The relationship between TMT gender diversity, a female
CEO and financial performance: international
diversification as the mediating mechanism
Manon Allersma
University of Groningen and Newcastle University Business School, Dissertation Dual Award for MSc DDM Advanced International Business Management and Marketing
Key information Submission date: 06-12-2020 Assignment deadline: 07-12-2020 Manon Allersma 200421904 (Newcastle) S2963981 (Groningen) Dissertation supervisors: Dr. E. Mendiratta Prof. I. Munro University of Groningen
Newcastle University Business School Faculty of Economics and Business
Abstract
The purpose of this paper is to examine the relationship between gender diversity/a female CEO in top management teams (TMTs) and financial performance mediated by international diversification. The Upper Echelon Theory (UET) has been used as a theoretical frame to draw several ideas about female top executives. Based on this theory, there will be two approaches for the unit of analysis; the CEO and the TMT. To examine whether this mediation mechanism exists, a three-step mediated regression analysis has been conducted, based on a sample of 783 companies, headquartered in Sweden. The results of this study do not find a significant relationship between gender diversity/a female CEO and financial performance, and there is no mediating mechanism of international diversification. However, the outcomes show that, even though not formally hypothesized, there is a negative significant relationship between female CEOs and international diversification. This research contributes to the discussion about the business case for gender diversity. The results of this study challenge this argument and support gender diversity in TMTs based on more ethical arguments. Furthermore, the results imply that when organizations are highly focused on diversifying internationally, they are not advised to hire a female CEO.
Keywords: top management team, gender diversity, CEO, international diversification, financial performance
Acknowledgements
Table of contents
1. Introduction ... 4
2. Literature review and hypothesis development ... 7
2.1 Theoretical overview – Upper echelons perspective ... 7
2.2 International diversification and financial performance ... 9
2.3 TMT gender diversity and financial performance ... 13
2.3.1 Female TMT members ... 15
2.3.2 Female CEOs ... 18
2.4 TMT gender diversity and international diversification ... 19
2.4.1 Female TMT members ... 20
2.4.2 Female CEOs ... 23
3 Methodology ... 25
3.1 Sample and data collection ... 25
3.2 Measurement ... 26 3.2.1 Independent variable... 26 3.2.2 Dependent variable ... 27 3.2.3 Mediator variable ... 27 3.2.4 Control variables... 27 4 Results ... 30
4.1 Descriptive statistics and correlations ... 30
4.2 Regression analysis ... 31
5. Discussion ... 34
5.1 Implications ... 37
5.2 Limitations and future research ... 37
6. Conclusion ... 40
References ... 42
Appendices... 55
Appendix A: Effect size ... 55
Appendix B: Measurement variables ... 55
1. Introduction
“Why do so many incompetent men become leaders?” In 2013, Thomas Chamorro-Premuzic wrote a Harvard Business Review on the topic of leadership between men and women and he found that female leaders are equally, or even more competent to become adequate business leaders (Chamorro-Premuzic, 2013). If this is the case, what is the reason that top management positions are dominated by men? In 2018, 76 per cent of the chief executive officers (CEOs) were men (Ahmed, 2019) and for executive managers, in general, the numbers of female top management team (TMT) members are still low. From an international perspective, it is interesting to see that Japan still has one of the lowest ratio of females in management positions with a percentage of 13,2 per cent. Whereas Sweden has the highest ratio of female managers with a percentage of 39 per cent (White Paper, 2018). Even though workforces are becoming increasingly more diverse, regarding dimensions such as culture, age, and gender. The results over the past years do not show that such diversity is represented as abundantly in the upper echelons (i.e. TMTs), as it is for the lower levels in corporations (Baxter and Wright, 2000; Zhang and Qu, 2016).
organizational performance effects due to female leadership (e.g. Shrader, Blackburn and Iles; Lee and James, 2007; Adhikari, 2012).
Hence, is there an evident performance advantage for organizations with female representation in strategy making roles in a CEO position and on TMTs? And if there is an evident relationship, how does this occur? Despite years of research on this topic, the answer to this question remains inconclusive (Jeong and Harrison, 2017). Over the years, there has been a rise in the management literature dedicated to better comprehend the role of TMTs. This stream of research is derived from the Upper Echelons Theory (UET) of Hambrick and Mason (1984) and these scholars have studied the role of TMTs in important organizational decisions (e.g. Bantel and Jackson, 1989; Finkelstein and Hambrick, 1990; Jeong and Harrisson 2017). Most studies on TMTs have searched for factors that might explain under what conditions TMT diversity tends to affect a company’s performance (Carpenter, 2002; Joshi and Roh, 2009; Nielsen and Nielsen, 2013; Roberson, Holmes and Perry, 2017). Much insight has been provided into the workings of TMTs due to the focus on moderators, however, the identifying mechanisms that link TMT diversity to organizational performance have received little attention (Barrick, et al., 2007; Hambrick, 2007; Joshi, Liao and Roh, 2011). Jeong and Harrisson (2017) studied gender diverse TMTs and risk-taking indicators. They found that there are still numerous sets of risk-related strategic choice variables that could be studied and used as mediating variables, such as innovation intensity, mergers and acquisitions, strategic dynamism or international diversification. This study will focus on international diversification as the mediating mechanism.
research regarding the relationship between gender diversity and international diversification is mostly conclusive and scholars argue and find empirical evidence that female representation in TMTs is negatively related to international diversification (e.g. Fisher, Reuber and Dyke, 1993; Orser et al., 2010; Levi and Zhang, 2014; Chen, Crossland and Huang, 2016; Ossorio, 2020). Their rationale is based on that women are less risk-taking (Vandergrift and Brown, 2005; Wei, 2007), less over-confident (Lundberg and Punchchar, 1994; Barber and Odean 2001) and have less aggressive investment policies (Chen, et al., 2019) than their male counterparts.
Consistently with Jeong and Harrisson (2017), who argue that improved financial performance can be explained by the reduced strategic risk-taking of female executives as a mediating mechanism. This research will argue that there is a positive relationship between a female CEO/female representation in TMTs and financial performance and this beneficial relationship is negatively mediated by international diversification. Moreover, in light of the above-mentioned limitations of existing research, this dissertation will contribute to the current literature by trying to remove uncertainties regarding the mixed findings on the relationship between gender diversity and financial performance. Additionally, it extends prior research by examining if international diversification is a mediating mechanism for the relationship between gender diversity and financial performance. Hence, this paper aims to extend this stream of research by addressing two related questions: (1) What is the relationship between a female CEO/female representation in TMTs and organizational financial performance? and (2) is this relationship mediated by international diversification? A distinction will be made between female CEOs and female representation in TMTs.
labour union contracts that guarantee women’s equal career opportunities, a subsidized network of day-care centres and a parental leave policy for both man and women, which creates gender equity at home (Haas and Whang, 1999; Thomas and Hildingsson, 2009). By removing these obstacles, Sweden ensures that women have the same opportunities to become active business leaders (Johnson and Tunheim, 2016). Moreover, if there is a relationship between the variables, it will be detected in a sample from Sweden, as it is most likely that there is gender diversity in the TMTs.
