Moving beyond the frontier: The role of internationalization in overcoming the descriptive social norm for board gender diversity

(1)

Moving beyond the frontier:

The role of internationalization in overcoming

the descriptive social norm for board gender

diversity

Master Thesis

Author:

Julian Hemmersbach

Student Number:

3796493

Date of Submission:

20-01-2020

Supervisor:

Dr. E. Mendiratta

Co-Assessor:

Dr. M. C. Sestu

Faculty of Economics and Business

University of Groningen

(2)

2

Abstract

Recent research has discovered a descriptive social norm for the number of women directors in American companies. This norm limits the number of women directors, thus constituting a barrier in the pursuit for gender equality in the boardroom and business in general. While antecedents for sticking to the norm have been discussed in literature, studies are yet to cover antecedents for firms to overstep the norm. Using a resource dependence theory-lens, this study examines three possible antecedents, all arising from a firm’s degree of internationalization, another understudied field of research in board gender diversity literature: Foreign institutional shareholders and foreign board members from countries with higher gender than the focal country, and a company’s share of international operations. Results from a sample of 178 American S&P 500 companies reveal that foreign institutional ownership from countries with higher gender equality increase a firm’s propensity to go beyond the descriptive social norm.

Keywords: Board gender diversity, Descriptive social norm, Resource dependence theory,

foreignness, institutional investors, foreign board members

(3)

3

Preface

“Have not the strongest And closest ties been bound so long

I’ve called upon them They remind me where I’m from

With deep conviction I am connected ‘cross the miles Without them nothing worth relying on.”

P. Labonte – Chiron, 2008.

(4)

4

List of Figures

Figure 1: Conceptual Model ... 24

List of Tables

Table 1: Sample Description ... 25

Table 2: Variable Description ... 28

Table 3: Descriptive Statistics ... 30

Table 4: Pairwise Correlations ... 31

(6)

6

List of Abbreviations

AR1 Autoregressive Within-Group Correlation Structure BGD Board Gender Diversity

CSR Corporate Social Responsibility DSN Descriptive Social Norm

EPS Earnings Per Share FBM Foreign Board Member FIS Foreign Institutional Investor

FR/TR Foreign Revenues To Total Revenues-Ratio GEE Generalized Estimating Equations Model GGGR Global Gender Gap Report

GICS Global Industry Classification Standard IB International Business

OR Odds Ratios

RDT Resource Dependence Theory RoA Return On Assets

(7)

7

1. Introduction

In Western societies, topics of gender equality have increased in importance in public debate. As a part of public life, this discussion has also found its way into the business world and public demands for more gender equality in companies have sparked an avid debate on the advantages and disadvantages of higher gender diversity in the workplace (see for example, Wong, 2019; Paul, 2019). A subject of discussion that has drawn particular public scrutiny here is the gender diversity on a firm’s board of directors. Through its high visibility, board gender diversity (BGD) is often an indicator and catalyst for the overall gender equality in a company and therefore receives particular public attention (Bilimoria, 2006; Chang, Milkman, Chugh & Akinola, 2019). The perceptions of what constitutes an adequate minimum level of gender equality in the workplace and in the boardroom differ among countries, which in turn influences the interaction of players from different countries, also in the business world (World Economic Forum, 2018; Terjesen, Aguilera & Lorenz, 2015; Mun & Jung, 2018).

In general, prior research suggests that one factor in a company’s decision to appoint more women to their board is the consideration to respond to international actors and increase the firm’s ability to survive in international operations. Mun & Jung (2018) for example find that BGD is an important criterion in the selection of investment targets among Japanese firms for foreign institutional investors (FIS) and that these investors are likely to maintain this pressure, once they have invested. Thus, in an economic environment that is otherwise characterized by low gender equality, firms can unlock and secure the financial capital of this investor-group by promoting women to the board. There is also evidence that engaging in international operations has influence on the number of women directors. The importance of gender equality in the societies of a firm’s foreign local staff for instance is a component that influences gender diversity-decisions, as it determines which gender is best suited to link to the new constituents (Ng & Sears, 2017). Lastly, firms increase BGD, because they value women’s ability to provide new perspectives to thus help create a more holistic picture in a complex international environment (Saeed, Yousaf & Alharbi, 2017). While these studies provide first inroads, BGD is still an underresearched topic in International Business (IB) literature, which may however change, given the general increase in research on the topic (Kirsch, 2018).

(8)

8 board can go. In a recent study, Chang et al. (2019) find that firms cumulate around a certain threshold-number of female directors. In their study, the majority of American firms would neither have more nor less than two women directors, in fear of losing its legitimacy with stakeholders if falling short of this number and losing other competitive advantages if going beyond it. This descriptive social norm (DSN) virtually caps the number of women on the board and makes it hard for women to obtain a director-seat once a company has hit this threshold. This is important, because the DSN’s existence thus signifies a salient barrier in the pursuit of gender equality in the boardroom. Yet, other studies evidence that companies going beyond it do exist, with a whole stream of research focusing specifically on the contributions a “critical mass” of three or more women can make to a board (Konrad, Kramer & Erkut, 2008; Joecks, Pull & Vetter, 2013). Consequently the question arises which factors lead companies to overstep the salient line and appoint more women directors than is expected by their domestic audience. While understanding these mechanisms is an important first step to overcoming this barrier and towards more indiscriminate employment practices (Bilimoria, 2006), to my best knowledge, no study in business literature to date has confronted this issue.

As the DSN for BGD constitutes a norm shaped by a firm’s domestic environment, a possible first starting point in tackling this question may be factors that do not originate in the home setting of the firm but have their roots in its international involvements. A firm operating in an international environment can often find itself exposed to different requirements than purely domestic companies (Sanders & Carpenter, 1998). In these cases, firms may need to reconsider their behavior to adapt to the altered business environment and remain in a position allowing them to attain vital resources (Hillman, Withers & Collins, 2009). Among others, international stakeholders can introduce new and foreign values and practices to a firm or even actively pressure them to adhere to foreign expectations, thus reducing the presence and relevance of local norms regarding the boards’ decisions (Oxelheim & Randøy, 2005; Mun & Jung, 2018). As noted before, international stakeholders can also generally influence a firm’s propensity to adapt BGD. In an attempt to pave the way for future research on DSNs in BGD and alleviate the general dearth of BGD-studies in IB-literature, this paper examines the research question:

How does a firm’s degree of internationalization influence its propensity to go beyond DSNs in BGD?

