J. V OCHTELOO T HE EFFECT OF BOARD NATIONAL DIVERSITY ON FIRM INTERNATIONALIZATION AND THE MODERATING EFFECTS OF GENDER DIVERSITY AND CEO TENURE

(1)

T

HE EFFECT OF BOARD NATIONAL DIVERSITY ON

FIRM INTERNATIONALIZATION AND THE

MODERATING EFFECTS OF GENDER DIVERSITY

AND CEO TENURE

J. V

OCHTELOO

Supervisor: Dr. E. Mendiratta Co-assessor: Dr. K. van Veen Student number: S2365332

Date: 18-6-2018

(2)

Abstract

Globalization has changed the business landscape dramatically, and companies need to expand to international markets to remain successful. Executives consider internationalization to be complicated and as a result, the question of how to deal with this becomes crucial. Despite the vital role of the board of directors on key organizational outcomes, their influence on the firms’ degree of international diversification has been researched less. In the last years, most discussions were related to composition characteristics of the board of directors,

resulting in much research on directors’ characteristics such as age, ethnicity, tenure,

educational and functional background on firm internationalization but less towards national diversity within a board. Nationality diversity is important because it has been researched less in comparison with other executives’ characteristics, but at the same time, the level of

national diversity itself has increased amongst board members in the last years. Therefore, this research aims to determine the effect of board national diversity on the degree of firm

internationalization. Furthermore, it contributes to the literature by looking at two moderating effects, board gender diversity as well as CEO tenure, which has not been used as moderators that often. The theoretical framework relies on the Upper Echelon Theory to explain the differences in the degree of internationalization of 196 companies classified in the S&P 500. The data of the years 2012 to 2015 is used and is sourced from Orbis, Datastream database, BoardEx and annual reports. The hypotheses formulated were tested using Fixed Effect Panel Data analysis. Even though the data showed insignificant results and no claim could be made based on this sample, it contributes to the understanding and could function as an extension of prior research of national diversity and the degree of firm internationalization.

Keywords: Upper Echelon Theory, Internationalization, Board National Diversity, Board

(3)

List of figures and tables

Figure 1: Conceptual model

Figure 2: Graphical analysis of the normality; Pnorm Figure 3: Graphical analysis of the normality; Qnorm

Table 1: Summary of the variables, description, and source Table 2: Descriptive statistics of the sample

Table 3: Results of regression analysis (part 1) Table 4: Results of regression analysis (part 2) Table 5: Results of robustness check

(5)

1

1. Introduction

In our fast pacing modern economy, internationalization has never been more important for companies. Globalization has changed the business landscape dramatically, and companies need to expand to international markets to remain successful (Caligiuri, et al., 2004). The international expansion enables an organization to obtain economies of scale, utilize the best technology available and increase the knowledge pool by hiring talent worldwide (Bartlett & Ghoshal, 1992). Executives consider internationalization activities across national borders to be complex (Tihanyi, et al., 2000; Oxelheim, et al., 2013). As a result, the question of how to deal with this uncertainty and complexity related to international expansion becomes crucial. This complexity may be due to the opportunities, as well as threats, that can be encountered internationally. However, it is certain that decisions related to international diversification have an impact on organizational outcomes (Tihanyi, et al., 2000).

One way to successfully address the complexity of international expansion is by using the experience of the board of directors (Sanders & Carpenter, 1998), who make the decisions related to international expansion and are the key drivers behind the success of international diversification (Lee & Park, 2006). The board of directors, selected by the shareholders of the firm, set the strategy of the company and functions as a monitoring and governance

mechanism, thereby guiding and authorizing the companies’ strategic decisions, which include M&As, alliances, as well as selecting and replacing the CEO (Terjesen, et al., 2016). These powerful individual actors influence strategic choices because of, among others, their diversity of characteristics, such as values, personalities, nationality as well as past

experiences (Carpenter & Fredrickson, 2001). This argument is based on the Upper Echelon Theory (UET), which states that organizational outcomes such as strategic choices and performance levels are partially predicted by powerful actors in the organization, thus a reflection of their values and cognitive bases (Hambrick & Mason, 1984).

(6)

2 background on firm internationalization (Rivas, 2012), but less towards national diversity within a board. Even though the diversity of the backgrounds of the members (e.g.,

nationality) should be a reasonable proxy for the international orientation of the company’s board (Caligiuri, et al., 2004), it has been researched significantly less. Therefore, one of the contributions of this research is to add more understanding whether, and if so, how, board national diversity matters for successful internationalization.

Furthermore, Hitt, et al. (2006) argued that most research that focused on the effect of

international diversification is related to a narrow set of constructs and less attention has been given towards moderators of international diversification. One characteristic which has not been researched a lot as a moderator is gender diversity. Studying this characteristic is

important because countries have set up quotas to promote gender diversity. In 2003, Norway introduced a law that required companies to have at least 40% of the boardroom positions to be held by women. Many countries followed this initiative. In 2016, Deloitte published a report that highlighted the progress that has been made for women and minorities on

corporate boards in America. They concluded that there has been a slow change of diversity in the boardroom and that women and minorities remain underrepresented in the boardroom. In 2016, the representation of women and minorities in the Fortune 500 is at 30.8%, an increase of 0.7 in 4 years’ time (Deloitte, 2016). The underlying idea in this research is that individuals’ characteristics influence the strategic decision making of the company.

Therefore, it is interesting to research if adding more women to the board has any impact on internationalization, especially as a moderator which can alter the relationship between board national diversity and internationalization. This can function as a contribution to the

discussions related to gender diversity in the boards.

The second board characteristic which has been researched less is CEO tenure and will serve as the second moderator of this research. The board of directors plays an important role in influencing firm internationalization, but their main role is monitoring and strategy involvement, not implementation. The CEO, as the head of the company, coordinates the formulation and implementation of the strategy, including the choices of which markets to emphasize (Canella, et al., 2008). The CEO’s opinion has a significant impact on the

implementation of the internationalization strategy, especially when CEOs are longer-tenured. The reason for this is that literature shows that because of the embeddedness and

(7)

3 that are divergent of the current strategy and status quo (Simsek, 2007; Barroso, et al., 2011). This can, as a result, alter any benefits of nationality diversity in the presence of a longer tenured CEO. Therefore, the moderating effect of CEO tenure on the relationship between board national diversity and internationalization will be another contribution of this research.

By using a sample of 196 companies from the S&P 500 over the period of 2012-2015, this research studies the effect of a different type of diversity (nationality) in the board of

directors, as well as adding two moderating effects of board gender diversity and CEO tenure, thereby contributing to the calls for the study of different types of diversity in the groups (Rivas, 2012) as well as more attention towards moderators of international diversification (Hitt, et al., 2006). This leads to the following research question:

“Does board national diversity affect the degree of firm internationalization, and what is the moderating effect of having more women on the board, as well as CEO tenure?”

The data was sourced from Orbis, Datastream, BoardEx and annual reports. The formulated hypotheses were tested using Panel Data Fixed Effect regression analysis to see if there is a relationship, and if so, to what extent. The paper has been structured as follows; it first presents the literature review and hypotheses. After that, the methodology will be described, consisting of the sample, data collection, measurement of variables and the empirical

(8)

4

2. Theoretical framework and hypotheses

This section will introduce a review of the available literature on this topic. First, firm internationalization will be explained. Second, the Upper Echelon Theory will be discussed. Third, the relationship between board national diversity and internationalization is

highlighted. Then, the two moderating effects (board gender diversity, followed by CEO tenure) on the relationship will be discussed. Finally, this section will end with a graphical explanation of the conceptual framework.

