N ATIONALITY DIVERSITY OF EXECUTIVE BOARDS

Master’s Thesis

N ATIONALITY DIVERSITY OF EXECUTIVE BOARDS

Evidence from Norway and Denmark

Submitted by: Tomáš Janata Student No.: s1823817 Supervisor: Mr. H. Stek Co-referent: Mr. J. Hotho

International Business and Management Faculty of Economics and Business

University of Groningen

June 3, 2009

A

This research investigates the effect of company specific variables, namely: size of company, size of executive board, cross-border mergers and acquisition, and foreign ownership on nationality diversity of executive boards in Norway and Denmark. Based on a database of 30 companies and 232 executive board members from Norway and 30 companies and 141 executive board members from Denmark, I find that the considered company specific variables have different explanatory strength of the dependent variable in each country.

Furthermore, I conclude that in Norway the company specific variables affecting nationality diversity of executive boards are only size of company and foreign ownership while in Denmark the situation is opposite, meaning that only size of executive board and cross-border merger and acquisition activity influence the nationality diversity of executive boards. Lastly, I compare the internationalization level of Norwegian and Danish executive boards with executive boards coming from other European countries. I find that Norwegian executive boards are some of the most internationalized boards in Europe. Contrariwise, Danish executive boards belong to one of the least internationalized boards in Europe.

Key words: Executive board, nationality diversity, Norway, Denmark, company variables, Blau’s index of diversity.

A

First of all, I would like to thank my supervisor Drs. Huib Stek for the time he invested into the discussions of this thesis and for his very useful commentary. I also want to thank my referent Mr. J. Hotho for his suggestions on this thesis.

Furthermore, I would like to thank all lecturers from IB&M master’s program and all my fellow students from University of Groningen for their constructive remarks on my papers during the entire program and their help in my integration into the Dutch education system.

I would also like to thank all the fellow exchange students at the Umea University, Sweden in spring semester 2007 because they were the trigger in my internationalization. Without them I would never contemplated applying for a master’s degree program in a foreign country.

Last but not least, I would to express my gratitude to my parents Jaromír Janata and Miroslava Janatová for their moral and financial support during all the years of my studies. I also want to thank my sister Lenka Janatová for her valuable advice in my decisions regarding student and career track.

This thesis is dedicated to my family.

Groningen, June 2009

T

C

1. Introduction ... 6

1.1. Globalization ... 6

1.2. Top Management Teams ... 6

1.3. Problem Indication ... 7

1.4. Research Question... 9

1.5. Structure of the Thesis... 10

2. Theoretical Background ... 11

2.1. Literature Review ... 11

2.2. Conceptual Model ... 14

3. Hypotheses Development... 15

3.1. Size of Multinational Company ... 15

3.2. Size of Executive Board ... 17

3.3. Cross-border M&A Activity ... 18

3.4. Foreign Ownership ... 19

4. Data and Methodology... 21

4.1. Analysis ... 21

4.2. Data Sample ... 21

4.3. Data Sources... 22

4.4. Dependent Variable... 23

4.5. Independent Variables... 23

5. Research Results... 25

5.1. Descriptive Results... 25

5.2. Results of Multiple Regression ... 28

5.2.1. Norway ... 28

5.2.2. Denmark ... 29

6. Conclusion and Discussion ... 31

6.1. Internationalization of executive boards ... 31

6.2. Company specific variables ... 33

6.3. Limitations and future research... 37

References ... 39

Appendix ... 44

1. I

1.1. Globalization

Although there is a general agreement that globalization involves cross-border flows of goods, services, money, people, information and culture, the researchers in the social sciences do not reach a consensus on what effect globalization has (Held et al 1999).

On one hand, thanks to advanced technology globalization allows multinational companies to benefit from economies of scale in production, distribution, marketing and management (Levitt 1983). Thus a multinational company produces higher performance in comparison to domestic companies (Hitt et al., 1994; Kim et al. 1993).

On the other hand, Gilpin (2000) and Mittelman (2000) show concerns about increasing international economic and financial flows. Their concerns might be related to the current global financial crisis. Besides, increased internationalization breeds organizational problems as the size and complexity of a multinational company increases (Fatemi, 1984; Geringer et al., 1989).

Despite the fact that many studies have indicated the importance of top management teams (TMTs) in the internalization process, TMTs as an important indicator of internationalization of companies has often been neglected.

1.2. Top Management Teams

The foundation of literature written on top management teams is the ‘upper echelons theory’

developed by Hambrick and Mason (1984). The theory argues that manager characteristics such as sex, age, tenure, values, and cognitive base have a direct impact on the decision making process of TMTs, which subsequently affects corporate performance. According to Carter et al (2003) diversity of TMTs is beneficial for a better understanding of marketplace, creativity and innovation, effective problem-solving, effectiveness of corporate leadership, and effective global relationships. This argument gained scholars attention and many empirical studies were carried out to analyze the relationship between demographic characteristics of TMTs and company performance (Ferrier and Lyon, 2004; Godthelp and Glunk, 2003; Carson et al., 2004; Simeon, 2001; Herrmann and Datta, 2005; Carpenter, 2002;

Ferrier, 2001; Carter et al., 2003).

In spite of the general conviction regarding the positive impact of board diversity on company performance, there are some studies which found contrasting evidence (Ancona and Caldwell, 1992). Likewise, Hambrick et al (1998) point out that integration, trust, and ease of communication in heterogeneous groups may cause serious problems which in the end exceed the advantages of group diversity.

1.3. Problem Indication

So far there have been done studies on the diversity of TMTs related mainly to demographic characteristics such as age, tenure, and gender. The existing studies incorporating those demographic characteristics mainly focus on USA, UK, and some of the western European countries. Although nationality is another demographic characteristic relevant to this study (Milliken and Martins, 1996), while reviewing the academic literature up to now, I found that it has been rather omitted so far.

The nationality diversity of executive boards is considered to have a positive impact on performance of a company (Nielsen and Nielsen, 2008). It is also according to Van Veen and Marsman (2008) important for international management theories in two other aspects. First, the nationality diversity is believed to be a reliable gauge for the level of internationalization of a company. Second, the board’s nationality diversity might indicate the level of so-called

‘transnational mindset.’

