The influence of Executive Board Diversity on firm performance : the moderating effect of the degree of internationalization

(1)

The influence of Executive Board Diversity on

firm performance: the moderating effect of

the degree of internationalization.

MSc Business Administration: International Management

O.P.J. Nijhuis - 10871470

Name: Olivier Nijhuis

Supervisor: Francesca Ciulli

Second Supervisor: Michelle Westermann-Behaylo

Date: 29 January 2016

(2)

2 | T h e s i s O . P . J . N i j h u i s

Statement of Originality

This document is written by student Olivier Nijhuis who declares to take

full responsibility for the contents of this document.

I declare that the text and the work presented in this document is original

and that no sources other than those mentioned in the text and its

references have been used in creating it.

The Faculty of Economics and Business is responsible solely for the

supervision of completion of the work, not for the contents.

(3)

Abstract

Studies about the relationship between board diversity and firm performance have found very ambiguous results during the last decade. However, the literature has been neglecting the effects of relations-oriented board diversity. Relations-oriented diversity types (age, gender and nationality) are becoming more important due to both social and institutional pressures and they are directly related to board performance. In addition, International Business literature has not yet focused on the role of the increasing trend of internationalization on the relationship between diversity and firm performance. This thesis aims to investigate the effect of relations-oriented board diversity on firm performance in multinational enterprises (MNEs). Moreover, we investigate the moderating effect of internationalization on the relationship between diversity and firm performance. We argue that, following the Resource-Dependence theory, there is a positive relation between diversity and firm performance and this relation is positively moderated by internationalization.

Using a sample of 253 MNEs, our study reveals significant empirical support for the direct effect of nationality diversity on firm performance. Secondly, we find empirical support for the moderating role of internationalization on the effect of gender diversity and nationality diversity on firm performance. This thesis is the first study to analyze the moderating effect of internationalization on the relationship between board diversity on firm performance. The associated managerial practices of this study show the importance of diversity within executive boards when a firm is internationalizing. MNEs can use this information to take on a strategic position and gain a competitive advantage when they are expanding their operations cross borders.

(4)

List of figures

Figure 1: Theoretical framework ... 21

Figure 2: Overview results - Cross industry ... 38

List of tables

Table 1: Overview dataset ... 23

Table 2: Summary Variables... 26

Table 3: Correlation Coefficients ... 28

Table 4: Correlation Coefficients - industry specific... 31

Table 5: Overview steps regression analysis ... 32

Table 6: Overview results - Direct effects ... 35

(6)

Acknowledgements

First of all, I would like to thank my supervisor, Ms. Francesca Ciulli, for all her efforts. With her patience and excellent guidance, she has stimulated me to continuously improve my work. She always wanted to give additional feedback on my work, even if the next deadlines was weeks ahead and I had send her a revised piece of my thesis. I have learned a lot from her supervision and feedback. In addition, I would like to extend my appreciation to my second supervisor, Dr. Michelle Westermann-Behaylo, for taking the time and the effort to evaluate this thesis. Secondly, I would like to thank my friends and fellow students. The last two years were hectic and I am glad that they stuck with me, when I could only think and talk about my study. Finally, I would like to thank my parents, as they managed to keep me on track during my bachelor and master period. Their endless support has encouraged me to finish my studies succesfully.

(7)

1. Introduction

The moment businesses cross national borders and start internationalizing their operations, business environments become increasingly more complex (Ruigrok, Peck & Tacheva, 2007). The uncertain and competitive environment of a multinational enterprise (MNE) has made it difficult for MNEs to rely solely on the capabilities of their CEOs. Rather, it is the combined capacity of the executive board that influences long-term success (Cannella, Park & Lee, 2008). The actions and roles of the entire executive board are related to the profitability of firms and are one of the major sources for super-normal profits (Cannella et al., 2008; Nielsen, 2010). Due to the growing internationalization of firms, directors are required to respond to a variety of different problems within their business operations, from both home and host countries (Ruigrok et al., 2007; Mahadeo, Soobaroyen, Hanuman, 2012). A high level of diversity, i.e. the degree of characteristic heterogeneity amongst board members, allows them to deal with the different aspects of managing a firm internationally (Miller & Triana, 2009). Differences in board composition and diversity lead to different strategic choices and ultimately to different performance outcomes (Cannella et al., 2008). Diversity is therefore an interesting aspect, when analyzing the impact of the board on firm performance (Mahadeo et al., 2012)

In the last decade, diversity within boards has increased, due to either practical reasons e.g. globalization, social pressures e.g. reputation or institutional pressures, e.g. laws (Nielsen & Nielsen, 2013; Ali, Ng & Kulik, 2014). For example, some countries have introduced gender diversity quotas or other types of regulations to indirectly push companies towards a more balanced gender proportion at board level (Ali et al., 2014). Gender diversity is not the only type of diversity that is on the rise, nationality diversity also has been increasing (Van Veen & Marsman, 2008). Indeed, in their research, Van Veen & Marsman (2008) found that in 1993 only 36% of the companies they analyzed had at least one foreign board member. During the follow-up study in 2007, of the same companies 75% now had at least one foreign board member. This can be seen as a consistent increasing trend in nationality diversity. A foundation of this trend is according to Nielsen and Nielsen (2013) the acceleration of globalization and executive search that is crossing national borders.

As the importance of directors’ characteristics increases, an increase in academic recognition is also visible (Ruigrok et al., 2007). However, extant literature on board diversity has been more focused on task-related characteristics of board members, such as educational or functional background, as well as organizational and board tenure factors (Westphal & Zajac, 1995; Golden & Zajac, 2001; Ruigrok et al., 2007; Ali et al., 2014). Yet, next to task-related characteristics, there are relation-oriented diversity characteristics, i.e. age, gender and nationality. According to Ruigrok et al. (2007), these relation-oriented diversity characteristics are more interesting, as they pose a different type of challenge to firms and boards, compared to task-related characteristics. This is because the relations-oriented diversity

(8)

attributes, more than task-related characteristics, have an impact on the performance of the executive boards as it is related to communication and decision speed (Ruigrok et al., 2007). A board with high relations-oriented diversity is also more sensitive for misunderstandings and conflict (Ruigrok et al., 2007). That is why the focus on gender, age and nationality diversity of board members is increasing (Smith, Smith and Verner, 2005). The findings on the relationship between relations-oriented diversity and on firm performance however are very ambiguous. Both weak and strong effects between relations-oriented diversity and firm performance have been found as well as negative and positive effects (Carter, D’Souza, Simkins & Simpson, 2010; Mahadeo et al., 2012; Nielsen & Nielsen, 2013; Ali et al., 2014). Since the implications of the former on the latter are not clearly established yet (Bear, Rahman and Post, 2010; Nielsen & Nielsen, 2013), the inconsistent findings suggest the need for additional research (Ali et al., 2014).