(9)

9 and BGD, the study focusses on actors from countries that put a higher value on reaching gender equality than the United States (US) and may due to their domestic values therefore positively influence gender diversity in American firms. By examining the impact of foreign board members (FBMs) and FIS from according countries, it complements previous papers on general BGD in this field, as previous research has not considered the gender equality in the FIS’ home country (Oehmichen, Rapp & Wolff, 2012). The final hypothesis then explores the possible influence overall internationalization of a firm’s operations may have on overstepping the DSN.

Resultwise, the paper demonstrates that FIS from countries with higher gender equality compared to the US have a positive influence on a firm’s propensity to go beyond the local DSN for BGD. Firms appoint more than two women directors to signal a commitment to corporate social responsibility (CSR) and values of gender diversity to the FIS and its clients. In doing so, they consider the FIS’ expectations and secure legitimacy. It also shows that, regardless of an FBM’s ability to connect firms to a wider pool of capable potentially female directors and the improved advice and counsel firms with internationalized operations can gain from women, these two factors have no effect on a firm’s propensity to overstep the DSN.

By regarding directors as linkages between firms and the resources in their environment, this study expands literature examining BGD through the lens of resource dependence theory (RDT). It allows examining the antecedents of going beyond a DSN from a firm-perspective and secondly permits to adequately capture interrelationships and benefits of different directors, like legitimacy-enhancement, the directors’ ability to give advice and counsel, communicate with constituents and secure resource access (Hillman et al., 2009).

To examine the research-question, the study uses a sample of 178 American firms from the 2013 S&P 500-index, with observations stretching over six years. The sample is used in conjunction with the US DSN for women on the board, which is two.

(10)

10

2. Literature Review

The term diversity relates to the recognition of categorical or qualitative differences in the characteristics of members of the same group (DiTomaso, Post & Parks-Yancy, 2007). It is based on structural and institutional differences, which can vary respectively within and among countries and regions (Lin, 2001; Kang, Cheng & Gray, 2007). Due to the dynamics the acknowledgement of these differences creates among group members, group heterogeneity has an influence on the execution of group processes and consequentially on the group outcomes (DiTomaso et al., 2007). While diversity as a concept in business literature also entails other forms, such as in-group heterogeneity in ethnicity, race, age, tenure and functional background, specifically the topic of gender diversity in firms has gained significance in recent years (Ruigrok, Peck & Tacheva, 2007). In a business environment, increasing gender diversity encompasses the inclusion of more women in the workforce. This especially pertains to the layers of management and governance- spheres that have traditionally been dominated by men. With regards to the board of directors, research separates into two different streams: In the first one, factors preceding board gender diversity (BGD) are investigated, while the second stream of literature analyses the consequences of BGD on firm outcomes (Oehmichen, et al., 2012). It is important to note that while both streams focus on two different concerns, they are intertwined in the sense that factors that are a consequence of BGD also exercise influence onto the creation of BGD. Simply said, decision-makers are likely to consider the influence a potentially gender diverse board may have on their organization in their selection of director-candidates (Kirsch, 2018). Under an RDT-lens, the next passage shall direct attention towards the antecedents of BGD, the field of research with which this study associates.

2.1 Advantages and antecedents of BGD

The reasons for why female directors have an impact on the work of the board and why henceforth companies choose to appoint female directors can be grounded in resource dependence theory (RDT) (Pfeffer & Salancik, 1978; Hillman, Shropshire & Canella, 2007). Under this lens also the present study seeks to analyze which external factors and which adjacent benefits of BGD could induce American firms to go beyond the DSN of having two women on the board.

2.1.1 Resource dependence theory

(11)

11 networks with other organizations, they are limited in their ability to obtain resources and depend on the actions of these other organizations in order to survive and prosper. Ambiguity about the actions of other players consequently evokes uncertainty concerning the own resource acquisition, thus threatening firm survival (Pfeffer, 1987). To reduce uncertainty and the concomitant costs and challenges, firms seek to gain and retain two types of resources from their environment: Such that decrease their dependence on other organizations, and those that increase the dependence of others on them (Pfeffer, 1981). Through the strive for superior resources, new interdependencies between actors may supersede old ones and full autonomy can never be reached. Inherent in the RDT is thus that firms must respond to their external environment and changes in it to survive and prosper (Hillman et al., 2009).

In their seminal work on the RDT, Pfeffer and Salancik (1978) propose the board of directors to be one of five options to create linkages with other entities to reduce dependencies and secure resources. The environment of the firm determines the features that a board must have to link the company most efficiently to critical resources (Hillman et al., 2009). While bigger boards can bring more resources to the table, it is also the type of director that is determined by the environment. Depending on the situation of the company and its strategy, some types of directors may be better able to provide links to relevant resources than others (Pearce & Zahra, 1992). According to Hillman et al. (2007), boards linked to the environment make three contributions to the firm: a) advice & counsel, b) legitimacy, and c) preferential access to external resources and enhanced access to channels for communication with key-constituents.

Boards can provide advice and counsel to the firm’s management, influencing strategy (Daily, Dalton & Cannella, 2003). Boards receive information from all parts of the firm and thus personally deal with the complexities of a firm’s environment (Hambrick, Cho & Chen, 1996). The personal and business networks board members bring to a company can provide them with unique information on the company’s environment, which may help in strategic decision-making (Barroso, Villegas & Perez-Calero, 2011). Based on this information, their perspectives and their personal and professional experience, they create the strategic frame in which the firm operates (Ruigrok et al., 2007). Boards further counsel executive managers in strategic questions (Kor & Misangyi, 2008), despite visible limitations to this, such as their obligation to also monitor executives in the name of shareholders (Daily et al., 2003).

(12)

12 & Jung, 2018). This gain in reputation also allows firms to differentiate themselves from competitors (Oh, Chang & Martynov, 2011). Companies can therefore satisfy stakeholder demands and gain their support and resources in competition with contending organizations.

Lastly, companies can unlock preferential resource-access and informative

communication channels to constituents, depending on their board composition. Through

interlocks with other boards, business networks that span important suppliers and customers, board members open channels that transmit valuable business intelligence. It further enables companies to access stakeholder-resources that would have not been attainable with a different board (Hillman et al., 2007).

2.1.2 Resource dependence theory and female directors

As argued above, different board members can link and provide firms with different resources. The next section investigates when, under the RDT, firms choose women directors because their characteristics positively affect the firm’s ability to link to certain resources and constituents.

(13)

13 and innovative than those of their competitors (Miller & Triana, 2009). Firms operating in complex environments with a variety of linkages are henceforth more likely to establish BGD. Empirical evidence for this exists through Hillman et al. (2007), who show that firms that operate in a complex multi-environmental setting of different product markets are more likely to have gender diverse boards, due to the breadth of perspectives and insights these enjoy through BGD and its linkages. Similarly, Saeed et al. (2016) find that family firms from emerging markets are more likely to entertain BGD, if they operate in complex international markets.