Current literature indicates that there is still a lack of empirical evidence on the nationality diversity of executive boards explained on a firm level. Ruigrok and Greve (2008) have investigated the relationship between nationality diversity of executive boards and a number of company variables in companies headquartered in Nordic countries, Switzerland, UK and the Netherlands. However, in the cases of Norway and Denmark, their database was rather small and unbalanced¹ to draw reliable conclusions. Similarly, Van Veen and Marsman (2008) investigated executive boards coming from the first 15 European Union member states. Nevertheless, they operationalized the nationality diversity in a rather straightforward way. They calculated the nationality diversity as a proportion of foreigners in boards, which means that they did not control for the breadth of diversity on international executive boards.

The consequence of using proportion of foreigners in boards as a measurement of board nationality diversity is that the final score is not a true reflection of board nationality diversity.

1 For Norway 7 companies and for Denmark 19 companies

According to ‘upper echelons theory’ values and cognitive base have impact on corporate performance (Hambrick and Mason, 1984). As values and cognitive base of upper echelons differ based on upper echelons’ nationality affiliation (Hambrick et al., 1998), I argue that the measurement of nationality diversity used in TMT studies should take into account all the nationalities present in the executive board and not only 2 sets: locals, and foreigners.

My main objective is to analyze the relationship between nationality diversity of TMTs and company specific variables. When selecting the company specific variables for my research I follow previous studies Van Veen and Marsman (2008), and Van Veen and Elbertsen (2008).

They recommended further tests of 3 company specific variables: number of countries a company is active in, size of company, and cross-border M&A activity. I do not incorporate the first variable (number of countries a company is active in) since I find it unfeasible due to the characteristics² of the companies I focus on. Furthermore, based on the suggestion of Van Veen and Marsman (2008) to further research the relation between ownership and the nationality diversity of executive boards, I add a third company specific variable (foreign ownership) to my research. The last relation I am interested in is a relation between executive board size and its nationality diversity. Based on Hambrick’s et al. (1998) argument regarding multinational group effectiveness, I assume that larger executive boards are more likely to be nationally diverse than smaller executive boards.

My secondary objective is to explore the differences in the relationship between nationality diversity of TMTs and considered company specific variables in a comparative study of two Nordic countries: Norway and Denmark. The rationale behind contrasting Norway and Demark is their different industry structure: MNCs in Denmark are more engaged in services while the Norwegian MNCs in manufacturing (Ericson et al., 2001). Following numbers demonstrate the latest industry composition in each country: Norwegian GDP is predominantly (31 %) generated by the energy industry while in Denmark the largest sector (45.6 %) is the private sector services, including trade, transport and financial institutions (http://www.regjeringen.no; http://www.um.dk). Those two countries should provide contrasting evidence in TMTs nationality diversity as Venzin et al. (2008) argue that internationalization of services sector has been lagging behind the nationality diversity of technology and manufacturing industry.

2 I focus on top 30 Norwegian and top 30 Danish companies. Those companies are relatively small in comparison to usually tested largest European companies meaning that they do not provide sufficient information on their web pages or in their annual reports. For instance, only 50 % of Danish companies under my scrutiny explicitly state in how many countries they are active in.

The governance structure is in both countries two-tier, meaning that the companies have executive and supervisory boards (Christensen and Westenholz, 1999; Ruigrok and Greve, 2008). While the executive managers run the daily management tasks of the company, the supervisory board is responsible for controlling the executive board (Glunk et al., 2001).

Based on the fact that the executive board represents the top management team (Heijltjes et al., 2003), I follow the earlier study by Van Veen and Marsman (2008) and restrict my focus only on the executive boards.

Nielsen and Nielsen (2008) argue that nationality diversity positively affects decision-makers’

ability to interpret the complexity of company’s international environment, thus, it is important to take into account the spread of executive board members across different nationality categories. The measurement that considers possible nationality categories and proportion of executive board members in each nationality category is Blau’s index of diversity (Harrison and Klein, 2007). Therefore, in order to be able to accurately determine nationality diversity of a particular executive board, I employ Blau’s index of diversity. This operationalization is in line with some of the existing literature, such as Ruigork and Greve (2008), or Carpenter (2002). Moreover, Van Veen and Marsman (2008) recommend Blau’s index of diversity as a way to improve the operationalization of nationality diversity.

In conclusion, the contribution of this empirical research is on several levels. First, this research aims to explain the nationality diversity with a complex analysis of company specific variables. Furthermore, this study enlarges the extant literature on top management teams by comparing empirical evidence from two countries which were so far rather omitted.

1.4. Research Question

How is the internationalization of Norwegian and Danish executive boards related to following company specific characteristics: size of company, size of executive board, cross- border M&As activity, and foreign ownership?

In order to be able to answer my research question, I use deductive approach (Gill and Johnson, 2002).

1.5. Structure of the Thesis

The remainder of this thesis is structured as follows. Firstly, a review of extant literature on the subject of top management teams and the nationality diversity of top management teams is presented. Secondly, I introduce employed research methods together with data issues.

Thirdly, I present my results. Lastly, I propose a discussion and provide my conclusions with regard to my analysis, while also mentioning the limitations and possible directions for future research.

2. T

B

2.1. Literature Review

In 1984 Hambrick and Mason presented the ‘upper echelons theory’ which suggests that a strategic choice of upper echelons is dependent upon their characteristics: cognitive base, values, age, functional tracks, other career experiences, education, socioeconomic roots, financial position, and group characteristics. The final strategic choice is then directly influencing the performance of a company. Furthermore, they developed a set of propositions related to the above mentioned upper echelons’ characteristics. This set of propositions was subsequently employed by many scholars, and the relationship between group heterogeneity and firm performance was tested. The empirical studies acknowledged the positive direction of the relationship between group heterogeneity and firm performance (Ferrier and Lyon, 2004; Godthelp and Glunk, 2003; Carson et al., 2004; Simeon, 2001; Herrmann and Datta, 2005; Carpenter, 2002; Ferrier, 2001; Carter et al., 2003).