2.1 Internationalization

Internationalization is a process where a firm increases its involvement in international operations (Welch & Luostarinen, 1988) and starts to rely more on foreign markets for their up and downstream factors (Bartlett & Ghoshal, 1998). It often evolves as a sequence of steps, gradually increasing its international involvement from export towards the creation of final units in a certain country (Johanson & Vahlne, 1977; Melin, 1992). International expansion is becoming more important for companies to survive. Companies that do not internationalize are at risk of reducing their competitiveness (Caligiuri, et al., 2004; George, et al., 2005). Determining a company’s degree of internationalization is a way the stakeholders use to assess a firm’s overall performance and future competitiveness (Ramaswamy, et al., 1996). In most industries, the major companies have realized significant degrees of

internationalization (Mattson, 1998), but also for smaller companies it can be important to have a clear growth strategy to be able to internationalize, especially for firms whose home market is limited. However, executives consider internationalization activities across national borders to be complex (Tihanyi, et al., 2000; Oxelheim, et al., 2013). This complexity may be due to the opportunities, as well as threats, that can be encountered internationally. Many scholars found that international expansion affects firms. However, a consensus on the direction has not yet been found (Hitt, et al., 1997).

An extensive body of research suggests significant benefits when a firm diversifies

(9)

5 International expansion, however, also has some risks. Increasing the degree of

internationalization can result in organizational problems (Tihanyi, et al., 2000) because if the company grows, this can lead to higher monitoring costs in comparison with domestic firms (Geringer, et al., 1989). Another important cost of internationalization is the liability of

foreignness and newness (Zaheer & Mosakowski, 1997) or liability of outsidership (Johanson, et al., 2014). The latter one, the liability of outsidership, occurs when a firm tries to enter a market abroad where it has no relevant networks. Because the company is also new, the liability of foreignness complicates the process of being an outsider.

As a result, internationalization can be a complex process but it is certain that decisions related to international diversification have an impact on organizational outcomes (Tihanyi, et al., 2000). These decisions are made by top executives of the company, who are the key driver behind the success of international diversification (Lee & Park, 2006). Because of the fact that the composition of executive teams changes often, different outcomes can be expected. These powerful individual actors influence strategic choices because of their diversity in, among others, characteristics such as values, personalities, nationality as well as past experiences (Carpenter & Fredrickson, 2001). This argument is based on the Upper Echelon Theory (UET), which states that organizational outcomes such as strategic choices and performance levels are partially predicted by powerful actors in the organization, thus a reflection of the values and cognitive bases (Hambrick & Mason, 1984). The next section elaborates more on the UET.

2.2 Upper Echelon Theory

(10)

6 Even though there are scholars that do not agree with the Upper Echelon Theory (Pettigrew, 1992; Lawrence, 1997), there is an abundance of research that has relied primarily on Top Management Teams (TMTs) demographic characteristics that have led to results that were significant (Tihanyi, et al., 2000; Lee & Park, 2006). Therefore, the Upper Echelon’s basic premise that managerial background characteristics are important for organizational outcomes will be used while examining the relationship between board national diversity and firm internationalization.

The UET originally focused only on TMTs, but recent research has applied and extended the UET to the board of directors to study the relationships between board diversity and

(11)

7 Due to all the complexities and risks related to internationalization, more diversity at the board and TMT is required. The internationalization of the firm’s operations increases the information asymmetry between TMT members and the board, which lead to higher monitoring costs (Oxelheim, et al., 2013). This is further driven by the need to coordinate resources from a variety of cultural, institutional and competitive environments (Sanders & Carpenter, 1998). Therefore, a broad range of skills is needed to succeed (Mishra &

Jhunjhunwala, 2013) and a diverse board could help here. Mishra & Jhunjhunwala (2013) define board diversity as the heterogeneous composition of the board on, among others, age, education, race, experience, culture, values, and gender. Boards must understand the global environment as well as the local environment where they operate, requiring a transnational mindset. Furthermore, having diversity on the board could serve as a signal to stakeholders of the company that it makes a serious effort to draw candidates from larger resource pools, thereby, trying to diminish the level of uncertainty that the international arena poses and improving decision making capability (Rivas, 2012).

Scholars researching board diversity have primarily focussed on the link between gender, age, ethnicity, tenure, educational and functional background on firm internationalization (Rivas, 2012). Even though the diversity of the backgrounds of the members (e.g. nationality) should be a reasonable proxy for the international orientation of the company’s board (Caligiuri, et al., 2004), it has been researched significantly less. Research showed that the nationality of an executive accounted for far more variation in strategic actions of a firm than other executives’ characteristics (Laurent, 1983) and the social values are among the most influential ones that affect strategic leadership and decision making (Finkelstein & Hambrick, 1996). Therefore, the following section will focus on illustrating the potential effects of the nationality diversity of the board of directors on firm internationalization.

2.3 Board national diversity and internationalization

National diversity has been researched less in comparison with other executives’

(12)

8 is detached from its home country (Ruigrok & van Tulder, 1995), as well as a signal how transnational the firm is (Bartlett & Ghoshal, 1998). Nowadays, to succeed in the global environment, board members need more specific skills. One of those skills is networking. Board members from a different nationality can have different (foreign) networks, which can bring knowledge and additional networks towards the firm. This can lead to better operating and investment decisions, acquiring foreign resources and opening up to new opportunities (Oxelheim, et al., 2013). Furthermore, the board of directors provides connections with other firms, as well as the resources needed to support the internationalization process (Pfeffer & Salancik, 1978). This is not only true for networks but also for knowledge. A higher level of nationality diversity is assumed to be an important requirement for the quality of strategic decision making. Strategic decision making goes hand in hand with complexity and uncertainty. Having a diverse board regarding nationality is likely to improve the

comprehensiveness and quality of strategic decisions, increased creativity and more dynamic problem-solving (Nielsen & Nielsen, 2011). Next, to knowledge, the interpretation (regarding decision making) and integration of information and knowledge is another necessary skill. Recent research showed that having diversity in a team helps integrating information to make better-informed decisions (Ali, et al., 2014) and leads to better problem-solving skills because they have more information available (Kaczmarek & Ruigrok, 2013), contributing to more input for discussions. Furthermore, research from Nielsen & Huse (2010), consisting of a survey with 200 Norwegian firms, showed that having more diversity in the board has a positive effect on essential board processes (leading to more strategic and operating control) as well as board effectiveness (it reduces the level of conﬂicts and ensures high quality of board development activities). Research from Carpenter & Fredrickson (2001) showed when firms have diverse boards (regarding the heterogeneity of their backgrounds), they were most likely to be more global and have a higher capability to internationalize, which consequently results in more internationalization. Research from Lee & Park (2006) concluded that top teams and boards with greater international exposure diversity have higher levels of internationalization.

(13)

9 integration is reached, and groups can obtain beneficial results from the variety of

perspectives (Hambrick, 1994). Furthermore, as argued already by the UET, board members bring individual interests and commitments which can lead to different ideas about proposed strategic changes, which can be a source of potential conflicts. However, though national diverse teams may experience more conflict, it is still positively linked towards the overall functioning of the group (Elron, 1997), thereby positively contributing to the boards’ two primary responsibilities (monitoring and strategy involvement). In the end, having national diversity in the board can be a source to help companies overcome the mentioned challenges related to internationalization.

