M
ULTINATIONAL BOARDS
,
FIRM INTERNATIONALISATION AND
INTER
-
COUNTRY DIFFERENCES IN BOARD STRUCTURE
-
T
HE IMPACT ON FIRM PERFORMANCE
Master’s Thesis Finance
(EBM866B20)STUDENT NUMBER: S2066211
NAME:MARIJN VONK
STUDY PROGRAM:MSC FINANCE
OTHER PROGRAMS FOR WHICH YOU HAVE SUBMITTED THIS THESIS:NONE
SUPERVISOR:PROF.DR.HERMES
FIELD KEY WORDS: CORPORATE GOVERNANCE, ETHICS & SRI
SPECIAL RESEARCH PROJECT:BOARD NATIONALITY DIVERSITY (SUPERVISOR:PROF.DR.HERMES)
Abstract
This research contributes to find relations between nationalities compositions of executive as well as non-executive boards and firm performance. By making use of a sample of 7947 West-European firms spanning the period 2006-2011, we obtain no consistent statistical support for a relationship between nationality diversity in executive or non-executive boards and ROA or Tobin’s Q. On average, multinational boards prove more often detrimental than beneficial to business. Second, this study demonstrates that a positive influence of executive board nationality diversity on corporate performance may emerge with higher degrees of firm internationalisation, albeit such a moderating influence remains absent in the non-executive board. Third, we take notice of inter-country variations, chiefly by accounting for differences in jurisdictions regarding board structure. One-tier corporations (with high percentages of foreign sales) benefit (most) from multinational executive boards. In addition, separate regressions for two countries show that German firms incur more significant negative effects from board nationality diversity, while higher foreign sales ratios invigorate British multinational executive boards. We obtain largely similar findings from firm-fixed OLS, instrumental variable analysis, and from using either a panel or cross-sectional data structure. The outcomes may designate that firms need to develop effective strategies that enable them to capitalise on the concomitant benefits of multinational boards and manage their inherent threats. Finally, the results imply that companies had better bear in mind firm characteristics (e.g. board structure) when considering foreign directors in order to expose their alleged benefits.
1. I
NTRODUCTION
To become and remain a major player in today’s complex and competitive markets, a firm’s board must understand its global environment. Hence, the board must comprehend the industry’s strengths and weaknesses, the market situation, regional regimes regarding economics and politics, financial and legal aspects, and the latest technology, while being able to overcome challenges during less prosperous times (Mishra & Jhunjhunwala, 2013). To this end it is necessary for the board to possess a variety of knowledge, expertise, skills, and perspectives. In other words, there ought to be diversity in board composition.
Board composition with regard to members’ backgrounds and characteristics has increasingly received attention in corporate governance (CG) research, even more after the global financial crisis, which resulted in a diminished base of support for existing CG practices (Eulerich, et al., 2013) (García-Meca, et al., 2015). While initial research on board composition chiefly addresses the educational and functional background of directors (e.g. Bilimoria & Piderit (1994), Boyd (1990) and Hillman, et al. (2000)), recent studies also include demographic characteristics like age, gender and nationality (e.g. Adams & Ferreira (2009), Hillman, et al. (2002) and Ruigrok, et al. (2007)).
A vast amount of empirical studies demonstrates how board diversity can be an indicator of success for the international corporate practice (e.g. Dahya & McConnell (2007), Kiel & Nicholson (2003) and Rose (2007)). Diversity in boards, comprising executive and non-executive directors, can create economic value by stimulating independence, increasing diversity of knowledge and ideas, providing better access to scarce resources (Hermes, 2015) and helping firms enhance their reputation as responsible citizens (Mishra & Jhunjhunwala, 2013). Other scholars have indicated that more diverse boards can actually inhibit board effectiveness (Eulerich, et al., 2013) (Hermann & Datta, 2005). For instance, Eulerich, Velte, & van Uum (2013) argue that diversity on boards complicates decision-making processes, and decreases communication and incites conflicts between members.
Surprisingly, within the field of board diversity, board nationality diversity is a dimension that has largely been omitted. This is astonishing considering that many other indicators of diversity, such as gender, educational and occupational background attained plentiful
attention by researchers (e.g. Goodstein, et al. (1994), Golden & Zajac (2001) and (Westphal & Zajac (1995). The limited number of studies that do address nationality diversity are usually based on US company data (e.g. Caliguri, et al. (2004), Greve, et al. (2010), Masulis, et al. (2012), Reuber & Fischer (1997), Sambharya (1996) and Tihanyi, et al. (2005)). An evolution of research towards nationality diversity would be a logical consequence of the globalising economy where more companies take part in foreign markets and multinational corporations grow bigger (Karadagli, 2012) (Thoumrungroje & Tansuhaj, 2007).
The central questions this thesis pertains to address are whether nationality diversity on the corporate board brings about effects in firm performance and whether these effects vary with the extent of firm internationalisation. At least in this context, our measure of nationality diversity (which we define below) has not been used in prior research papers. The study is based on a unique sample of West-European companies, where we pay special attention to cross-country differences, such as CG regimes. Corporate boards contain both executive and non-executive directors ((N)EDs) and their performance is interlinked (Chahine & Goergen, 2014) (Millet-Reyes & Zhao, 2010) (Shleifer & Vishny, 1997). Hence, our analysis addresses NEDs and EDs where each relies on a different theoretical underpinning. The results of this study may be used to optimize board composition in order to maximize firm performance.
The contribution of this thesis is fourfold. First, we use a different measure of nationality diversity. An overwhelming majority of researchers use similar measures for board internationalisation, which they often (incorrectly) denote as “nationality diversity”. The most frequently used measures are international experience, the proportion of foreigners on the board, or dummy variables that account for the presence of a foreigner on the board (Alli, et al., 2010) (Carpenter & Fredrickson, 2001) (García-Meca, et al., 2015) (Masulis, et al., 2012) (Oxelheim & Randøy, 2003) (Oxelheim, et al., 2013) (Sambharya, 1996) (Ujunwa, et al., 2012). Unfortunately, investigators often fail to specify the diversity type they are studying (Harrison & Klein, 2007). Instead, we consider the variety of nationalities on the board using Blau’s index (Blau, 1977) and evaluate its effect on firm performance. Although this formula is commonly applied in diversity studies relating to executive boards (e.g. Bunderson & Sutcliffe (2002) Greve, et al. (2010) and Tihanyi, et al. (2005)), it is far less popular in empirical work regarding non-executive board diversity. This is astonishing, since unlike
other measures, Blau’s index implies an unambiguous definition of diversity, which lets a clearer and ampler understanding of diversity influences in firms (Harrison & Klein, 2007).
Second, we incorporate the potential effect of firm degree of internationalisation on the relation between board nationality diversity and firm performance. We witnessed no prior empirical work that touches upon this moderating influence. A firm’s international entanglement could affect particular variables through various relationships and these relationships may be bilateral (see Figure 5 in Section 3.3.2). Yet, so far, researchers have only explored a few of these links (e.g. Masulis, et al (2012), Oxelheim, et al. (2013) and Ruigrok & Wagner (2003)). For this reason, we attempt to set up our study in such a way that the degree of internationalisation serves as a moderator of the association between board nationality and firm performance (i.e. arrow 2 in Figure 4 in Section 3.3.1).
