&A is of a significant size, there could still be political and socio-economic reasons for not allowing foreign members to join the Board.

11

Furthermore, whereas the previous two arguments have assumed that there will be no Board nationality diversification, there is also the possibility that Board nationality does occur, but that it appears to be unrelated to the cross-border M&A under investigation. There are a number of arguments which could be put forward in favour of this statement. Firstly, it could be the case that a certain cross-border M&A is “overshadowed” by another, more important, cross-border M&A which occurred right before the M&A of interest, which lies outside my sample period. In that case we would not expect to see any Board nationality diversity changes related to that particular M&A. Secondly, it could be the case that the Board nationality diversity is the result of a strategic change, which is not related to the cross-border M&A itself (the two events would be a mere coincidence in such a case). In his Board Globalisation working paper, Staples (2006, p. 17) mentions that: “cross-border M&As accounted for the majority of non-national appointments, but there were a number of cases where corporations made isolated, non-national appointments that were not directly related to a M&A—at the very least, there was no evidence that the incoming director was associated in any way with a recently acquired or merged firm.”