Effect of board’s gender diversity on Corporate Social Responsibility performance– comparison within Europe

(1)

Effect of board’s gender diversity on Corporate

Social Responsibility performance– comparison

within Europe

Juul Driessen, S446557

Supervisor: Jacqueline Drost

14-06-2019

Radboud University Nijmegen

(2)

1

Abstract

In 2012, the European Commission proposed a directive under which women must represent at least 40% of the total board members. This research aims to demonstrate the importance of the gender diversity in boards, examining its effect on the Corporate Social Responsibility (CSR) performance. To find the effect of the board’s gender diversity on CSR performance, a regression analysis with a random effects model is performed. The used unbalanced panel-dataset consists of 462 EUROSTOXX-listed firms from 2003-2017, across the eight European countries with the highest gender diversity. The results show that there is a positive significant effect of the board’s gender diversity on the ESG-score, a well-defined measurement of overall CSR-performance (Thomson Reuters, 2019). Contrary to previous studies, no effect of the binding gender quota on the relationship between the gender’s diversity and ESG-performance is found.

Keywords: gender diversity, gender quota, Corporate Social Responsibility, Environmental

(3)

2

Table of figures and tables

Figure 1: Representation of women and men on boards of the largest listed EU companies 16 Figure 2: Conceptual model………... 26

Figure 3: Average ESG score per country……….. 29

Figure 4: Board's gender diversity average per country………. 30

Figure 5: Kernel density plot………... 35

Figure 6: R variables against normal distribution……….. 35

Figure 7: Overview of method to achieve the data quality goal………. 58

Figure 8: ESG score determinants……….. 58

Table 1: ESG categories and their definitions……… 21

Table 2: Summary of all variables included in this research……….. 25

Table 3: Summary of statistics of variables……… 28

Table 4: Companies and quota per country……… 31

Table 5: PROFILE dummies created by SIC code………. 32

Table 6: Pearson correlation matrix……… 34

Table 7: Variance Inflation Factor……….. 36

Table 8: OLS regression with ESG as dependent variable………. 42

Table 9: OLS regression with ESG as dependent, moderating quota……… 43

Table 10: Regression with lagged variables………... 46

Table 11: Overview relevant CSR literature……….. 56

(5)

4

1. Introduction

In 2012, the European Commission proposed a directive under which women must represent at least 40% of non-executive directors in a firm, and 33% if the legislation is also directed at the supervisory boards (European Commission, 2018). In its report of 2018, it is stated that ‘equal treatment between women and men implies that women must have the same opportunities as men to reach leadership positions in the economic area’. Nevertheless, data collected in this report showed that at the end of 2017, women accounted for only 25.3% of board members in the largest publicly listed companies registered in EU member states (European Commission, 2018). This also holds for the Netherlands, where the stated 30% quota regarding the number of women at the top is not reached (Accountant.nl, 2019). However, why does the gender diversity in boards even matter? This research paper aims to demonstrate the importance of gender diversity in boards, examining the effects of the board’s gender diversity on Corporate Social Responsibility (CSR) performance.

At first, the board members are important in monitoring, controlling, and governance functions (Carter et al. 2010). The composition of the board plays an important role in these functions; the board needs an appropriate mix of experience and capabilities to evaluate management and assess business strategies (Bear, Rahman and post, 2010). Within the board composition, it seems that board diversity enhances the management effectiveness. The board’s diversity influences decision making through involvement of directors with diverse knowledge, perception and ideas (Bear, Rahman and Post, 2010). Gender composition of the board has become widely recognized as a feature of differences of director’s decisions. Women tend to enhance trust, reduce risk seeking and better monitor the agency problem (Huang and Kisgen, 2013). Answering the above question, a more gender diverse board enhances the board’s diversity, and as such the monitoring-, decision-making-, and governance-functions of the board.

This can also be found in prior literature, examining the board-composition’s effect on the board performance by measuring the financial performance of that firm (Carter et al. 2010; Rampling, 2011). However, measuring the importance of the board’s gender diversity only focusing on the financial performance does not fully capture the effect on the firms performance and gives some contradictory results (Ionascu et al, 2018). To find out why the gender diversity in boards does matter in Europe, prior literature on financial performance can be expanded by measuring the firm’s performance on other criteria (Carter et al. 2010; Ionascu et al. 2018).

For example, measuring the influence of the board’s gender diversity on the non-financial variable ‘Corporate Social Responsibility (CSR)’ can fill the research gap. CSR is defined as ‘all activities related to economic, social and environmental responsibilities resulting from the firm’s impact on society’ (EC, 2016). As stated by Galant and Cadez (2017), ‘managers focus more on multi-stakeholder’s welfare instead of concentrating only on maximising the wealth of shareholders’. For these multi-stakeholders, not only the financial performance is important, but they also benefit from the company being socially responsible (Galant and Cadez, 2017).

(6)

5 Therefore, the CSR performance should be considered as an important measurement of the overall firm performance as well. Measuring the effect of the board’s gender diversity on the CSR performance expands the prior literature in a way that the firm performance is now measured in a broader term, instead of only focusing on financial performance (Carter et al. 2010; Ionascu et al. 2018; Carter, Simkins and Simpson, 2003; Rampling, 2011).