In order to provide clarity on the ambiguous finding regarding the effect of executive board diversity on firm performance, separate types of diversity on firm performance needs to be analyzed. By analyzing only relations-oriented diversity instead of analyzing the general concept as a whole, more clear results about the relation between relations-oriented diversity and firm performance can be found (Nielsen & Nielsen, 2013).

In addition, the increase of the internationalization of MNEs, i.e. the increase in operations crossing national boundaries (Zweig, 2002), requires diverse board directors with the necessary experiences to cope with the different context of the countries the company has operations in (Ruigrok et al., 2007). Unfortunately, the literature on the relation between internationalization and firm performance has largely underplayed the role of managers (Nielsen, 2010). To the author’s knowledge, only one Swiss study has focused on the effect of internationalization on the relationship between diversity and firm performance (Nielsen & Nielsen, 2013). In order to successfully manage and exploit the benefits of internationalization, it is vital to explore the interaction between the increasing trends of internationalization and relations-oriented diversity. Therefore, In order to fill this gap in the literature, the research question this study aims to answer is:

What is the influence of the degree of internationalization, as moderating variable, on the relation between age, gender and nationality diversity within boards and firm performance?

This thesis will shed light on the influence of internationalization on the effect of relations-oriented diversity attributes, i.e. age, gender and nationality, within executive boards on firm performance. We believe that an increase of experience and knowledge from diversity becomes more beneficial as the business environment of a company becomes more complex due to the different

(9)

countries the company has operations in. As a result, we argue that internationalization positively moderates the effect of these three types of board diversity on firm performance.

This study has multiple contributions to the literature on board diversity and internationalization. First of all, an in-depth analysis of the effect of relations-oriented diversity on firm performance will be presented, which gives insights on the influence of age, gender and nationality as single attributes on firm performance. This thesis uses a large sample of firms based in the United States (US), over a 3 year observation period. Multiple studies analyzed this relationship with a follow-up period limited to 1 year (Richard & Shelor, 2002; Carter, Simkins, & Simpson, 2003; Ruigrok et al., 2007; Mahadeo et al., 2012; Ali et al., 2014), whereas a longitudinal research design provides stronger evidence (Ali et al., 2014). Other studies that used a longitudinal research, did not specifically focus on the relationship between executive board diversity and firm performance. For example, Richard and Shelor (2002) focused on diversity and sales growth. Bear et al. (2010) analyzed how gender diversity affects a firm’s corporate social responsibility, ratings and reputation. In addition, Ruigrok et al. (2007) analyzed how board members’ nationality and gender interacts with a directors’ level of independence, number of other directorships and demographic characteristics.

Secondly, this study will discuss the impact of internationalization on the diversity-performance relationship, which has lacked the academic attention of scholars. To the author’s best knowledge and understanding, only one study has analyzed internationalization as a moderator on the relationship between diversity and firm performance. Nielsen and Nielsen (2013) analyzed the moderating role of internationalization on board diversity and firm performance, but they solely focused on nationality diversity and limited their research by using a dataset with only Swiss based companies. Since Swiss firms have a relatively high degree of internationalization by virtue of the small home market size and long history of international operations, studying the moderating role of internationalization in another context or country (e.g. US) can lead different findings (Nielsen & Nielsen, 2013). As a final theoretical contribution, by focusing on characteristics of an individual (board member) level rather than firm level, this study emphasizes the role of executive board members in shaping and improving the internationalization strategy and performance of their firm. Next to theoretical contributions, this thesis has managerial implications and will provide practical information for executive boards, managers and human resource departments of MNEs. MNEs can take on a strategic position and gain a competitive advantage by increasing the degree of board diversity, depending on the level of internationalization and the type of industry they are competing in. For example, when a MNE wants to increase its performance during a period of internationalization and of expansion of its scope of operations, a diverse top management team could be most sufficient to do so. Consequently, recruitment guidelines can be set out and used to evaluate their degree of diversity in relation to the firm’s degree of internationalization.

(10)

This thesis is structured as follows: firstly, the main theoretical concepts of this study will be described. In this section, an in-depth review of the literature on the relations-oriented attributes will be addressed, including the relationship between these attributes and firm performance. Secondly, the concept of internationalization is introduced. Thirdly, the research model and hypotheses are presented in the theoretical framework section, which are supported by the literature. In the methodology section, the data used in this study is described. The data set and sample are described and the variables for this study are explained, including their sources, collection method and how they are measured. In the chapters following the methodology section, we present and discuss the findings that emerged from the empirical analysis. Finally, we bring forward the academic and managerial implications of this study and mention the limitations and suggestions for future research.

2. Literature review

This section will review the current theoretical debate about the impact of diversity on firm performance as well as the literature on the topic of internationalization. Firstly, diversity among executive board members is described as a general concept, followed by two different opposing theories that provide an explanation on how diversity influences performance. After this, the influence of specific types of relations-oriented diversity (i.e. age, gender and nationality) on firm performance are discussed. Finally, the literature about the relationship between internationalization and firm performance is reviewed.

2.1 Board Diversity

Board diversity as a general concept is defined by Miller and Triana (2009) as “the degree of heterogeneity among characteristics of board members”. This heterogeneity is categorized by Mahadeo et al. (2012) in two categories: directly observable and less observable. The first category encompasses directly observable characteristics, such as nationality, age, gender and ethnicity. The less observable characteristics are education, functional and occupational backgrounds. This categorization is similar to the relation-oriented versus task-related diversity, described by Ruigrok et al. (2007). In this classification, the directly observable characteristics match the relation-oriented diversity characteristics and the less observable characteristics the task-related diversity characteristics. Extant literature has given limited attention to the effect of board diversity on firm performance; the few empirical studies conducted have led to the identification of conflicting findings, and thus to the impossibility of ascertaining whether this relationship is positive or negative (Gong, 2006).