The arguments mentioned above lead to the following hypothesis:

Hypothesis 1: Board national diversity is positively related to the degree of internationalization of a company

2.4 The moderating effect of board gender diversity

(14)

10 making. Talents and competencies are not evenly distributed between men and women. As a result, when a specific part of the society is excluded from participating in boards, not because of their skills but because of gender, the board is functioning sub-optimaly (Grosvold, et al., 2007).

A number of empirical studies found that females have a fresh view on difficult subjects (Francoeur, et al., 2008), mainly because women have professional experiences which are from non-traditional backgrounds (Singh, et al., 2008). Furthermore, women possess more advanced diplomas (Hillman, et al., 2002), which can enhance the effectiveness of the boards of directors (Terjesen, et al., 2016). This could mean that by adding (more) women to the board, new internationalization possibilities can be brought in and taken into consideration, because they approach things differently due to their backgrounds. This can strenghten the positive effects related to knowledge which may be present in the board due to the nationality diversity.

Higher levels of gender diversity will increase the creativity and therefore, also problem-solving capabilities (Richard & Shelor, 2002), as well as reduced information processing costs of globalization (Luo, 2005). This is because a more diverse board can better deal with

foreign liability problems and have a larger processing capacity. Besides that, females tend to be more patient and open-minded, which makes communication more accessible and helps to diffuse information better (Gul, et al., 2011), for example from the board to the TMT, as well as internally in the board. Moreover, scholars found that female directors are more likely to have managerial skills and hold positions related to the soft managerial issues rather than operational and marketing functions (Zelechowski & Bilimoria, 2006). These skills can help with thinking outside the box and approaching problems from a different angle, thereby complementing the board’s current pool of skills, which is needed to coordinate resources from a variety of cultural, institutional and competitive environments (Sanders & Carpenter, 1998).

(15)

11 are exposed. Having people from different backgrounds can add more and unique

information, skills, as well as networks to the current knowledge pool. By increasing gender diversity, it may facilitate the development of new relation-based resources by providing access to previously untapped networks which may help adapting to new international markets (Thomas, et al., 2012) and intensifying the effects of nationality diversity on internationalization.

The arguments mentioned above lead to the following hypothesis:

Hypothesis 2: Board gender diversity positively moderates the relationship between board national diversity and the degree of internationalization. Board national diversity will have a stronger positive impact on the degree of internationalization when there are more women on the board.

2.5 The moderating effect of CEO tenure

The previous discussion showed that the board of directors play an important role in influencing firm internationalization. However, their primary functions are monitoring and strategy involvement, not implementation. The CEO, as the most comprehensive decider of important strategies within a company, coordinates the formulation and implementation of the strategy, including the choices of which markets to emphasize (Canella, et al., 2008).

Therefore, the CEO has a significant impact on the implementation and can alter the strength of the relationship between board national diversity and internationalization.

(16)

12 (2006) noted that CEOs become more conservative as their tenure increases, even lessening the influences of other CEO variables that might encourage a more liberal attitude. This was the same conclusion in the paper by Hambrick & Fukutomi (1991), who researched that the longer the CEO is in place, the more different their perceptions to change will be. In their paper, Hambrick & Fukutomi created a model consisting of five seasons (stages), which showed that there are specific stages in a CEO’s tenure which results in different executive attention, behavior, and actions. In the first years as a CEO, they rely on their previous experience, their study, and functional backgrounds. However, this may become less strong the longer he/she remains a CEO. A potential explanation is that long-tenured CEOs (I) become committed to their way of running the company, and therefore, may listen less to others’ opinions, and (II) build up more resistance for change (Miller, 1991). Furthermore, as the length of the CEO’s tenure increases, the CEO will start relying more on familiar

information sources and gradually restricts his/her information processing. A reason for this could be that CEOs that are longer tenured may operate from the basis of routines that are built up over time which can result in the fact that their company knowledge can become a less valuable resource (Barroso, et al., 2011). So, when suggestions are made which do not fit with the routines, it could lead to CEOs doing nothing with this information in order to avoid potential risks, because longer-tenured CEOs narrow down their perspectives and become less open-minded (Hambrick & Fukutomi, 1991).

In summary, because of the embeddedness and entrenchment of a longer tenured CEO, it can influence how likely he/she listens to opinions that are divergent of the current strategy and status quo. In section 2.3 (hypothesis 1), it has been argued that different skills and opinions (because of heterogeneity of nationalities) could lead to more internationalization. However, when the CEO is not listening to this input and opinions that could be divergent to his/her way of thinking, any benefits of nationality diversity might become lower in the presence of a longer tenured CEO.

Therefore, the arguments mentioned above lead to the following hypothesis:

(17)

13 2.6 Conceptual framework

Figure 1: Conceptual model

3. Methodology

This section will introduce the methodology, consisting of the sample, data collection, measurement of variables including the assumptions, as well as the empirical approach used.

3.1 Sample and data collection

This research aims to examine the influence of board national diversity on the degree of internationalization. Theoretically, an explanation is given as to what the impact of board national diversity is on the degree of internationalization, as well as the two moderating effects. This will be tested via the gathered data to see what, if any, correlation can be derived. The sample consists of companies classified in the Standard & Poor’s 500 index (S&P 500), which includes the 500 largest companies with a market capitalization of at least $6.1 billion. These firms are all publicly listed, which means that the information about the board is widely available. The data for the years 2012-2015 will be used. One of the reasons to choose

America, and specifically the S&P 500, is that a lot of the world's economies like Japan and Germany have slipped down global ranking (Martin, 2017). However, America is still in the top 3 with $34.102 trillion. Another reason is the fact that American samples are more likely to generate results in support of UET, mainly because top executives in the US have more discretion in comparison with their counterparts in other advanced economies due to the formal institutions that allow and encourage these top executives such as the CEO, to take any action appropriate in order to reach success (Hambrick, 2007). Furthermore, American

companies are required to provide annual reports which result in data availability (Miller &

(18)

14 Triana, 2009). Finally, multiple studies have used this sample for diversity studies, which validates the data selection (Adams & Ferreira, 2009; Simpson, et al., 2010).

The data was collected via multiple databases. One of the primary sources was ‘Datastream’, a database by Thomson Reuters, which contains information on financial markets, companies, international trade and national economies as far back as 1970. The second primary source was the BoardEx database, which includes extensive biographical and relationship data of more than 17.000 companies on the boards of notable public and private companies globally since 1999. In case of missing data, it was completed via researching annual reports and the Orbis database. This database, provided by Bureau van Dijk, contains information on more than 275 million private companies worldwide and gives company information to make easy cross-border comparisons (Bureau van Dijk, 2017). To merge the data from the different databases, company identifiers such as ISIN (International Securities Identification Number) were used. After removing companies with missing data, the final sample consists of 196 companies. This resulted in 784 observations (4 years, 196 companies). It consisted of a proper distribution of all the different industries which allows variability regarding the degree of firm internationalization based on previous studies (Sullivan, 1994).