The CSR performance will be measured by the overall company socially responsible profile, as done by Nadeem, Zamam and Saleem (2017). Measuring the effect of the board’s gender diversity on the CSR performance, they were first to use an Environmental, Social and Governance (ESG) score to examine the CSR performance. Yet, this was only limited to Australian listed firms, and as they stated; ‘Different countries have their own legal structure and unique cultural orientations, so a cross-country comparison is recommended for future research’ (Nadeem, Zamam and Saleem, 2017). Therefore, this thesis expands prior literature by focusing on firms in multiple EU countries.

The aim of this research is twofold. First this research examines the influence of the higher board’s gender diversity on the CSR performance. This can fill the research gap as stated by Ionascu et al. (2018) and Carter et al (2010), stating that prior literature on the influence of the board’s gender diversity on firm performance should be expanded by other measurements than financial performance. By focusing only on the European countries with the highest board’s gender diversity, this effect is expected to be positive: a more gender diverse board is more sensitive to stakeholders than just to shareholders, so it enhances the firm’s non-financial performance in longer term (Gupta et al., 2015). Measuring the relationship between the gender diversity and the CSR performance should answer the following research question: What is the effect of the board’s gender diversity on Corporate Social Responsibility performance?

Secondly, the effect of the gender quota laws as stated by the European Commission is taken into account. The binding gender quota might positively influence the CSR performance, by having a higher board’s effectiveness and strategic control (Nielson and Huse, 2010). However, the gender quota law can also negatively affect the board, in a way that the binding quota leads to younger and less experienced boards, changing the decision making and affecting the monitor- and advisor roles (Ahern and Dittmar, 2012). To test if the legislative gender quota affects the relationship between the board’s gender diversity and CSR performance, a moderating effect of the binding gender quota is tested.

Using data on the period 2003-2017 for the eight European countries with the highest gender quota, the results provide support for a positive effect (European Commission, 2018). The sample consists of 462 firms in eight different sectors located in France, Sweden, Italy, Finland, Germany, Belgium, Denmark and The Netherlands. The board’s gender diversity is the main independent variable to be tested on the firm’s CSR performance. This CSR performance is measured by using Thomson Reuters’ ESG score, an overall measurement of the ESG performance, commitment and effectiveness (Thomson Reuters, 2019). The Ordinary Least Squares regression method is used to predict relationship between the unbalanced panel data obtained, in which the regressors are threatened as random variables.

(7)

6 This study contributes to the literature in different ways. At first, most research focused on the effect of gender diversity on firm value, not taking into account the social and environmental aspects (Carter et al., 2010; Rampling, 2011). However, social responsibility of a firm becomes more important to multi-stakeholders, therefore the non-financial firm performance measured by CSR performance should be taken into account as well . (Gupta et al. 2015). This is also stated by Terjesen, Couto and Fransisco (2016), saying that their research can be expanded by measuring firm performance in non-financial terms, like social performance or environmental sustainability practices. Also Carter et al. (2010) and Ionascu et al. (2018) imply further research is needed, looking at other criteria like social cohesion and sustainable development. Overall the effect of the board’s gender diversity on firm performance can be expanded by measuring the effect on non-financial performance like the CSR performance. Secondly, results in previous research on the relationship between gender diversity and CSR reporting or performance is often country-specific and can therefore differ. This research expands previous research by focusing on multiple European countries. By focusing on Europe, it provides an alternative to the limitations noticed by Majeed, Aziz and Saleem (2015). Looking only at emerging countries within their research, they state that no significant effect can be found because of its lack of women representation in the board. For the chosen European countries however, the average women’s representation is much higher with an average of 25.3% which enhances the probability of finding a significant result (European Commission, 2018).

Also, this research is the first transferring this focus to multiple European countries (excluding the U.K.) and to combine this with the moderating effect of the gender quota laws. Previous research mostly focused on individual countries like the U.S. (Bear, Rahman and Post, 2010) and the U.K. (Al-Shaer and Zaman, 2016; Frias-Aceituno et al., 2013). The effects of gender diversity for U.K. cannot be generalized to the rest Europe because of its different institutional environment. In comparison to the U.S., Europe is much further in implementing the gender quota laws and boards are already more gender diverse (Jourová, 2016).

Adding the gender quota laws is new, and also relevant in a political way. Examining the moderating effect of legislative gender quota might help in the discussion of whether the gender quotas for the countries should be binding or not. For example the Netherlands does not have a binding quota, however the minister of education strongly advocates for sanctions if this is not met (Accountant.nl, 2019). If a positive moderating effect of the legislative quota on the relationship between gender diversity and CSR is found, this will support her view. The remainder of this thesis is structured as follows. In part two, the theoretical framework and literature review both lead to the hypothesis development. Next, this study’s data is described and also the methodology including the regression models will be explained. In the fourth section the data is summarized and the results are described and compared to prior literature and theory. Finally in the conclusion the limitations of this research and directions for further research are given.

(8)

7

2. Literature review and hypothesis

In this chapter an overview of the relevant literature is given. At first the agency theory is used to describe how the board of directors is an important corporate governance mechanisms in big companies. By looking at different theories and prior literature it becomes clear that gender diversity is important for the board and their decision making. However, measuring the board’s gender diversity only on financial performance does not fully capture the effect on the overall firm performance. Considering Corporate Social Responsibility as an important factor driving firm performance, this research gap can be filled and the effect of the board’s gender diversity on overall firm performance is examined.