Third, we base our findings on a unique dataset of West-European firms. Therein, we account for country-specific effects for the analysis of our results, since we expect significant differences between nations in terms of the effect on the variables of interest. In particular, we distinguish countries on the basis of their dominant board structure. To our knowledge, no other researchers have considered the country’s prevalent board structure in the context of board internationalisation. These cross-country variations in board structure influence the content and realisation of director duties and the expectations that they face (Adams & Ferreira, 2009) (Belot, et al., 2014), and may thus affect the studied relations in this thesis.
Finally, we perform our analysis and assessments for non-executive, as well as executive directors, since both groups constitute the corporate board and are highly interdependent (Adams & Ferreira, 2007) (Millet-Reyes & Zhao, 2010) (Shleifer & Vishny, 1997). The structure and processes through which boards deliver value must enable optimum engagement and communication between the non-executive and the executive directors (Heidrick & Struggles, 2011) (Shleifer & Vishny, 1997). To this end, it is interesting to investigate effects for both groups (based on the same firm sample) and perceive the results in parallel. Consequently, we develop a theoretical foundation for each group. Afterwards, we acquire and reflect on the results for NEDs and EDs separately.
This thesis is structured as follows. Chapter two outlines the theoretical background for this study, resulting into a set of hypothesis statements. The third chapter elaborates on the methodology and input data. Subsequently, chapter four provides an overview and discussion of the results. These findings are coupled to the hypotheses and theory given in chapter two. Chapter four ends with some suggestions for future research. The final chapter summarizes the main conclusions of this study.
2. T
HEORETICAL
B
ACKGROUND AND
H
YPOTHESES
Ingley & van der Walt (2003) state that “diversity relates to board composition and the varied combination of attributes, characteristics and expertise contributed by individual board members in relation to board process and decision-making.” Nationality diversity is an important element of board diversity. This chapter provides the theoretical foundation for this research on the influence of board nationality diversity on corporate performance and examines whether anticipated effects may vary with the firm degree of internationalisation. In addition, we devote a separate section to explaining prevailing board structures and their possible impact on a relationship between board nationality diversity and firm performance. We address board nationality diversity through assessing its impact on three primary roles that non-executive directors need to fulfil. These roles entail: (1) monitoring, (2) advising, and (3) providing access to key resources (Pearce & Zahra, 1992). The first two roles arose with the introduction of agency theory, while the third role originated from
resource dependence theory. We adopt resource dependence and upper echelon theory to
explain potential performance effects of nationally diverse executive board members.
2.1. NON-EXECUTIVE DIRECTORS
2.1.1. AGENCY THEORY
Most scholarly research projects in corporate governance start from the understanding that in public companies with dispersed ownership an agency relation exists between the managers as agents whose decisions affect the shareholders as principals. Corporate governance practices aim to address the complications that emerge from this relationship, in order to ensure that managers act in the interests of shareholders and overcome agency
problems (Fama & Jensen, 1983) (Jensen & Meckling, 1976).
The board is entrusted with the responsibility to monitor and ratify management decisions so as to align the interests of executives and shareholders (Jensen & Meckling, 1976). From this perspective, independent directors are expected to be more objective assessors of management decisions and therefore better capable of protecting the interests of shareholders (Arnegger, et al., 2014). Foreign directors are less likely to get entangled in
the social network of board directors, thanks to their socio-cultural or geographical distance from other board members (Jayne & Dipboye, 2004) (Fogel, et al., 2013) (Masulis, et al., 2012) (Oxelheim & Randøy, 2003). Therefore, they will be more capable of making independent decisions, which contributes to a better fulfilment of their advisory and supervisory role (Choi, et al., 2007) (Choi, et al., 2012) (Oxelheim & Randøy, 2003).
Board diversity increases diversity of knowledge and ideas. The same argument applies to board nationality diversity (Ezat & El-Masry, 2008) (Ruigrok, et al., 2007) (Samaha, et al., 2012) (Ujunwa, 2012). Board members from different countries may better understand how businesses, markets and people in these countries operate and compare the firm with its foreign competitors (Oxelheim, et al., 2013). Hence, they bring unique expertise and information on the business environments of other countries, that domestic constituents are unable to acquire (Berger & Neugart, 2010) (Carter, et al., 2010) (Maznevski, 1994) (Oxelheim, et al., 2013) (Pearce & Zahra, 1992). This encourages better decision-making with regard to investments and operations (Oxelheim, et al., 2013).
Foreign directors also have different backgrounds with respect to culture and lifestyle that promote other perspectives and solutions among board members (Maznevski, 1994) (Mishra & Jhunjhunwala, 2013). Maznevski (1994) contends that this diverse set of perspectives and solutions adheres to the board’s “culture-generic” competence, i.e. “knowledge and skills that are applicable across countries or ethnic groups”. Consequentially, this allows the board to produce better advice and a broader decision set for top management. An expansion of decision opportunities increases the probability of a positive company result (Eulerich, et al., 2013) (Ruigrok, et al., 2007). In addition, Bantel & Jackson (1989) Eulerich, et al. (2013), Ferreira (2010) and Ujunwa, et al. (2012) contemplate that varying perspectives, backgrounds and viewpoints, especially international ones, can insulate against groupthink. Chiefly, this should increase the non-executives’ quality and quantity of advice to their counterparts.
A wide range of knowledge, ideas and points of view are especially useful in settings of high complexity (Mohammed & Ringseis, 2001) (Sauer, et al., 2006). Firms operating in international environments are inherently more complex (Celo, et al., 2015) (Gaur, et al., 2011) (Hymer, 1976) (Oxelheim & Wihlborg, 2008) (Zaheer, 1995). The added value of incumbent foreign directors may concomitantly become more noticeable in such intricate sceneries, where boards have to perform increasingly non-routine and complex tasks (Hage, et al., 1971) (Lublin, 2005). Particularly, firms will more likely seek the wide variety of knowledge, ideas and perspectives from their multinational boards in order to derive a solution that overcomes its “liability of foreignness” (Hymer, 1976) (Zaheer, 1995). On the other hand, a narrow-minded group could easily run out of alternative solutions and stagnate (Keen & Sol, 2008). In addition, foreign directors’ knowledge about other countries’ business and institutional environment becomes particularly useful when the firm has operations in these, or similar countries. In other words, the degree of firm internationalisation should reinforce any positive effects caused by nationality diversity among NEDs on overall board effectiveness, firm decision-making and performance.
The former discussion elaborates on how foreign NEDs may enforce board effectiveness, especially by means of better advising. However, foreign board members may also impede board effectiveness. Research indicates that nationality diversity provokes reduced and ineffective communication, particularly due to language and cultural barriers (Hermann & Datta, 2005) (Piekkari & Oxelheim, 2014). This could engender slower decision-making, misunderstandings and conflicts (Eulerich, et al., 2013) (Konrad & Kramer, 2006) (Ruigrok, et al., 2007). For example, Piekkari & Oxelheim (2014) show that NEDs find it difficult to contribute to board meetings and articulate disagreement when using a non-native language. Persistent anti-effective communication could make NEDs reluctant to provide advice to executive directors, and management unwilling to consider the recommendations made by the NEDs. These communication inefficiencies are found most disruptive to group-decision making processes (Adler, 1991).