The group heterogeneity can be measured in terms of observable (easier to track) and unobservable (more difficult to detect) characteristics (Milliken and Martins, 1996).

‘Observable diversity includes race/ethnic background, nationality, gender and age’ (Kang et al., 2007: p.195). On the other hand, the underlying attributes include characteristics such as technical abilities, socioeconomic background, personality characteristics, and values (Cummings et al., 1993; Jackson et al., 1995; Tsui et al., 1992).

While board diversity has been an increasing area of research in recent years, nationality diversity as one of the demographic variables has been rather omitted. Nationality diversity of TMTs belongs to the group of characteristics which are easier to track. Surprisingly, there are only few studies which investigate what the internationalization level of TMTs in different countries or different industries is. To my knowledge, the existing empirical studies on nationality diversity of executive boards focus mainly on Western Europe or a certain selection of the world’s largest companies.

Staples (2007) carried out a study on globalization of boards of 80 transnational companies from the UNCTAD (2005) list. He found that 75 per cent of the investigated companies had at least one non-national board member in 2005. However, he also states that only 10 per cent of companies were composed of more than 50 per cent of non-national board members. Ruigrok

and Greve (2008) conducted an analysis of Swiss, British, Dutch, Danish, Finnish, Norwegian and Swedish top management teams. They concluded that in 2005 the most internationalized TMTs were Swiss TMTs (48.8 per cent) followed by Dutch TMTs (34.53 per cent). In contrast, Danish (2.34 per cent) and Norwegian (8.00 per cent) turned out to be the least nationally diverse TMTs. The question is to what extent are their results from Norwegian and Danish companies reliable. The number of companies in their dataset from Norway (7) and Denmark (19) was rather small to draw robust conclusions.

Although there is empirical evidence on how nationally diverse executive boards of multinational companies are in some countries, there is still a lack of studies into how the differences in nationality diversity of executive boards can be explained (Van Veen and Marsman, 2008). While looking for an inspiration to derive potential explanatory factors, the existing body of literature can be divided into two groups (Van Veen and Marsman, 2008).

First, country differences are being investigated to explain the different levels of accessibility of executive teams for foreigners. One of the main country differences which play a significant role in the field of TMTs composition are differences in governance regimes such as structure of managerial labor market, ownership structure and corporate governance systems (Maclean et al. 2005). To my knowledge there are two empirical studies which tested the relationship between nationality diversity of executive boards and country specific variables (Van Veen and Marsman, 2008; Ruigrok and Greve, 2008). The former study based on the sample of 363 multinational companies coming from first 15 European Union member states found that executive boards in Liberal Market Economies have higher nationality diversity than boards from Coordinated Market Economies. Furthermore, they argue that the length of European Union membership has a positive impact on nationality diversity of executive boards. The latter study shows that size of linguistically related neighboring markets is positively associated with nationality diversity in executive boards. On the other hand, they failed to find a support to accept another proposition suggesting that size of domestic market is negatively associated with nationality diversity of executive boards.

Second, the characteristics of the multinational company itself may have an impact on the composition of executive boards. According to Van Veen and Marsman (2008) strength of the internal labor market for managers within the company, relative level of international activity, size of the company, and cross-border M&As are the company differences which might be the most relevant for research in the field of nationality diversity of executive boards. Staples

(2008) seems to be the first one who investigates the effect of cross-border M&A activity on globalization of boards. Based on his research of 148 largest transnational companies and commercial banks, he argues that cross-border acquisitions create more multinational executive boards. Similarly, Van Veen and Marsman (2008) found a significant relationship between cross-border M&As and board nationality diversity. They also show that size of the multinational company affects nationality diversity of boards, no matter how the size of company is operationalized, as either a total number of employees or total sales. However, they conclude that size of MNC expressed in terms of total number of employees is a better predictor. Other company specific variables which were studied are foreign-based ownership and industry affiliation (Ruigrok and Greve, 2008). They concluded that higher foreign-based ownership leads to higher proportion of non-national members in boards. As far as the industry affiliation is concerned, they demonstrated the existence of a positive relationship between board internationalization and the following industry categories: automotive, chemicals, consumer durables & household products, food & beverages, health care, materials

& fixtures, metal & mining, and oil & gas.

While reading the literature review one might speculate about what the direction of the relationship between company characteristics and nationality diversity of boards is. There are certainly two possible directions which might be taken into account. In this section I show an example with cross-border M&A activity, however, the causal relationship of each firm variable is discussed in section 3 (Hypotheses development).

Does higher cross-border M&A activity cause higher nationality diversity of executive board?

Or does higher nationality diversity of executive board cause higher cross-border M&A activity?

To answer this question I refer to the only three studies discussing the issue of direction. First, Staples (2008) supports the previous speculation of Robinson (2004) that cross-border M&As are very likely to result in more multinational boards. Second, Van Veen and Elbertsen (2008) argue that after a cross-border M&A a new executive board is formed as a combination of the two former boards. Since the two merging companies belong to different countries, the newly formed executive board should be more internationalized. Additionally, they argue that when a company expands its activity in foreign countries, it results in an increase of its internal labor market. Subsequently, the internal pool of potential managers becomes more nationally diverse.

Based on the above mentioned arguments I intend to test the following direction:

internationalization of executive board is dependent upon cross-border M&A activity.

2.2. Conceptual Model

Hence, the overall relationship suggested by the existing academic literature on the subject of nationality diversity of top management teams can be depicted as follows:

Figure 1 Conceptual Model

3. H

D

From the conceptual model (see Figure 1) I choose four company specific variables: size of multinational company, size of executive board, cross-border M&A activity, and foreign ownership. Previous empirical research (Van Veen and Marsman 2008; Van Veen and Elbertsen, 2008; Ruigrok and Greve 2008; Staples, 2007) has concluded that size of multinational company, cross-border M&A activity, and foreign ownership play a significant role in explanation of nationality diversity of executive boards. Moreover, Van Veen and Marsman (2008) and Van Veen and Elbertsen (2008) recommended those variables for further research. I also incorporate size of executive board as another independent variable. I assume that size of executive board is a relevant variable for this study as especially Danish TMTs are characterized as small TMTs with average size of 3 managers with a range from 1 to 8 (Glunk et al., 2001). This may subsequently influence the possibility for foreigners to be assigned to an executive position.