2.1 Agency theory and corporate governance

Corporate governance is defined as ‘the system of rules, practices and processes by which a company is directed and controlled’ (ICSA, 2018). In this system, the board of directors is a potentially important instrument of internal control (Baysinger and Hoskisson, 1990). This importance of the board of directors can be explained by the agency theory. Agency theory attempts to describe the relationship between one party (the principal) delegating work to another party (the agent) (Jensen & Meckling, 1976). In case of an agency problem, the desires or goals of the principal and agent conflict and the principal can hardly verify what the agent is actually doing (Eisenhardt, 1989). This can also be applied to the shareholder-manager relationship, as examined by Barnea & Rubin (2010). They stated that there might be a conflict of interest between the top management and the stockholders of a firm, so internal versus external ownership. For example, by having more information about the company, managers can make decisions for their own benefit, which cannot be as beneficial for shareholders (Boshkoska, 2015).

According to Saeid and Sakine (2015), the board of directors is elected to manage the potential conflicts of interest between the managers and the stockholders. In their study they suggest that corporate governance can reduce the agency costs, and in this way improve the firm’s performance. The ‘board’ and ‘corporate governance’ terms are used interchangeably, however De Andres and Vallelado (2008) clarify this by stating that the role of boards work as a mechanism for corporate governance, so the board is a part of the corporate governance system. They come to the same conclusion, stating that the board of directors and senior management enhances corporate governance, which increases monitoring efficiency.

To avoid the conflict of interest between regulator and shareholders, the board fulfills the monitoring and advising function in an efficient matter (De Andres and Vallelado, 2008). In other words, the board can monitor the management so it becomes less likely the interest of shareholders is subverted (Carter, Simkins and Simpson, 2003). Concluding from these articles, the board of the firm can reduce the conflicts of interest between managers and stockholders, and thus enhance the corporate governance and reduce the agency problem.

(9)

8 2.2 Board composition

‘Discussing the role of the board in a theory of corporate governance without discussing board composition is as appropriate as discussing the theory of the firm and ignoring the internal structure of the organization’ (Baysinger and Butler, 1985, p. 121). We have to look deeper into the board, stating that its composition does have an effect on the role of the board in corporate governance and in that way affecting the firm’s performance.

By looking at the board composition, the main issues are related to the board independence and diversity of board members (Tamang, 2014; Hillman and Dalziel, 2003). One of the first research examining the board independence importance is written by Baysinger and Butler (1985), focusing on the dependent versus independent directors in the board on the firm’s performance. When a board is said to be ‘independent’, this means the majority of the board’s members consists of independent, outside directors which are believed to better monitor the managers and decision-making of the firm (Tamang, 2014). Baysinger and Butler (1985) concluded that a firm with higher proportion of independent board members ended up with better financial performance compared to firms with lower independence.

Also on the board’s member diversity Baysinger and Butler (1985) examined that a mix of insiders and outsiders, plus a component of instrumental directors provides an appropriate board structure for most corporations (Baysinger and Butler, 1985). Ten years later, Treichler (1995) expanded the research on board composition by measuring the effect of the board’s function, board size, outsiders and experience on the strategic, governance and institutional function of the board. He found that a specific board composition positively affects these functions, however this only held when the trade-off between the diversity and its enhanced costs of coordination and conflict was positive (Treichler, 1995).

Nowadays, the effect of the board’s composition is further examined by e.g. McIntyre, Murphy and Mitchell (2007), regarding more components of the board composition on firm performance. They found that higher levels of experience, appropriate team size, moderate levels of variation in age and team tenure were correlated with firm performance (McIntyre, Murphy and Mitchell, 2007).

The main theoretical point in Baysinger- and Butler’s (1985) research is that the most appropriate board composition includes a mixture of various types of directors. When this board’s diversity holds, it serves to resolve conflicts of interest among the managers and shareholders in a better way and agency costs are reduced (Baysinger and Butler, 1985). This is also stated by Bear, Rahman and Post (2010), examining that the board needs the appropriate mix of experience and capabilities to evaluate management and assess business strategies to fulfill the monitoring and governance function of the board.

In short, the monitoring and advising capability of the board to reduce the agency problem depends on the composition of the board. For example a proper mix of insiders and outsiders (independence), and more diversity can increase the monitoring and advising capability, and thus increase the corporate governance mechanism of a firm.

(10)

9 2.3 Board’s diversity

Within the board composition, it seems that board diversity enhances management effectiveness and is an important mechanism for corporate governance. Board diversity enhances board independence, and next influences better decision making through involvement of directors with diverse knowledge, perception and ideas (Post et al., 2011). This is also stated by Harjoto, Laksmana and Lee (2015), measuring the board diversity in seven dimensions: race, outside directorship, gender, expertise, power, age and tenure. These factors all contribute to a higher board diversity, which are found to enhance the firm’s ability to satisfy the needs and interests of different groups of stakeholders. So given that group dynamics and decision making vary depending on the background of the individuals serving on corporate boards, a diverse group of directors can better enhance the monitoring and advising capability in a firm (Harjoto, Laksmana and Lee 2015).