Two main theories have been adopted as framework to explore how and why board diversity affects firm performance; the Social-Identity theory (Tajfel, 1978) and the Resource-Dependence theory (Pfeffer and Salancik, 1978). Each one of these two theories provides a different underlying perspective

(11)

for the diversity-performance discussion. According to Tajfel (1978), the Social-Identity theory suggests that “individuals use demographic attributes, such as age and gender, to categorize self and others into social groups. In order for individuals to maintain a positive self-identity, they maximize the differences between in-group members (similar others) and out-group members (dissimilar others)”. Social-Identity theory thus provides an explanation to why diversity could lead to negative outcomes on firm performance. Due to the process of self-categorization, individuals are creating stereotypes. These stereotypes are likely to create in-groups and outgroups, which decrease the performance of the group (Turner et al., 1987). Therefore, according to this theory, diverse groups are more likely to experience difficulties working together, decreased team dynamics and report more group dissatisfaction (Nielsen & Nielsen, 2013; Ali et al., 2014).

Resource-Dependence theory, as introduced by Pfeffer and Salancik in 1978, provides an opposing view on the relation between board diversity and firm performance. According to the Resource-Dependence theory, the external environment influences a MNE’s firm performance. In this context, board members play an additional role. Next to their day-to-day board obligations, they are linking their firm to its external environment to secure critical resources (Ruigrok et al., 2007; Ali et al., 2014). According to this perspective, board members (and their skills, networks and experiences) are important for the ability to perform as boundary spanners for their companies. The Resource-Dependence theory argues that, as the board makes strategic decisions and has to cope with external stakeholders, a firm requires a diverse board. A diverse board can facilitate each of these functions through the increase of different resources and abilities needed for organizational survival (Pfeffer, 1972; Ali et al., 2014).

Throughout the literature on board diversity, many studies adopt these two theories. Among the research that is done, the studies by Murray (1989), Miller and Triana (2009) and Richard and Shelor (2002) are consistent with the Resource Dependency theory. Murray (1989) investigated 84 US food and oil companies and found that a heterogeneous group will be more able to implement and handle change than a homogeneous group. Miller and Triana (2009) add to Murray that diversity between board members lead to unique views and backgrounds brought to a firm. This allows executive board members of a heterogeneous board to better understand the global market of MNEs than a homogeneous board. Richard and Shelor (2002) found that diversity could lead to new approaches and better insights in the way other people act and think. This leads to an increase in considerations of alternative solutions. These studies are primarily focused on the abilities and resources of board members, which are increased by diversity. Due to this increase in abilities and resources, firm performance is increased.

On the other hand, scholars have also highlighted that, consistent with the Social-Identity theory, diverse groups come across problems, which do not appear within more homogeneous groups.

(12)

Diverse groups undergo negative experiences, such as decreased satisfaction within the group, lower levels of cohesiveness, reduced within-group communication, and higher conflict levels (Richard & Shelor, 2002). Therefore, this stream of research has found that dissimilarity often results in performance loss. Studies showing a negative relationship between board diversity and firm performance tend to focus more on the difficulties that arise when people with different characteristics have to work closely together, cooperating in an executive board. Ali et al., (2014) describe the existence of subgroups when they examined the effects of age diversity on firm performance, which led to negative firm performance.

The presence of conflicting results regarding the relation between board diversity and firm performance requires research to focus on specific features of board diversity. This study focuses on the direct effect relations-oriented diversity, i.e. age, gender and nationality diversity on firm performance. In the following section, the available literature on these three types of board diversity will be reviewed.

2.1.1. Board age diversity

Board age diversity is defined as the difference in age distribution amongst board members (Mahadeo et al., 2012). Board age diversity has not yet attracted much attention of research on board diversity (Ali et al., 2014). In an older study, Houle (1990) found that there are specific age groups (young, middle and old) that could lead to an increase in firm performance in their own specific way. In this differentiation, the old group (over 60) provides experience, which they accumulated throughout the years (Ali et al., 2014). Next to the old group, there is a middle group (between 35 and 60) that is in charge of main executive functions and a younger group (below 35) that is developing its knowledge and experience. According to Ali et al. (2014), younger directors are not only learning and gaining experience, but they can increase firm performance as well, as they tend to be higher educated and are more able to cope with new technologies. Although each age group has specific benefits, the existence of three subgroups are quite consistent with Social-Identity theory. Ali et al. (2014) found that directors are more likely to interact with other board members from the same age group. The clustering of older directors with each other and younger directors with each other can lead to the negative in-group/out-group behaviors (Ali et al., 2014).

In one of the few more in-depth reviews on age diversity on firm performance, Richard and Shelor (2002) performed a quantitative meta-analysis research on the age diversity within top management. They stated that in the research that had been done, no clear relationship between board age diversity and firm performance was found. In their review, they analyzed multiple studies in which the inconsistency in results is showed. Several studies showed a significant negative relationship between age diversity and firm turnover. For example, Zaject et al. (1991; as cited in Richard & Shelor,

(13)

2002) found that age diversity relates negatively to firm outcomes. Zenger and Lawrence (1989; as cited in Richard & Shelor, 2002) suggested that age diversity in group projects relates negatively to frequency of technical communication which they in turn link to turnover. Richard and Shelor also describe studies that find a clear positive relationship (Bantel & Jackson, 1989; Murray, 1989; Jackson, 1992). Bantel and Jackson (1989) and Murray (1989) both found that board members with differences in age produce more innovative and higher quality decisions than board members who were more similar in age. Jackson (1992) adds that boards composed of members with diverse ages classes produced a wider variety of ideas, alternatives and solutions than teams composed of people with similar ages. To make the results even more inconsistent, the study from Ali et al. (2014) found a negative relationship between age diversity and Return on Assets (ROA), but a positive relationship between age diversity and sales growth.

It is evident that previous research did not find consistent results. However, the studies cited in the article by Richard and Shelor (2002) were performed over 20 years ago. In addition, the research by Ali et al. (2014) was focused only on 1 year and limited to Australian firms. The unique aspects of the Australian context could have impacted the study findings, as they see this as one of the limitations of their study. Therefore, it is interesting to investigate the effect of age diversity on firm performance, especially since studies proved that age diversity affects firm performance, but on a more recent dataset with observations that exceeds the 1 year time lap.