Foreign directors can also be inhibited in their monitoring role. Masulis, et al. (2012) state that foreign directors are less familiar with national accounting rules, laws and regulations, governance standards and management methods, and find evidence backing up their statements. This makes it tough for foreign NEDs to perform accurate monitoring and convey suggestions credibly. Second, foreign board members less frequently attend periodic board meetings due to directors’ geographic distance from corporate headquarters (Lerner, 1995) (Masulis, et al., 2012). Logically, the absentees’ advisory and monitoring efforts remain absent with them. Third, as mentioned before, non-national directors can be cut off from local networks. Notwithstanding a higher independence, it may lead to passing on important soft information (Coval & Moskowitz, 2001). Thus, foreign NEDs may miss crucial information about the firm’s status on performance and latest activities (Masulis, et al., 2012).
In addition to all the foregoing, foreigners on the board are found more likely to exhibit less commitment or counterproductive behaviour to the firm, as a result of social and cultural distances (Jayne & Dipboye, 2004). Related to these findings, literature shows that diversity results in less social cohesion and that social barriers reduce the amount of successful influence efforts by minority directors in the board’s decision-making (Westphal & Milton, 2000). Each of these elements could incite a less adequate fulfilment of board roles.
2.1.2. RESOURCE DEPENDENCE THEORY
Emerging criticism on corporate governance research includes a too narrow focus on agency theory and associatively board member independence, thereby neglecting board functions beyond monitoring and advising (Daily, et al., 2003). Taking this into account, researchers more often consider the importance of board member characteristics, experience and traits other than those stimulating independence (Hillman, et al., 2000).
One theory that serves as an alternative to agency theory is resource dependency theory. Since resource dependence theory is closely related to the third role of the board (i.e. providing access to key resources) this theory serves as an extension rather than an alternative to agency theory. Pfeffer & Salancik (1978) introduced resource dependence theory and stipulated that organizations are affected by and dependent on resources cultivated within other organizations. Due to that resources needed by one organization are
often in possession of other organizations, it makes sense for a company’s board to build a network and establish relationships with parties that own these key resources. This example of providing access to key resources is exactly the third role that boards need to fulfil. In particular, resource dependence theory points to resources such as communication channels, expertise, support from external organizations, and legitimacy that directors might bring to the boardroom (Arnegger, et al., 2014) (Pfeffer & Salancik, 1978).
Researchers identified numerous potential advantages linked to resource dependence theory, which could be instigated by foreign board members. First of all, foreign directors provide access to networks of foreign companies that own strategic resources (Oxelheim, et al., 2013). Access to these networks is especially important during turbulent times when the firm’s competitive strategic advantages may be jeopardised (Carpenter & Westphal, 2001) and enables the firm to (re-)establish strategic partnerships. Hillman & Dalziel (2003) categorise this source of benefits accrued from the appointment of foreign NEDs as “relational capital”. Second, foreign directors should bring intangible resources like intellectual skills, technological know-how and reputation (Gupta, 1987) (Oxelheim & Randøy, 2003) (Shirodkar & Mohr, 2015), i.e. “human capital" (Hillman & Dalziel, 2003). Hereby, the firm will improve its competitive position.
Foreign operations pose significant challenges to a firm and its board. Hymer S. H.(1976) first identified the associated “costs with doing business abroad”. Over time, researchers refined this concept by highlighting the costs incurred by firms due to dissimilarities in the institutional environments between countries, which became known as the “liability of foreignness” (Zaheer, 1995). Companies that have foreign NEDs in place will more often consult their unique networks, expertise and information channels in order to help overcome these costs of liability of foreignness (Celo, et al., 2015) (Gaur, et al., 2011) (Hymer, 1976) (Oxelheim & Wihlborg, 2008) (Zaheer, 1995). As such, when the firm has a larger degree of internationalisation, it is more likely to notice the associated benefits of its foreign directors. In view of that, firm degree of internationalisation should positively moderate the contributions of foreign NEDs to firm performance.
2.2. EXECUTIVE DIRECTORS
2.2.1. UPPER ECHELON THEORY
According to Hambrick (2007) the central premise of upper echelon theory is that “executives’ experiences, values, and personalities greatly influence their interpretations of the situations they face and, in turn, affect their choices”. The main choices that EDs must make dedicate to setting the firm’s strategic direction to develop and sustain competitive advantages in order to enhance performance (Hambrick & Mason, 1984) (Marimuthu & Kolandaisamy, 2009). Hambrick & Mason (1984) thus consider corporate performance to be a reflection of the characteristics and decisions of the executive board. Although the theory has been refined over the years, it has always recognized the influence of diversity in managerial background, characteristics and previous experiences on these strategic choices and performance (Carpenter & Fredrickson, 2001) (Carpenter, et al., 2004) (Certo, et al., 2006) (Finkelstein & Hambrick, 1996) (Finkelstein, et al., 2008) (Hambrick & Mason, 1984) (McGrath, 1984) (Murray, 1989). Virtually all work that shows a primary interest in foreign EDs makes use of this theory and offers evidence for its significance (e.g. Greve, et al. (2010), Heijltjes, et al. (2003), Hermann & Datta (2005) and Nielsen & Nielsen (2011)).
Upper echelon theory’s underlying perceptual model of firm strategic choice suggests that top executives have a limited field of vision, selective perception and interpretation bias, which create the basis for the decisions that they make (Hambrick & Mason, 1984). The leading argumentation that links upper echelon theory to the characteristics of EDs (like age, gender and nationality) entails that these characteristics influence the decisions made by these directors and as such the actions adopted by the company that they lead (Child, 1972) (Hambrick & Mason, 1984). This line of reasoning is valid, since demographic characteristics are associated with many cognitive biases, values and perceptions that influence the decision-making of top management (Marimuthu & Kolandaisamy, 2009) (Michel & Hambrick, 1992). Research confirms that diversity introduces heterogeneity of ideas, knowledge and perspectives, which promote creativity, innovativeness, quality decision-making and competitiveness (Hambrick, et al., 1996). Over time, this discussion initiated the development of six specific influence processes that pave a way for strategic decision-making and performance enhancement in the firm (Nahavandi, 2006).
Again, building on the former reasoning behind upper echelon theory, we contend that in firms facing high environmental uncertainty, EDs’ decisions and actions are strongly influenced by their backgrounds, characteristics and experiences (Child, 1972) (Hambrick & Mason, 1984). Internationally-developed firms are often larger and have to deal with different countries’ regulations and CG regimes (Hymer, 1976) (Zaheer, 1995). These firms concomitantly experience more environmental uncertainty and are more difficult to manage (see Section 2.1.1) (Celo, et al., 2015) (Hymer, 1976) (Oxelheim & Wihlborg, 2008) (Zaheer, 1995). Former research advocates that the likelihood of a successful outcome in situations facing substantial complexity rises considerably, as diversity among executives’ experiences, values and personalities increases (Mohammed & Ringseis, 2001) (Sauer, et al., 2006). In contrast, simple tasks barely benefit from more diversity (Mohammed & Ringseis, 2001) (Sauer, et al., 2006). Accordingly, the increased complexity that comes with a higher degree of firm internationalisation should help determine to what extent a firm is able to exploit potential benefits of having more EDs from different countries.
2.2.2. RESOURCE DEPENDENCE THEORY
International executive directors can also bring resources (i.e. networks, expertise, access to external resources and reputation) from abroad that complement or supplement available resources in the firm (Clarysse, et al., 2007) (van Veen & Marsman, 2008). Besides, unique experiences of foreign managers can help the executive board to better understand the local market and institutions (Kobrin, 1984). Where foreign EDs provide resources that are complementary to those of the domestic directors, firm performance may increase. Thus, analogous to the debate on NEDs as regards resource dependence theory, foreign executive directors make it easier for the firm to acquire strategic resources (Mizruchi, 1997).