Again, the agency theory provides some arguments to explain the impact of a board’s diversity on the corporate governance of the firm. Carter, Simkins and Simpson (2003) state that board independence is critical for boards to function in the best interests of shareholders. With board independence they do not refer to the difference between inside and outside board members as Baysinger and Butler (1985) did, but Carter, Simkins and Simpson (2003) focused more on increasing diversity. They stated that people with different gender, ethnicity or a cultural background might ask different questions than directors with a traditional background (Carter, Simkins and Simpson, 2003). The higher the differences, the more independency and thus the better the board can reduce the agency problem.

2.4 Board’s gender diversity

Gender composition of the board has become widely recognized as a feature of differences of director’s decisions. The role of this gender difference is a significant aspect of the board’s diversity, because women are different from men (Terjesen, Sealy and Singh, 2009). This was also found by Harjoto, Laksmama and Lee (2015), examining a positive effect of gender diversity on the board diversity, and in this way on the board’s decision making. Also Perrault (2015) examines why board gender diversity matters in the board’s effectiveness. She finds that women play an important role in the board because they enhance the perception of the board’s legitimacy and trustworthiness, thus fostering shareholder trust in the firm and contributing to its market performance.

Women can enhance trust in the firm, but can also affect the quality of monitoring. This is examined by Adams and Ferreira (2009), finding that women are more likely to join monitoring committees. Also, female directors have better attendance records than male directors and the more gender-diverse the board is, the less attendance problems male directors have. Men and women are thus different in their behavior and can influence the board’s composition, having a significant impact on board inputs and firm outcomes (Adams and Ferreira, 2009).

(11)

10 Other discrepancies between male and female directors are differences in overconfidence and risk attitudes (Huang and Kisgen, 2013). It is stated that women tent do be less risk seeking than men and eventually men seem to be more overconfident. This overconfidence can also be associated with a higher self-attribution bias for men. This bias states that men have a higher tendency to attribute bad outcomes to external factors, but successful outcomes to their own action (Dowling and Lucey, 2010).

Another behavioral way in which men differ from women is that female directors are less driven by self-interest. According to Ibrahim and Angelidis (1994), female directors are less economically oriented and more philanthropically driven than male. Also, women are different in terms of ‘personality, communication, style, education and professional experience and expertise’ (Buss, 2005).

Overall, the role of gender in the boardroom seems to be important because women are both ethically and socially different from men. Women can enhance trust, reduce risk seeking, and better monitor different interests. To reduce the agency problem, a more gender diverse board can enhance the board diversity, and so improve the corporate governance mechanisms of the firm.

2.4.1 Gender diversity on firm financial performance

To examine the importance of the gender diversity on the board’s decisions and the corporate governance of the firm, previous research mostly focused on the effect on financial performance, or called firm value. Financial performance is measured using the firm’s Tobin’s Q (Carter, Simkins and Simpson, 2003; Rose, 2007), Ebitda (Rampling, 2011), return on equity (Carter et al. 2010) or return on assets (Ciavarella, 2017; Terjesen, Couto & Francisco, 2016), all financial factors.

Carter, Simkins and Simpson (2003) presented the first empirical evidence examining that a higher fraction of women on the board improved financial value of the firm. By regressing mostly U.S. firms they found that a higher fraction of women in the board positively affects the Tobin’s Q of the firm. Also Campbell and Minguez-Vera (2008) came to the same conclusion; however this research is only focused on Spain. Again only the Tobin’s Q was used as measurement of financial performance. They emphasize that the presence of women on the board does not in itself affects the firm value, but only the diversity of the board (the percentage of women) has a positive effect on firm value. This is also examined by Ciavarella (2017), stating that the existence of female directors has no effect on econometric specifications, however the gender diversity does have an effect.

A positive effect is also found by Terjesen, Couto and Francisco (2016). They used worldwide data of almost 4000 public firms, finding firms with more female directors have higher firm performance. However, firm performance is only measured by the Tobin’s Q and the return on assets, two financial factors. They imply that this research can be expanded by measuring firm performance in non-financial terms, like social performance or environmental sustainability practices (Terjesen, Couto and Fransisco, 2016).

(12)

11 In contrast, focussing on the Danish market Rose (2007) found no significant effect of the gender diversity on the Tobin’s Q. A plausible reason for this is that there might be a process of socialisation, in which unconventional board members adopted the norms of conventional board members. So in that case, the gains from having more female in the board are never reflected in the chosen performance measure. Also, the data is collected up to 2001, a period in which only four percent of Danish boards were represented by female (Rose, 2007). Also Carter et al. (2010) found no evidence of any effect of the number of women directors on the firm’s financial performance based on US corporations. However, they also found no evidence of any negative effect either. Based on this, they state that decisions concerning more or less women in corporate boards should be based on criteria other than future financial performance (Carter el al. 2010).

As already implied by Carter et al. (2010), measuring the effect of the board’s gender diversity on financial performance does not capture the effect of the board’s gender diversity on the corporate governance in general. This is also showed by Ionascu et al. (2018), measuring the percentage of women on boards on financial performance for Europe. They also found no significant impact on the financial performance; however they found that higher gender diversity is still beneficial for the major part of listed companies because it balances the social cohesion and economic components of sustainable development.