2.1.2 Board gender diversity

The past decade, an increase of gender diversity within executive boards is seen. Gender diversity is (the increasing of) the number of female board members in an executive board (Bear et al., 2010; Ali et al., 2014). In order to discuss the relationship between gender diversity within executive boards and firm performance, it is relevant to discuss the differences between male and female directors identified by extant literature. In general, female directors have appeared to be more participative and democratic than men (Bear et al., 2010). When a larger number of board members is female, their presence can lead to more open conversations within the board (Bear et al., 2010). Also, women tend to be more detail-focused and risk averse than male directors (Graham et al. 2002). Regarding education and experience, it has been found that women in a board are more than twice as likely as men to hold a doctoral degree, which brings expert knowledge from outside the business (Bear et al., 2010). This allows women to bring additional perspectives to meetings and they can provide different input during decision making processes. In addition, female directors are more likely to gain board experience within smaller firms, and have less prior CEO experience (Bear et al., 2010). This can be explained by the fact that female directors are more likely to be given operational responsibilities rather than strategic ones (Peterson and Philpot, 2007). Finally, the number of female directors on a

(14)

board has shown to increase the reputation of MNEs. Studies have shown that gender diversity can act as a signal of firm value and strive for equality towards external stakeholders, which directly influences e.g. investors, media and consumers (Bear et al., 2010). When male and female directors take decisions together, risks are more critically weighed and investments are more thought through (Ali et al., 2014). As male and female directors possess different skills, knowledge and experience, the combination of these different perspectives could result in higher quality of decisions as well as a more successful functioning of the board as a whole (Bear et al., 2010, Ali et al., 2014).

However, these different point of views could lead to more conflicts, which will lead to negative effects. Ruigrok et al. (2007) state that boards with a higher number of women had more debates and disagreement. This effect is larger when the industry of the firm requires quick responses (Smith et al., 2005). In addition, as with age diversity and consistent with Social-Identity theory, boards with a high level of gender diversity can develop a ‘‘us vs. them’’ perceptions among its members (Ali et al., 2014). The relationship between gender diversity and firm performance is gaining more academic attention. Carter et al. (2010) found seven empirical studies that specifically tested the link between gender board diversity and the financial performance of the firm, which is a rather limited amount of evidence given the amount of discussion on the topic. The studies took place in a period from 1997 up to 2008 (respectively, 1997, 2003, 2005, 2006, 2007, 2008, and 2009). However, similarly to age diversity, no clear relationship between gender diversity and firm performance has been found in the literature. Of the seven studies Carter et al. (2010) analyzed, two studies illustrated a positive relation between gender diversity and firm financial performance (Carter et al., 2003; Campbell & Minguez-Vera, 2008), and three studies found no direct relation (Shrader, Blackburn & Iles, 1997; Farrell & Hersch, 2005; Rose, 2007; as cited by Carter et al., 2010). In addition, two studies found a negative relationship between gender diversity and firm performance (Smith et al., 2006; Adams & Ferreira, 2009; as cited by Carter et al., 2010). In their own research on US firms between 1998 and 2002, Carter et al. (2010) did find a positive, significant relationship between the number of women on the board and the ROA.

The conflicting findings of previous studies raise questions to why such different results can be found. One explanation can be that some of these studies are older, Carter et al. (2010) for example used a dataset from 1998 up to 2002. Another explanation is that some studies did not specifically focus on the effect of gender diversity on firm performance. For example, Bear et al. (2010) mainly examined how the number of women on a board affects a firm’s corporate social responsibility, ratings and reputation. In addition, some of the studies analyzed the effect of gender diversity on firm performance on firms, but focused on Danish (Smith et al., 2006; Rose, 2007; as cited by Carter et al., 2010), Spanish (Campbell & Minguez-Vera, 2008) or Swiss firms (Ruigrok et al., 2007). Diversity can lead to different results when analyzed in a different context or country (Nielsen & Nielsen, 2013). It is evident that

(15)

investigating the recent effect of gender diversity on firm performance of US firms will provide new insights to the previous findings.

2.1.3 Board nationality diversity

The third type of diversity studied in this research is nationality diversity. Nationality diversity is the amount of executive board members with different nationalities in a board (Nielsen & Nielsen, 2013). This is a relevant type of diversity to be studied, since in the last years national diversity has increased amongst board members (Van Veen & Marsman, 2008). However, empirical studies on the role of board nationality on firm performance are still rare (Nielsen & Nielsen, 2013). It has been acknowledged that, when teams are nationally diverse, they bring a wide range of knowledge of, and experiences with different institutional environments (Nielsen & Nielsen, 2013). Ruigrok et al. (2007) add to this that having different types of nationalities on a board may lead to the introduction of new ideas. This is confirmed by Nielsen (2010) as she found that nationality diversity is associated with an increase of creativity and the generation of more numerous and higher quality solutions. However, board members with a different nationality than the main group (i.e. Spanish national in an American board) may easily be trapped in a minority position. In addition, nationality is often associated with increased team conflict (Nielsen, 2010).

In the limited academic attention, there are 3 relevant studies that focused on the relationship between nationality diversity and firm performance. Nielsen (2010) found that diversity in nationalities is a reliable proxy for international orientation in MNEs that lead to valid foreign market entry decisions, which positively impacts firm performance. In addition, Miller and Triana (2009) found a positive relationship between board racial diversity and firm reputation. These two studies however, did not specifically focus on the relationship between nationality diversity and firm performance. To the author’s knowledge, only one study focused on the relationship between nationality diversity and firm performance. Nielsen and Nielsen (2013) investigated this relationship and found a positive relationship. Although this is a good indication for the effect of nationality diversity on firm performance, Nielsen and Nielsen (2013) used only Swiss firms in their study. As Swiss firms have a relatively high degree of internationalization by virtue of the small home market size, nationality diversity can have a different effect when analyzed in a different context (Nielsen & Nielsen, 2013).

Regarding the relationship between all three types of diversity and firm performance, no clear effects are described so far. This is either due the high ambiguous results from previous older studies, or due to the overall lack of studies. As in the last decade diversity within boards is increasing, it remains important to bring clarity to these discussions.

(16)

2.2 Degree of internationalization

Today, firms are increasingly expanding their geographic scope (Lu & Beamish, 2004; Rugman, Verbeke & Nguyen, 2011). Zweig (2002:3) defined internationalization as “the expanded flow of goods, services, and people cross state boundaries, thereby increasing the share of transnational exchange relative to domestic ones, along with a decline in the level of regulation affecting those flows”. Internationalization is the process of increasing involvement in international operations (Daszkiewicz & Wach, 2012). Over 100 empirical studies have tried to find a relationship between internationalization and firm performance (Hennart, 2007). In some of these studies, a small positive and monotonous relationship between internationalization and performance has been found. For example, the study by Vermeulen and Barkema (2002) state that as the amount of subsidiaries increases, firm performance increases. Chang and Rhee (2011) find that the expansion of foreign operations increases sales growth and firm performance, depending on the pace of the expansion. The gradual approach (slow pace) is intuitively appealing and follows a slow learning curve. The fast pace of internationalizing may be more appropriate when the industry of an MNE is globalizing rapidly and time is limited. In such industries, the competitive risks of gradual expansion (slow pace) may outweigh any gains from reduced market uncertainty (Chang & Rhee, 2011). Kogut and Kulatilaka (1994) argue that international expansion positively influences the performance of firms that have strong internal resources and capabilities. Other studies found a negative relationship between internationalization and firm performance (Click & Harrison, 2000; Denis et al., 2002). Click and Harrison (2000) report that internationalization results in inefficient use of assets, resulting in a reduction in the return on assets for MNEs. Denis et al. (2002) find that as MNEs are expanding and diversifying its operations throughout different countries, the risk of inefficient investment decisions increases, which negatively affects firm performance.