Furthermore, when developing my research model I have to take into account whether a potential variable is accessible to me or not. As the average response rate of direct contact with a company is in general rather low, I have to rely mainly on secondary data such as annual reports, corporate websites and databases.

Based on the above, I construct my research model. I investigate to what extent the nationality diversity of executive boards can be explained by considered company specific variables.

Moreover, those variables should have a high explanatory power and be accessible from secondary data.

3.1. Size of Multinational Company

The growth of companies is limited to the size of their domestic market. If companies have intentions to overgrow their competitors, they are forced to expand to foreign markets. Then, such multinational companies have complex organizational structures with a great number of subsidiaries abroad. The expansion of activities abroad breeds an increase in the number of foreign employees. Therefore, according to Van Veen and Elbertsen (2008) the odds for a foreigner to become a board member increase since the internal labor market stretches beyond the country’s borders. In addition, Van Veen and Marsman (2008) argue that large

multinational companies which are often listed on stock markets attract public awareness.

Subsequently, the pool of international candidates and potential management talents grows.

To sum it up, the increase in size of company is believed to result in the greater number of international candidates, both external and internal, available for management positions.

Additionally, the executive board is more likely to increase its nationality diversity.

As far as the causality between the size of company and nationality diversity of executive board is concerned, the above mentioned arguments indicate the direction of the relationship:

a company grows in size and then the amount of non-national executive board members increases.

Despite the fact that Van Veen and Elbertsen (2008) found a negative relationship between the size of company and the level of internationalization of its executive board, the general conviction is that the relationship is positive. Van Veen and Elbertsen (2008) supposed that their contradicting results were caused by the characteristics of their database³. Therefore, they decided to test the relationship on Staples’s (2008) database. However, they found again a negative relationship between nationality diversity and size expressed in total number of employees. In contrast, Van Veen and Marsman (2008) tested on a sample of 363 companies headquartered in the first 15 European Union member countries suggested relationship between company size and executive boards’ internationalization. They concluded that the relationship is positive no matter how the size of a company is expressed, whether in total sales or in total employees.

As the empirical research which found the negative relationship between board nationality diversity and company size were focused mainly on supervisory boards, I assume that I will not encounter the same opposite evidence. I assume that my evidence will be in line with Van Veen and Marsman (2008) since their database consists of only executive boards as mine does.

As Van Veen and Marsman (2008) concluded that total employees is a better predictor than total sales, I express size of a company in total employees.

3 Their database composes of 25 Dutch MNCs, 29 British MNCs and 30 German MNCs, where the Dutch MNCs are the most internationalized despite their relatively small size in comparison with British and German MNCs.

This leads to the following hypothesis:

H1: The larger the company in terms of total employees, the more nationally diverse its executive board is.

3.2. Size of Executive Board

In the extant academic literature there is no consensus on whether nationally diverse groups are more effective than single-nationality groups. Hambrick et al. (1998) summarized studies which deeply focused on the effectiveness of international groups. He concludes that a group consisting of several nationalities does not have a significant advantage over single-nationality groups. Furthermore, he argues that in some cases, depending on task and amount of nationality diversity, international groups can be ineffective.

Thus, having in mind Hambrick’s arguments, I assume that nationality composition of executive boards is dependent on the size of executive board. I argue that appointing a non- national manager to a smaller executive board is riskier than appointing him or her into a larger executive board. In larger executive boards the negative aspects of nationality diversity might be moderated by the larger number of executive board members. Therefore, the larger groups give non-national managers more opportunity to enter an executive board.

Furthermore, Jaw (2009) suggested that there is a relationship between board size and firm’s internationalization. Since Dörrenbächer (2000) indicates that a firm’s internationalization could be measured in terms of proportion of non-nationals in boards, I assume that there is a direct positive link between executive board’s size and nationality diversity of executive boards.

Therefore, the causality between the executive board size and nationality diversity of executive board is as following: nationality diversity of executive board is dependent upon executive board size.

This leads to the following hypothesis:

H2: The larger the executive board, the more nationally diverse the executive board is.

3.3. Cross-border M&A Activity

Shimizu et al. (2004) define cross-border mergers and acquisitions (M&As) as those M&As which involve an acquirer and a target company headquartered in different home countries.

Cross-border M&As represent one of the options firms have for strategic expansion in the turbulent globalizing world. The period of the 1990s noted extreme growth in the number of cross-border M&A deals. According to Hitt et al. (2001a, b) the number of M&As completed in 1997 was higher than number of all M&As during the 1980s. Another wave of increase in M&A activity took place between 2003-2005 (World Investment Reports, 2006). When an M&A is completed the acquiring company obtains the target’s company resources such as knowledge base, technology and human resources (Shimizu et al., 2004). Subsequently, new executive board positions are created which are further filled by managers coming from both companies: an acquirer and a target company. Although Hambrick and Cannella (1993) found that 27 per cent of executive managers from target companies leave in the first year after the acquisition is completed, scholars argue that top management teams become and stay more nationally diverse after a cross-border acquisition.

Therefore, the causality between the cross-border M&A activity and nationality diversity of executive board what I intend to test is following: internationalization of executive board is dependent upon cross-border M&A activity.

Based on this argument Staples (2007) examined cross-border acquisitions (valued at $1 billion or more) of 148 transnational companies. After filtering the total amount of cross-border acquisitions that occurred between 1995 and 2004, he acquired a database containing 239 ‘mega deals’ worth almost $1,5 trillion. His findings are in line with the general argument that cross-border acquisitions create more internationalized executive boards. Additional supporting evidence on the relationship between cross-border M&As and internalization of executive boards found Van Veen and Marsman (2008).

This leads to the following hypothesis:

H3: Cross-border M&A activity in the period from 2000 to 2008 is positively associated with nationality diversity of executive boards.