Firmsincur significant “social and economic costs” from foreign market operations (Celo, et al., 2015) (Gaur, et al., 2011) (Hymer, 1976) (Oxelheim & Wihlborg, 2008). For example, companies that operate in different countries rely on strategic resources that are situated in those countries, but are not easily accessible or readily available (Gaur, et al., 2011) (Gupta, 1987) (Shirodkar & Mohr, 2015). Existing local corporations may regard foreign corporations intrusive and be unwilling to cooperate, while local governments may mandate restrictions on multinational corporations so as to protect local firms (Hoskisson, et al., 2000)
(London & Hart, 2004). Typically, incumbent foreign directors possess valuable knowledge and expertise linked to these resources, and have the ability to develop networks of cooperative relationships (Clarysse, et al., 2007) (Guillén & García-Canal, sd). Concomitantly, internationalised firms will regularly appeal to the foreign EDs’ unique abilities to secure access to these resources. Hence, similar to our reasoning for NEDs (see Section 2.1.2), variation between the nationalities of EDs could cultivate firm strategic advantages, which are intensified by the extent of firm internationalisation.
In non-executive boards, foreigners may introduce issues connected to communication and team dynamics (see Section 2.1.1). These problems could also apply to executive boards (Hambrick, et al., 1996) (Hermann & Datta, 2005) (Thomas, et al., 1996). For example, Hambrick, Cho, & Chen (1996) observed that heterogeneous top management teams were slower in their actions and responses and less likely to respond to competitor’s initiatives. Hence, communication inefficiencies, conflicts, low cohesion and slow decision-making in executive boards could also seriously reduce their effectiveness, while persistent anti-effective communication could make management even more unwilling to cooperate. In confirmation with these statements, Westphal & Milton (2000) document that higher diversity results in lower social cohesion, and that social barriers reduce the amount of successful influence attempts by minority board members in group decision-making.
Furthermore, foreign EDs may be unfamiliar with national business practices, laws and regulations, and management methods (Masulis, et al., 2012). This could hinder their ability to do the right things and do the things right. Additionally, foreign directors may be cut off from local networks due to intercultural distances (Jayne & Dipboye, 2004). This could give occasion to passing on important business-related information (Coval & Moskowitz, 2001). Persistent social awkwardness could eventually bring about less commitment and/or counterproductive behaviour (Jayne & Dipboye, 2004). So far, our review of literature on board nationality diversity in relation to firm performance shows concomitant benefits and costs with foreign directors. We want to find out how the costs and benefits of multinational boards balance and whether firm degree of internationalisation affects a relationship between board nationality diversity and firm performance. In regard of these relationships, the next section also contemplates influences of cross-country differences in board structure.
2.3. CROSS-COUNTRY DIFFERENCES IN BOARD STRUCTURE
There may exist significant inter-country differences in the results of our investigation. Such differences likely originate from cross-national variations in ownership structures and composition of corporate boards (Li, 1994). Government regulations affect the manner in which they are controlled and the processes by which changes in ownership and control take place (Jenkinson & Mayer, 1992) (Prowse, 1990). These regulations influence corporate governance and control by means of two factors: (1) the structure of corporate ownership, and (2) the structure of the board of directors (Li, 1994). Subsequently, these factors shape the nature of executive and non-executive director duties and the expectations that they face (Heidrick & Struggles, 2011). Many variables exists which help to explain inter-country differences (Ferreira & Kirchmaier, 2013) (Jenkinson & Mayer, 1992) (Li, 1994). Yet, for the sake of clarity and its anticipated significance, our research explores only the variations in the country’s prevailing board structure. Regulations on board structure vary considerably across countries and evidently shape the corporate board and ownership structure (Adams & Ferreira, 2009) (Ferreira & Kirchmaier, 2013) (Gillette, et al., 2004) (Li, 1994). Understanding the influences of differences in board structures throughout Europe is important, because many of the legislative proposals that have developed since the crisis focus on restructuring European boards (Ferreira & Kirchmaier, 2013). Hence, we use the differences in board structure to separate countries for which we perform our analysis independently (see Section 4.2.3). In addition, we apply the variable “board structure” into our original regression models (see Section 3.3.1 and 4.2.3).
Figure 1 – Board structures in Europe. Source: Heidrick & Struggles (2011)
The most clear-cut distinction between national legal structures influencing boards can be made in terms of the country’s prevalent board structure (see Figure 1) (Ferreira & Kirchmaier, 2013) (Gillette, et al., 2004) (Jenkinson & Mayer, 1992) (Millet-Reyes & Zhao, 2010). In 2011, i.e. the end period of our sample, several West-European countries have a
unitary board structure with a single board comprised of at least three and no more than 24
directors, both executive and non-executive. In single-tier jurisdictions such as the United Kingdom, the board is empowered to both manage and supervise a corporation (see Figure 2). Moreover, many West-European countries have a two-tier board structure with a distinct supervisory and executive board, and separate chairmen. The management board consists of three to five members, while the supervisory board consists of three up to a maximum of 24 members (Millet-Reyes & Zhao, 2010). This board structure is found in all German and Austrian companies (see Figure 2). The mixed system has an executive and a non-executive board, but usually with a combined chairman and CEO and some executives on the non-executive board. This system is predominant in Belgium (see Figure 2).
Figure 2 – The prevalence of board structures of European firms in 2011. The figures are derived from the content provided by Heidrick & Struggles (2011), who base their findings on a European dataset of 400 firms. We also compared their findings with those of Belot, Ginglinger, Sloyin, & Sushka (2014) and Millet-Reyes & Zhao (2010), which bear close resemblance. 68% 12% 83% 5% 100% 27% 100% 5% 89% 75% 23% 100% 8% 84% 65% 84% 26% 60% 42% 95% 11% 25% 77% 24% 4% 35% 16% 17% 74% 35% 31% Austria Belgium Denmark Finland France Germany Italy Netherlands Norway Portugal Spain Sweden Switzerland United Kingdom 2011 European Average
One-tier board structure Two-tier board structure Mixed system
Compared to unitary boards, one can identify a number of benefits associated with two-tier boards. First, the dual board makes a clear division between the tasks and responsibilities of the directors on the management and supervisory board (Adams & Ferreira, 2007) (Belot, et al., 2014) (ecoDa, NautaDutilh, 2014) (Millet-Reyes & Zhao, 2010). Unlike one-tier boards, this reduces the risk of the intertwining of implementation and supervision, because the supervisory board is no longer outside the board (Belot, et al., 2014) (Millet-Reyes & Zhao, 2010). In unitary structures non-executive directors are less independent of the management because they now form part of it themselves (Adams & Ferreira, 2007). Another advantage arises from the small size of the executive board (three to five members), which allows for quick decision-making (Belot, et al., 2014) (Millet-Reyes & Zhao, 2010). Moreover, the two-tier structure helps non-traditional candidates to rise to the ranks of directors, creating more diversity in the boardroom (Millet-Reyes & Zhao, 2010). An elevation of diversity may stimulate the entire board’s functioning owing to a greater base of complementary knowledge, ideas and perspectives (see Section 2.1.1). Finally, two-tier board typically generate good publicity and help companies to attract foreign capital (Millet-Reyes & Zhao, 2010). With more available funds and “external support”, (foreign) directors may be hindered less in their actions and decision-making.