The studies describing a positive relation highlight the advantages of operating internationally. Increasing a firm’s operating countries will lead to the exploration and exploitation advantages. This can lead to economies of scale and scope, which reduce the costs of operations (Lu & Beamish, 2004). Increasing the operating countries also helps to reduce fluctuations and variation in revenue as investment risks is spread over different countries (Kim, Hwang, & Burgers, 1993). The studies that found a negative relation prove that internationalization does come with additional costs and negative consequences for firm performance. International expansion comes with a set of attendant costs (Lu & Beamish, 2004). When a firm expands in a foreign country, the board and the country manager are faced with many challenges. These can be operational (purchasing and installing new facilities), social (staffing, new external business networks) or cultural difficulties. Overcoming these challenges can be more costly and can require more time, which eventually could lead to performance loss (Lu & Beamish, 2004). Therefore, the potential benefits of operating in another country need to outweigh the costs and

(17)

the risks of operating in a foreign country and culture. The Uppsala process model by Johanson and Valne (1977) for internationalization argues that firms tend to initially expand in markets that are geographically and culturally close (Nielsen, 2010). As a result, firms expand in nearby geographic countries that may have similar aspects in culture, therefore decreasing the risks. As the firm increases its experience and knowledge from neighboring countries, it can overcome their liability of foreignness. This allow MNEs to slowly increase to even more distant country markets (Rugman et al., 2011). The role of learning and experience is therefore vital (Nielsen, 2010).

As shown by extant literature, there are a lot of benefits when a firm increases its level of internationalization, but internationalization also brings additional difficulties. In the next section, the effect of internationalization on the diversity and firm performance relation is discussed.

3. Theoretical Framework

The literature review shows that two central, yet opposing views can explain how diversity is affecting firm performance and that the discussion regarding the synergy between the two is still ongoing. It is evident that there is a need for clarifying the relationship between diversity and firm performance (Nielsen & Nielsen, 2013). Although firms are increasingly expanding their geographic scope (Lu & Beamish, 2004; Rugman et al., 2011), almost no literature has yet analyzed how this might affect the relationship between diversity and firm performance. As MNEs expand their flow of goods, services, and people cross state and National borders, the complexity of their operations increases and it requires diverse board directors with the necessary experiences to cope with the contextual diversities of the countries the company operates in (Zweig, 2002; Daszkiewicz & Wach, 2012).

Using this background and the connection between the different variables, we first investigate the general relationship between the different types of diversity and firm performance. This will be followed by an examination of the moderating effect of internationalization on these relationships.

3.1 Board diversity and firm performance

As discussed above, internationalization has benefits and consequences. The benefits mentioned are mainly operational or financial e.g. spreading risks, economies of scale and scope (Kim et al., 1993), whereas the difficulties of internationalization are not only operational. There are additional difficulties as cultural difficulties and lack of knowledge of both formal and informal institutions (Kim et al., 1993). In order to overcome these challenges, organizations need diversity among their board members. Diversity can help integrate a wider range of information to make more and better informed decisions (Ali et al., 2014). Organizations can reduce uncertainties and dependencies, if they capitalize on the full range of connections delivered by a diverse board (Miller and

(18)

Triana 2009; Pfeffer and Salancik 1978). Therefore, when the effect of internationalization on the relation between the relations-oriented diversity and firm performance is examined, the Resource-Dependence theory will provide a better foundation for the theoretical framework of this study than the Social-Identity theory. As internationalization requires different points of view and different skills sets and traits to achieve high performance, Resource-Dependence theory matches the importance of different skills of executive board members for performance. In addition, Resource dependence theory is said to provide the foundation for some of the most convincing theoretical arguments for a business case for board diversity (Carter et al., 2010). Therefore, as theoretical foundation for the theoretical framework of his thesis, Resource-Dependence theory is used.

3.1.1. Age diversity and firm performance

Even though the literature on the effect of age diversity on firm performance provides little or no consensus, I believe that by using the Resource-Dependence theory, a valid proposition regarding the effect of age diversity on firm performance can be made. Houle’s study (1990) showed that three age groups all provide their own type of resource. This might be experience (old age group), operational know-how (middle age group) or new knowledge and innovations (younger age group). According to Mahadeo et al. (2012), a lack of diversity in age could lead to a lack of interest in new ideas. This is supported by Richard and Shelor (2002), as they state that a higher age diversity leads to an increase in innovation. A high diversity could therefore provide more flexibility, which leads to more adaptive and agile strategies. In addition, diversity in age can lead to a more efficient division of labor than could be found in more homogenous boards (Ali et al., 2014). Although Social-Identity theory states that age could lead to negative performance, I believe that, following the Resource Dependence theory, diversity in age increases the availability of resources, thus improving executive board performance. Therefore, the first hypothesis is proposed as:

H1: Age diversity has a positive effect on firm performance.

This hypothesis leads to the statement that as the diversity in age increases, firm performance will increase as well.

3.1.2. Gender diversity and firm performance

Similar to age diversity, no clear relation between gender diversity and firm performance has been described in the current literature. From a Resource-Dependence theory point of view, it is possible to hypothesize a direct effect between gender diversity and firm performance. Firstly, as women are twice as likely as men to hold a doctoral degree (Bear et al., 2010), they bring more expert knowledge to the table, compared to a male-only executve board. Secondly, women are more frequently given operational responsibilities, whereas men are given more strategic ones (Peterson and

(19)

Philpot, 2007). As a result, women provide unique sets of information that are available to management for better decision making (Carter et al., 2010). I therefore conclude that from a Resource-Dependence theoretical base, it is evident that having more women on the executive board will increase the resources available. Combining the different perspectives from both male and female directors should result in higher quality of decisions as well as a more successful functioning of the board as a whole (Bear et al., 2010; Ali et al., 2014). As male and female directors possess different skills, knowledge and experience (Bear et al., 2010), which lead to an increase of resources and thus the ability to achieve better informed decisions, the following is hypothesized:

H2: Gender diversity has a positive effect on firm performance.