Overall, this research aims to make three contributions. First, this study contributes to the discussion of TMT gender diversity, whether it is positively related to the financial performance of an organization, thus if it supports the business case for diversity. However, the results of this study challenge this positive relationship and argue for more ethical based reasons to support gender diversity in TMTs. Second, this research extends existing research by trying to identify international diversification as a mechanism that links TMT diversity to organizational performance. The results of this study indicate that international diversification is not a mediating mechanism. Third, this research provides a thorough overview of prior literature and highlights key outcomes on the different relationships. Therefore it assists scholars in deciding whether there is a need for additional research and if adjustments in methodology are necessary.
The dissertation will be structured as follows. First, the issues regarding the inconclusive results on a gender diverse TMT, international diversification and financial performance will be addressed in section 2. Besides, the hypotheses will be formed in this section. Furthermore, in section 3 the methodology of this research will be described and the results of this study will be reported in section 4. Additionally, in section 5 the results will be discussed and the limitations, future research and implications of this study will be provided. Finally, section 6 will summarize the conclusions.
2. Literature review and hypothesis development 2.1 Theoretical overview – Upper echelons perspective
decisions, due to their characteristics, cognitive biases, and bounded rationality (Nielsen, 2009; Hoffmann and Meusburger, 2018). Additionally, Hambrick and Mason (1984, p. 193) note that “organizational outcomes – both strategies and effectiveness – are viewed as reflections of the values and cognitive bases of powerful actors in the organization.” Overall, the composition of a TMT is of great importance to organizational outcomes. The UET will serve as the underlying framework of this research since it gives a logical theoretical foundation for linking organizational outcomes with TMT gender diversity.
CEO perspective separately when proposing factors that shape an organization’s international diversification strategy (Jaw and Lin, 2009).
2.2 International diversification and financial performance
Even though the existing literature on international diversification may be one of the largest bodies of work in business strategy, outcomes have been mixed (Tallman & Li, 1996). In this section, the main findings and key issues of several articles regarding the relationship between international diversification and financial performance will be summarized.
International diversification may be defined as: “expansion across the borders of global regions and countries into different locations or markets” (Hitt, Hoskisson and Kim, 1997, p. 767). Managing different operations in several foreign countries and creating strategies for the international marketplace has become an important task for executives. Moreover, international diversification is a complex TMT-level strategy that offers a successful alternative to product diversification and other strategies (Hitt, Tihanyi and Miller, 2006). Organizations that focus on international diversification have several motives, including performance improvements, knowledge acquisitions, and location advantages (Hitt, Hoskisson and Kim, 1997). Due to access to foreign institutions, stakeholders, and resources, international diversification provides new opportunities for value creation. Additionally, this kind of diversification allows corporations to exploit growth opportunities in the product markets of the host countries, to gain unique knowledge, and to emphasize their core competencies (Hitt, Tihanyi and Miller, 2006).
TMT members consider the diversification of organizational activities to foreign countries a complex strategic decision. This complexity could be due to the increased threats and opportunities that managers endure in the international markets (Tihanyi, et al., 2000). In prior literature, there has been concurrence that international diversification does affect organizations. However, these empirical results have varied regarding the outcomes of these influences (Tallman and Li, 1996; Hitt, Hoskisson and Kim, 1997). Earlier studies started by researching differences in the performance of domestic and multinational firms (Vernon, 1971; Brewer, 1981; Shaked, 1986). However, later research emphasized more on comprehending the nature of the relationship between the two (Hitt, Tihanyi and Miller, 2006).
On the one hand, there are some downsides to international diversification and it does not come without risk. When expanding to foreign countries the level of uncertainty increases. Costly mistakes could arise when organizations do not prepare for misunderstanding factors, such as different regulations, consumer tastes, or access to distributors (Mitchell, Shaver and Yeung, 1992). Besides, due to the increased complexity and size, organizational problems may arise, e.g. due to increased monitoring costs (Fatemi, 1984; Geringer, Beamish and daCosta, 1989). Additionally, research also found that performance declines are associated with increased international diversification. For instance, Michel and Shaked (1986) have reported that performance is higher for domestic organizations than for multinational companies. Furthermore, Lee, Ahn and Shin (2016), reported that international diversification in the construction sector in Korea undermines the financial viability of organizations. Due to the increase in their debt and decrease in their current ratio. Thus, some scholars argue that investing in international diversification does not always result in value creation because of the firms’ liability of foreignness – “the costs of doing business abroad that result in a competitive disadvantage for an MNE subunit” – (Zaheer, 1995, p. 342). Nonetheless, the level of the liability of foreignness depends partly on the structural dimensions of the foreign markets. For instance, operating in countries with different levels of development, institutions, or different cultural values. Besides, the skills as a company to manage the entry into, and the operations in the foreign market, are of great importance as well (Hitt, Tihanyi and Miller, 2006).
On the contrary, several scholars also found positive empirical results. For example, Hughes, Lounge and Sweeney (1975) reported that multinational firms delivered higher returns than domestic organizations. Furthermore, Kim, Hwang and Burgers (1993) and Hitt, Hoskisson and Ireland. (1994) reported that international diversification results in higher firm performance than domestic product diversification. Additionally, Tallman and Li (1996) and Delios and Beamish (1999) stated that the performance of an organization and the scope of international diversification are positively related due to the location-based advantages and the economies of scale. This increases market power and opportunities, and it diversifies risk (Kogut, 1985; Kim, Hwang and Burgers, 1993). Also, Horta, et al. (2016) reported that international diversification is positively related to financial performance, even though this positive impact was only statistically significant for contractors in Spain.