3.4. Foreign Ownership

At first, it might seem that there is an overlap between cross-border M&A activity and foreign ownership. However, each of those firm variables involves a different perspective. The cross-border activity considered in this research is only the activity where a company from the dataset was an acquiring company. In contrast, foreign ownership focuses on how executive boards are affected by the ownership structure of companies. Therefore, there is no overlap between those two firm variables.

As I have already indicated, a company’s internationalization has an effect on the composition of its top management teams. An interesting question is how to measure firm’s internationalization.

According to Dörrenbächer (2000) there is no universal concept to measure a firm’s internationalization. However, there are three different approaches: structural indicators, performance indicators, and attitudinal indicators. One of the structural indicators is the amount or proportion of shares owned by foreigners (Dörrenbächer, 2000). The level of foreign ownership is subsequently associated with the governance aspects of companies (Hassel et al., 2003). Corporate governance concerns various mechanisms (legal, institutional and cultural) which enable shareholders to control corporate insiders and management (John and Senbet 1999; Peace and Osmond, 1999). Since foreign owners may want their national representative to be present in management teams, foreign ownership increases the likelihood of appointing foreign executive managers. Moreover, Oxelheim and Randøy (2003) suggest that having non-national board members (especially Anglo-American citizen) signals both openness to foreign investors and a commitment to corporate transparency.

Therefore, the causality between the foreign ownership and nationality diversity of executive board is following: internationalization of executive board is dependent upon foreign ownership.

Based on the above argument Ruigrok and Greve (2008) examined the relationship between foreign-based ownership and top management team nationality diversity. To my knowledge their study is the only one so far providing us with empirical evidence. They concluded that companies with higher proportions of foreign owners are more likely to appoint an executive member from outside the home country.

This leads to the following hypothesis:

H4: Foreign ownership is positively associated with nationality diversity of executive boards.

4. D

M

4.1. Analysis

My analysis explores the relationship between considered company specific variables and nationality diversity of executive boards coming from Norway and Denmark. In order to be able to find the differences in the relationships between the two countries, I divide my observations into two databases: Norwegian and Dannish. Subsequently, each of the databases is separately analyzed by correlation and multiple regression analysis.

First, I search if there is a correlation between independent variables. In case of high correlation of certain independent variables, one of the highly correlated independent variables will be removed from the model. Thus, multicolinearity will be avoided and the least squares estimates will explain the model in the best way in sense of Gauss-Markov theorem (Kleinbaum and Kupper, 1978).

Second, I assume that the relationship between all the variables is linear. Therefore, in order to control for relations between all the independent variables and dependent variable at once, I run regression analysis with ordinary least square method. I choose multiple regression analysis because it is the most suitable method for description of the extent, strength and direction of the relation between several independent variables and a continuous dependent variable (Kleinbaum and Kupper, 1978).

Tabachnick and Fidell (1989) argue that in order to run standard multiple regression a bare minimum requirement is to have at least 5 times more observations than independent variables. This 5-1 ratio is sufficient to develop a sample size that ensures a chance to reject a null hypothesis.

4.2. Data Sample

Following earlier studies (Glunk et al., 2001) I focus on the top 30 companies measured by total sales figure in each country. According to Benito et al. (2002) these largest companies are dealing with a dilemma whether to retain their local roots or not in their globalization process. As I test four independent variables, the number of companies in each country in my database is in accordance with the recommended minimal data sample size (5 cases per independent variable) by Tabachnick and Fidell (1989). I have not involved companies which

have just one executive manager. Furthermore, I excluded those companies being in a parent company – subsidiary relationship and simultaneously showing completely the same financials in Amadeus database. However the websites and further information about the subsidiaries were impossible to track down. I assume that in this case the financials in Amadeus database are reported as consolidated in parent company as well as in the subsidiary.

The governance structure in both countries is two-tier (Christensen and Westenholz, 1999;

Ruigrok and Greve, 2008). In my research I focus only on executive board. This group comprises of top managers who are responsible for daily management of the company. In cases when a company does not explicitly state who belongs to executive management team, I incorporate team which is closest to the CEO.

4.3. Data Sources

As primary data source for detecting executive managers’ nationality I use corporation’s websites. In case there is no information on the composition of executive board, I retrieve annual reports which provide supplementary evidence on the executive board members. In spite of the fact that in most cases companies publish information on their top management teams, the nationality of individual board members is still often missing. There is only 15 % of nationalities detected from corporate web sites or annual reports. Therefore, I contact the remaining 85 % of companies directly. Since the top managers might be reluctant to respond to my phone calls or e-mails, I contact Investor Relations Department or Communication Department. In cases where those departments do not exist, I turn to a secretary of a particular manager. The phone calls enabled me to determine the nationality of 74 % of board members.

I have to mention that, in general, the response rate to the phone calls was really high, some 95 %. Although I attempted to determine nationality of the remaining managers through various on-line sources, such as Zoom Information, and press releases, I did not manage to identify the nationality of any of the managers. Therefore, there was still 11 % of nationalities unknown. Subsequently, I made following steps. First, I assumed that managers’ nationality is associated with the country where they received their first academic degree. Second, I made an assumption of remaining managers’ nationality based on their surname.

Taking into account the fact that 89 % of managers are assigned nationality based on information conveyed by corporation, either web site or phone call, I am really confident that

the data in my database is credible. There might be some individuals who are assigned wrong nationality in the remaining 11% of managers, however, the percentage of an error is minimal.

For my analysis of the link between company specific variables and nationality diversity of executive boards, I make use of several data sources. The ranking of companies in each country according to total sales figure in 2007 is retrieved from Amadeus database. Number of employees of a particular company is gained either from Amadeus database or corporation sources such as annual reports. Size of executive board is computed based on either direct phone call with company representatives or corporation web page. For identifying ownership structure of companies I use either Amadeus database or corporation annual report. Lastly, the data on cross-border M&As activity are retrieved from Zephyr database.