Overall, measuring the importance of the board’s gender diversity only focusing on the financial performance does not fully capture the effect on corporate governance. The contradictory results show a research gap when only measuring on financial performance. This is why Carter et al. (2010) and Ionascu et al. (2018) imply further research is needed, looking at other criteria like social cohesion and sustainable development, also influenced by the corporate governance mechanisms of the firm.

2.5 Corporate Social Responsibility

Considering financial performance is not sufficient to measure the impact of the board’s gender diversity on corporate decision making, the non-financial dimension ‘Corporate Social Responsibility (CSR)’ is taken into account (Gupta et al. 2015). At first is explained what CSR is and how this can be measured. Next some previous literature shows gender diversity in the board does affect CSR. The Corporate Social Responsibility is in most previous literature measured by the CSR disclosure or reporting. In section 2.5.3. is showed that CSR performance can also be measured by the Corporate Social practices or reputation. In the last part the gender quota laws are taken into consideration.

2.5.1 What is CSR performance

To examine the factors influencing Corporate Social Responsibility (CSR), at first it is important to define this term. According to McWilliams and Siegel (2001), CSR is described as ‘actions that appear to further some social good, beyond the interest of the firm and going beyond obeying the law’. Firms can benefit from achieving a higher CSR status, considering the ‘profit-maximizing CSR perspective’ (Siegel and Vitaliano, 2007).

(13)

12 This was also emphasized by Carroll (1999), stating that there are four components of social responsibilities, all contributing to the firm’s performance. He saw it as a kind of pyramid, starting with the economic-, then the legal-, next the ethical- and on top the philanthropically-component. He summarized the true social responsibility as meeting all four components, so the Corporate Social Responsibility of the firm should enhance the profit, obey the law, be ethical and be a good corporate citizen (Carroll, 1999, p. 43). Important is that the firm’s performance is not only increased in financial terms, but also social and environmental. More recently, CSR is defined by the European Commission as ‘the responsibility of enterprises for their impact on society.’ They state that a company becomes socially responsible by ‘integrating social, environmental, ethical, consumer and human rights concerns into their business strategy and operations’ (EC, 2016). Considering these broad definitions, at its core the CSR activities are related to economic, social and environmental responsibilities resulting from the company’s impact on the society. CSR has become popular not only because of increasing the firm’s economic value; it will also increase the firm’s social and environmental value, and eventually increasing the overall firm’s performance. Because CSR has such a broad definition, measuring CSR differs a lot in previous literature. In most cases, a content analysis is used to measure the overall CSR performance of a firm. Important to notice is, that by using this analysis only the CSR-reporting quantity is measured, and the methodology presupposes that social disclosure is a good proxy of overall Corporate Social Performance (Soana, 2011). Following the theory, disclosing CSR activities in the firm’s report is used to signal the superior performance to the market, and thus reduce the information asymmetry between managers and shareholders. So focusing on the reporting part of CSR, disclosing CSR activities in the reports is a way of reducing the agency problem and increases the firm’s performance (Guidry and Patten, 2012).

However, as shown by Soana (2011), the CSR performance can exist of more than only the reporting practices. For example reputational measures worked out by researchers or specialized journals exist of more than CSR reporting. These reputational measures are calculated by a score on ‘goodwill’, associated with the ‘company reputation on the basis of a subjective definition of social performance’ (Soana, 2011). The Corporate Reputational Index (CRI) is an example of measuring the CSR performance not only based on reporting, but as a proxy of the real company socially responsible profile.

Overall, the CSR performance is often measured by the CSR-reporting because the content analysis makes it easy to quantify these variables. However, the CSR performance consists of more than only the reporting practices, for example the firm adopting environmentally friendly practices and become socially responsible through good governance practices (Nadeem, Zaman & Saleem, 2017).

(14)

13 2.5.2 Board’s gender differences on CSR-reporting

Regarding the firm’s performance not only entails the firm’s economic value but also social and environmental aspects, measuring the effect of the board’s gender diversity on only financial performance seems to be limited as well. The board of directors has an important impact on monitoring and steering decisions of managers, which in return can affect the corporate governance of the board. So finding the effect of board’s gender diversity on corporate governance is limited as long as the corporate governance outcomes are only measured in terms of financial factors.

As shown by appendix I, most non-financial measurements on Corporate Social Responsibility are based on the reporting of CSR practices. Because the term CSR differs a lot in literature, there is no clear definition of what it should entail. Therefore, more than 36 different metrics have been used to assess CSR (Mellahi et al., 2015), in which reporting of CSR is the most easy and best-accessible measurement (Majeed, Aziz and Saleem, 2015). Focusing on the CSR reporting practices, for most research a positive effect of the board’s gender diversity on CSR disclosure is found. For example for the U.S., Frias-Aceituno et al. (2013) found a positive significant effect of the board’s gender diversity on the financial statements and integrated reporting. They explain this relationship by using the stakeholder theory; when the board of directors is more diverse, more diverse knowledge and opinions make it easier to supervise information documents and decision making processes of larger group of users is supported.