3.2 Variables Dependent variable

The dependent variable in this research is the degree of internationalization. Many studies were completed about internationalization. One of the first researchers covering this topic was Sullivan in 1994. He pointed out that the degree of internationalization has three attributes: performance (what goes on overseas), structural (what resources are overseas), and attitudinal (what is top management's international orientation). Scholars (e.g., Ruigrok & van Tulder, 1995) have used several measurements of this, including the Transnationality Index (TNI), created by the United Nationals Conference on Trade and Development (UNCTAD). This measurement makes use of different ratios to measure the degree of internationalization. It consists of the following three parts:

(19)

15 FSTS is a famous definition in the internationalization research because it reflects dependence on international markets (Barroso, et al., 2011). The data for the FSTS and FATA was

gathered via Datastream. Unfortunately, with the available databases, annual reports, and company websites, the information for the third part of the TNI was not fully available. In recent research (e.g. Kaczmarek & Ruigrok, 2013; Nielsen & Nielsen, 2013), and as

suggested by Sullivan (1994), the degree of internationalization was interpreted as the average of (1) FSTS, (2) FATA, and (3) geographic dispersion of foreign sales (sometimes used as the number of employees abroad). Because the data for the third part of the TNI (either number of employees abroad / total employees, or geographic dispersion of foreign sales) was missing, in this dissertation the following build up will be used:

Model 1: regression with the sales part (FSTS) Model 2: regression with the assets part (FATA)

Model 3: regression with the growth rate of the sales part (RFSTS) Model 4: regression with the growth rate of the assets part (RFATA)

The difference between the first two and the last two models is the ratio component. Model 1 and model 2 will use the separate two (available) parts of the TNI to measure

internationalization. Most studies have relied on measures such as FSTS to characterize the internationalization construct (Ramaswamy, et al., 1996; Barroso, et al., 2011). Model 3 and model 4 used growth rate between two years as in; (new year – old year) / old year. By doing this, the change in internationalization can be better allocated to the decisions from last year made by the board of directors.

For the robustness check, a different operationalization will be used. Scholars researching the internationalization of a company suggested using a multidimensional measure to improve validity (Sullivan, 1994; Lee & Park, 2006). Therefore, as a robustness check, firm

(20)

16 Independent variable

The independent variable is board national diversity (NAT). Nationality can mean the country of birth or where a person has lived most of his/her life. In this research, board national diversity is defined as the number of foreign nationalities to the total number of nationalities of the board (Caligiuri, et al., 2004) based on their passport and is measured via the Blau Index of heterogeneity, as already used by earlier studies (e.g. Lee & Park, 2006; Nielsen & Nielsen, 2013), and suggested by van Veen & Marsman (2008). This Blau Index is measured with the following formula (see formula 1):

𝐵𝑙𝑎𝑢 𝐼𝑛𝑑𝑒𝑥 = [1 − ∑(𝑝_𝑖)2_{] (1)}

Where ‘𝑝𝑖’ is the percentage of board members in the ith category (nationality). The functional categories are (I) domestic and (II) all other nationalities. In total, this sample consisted of 122 unique nationalities. The Blau Index ranges from 0 (homogenous) to 1 (heterogenous), where 0,5 represents an equal amount of nationalities and the higher the value, the higher the degree of national diversity will be in that team. In the dataset, per company, the unique board members have been selected per year and duplicates have been dropped. The information is gathered via the BoardEx database, annual reports and Orbis database.

A second way of operationalizing the independent variable is by using the proportion of foreign nationalities in the team (PNAT) as in foreign unique nationalities / total board size, as used by Caligiuri, et al. (2004). This way of operationalizing will be used in the robustness check.

Moderators

(21)

17 Mínguez-Vera, 2008). This Blau Index is measured by the same formula as with the

independent variable (see above), where ‘𝑝_𝑖’ is the percentage of board members in every category (male/female). The Blau Index ranges from 0 to 1, where 0.5 represents an equal amount of man and women. The data for the second moderator, CEO Tenure (TENURE), is gathered via the Datastream database. It shows how long the CEO is in their position and is measured in years.

Control variables

There are several control variables in this research, which are the firm size, firm performance and firm age. Research from Finkelstein & Hambrick (1996) suggested that firm size and firm performance have impacts on various organizational outcomes, so these two variables are included as first controls.

(I) Firm size (logEMPL) has an impact on various organizational outcomes, as explained by Hambrick & Finkelstein (1997) and Lee & Park (2006). One of them is that larger firms may have more resources to internationalize. This variable is measured by the number of

employees and is the logarithm of firm employees, as used by Nielsen & Nielsen (2013). The reason to log the variable was that the variable was not normally distributed. Logging

variables has the characteristic that it changes it towards the ‘rate of change’, resulting in the fact that the distribution becomes more normal. Furthermore, employees have been chosen in this research as a characteristic of firm size instead of Total Assets, because Total Assets is closely linked to Return On Assets, which is the next control variable.

(II) Firm performance, measured as the Return on Assets (ROA), is relevant because decisions and strategies related to internationalization, as explained in the theoretical framework, are based on the search for economies of scale and scope (Bartlett & Ghoshal, 1992). As a result, it shows how efficient a company is using their assets for its’

internationalization strategies.

(22)

18 Below in table 1, a summary of the variables, description, and source can be found.

Variable Description Source

Degree of

Internationalization

FSTS (%) – foreign sales to total sales Datastream FATA (%) – foreign assets to total assets Datastream RFSTS (%) – growth ratio; foreign sales to total sales Datastream RFATA (%) – growth ratio; foreign assets to total assets Datastream Board Nationality

Diversity

NAT (Blau index) – foreign board members to total board size

BoardEx, annual reports PNAT (proportion of foreign nationalities) BoardEx,

annual reports Board Gender

Diversity

GEN (Blau index) – female to total board size Datastream

CEO Tenure “Time in role” (in years) BoardEx,

annual reports

C1; Firm size Number of employees (Logarithm) Datastream

C2; Firm age Difference between 2018 and incorporation year Datastream

C3; ROA ROA (Return on Assets) Datastream

Table 1: Summary of the variables, description, and source

All the models (model 1 until 6) have the following build up:

Model A; Dependent variable (Y) independent variable (𝑋₁) Control variables (C1, C2, C3) Model B; Add moderators 𝑀1 + 𝑀2 (or 𝑀3 and 𝑀4)

Model C; Add interaction effect 𝑋₁*𝑀₁ (or 𝑋₁*𝑀₃) Model D; Add interaction effect 𝑋₁*𝑀₂ (or 𝑋₁*𝑀₄) Model E; Complete model

3.3 Empirical approach

(23)

19 To investigate the relationship between nationality diversity within a company’s board of directors and the firm’s degree of internationalization, as well as the interaction effects of board gender diversity and CEO tenure, a decision needed to be made which specific panel data test was applicable. The regression model could be analyzed by using either a fixed or the random-effects model. The former one, fixed effects (FE), assumes that the firm-specific residual has a variance of zero, whereas the latter one, random-effects (RE), assumes that the firm-specific residual has a distribution with a variance of 𝜎2 (Lee & Park, 2006).

To identify which test is most appropriate the ‘Hausman test’ was used (Greene, 2012). Essentially, it tests if there is a correlation between the unique errors and the regressors. The null hypothesis is that the preferred model is random effects (so no correlation between the unique errors and the regressors). The alternate hypothesis is that the model is fixed effects. The Hausman test rejected the null hypothesis (Prob>chi2 = 0.0000), so the fixed effect is preferred. Another possibility would be to do a regular regression for each year by itself and see if each year generates similar results. However, this approach identifies differences

between firms in a given year, whereas the FE panel approach identifies differences within

firms over time. The rationale behind the FE is the assumption that something within the company may impact the dependent variable and one needs to control for this. FE removes this effect so one can assess the true effect of the predictors on the dependent variable. Looking at the assumptions for the data, some possible problems could arise. After checking for normality of the variables, one of them (firm size) was considered far from normally distributed. If one does not ‘log’ variables when necessary, the models will be biased (p-values, standard errors). Therefore, the logarithm was taken in order to control for that, as explained in control variables section. Furthermore, the normality of the residuals was checked via Qnorm and Pnorm, which turned out to be somewhat normally distributed (see Appendix A).