Similarly, it is harder for companies in these environments to find suitable director-candidates, because less director-candidates may exhibit the according special skills and perspectives in general (Daily & Dalton, 2003). In their search firms are driven to widen their recruitment pool to candidates from diverse backgrounds to ensure that the search for an optimal fit between director benefits and environment is not obstructed by clinging to traditional employment practices, thus increasing the probability of their appointment (Ruigrok et al., 2007).

Indeed, research suggests that firms create BGD where the specific female director-candidate can uniquely provide the skills a company lacks (Singh & Vinnicombe, 2004; Hillman, Cannella & Harris, 2002). Female directors are less likely chosen purely to their business background, but rather from occupational fields that grant them expertise in other areas of society (Burke, 1997; Hillman et al., 2002). Indiscriminate of the environment firms thus appoint female leaders predominantly from business support functions, such as law, or societal leaders from community-related fields like politics or higher education, because they desire advice and counsel complementary to that of their male directors from traditional management backgrounds (Hillman et al., 2002).

(14)

14 directors (Fryxell & Lerner, 1989). In contrast, studies in industries that are less dependent on groups for which gender diversity is an important social issue, have shown that these industries also rely much less on female directors (Adams & Kirchmaier, 2016).

Societal values and more abstract societal actors also drive the appointment of female directors. Since gender equality and gender diversity have become avid issues in social discourse, more visible companies with wider reach and bigger size appoint female directors to gain legitimacy and avoid falling under scrutiny by actors like the media (Farrell & Hersch, 2005; Hillman et al., 2007). Pressure from the broader public also compels other constituent groups to demand BGD (Mun & Jung, 2018). Institutional investors are likely to call for more women on the boards of companies they have invested in (Farrell & Hersch, 2005), to ensure that their investments do not fall under public scrutiny, which may potentially decrease their share prices (Chang et al., 2019). Additionally, as institutional investors, such as banks, insurance companies and government funds, they themselves have to achieve legitimacy with their own constituents. They therefore use their engagement in social change in the company as a tool to show their own clients that they take social issues serious (Oh et al., 2011). Therefore, if a company adds female directors to achieve legitimacy is depending on the characteristics of its stakeholder-groups and on how important these groups find a gender diverse board or gender diversity in general (Brammer, Millington & Pavelin, 2007).

Women directors can also be a channel of communication to external contingencies (Pfeffer & Salancik, 1978). Like with all directors, women directors are an entry-point for the perspectives and values of their social environment and networks into the firm’s boardroom activities (Jackson, 1992) Companies appoint heterogeneous boards to collect valuable feedback on the company’s actions and let the different reference groups’ input enrich the work of the board (Jackson, 1992; Hillman et al., 2007).

(15)

15 gender equality in its employment practices and career advancement for women is possible (Hillman et al., 2007). Next to female employees, communicating gender equality-commitments also attracts institutional investors. Institutional investors prefer socially responsible companies for investing and perceive BGD as part of such CSR (Mun & Jung, 2018). While some institutional investors also consider BGD as a prerequisite to invest into a company, corporate social issues such as BGD can particularly help foreign institutional investors come to a conclusion concerning the trustworthiness of their partner (Oh et al., 2011; Mun & Jung, 2018). Thus, organizations that seek or have established business-links to institutional or foreign investors and those aiming at retaining such investor groups create BGD to showcase commitment to socially responsible practices and secure preferential access to

their financial resources (Oehmichen et al., 2012; Mun & Jung, 2018).

Networks can also be a driver of BGD. These links communicate the idea of gender diversity as a resource-enhancing practice. It also simplifies implementation on the own board by providing information on women in leadership and implementation itself. If a board is for example linked to boards of firms with female directors, it is more likely to accept female directors and obtain BGD itself. When the other firm already entertains BGD, it also unlocks a readily available pool of potential candidates to the own organization. Board members themselves can therefore have an effect on the acceptance of BGD through their networks (Hillman et al., 2007).

(16)

16 level of gender equality or BGD, an actor from a country with higher gender equality may perceive the same level as insufficient and find it important to raise it. It is those actors that can bring a firm to increase BGD, for instance to connect to certain resources (Mun & Jung, 2018).

2.2 Twokenism - The descriptive social norm for BGD

The previously described relationships illustrate a plethora of circumstances in which firms hire female directors as suitable means to an end. Recent research however suggests that the degree to which companies engage in BGD may be relatively homogenous. A 2019-study by Chang et al. reveals that most large American companies’ degree of BGD accumulates around a number of two women on the board. Most firms do not deviate from this norm, with most of them neither having significantly more nor less female directors. A possible explanation for this is an explicit or implicit desire by the company to only acquire a minimum number of female directors, to suffice public demands, but -if not necessary- avoid going beyond those social minimum standards (Chang et al., 2019). Such a concept is called a descriptive social norm (DSN), defined as the average behavior of other groups that are similar to the own group (Prentice & Miller, 1993). Firms orientate towards DSNs for diversity to avoid negative public scrutiny from their stakeholders. In the case of BGD, firms orientate towards a number of women directors on the board. If an American company has at least two women on the board, it is safe from public scrutiny, because it has a number of women on the board that covers society’s expectations. Because firms are uncertain about the correct number of women directors, they attain the DSN by copying the behavior of the majority of their peers who have proven to be safe from the negative effects of public scrutiny, concerning a too low number of female directors (Chang et al., 2019). In this sense, companies engage in impression management to alter the public perception of themselves in their favor (Elsbach & Sutton, 1992) and achieve greater legitimacy with stakeholders.

(17)

17 et al., 2013). If the number of women on the board is below, their differences to the rest of the group are more visible, which impairs their ability to be accepted as partners on eye-level by their male counterparts and exercise their duties as usual (Kanter, 1977). However, if three or more women are on the board, the number of women has reached critical mass. Their presence is normalized and the board can now fully benefit from the advantages of BGD (Konrad et al., 2008). These studies show that having more than three women on the board can have visibly positive effects on firm outcomes, like financial performance and innovation (Torchia, Calabro & Huse, 2011). More importantly to this paper however, they are proof that firms exist that do go beyond the BGD-DSN and that there may be factors that influence the firm’s propensity to look for more female directors. To my best knowledge however, no research has yet examined the antecedents of going beyond Twokenism. In an effort to bring light into the dark, the next sections shall outline three plausible antecedents, based on the previously reviewed literature.