For the U.K. (Al-Shaer & Zaman, 2015) and Ghana (Agyemang and Konady, 2017) also a positive effect of the board’s gender diversity on CSR reporting quality and quantity is found. Even though the number of female in boards is very low for Ghana, Agyemang and Konady (2017) found a positive significant effect of women as chairperson on board in determining the level of CSR disclosure.

However, not all prior research finds a positive effect of the board’s gender diversity on CSR reporting. In appendix I can be found that for Indonesia (Handajana et al., 2014), Pakistan (Majeed, Aziz and Saleem, 2015; Mahmood et al., 2018), the U.S. (Giannarakis, 2014), Poland (Dyduch & Krasadomska, 2017), Bangladesh (Khan, 2010), Iran (Kamangari and Gerayli, 2017) and Malaysia (Rahman, Zain &Al-Haj, 2011) the board’s gender diversity did not – or did negatively- influence the CSR of a firm. Finding no effect can be due to various factors like culture, social education, economic problems or personality traits can affect the social performance not taken into account in their research (Kamangari and Gerayli, 2017). Also for companies in Indonesia the expected positive relationship between gender diversity and Corporate Social disclosure was not found (Handajana et al. 2014). They explain this differently, stating that the presence of women in the board is not due to their experience or expertise, but rather driven by family-ties to control the shareholder. Together with the very low number of female board members, the lack of their competence made it unable to encourage and improve corporate ethical behavior and even negatively affecting the impact on CSR disclosure.

(15)

14 The negative effect is not in line with the expectation derived from the theory; however this can be explained by other factors influencing the research. For example for Poland, Bangladesh and Pakistan, the overall female participation in decision-making is very limited. Even though women are more tend towards sustainability, their power to influence the board’s decision (and in that way the corporate governance) is low and is dictated by orientation of male directors (Majeed, Aziz and Saleem, 2015).

2.5.3 Board’s gender differences on CSR-performance

Whether previous research measured the non-financial firm performance mostly based on CSR reporting, Gupta et al. (2015) expands this by measuring the non-financial performance on overall CSR performance. Focusing on firms in the U.S. from 2003 till 2012, they find a positive effect of gender diverse board on social, environment and governance dimensions. No significant effect is found on financial performance. Following from its theory, they state that boards with higher gender diversity are more sensitive to stakeholders than just to shareholders, so it enhances the firm’s non-financial performance and general firm performance in longer term (Gupta et al. 2015).

Also Bear, Rahman and Post (2010) focus on the broader aspect of CSR performance, measuring the effect of the board’s gender diversity on the firm’s reputation, moderated by the CSR strength. In their research, they measure the CSR performance by using the KLD social ratings and develop two constructs. The first CSR performance-construct is the institutional strength, composed of positive actions towards the community and diverse stakeholders. For example, charitable giving, volunteer programs or promotion of minorities. The second construct is the technical strength and is composed of positive actions toward customers, stockholders and employees. The transparency in reporting social and environmental performance is only part of this construct, also actions like products with social benefits and health and safety programs are taken into account (Bear, Rahman and Post, 2010).

Even though Bear, Rahman and Post (2010) use different measurements for the CSR performance, they find the same result: the higher the board’s gender diversity, the higher the CSR performance. This is explained by the fact that ‘women bring an increased sensitivity to CSR and have participative decision-making styles’, both enhancing CSR strength ratings (Bear, Rahman and Post, 2010).

Another positive effect of the board’s gender diversity on decision making in overall CSR performance is found by Nadeem, Zaman & Saleem (2017). Instead of measuring the CSR performance by only measuring the CSR reporting quality, this paper expands its research by using a third party sustainability rating of Environmental, Social and Governance (ESG) scores from the Bloomberg database. With environmental aspects for example the water consumption, energy use, wastes management and greenhouse gases emission are measured. For the social aspect, for example the employee turnover, workforce diversity and workplace fatalities are taken into account to receive an overall ESG score. In this way, not only the CSR reporting is measured, but also the actual CSR practices are taken into account by looking at the impact of the board’s gender diversity.

(16)

15 Overall, the way the CSR performance is measured differs between the relevant literature. To find the effect of the board’s gender diversity on non-financial firm performance, in most cases only the CSR reporting quality is used as dependent variable (Frias-Aceituno et al. 2013; Agyemang and Konady, 2017; Handajana et al. 2014; Majeed, Aziz and Saleem, 2015, Mahmood et al. 2018; Giannarakis, 2014; Dyduch and Krasadomska, 2017; Khan, 2010; Kamangari and Gerayli, 2017; Rahman, Zain & Al-Haj, 2011). Papers by Gupta et al. (2015), Bear, Rahman and Post (2010) and Nadeem, Zaman & Saleem (2017) expand this research by also looking at CSR practices and reputation.

Previous research mainly focused on the U.S. and the U.K., but also some emerging countries like Pakistan and Ghana are considered. However, for these emerging countries no significant- or even a negative- effect is found, because in these countries the gap between men and women is even higher. There are not only less female present in the board, but they are also not as highly educated as men (Majeed, Aziz and Saleem, 2015).