international diversification. These relationships resemble a U-shaped (e.g. Lu and Beamish, 2001; Zhou, 2018); inverted-U-shaped (e.g. Ali, et al., 2016; Tang, Tang and Su, 2018) and S-shaped (e.g. Riahi-Belkaoui, 1998; Lu and Beamish, 2004) curve. In a U-S-shaped curve, international diversification initially may decrease an organizations’ profitability due to the difficulty of unrelated strategies. However, when an organization learns about the environment of the host country and the foreign market becomes familiar, the profitability begins to rise (e.g. Lu and Beamish, 2001; Ruigrok and Wagner, 2003; Wu and Wu, 2017; Zhou, 2018; Feng, et al., 2019). For instance, Feng, et al. (2019) measured the benefits and costs of Chinese companies regarding their internationalization. They found that in the early stage, international diversification entails high costs since the organizations have little knowledge and understanding of the foreign country and they lack experience when operating internationally (Kirca, Fernandez and Kundu, 2016). Besides, they need to conquer the aforementioned uncertainties such as institutional obstacles, cultural disparity and regional differences (Hitt, Tihanyi and Miller, 2006; Feng, et al., 2019). Consequently, firms may struggle to overcome the costs of diversifying their operations internationally. Even though organizations may gain profits in the first stage, it may not be enough to counterbalance the costs (Prange and Pinho, 2017). However, when internationalizing there are several advantages. For instance, companies can keep improving their international competitiveness and keep enhancing their learning ability (Assaf, et al., 2012). Besides, they can use the gained knowledge from one country to other subsidiaries (Cardinal, Miller and Palich, 2011). Finally, they can slowly enjoy the advantages of economies of scale and scope. Moreover, Feng, et al. (2019) found that the enterprises in their sample are still in the early stage of internationalization. Nonetheless, the U-shaped curve suggests that when the company passes the threshold, they will be able to benefit from the success factors of internationalization. Other examples of scholars who found U-shaped curves are the following; Zhou (2018) revealed a U-shaped relationship in large firms when he tested the international diversification and performance relationship in China. Additionally, Lu and Beamish (2001) documented a U-shaped curve in a sample of small- and medium-sized enterprises (SMEs), who are engaged in international diversification. In both cases, the large firms and the SMEs encounter profitability decreases in the initial stage of international diversification, but once they overcome the liability of foreignness the organizations recover.
access to resources and economies of scale and scope. Eventually, in a later stage, when it reaches the “internationalization threshold”, the increased transaction costs will start to decline regarding their profitability and performance (e.g. Gomes and Ramaswamy, 1999; Ali, Hashmi and Mehmood, 2016; Brida, et al., 2016; Tang, Tang and Su, 2018). Gomes and Ramaswamy (1999) argue that when an enterprise starts, it only has a handful of subsidiaries. Meaning, they can manage resource flows and information across its network quite easily and are thus able to enjoy the benefits of internationalization. Nevertheless, when enterprises continue to expand, managing resource and information exchanges among worldwide subsidiaries may result in costs which are increasing faster than their returns. Several scholars support this view. For instance, Tang, Tang and Su (2018) reported an inverted U-shaped relationship between the performance of Chinese listed companies and the research and development firms that are engaged in international diversification. Furthermore, Capar and Kotabe (2003) supported the outcome of the inverted U-shape, as they investigated the relationship between the international diversification of German service firms and organizational performance. Moreover, an inverted-U-shaped curve suggests that international diversification leads to increased organizational performance up to a threshold. After reaching this threshold, the complexity and corporate costs related to managing internationally dispersed subsidiaries will begin to outweigh the benefits of further international diversification (Contractor, Kundu and Hsu, 2003).
positive at moderate levels but negative at high and low levels. Because initially the costs and liabilities are lessened by experimental learning and the organizations’ benefit from economies of scale and scope. However, when international diversification increases, coordination and governance costs rise (Hitt, Tihanyi and Miller, 2006).
Moreover, the relationship between international diversification and performance has been a popular research topic among scholars (Hitt, Tihanyi and Miller, 2006), and as the information above indicates, the results on this relationship have been very mixed and inconclusive. The review of Hitt, Tihanyi and Miller (2006) shows that this particular relationship tends to be context-dependent, regarding the samples that scholars use to perform their research.
2.3 TMT gender diversity and financial performance
Long-term financial performance is the most frequently studied outcome in the literature on females in the upper echelons (Jeong and Harrisson, 2017). The results regarding this relationship have been mixed as well, therefore in the following section, an overview is provided of key findings. Additionally, a distinction will be made between female TMT members and female CEOs and based on the findings hypothesis 1 and 2 will be formed.
expertise and skills, along with other social diversity aspects as age and race/ethnicity, are centralized (Creary, et al., 2019).
This research will continue to give rise to the upper echelons perspective and its emphasis on how individual characteristics, in this case, the gender of TMT members compose firms’ strategic choices. As aforementioned, according to the UET, managers do not always make rational decisions, due to their bounded rationality, characteristics, and cognitive biases (Nielsen, 2009; Hoffmann and Meusburger, 2018). Moreover, executives’ cognitive frames – i.e. their information-evaluation and information-seeking processes – are dependent on their values, experiences, and knowledge. These values, experiences, and knowledge shape how executives interpret and gather information. In turn, the executives’ cognitive frames form the decision-making processes, decisions, and eventually, organizational outcomes (Carpenter, 2002). In this research, it will be argued that TMT gender diversity is likely to influence organizational performance since male and female managers vary in their cognitive frames.
There are different outcomes on the relationship between financial performance and a gender-diverse TMT. Some scholars found an insignificant relationship (Manner, 2010; Opstrup and Villadsen, 2015). For instance, Opstrup and Villadsen (2015) proposed that increased gender diversity is related to higher financial performance, however, this was not supported by their results. However, they did found that gender diversity does have a positive impact on financial performance when the municipality’s top management is organized as a board. Whereas others (e.g. Shrader, Blackburn and Iles; Lee and James, 2007; Adhikari, 2012) found a negative relationship. For instance, the results of Adhikari (2012) indicate that an organization headed by a female CEO has lower industry adjusted return on assets (ROA). Besides, they are more likely to hold more cash and to maintain lower financial leverage.
This will ultimately lead to high financial performance, due to their multitasking approach and noticeable problem-solving attitude (Welbourne, Cycyota and Ferrante, 2007) and their “feeling” cognitive style, empowering and relation-building skills (Hoffman and Hurst, 1990).
Moreover, there are mixed results on this relationship and therefore Hoobler et al. (2018) provide a more in-depth examination of the theoretical and methodological foundations for the business case for female leaders. One of their research questions was: “What is the relationship between women’s leadership and organizational financial performance? (p. 2476)” Through a meta-analysis with k=78 and n=117,639 organizations, they wanted to find an answer to this question. Within their sample, they included women in organizational leadership positions as CEOs, TMT members and board members. They did not found a direct relationship between women’s representation in TMTs and women CEOs on financial performance. However, when they moderated for gender supportive climates they found that female representation in leadership positions is positively related to organizational performance. This is not only the case in countries where women had access to outcomes such as health care and education but also in countries where attitudes towards women are more progressive. Moreover, the findings suggest that women in leadership positions – overall, and in particular the presence of a female CEO – tend to positively relate to organizational financial performance in more gender-egalitarian cultures. Thus, countries that have a more progressive attitude towards women’s equality, will have the culture and/or organizational support for women to influence their firm’s performance. Additionally, Jeong and Harrison (2017) conducted a meta-analysis as well on how female representation in the upper echelons may affect organizational performance. They used a sample of 146 primary studies conducted in 33 countries, and their results suggest that female representation in TMTs is positively related to long-term financial performance.
As mentioned, in this study a distinction will be made between TMT members and CEOs as two separate unit of analysis. Below they will be discussed, respectively, regarding their relation to financial performance.