2.3 Firm internationalization, female directors and Twokenism

This study aims to reflect upon the increasing globalization of business and the possible influence of internationalization on a firm’s business practices. As previously mentioned, foreign influences can have a positive impact on BGD: Foreign institutional investors push firms for BGD to acquire legitimacy, internationalizing firms from emerging markets appoint women to handle the complexities of their new environment. and actors from different countries have different perceptions of what constitutes an adequate degree of BGD, which is likely to influence their decisions in their interactions with other actors (Mun & Jung, 2018; Saeed et al., 2016; Terjesen & Singh, 2008). To acquire a wider understanding of international influences on director-appointments, three different types of foreignness are chosen: 1) Foreign ownership, 2) foreign board members (FBMs), and eventually 3) foreign operations. According to Sullivan (1994), all three constitute important indicators of a firm’s degree of internationalization.

2.3.1 Foreign institutional ownership and BGD

(18)

(19)

19 in which their positive influence on company gender equality has highest visibility to their domestic audience: The board (Takeishi, 2007).

Due to their home-market affiliations and the need to obtain legitimacy with their domestic clients, FIS hold values and consider business practices to be implemented in their foreign holdings that are shaped by their home-countries (Ahmadijan & Robbins, 2005). FIS from countries that are closer to equality in their society may thus be more open towards women in leadership, because it is a more regular appearance in their domestic environment as well (Kirsch, 2018; Hillman et al., 2002). Free of social norms from the host country, these FIS may more likely consider the entire employment pool and more likely indiscriminately recommend female candidates to the board, regardless of if the current number surpasses local DSNs or not. Because FIS are large, influential shareholders, firms should follow suit, to avoid losing a preferential access to resources and retain legitimacy with the wishes of the FIS.

Furthermore, BGD is a component of CSR (Mun & Jung, 2018; Boulouta, 2013). FIS are long-term investors that are inclined to increase CSR beyond standard levels in the host-country, because -as a good management practice- it positively influences long-term performance (Hillman, 2015; Graves & Waddock, 1994; Oh et al., 2011). For BGD, a performance-impact is however only visible after adding more than two women to the board (Joecks et al., 2013). Because FIS from countries that are closer to gender equality are long-term investors, hold superior knowledge on business practices and may have less constraints towards women in leadership, they may be more likely to push for at least a third woman on the board, to yield positive long-term performance.

(20)

20 and its clients now perceive it as closer to gender equality than other local firms. Therefore, the FIS can demonstrate its positive influence to its clients, because the firm has demonstrated that it is now apparently more closely aligned to the values of the FIS’ domestic audience than other local companies are. Thus, through its successful impression management, the firm in return achieves higher legitimacy with the FIS and secures its resources.

Hypothesis 1: Firms with more shares held by foreign institutional investors from countries closer to gender equality have a higher propensity to go beyond Twokenism. 2.3.2 Foreign board members and BGD

FBMs are board members with a different national background than that of the firm on whose board they serve (Masulis, Wang & Xie, 2012). They are another group that can bring a similar kind of diversity to the board like women. Therefore, the two groups also share a number of characteristics in the potential qualities and drawbacks they bring to the board. Similar to women’s, FBM’s diversity is based on a different background (Ruigrok et al., 2007). However, in this case the different background is primarily seen as owed to the deviation in nationality (Choi, Sul & Min, 2012). For instance, just like women, FBMs bring a new set of perspectives and values to the board’s strategic decision-making process. They are usually recruited from foreign firms from which they introduce new progressive management styles to the local firm that are appropriate in their home country, but are not the norm in the local environment (Oxelheim & Randøy, 2005). This is related to the fact that they are usually recruited from foreign firms. Situated at a distant location, they are also not as embedded in the local business-culture and respective norms. They infer more of their judgement from linkages to home-country constituents (Estélyi & Nisar, 2016): Their decisions are influenced by the cultural values, set of beliefs and firm practices that they are used to from their domestic environment (Miletkov, Poulsen & Wintoki, 2016; Rhee & Lee, 2008). Nationality also influences the ethicality of an FBM’s decision-making, and the importance ethicality has in their decisions. Pearce II (2013), finds that non-US managers are more likely to emphasize ethicality in their judgement, while what they consider as ethical relates to the norms of the groups, here their nationality, with which they identify.

(21)

21 candidates than their American colleagues, as long as the woman director can provide the same link to new resources as other candidates. As FBMs seek to push matters that are important in their home environment forward in new environments as well, they may also be eager to actively promote the concept of gender equality in the American firm (Zhang, Kong & Wu, 2018; Oxelheim & Randoy, 2005; Estélyi & Nisar, 2016). Just like FIS, these FBMs will be free from local norms in their considerations, due to their different background and limited information on the American business-environment. They may therefore also be independent from the local BGD-DSN, more indiscriminately supporting female director-candidates, even if the local DSN has been met.

Additionally, countries with higher overall gender equality also usually have more women in leadership positions (World Economic Forum, 2018). This does not only pertain to women in management, but also in business support functions and community-domains (World Economic Forum, 2018; Hillman et al., 2002). As reported, firms predominantly recruit women directors from these fields, because they hold certain skills that may uniquely complement those of their male managers in board-tasks like advising and counselling. FBMs are often senior managers or social leaders with affiliations to other senior leaders in their home country (Hillman et al., 2002). This means FBMs as a linkage to their home country with higher gender equality represent an extended knowledge pool on how to install women in leadership positions. They provide their boards with preferential access to a bigger pool of female candidates, because their home societies have more capable females in leadership positions, making it more likely for the firm to find a qualified female candidate (Hillman et al., 2007; Tharenou, 2008).

Lastly, by entertaining FBMs, a company has already established diversity in its board. While increasing in numbers, FBMs are still a rare site on boards and boards that appoint them demonstrate that they are less restrained in their appointment-process by normative recruitment practices (Ruigrok et al., 2007). They are less afraid to overstep local recruitment-norms. Additionally, diversity may also contribute to a reduction of prejudice and bias towards new types of directors (Estélyi & Nisar, 2016). Combined with the greater openness of such FBMs towards female senior leaders, board-appointment decisions in companies with these FBMs may more likely concentrate more on the resource-fit director-candidates bring and less on the adherence to local norms based on a bias towards the different gender, as all involved parties are more used to unconventional, diverse board compositions.

(22)

22 women may have skills and characteristics suited for the company to connect to constituents’ resources.

Hypothesis 2: Firms with more foreign board members from countries closer to gender equality have a higher propensity to go beyond Twokenism.