As the gender diversity increases and the amount of female board members increases (until the point where absolute gender parity has been achieved), firm performance will increase.

3.1.3. Nationality diversity and firm performance

Regarding the third relation-oriented diversity trait, nationality diversity, only a few studies address nationality diversity in top management teams as related to firm performance (Nielsen, 2010). It is worth considering however that, just as with the previously discussed diversities, Resource-Dependence theory can provide a good indication on what to expect from the relationship between nationality diversity and firm performance. Nielsen & Nielsen (2013) state that high nationality diversity leads to the ability of exploiting different institutionally embedded experiences. This decreases uncertainties and it will increase the range of connections, overcoming the difficulties of managing new stakeholders as employees, suppliers, and customers (Miller and Triana 2009; Pfeffer and Salancik 1978). It is evident that different nationalities give access to a lot of new resources, which can increase the performance of the board. Nielsen & Nielsen (2013) state that nationally diverse boards even have greater processing capacity and can attend to more environmental cues. This leads to the third hypothesis:

H3. Nationality diversity has a positive effect on firm performance.

Therefore, the more an executive board consists of members with different nationalities, the more firm performance will increase due to the increase in resources.

3.2 The moderating role of internationalization

As the relationship between specific types of diversity and firm performance is hypothesized, we can address the bigger problem. As said in the introduction, firms are increasingly expanding their geographic scope (Lu & Beamish, 2004; Mahadeo et al., 2012). However, to the author’s knowledge, only one study has addressed the influence of internationalization as a moderator in a relation between two other variables (diversity and firm performance). It is noted that this was a study based exclusively

(20)

on Swiss firms. Most studies use internationalization as a predictor or measure of firm performance (Nielsen, 2010). Kogut and Kulatilaka (1994) argue that international expansion positively influences the performance of firms, but only if they have strong internal resources and capabilities. Another interesting study examines the effect of age diversity on internationalization. They find that age diversity of boards may lead to an increase in internationalization of American and European firms (Rivas, 2012). This shows there is a relationship between internationalization and age diversity. It is interesting to examine how the relationship between diversity and firm performance is influenced when a firm increases its operations in foreign countries. If a positive moderating effect is noticeable, firms with a highly diverse board can take the strategic step of increasing their level of internationalization thus benefitting their firm performance. The available literature shows that diversity in a board will not only reduce costs, but also reduce the uncertainty involved when operating in foreign countries (Benito & Gripsrud, 1992). Mahadeo et al. (2012) add to this by stating that when a business becomes more international, it needs to respond increasingly more to social and environmental problems from home and host countries alike. This is sufficiently more simple to deal with if a board is truly diverse (Mahadeo et al., 2012). Therefore, it is posited that the relation between diversity and firm performance is positively moderated by internationalization. The hypothesized moderating effect of internationalization on three different types of relation-oriented characteristics and firm performance are explained below.

3.2.1. Moderating age diversity-Firm performance

As age diversity increases, a better combination of experience, knowledge and flexibility can be found. This mix between experience and flexibility could prove to be very useful in an uncertain environment or when the firm needs to cope with new, unforeseen problems on a social or environmental level (Mahaedo et al., 2012). Therefore, as internationalization increases, age diversity will have a larger positive impact on firm performance. It is proposed that:

H4: Ceteris paribus, the degree of internationalization positively moderates the effect of age diversity on firm performance.

The additional benefits of age diversity (experience and flexibility) are more useful, when the operational environment of the firm expands and becomes more difficult.

3.2.2. Moderating gender diversity-Firm performance

Similar to age diversity, the effect of gender diversity on firm performance will increase when the operational environment becomes more difficult. A firm’s reputation is important when operating internationally. Therefore, as reputation becomes more important for international expanding firms, combined with the fact that having female directors on the executive board increases firm reputation

(21)

(Bear et al., 2010), firm performance will benefit from diversity. In addition, when men and women take decisions together, risks are more critically weighed and investments are more thought through (Ali et al., 2014). In an uncertain and risky environment, this can be very useful. Both statements lead to the following hypothesis:

H5: Ceteris paribus, the degree of internationalization positively moderates the effect of gender diversity on firm performance.

3.2.2. Moderating nationality diversity-Firm performance

Last but not least, nationality diversity allows for valuable knowledge of both formal and informal institutions of different countries to be integrated into a firm’s strategic decisions. This will lead to a higher quality of strategic decisions, which in turn will lead to higher performance (Nielsen & Nielsen, 2013). Therefore, as internationalization increases, the benefits of nationality diversity exceed the risks of diversity even more and it is proposed:

H6. Ceteris paribus, the degree of internationalization positively moderates the effect of nationality diversity on Firm Performance.

To summarize the hypotheses, I propose an effect as described in the theoretical framework presented figure 1. This framework, and its hypotheses, follow the Resource-Dependence theory as it is focused on the added value of different skills, knowledges and experience that can result from board diversity. When the degree of internationalization increases, the effect of diversity of firm performance will be strengthened as the increase of complexity matches the increase of experience and knowledge.

H2 H3 H4 H5 H6 H1 Relations -oriented Board Diversity Age Firm Performance Degree of internationalization Gender Nationality

(22)

4. Methodology

For this study, a panel research design has been used to measure the effects of board diversity and firm performance, with internationalization as a moderator. The method that was used to analyze and test the hypotheses is a multiple linear regression analysis. The empirical results are based on the correlations and regression outcomes amongst the variables.

4.1 Sample and Data collection

The sample of this study consists of US MNEs that are listed in the Fortune 500. The choice to use US Fortune 500 firms as the target firms is based on 3 main reasons. Firstly, the Fortune 500 companies represent leaders within their industry (Miller & Triana, 2009). Therefore, they are seen as proper examples for other companies, allowing to generalize the findings of this study to other firms. Secondly, these firms are required to provide annual reports on their websites and keep their previous annual reports in their online database. This improves both the data quality as the data availability. Finally, multiple studies (e.g. Carter et al., 2003; Miller & Triana, 2009; Carter et al., 2010) have used these types of firms for diversity studies, which validates the data selection. The firms that have been selected had to be continuously listed, publically traded without being acquired, active between 2012 and 2014 and have cross-border operations. This selection leads to a data set with stable and similar firms, which improves the quality of this study (Miller & Triana, 2009).