2.3.1 Female TMT members
view more often than their male colleagues (Chen, et al., 2019). Hence, an increased female TMT representation results in a greater diversity of viewpoints, which can help in solving complex issues.
Previous literature reported that heterogeneous teams outperform homogenous teams, because they tend to engage more in discussions of diverse information and knowledge, and to incorporate this information and knowledge in its strategy (Van Ginkel and Van Knippenberg, 2008; Loyd, et al., 2013). Additionally, female TMT members are expected to value tolerance, interdependence, and benevolence, more than their male colleagues (Adams and Funk, 2012). Whereas, male managers are more likely to have a decision-making process based on traditional ways of doing and regulations, getting along, or using rules (Bart and McQueen, 2013). This does not mean that a female director does not make use of rules, however, Bart and McQueen (2013) suggest that female managers are less constrained within these boundaries, and are more willing to “rock the boat” than their male colleagues.
markets, and these new understandings can be added to the knowledge pool of the TMT (Bilimoria and Wheeler, 2000).
In conclusion, these diverse perspectives that are provided by female managers, give access to important information that could potentially improve organizational performance (Peterson and Philpot, 2007). Several scholars support this evidence, as the cognitive resource model suggests that the available cognitive resources will expand, when (gender) diversity increases (Jackson, May, Whitney, 1995; McLeod, Lobel, Cox, 1996). These different cognitive frames can contribute to enhanced discussions, and consequently, a more thorough search for other solutions since they introduce new perspectives to the TMT (Watson, Kumar and Michaelsen, 1993). Additionally, the different points of view stimulate innovative and creative solutions (Bassett-Jones, 2005) and develop a critical analysis of difficult problems, which prevents premature decision-making (Van Knippenberg, De Dreu and Homan, 2004; Carter, Souza and Simkins, 2010). Furthermore, Chen et al. (2019) found that female executives can help temper male CEOs’ overconfidence. When a CEO is overconfident (this characteristic is more prevalent among male CEOs), it leads to an underestimation of risk and an overestimation of returns, which can lead to excessive risk-taking and overinvestment. A TMT with female executives might challenge the male CEO to consider other options and take the pros and cons into consideration while making a strategic decision. Furthermore, Creary, et al. (2019), agrees with this line of reasoning and suggests that having female executives on TMTs helps temper the overconfidence of male CEOs, which improves organizational decision making. Moreover, the different perspectives of female managers – i.e. the added available cognitive resources – may help temper CEO overconfidence, which results in better decisions regarding investments (Huang and Kisgen, 2013), and consequently in better firm performance.
will argue for a positive relationship between female presence in TMTs and long-term financial performance. Moreover, this research will support the abovementioned business case for diversity (Robinson, 1997). Hence, the following is proposed:
Hypothesis 1: Female presence in TMTs has a positive relationship with long-term financial performance.
2.3.2 Female CEOs
Mackey (2008) and Hambrick and Quigley (2014) find that executives who fill CEO positions have the largest opportunity to influence long-term financial profitability. According to some psychological and behavioural studies (Nadkarni and Herrmann, 2010; Peterson, Galvin and Lange, 2012) and the UET (Hambrick and Mason, 1984), these influences can be traced back to individual characteristics and cognitive frames of the CEO.
Jeong and Harrisson (2017) argue that female CEOs will make strategic decisions more cautiously, which would impact the long-term financial performance positively through its influence on strategic risk-taking. Their results indicate as well that female CEOs will be beneficial for long-term financial performance. Furthermore, studies show that women extract less personal benefits from the organization than male executives and worry more about the way money is spent in an organization (e.g. Barber and Odean, 2001; Bliss and Potter, 2002). Additionally, psychological literature reports that women are likely to perform better in decision-making and group problem-solving tasks and that females have better communication skills than men (e.g. Robinson and Dechant, 1997; Dallas, 2002; Schubert, 2006). Besides, previous studies indicated that women are less individualistic, less tolerant for opportunistic behaviour, and they act more decisively since they are more averse to reputation loss and litigation (Thorne, Massey and Magnan, 2003; Srinidhi, Gul and Tsui, 2011). Furthermore, as mentioned earlier, the business world is mostly controlled by male CEOs. Therefore, women who want to fulfil a managerial position in a TMT or strive to become a CEO have to show special talent. Hence, female managers may have to work harder than male executives to be recommended as a CEO. Consequently, the women in these functions tend to be more devoted and talented than men, which could lead to increasing firm performances (Fondas and Sassalos, 2000; Eagly and Carli, 2003).
Additionally, some of the arguments noted above, regarding the section 2.3.1 female TMT members, apply here as well. For instance, female directors may have a more complete and better understanding of customers’ needs and consumer behaviour, which provides a competitive advantage for firms who are controlled by a woman (Brennan and McCafferty 1997). Furthermore, as mentioned, male CEOs tend to be overconfident (Huang and Kisgen, 2013; Chen et al., 2019; Creary, et al., 2019). Bonner (2008) confirms this reasoning, he found that men are more likely to be overconfident than women and this is especially significant in masculine environments, as the business environment. These differences in levels of overconfidence, based on gender, can be seen in the way in which executives operate, which can result in differences in firm performance of male- and female-controlled organizations.
Moreover, several scholars have argued that in the current business world, the cooperative leadership style of female managers may provide a larger advantage than the competitive leadership style of men (e.g. Eagly and Carli, 2003). Also, Peni (2014), found, that in general, organizations with female CEOs are more likely to outperform firms with male CEOs. This research hypothesizes in favour of this result, as it argues that female presence in CEO positions will have a positive relationship with financial performance. Consequently, the following is hypothesized:
Hypothesis 2: Female presence in CEO positions have a positive relationship with long-term financial performance.
2.4 TMT gender diversity and international diversification
the trajectories and fortunes of firms are often traceable to the actions or inactions of their boards and top executives.”
Previous research has shown different aspects of diversity and its relationship with international diversification. For instance, Rivas (2012) explored the role of diversity between tenure, age, and functional background at TMT level and its exercise on internationalization. Additionally, Tihanyi, et al. (2000) also focused on demographic variables, such as average age, tenure, elite education and international experience of TMT members and the relation with international diversification. To find out whether there is a relationship between gender diversity in TMTs and international diversification, this research will take the capability of strategic decision-making into account. Because according to Tihanyi, et al. (2000), business activities across foreign borders are considered to be complex strategic decisions among corporate executives. Therefore, it is up to the TMT to deal with the uncertainty surrounding business decisions and to provide counsel and advice in shaping strategies (Finkelstein, Hambrick and Cannella, 2009). Moreover, developing strategies – and thus making difficult strategic decisions – and managing operations in different countries, have become an important task for TMT members (Hitt, Tihanyi and Miller, 2006). This study will elaborate further on the finding of Jeong and Harrison (2017) as their results of the meta-analytic path analysis indicate that reduced strategic risk-taking is the explanation to why female representation is connected with enhanced long-term financial performance.