2.3.3 Internationalization of operations and BGD

The extension of a firm’s operations to foreign countries can bring companies into the position to acquire new skills and knowledge and expand their customer base (Zhang, Ma, Wang & Wang, 2014). Companies can create economies of scale and connect to new resources (Kogut, 1985; Child & Rodrigues, 2005). It can thus be beneficial for a company to internationalize operations. However, the environment in the form of constituents, such as customers, competitors and regulators becomes more diverse and complex, sending a myriad of often conflicting information to the company (Brahm, 1994; Weick & Van Orden, 1990), all while the company becomes more dependent on these heterogeneous environments for its survival (Sanders & Carpenter, 1998). The board is considered a critical component in dealing with the information sent to the firm and in dealing with ensuing strategic complexities, because information from all company parts converge to its hands (Hambrick et al., 1996). It makes decisions based on the information and extracts new solutions for the problems at hand (Baysinger & Hoskisson, 1990).

(23)

23 dependence on international operations rises. If a woman can now offer the characteristics or skill-set a company needs, this constitutes a better fit to the company than sticking to the norm would. The company is more likely to hire her, irrespective of if the company is beyond the local DSN or not.

Characteristics and skills that are relevant in a complex environment, such as the international, are for example linkages to new diverse actors and opinions to improve the input into strategic decision-making. Indeed, Hillman et al. (2007) find that BGD, helps counter heterogeneity of the environment through a higher diversity in views among the directors. BGD also adds creativity to the boardroom and enriches strategic decision-making by heterogeneity in perspectives, values, experience and skills. Women can also connect to new resources and are likely to connect to different resources and constituents in foreign countries than men, thanks to their backgrounds (Adler, 1997). Thus, companies in international environments will appoint more female directors, because these can provide the benefits the company needs.

Additionally, because it is harder to find directors that can connect the company to resources and perspectives valuable in this uncertain and complex environment, rationally-operating firms with international operations are expected to consider a wider pool of possible candidates, which is not constrained by gender-affiliations (Daily & Dalton, 2003; Broome, Conley & Krawiec, 2011).

To sum up, the more internationalized the firm is, the less likely it is to be influenced by the American DSN. At the same time, the need to select a board suited to dealing with a complex and diverse environment rises, which is where gender diverse boards are proven to have an edge over homogeneous boards (Hillman et al., 2007). Firms will more indiscriminately search for new directors, because the employment pool is already small due to the requirements the environment sets for new directors to be effective. This will lower the barrier for firms to go beyond the DSN. Lastly, the impact of women on the board is notably more pronounced when three or more directors are women and their voices, perspectives and opinions are heard (Torchia et al., 2011, Kanter, 1977). A number of women beyond two is therefore particularly attractive for firms in a complex, international environment.

(24)

24 Hereinafter, the conceptual model is presented:

(25)

25

3. Methodology

3.1 Sample

The initial sample consists of the constituents of the S&P 500-index in the year 2013. To allow for a balanced sample that simultaneously provides a rich number of observations, I observed the companies of the year 2013 over a period of six years, namely from 2013 to 2018.

The S&P 500 provides a good representation of the United States’ public companies, because it assesses more than 80% of the American equity market and comprises firms from all of the eleven Global Industry Classification Standard (GICS) industry-categories. Containing 500 of the biggest American corporations, it displays companies that operate with both national and international scope (Brzenk, 2018) and are the likeliest to attract foreign investors (Mun & Jung, 2018). International Business literature has thus recurringly used it as a sample-source. Lastly, using high-profile American corporations as my sample-set also allows for the collection of detailed information across the different measures, broadening the pool of publicly available information on items such as firms’ board members, shareholders and performance.

The dependent variable rests upon the descriptive social norm for the number of women directors in the United States. I therefore removed all companies with foreign headquarters, as these may be subject to different descriptive social norms (Chang et al., 2019). In order to not count shareholder data twice, I focused on the company’s common stock and did not measure a company’s preferred stock, if existent. In the same manner, I excluded companies cross-listed at foreign stock exchanges. I excluded all companies for which data was not consistently available throughout the full six-year range, for instance as a result of acquisition by another company, bankruptcy, delisting or due to a company split that created two entirely new entities. After accounting for missing data in my variables, the final sample consisted of 178 companies. Each company has full data available for the entire six years and the final data set thus comprises 1068 observations.

Table 1: Sample Description

Directors Total Female Directors

Year Total Average Seats Total Percentage of Seats

(26)

26 3.2 Measures

3.2.1 Dependent variable: Female board representation

Reflecting the nature of my research question, the dependent variable seeks to express whether a company’s board of directors surpassed the integer-version of the social descriptive norm for female board members. To answer such a polar question, I used a binary format. Therefore, if a board had more than two female members, the dependent variable was coded as a 1 and otherwise as a 0. The necessary raw data for the total number of female board members per company was taken from the BoardEx database. BoardEx is a database that covers the profiles of global business leaders and the companies they work for. I used it due to its high data availability in this field (BoardEx, 2019).

3.2.2 Independent variables

1. Institutional ownership from more gender equal countries

The first independent variable is institutional ownership from more gender equal countries, measured as the percentage of shares held by foreign institutional investors from countries that are closer to overall gender equality than the US. I derived data on shareholder-type, -nationality and number of shares held per investor from the Thomson Reuters Eikon database (Eikon), which serves as my main source for all company data that is not board-related. In line with previous literature, I calculated percentages and defined institutional owners as pension funds, insurance companies, banks, securities firms, mutual and hedge funds, venture capitalists and private equity investors (Mun & Jung, 2018; Oh et al., 2011; Oehmichen et al., 2012). In this group, I selected the investors from countries whose societies exhibited higher gender equality than the US according to the ranking of the World Economic Forum’s “Global Gender Gap

Report” (GGGR) and computed their overall stock share in percent. The GGGR ranks countries

(27)

27

2. Board members from more gender equal countries

This measure computes the percentage of board members from countries closer to gender equality compared to the US. Once again, I use the GGGR-rank to sort nationalities. Data for board size and director nationality came from BoardEx. I complemented missing director nationalities by data hand-collected from annual reports, company websites, newspaper articles, online databases (NNDB, Prabook and Equilar) and recorded self-introductions by the directors in question. For cases in which such sources could not be found, I took consistent higher education paired with a consistent occupational history in the same country as a proxy. In all other cases, nationality is defined as country of birth.

3. Degree of internationalization of operations

To measure the internationality of a firm’s operations, I used the ratio of foreign revenues to total revenues (FR/TR), as this is a commonly used ratio in internationalization literature (Dörrenbächer, 2000; Ramsey, Barakat, Mitchell, Ganey, Voloshin, 2016; Ng & Sears, 2017). Foreign revenue is the revenue that is attained from all foreign operations, it is used in order to determine the degree to which a company relies on its foreign operations to drive its business. To retrieve the data I used Eikon. Firms for which the ratio could not be established, due to a lack of clarity about the domestic revenue for example, were excluded from the study.