There are also a number of drawbacks of two-tier relative to one-tier board structures. First, dual board structures encumber directors to build relationships of trust, due to the separation of boards, thereby potentially undermining communication between the two boards (ecoDa, NautaDutilh, 2014). Second, an important advantage of unitary boards, is that directors receive earlier and more information than in a two-tier board structure (Belot, et al., 2014) (ecoDa, NautaDutilh, 2014). Third, the (direct) involvement of non-executive directors should be greater in one-tier boards (Belot, et al., 2014) (ecoDa, NautaDutilh, 2014). Non-executive directors can timelier influence the direction and strategy of the company in their supervisory role, accordingly (Belot, et al., 2014) (ecoDa, NautaDutilh, 2014). They are after all responsible for the actions and decisions of executive directors in the daily business. Furthermore, companies often comply with a dual board structure for political rather than economic reasons. For example, an executive who has been directed to abandon the executive board can subtly be re-installed on the non-executive board (Millet-Reyes & Zhao, 2010). We
now proceed to discuss how the board structure may moderate or incite a relationship between board nationality diversity and firm performance. Such an influence may funnel
through the theories as discussed in Sections 2.1 and 2.2. For instance, in the case of NEDs, a
particular board structure may incite manifestation of effects that originate from board nationality diversity that cause a better fulfilment of their board roles.
The characteristics of either board structure may affect the extent to which directors have an impact on firm performance (see Section 2.1 and 2.2 for an exposition on the influence of board nationality variety on firm performance). In theory, single-tier boards concentrate decision-making power, while two-tier boards favour collective decision-making (Adams & Ferreira, 2007) (Enriques, et al., 2009) (Moerland, 1995). For instance, unitary boards permit firms to combine the roles of chairman and CEO, as commonly occurs in the US (Enriques, et al., 2009). By contrast, dual boards as in Germany prohibit supervisory boards from making managerial decisions and require that management boards make decisions by majority vote (Dejure, sd) (Enriques, et al., 2009). Hence, foreign directors on a two-tier board may be more
regularly involved in the board’s decision-making. This is especially important for foreign
directors, because they seldom appear on top board positions (Egon Zehnder, 2014) (Hunt, et al., 2015) (Osawa, 2010), which makes them less likely to have a decisive influence. Alternatively, sole boards concentrate decision-making power, such that the magnitude of influence exerted by foreign directors on decisions should be greater (Enriques, et al., 2009). One-tier jurisdictions allow for a greater spirit of partnership and mutual respect between directors, which promotes interaction amongst board members (Belot, et al., 2014) (ecoDa, NautaDutilh, 2014). This should rouse partaking of foreign directors and enlarge their contributions, accordingly. Adams & Ferreira (2007) contend that both monitoring and advising by the board become more effective, since proximity between NEDs and EDs augments exchange of information about the firm's investment opportunities. Subsequently, multinational executive boards may prosper in the course of processing NEDs’ advice for decision-making. In contrast, two-tier board structures make it more difficult for directors to build and sustain trust (Adams & Ferreira, 2007) (ecoDa, NautaDutilh, 2014). This could hinder the internationalised board to flourish, since NEDs and EDs are mutually reliant (e.g. through information and approval provision).
Building on the characteristics of sole boards, NEDs have direct access to information, which grants them the opportunity to take necessary interventions (Adams & Ferreira, 2007) (Belot, et al., 2014) (Boot, et al., 2005) (ecoDa, NautaDutilh, 2014). The degree to which such interferences are successful is attributable to the quality of solutions brought by NEDs. The aptness of NEDs’ solutions typically relies on the amount and diversity of knowledge, ideas and perspectives, which should be greater for multinational boards (see Section 2.1.1). In other words, improved access to information in a unitary board prompts more intervention (i.e. advice or monitoring) opportunities for NEDs. Any benefits of nationality diversity should manifest in these interventions and could eventually equip EDs with new opportunities as well. For instance, EDs may be offered early advice enabling them to identify and tackle problems head on. The degree to which EDs properly utilise NEDs’ advice depends partially on executives’ ability to conquer their limited fields of vision, selective perceptions and interpretation biases (see Section 2.2.1). Contrarily, in two-tier boards NEDs only receive limited information from executives and at a later stage (Adams & Ferreira, 2007) (Boot, et al., 2005) (ecoDa, NautaDutilh, 2014). This magnifies the risk of NEDs not (or too late) discovering deficiencies and not fully understanding and ratifying strategic initiatives (ecoDa, NautaDutilh, 2014). Similar to our prior argumentation, this can inhibit the extent to which nationality diversity among NEDs as well as EDs affects corporate performance.
Furthermore, board meetings take place on a more regular basis in sole boards, due to the pooling of responsibilities related to management and supervising (ecoDa, NautaDutilh, 2014). Although frequent meetings are usually beneficial, it may make it (even) harder for foreign NEDs to stay up-to-date about business developments, because of their recurrent absence at board meetings (Masulis, et al., 2012). While in dual boards the decision-making process is delayed due to less frequent supervisory board meetings (ecoDa, NautaDutilh, 2014), non-executives may attend these meetings relatively more often. In line with the former argumentation, decision-making processes can be accelerated in one-tier boards, and slowed down in two-tier boards (Belot, et al., 2014) (ecoDa, NautaDutilh, 2014) (Gillette, et al., 2004). In accordance, EDs may be granted early permission to take (strategic) actions and decisions, and NEDs may intervene more timely. Either occasion provides the opportunity for a manifestation of the effects of board nationality diversity on firm performance.
Dual boards separate supervisory and management functions and liabilities stimulating independence. Independent boards monitor management more intensively (Adams & Ferreira, 2007) (Rose, 2005), allowing foreign NEDs to excel in this role. However, as a result of invigorated monitoring, NEDs may interfere in the activities of the executive board, diluting any efforts made by foreign EDs (Adams & Ferreira, 2007) (Gillette, et al., 2004). In response, EDs will not communicate firm-specific information to NEDs that are too independent, because of the risk to the EDs that the NEDs will interfere in decision-making (Adams & Ferreira, 2007). This may inhibit foreign NEDs in their advisory role, because they forego essential information. Consequently, EDs will miss out on valuable advice. Moreover, dual boards empower the management board, such that professional executives can gather information on projects without conflicting with the incentives of the leading shareholders on the non-executive board (Belot, et al., 2014) (Graziano & Luporini, 2012). This highlights benefits from nationality diversity on one of EDs’ core tasks, i.e. gathering information on projects to make investment decisions. One final advantage of dual boards relates to the small size of the management board. An executive board composed of merely three to five members makes the influence of foreign EDs more paramount (Millet-Reyes & Zhao, 2010).
Mixed boards generally give expression to a subset of the advantages and disadvantages of two-tier boards relative to one-tier boards (Belot, et al., 2014). Unlike sole and dual boards, mixed board structures vary across countries per se (Belot, et al., 2014) (Gillette, et al., 2004) (Millet-Reyes & Zhao, 2010). This makes it extremely difficult to develop a clear and reliable theoretical foundation for its conceivable effects and draw accurate conclusions. Hence, we limit our study to the effects of one-tier and two-tier board structures. The former elucidation substantiates that unitary boards should chiefly incite any influences of (non-)executive board nationality diversity on firm performance, though we find no univocal arguments for such an effect of dual boards. Internationalised firms perish from “liability of foreignness”. The concomitant complexity may more likely be overcome in sole boards owing to improved access to information, greater interaction and faster decision-making. Dual boards express the opposite of these advantages and may thus inhibit effects of nationality diversity. Table 13 lists the leading arguments that provide for the theoretical backbone of this study. This table also illustrates that there are a number of overlapping arguments for NEDs and EDs.