2.4.1 Female TMT members
Chen, Crossland and Huang (2016) did research on the relationship between the impact of female board representation on firm-level strategic behaviour regarding mergers and acquisitions (M&As). Even though M&As may provide many advantages, such as increased economies of scale and scope, the actual returns of these acquisitions differ per deal. In reality, research suggests that acquisitions may destroy, instead of increasing value of the acquiring company (King et al. 2004; Haleblian et al., 2009). A possible reason for this result is that the acquiring executives benefit disproportionately from M&As in the short-term (via compensation and status) but that the success of the M&A can only be evaluated successfully years afterwards (Haleblian, et al., 2009). Furthermore, executives are often overconfident regarding prospective synergies (Hayward and Hambrick, 1997) and M&As are often executed without enough due diligence (Puranam, Poweel and Singh, 2006). Chen, Crossland and Huang (2016) argued that increased female representation on boards would be associated with a better oversight in evaluating managers’ recommendations and a more thorough intra-board discussion. This oversight and comprehensive decision-making will result in an increased time to reach a decision. During this extended decision-process, the boards will recognize the uncertainty of the M&As payoff (Haunschild, 1994). Moreover, they will become aware of the complex nature of M&As (Haspeslagh and Jemison, 1991). Whereas boards who are all male, tend to sign off more rapidly on any given acquisition. This is in line with other scholars (Van Ginkel and Van Knippenberg, 2008; Loyd, et al., 2013), as they argue female executives may help increase an organization’s ability in the decision-making process, due to the different perspectives that female executives bring to the table and their enhanced deliberativeness in decision making. Moreover, Chen, Crossland and Huang (2016) found that female board representation is negatively significantly related to acquisitiveness, meaning increased female board representation is associated with fewer acquisitions, thus less international diversification.
beneficial for several reasons. Among other things, Ossorio (2020) argued that female directors are more likely to show specific behaviours, such as socialization and better networking skills, that will help organizations reduce uncertainty about their dependence on the external environment (Bear, Rahman and Post, 2010). This is necessary for profitable acquisitions, as the company needs to open up their social networks (Gómez-Mejía, Patel and Zellweger, 2015). Second, Ossorio (2020) hypothesized that an increased representation of female directors would negatively affect cross-border acquisitions. Among other things, she argued that females perceive bad outcomes more negatively than their male counterparts do and thus women are less inclined to make risky decisions. Additionally, the author follows the rationale that women are less confident than men, as females are less secure about their predictions of the future to be precise (Barber and Odean, 2001). The results of Ossorio’s (2020) study shows that the ratio of female directors was negatively significant, meaning increased representation of female directors would negatively affect cross-border acquisitions.
Furthermore, the results of Levi and Zhang (2014) show that female directors tend to make fewer acquisitions, and when they do acquire the bid premia will be lower. Besides, they found that the influence female directors have on acquisition decisions will help create shareholder value. Finally, Parola, Ellis and Golden (2015) state that gender diversity within TMTs will enhance the performance expectations on the short-term. Since a gender diverse TMT is more likely to select less risky targets and they will have an increased reach in identifying potential companies for target selection. Because a gender diverse TMT holds qualities that lead to the selection of better targets and the target selection is a primary driver of performance expectations (e.g. Capron and Shen, 2007; Zaheer, Hernandez and Benerjee, 2010). However, the authors expect the former positive performance implication to decrease in the long-term, and to become negative. Because the initial market expectations are declined since value creation target are not met and several integration challenges are exposed (Oler, Harrison and Allen, 2008). Moreover, Parola, Ellis and Golden (2015) expected this negative relationship because gender diverse TMTs reach limited strategic consensus and take more time to make decisions. Hence, both aspects are likely to create integration issues that hinder value creation.
members will be negatively associated with international diversification. Furthermore, as the results of Jeong and Harrison (2017) indicate; reduced strategic risk-taking is the explanation to why female representation is connected with enhanced long-term financial performance. This research will propose the following:
Hypothesis 3: The beneficial relationship between female TMT members and long-term financial performance is negatively mediated by the strategic risk: international diversification.
2.4.2 Female CEOs
The CEO is eventually accountable and responsible for action and reaction to a firm’s strategic change (Dalton and Kesner, 1983). Jackson (1992) reported that scholars tend to treat the CEO the same as other TMT members. However, the CEO might have a huge influence on the other TMT members. Likewise, Hambrick (1994) noted that the CEOs’ behaviour can have a large impact on the behavioural integration of the TMT. Simsek, et al. (2005) find that because of the experience and power of the CEO, they are in a unique position to influence team processes.
There are some mixed results regarding this relationship as well. First, Jeong and Harrison (2017) find that a female CEO tends to engage less in strategic choices that may involve harmful risk-taking. This is in line with other scholars (Vandergrift and Brown, 2005; Wei, 2007). For instance, Ramón-Llorens, García-Meca and Duréndez (2017) studied CEO characteristics that may influence family firms in their decisions to internationalize. One of the characteristics they studied was gender and they hypothesized that: “the higher the female presence in the running of the family firm, the lower the probability of reaching out to international markets” (Ramón-Llorens, García-Meca and Duréndez, 2017, p. 788). They came to this hypothesis, because women are more risk-averse (Vandergrift and Brown, 2005; Wei, 2007) and could be less inclined to make risky decisions in international markets. Additionally, Fisher, Reuber and Dyke (1993) found that in SMEs the female-owned companies are less likely to internationalize than organizations who are owned by men. However, the results of the hypothesis of Ramón-Llorens, García-Meca and Duréndez (2017) show to be insignificant. Meaning, they found that CEO gender does not predict the tendency to export.
firms, the female-owned organizations are less growth-oriented, often smaller and less likely to operate in sectors with a high exporting rate. Moreover, female-owned organizations are less likely to diversify internationally.
Furthermore, prior literature has reported that in general, women are less individualistic and less tolerant for opportunistic behaviour (Thorne, Massey and Magnan, 2003; Srinidhi, Gul, Tsui, 2011). Additionally, Simsek, et al. (2005) noted that collectivistic CEOs will support and encourage joint decision-making by all TMT members. Rosener (1995) found that female leadership styles involve empowering employees, emphasize collaboration, and interaction. Whereas they labelled male leadership as command-and-control, which involves the declaration of authority and the gathering of power. Besides, as mentioned previously, women are less overconfident than men (Chen et al., 2019). This could be a reason as to why female CEOs make less risky decisions and emphasize more on group harmony and cooperation (Finkelstein and Hambrick, 1996), as they are not as confident as men concerning how successful the outcomes of their decisions will be (Lundberg and Punchchar, 1994; Barber and Odean 2001). Meaning, they might be less inclined to make risky diversification decisions compared to male CEOs. Additionally, when the company has a female CEO, the risk level of the organization is lower than when the CEO position is filled by a man (Khan and Vieito, 2013). Moreover, when an organization needs to invest, females are more risk-averse then their male counterparts and this difference is increased with a higher level of uncertainty of the investment (Schubert, Brown and Brachinger, 2000).