3.2.3 Control variables

(28)

28 dummy variables coded as 1 if in the industry and otherwise as 0. I categorize firms along the eleven GICS-categories: Energy, materials, industrials, consumer discretionary, consumer staples, health care, financials, information technology, telecommunication services, utilities and real estate. I use industrials as the reference category, because it includes the most companies (180). I rely on GICS, because of its proven supreme consistency in assigning firms to industry-classes (Crossland, Zyung, Hiller & Hambrick, 2014; Bhojraj, Lee & Oler, 2003). Because data stretches across six years, binary dummies account for year-fixed effects. The reference year is year 1, 2013. All data is from Eikon, except for board size, which was retrieved from BoardEx.

Table 2: Variable Description

Variable Logit-Formula

Labels Description Value Source

Female Board

Representation FBR Indicator of female board representation surpassing the

descriptive social norm (>2 female boardmembers)

Dichotomous BoardEx

Foreign

Shareholders Shareholder Percentage of shares held by institutional investors from countries closer to gender parity than the US

Continuous Thomson Reuters

Eikon Foreign

Directors Boardmember Percentage of directors from countries closer to gender parity than

the US Continuous BoardEx, annual reports, newspaper articles, recordings, databases Foreign

Revenue ForeignRevenue Percentage of foreign revenue from total revenue Continuous Thomson Reuters Eikon

Board Size BoardSize Total Number of Board members Continuous BoardEx

Firm Size FirmSize Market capitalization, billion dollars Continuous Thomson Reuters

Eikon Firm

Performance Performance Return on Assets Continuous Thomson Reuters Eikon

Firm Age Age Years since incorporation Continuous Thomson Reuters

Eikon

Year Dummy Year Year of observation Dichotomous According

database of observation Industry

Dummy Industry Industry of company based on GICS Dichotomous Thomson Reuters Eikon

3.3 Analysis

(29)

29 between the usual fixed- and random-effects models, because a fixed-effects model cannot take into account the consistent values of the dependent variable (Crossland et al., 2014). Based on corporate governance literature that encountered similar complexities (Hillman et al. 2007; Crossland et al., 2014), I chose to use a Generalized Estimating Equations Model (GEE). GEE estimates population-averaged model parameters and standard errors, rather than subject specific ones. This makes it possible to adequately deal with the above-mentioned peculiarities (Allison, 2009). Because my dependent variable is dichotomous, I chose a logistic link function. A working covariance structure needs to be specified in order to run a GEE. Relying on another corporate governance study by Crossland et al. (2014), I chose an autoregressive within-group correlation structure (AR1). Such a correlation structure assumes that with measurements taken on a regular basis, the further apart in time measurements are taken, the weaker their correlation is (Shults, Sin, Tu, Kim, Amsterdam, Hilbe & Ten-Have, 2009). This seems appropriate, as the correlation between being above the descriptive norm in year one may be higher with being above the descriptive norm in year two than that in year six.

To check for heteroscedasticity, I convert the dependent variable into total number of female board members and run a regression with my variables, to plot residuals against the fitted values. In homoscedasticity, points should be centered around a horizontal line, because residuals should be independent of the fitted value. In my residual-plot, this is not the case. Residuals on average seem to be higher for larger fitted values. It thus suggests heteroscedasticity among the observations (See Appendix 2). To control for it, I run the GEE-model with clustered standard errors, a standard tactic in working with panel data with a small time series and many observed individuals (Field, 2014). Using clustered standard errors in a GEE-model furthermore has the advantage that standard errors remain valid, even if the aforementioned correlation structure is misspecified (Stata, 2013).

I create variance inflation factors (VIFs) for my original model to investigate possible multicollinearity (See Appendix 1). Multicollinearity occurs when an independent variable is significantly correlated with at least one other independent variable and both can predict the outcome of each other (Field, 2014). In this case, multicollinearity is ruled out, because values for VIFs remain between 1 and 2, significantly lower than the threshold (>10) (Field, 2014).

This results in the final model equation1_{for a logistic regression GEE-model with AR1-correlation}

structure and clustered standard errors: Logit(FBRit) = ln { 𝜋

1−𝜋} = αi + β1Shareholderit + β2Boardmemberit + β3ForeignRevenueit +

β4BoardSizeit + β5FirmSizeit + β6Performanceit + β7Ageit + β8Yearit + β9Industryit + εit

(30)

30

4. Results

4.1 Descriptive statistics and pairwise correlation matrix

Table 3 summarizes the descriptive statistics of my panel dataset consisting of 1068 observations of 178 companies over 6 years. This includes the observations, the mean, standard deviation, minimum and maximum for each variable.

Table 3: Descriptive Statistics

Variable Obs. Mean Std.Dev. Minimum Maximum

IV Female Board Representation 1068 .411 .492 0 1 DVs Foreign Shareholders 1068 4.359 3.636 0 23.218 Foreign Directors 1068 8.167 9.197 0 44.444 Foreign Revenue 1068 39.926 22.398 0 99.283 CVs Board Size 1068 11.093 1.913 6 19 Firm Size* 1068 55.699 93.807 .242 868.88 Firm Performance 1068 8.649 7.651 -5.42 169.92 Firm Age 1068 47.955 35.424 1 165

Note: *Market Capitalization in billion dollars. Industry and Year dummy results reported in Appendix 3.

(31)

31

Table 4: Pairwise Correlations

Variables (1) (2) (3) (4) (5) (6) (7) (8)

DV Female Board

Representation 1.00

IVs Foreign Shareholders

0.06* 1.00 Foreign Directors 0.06 0.07* 1.00 Foreign Revenue 0.01 0.05 0.16* 1.00 CVs Board Size -0.02 -0.29* 0.02 -0.04 1.00 Firm Size 0.07* -0.29* 0.04 0.10* 0.16* 1.00 Firm Performance 0.04 0.04 -0.02 0.17* -0.11* 0.18* 1.00 Firm Age 0.04 -0.15* -0.07* 0.03 0.24* 0.08* -0.04 1.00

* shows significance at the 0.05 level. Industry and Year Dummy results excluded from table to aid conciseness.

(32)

32 4.2 Regression model results

The next chapter describes the results of the logistic regression run with a GEE-model with clustered standard errors and AR1-correlation structure. Table 5 reports the results of the analyses, in which 1068 observations from 178 companies over 6 years were used.