2.4. HYPOTHESIS STATEMENTS
In line with the theoretical foundation (see Section 2.1-2.3) of this study, we unfold eight hypotheses. Hypothesis 3-8 complement hypothesis 1 and 2 with the purpose to offer a more comprehensive story on the influence of board nationality variety on firm performance. With this study, we want to identify the net effect of board nationality diversity on firm performance. Prior empirical and theoretical research offer widely diverging evidence for the direction or magnitude of this effect. Thus, we state the first hypothesis in its null form (H0):
Hypothesis 1. Nationality diversity in non-executive boards does not affect firm performance.
Based on the theory and empirical evidence on executive board nationality diversity (see Section 2.2), we conjecture that the net performance effect of foreigners on the company board is neutral, negative or positive. Consequently, we pose hypothesis two as follows:
Hypothesis 2. Nationality diversity in executive boards does not affect firm performance.
Section 2.1 explains, through agency and resource dependence theory, that the degree of firm internationalisation is expected to provide the right setting for foreign directors in order for them to excel. This results into the following hypothesis:
Hypothesis 3. Firm degree of internationalisation moderates and/or triggers a positive relationship between nationality diversity in non-executive boards and firm performance.
Similar to their non-executive counterpart, we predict that firms receive (more) benefits from board nationality diversity, as the firm’s extent of internationalisation increases:
Hypothesis 4. Firm degree of internationalisation moderates and/or triggers a positive relationship between nationality diversity in executive boards and firm performance.
Our literature review for the most part indicates that unitary boards should stimulate the manifestation of effects from board nationality diversity on firm performance. Prior research
do not consistently point to such an influence of two-tier board structures:
Hypothesis 5a. One-tier boards moderate and/or trigger a positive relationship between nationality diversity in non-executive boards and firm performance.
Hypothesis 5b. Two-tier boards do not moderate and/or trigger a relationship between nationality diversity in non-executive boards and firm performance.
In the same way, we formulate the hypothesis for executive directors like so:
Hypothesis 6a. One-tier boards moderate and/or trigger a positive relationship between nationality diversity in executive boards and firm performance.
Hypothesis 6b. Two-tier boards do not moderate and/or trigger a relationship between nationality diversity in executive boards and firm performance.
We also explore effects of board structure interacting with degree of internationalisation:
Hypothesis 7a. One-tier boards in combination with a higher degree of firm internationalisation moderate and/or trigger a positive relationship between nationality diversity in non-executive boards and firm performance.
Hypothesis 7b. Two-tier boards in combination with a higher degree of firm internationalisation impair or restrain a positive relationship between nationality diversity in non-executive boards and firm performance.
By the same token, we frame the ED equivalent of hypothesis 7 as follows:
Hypothesis 8a. One-tier boards in combination with degree of firm internationalisation moderate and/or trigger a positive relationship between nationality diversity in executive boards and firm performance.
Hypothesis 8b. Two-tier boards in combination with a higher degree of firm internationalisation impair or restrain a positive relationship between nationality diversity in non-executive boards and firm performance.
3. D
ATA
&
M
ETHODOLOGY
3.1. SAMPLE SELECTION AND DATA COLLECTION
We start with a large data sample on board characteristics for 7947 listed West-European firms (see Table 21) during the period 2006-2011 retrieved from the BoardEx database. The level of our analysis respects company board average values per year. Accordingly, we do not need to derive the values for our main independent variable (see Section 3.2.2.1) from data at the individual level. The initial amount of company-year observations is 47713. We match these data with firm-specific data from ORBIS. ORBIS does not readily provide data on foreign sales or assets. Therefore, we derive all values for these firm internationalisation proxies (see Section 3.2.2.2) from the total amount of sales and assets per geographic segment. Unfortunately, there is no homogenous way as to how companies define these geographic segments. For instance, some firms mention the amount of assets or sales for subsidiaries per
country, while other companies state these values for specified regions, which typically
include multiple countries. Accordingly, we analyse all possible geographic segment denominations, in order to correctly allocate sales (or assets) to either foreign or domestic business establishments. In order to preserve data validity, we attach an “n.a.”-tag to the firm internationalisation variables of those companies that offer no clear distinction between the total amount of sales (assets) ascribed to domestic and foreign business operations.
A final panel-dataset of 47682 company-year combinations remains after deletion of double or empty occurrences. This dataset of 47682observations is inherently unbalanced. Particularly, we find a lot of missing observations for our main independent variable (see Table 20). Although this does not pose a problem for our regression estimation in terms of the total amount of remaining observations, it does affect representativeness per country (see Table 21). Nearly half of the observations are from the United Kingdom, while other countries are underrepresented. For example, after correcting for missing values on nationality diversity, we are left with less than 30 observations for Finland and Luxembourg. We account for this issue in our analysis of country-specific characteristics (see Section 4.2.3). In addition, we repeat the main regressions after exclusion of all UK observations (see Section 4.2.1 and 4.2.2).
In order to secure validity of the data, we verify the BoardEx and ORBIS statistics for a selection of 20 firms that are among the top 10 largest corporations of each country in our dataset (see Table 15). We compare the BoardEx data with the information provided through the respective company website (e.g. we compute the ED and NED nationality mix by using the data from the company website and compare it to the BoardEx data). We verify the same selection of firms from the ORBIS data with the information provided on the company website and Compustat. The data is deemed valid if the respective values do not deviate more than 10% from the verification data. In particular, we check whether the values for our dependent variables (see Section 3.2.1) are accurate, which happens to be the case.
3.2. VARIABLES
3.2.1. DEPENDENT VARIABLE
Firm profitability is the main indicator of firm performance. Indicators of firm profitability
can be classified into two broad sets: accounting-based and market-based measures. Kim, Hoskisson, & Lee (2015) propose three accounting-based measures of firm profitability: return on assets (ROA), return on sales (ROS) and return on equity (ROE). We abandon ROE as the proxy of profitability because this variable is sensitive to company's capital structure. With respect to ROS, since the sample consists of a large number of companies from different industries, the industry factor may cause substantial differences in total sales between the research targets and culminate biased coefficients. Oxelheim, et al. (2013) and Ujunwa (2012) use ROA, because it is directly related to the management’s ability to efficiently utilise corporate assets, which ultimately belong to shareholders. For the same reason, we elect ROA as the main accounting-based dependent variable. ROA is defined as the firm’s net income divided by its total assets (book value), at the end of the firm’s financial year (García-Meca, et al., 2015) (Kim, et al., 2015).