4.4. Dependent Variable

Nationality diversity

I follow earlier studies by (Ruigrok and Grepe, 2008; Carpenter, 2002) and operationalize nationality diversity in Blau’s Index of Diversity (Blau, 1977). It is a commonly used measurement for diversity as variety (Bunderson and Sutcliffe, 2002). The formula is following (Blau, 1977):

∑

−

= ^N

i

pi

D

1

1 2 (1)

Where p is the proportion of unit members in ith category. The values of Blau’s index are spread from zero to (I-1)/I. The maximal diversity occurs when the value of Blau’s index is equal to (I-1)/I, which means that unit members are spread equally. The higher the Blau’s index is the higher the diversity of a group.

4.5. Independent Variables

In this section I present how individual independent variables were operationalized.

Size of Company

Size of company is defined as total number of employees of a company. The reason for operationalizing size of company as total number of employees is that Van Veen and Marsman (2008) argue that total number of employees is a better predictor than total sales.

Size of Executive Board

Size of executive board is defined as total number of executive managers in a company.

Cross-border Merger and Acquisition Activity

A dummy variable is employed to state whether a company took part in either merger or acquisition worth at least € 100 million between years 2000 and 2009. The activity which is considered in this research is only the activity where a company was the acquirer. This operationalization is based on Staples (2007). Nevertheless, since the companies in UNCTAD list are significantly larger in terms of total sales than companies in my data sample, I did not follow Staples (2007) criterion on the deal value. I assume that deal value of $ 1 billion would be too high for Danish and Norwegian companies. Therefore, I include completed deals of at least € 100 million value.

Foreign Ownership

Foreign ownership is defined as a percentage of non-national shareholders. Only direct shareholders were considered, I have not incorporated ultimate owners. According to Benito et al. (2002) companies which are state owned are limited in their scope of strategic choices. I assume that this limitation has an impact on the access of non-national managers to executive boards. Therefore, the companies in which the government was the major shareholder were marked as company with solely domestic shareholders.

5. R

R

5.1. Descriptive Results

In the Appendix 1 and 2 (see Appendix part) I present descriptive results on a company level, for a country level see Table 1. First, the appendices show what is the percentage of foreigners in executive boards and how nationally diverse the executive boards are. As we can see, Danish executive boards (0.05) are on average way less nationally diverse in comparison to Norwegian ones (.240). What is interesting is that in each country the average percentage of foreigners and average nationality diversity is almost identical. However, comparing the scores for each company, we can see that there are usually significant differences, e.g. the largest difference is in NCC ROADS AS company which scores 1.00 on percentage of foreigners in executive board members and .00 on Blau’s index of diversity. On the other hand, for some companies the scores are the same, e.g. NORSKE SKOGINDUSTRIER ASA company reaches .50 on both measurements. Second, the appendices provide information on companies’ total sales figures. The Norwegian companies are on average bigger (milEUR 6,754) compared to Danish ones (milEUR 5,811). Third, there was a similarity in cross- border M&A activity of Norwegian and Danish companies. There were 6 companies involved in cross-border M&As in Norway and 7 companies in Denmark. Fourth, regarding to foreign ownership I detected that on average 47 % of each Norwegian company is controlled by foreign shareholders. In contrast, non-national shareholders hold on average only 9 % of each Danish company. Fifth, the appendices disclose information on state’s presence in companies.

I found that state is a significant shareholder in 6 Norwegian companies, which is a double amount compared to Danish companies (n=3). Lastly, the appendices also provide an overview of firms’ industry affiliation. The top 30 Norwegian companies are dominated by following resources orientated industries: ‘crude petroleum and natural gas’, ‘primary smelting and refining of nonferrous metals’, ‘wholesale of solid, liquid and gaseous fuels and related products’, and ‘abrasive, asbestos and miscellaneous nonmetallic mineral products’.

Contrarily, the Danish companies are most often affiliated to ‘offices of bank holding companies’ industry followed by various industry types within department stores services or food production chain.

There are some other substantial differences between the Norwegian and Danish companies (see Table 1). As far as the average size of companies in terms of total employees is concerned, the Danish companies were on average larger by 36.945 employees. On the other hand, although the Danish companies are on average larger in terms of total employees, their executive boards are smaller compared to Norwegian ones. The average size of Danish executive board is 4.7 executive board members while the same variable reaches 7.7 in Norway. Moreover, Table 1 also contains data on the number of companies with at least one foreign executive board member in each country. There are 17 companies with at least one non-national executive manager among the Norwegian companies which is approximately 4 times more than in case of Danish companies (n=4).

Table 1 Comparison of Norwegian and Danish executive boards

Number of companies with Average

Country

M&A Activity

State Ownership

≥1 Foreign Manager

Foreign Owners

Total Sales in milEUR

Total Employees

Average Blau Index

Average Executive

Board Size

Norway 6 6 17 17 6 754 5 677 .240 7.7

Denmark 7 3 4 5 5 811 42 662 .005 4.7

Table 2 shows statistics regarding nationality diversity in Norway. We can see that 174 executive managers out of 232 are Norwegian, which is 76 %. The most frequent foreign nationality in Norwegian executive boards is Swedish with 6 % share. Other nationalities with shares above 1 % are following: Denmark, Italy, USA, Finland and UK.

Table 3 shows statistics regarding nationality diversity in Denmark. The empirical evidence shows that only 9 out of 141 executive board members are non-national, which means that 93 % of executive board members are Danish. Furthermore, the most frequent foreign members of executive boards are from Sweden followed by France.

Table 2 Origins of Executive Board Members (Norwegian Companies)

Frequency

Norway 174

Sweden 15

Denmark 11

Italy 7

USA 6

Finland 4

UK 4

Germany 3

France 2

Korea 2

Chile 1

Litva 1

Belgium 1

Canada 1

Total 232

76%

2% 1%

2%

3%

3%

6%

5%

Norway Sweden Denmark Italy USA Finland UK Chile Germany France Korea Litva Belgium Canada

Table 3 Origins of Executive Board Members (Danish Companies)

Frequency

Denmark 132

Sweden 3

France 2

Finland 1

Netherlands 1

Russia 1

UK 1

Total 141 93%

1% 1% 1%

1% 1%

2%

Denmark Sw eden France Finland Netherlands Russia UK

To sum it up, both Table 2 and Table 3 show that executive board members from other Scandinavian countries are the ones who are assigned executive board positions most frequently.