Several different databases have been used in order to obtain the data on all variables. For the collection of the dependent variable, firm performance, I used DataStream, the Wharton Research Data Services (WRDS) and annual reports. DataStream is an Excel add-on, supported by the University of Amsterdam, and is a financial database containing company data, equities and macro-economic data. It holds time series information for over two million financial instruments, securities and indicators for over 175 countries and 60 markets worldwide, with up to 50 years of historical depth for some series. WRDS is a research platform and business intelligence tool for over 30,000 corporate, academic, government and nonprofit clients at over 375 institutions in 33 countries. It provides access to over 200 terabytes of data across multiple disciplines as Economics, Finance, ESG, and Statistics. In addition, annual reports are used to find additional information for the dependent variables. For the independent variables, age, gender and national diversity amongst board members, annual reports, WRDS and BoardEx were used. BoardEx is a business intelligence service which is used as a business development tool. Additionaly, it is a source for academic research concerning corporate governance and boardroom processes. BoardEx holds over 400,000 in-depth profiles of the world's business leaders, which is updated daily. For the moderating variable, internationalization, the WRDS database was also used. In addition, for the control variables, firm size and age, DataStream, WRDS and annual reporting was used.

(23)

For the industry effects as a control variable, the industries were classified by using the Global Industry Classification Standard (GICS®). The GICS consists of 10 industry sectors which are presented in table 1. The GICS is universal, reliable and annually reviewed. The GICS methodology has been widely accepted as an industry analysis framework for investment research, portfolio management and asset allocation. The WRDS provides a code to each MNE, which classifies a company within one of the 10 sectors.

From the original fortune 500 firms, 452 MNEs matched the sample selection (continuously listed, publically traded without being acquired, active between 2012 and 2014 and cross-border operations). From the 452 firms of this selection, 199 firms had little or no data on the variables available and were excluded from the selection. Therefore, the final dataset consists of 253 unique MNEs, with observations over a 3 year period (2012, 2013 and 2014), resulting in 759 observations. The final dataset is presented in table 1.

Table 1: Overview dataset

AGE DIVERSITY GENDER DIVERSITY NATIONALITY DIVERSITY

N Mean Min/Max SD N Mean Min/Max SD N Mean Min/M

ax SD CROSS INDUSTRY 759 .234 .00/ .50 .169 759 .249 .00/.50 .147 759 .141 .00/.60 .171 ENERGY 54 .235 .00/.49 .165 54 .173 .00/.42 .152 54 .071 .00/49 .149 MATERIALS 51 .231 .00/.48 .182 51 .200 .00/.50 .149 51 .139 .00/.56 .187 INDUSTRIALS 94 .169 .00/.50 .176 94 .244 .00/.48 .134 94 .126 .00/.50 .163 CONSUMER DISCRETIONARY 173 .231 .00/.50 .173 173 .252 .00/.50 .158 173 .136 .00/.54 .162 CONSUMER STAPLES 63 .235 .00/.50 .188 63 .295 .00/.49 .162 63 .245 .00/.60 .230 HEALTH CARE 78 .302 .00/.50 .127 78 .309 .00/.50 .106 78 .142 .00/.44 .137 FINANCIALS 99 .285 .00/.49 .134 99 .265 .00/.44 .124 99 .094 .00/.39 .125 INFORMATION TECHNOLOGY 90 .192 .00/.50 .165 90 .222 .00/.49 .152 90 .206 .00/.53 .169 TELECOMMUNICATIONS 9 .219 .00/.34 .164 9 .177 .00/.22 .037 9 .177 .00/.22 .037 UTILITIES 48 .232 .00/.48 .172 48 .255 .00/.49 .134 48 .101 .00/.59 .190

4.2 Variables

In the following section, the variables used in this study are explained and described how they are measured. The summary of the variables, their sources and measurements can be found in table 2. 4.2.1 Dependent Variable

The dependent variable in this panel research design is firm performance. Firm performance is predominantly based on financial criteria (Kennerley & Neely, 2003). There are multiple financial measures available to measure firm performance. Miller and Triana (2009) use the Return on Sales (ROS) as an predictor for firm performance, which is measured as the net income divided by net sales. Return on Investment (ROI)is another way to measure firm performance (Woo & Willard, 1983), measured as

(24)

the net income of a firm divided by its invested capital (Erhardt et al., 2003). A final financial measure for firm performance is the Return on Assets (ROA) (Mahadeo et al., 2012) and is calculated as follows:

ROA = net income + (1 - tax rate) (interest expense) / average total assets.

ROA is a measure of the success of a firm in using assets to generate earnings independent of the financing (debt versus equity) of those assets. Therefore, ROA can provide insight into how board diversity impacts an organization’s use of financial resources (Erhardt et al., 2003; Ali et al., 2014). ROA is used in multiple recent diversity studies as a predictor of firm performance (Carter et al., 2003; Erhardt et al., 2003; Campbell & Mínguez-Vera, 2008; Carter et al., 2010; Mahadeo et al., 2012; Ali et al., 2014). This validates the use of ROA over other financial criteria as predictor for firm performance. Therefore, this study is using ROA as equivalent for firm performance, the dependent variable. ROA is collected from WRDS and annual reports. The definition and calculation for ROA are summarized in table 2.

4.2.2 Independent Variable

The independent variables of this study are relations-oriented board diversity. As illustrated in chapter 2.1, board diversity can be defined as the degree of heterogeneity within characteristics of board members (Miller and Triana, 2009). Diversity can be measured by using Blau’s Index (1977), calculated as B = [1 - Σ(pi)2], where P is the proportion of individuals in a category and i is the number of categories (Richard, Barnett, Dwyer and Chadwick, 2004). This has been suggested by Harrison and Klein (2007) as an optimal measure for diversity, as it meets multiple important criteria:

1. Zero point which is complete homogeneity, whereas an increase in numbers indicates an increase in diversity.

2. The index has no negative value

3. The index is not unbounded, therefore limited

4. The index has a track record of being frequently used and noted as a proper measurement for diversity and it is seen as the standard measurement of board diversity (Carpenter, 2002; Finkelstein & Hambrick, 1996; Nielsen, 2010).