2.5.4 European gender quota

As stated in the introduction, the European commission reported that on average women account for 25.3% of board members in the largest publicly listed companies in the EU (European Commission, 2018). This implies that the board’s gender diversity is higher than for example Ghana, so we are better able to find an effect of this diversity on the CSR by focusing on Europe. However, for Europe not much relevant literature is found for the specific effect of the board’s gender diversity on CSR performance. Fernandez-Feijoo, Romero and Blanco (2014) did involve some European countries in their cross-country research, and found a positive effect of diversity on CSR. By controlling for the different legal systems they try to capture the institutional effect of different countries.

Even though there is some evidence for a positive effect of gender diversity on CSR in Europe, this is not controlled for the gender quota laws, varying across the European countries. For the eight countries above the European average of 25.3% women in the board, only France, Italy, Germany and Belgium had a binding legislated quota target in 2017, see figure 1 below (European Commission, 2018). For the Netherlands, a quota was set, however there are no hard measures when not meeting this quota (Accountant.nl, 2019).

As shown by the red lines in figure 1, the legislated quota target is not the same for France, Italy, Germany and Belgium. In 2017, France set the highest binding quota regarding 40% for both non-listed as large listed companies. For Italy, 33% was set for listed companies and state-owned companies. Germany set the quota to be 30%, only for supervisory boards of the 110 biggest listed companies. Other companies being listed had to set individual quantitative objectives of women on boards. For Belgium, there is a legislated quota stated to be 33% for (non-) executives in state-owned and listed companies (Jourová, 2016, p.6). Nevertheless, as also stated by the Dutch minister of education, culture and science, lots of firms do not meet this quota. So the quota does not immediately imply that there are actually more women on the board (accountant.nl, 2019). But, when the quota is binding and there are so-called ‘hard measures’ when not meeting this quota, the board diversity will be higher.

(17)

16 There already is some literature about the effect of gender quotas on boards. For Norway, Wang and Kelan (2013) found that the binding gender quota increased the number of women being appointed to top leadership roles, including an increase in their independence status and qualification. Terjesen, Aguilare and Lorenz (2015) found a positive effect of the binding gender quota on the effectiveness of corporate governance, via the presence of more women. Also for Nielson and Huse (2010), a positive effect of the gender quota on the board’s effectiveness and strategic control is found.

However, the gender quota law can also negatively affecting the board’s effectiveness, as shown by Ahern and Dittmar (2012). They found that the legislative quota in Norway led to younger and less experienced boards, which changed the decision making of that board through less effective monitor- and advisor roles (Ahern and Dittmar, 2012). This was also examined by Adams and Ferreira (2009), stating that when the gender quota became mandatory the new board members had to be chosen out of a limited pool of available female directors. These (female) board members lacked experience and were younger than their male counterparts, leading to a negative effect on firm performance

Figure 1: Representation of women and men on boards of the largest listed EU companies (p.31 European Commission, 2018)

(18)

17 2.6 Hypothesis development

Considering the agency theory, the board of directors can reduce conflicts of interest between managers and stockholders (De Andres and Vallelado, 2008). In this way, monitoring and advising the management can enhance the corporate governance, and eventually the firm’s performance (Carter et al. 2003). The monitoring and advising capability of the board depends on the composition of the board, in which board diversity enhances management effectiveness and is an important mechanism for corporate governance (Carter, 2003)

Focusing on this board’s diversity, gender composition is a significant aspect. Women enhance the perception of the board’s legitimacy and trustworthiness, fostering shareholder trust in the firm (Perrault, 2015). Also the quality of monitoring is affected by more women in the board, because they are more likely to join monitoring committees (Adams and Ferreira, 2009). Female tent to be less risk seeking, less overconfident and less driven by self-interest (Huang and Kisgen, 2013; Ibrahim and Angelidis, 1994). In this way, female director are less economically driven than men. Overall, women tend to enhance trust, reduce risk seeking and better monitor the agency problem. A more gender diverse board enhances the board’s diversity, improving the corporate governance mechanism of the firm.

However, this effect on firm’s performance is mostly measured by looking only at financial performance. Taken this into account, the effect of the board’s gender diversity on the corporate governance can be examined by looking at the effect on CSR performance, because this enhances the firm performance as well (Gypta et al., 2015). Prior literature however does show some contradictory results, mainly for the U.S. and U.K. a positive effect of the board’s gender diversity on the CSR performance is found (Frias-Aceituno et al., 2013). Even though this is not the case for most emerging countries, this positive effect is also expected for Europe, because overall the women representation in the boards is much higher than for the emerging countries (European Commission, 2018). This leads to the first hypothesis:

H1: Higher gender diversity in a company’s board positively affects the CSR performance of the European firm

Although many studies examined the importance of female representation in the board, the board is still dominated by male directors (Wang and Kelan, 2013). Gender quota laws are introduced to diminish this effect. When the quota is binding, the board’s gender diversity indeed will increase. However, following both the implied negative (Ahern and Dittmar, 2012; Adams and Ferreira, 2009) and positive (Terjesen, Aguilare and Lorenz, 2015; Wang and Kelan, 2013; Nielson and Huse, 2010) effects of the gender quota laws on the board’s effectiveness, the relationship between more women in the board and the firm performance can be affected by that legislative quota. To test whether the legislative gender quota moderates the effect of the board’s gender diversity on firm performance, specifically the CSR performance, the following hypothesis is stated:

H2: A legislated gender quota does have a moderating effect on the relationship between the board’s gender diversity and CSR performance

(19)

18

3. Methodological approach

This chapter describes the methodology to test the hypothesis and answer the following research question. At first the data is described by dividing it into the dependent-, the independent- and the control variables. These variables are used in a multiple regression, explained in the last part.