Before the analysis was done, panel data assumptions related to autocorrelation,

(24)

20 Inflation Factors (VIF) values after running a normal regression, as well as via a correlation table. Multicollinearity can negatively affect a conclusion because the 'actual' effect of two or more variables cannot be determined for sure. Doing this normal regression, and calculating the VIFs, no variables were higher than 10 (Hair Jr., et al., 1995) or the more conservative 5 (Rogerson, 2001). After the moderators were added, two variables had high VIFs (NAT 8.47 and M2 = GEN*NAT 7.16). By looking at the outcome of the regression, a possible

explanation of the high VIFs was that interaction terms are correlated with the main effect terms because they include the main effects terms. However, according to Allison (2018), it is not something to be concerned about because the p-value for the interaction term is not

affected by the multicollinearity. Furthermore, Brambor, et al. (2006) argued that high VIFs values (so multicollinearity) represent the fact that there is not enough information in the dataset to estimate the model parameters accurately. In the past, scholars (e.g. Allison, 2009) argued that centring the constitutive terms can solve this problem. However, other research (Brambor, et al., 2006; Kam & Franzese JR, 2003) rejected this because centring adds nothing important in statistical terms because subtracting a constant from each of the variables cannot affect a linear regression. Therefore, as a second check, a correlation table was used to check for high correlations. As can be seen in Appendix B, only the moderators are highly

correlated.

4. Empirical results

In this section, an elaboration on the descriptive statistics is given, followed by the regression results. After that, a robustness check will be presented.

4.1 Descriptive statistics

Table 2 shows the descriptive statistics of 196 S&P 500 companies, sampled over 4 years. In total there were 784 observations. The main industries present in the sample were the

Electronic & Electrical Equipment (29), Food Producers & Processors (11), Oil & Gas (12), and Software & Computer Services (14). These categories were made based on the SIC

(25)

21

Variable Name Observations Mean SD Min Max

FSTS 784 44.977 21.280 0.000 100.380 RFSTS 784 -0.105 2.698 -19.320 27.875 FATA 784 15.515 18.234 0.000 96.920 RFATA 784 0.031 0.766 -1.000 16.119 DOI 784 30.246 16.292 0.000 91.780 RDOI 783 1.038 33.573 -100.000 647.301 NAT 784 0.209 0.237 0.000 1.000 PNAT 784 11.587 14.022 0.000 75.000 TENURE 784 5.002 4.310 0.000 22.600 GEN 784 0.283 0.107 0.000 0.500 AGE 784 51.265 37.711 2.000 183.000 ROA 784 8.605 6.720 -61.350 32.530 logEMPL 784 10.018 1.268 5.451 12.995 M1 (TEN*NAT) 784 0.973 1.566 0.000 15.600 M2 (GEN*NAT) 784 0.060 0.073 0.000 0.377 M3 (TEN*PNAT) 784 53.381 88.377 0.000 780.000 M4 (GEN*PNAT) 784 3.344 4.364 0.000 24.000

Table 2: Descriptive statistics of the sample

The descriptive statistics, as presented above, showed the observations, mean, standard deviation, minimum and maximum for this sample. Looking at the first two

(26)

22 small proportion. The mean value of gender was considered low, indicating that there are fewer women than men on the board. The average company age is 51 years.

4.2 Regression results

As mentioned in the methodology section, a separation between four operationalizations of the dependent variable was made:

Model 1 = FSTS Model 2 = FATA

Model 3 = growth rate of FSTS Model 4 = growth rate of FATA

The build-up of the model was as follows: in the first model (A), the model estimated the effect of board nationality diversity and the control variables on the degree of

internationalization. After that, in model B, the two moderators were added. Furthermore, in model C and D, the interaction terms were added separately. Finally, the last model (E) contains the complete model.

In the first model, where FSTS was used as the variable to measure the degree of

internationalization, none of the models showed a significant result. Furthermore, looking at the final model (1E), only 1% of the amount of variance of Y was explained by the variables. Also, the F-test is insignificant (0.8332 > 0.05). The F-test checks whether all the coefficients in the model are different from zero, and gives an indication of how the overall model fits the data well. So, one can reject the null hypothesis of all coefficients being 0.

(27)

23 one, which is significant at the 10% level. The R-square improved as well, from 1.1% to 5.6%.

In the third model, using the growth rate, none of the models showed a significant result. Furthermore, looking at the final model (2E), only 2% of the amount of variance of Y explained by the variables and the F-test is insignificant (0.6966 > 0.05).

For the last model (4), using the growth ratio of FATA, only AGE is significant at the 10% level. The number has a positive indication, so the older the company, the higher the degree of internationalization, which is in line with expectations. Furthermore, only 1.7% of the amount of variance of Y explained by the variables, and the F-test is insignificant (0.1977 > 0.05).

Comparing the ‘A’ models (hypothesis 1), the results show that all outcomes are insignificant (model 1: β = -0.328; model 2: β = 0.571; model 3: β = 3.727; model 4: β = 0.460). Thus, the hypothesis that board national diversity leads to a higher degree of internationalization is rejected. Looking at the ‘E’ models, one can make a statement about the second and third hypotheses. Again, the results show that all outcomes are insignificant, thereby rejecting the second hypothesis that board gender diversity positively moderates the relationship between board national diversity and the degree of internationalization. Also, the third hypothesis, suggesting that CEO tenure negatively moderates the relationship between board national diversity and the degree of internationalization, has been rejected.

Table 3

Results of regression analysis (part 1)

FSTS FATA

(28)

24

Table 4

Results of regression analysis (part 2)

RFSTS (ratio/growth) RFATA (ratio/growth)

(3A) (3B) (3C) (3D) (3E) (4A) (4B) (4C) (4D) (4E) NAT 3.727 3.738 2.538 5.085 3.673 0.460 0.465 0.393 -0.226 -0.379 (1.29) (1.30) (0.90) (1.15) (0.82) (1.32) (1.34) (1.10) (-0.28) (-0.45) logEMPL 1.500 1.484 1.538 1.501 1.548 -0.082 -0.085 -0.082 -0.094 -0.089 (1.00) (0.98) (1.01) (0.99) (1.02) (-0.43) (-0.43) (-0.41) (-0.46) (-0.44) AGE -0.129 -0.094 -0.099 -0.094 -0.099 -0.000 0.025* 0.025* 0.025* 0.025* (-1.38) (-0.89) (-0.93) (-0.89) (-0.92) (-0.00) (1.74) (1.73) (1.73) (1.70) ROA -0.020 -0.022 -0.020 -0.019 -0.019 0.004 0.003 0.003 0.002 0.002 (-0.62) (-0.65) (-0.61) (-0.60) (-0.57) (0.62) (0.48) (0.50) (0.33) (0.35) TENURE 0.015 -0.020 0.017 -0.017 0.008 0.005 0.007 0.003 (0.49) (-0.65) (0.53) (-0.54) (0.89) (0.51) (0.82) (0.32) GEN -2.043 -2.127 -1.005 -1.303 -1.473 -1.478 -2.006 -2.039 (-0.90) (-0.95) (-0.38) (-0.50) (-1.43) (-1.44) (-1.28) (-1.30) M1 (ten*nat) 0.209 0.197 0.013 0.021 (1.28) (1.15) (0.54) (1.07) M2 (gen*nat) -5.007 -3.946 2.570 2.685 (-0.65) (-0.52) (0.92) (0.96) _cons -9.105 -10.237 -10.269 -10.712 -10.641 0.719 -0.163 -0.165 0.081 0.089 (-0.64) (-0.72) (-0.72) (-0.74) (-0.73) (0.27) (-0.07) (-0.07) (0.03) (0.04) N 784 784 784 784 784 784 784 784 784 784 R-Square 0.015 0.017 0.019 0.018 0.020 0.002 0.014 0.014 0.017 0.017 Adj.R-Square 0.01 0.01 0.01 0.01 0.01 0.00 0.01 0.01 0.01 0.01 F-statistic 1.02 0.82 0.79 0.72 0.69 1.64 1.73 1.60 1.44 1.40 Prob > F 0.3977 0.5544 0.5978 0.6539 0.6966 0.1653 0.1166 0.1379 0.1930 0.1977 t statistics in parentheses * p < 0.1, ** p < 0.05, *** p < 0.01 4.3 Robustness check