Consistent with the above-mentioned rationale and in line with Fisher, Reuber and Dyke (1993) and Orser et al. (2010), the following will be predicted:
Hypothesis 4: The beneficial relationship between a female CEO and long-term financial performance is negatively mediated by the strategic risk: international diversification.
Figure 2.1 Conceptual model
3 Methodology
3.1 Sample and data collection
This quantitative study focuses on TMT members of Swedish companies. Due to the limited timeframe and the COVID-19 pandemic, it was difficult to obtain primary data. Therefore, this study used secondary based data from the database Amadeus. This research has made use of the non-probability sampling technique. This sampling technique involves a specifically chosen sample based on specific characteristics, which are important to the study. There are different types of non-probability sampling; this study has used the purposive type. Meaning, the researcher will decide who and what study units will be involved in the study (Taheri, et al., 2015).
active business leaders (Johnson and Tunheim, 2016). Moreover, if there is a relation between the variables, it will be detected in a sample from Sweden, as it is most likely that there is gender diversity in the TMTs.
Within this population of Swedish companies with foreign subsidiaries, a frame was selected that included the information on these companies and their TMTs. After filtering these companies, there was a frame of 3,110 companies. However, due to the time limitation and the missing values of companies, not all these companies have been used. The sample size has been calculated according to the rationale of Cohen (1992). The following guidelines can be used when determining a sample size: if a standard 𝛼-level of 0.05 will be used and the required recommended power of 0.8, then 783 participants are needed to detect a small effect size of (r = 0.1) (Field, 2009). This study needs a sample size of 783 companies because Jeong and Harrisson (2017) found r = 0.14 in their meta-analytic study (see Appendix A). Therefore, to detect a small effect size, the sample size of 783 is appropriate for this study. Thus, within this frame of 3,110 companies, 783 companies have been used for the sample.
3.2 Measurement
3.2.1 Independent variable
To measure the gender diversity in a TMT two variables are measured: Female TMT members and Female CEOs. The information regarding the TMTs in Amadeus was not complete, as it included several members that were not in the management team but on the board of the company and it also included lower-level managers. To find the information about the TMTs, there has been manually searched for the management teams of the companies on their websites. When this information was not available, this company has been removed and was therefore not included in the final sample.
Second, to measure whether a CEO is female or male, this study will follow the same method that e.g. Jeong and Harrisson (2017) and Rivas (2012) have used. A value of 1 has been used for a female CEO and a value of 0 for a male CEO. There will be a small, but inevitable, overlap between the variable of the ratio of female TMT members and female CEOs since the CEO is also a member of the TMT.
3.2.2 Dependent variable
The dependent variable long-term financial performance has been measured with an accounting-based measure. The ROA of 2018 has been used to measure the long-term financial performance (Adams and Ferreira, 2009; Jeong and Harrisson, 2017). ROA measures long-term performance as it is a “backwards-looking” accounting-based measure, this means that it is based on the organization’s self-reported financial performance in the recent past (Haslan, et al., 2010). This measure is obtained from the database Amadeus. Companies who had a missing value for ROA have not been included in the final sample.
3.2.3 Mediator variable
Previous studies have used several different measurements regarding international diversification (Hitt, Tihanyi and Miller, 2006), including the ratio of foreign assets to total assets (FATA), foreign sales to total sales (FSTS) and foreign employees to total employees (FETE) (Contractor, Kundu and Hsu, 2003; Li and Qian, 2005). This study makes use of the ratio FATA, as the components of this ratio are available in the database Amadeus. However, there is a side note to this, because in Amadeus only the assets of the subsidiaries and the total assets were available. The assets of subsidiaries did not only include the foreign subsidiaries, the Swedish subsidiaries were included as well. These Swedish subsidiaries have been filtered out. Moreover, to calculate FATA, the foreign assets of the foreign subsidiaries were divided by the total assets. Furthermore, there were a lot of missing values regarding the number of assets a foreign subsidiary had, these missing values have been removed.
3.2.4 Control variables
First, the size of a TMT can influence how the ratio between male and female positions are divided. A larger group has more potential to have a more diverse TMT (Wiersema and Bantel, 1992). Therefore, to account for possible effects related to the size of the TMT, this variable is included in the model. The size of the TMT has been found manually when information regarding the independent variables was searched. TMT size is measured by the number of individuals on each organization’s TMT.
Second, the firm’s size can influence the financial performance of the firm and is therefore a common control variable (Michel and Hambrick, 1992; Finkelstein and Hambrick, 1996). This will be measured by the number of employees in the firm (Kogut and Singh, 1988). The number of employees has been found in the database Amadeus.
Third, the number of subsidiaries can show how many foreign operations an organization has, this does not take the entry mode of the subsidiary into account. As aforementioned, Gomes and Ramaswamy (1999) argue that when an enterprise starts, it does not have that many subsidiaries. Meaning, they can manage resource flows and information across its network quite easily and are thus able to enjoy the benefits of internationalization. Following this rationale, it means that the number of subsidiaries will influence financial performance. The number of subsidiaries is collected in Amadeus. As mentioned above, the missing values of the foreign subsidiary assets have been deleted. Therefore, the number of subsidiaries may have decreased for certain companies, if that company had a subsidiary with missing values regarding the assets.
A quick overview of the measurements of the variables can be found in Appendix B.
3.2.5 Statistical analysis method
multiple regression has been used. To establish mediation, three conditions have to hold. First, the independent variable has to affect the mediator in the first step. Second, in the second step, the independent variable must affect the dependent variable. Finally, the mediator has to affect the dependent variable in the third step. Moreover, the effect of mediation exists if international diversification affects financial performance and if the independent variables (gender diversity and CEO) are related to the dependent variable and the mediator. The relationship between the dependent and the independent variables has to be weaker when the mediating variable is added to the model. Complete mediation effects exist when the independent variables do not have a statistically significant effect on the dependent variable when the mediator is controlled for. Partial mediation will exist if the statistically significant effects of the independent variables on the dependent variable are smaller when the mediator is controlled for (Baron and Kenny, 1986).
Before the regression analyses have been performed, some tests of assumptions have been executed. First, to test for normality, a histogram and a Q-Q plot have been made. They both showed that the dependent variable is overall normally distributed. However, the data seems to have a slightly left-skewed distribution, which indicates that the data is not completely normally distributed. Additionally, the Shapiro-Wilk test has been used to confirm whether the data is normally distributed and the results show that the dependent variable is not normally distributed. Even though the data is not normally distributed, Altman and Bland (1995), found that the violation of the normality assumption should not cause crucial problems if a study has a large enough sample size (greater than 30 or 40). They find that if a study consists of hundreds of observations, the distribution of the data can be ignored. Moreover, this suggests that parametric procedures can be used even though the data is not normally distributed (Elliot and Woordward, 2007).
significance nor the estimates. Therefore, these outliers have not been excluded from the sample.