Subsequently, I calculate Pearson correlation coefficients between considered company specific variables and nationality diversity of executive boards.

In both countries there is no significant correlation between the independent variables:

company size, executive board size, M&A activity and foreign ownership. Thus, multicolinearity is avoided..

Table 4 Descriptive Statistics and Pearson Correlation Coefficients (Norway)

Norway: variables Mean S.d. 1 3 4 5 6

1 Blau's index .240 .265 1 .262 .006 -.238 .384**

2 Company size° 5 677 8 369 .262 1 .097 -.162 -.163 3 Executive board size 7.733 2 180 .006 .097 1 -.540 -.019 4 M&As activity .200 .406 -.238 -.162 -.054 1 -.099 5 Foreign ownership .466 .450 .384** -.163 -.019 -.099 1

** Correlation is significant at the 0.01 level (2-tailed).

* Correlation is significant at the 0.05 level (2-tailed).

° In thousands of employees

Table 5 Descriptive Statistics and Pearson Correlation Coefficients (Denmark)

Denmark: variables Mean S.d. 1 3 4 5 6

1 Blau's index .045 .120 1 .009 .556** .696** .061 2 Company size° 42 622 104 055 .009 1 -.149 .254 -.120 3 Executive board size 4 700 2 292 .556** -.149 1 .353 .004 4 M&As activity .233 .430 .696** .254 .353 1 -.041 5 Foreign ownership .092 .250 .061 -.120 .004 -.041 1

** Correlation is significant at the 0.01 level (2-tailed).

* Correlation is significant at the 0.05 level (2-tailed).

° In thousands of employees

5.2. Results of Multiple Regression

In order to test the hypotheses regarding size of company, executive board size, mergers and acquisitions activity, and foreign ownership, I conduct regression analysis.

5.2.1. Norway

Table 6 shows the results of multiple regression. The R of the multiple regression model is .527, the R² is .277 and the adjusted R² is .162. Therefore, the model explains 16.2 % of the dependent variable.

Table 6 Regression Analysis, Norway

Unstandardized Coefficients

Standardized Coefficients

Model B Std. Error Beta t Sig.

(Constant) ,113 ,179 ,631 ,534

Employees 9,78E-006 ,000 ,309 1,757 ,091

BoardSize -,003 ,021 -,024 -,139 ,891

MAsActivity -,096 ,113 -,147 -,848 ,405

1

ForeignShareholders ,243 ,101 ,420 2,415 ,023

a Dependent Variable: Nationality diversity R = ,527

R²=,277

Adjusted R²= ,162

Moreover, Table 6 also shows expected positive significant relation between company size and nationality diversity of executive boards (β = 0.309, p/2 value = 0.046). Additionally, the effect of foreign shareholders on nationality diversity of executive board is also significant and positive (β = .420, p/2 value = .012). Therefore, hypotheses 1 and 4 are supported.

Surprisingly, hypotheses 2 and 3 are rejected. Involvement in M&A activity has no significant effect on nationality diversity. Similarly, there is no significant relation between size of executive board and its nationality diversity.

5.2.2. Denmark

Table 7 presents the results of multiple regression model for Danish companies. The R is .779, the R² is .607 and the adjusted R² is .544, which means that 54.4 % of the dependent variable is explained by the set of employed independent variables.

Table 7 Regression Analysis, Denmark

Unstandardized Coefficients

Standardized Coefficients

Model B Std. Error Beta t Sig.

(Constant) -,074 ,037 -1,991 ,058

Employees -9,93E-008 ,000 -,086 -,639 ,529

BoardSize ,017 ,007 ,329 2,366 ,026

MAsActivity ,168 ,039 ,604 4,255 ,000

1

ForeignShareholders ,036 ,060 ,075 ,591 ,560

a Dependent Variable: Nationality diversity R = ,779

R²= ,607

Adjusted R²= .544

Table 7 also reports support for hypotheses 2 and 3. The size of executive board has a significant positive impact on the national composition of the executive board (β = .329, p/2 value = .013). Likewise, I observe strong positive relation between the involvement in cross

border M&As and the nationality diversity of executive board (β = .604, p/2 value = .000). On the other hand, hypotheses 1 and 4 are not supported meaning that in case of Danish companies there is no relation between nationality diversity of executive boards and both size of the executive board and foreign ownership.

6. C

D

The main objective of this study was to investigate to what extent can be nationality diversity of Norwegian and Danish executive boards explained by the following company specific variables: size of company, size of executive board, cross-border M&A activity, and foreign ownership. The secondary objective of this research was to explore the differences in relationship between nationality diversity of TMTs and considered company specific variables in a comparative study of Norway and Denmark.

6.1. Internationalization of executive boards

First, I compare the nationality diversity of Norwegian and Danish executive boards what I determined with previous research. Second, I provide an explanation why there are differences in the level of internationalization between the studies. Third, I compare the internationalization of Norwegian and Danish executive boards to other European countries.

Last, I explain why the level of internationalization of executive boards differs.

Norway and Denmark

My results show that 24.2 % of executive board positions in Norwegian companies were occupied by non-national executive managers, while Danish executive boards were represented by only 5 % of non-national members. Table 8 presents a comparison of my results with Ruigrok and Greve’s (2008) findings. Both empirical studies suggest that Norwegian companies have more nationally diverse executive boards. According to Ruigrok and Greve (2008), the difference between nationality diversity in Norway and Denmark was only .04. The difference in my findings concerning nationality diversity between those two countries is .19. The fact that they detected quite lower nationality diversity scores, particularly in Norway, may be caused by their sample size. They included only the seven most profitable companies which were most likely wholly/partly-owned by Norwegian state, e.g. companies affiliated to oil and gas industry. According to The Norwegian Ministry of Labor and Administration report (1997) management of those wholly/partly state owned companies is appointed by the respective ministries. In addition, the report states that the top management positions are often occupied by former Norwegian politicians. Therefore, I argue that it is very unlikely for a non-Norwegian manager to be appointed to the executive board position in wholly/partly state owned companies.