Robustness checks are a way to examine if the core regression coefficients change or not when some regressors are added or removed. If the coefficients do not significantly change, there is evidence of structural validity (Lu & White, 2014). In order to do this, a different operationalization of the dependent variable degree of internationalization was used. In the main four models (model 1 to model 4), this variable was measured as FSTS, FATA, RFSTS, and RFSTS. Some researchers have used either the Transnationality Index (TNI), which consists of an average of the FATA, FSTS, and number of employees abroad/total employees. Other researchers have substituted the last variable (employees) for the geographic dispersion of foreign sales. Because the data for the third part (either number of employees abroad / total employees, or geographic dispersion of foreign sales) was missing, for the robustness check, the average of only the first two (FSTS + FATA, and RFST + RFATA) was used.

(29)

25

Table 5 Results of robustness check

DOI (FSTS / FATA) RDOI (ratio/growth = RFSTS / RFATA) (5A) (5B) (5C) (5D) (5E) (6A) (6B) (6C) (6D) (6E) PNAT 0.004 0.002 0.036 -0.024 0.016 0.315 0.323 0.308 0.297 0.275 (0.06) (0.04) (0.61) (-0.28) (0.19) (1.39) (1.40) (1.07) (0.46) (0.49) logEMPL 2.166 2.436 2.360 2.412 2.345 3.674 2.062 2.099 2.037 2.074 (0.68) (0.77) (0.75) (0.76) (0.75) (0.41) (0.23) (0.24) (0.23) (0.23) AGE -0.758*** -0.811*** -0.802*** -0.812*** -0.803*** 0.859 1.133 1.129 1.133 1.127 (-3.21) (-3.37) (-3.37) (-3.37) (-3.38) (0.86) (1.48) (1.44) (1.48) (1.44) ROA -0.119** -0.118** -0.120** -0.121** -0.122** 0.473 0.471 0.472 0.468 0.469 (-2.43) (-2.40) (-2.42) (-2.46) (-2.48) (1.40) (1.32) (1.31) (1.22) (1.21) TENURE -0.117 -0.062 -0.119 -0.065 0.688 0.662 0.687 0.657 (-1.51) (-0.75) (-1.53) (-0.79) (1.12) (0.77) (1.17) (0.83) GEN 2.895 3.026 1.779 2.241 -14.701 -14.762 -15.810 -16.064 (0.69) (0.73) (0.34) (0.42) (-0.30) (-0.31) (-0.21) (-0.22) M1 -0.323 -0.311 0.152 0.171 (-1.08) (-1.04) (0.09) (0.12) M2 5.374 3.759 5.331 6.218 (0.46) (0.32) (0.04) (0.05) _cons 48.375* 48.180* 48.097* 48.790* 48.527* -87.542 -84.833 -84.798 -84.214 -84.070 (1.92) (1.90) (1.90) (1.90) (1.88) (-0.97) (-0.99) (-0.99) (-0.94) (-0.93) N 784 784 784 784 784 783 783 783 783 783 R-Square 0.052 0.058 0.060 0.058 0.061 0.006 0.009 0.009 0.009 0.009 Adj.R-Square 0.05 0.05 0.05 0.05 0.05 0.00 0.00 0.00 0.00 -0.00 t statistics in parentheses * p < 0.05, ** p < 0.01, *** p < 0.01

Looking at the robustness check, the results are robust to using a slightly different dependent variable, one that captures the general idea of internationalization but as a different measure. As different models were utilized, the robustness of the results was checked if they hold or not. Looking at the first model of the robustness check (model 5), it is almost similar to model 2, where only AGE (1%) and ROA (5%) are significant. Those two coefficients did not change sign and magnitude significantly following the modifications. The second robustness model, model 6, differs a lot compared to the other models, so is less robust.

5. Discussion and conclusion

In the past, scholars have researched the relationship between board diversity and indices of internationalization (e.g. foreign sales ratio, foreign assets ratio, or a mixture). Previous research has focused more on board demographics such as international experience, tenure, and gender, whereas nationality diversity has received less attention. Therefore, by using the Upper Echelon Theory, the variable board nationality diversity present in the boards of 196 American companies, sampled over four years, on the firm’s degree of internationalization was tested.

(30)

26 a relatively similar sample, also consisting of the biggest companies in the USA, but instead of the S&P 500, they used the Fortune 500. In their research, with a sample size of 76 companies, they found that national diversity ratio was positively and significantly related to indices of internationalization. However, they only took it for one specific year (no panel data). In the research by Lee & Park (2006), they argued that more diverse boards are more likely to provide companies with the necessary capabilities for firm internationalization. They found that board or TMT international diversity, in general, has a significant relationship with firm internationalization. They researched 226 US firms during the period 1988–1994,

originating from a variety of industries to ensure generalizability. These 226 companies were classified in the ‘Ward’s 50,000 Largest Corporations’. Within this dissertation, the research showed that for those companies sampled from the S&P 500, there is no effect. Possibly the relationship between nationality diversity and internationalization is more prominent in companies that do not belong in the top of the S&P 500 but more in the middle.

The second hypothesis, which implies that board gender diversity positively moderates the relationship between board national diversity and the degree of internationalization, is also rejected because of insignificant results. In the literature section, it is described that board gender diversity positively moderates the relationship between board national diversity and firm internationalization. To the author’s knowledge, almost no research has been completed about the specific relationship between board gender diversity and internationalization where the board gender diversity serves as a moderator. Potential reasons why the results were insignificant could be related to the sample of this research, in specific the size of the

(31)

27 Hypothesis 3, which implies that CEO tenure negatively moderates the relationship between board national diversity and the degree of internationalization, is also not confirmed. Looking at the literature, it is less likely that the insignificant results of this hypothesis are a function of theoretical problems because most work looking at CEO tenure argues the same as

hypothesized in this research. So, a potential reason could be related to a method problem regarding operationalization. CEO tenure was operationalized as how long the CEO is in their position and is measured in years. Perhaps, one needs to consider the predecessor of the current CEO when making claims about the impact CEO tenure has on the relationship. For example, if the predecessor performed poorly, it can have a different effect than when he/she performed well (Miller, 1991), because the company needs to ‘recover’ which can delay the internationalization speed.