Many researchers tend to consider Tobin’s Q as the most reliable market-based indicator of firm performance (Alli, et al., 2010) (Arnegger, et al., 2014) (Eulerich, et al., 2013) (García-Meca, et al., 2015) (La Porta, et al., 1999) (Masulis, et al., 2012) (Oxelheim, et al., 2013). García-Meca, et al. (2015) argue that ROA and Tobin’s Q complement one another, because Tobin’s Q is a market-based measure, whereas ROA is an accounting-based measure. In
accordance with this last statement, we use Tobin’s Q to account for any differences between accounting-based and market-based measures. Differences between values for ROA and Tobin’s Q may, for instance, originate from the presence of foreign board members signalling a commitment to shareholder rights, less resistance to possible takeovers and less exposure to managerial entrenchment, all of which appeal to investors (Oxelheim & Randøy, 2003). This could initiate an upsurge in market value, and thus Tobin’s Q, while leaving ROA unaffected. We define Tobin’s Q as the market value over the book value of assets at the fiscal year end. The market value is also known as the enterprise value, which is calculated as:
Enterprise value = market value of common stock + market value of preferred equity + market value of debt + minority interest − cash and investments
We use either firm performance proxy to test the hypotheses regarding NEDs as well as EDs.
3.2.2. INDEPENDENT VARIABLES
3.2.2.1. Board Nationality Diversity
Following a well-established framework in literature, Harrison & Klein (2007) distinguish three fundamental types of diversity: separation, variety and disparity (see Figure 3). They argue that “failure to recognize the meaning, maximum shape, and assumptions underlying each type has held back theory development and yielded ambiguous research conclusions”. The diversity type “separation” refers to the composition of differences in position or opinion among unit members, mainly of value, belief or attitude (Harrison & Klein, 2007). Factors of team dynamics typically relate to this kind of diversity. Variety describes “the differences in kind, source, or category of relevant knowledge, background or experience among unit members” (Harrison & Klein, 2007). Example attributes are content expertise, educations and industry experience. On the other hand, disparity means that an unequal distribution exists in the proportion of socially valued assets (e.g. income, prestige or authority) or resources held among unit members (Harrison & Klein, 2007).
The main variable of interest in our study is nationality of board members. According to Harrison & Klein (2007), such a variable is dispersed across members who might be in one of K possible categories (i.e. the number of possible nationalities), and hence refers to diversity as variety. Harrison & Klein (2007) recommend the use of Blau’s index (also known
as Gibbs-Martin (1962) index) for measuring this type of diversity. Measures of the dispersion of a group over specified categories, such as Blau’s index, rather than central propensities, like the mean or proportion, are essential for understanding the impact of demography on organizational outcomes (Blau, 1977) (Pfeffer, 1983). In our situation, the equation for Blau’s index is:
𝐵𝐵 = 1 − ∑ 𝑝𝑝𝑘𝑘2,
where pk is the proportion of board members that represent a single nationality. This
concentration index for board nationality diversity as a variety ranges between zero and nearly one. In this structure, a zero indicates a nationally homogenous board, while a higher value represents more board members having a different nationality. The formula manages to capture the presence of foreign directors (i.e. similar to a dummy variable), as well as the
weight of international diversity on the board. We employ Blau’s index in either regression
related to executives or non-executives, where it is measured at the beginning of the year. Since ROA and Tobin’s Q are measured at the fiscal year end, this lagged structure ascertain that the explanatory variable precedes the dependent variable, even if the fiscal year ends prematurely. We mitigate causality concerns accordingly (Hambrick, 2007).
Figure 3 - Representation of the types and amounts of three types of diversity. Source: Harrison & Klein (2007)
3.2.2.2. Firm Internationalisation
Prior literature makes a distinction between structural, performance and attitudinal indicators measures of firm internationalisation (Dorrenbacher, 2000) (Heenan & Perlmutter, 1979) (Sullivan, 1994). Structural indicators represent the international involvement of a firm at a moment in time and can either be related to foreign activities of the corporation (e.g. the distribution of assets or affiliates across countries) or internationalisation of governance structures (e.g. the amount or proportion of shares owned by foreigners). Performance indicators illustrate to what extent the success or failure of business activities over time is thanks to operations in foreign countries (e.g. the amount of foreign sales or operating income abroad). Finally, attitudinal indicators can be used to show how firms perceive foreign countries and treat their foreign affiliates. For example, Perlmutter proposed four alternatives of headquarters management’s orientation towards foreign subsidiaries (Perlmutter, 1969) (Perlmutter & Heenan, 1974); Managers may be mainly home-country, host-country, regionally or globally oriented. We concentrate on the two former characteristics of firm internationalisation, since there are doubts whether attitudinal indicators can be measured reliably (Sullivan, 1994).
Foreign sales as a percentage of total sales (FSTS) is by far the most applied estimator for
firm degree of internationalisation. Sullivan (1994) expressed this fact by presenting that “each of the seventeen studies [used in his paper] designated this ratio as the independent variable… By no means did [he] selectively include only studies that set foreign sales to total sales as the independent variable.” FSTS is a performance indicator of firm internationalisation (Dorrenbacher, 2000), and elected for this study in this respect. Daniels & Bracker (1989) and Sullivan (1994) recommend foreign assets as a percentage of total
assets (FATA) in order to capture the structural attribute of the degree of
internationalisation. Accordingly, we adopt this ratio as well in a separate regression model. To explore whether the effects of board nationality diversity on ROA and Tobin’s Q vary with the extent of firm internationalisation, we add an interaction term (see Section 3.3.1) between the indicator for board nationality diversity and firm internationalisation. In summary, each of following terms is used as an alternative measure of firm internationalisation within this construct:
Performance indicator: Sales of foreign affiliatesTotal amount of sales
Structural indicator: Assets of foreign affiliatesTotal amount of assets
We incorporate the interaction into the regression model as an additional independent variable (see Section 3.3.1) to test hypotheses three and four.
3.2.2.3. Board structure
In Sections 2.3 and 2.4, we conjecture that the board structure may influence the impact of the relationship between nationalities compositions of executive as well as non-executive boards and firm performance. Those variations may, for instance, originate from differences in jurisdictions ordaining a one-tier or two-tier board structure. We utilise a separate variable that indicates whether the firm may feature a unitary or dual board system and operationalise it as an interaction term into the regression model (i.e. similar to our construct for firm internationalisation). Due to a lack of reliable data per year, we assume that the prevalence of board structures for each country did not change over the study period. This assumption may largely hold, since European board structures barely change over time, indeed (Adams & Ferreira, 2007) (Heidrick & Struggles, 2011) (Millet-Reyes & Zhao, 2010).
3.2.3. CONTROL VARIABLES
In this study we control for firm financial and governance characteristics and other director attributes, to reduce the probability that another variable drives the results.
3.2.3.1. Board Size
Jensen M. (1993) states that a higher board size engenders less effective monitoring, due to freeriding issues and a higher decision-making time. In confirmation, Harris & Raviv (2008) and Guest (2009) find that the number of non-executive and executive directors have a negative impact on profitability. On the other hand, other researchers demonstrate that large boards increase the amount of expertise and resources available to the firm and have a positive impact on firm value accordingly (Dalton, et al., 1999) (Finkelstein, 1993). As such, we control for board size, represented by the natural logarithm of the total number of directors on the board (García-Meca, et al., 2015) (Guest, 2009) (Masulis, et al., 2012).
3.2.3.2. CEO-and-Chair Duality
Empirical research indicate significant differences in the performance of companies with CEO duality and those without a duality role of the CEO (Goyal & Park, 2002) (Yanga & Zhaob, 2014) (Ugwoke, et al., 2013). A CEO that is also chairman of the board presumably compromises the governance system for control, while it may not always have the objective to maximize shareholder value (Krause, et al., 2014). For instance, he could abuse his position and conceal from the NEDs any problems triggered by executives. As the CEO may have the best “knowledge of the strategic challenges and opportunities facing the firm” (Jensen & Meckling, 1995) dual leadership could also benefit the firm (Yanga & Zhaob, 2014). Accordingly, we include a dummy variable that takes the value one if the CEO is the chairman of the board and zero otherwise (García-Meca, et al., 2015) (Masulis, et al., 2012).