Furthermore, to test for heteroscedasticity, scatterplots and the Breusch-Pagan test have been performed. First, a scatterplot of the independent variable long-term financial performance shows that there are some outliers and the values are a bit spread out, but overall the values are clustered together. Meaning there is a sign of heteroscedasticity. The Breusch-Pagan test confirms this result with (p<0.05). Second, the values of the independent variables gender diversity and female CEO show dispersed values and these are less clustered than the dependent variable, therefore there is no real sign of heteroscedasticity. When the Breusch-Pagan test is performed, and a threshold of 0.5 is chosen, for both variables there is no sign of heteroscedasticity (p>0.05). However, when a threshold of 0.10 is chosen, there is, for both variables, a sign of heteroscedasticity (p<0.10). Thus, in this case, it is debatable if there is a sign of heteroscedasticity. Finally, for the mediating variable international diversification, there is no sign of heteroscedasticity when the Breusch-Pagan test is performed (p>0.10). Heteroscedasticity is a problem when performing a regression analysis because the error term in the performed regression will not have a constant variance, whereas with homoscedastic data the error term will have a constant variance (Berry and Feldman, 1985). Moreover, to correct for heteroscedastic data, a robust standard error has been included in the regression analyses (Cribari-Neto, da Gloria and Lima, 2014).
Finally, to test for multicollinearity the variance inflation factors (VIF) were taken into consideration. The VIF for all the variables is well under the threshold of 10, as they are strictly less than 2, suggesting that multicollinearity is not an issue in the models.
4 Results
4.1 Descriptive statistics and correlations
(SD=0.319). Out of the 783 observations, 90 observations are female, moreover, in this sample 11,49 per cent of the CEOs are female. The variable international diversification has a mean of 0.298 (SD=0.25), with a minimum of 0 and a maximum of 0.995. The data is controlled for the number of TMT members, employees and subsidiaries. The number of TMT members has a range between 1 and 33 members, the size of the firm has been accounted for in terms of the numbers of employees with a minimum of 1 and a maximum of 123.283 employees. Finally, the number of subsidiaries ranged from 1 to 66 subsidiaries.
The correlations are presented in table 4.2. None of the correlation coefficients is high enough (>0.80), therefore, it can be concluded that the models are free of multicollinearity problems.
Table 4.1: Descriptive statistics
Variables Obs Mean Std. Dev. Min Max
Financial performance 783 5.705 13.188 -71.774 63.915 Gender diversity 783 0.292 0.178 0 0.5 Female CEO 783 0.115 0.319 0 1 International diversification 783 0.298 0.25 0 0.995 TMT size 783 7.573 3.593 1 33 Firm size 783 1501.701 6359.468 1 123283 Number of subsidiaries 783 4.65 8.069 1 66
Table 4.2: Correlation matrix
Variables (1) (2) (3) (4) (5) (6) (7) (1) Financial performance 1.000 (2) Gender diversity 0.021 1.000 (3) Female CEO -0.006 0.246*** 1.000 (4) International diversification 0.026 -0.053 -0.043 1.000 (5) Number of TMT members 0.026 0.276** 0.050 0.010 1.000 (6) Number of employees 0.001 0.080** 0.142*** 0.082** 0.179*** 1.000 (7) Number of subsidiaries 0.017 0.030 0.063* 0.351*** 0.151*** 0.445 1.000 *** p<0.01, ** p<0.05, * p<0.1 4.2 Regression analysis
Tables 4.3 and 4.4 show the results of the regression analyses. In model 1 and 2 (table 4.3) only the control variables are entered and regressed on international diversification and financial performance respectively. Models 3-5 are based on gender diversity (table 4.3) and models 6-8 are based on whether a company has a female CEO (table 4.4).
Hypothesis 2 suggests that female presence in CEO positions has a positive relationship with long-term financial performance. The results indicate that a female CEO has no relation to the financial performance of an organization (t=0.15) (p>0.1) in model 7 table 4.4.
Hypothesis 3 suggests that the beneficial relationship between female TMT members and long-term financial performance is negatively mediated by international diversification. This mediating mechanism has been examined by performing the three-step mediated regression analysis of Baron and Kenny (1986). In the first step, the mediator was regressed on the independent variable gender diversity. The results show that gender diversity is not significantly related to international diversification as (t=-1.50) (p>0.1) in model 3, table 4.3. In the second step, the dependent variable financial performance was regressed on the independent variable gender diversity (H1). These results have been provided above. In the third step, the dependent variable was regressed simultaneously on the independent variable gender diversity and the mediator international diversification. The results indicate that both gender diversity and international diversification are not significantly related to firm performance with gender diversity (t=0.44) (p>0.1) and international diversification (t=0.65) (p>0.1) in model 5 table 4.3. Hence, since all three steps do not hold there is no sign of a mediating mechanism.
Even though the focus is not on the control variables, it is interesting to note that only the number of subsidiaries and the firm size are statistically significant when regressed on international diversification in model 1 and 3 in table 4.3 and model 6 in table 4.4. Additionally, only these models have an adjusted R-squared greater than 0.10 and are therefore sufficient according to Falk and Miller (1992).
Table 4.3 Regression analysis with independent variable gender diversity regressed on: ID = international diversification and FP = financial performance
VARIABLES (1) ID Controls (2) FP Controls (3) ID Dependent variable (4) FP Dependent variable (5) FP Mediator Gender diversity -0.074 (0.049) 1.106 (2.761) 1.207 (2.760) International diversification 1.371 (2.112) Number of subsidiaries 0.012*** (0.001) 0.023 (0.035) 0.012*** (0.001) 0.024 (0.035) 0.007 (0.042) Firm size -3.44e-06*** -2.22e-05 -3.35e-06*** -2.36e-05 -1.90e-05
(1.27e-06) (3.60e-05) (1.27e-06) (3.63e-05) (3.58e-05)
TMT size -0.002 0.104 -0.001 0.089 0.091 (0.002) (0.136) (0.002) (0.143) (0.144) Constant 0.265*** 4.845*** 0.279*** 4.845*** 4.251*** (0.019) (1.147) (0.021) (1.147) (1.284) F-statistic 25.56*** 0.38 20.51*** 0.35 0.37 Observations 783 783 783 783 783 R-squared 0.131 0.001 0.133 0.35 0.002 Adj. R-squared 0.128 -0.003 0.129 -0.004 -0.005
Robust standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1
Table 4.4 Regression analysis with independent variable CEO regressed on: ID = international diversification and FP = financial performance
VARIABLES (6) ID Dependent variable (7) FP Dependent variable (8) FP Mediator Female CEO -0.043* (0.024) -0.280 (1.842) -0.224 (1.864) International diversification 1.306 (2.137) Number of subsidiaries 0.012*** (0.001) 0.023 (0.035) 0.007 (0.0419)
Firm size -3.14e-06** -2.02e-05 -1.61e-05
(1.30e-06) (3.72e-05) (3.63e-05)
TMT size -0.002 0.104 0.107 (0.002) (0.137) (0.137) Constant 0.269*** 4.870*** 4.519*** (0.019) (1.120) (1.224) F-statistic 21.73*** 0.90 0.30 Observations 783 783 783 R-squared 0.134 0.001 0.002 Adj. R-squared 0.129 -0.0041 -0.005
Robust standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 5. Discussion
Overall, the results of the regression analyses do not support the hypotheses, meaning both gender diversity and a female CEO are not related to the financial performance of the firm and this relationship is not mediated by international diversification. In the following section, these results will be discussed.