In summary, by comparing all the final models (1E, 2E, 3E and 4E) stated in the regression results section (4.2), there is no statistical evidence to make a claim about the created hypotheses for this sample, by using either FATA, FSTS, RFATA and RFSTS, that board national diversity has an impact on the degree of internationalization of a company and

respectively the moderating effects (board gender diversity as well as CEO tenure). Therefore, answering the overall research question based on this sample is not possible, due to the

insignificant results.

Even though the results were insignificant, it contributes to the understanding and could function as an extension of prior research of national diversity and the degree of firm internationalization. Another contribution is the addition of the two moderators on the

relationship between board nationality diversity and firm internationalization. By the author's knowledge, not much research used board gender diversity as a moderator. Even though the results were not significant, it provided scholars with substantially different and new

perspectives which could influence internationalization.

5.1 Limitations and future research

Like all studies, this research has left some unanswered questions that suggest possible input for future research. First, the sample for this study consists of US ﬁrms only, which makes it harder to generalize to Europe or other geographical areas. Therefore, a possible future study could look at European firms, or a mixture of global companies to make cross-country

(32)

28 S&P 500. As suggested, maybe the relationship between nationality diversity and

internationalization is more prominent in companies that do not belong in the upper echelon of the market but more in the middle. A third limitation is the possibility of causality between board national diversity and internationalization. In this study, it has been assumed that board national diversity influences internationalization. However, it can be possible that a company that has diversified internationally is able to attract more international directors. This could result in the opposite direction as has been assumed in this study, where the degree of internationalization can influence the degree of board national diversity (Oxelheim, et al., 2013). Another limitation is that this research focussed on the S&P 500 in general, not into some specific industries. Internationalization might be contextual dependent, and some

sectors show more internationalization than others. Future research could explore this to see if it holds for specific sectors. A final limitation is the operationalization of the dependent variable; internationalization. This research used two of the three parts of the Transnationality Index. Future research could include the complete Transnationality Index, as well as changing the ‘number of employees abroad / total employees’ by geographic dispersion of foreign sales. Finally, as mentioned, some countries are introducing laws that require companies to have a certain amount of female board members. Companies can already take this into account while selecting and developing top executives. It would be interesting to see if there is a difference in companies choosing to have more women on the board on a voluntary basis, compared with quotas. So, another suggestion for future research is to find out if female directors behave differently in the context of quotas.

6. References

Adams, R. B. & Ferreira, D., 2009. Women in the boardroom and their impact on governance and performance. Journal of Financial Economics, Volume 94, pp. 291-309.

Adams, R. & Funk, P., 2011. Beyond the glass ceiling: Does gender matter?. Management Science, 58(2), pp. 219-235.

Ali, M., Ng, Y. L. & Kulik, C. T., 2014. Board Age and Gender Diversity: A Test of Competing Linear and Curvilinear Predictions. Journal of Business Ethics, Issue 125, pp. 497-512.

Allison, P., 2018. When Can You Safely Ignore Multicollinearity?. [Online] Available at: https://statisticalhorizons.com/multicollinearity

[Accessed 10 4 2018].

Allison, P. D., 2009. Fixed Effects Regression Models. 1 ed. London: SAGE Publications.

(33)

29 Barroso, C., Villegas, M. & Pérez-Calero, L., 2011. Board Influence on a Firm’s Internationalization.

Corporate Governance: An International Review, 19(4), pp. 351-367.

Bartlett, C. & Ghoshal, S., 1992. What is a global manager?. Harvard Business Review, 70(5), pp. 124-132.

Bartlett, C. & Ghoshal, S., 1998. Managing Across Borders. 2 ed. Boston: Harvard Business School Press.

Brambor, T., Clark, W. R. & Golder, M., 2006. Understanding Interaction Models: Improving Empirical Analyses. Political Analysis, Issue 14, pp. 63-82.

Bureau van Dijk, 2017. Orbis - Overview. [Online]

Available at: https://www.bvdinfo.com/en-us/our-products/company-information/international- products/orbis?gclid=CjwKCAiA9f7QBRBpEiwApLGUipopvB14REF0BWW4O6SzFKUMQdkfMJge9-IUJut8HGkUZyahCjVbXBoCShUQAvD_BwE

[Accessed 30 11 2017].

Caligiuri, P., Lazarova, M. & Zehetbauer, S., 2004. Top managers’ national diversity and boundary spanning; attitudinal indicators of a firm’s internationalization. Journal of Management

Development, 23(9), pp. 848-859.

Campbell, K. & Mínguez-Vera, A., 2008. Gender Diversity in the Boardroom and Firm Financial Performance. Journal of Business Ethics, 83(3), pp. 435-451.

Canella, B., Finkelstein, S. & Hambrick, D., 2008. Strategic Leadership: Theory and Research on

Executives, Top Management Teams, and Boards. 1 ed. New York: Oxford University Press.

Carpenter, M. A. & Fredrickson, J. W., 2001. Top Management Teams, Global Strategic Posture, and the Moderating Role of Uncertainty. The Academy of Management Journal, 3(533-545), p. 44. Chao, M. C.-H. & Kumar, V., 2010. The impact of institutional distance on the international diversity– performance relationship. Journal of World Business, Issue 45, pp. 93-103.

Delios, A. & Beamish, P. W., 1999. Geographic scope, product diversification and the corporate performance of Japanese firms. Strategic Management Journal, 20(8), pp. 711-727.

Deloitte, 2016. Missing Pieces Report: The 2016 Board Diversity Census of Women and Minorities on

Fortune 500 Boards, New York: Alliance for Board Diversity.

Dunning, J. H., 1993. Multinational Enterprises and the Global Economy. 1 ed. Wokingham: Addison-Wesley.

Elron, E., 1997. Top management teams within multinational corporations effects of cultural heterogeneity. Leadership Quarterly, 8(4), pp. 393-412.

Fama, E. & Jensen, M., 1983. Separation of ownership and control. Journal of Law and Economics, Volume 26, pp. 301-325.

Finkelstein, S., Cannella, B. & Hambrick, D., 2008. Strategic Leadership: Theory and Research on

Executives, Top Management Teams, and Boards. 1 ed. New York: Oxford University Press.

Finkelstein, S. & Hambrick, D., 1996. Strategic Leadership: Top Executives and Their Effects on

(34)

30 Francoeur, C., Labelle, R. & Sinclair-Desgagné, B., 2008. Gender Diversity in Corporate Governance and Top Management. Journal of Business Ethics, 81(1), pp. 83-95.

Gabarro, J. J., 1987. The dynamics of taking charge. 1 ed. Boston: Harvard Business School. George, G., Wiklund, J. & Zahra, S., 2005. Ownership and the internationalization of small firms.

Journal of Management, 31(2), pp. 210-233.

Geringer, M. J., Beamish, P. W. & daCosta, R. C., 1989. Diversification Strategy and

Internationalization Implications for MNE Performance. Strategic Management Journal, 10(2), pp. 109-119.

Greene, W. H., 2012. Econometric analysis. 7 ed. London: Pearson.

Grosvold, J., Brammer, S. & Rayton, B., 2007. Board diversity in the United Kingdom and Norway: an exploratory analysis. Business Ethics: A European Review, 16(4), pp. 344-357.

Gul, F. A., Srinidhi, B. & Ng, A. C., 2011. Does board gender diversity improve the informativeness of stock prices. Journal of Accounting and Economics, 51(3), pp. 314-338.

Hair Jr., J., Anderson, R., Tatham, R. & Black, W., 1995. Multivariate Data Analysis. 3 ed. New York: Macmillan.

Hambrick, D., 1994. Top management groups: A conceptual integration and reconsideration of the "team" label. Research in organizational behavior, Issue 16, pp. 171-213.