Furthermore, Van Veen and Marsman (2008) show that Danish executive boards contain 12.5 % of non-Danish executive board members. Their result contradicts mine since I noted only 5 % of non-Danish executive board members. Ruigrok and Greve (2008) argue that the internationalization patterns of Danish TMTs were decreasing in the period from 2000 to 2005, I assume that this tendency has continued until April 2009. Therefore, the difference between my results and Van Veen and Marsman (2008) might be caused mainly by the effect of time. The Van Veen and Marsman (2008) conclusions are based on their database from 2005 while my database reflects nationality diversity in March and April 2009.

Table 8 Comparison of my results with Ruigrok and Greve (2008) My results Ruigrok and Greve

(2008) Country

PF BI PF BI

Norway 0,242 0,240 0,057 0,080

Denmark 0,050 0,050 0,018 0,023

Note: PF = Proportion of foreigners in executive board; BI = Blau’s index of diversity

Comparison with other European countries

In this section I first compare my results from Norway and Denmark to previous research on other European countries. Furthermore, I explain why Norwegian companies score high on the nationality diversity of executive boards and why Danish companies score low.

In order to compare internationalization of executive boards in Norway and Denmark to other European countries I chose a study conducted by Van Veen and Marsman (2008) as they focus only on executive boards (as I did). They conclude that executive boards headquartered in Luxemburg are the most internationalized (76.2 %) followed by Dutch (46.4 %), British (25.2 %) and German (21.4 %). On the other hand they find that the least internationalized executive boards come from Spain (2.5 %), Portugal (5 %) and Italy (7 %). Contrasting my results with Van Veen and Marsman (2008) I argue that Norwegian executive companies are some of the most internationalized in Europe since they are almost as nationally diverse as British ones. On the other hand, Danish executive boards are one of the least internationalized boards. The Danish executive boards would occupy the next to last position on the internationalization ranking provided by Van Veen and Marsman (2008, Table 1). Therefore, the internationalization of Danish executive boards is lagging behind the most internationalized ones. There may be various reasons. First, Glunk et al. (2001) made a comparison of top 30 companies in Great Britain, Denmark and the Netherlands. They indicated that Danish companies are the smallest in terms of both annual turnover and number

of employees. As Van Veen and Marsman (2008) concluded, internationalization of executive boards is dependent on the size of the company no matter how the internationalization is expressed whether in total sales or total employees. I argue that the relatively lower size of Danish companies might affect the internationalization patterns of Danish executive boards.

Second, the Danish executive boards are characterized by an average size of 3 board members, which is 3 board members less than the average executive board in the United Kingdom, and 2 less than those in the Netherlands (Glunk et al., 2001). The lower average size of Danish executive boards may be responsible for lower internationalization patterns of Danish executive boards. Based on Hambrick et al. (1998), I assume that placing a non- national manager to a relatively small executive board may be riskier than placing him/her to a relatively big executive board.

Benito et al. (2002) argue that Norwegian internationalization patterns significantly increased in 1990s in order to catch up with the other Scandinavian countries. The importance of internationalization of Norwegian companies was also influenced by the fact that Norway is not a member country of European Union. Therefore, the Norwegian companies relocated their strategic activities to companies headquartered in European Union. Furthermore, Benito et al. (2002) also states that recently Norwegian companies tended to internationalize more in comparison to the other Scandinavian countries because of the weak clusters in Norway.

Thus, the companies had to search for an access to clusters in other countries (Benito et al., 2000). I assume that this necessity of Norwegian companies to internationalize resulted in the fact that 8 out of the top 30 Norwegian companies are direct subsidiaries of foreign MNCs.

Moreover, another 9 companies are to some extent held by foreign shareholders. The high proportion of foreign-based owners then results in more nationally diverse boards (Ruigrok and Greve, 2008).

6.2. Company specific variables

Subsequently, I tested the impact of a set of four company specific variables on the nationality diversity of executive boards. The results show that in each country the considered company specific variables play a different role in terms of both strength of the model and significance of individual variables. Furthermore, I compare my results to the extant academic literature.

Norway

In Norway the considered company specific variables explain only a low percentage of the variance of nationality diversity. Furthermore, only two out of the four company specific variables are concluded as having a significant positive effect on the dependent variable. The bigger the company in terms of total employees, the more nationally diverse its executive board is. Additionally, the more shares that are held by foreign shareholders, the more internationalized the executive board is.

The reason why the other two company specific variables, namely cross-border M&A activity and size of executive board are not supported might be following. In my database, 6 companies out of 30 are to some degree owned by Norwegian state or municipalities.

According to The Norwegian Ministry of Labor and Administration report (1997) only managers with prior political experiences are assigned to those top management positions. For instance, in case of Statoil AS the managing directors were formerly employed as state secretaries. Therefore, it is not a surprise that none of the companies with a state stake had a non-national executive board member. Looking closely at cross-border M&A activity, half of the companies completing an M&A deal in the observed period are state owned. I assume that the influence of the state on composition of executive board is the crucial factor which results in rejection of my hypothesis about cross-border M&As. As far as the size of executive board is concerned, I assume that the reason for not finding a significant relationship is following: in the database 46.7 % of companies showed either internationalized small executive board or non-internationalized large executive board. Those companies are presented in Table 9. On one hand, we can see that there are 7 companies with board size bellow average size (average size = 7.7; see Table 1) and simultaneously being nationally diverse. Looking at the ownership structure of those companies, 4 are direct subsidiaries of MNCs, the majority of the shares of another 2 companies are held by foreign companies and the last company is owned mainly by Norwegian corporate bodies. On the other hand, I detected nationally homogeneous boards in 7 companies having executive boards larger than average size (average size = 7.733, see Table 1). Out of those 7 companies, 3 are partly/wholly owned by state, 3 wholly owned by Norwegian corporate bodies and 1 is mainly held by owners outside Norway.