Table 5: Regression Results

Variables Model 1 Model 2 Model 3 Model 4 Model 5

Foreign Shareholders 1.04*** 1.04*** (.019) (.019) Foreign Directors 1.00 1.00 (.008) (.008) Foreign Revenue 0.99 0.99 (.005) (.005) Board Size 1.01 1.01 1.01 1.01 1.17 (0.35) (.035) (.035) (.035) (.035) Firm Size 1.00 1.00 1.00 1.00 1.00 (.000) (.000) (.000) (.000) (.000) Firm Performance 1.01* 1.01* 1.01* 1.01** 1.01** (.010) (.009) (.010) (.009) (.008) Firm Age 1.00 1.00 1.00 1.00 1.00 (.003) (.003) (.003) (.003) (.003)

Industry and Year

Fixed Effects Included Included Included Included Included

Wald χ2 44.51*** 51.11*** 44.63*** 45.41*** 54.40***

N 1068 1068 1068 1068 1068

N firms 178 178 178 178 178

***p<0.01, **p<0.05, *p<0.1. Clustered Standard Errors in parantheses. Odds Ratios are reported.

Following Hilman et al. (2007), I use odds ratios (OR) in the result description and not coefficients to enhance readability and simplify interpretation. ORs depict how a one-unit increase in the independent variable changes the likelihood of the dependent variable. ORs bigger than 1.00 signify an increase in likelihood of the dependent variable if the independent variable increase. ORs smaller than 1.00 represent a negative effect. ORs of 1.00 mean there is no effect.

(33)

33 The second model tests the first hypothesis, which is that firms with more shares held by foreign institutional investors from countries closer to gender equality have a higher propensity to go beyond Twokenism. It was measured as the percentage of shares held by this group of shareholders. In this case, the test finds a significant positive influence (OR=1.04, p=0.008). Thus, the findings support hypothesis

1.

Hypothesis 2 predicts that a higher percentage of foreign directors from countries closer to gender

equality than the US has a positive influence on the dependent variable. However, test results from model 3 do not support this notion. Neither the odds-ratio nor the p-value suggest an effect on my dependent variable (OR=1.00, p=0.322). This does not change with the inclusion of all variables in the full model 5.

Ultimately, I hypothesized that the presence of the company in foreign countries would increase its probability to take a step beyond social norms. FR/TR measures this independent variable. However is the related OR not just insignificant, it is also below zero (OR=0.99, p=0.667). This is adjacent to the original prediction. Thus, hypothesis 3 is not supported.

Model 5 is the full model, combining all hypotheses. ORs of independent variables are in line with results from the prior models: First, they provide strong support for Hypothesis 1. It indicates a 4% increase in likelihood to go above the social norm when going 1 unit above the mean of percentage of foreign institutional shareholders from more gender equal countries. At the same time, hypothesis 2 and

hypothesis 3 are not supported.

4.3 Robustness checks

To investigate the robustness of the regression-results, I conducted four different robustness tests. In these tests, I checked for the effect of different variables, outliers and the effect of a new dataset, simulated by the cube root transformation of my current sample. Again, data for new variables in robustness test 1 & 4 draw from Eikon.

First I tested for the influence of new control variables for firm performance and firm size. I use the annually averaged earnings per share (EPS) of a company as an indicator for firm performance. In literature, EPS is a commonly used indicator for firm financial performance and is thus suited to replace RoA (Cordeiro & Sarkis, 1997). The secondary measure for firm size, total assets (in billion dollars), is also an established firm size-variable in research (Orlitzky, 2001). While firm performance becomes insignificant to the model, the results validate the outcomes with regard to my hypotheses: Strong support is given for hypothesis 1 (OR=1.06, p=0.007) whereas hypothesis 2 and hypothesis 3 are unsupported as insignificant.

(34)

34 root transformation is used, because the data involves negative values for firm performance. In contrast to log-transformation, in cube root transformation it is not necessary to add a constant and thus alter the structure of the dataset. The constant itself has an impact on the results, as different constants yield different transformation results. This may have an impact on the regression analysis (Field, 2014). The results of robustness test 2 yield mixed results. When singularly testing hypothesis 2 and 3, the tests reap marginal support for hypothesis 2 (OR=1.14, p=0.085), and a positive, yet entirely insignificant result for hypothesis 3 (OR=1.01, p=0.918). The full model however confirms my result from the regression. Regression-results for hypothesis 1 are continuously validated.

Robustness test 3 checks for outliers by winsorizing afflicted variables (Firm Size, Firm Performance and Firm Age). Outliers may influence results, because their distinct deviation from other observations may produce inadequately large residuals. Winsorizing adjusts outliers to the rest of the sample data (Field, 2014). Robustness test 3 corroborated the results of my regression analysis. Hypothesis 1 was supported, while hypothesis 2 and 3 remain unsupported.

(35)

35

5. Discussion

5.1 Interpretation of results

This research sought to investigate determinants for gender diversity on the board of directors, a topic that has experienced increasing scholarly activity in recent times (Kirsch, 2018). It looked upon factors influencing American businesses’ propensity to break through the Twokenism-DSN. Specifically, it focused on the little studied effect of international influences (Kirsch, 2018). To gain ground in this sparsely illuminated research field, three internationalization-related hypotheses were postulated: 1. Companies with a larger portion of institutional shareholders stemming from societies that are closer to reaching gender equality than the US will also be more likely to having overcome Twokenism. 2. A larger portion of FBMs from these countries will have a similar effect. And 3. The company’s degree of operational internationalization, will also positively effect a firm’s likelihood to go beyond the DSN.

Results for tests on hypothesis 1 demonstrated that FIS from more gender equal countries have a positive influence on a firm’s tendency to have more than two women directors. To achieve legitimacy with these players and accordingly secure preferential access to their financial resources, firms appoint a number of women beyond their domestic expectations. In this manner they can satisfy the FIS and its clients, who expect an engagement in gender diversity beyond the local standards. Thanks to information asymmetry and the high visibility of the board, this allows them to keep the rest of the company’s structure intact. The FIS will push for more BGD, because it may have positive effects for itself as well, like higher long-term performance of its invested companies, increased legitimacy with its stakeholders and differentiation from other FIS (Oh et al., 2011; Mun & Jung, 2018). Lastly it would be able to provide the company with a larger candidate pool (Hillman et al., 2002; Kirsch, 2018). This group of FIS can henceforth be regarded as a valuable driver in the dissemination of higher levels of BGD beyond local norms.

(36)

36 hypothesis and did not reveal an influence of FBMs from more gender-equal countries on BGD surpassing the DSN.