Hambrick, D. C., 2007. Upper Echelons Theory: An Update. The Academy of Management Review, 32(2), pp. 334-343.

Hambrick, D. C. & Fukutomi, G. D., 1991. The Seasons of a CEO's Tenure. The Academy of

Management Review, 16(4), pp. 719-742.

Hambrick, D. & Mason, P., 1984. Upper Echelons: The Organization as a reflection of Its Top Managers. Academy of Management Review, 9(2), pp. 193-206.

Henderson, A. D., Miller, D. & Hambrick, D. C., 2006. How Quickly Do CEOs Become Obsolete? Industry Dynamism, CEO Tenure, and Company Performance. Strategic Management Journal, 27(5), pp. 447-460.

Herring, C., 2009. Does Diversity Pay?: Race, Gender, and the Business Case for Diversity. American

Sociological Review, 74(2), pp. 208-224.

Hillman, A. J., Cannella Jr, A. A. & Harris, I. C., 2002. Women and Racial Minorities in the Boardroom: How Do Directors Differ?. Journal of Management, 28(6), pp. 747-763.

Hitt, M., Hoskisson, R. E. & Kim, H., 1997. International diversification Effects on innovation and firm.

Academy of Management Journal, 40(4), pp. 767-798.

Hitt, M., Tihanyi, L., Miller, T. & Connelly, B., 2006. International diversification: Antecedents, outcomes and moderators. Journal of Management, Volume 32, pp. 831-867.

Houle, C., 1989. Governing boards. 1 ed. San Francisco:: Jossey Bass.

(35)

31 Joecks, J., Pull, K. & Vetter, K., 2013. Gender Diversity in the Boardroom and Firm Performance: What Exactly Constitutes a"Critical Mass?". Journal of Business Ethics, 118(1), pp. 61-72.

Johanson, J., Holm, U. & Forgsgren, M., 2014. Knowledge, Networks and Power: The Uppsala School

of International Business. 1 ed. London: Palgrave Macmillan.

Johanson, J. & Vahlne, J.-E., 1977. The internationalization process of the firm - a model of knowledge development and increasing foreign market commitments. Journal of International

Business Studies, 8(1), pp. 23-32.

Kaczmarek, S. & Ruigrok, W., 2013. In at the deep end of firm internationalization – Nationality diversity on top management teams matters. Management International Review, 53(4), pp. 513-534. Kam, C. & Franzese JR, R., 2003. Modeling and Interpreting Interactive Hypotheses in

RegressionAnalysis: A Brief Refresher and Some Practical Advice, Michigan: University of Michigan.

Laurent, A., 1983. The Cultural Diversity of Western Conceptions of Management. International

Studies of Management & Organization, 13(1), pp. 75-96.

Lawrence, B., 1997. The black box of organizational demography. Organization Science, Issue 8, pp. 1-22.

Lee, H.-U. & Park, J.-H., 2006. Top Team Diversity, Internationalization and the Mediating Eﬀect of International Alliances. British Journal of Management, Volume 17, pp. 195-216.

Luo, Y., 2005. How does globalization affect corporate governance and accountability? A perspective from MNEs. Journal of International Management, Volume 11, pp. 19-41.

Lu, X. & White, H., 2014. Robustness checks and robustness testsin applied economics. Journal of

Econometrics, Issue 178, pp. 194-206.

Martin, W., 2017. These will be the 32 most powerful economies in the world by 2050. [Online] Available at: https://www.independent.co.uk/news/business/these-will-be-the-32-most-powerful-economies-in-the-world-by-2050-a7587401.html

[Accessed 30 11 2017].

Mattson, L., 1998. Dynamics of overlapping networks and strategic actions by the international firm. 1 ed. Oxford : Oxford University Press.

Melin, L., 1992. Internationalization as a strategy process. Strategic Management Journal, Volume 13, pp. 99-118.

Miller, D., 1991. Stale in the Saddle; CEO Tenure and the Match Between Organization and Environment. Management Science, 37(1), pp. 34-52.

Miller, T. & Triana, M. d. C., 2009. Demographic Diversity in the Boardroom: Mediators of the Board Diversity–Firm Performance Relationship. Journal of Management Studies, 46(5), pp. 755-786. Milliken, F. J. & Martins, L. L., 1996. Searching for Common Threads: Understanding the Multiple Effects of Diversity in Organizational Groups. The Academy of Management Review, 21(2), pp. 402-433.

Mishra, R. K. & Jhunjhunwala, S., 2013. Diversity and the Effective Corporate Board. 1 ed. Cambridge: Academic Press.

(36)

32 Musteen, M., Barker III, V. L. & Baeten, V. L., 2006. CEO attributes associated with attitude toward change: The direct and moderating effects of CEO tenure. Journal of Business Research, Issue 59, pp. 604-612.

Nicholson, G. J. & Kiel, G. C., 2004. A Framework for Diagnosing Board Effectiveness. Corporate

Governance, 12(4), pp. 442-460.

Nielsen, B. B. & Nielsen, S., 2011. The role of top management team international orientation in international strategic decision-making: The choice of foreign entry mode. Journal of World Business, Issue 46, pp. 185-193.

Nielsen, B. B. & Nielsen, S., 2013. Top management team nationality diversity and firm performance: A multilevel study. Strategic Management Journal, Volume 34, pp. 373-382.

Nielsen, S. & Huse, M., 2010. The Contribution of Women on Boards of Directors: Going beyond the Surface. Corporate Governance: An International Review, 18(2), pp. 136-148.

Oxelheim, L., Gregorič, A., Randøy, T. & Thomsen, S., 2013. On the internationalization of corporate boards: The case of Nordic firms. Journal of International Business Studies, 44(3), pp. 173-194. Papadopoulos, N. & Martin, O., 2010. Toward a model of the relationship between

internationalization and export performance. International Business Review, Volume 19, pp. 388-406. Pettigrew, A., 1992. On studying managerial elites. Strategic Management Journal, Issue 13, pp. 163-182.

Pfeffer, J. & Salancik, G., 1978. The external control of organizations: A resource dependence

perspective. New York: Harper & Row.. New York: Harper & Row.

Pletzer, J. L., Nikolova, R., Kedzior, K. K. & Voelpel, S. C., 2015. Does Gender Matter? Female Representation on Corporate Boards and Firm Financial Performance - A Meta-Analysis. PLoSONE , 10(6).

Post, C. & Byron, K., 2015. Women on boards and firm financial performance: a metal-analysis.

Academy of Management Journal , 58(5), pp. 1546-1571.

Ramaswamy, K., Kroeck, K. G. & Renforth, W., 1996. Measuring the Degree of Internationalization of a Firm: A Comment. Journal of International Business Studies, 27(1), pp. 167-177.

Richard, O. C. & Shelor, R. M., 2002. Linking top management team age heterogeneity to firm performance: juxtaposing two mid-range theories. The International Journal of Human Resource

Management, 13(6), pp. 958-974.

Rivas, J. L., 2012. Diversity & internationalization: The case of boards and TMT's. International

Business Review, 21(1), pp. 1-12.

Rogerson, P., 2001. Statistical methods for geography. 1 ed. London: Sage.

Rose, C., 2007. Does female board representation influence firm performance? The Danish evident.

Corporate Governance: An International Review, 15(2), pp. 404-413.

Ruigrok, W. & van Tulder, R., 1995. The logic of international restructuring. In: Chapter 7: The Myth of