The theory of this research was in line with Jeong and Harrisson (2017) who argue that female CEOs and female TMT members will make strategic decisions more cautiously, which would impact the long-term financial performance positively. Additionally, they find that the improved financial performance can be explained by the reduced strategic risk-taking as a mediating mechanism. However, the results of this study do not support this line of reasoning. The results of the female TMT members and the female CEO will be discussed per hypothesis.
First, it was hypothesized that female presence in TMTs will have a positive relationship with long-term financial performance (H1). However, this result differs from prior research which found a positive significant relationship (e.g. Krishnan and Park 2005; Cambrea et al., 2017; Jeong and Harrison, 2017). The results of this study indicate that female representation
is not significantly associated, positively or negatively, with financial performance. This insignificant result could arise due to the small time-frame of one year of the variable financial performance. Additionally, only one measure of the dependent variable, ROA, has been used. However, if more measures of the dependent variable, such as return on equity, return on capital and return on investment (Hoobler et al., 2018) would have been used the outcomes could have been compared and different outcomes per measure could have arisen. Furthermore, a possible explanation for this insignificant result could be that gender differences might fade beyond the glass ceiling (Adams and Funk, 2012). Because female managers who operate in a male-dominated environment may amend their behaviour so gender differences vanish. Additionally, there is legal evidence that female executives were denied a promotion, because they acted too “feminine” (Branson, 2006). Meaning, women who apply for leadership roles may be similar to men (e.g., Niederle, Segal and Vesterlund, 2008). As a consequence, the benefits of having female TMT members may not be reflected or realised in any chosen performance measure.
Second, it was hypothesized that the female presence in CEO positions is positively related to long-term financial performance (H2). The results of this study found a different outcome as it was insignificant; a female CEO is not significantly associated, positively or negatively, with financial performance. This outcome differs from prior studies which found that organizations with a female CEO tend to outperform firms with a male CEO (e.g. Peni, 2014, Jeong and Harrisson, 2017). This result could have arisen since only 90 from the 783 CEOs are female, this is only 11,49 per cent. A different outcome could have been possible if there were more female CEOs. However, this also means that even though Sweden has been working hard the past years to decrease the gender equality gap in business (Johnson and Tunheim, 2016), they still need to work on the gap regarding female CEOs. Additionally, this result could also be the outcome of the small time frame of the variable financial performance or because only ROA has been used as the measure for the dependent variable. Furthermore, a possible explanation for this insignificant result could arise due to the pressures towards similarities in a leadership role. Hence, the differences between male and female leadership styles are likely to be relatively small (Engen, van der Leeden and Willemsen, 2001; Nieva & Gutek, 1981). Thus, the advantages of having a female CEO may not be reflected in any chosen performance measure.
diversification. The results do not support this hypothesis, and they will be discussed per step. In the first step (model 3, table 4.3), the mediator international diversification was regressed on gender diversity. There was a small negative, but not statistically significant relationship between gender diversity and international diversification. This result differs from prior research which found a negative significant relationship (e.g. Levi and Zhang, 2014; Chen, Crossland and Huang, 2016; Ossario, 2020). This different result may have arisen due to the partially incomplete variable of international diversification since the used FATA ratio only includes the assets of foreign subsidiaries. Whereas, the rationale of the scholars discussed in the literature review (Levi and Zhang, 2014; Chen, Crossland and Huang, 2016; Ossario, 2020) build their hypotheses on mergers and acquisitions, thus another aspect of international diversification. Furthermore, the result of the second step (model 4, table 4.3) has already been discussed above (H1). Finally, in the third step (model 5, table 4.3) the dependent variable financial performance was regressed simultaneously on the independent variable and the mediator. The results show that there is no significant relationship. Again, this result could have arisen due to the small time frame and the use of only one measure of the variable financial performance or to the partially incomplete variable of international diversification, as aforementioned.
5.1 Implications
Even though the hypotheses are not supported there are some key takeaways for managers and policy-makers, which will be discussed in the following section. First, for managers, this study provides evidence that female representation and a female CEO in TMTs are not significantly associated, positively and negatively, with the financial performance of the firm. This challenges the scholars who support the business case for diversity (e.g. Krishnan and Park, 2005; Cambrea et al., 2017). Hence, the results indicate that both female representation and a female CEO in TMTs do not increase firm performance, but more importantly, they do not decrease financial performance either. If this is the case, it seems reasonable to promote gender equality within TMTs and thus support the ethical reasons (Robinson, 1997) to promote fairness. Furthermore, even though it is not formally hypothesized, this study found a negative significant relationship between a female CEO and international diversification. This is noteworthy for companies whose main strategic focus is on increased international diversification, for whom it is advised to not hire a female CEO. Because there will be less international diversification with a female CEO, since female-owned companies are less growth-oriented (Orser et al., 2010), female CEOs make less risky decisions (Jeong and Harrison, 2017) and are less over-confident than their male counterparts (Lundberg and Punchchar, 1994; Barber and Odean 2001).
Second, as aforementioned, several countries (e.g. Norway and Spain) already set quotas to increase female representation in the TMTs (Adams and Ferreira, 2009). Even though this study does not provide evidence for the business case to increase gender equality within TMTs, it does provide evidence for ethical reasons as mentioned above (Robinson, 1997). Therefore, this study urges policy-makers to keep setting quotas to increase gender equality in TMTs.
5.2 Limitations and future research
as mentioned, there were a lot of missing values for foreign subsidiary assets. However, this is something that more scholars encounter when studying international diversification (Kim, Hwang and Burgers, 1989). Therefore, Hitt, Hoskisson and Kim (1997) developed an entropy measure of international diversification to take the extent of sales outside the home market and their distribution globally into consideration. As there was a lack of sales data at the country level, they used regional markets. However, due to time constraints, I was not able to create this entropy measure for the asset data using regional markets. Moreover, I urge researchers in this area to continue to improve measures of international activities. Scholars may also wish to conduct a more fine-grained analysis of the influence of TMT members in international operations. As this paper could focus on more specific characteristics of managers such as age, international experience and country of origin to determine how these characteristics influence executive decision-making regarding internationalisation.