3.2.3.3. Proportion of Independent Directors
Our second control for board independence entails the proportion of independent directors. Like García-Meca, et al. (2015), Masulis, et al. (2012) and Oxelheim & Randøy (2003), this research controls for the percentage of independent directors on the board. As the ratio of independent to total directors increases, so does the board’s knowledge, incentives, and abilities. This will foster better monitoring and advice (Harris & Raviv, 2008), and subsequently performance of the firm.
3.2.3.4. Firm size
Large firms obtain substantial advantages from economies of scale (Boyd & Runkle, 1993). There is also evidence indicating that the severity of managerial problems varies by firm size (Orser, et al., 2000). We account for firm size effects on firm performance by the logarithmic transformation of the market capitalization (i.e. equity market value) (García-Meca, et al., 2015) (Masulis, et al., 2012). Market capitalization is measured at the end of the fiscal year. We include this control variable for the regressions related to both NEDs and EDs.
3.2.3.5. Financial Leverage
The level of debt could affect the company’s financial performance (Brick, et al., 2006). It is more difficult for firms with high leverage to pay-off their debt compared to low leverage firms. Very high debt-to-assets ratios are often indicative for an increased probability of firm bankruptcy. Companies that are in default or threatened by bankruptcy are naturally associated with poor financial performance. Like Choi, et al. (2012) and García-Meca, et al. (2015) and Nielsen (2010), financial leverage was entered as a control and operationalized as the debt-to-assets ratio (at book values).
3.2.3.6. R&D Expenditures
Companies with more R&D activities can be less profitable presumably as a result of the immediate expensing of R&D expenditures (Masulis, et al., 2012) (Narayanan & Srinivasan, 2014). On the contrary, other researchers find that firms with an intensive investment strategy in R&D better satisfy customer needs and overcome competitive advances, and exhibit significant subsequent gains in performance (Gatignon & Xuereb, 1997) (Mizik & Jacobson, 2003) (Zhu & Huang, 2012). Measured as the firm’s research and development expenses scaled by its total assets at the fiscal year end, we control for R&D expenditures in the regression models for NEDs and EDs.
3.3. EMPIRICAL METHODOLOGY
We execute a multiple regression analysis with panel data to determine the net impact of board nationality variety on West-European firm profitability over a period of six financial years. Subsequently, we add an interaction term to the model that incorporates the firm’s degree of internationalisation.
3.3.1. OLSSPECIFICATION
First, we use an ordinary least squares (OLS) linear regression specification to test hypothesis one and two (see arrow 1 in Figure 4), relating to the effects of nationality diversity in boards on firm performance. The OLS regression model for testing this relationship is:
Firm performance = α + β0∗ board nationality variety + β1∗ board size + β2∗ CEO duality + β3∗ proportion of independent directors + β4∗ firm size + β5∗ financial leverage + β6∗ R&D expenditures + ε
where firm performance is either ROA or Tobin’s Q. We add a constant α, because otherwise the beta estimates can be biased and the R-squared negative (Brooks, 2014). Using Blau’s index, we compute board nationality variety in hypothesis one and two for NEDs and EDs, respectively. The other variables are control variables, while ε is the error term.
In hypotheses three and four (see arrow 2 in Figure 4), we expand our analysis on the anticipated effect of board nationality variety on firm performance. In particular, we want to test whether firm degree of internationalisation may alter the direction and/or magnitude of a potential relationship between board nationality variety and firm performance. Such moderating effects are generally indicated by the interaction of the independent and moderating variable in explaining the dependent variable. Accordingly, for hypotheses three and four, we add an interaction term between the foreign directors and the firm internationalisation indicator (see Section 3.2.2.2) to account for the degree of firm internationalisation. This results into the following OLS multiple regression equation for our performance-related indicator of firm internationalisation:
Firm performance =α + β0∗ board nationality variety +
β1∗ board nationality variety ∗ firm degree of internationalisation +
β2∗ firm degree of internationalisation + β3∗ board size + β4 CEO duality +
β5∗ proportion of independent directors + β6∗ firm size + β7∗ financial leverage + β8∗ R&D expenditures + ε
The unique effect of board nationality variety is represented by everything that is multiplied by board nationality variety in the model, i.e. β0 + β1 * firm degree of
internationalisation. β0 is now interpreted as the unique effect of board nationality variety
on firm performance only when “firm degree of internationalisation” = 0. This elucidation also implies that it is necessary to include the separate terms of the interaction in the regression model. If we abstain from doing this, we would implicitly assume that the regression coefficients of the main effect terms equal zero, which we do not have prior evidence of.
Hypothesis 5 (see Section 2.4) states our expectations with regards to cross-country differences. To account for national differences with respect to hypothesis 1 and 2, we add an interaction term between board nationality variety and board structure. Board structure either refers to the proportion of firms in a country that maintain a one-tier or two-tier board structure. As such, we test for two different effects. This construct helps to explore the influence of a country’s prevalent board structure on a relationship between board nationality variety and firm performance.
Firm performance =α + β0∗ board nationality variety +
β1∗ board nationality variety ∗ board structure + β2∗ board structure + β3∗ board size + β4∗ CEO duality + β5∗ proportion of independent directors + β6∗ firm size +
β7∗ financial leverage + β8∗ R&D expenditures + ε
By the same token, we cover the situation in which the board structure and the degree of firm internationalisation can jointly influence the regression of multinational boards on firm performance. The corresponding regression model exhibits a three-way interaction between board nationality variety, firm internationalisation (i.e. FSTS or FATA) and board structure:
Firm performance = α + β0∗ board nationality variety +
β1∗ Blau′s Index ∗ firm degree of internationalisation ∗ board structure +
β2∗ firm degree of internationalisation + β3∗ board structure + β4∗ board size + β5 CEO duality + β6∗ proportion of independent directors + β7∗ firm size + β8∗ financial leverage + β9∗ R&D expenditures + ε
Figure 4 – Main potential relationships that are investigated in this study
3.3.2. FIRM-FIXED EFFECTS REGRESSION
Figure 5 displays a number of potential relationships between the main variables of this study, which may contaminate the results of our regressions. The figure implies that endogeneity is a pertinent issue in our research, as in all studies on the subject of corporate governance and corporate boards. OLS regressions assume that the value of the error term is independent of explanatory variables. This supposition may not hold in the context of our analysis, because foreign directors are not randomly distributed over firms. It seems more likely that foreigners carefully consider which firm they join or shareholders wish to appoint them. Other factors that influence the firm’s demand for foreigner directors could also affect their eventual presence. For instance, high performance firms may more often appoint foreigners, which would suggest reverse causality. Another example of endogeneity concerns a CEO that is also chairman of the board, who may indulge in private benefits at the expense of firm value (Krause, et al., 2014). Typically, this CEO would try to avoid getting disciplined for his (her) bad practices. Hence, (s)he may encourage the appointment of foreign directors, because of their constrained monitoring ability (Masulis, et al., 2012).
We deploy two econometric approaches to mitigate endogeneity concerns. First, since we use a longitudinal panel dataset, there is a high possibility that firm-fixed effects are present. Firm-fixed effects regressions address endogeneity that emerges from omitted variables by