An explanation may lie in the differences in characteristics between foreign board members and female board members. Nationality constitutes a different form of diversity than gender. In their 2013 study, Nielsen and Nielsen (2013) find that the different forms of board diversity have different effects on firm performance and that these effects are dependent on the firm’s strategy and the environment the firm is facing. Research has found firms with internationalizing operations are specifically recruiting more FBMs (Greve, Nielsen & Ruigrok, 2009; Hitt, Tihanyi, Miller & Connelly, 2006). This enables them to connect to important resources and incorporate new viewpoints that are not only new and multiculturally diverse, but, by hiring local directors, also specifically tailored to the new environment the firm expands into (van Veen et al., 2014). Other forms of diversity may prove superior for other strategies, such as functional diversity for product markets (Michel & Hambrick, 1992). As reported, firms that undertake product diversification-strategies often hire women for their additional advice and counsel and women directors are often hired in environments in which their specific skills, learnt in a different functional background (e.g. business support or public leader), are valuable for the firm (Hillman et al., 2007; Hillman et al., 2002; Konrad et al., 2008). Thus, even if the FBM holds strong values for gender equality and connects the firm to a richer female candidate pool at home, this pool will remain untapped, because in its current environment, the company may not need the functions, skills and perspectives of an effective group of women directors specifically. More FBMs may be an expression that other forms of diversity are more effective (Nielsen & Nielsen, 2013). Current board members will notice this lack of fit between female directors’ resources and the needs of the company. RDT entails that firm decisions on board diversity are not just value-bound, but are a consequence of considerations about the potential benefits and costs of diversity for the firm (Estélyi & Nisar, 2016). Thus, as responsible for strategy and firm-performance, FBMs will not be any more or less likely than other directors to have an effect on the propensity of appointing an above average number of women directors.

(37)

37 If FBMs perceive even BGD at or below American DSN-level as a business practice that is above the norm at home, they may see BGD as sufficient. Thus, they may be content already with two women on the board.

Hypothesis 3 argued for a positive effect of the internationalization of a firm’s operations on the propensity of such a firm to overcome Twokenism. Firms would use beyond-average numbers of female directors to establish a number of new perspectives and external ties allowing them to reduce uncertainty and new complexity and would be more willed to do so, because the relevance of the local norm would decrease with an increase of the international share of operations (Oxelheim et al., 2013). However, the results were not only insignificant, but also opposed in their effect to the hypothesis.

One explanation may be that firms that internationalize do not place as much priority on the career advancement of women, which in turn may lead to a lower number of women on the board of directors. Ng and Sears (2017) uncover that firms with a higher share in revenues from international operations have a lower percentage of women in management positions, both in their home country and in the host countries. They explain this with existing gender biases against female managers in potential host markets and the ensuing difficulty for women to partake in career-advancing activities in an international company, such as an abroad assignment in such a vital country. In line with RDT, women are then less able to connect with resources in the international environment than men and are seen as less valuable resources. This leads to a greater demand for male managers and makes it in turn harder for female managers to reach higher management positions, because they cannot access the same career-advancement opportunities as men. If women cannot advance in their fields in the company, this also has negative consequences on their reaching of the most senior management positions in their company, namely the C-Suite and the board of directors (Kirsch, 2018). This may explain the slightly negative, yet insignificant effect: Firms rather appoint male directors that connect better to the environment. The internal supply of capable male managers is bigger and less females have a chance to join the board.

(38)

38 negative behavioural aspects of diversity overweigh the positive effects: In a fast paced-environment, the time to resolve different perspectives may simply be lacking and the information amassed by top teams simply become too great to be processible (Carpenter, 2002; Meyer, Mudambi & Narula, 2011). Fold-lines between female and male directors emerge more strongly, because views are more likely to overlap within these two groups than between them (Triana, Miller & Trzebiatowski 2014). This decreases team cohesion and thus the team’s ability to reach a consensus about the best strategic actions (Knight, Pearce, Smith, Olian, Sims, Smith & Flood, 1999), which in turn decreases board’s ability to set a long-term strategy (Triana et al., 2014). Firms which have internationalized and who have already experienced the effect of one or two women on the board and their differing opinions, will not increase the number of women anymore to not torpedo decision-making-swiftness and -efficiency. They will rather place a premium on more homogenous groups, which will come to decisions faster and generate less conflict (Carpenter 2002; Triana et al., 2014).

Another explanation may be based on institutional theory and the causes for why organizations embrace a DSN: Firms adopt DSNs to mimic the behavior of other successful companies in the same field. Because the competitors are successful, their behavior is seen as related to their achievements. Thus, in an effort to become as successful as them, other firms copy their patterns. This furthermore enables them to achieve a similar level of legitimacy with the environment (DiMaggio & Powell, 1983). This is especially the case in environments and situations in which firm actions and outcomes are connected through ambiguous processes. Uncertainty makes the firms copy each other, even more so, if the firms are highly visible (Chang et al., 2019). International firms are more visible than others and international operations pose an uncertain environment (Strike et al., 2006; Sanders & Carpenter, 1998). Irrespective of if more diversity would yield better resource-connections, firms that internationalize their operations will still mimic the BGD-levels of other American firms because this behavior has already proven to sustain and not to damage the organization in an uncertain environment.

5.2 Theoretical implications

(39)

39 provides a first entrance, which future research may follow. While previous research proved that FIS could bring a company to increase its BGD in general, this study proved that specifically FIS from countries with higher gender equality can induce firms to go beyond what is domestically expected to retain legitimacy with the shareholder and secure access to its financial resources.

Simultaneously, it advances the sparsely researched relationship between foreign influences and a firm’s BGD. Here it is one of the first to consider the potential influence of gender equality in the country of origin of different actors in its theorizing. It contributes to research by showing that more FIS from countries with higher gender equality increase a firm’s propensity to go beyond a DSN. It suggests that shareholders from countries with higher gender equality are prone to pressure American companies to increase BGD, in order to differentiate themselves and convince their domestic clients of their positive influence on American companies. The evaluation of their influence as positive, is based on their home-values, as for example higher gender equality. Based on RDT, I theorized that firms are likely to follow suit, to not lose access to valuable resources. Because firm and investor both prefer a channel with maximum visibility to communicate their gender equality-commitment to their respective stakeholders, they are more likely to appoint at least three female directors to the board. With its grounding in RDT, the study furthermore heeds the call by Hillman et al. (2007) to advance research on BGD under this particular lens. Its results yield a new perspective on what steps companies are willed to undertake to achieve legitimacy with its constituents.

5.3 Practical implications

Beyond the theoretical horizon, this study also has practical implications for managers. It proves that when international influences are involved, this can significantly alter the effectiveness of certain board structures in dealing with the environment. It thus requires more sophisticated answers than a well-established one-size-fits-all approach: While orientating towards the DSN is beneficial for companies in a national environment, foreign influences may make it more advisable for senior management to leave beaten parts and create governance structures that accurately reflect the individual environmental complexities the firm encounters. Leadership needs to consider which measures optimally connect the company to necessary resources and henceforth free their minds from constraints created by minimum social norms.

Moving beyond the frontier: The role of internationalization in overcoming the descriptive social norm for board gender diversity