Gender, Nationality and age have been retrieved with BoardEx and Annual Reports (table 2). In addition, for age diversity, I adopt the categorization used by Houle (1990), creating three different age groups; young, middle and old. Therefore, using the calculation of Blau, age has three categories (i =3). Gender diversity is calculated by using only two categories (i=2); male and female. Finally, nationality diversity has been calculated by capturing the dispersion of executive board members, where p is the percentage of members in the ith_{group (Nielsen, 2010).}

(25)

25 | T h e s i s O . P . J . N i j h u i s 4.2.3 Moderating Variable

For the moderating variable, the degree of internationalization, Sullivan (1994)’s composite measurement has been used. Sullivan’s composite measurement has already been selected in multiple studies as indication for internationalization which measures internationalization on three important and theoretically distinct dimensions (e.g. Sanders & Carpenter, 1998; Lu & Beamish, 2004). The first dimension is foreign sales, or the ratio of foreign sales to total sales. This reflects the MNE’s dependence on sales to foreign markets. The second dimension is foreign production, or the reliance on owned foreign stocks, measured by foreign assets expressed as a percentage of total assets. The final dimension is geographic dispersion, i.e. the number of countries in which a firm has subsidiaries (Sullivan, 1994). The theoretical range for each dimension is from 0 to 1. The three different dimensions (foreign sales, foreign production, and geographic dispersion) can be summed to form the composite measure of degree of internationalization, which has a validated range of 0 to 3 (Lu & Beamish, 2004). All three dimensions are retrieved from WRDS and annual reports (table 2).

4.2.4 Control Variable

The control variables used in this study are the size of a firm, age and industry effects, as these are related to firm performance more experience and resources (Deephouse & Carter, 2005; Miller & Triana, 2009).

Firm size. Firm size is commonly used as a control variable, due to its effect on firm performance. Since internationalization requires the ability to deal with uncertainty, firm size is an important control variable. Large firms are more likely to survive and sustain uncertain environments than smaller firms, but smaller firms tend to be more flexible and better able to cope with uncertainties (Tihanyi et al., 2000). For firm size, annual revenues and market capitalization (in Billion USD) have been used as measurement (Miller & Triana, 2009; Carter et al., 2010). Both WRDS as Annual reports are used to collect this data.

Firm age. Firm age has been selected as control variable, because older firms tend to be less flexible, and therefore could perform less in an uncertain and rapidly changing environment (Yli‐Renko, Autio & Sapienza, 2001). On the other hand, younger firms are limited in their resources, making them prone to liabilities of newness and adolescence (Yli‐Renko et al., 2001). Age diversity is measured as the difference between the years since the firm was founded until the year of reference for this study, 2015, is used as a measurement (Miller & Triana, 2009). For collecting this data, WRDS and annual reports are used.

Industry effects. The measurement of firm performance, ROA is affected by two aspects: operating leverage and product-life-cycle (Selling & Stickney (1989). Per industry, these phenomena differ. For example, when examining at the operating leverage, firms operate with different mixes of

(26)

fixed and variable costs. Differences between firms and industries in their ROAs might be related to economies and diseconomies of scale from operating leverage and to movements through product life cycles (Selling & Stickney, 1989), resulting in differences in ROA between industries (Boter & Holmquist, 1996). Therefore, industry effects need to be taken into account. The WRDS provides a GICS industry code to each firm, classifying companies within one of the 10 sectors, which makes it possible to separate industries. The industries as classified by GICS are presented in table 1. Each variable from this thesis is presented with a short summary, its’ source and measurement in table 2.

Table 2: Summary Variables

Variable Description Source

Firm Performance Return on Assets, calculated as net income + (1

- tax rate) (interest expense) / average total assets. ROA is often used as a predictor of firm performance.

Wharton Research Data Services; Annual aeports

Age diversity Difference in age distribution amongst board

members. Calculated by using Blau’s Index, where three age groups are categorized.

BoardEx; Annual reports, Houle (1990)

Gender diversity (Increasing) the number of female executive

board members within an executive board. Blau’s Index can be used to calculate gender diversity.

BoardEx; Annual reports

Nationality diversity The amount of executive board members with

different nationalities in a board. Blau’s Index is used to calculate the Diversity number.

BoardEx; Annual reports

Internationalization Increasing the share of transnational exchange

relative to domestic ones, by expanding the flow of goods, services, and people cross state boundaries. internationalization is calculated by using Sullivan’s composite measurement

Wharton Research Data Services; Annual reports

Firm Size Firm Revenues and Market Capitalization Wharton Research Data Services;

Annual reports

Firm Age Difference between year of study and the year

the firm was founded.

Wharton Research Data Services; Annual reports

Industry The industry sector is assigned to a company

whose definition most

closely describes the business activities that generate the majority of the firm’s revenues

Wharton Research Data Services;

Global Industry Classification

(27)

4.3 Method and model Specification

In this section, an overview is given of the method and model specification to analyze the hypotheses. The proposed hypotheses are tested with a multiple linear regression analysis. According to Field (2009), this type of analysis is a proper and valid way to predict the influence of one independent variables on one dependent variable. In order to test the first three hypothesis (i.e. the effect of age -, gender - and nationality diversity on firm performance) and to test the relationship between internationalization and firm performance to see whether a moderation effect would be visible, the following equation is used:

𝑌it= 𝑎 + 𝑠it𝛿 + 𝑧it𝛽 + 𝜀it

In this equation, Y represents the firm performance, followed by a subscripts i and t, which stand for the firm and the specific year respectively. Going through the rest of the equation, 𝑎 is the constant or intercept, the s for the independent variable (i.e. age diversity, gender diversity or nationality diversity) and the coefficient δ is added to consider the effect of the independent variable on the dependent variable, or firm performance. Here, δ is the number of steps by which firm performance increases, if the independent variable increases one step. The final two in the equation are the vector z, which measures the control variables and the 𝜀 will calculate for the error term of the regression (Field, 2009).

For the analysis of the moderation effect as is hypothesized for the final three hypotheses, a different equation is used. The equation including the moderating effect of internationalization is:

𝑌it= 𝑎 + 𝑠it𝛿 + 𝑧it𝛽 + 𝑒it𝜃 + 𝑒it𝑠it𝜇 + 𝜀it

The variables in this equation are similar to the variables in the previous equation, but a few variables are added. In this equation, 𝑒 is the direct effect of internationalization on firm performance and as most important new variable to this equation is, 𝑒it𝑠it which is the moderating effect of internationalization on the independent variable.

In model 1, 2 and 3, the first three hypothesis (H1, H2, H3) are tested. In addition, model 4 tests the direct effect of internationalization on firm performance. In Model 5, 6 and 7, the three hypotheses regarding moderating effect of internationalization are tested (H4, H5 and H6). As explained above, ROAs could differ per industry (Selling & Stickney, 1989). Therefore, besides a general analysis, results are likely to differ when examining industries separately. By selecting only one industry sector and repeating the analysis, industry effects can be seen (Selling & Stickeny, 1989). Therefore, each model consists of 11 steps. The first step is an analysis with all MNEs and industries together, combining all

The influence of Executive Board Diversity on firm performance : the moderating effect of the degree of internationalization