3.1 Data sample

By collecting the data only the biggest firms in countries in the European Union are taken into account, all listed at the STOXX Europe 600 index. This sample is chosen because it expands previous research. The effect of gender diversity on CSR is mostly measured for firms in the U.S. (Bear, Rahman and Post, 2010; Frias-Aceituno et al. 2013; Giannarakis, 2014) or emerging countries (Mahmood et al., 2018; Majeed, Aziz and Saleem, 2015; Khan, 2010; Kamangari and Gerayli, 2017), only some European firms are taken into account when doing a cross-country analysis for companies over the whole world (Fernandez-Feijoo, Romero & Blanco, 2014).

The European sample is also chosen because the European commission implemented gender quotas in different EU member states, and these are for example not considered in the United States (Jourová, 2016). As stated in the Introduction, the European Commission proposed a directive under which women must represent at least 40% of non-executive directors in a firm and 33% if the legislation is also directed at the supervisory boards (European Commission, 2018). However, in many European countries this quota is not met, as shown in figure 1. To find the relationship between the board’s gender diversity and the CSR performance, only the eight countries are taken into account which are above the EU board’s gender diversity average of 25.3%. These countries are: France, Sweden, Italy, Finland, Germany, Belgium, Denmark and The Netherlands . Latvia also has a gender quota above the 25.3%, nevertheless firms in this country are not taken into account because these are not stated on the STOXX Europe 600 index, so no data is available.

The United Kingdom is left out, because of its different institutional environment in comparison to the other countries. The U.K. is a shareholder oriented country, while the other countries are more stakeholder oriented (Greenley and Foxall, 1998). This can influence the effect of the role of women in corporate bonds, because stakeholder oriented countries tend to pay more attention to corporate governance issues compared to shareholder oriented countries like the U.K. (Russo and Perrini, 2010).

Data for these eight countries is obtained from year 2003 until 2017. The sample starts in 2003, because in that year the Council of Europe adopted the ‘Recommendation Rec (2003)’, in which member states committed themselves to achieve a minimum representation of 40% of women in political and public life (Counsil of Europe, 2016). The sample time ends in 2017, because this is the most recent year for which complete data for all variables is available.

(20)

19 Also, the European Commission aimed to reach the gender quota by the end of 2020 (EC, 2016), so the more recent the data is, the more pressure the EC can put on countries to implement the gender quota law. Considering the EUROSTOXX 600 companies listed in the eight selected countries within these years, a sample of 490 firms active in (some of the) years 2003-2017. After taking out the firms for which no ESG data is available in Eikon, 462 firms remain.

To collect the data, the first database used is Eikon. In Eikon, environmental, social and governance data can be found. The control variables firm size, profitability, financial leverage and the SIC codes (for industry) are obtained from Eikon. Also, the main dependent variable measuring the CSR performance is obtained from Eikon. On February 2019, Thomson Reuters offered a document including the newest ‘ESG Scores’. This is an overall score which measures Environmental, Social and Governance scores of a firm and puts this into quantitative data. This score is used for this study, because it gives a transparent and objective measure of the company’s ESG performance across 10 main teams (Thomson Reuters, 2019). How this score is designed is further explained in the ‘dependent variable’ section.

The other dataset used is BoardEx, consisting of data about the board composition and size. The board’s average age, gender diversity and board independency is described at BoardEx, so the board-specific independent factors can be taken from there.

3.1.1 Dependent variable

Measuring Corporate Social Responsibility performance differs a lot in previous literature, because there is no clear definition of what is should entail. More than 36 different metrics have been used to assess CSR performance (Mellahi et al. 2015), in which the content analysis is used the most (Majeed, Aziz & Saleem, 2015; Rahman, Zain& Al-Haj, 2011; Mahmood et al. 2018). This is a technique in which data out of the annual reports is coded into various categories, so this only focuses on the reporting part of CSR. However, this technique does not always include the quality of the CSR, it focuses mostly on quantity. For example Majeed, Aziz and Saleem (2015) use the ‘keywords count’ in the annual report as measurement unit, indicating that the more words related with CSR, the higher the quality of the CSR reporting.

As showed in chapter 2, CSR performance can also be measured by other variables than the reporting quantity or quality. Papers by Gupta et al. (2015), Bear, Rahman and Post (2010) and Nadeem, Zaman & Saleem (2017) expand the research by also looking at CSR practices and reputation. Following Nadeem, Zamam and Saleem (2017), the CSR performance is measured by using a third party sustainability rating of Environmental, Social and Governance (ESG) scores, also taking into account CSR practices and not only reporting. To measure the effect of the board’s gender diversity on the CSR performance, this variable measures the actual CSR practices and quality, so it fits better than only measuring the reporting quality or quantity.

(21)

20 The data available for this research consists of the ESG variable obtained by Thomson Reuters (Thomson Reuters, 2019). This is in line with the ESG measurement used by Nadeem, Zamam and Saleem (2017) and measures the company’s relative ESG performance, commitment and effectiveness. It expands the content analysis method, in a way that also the reporting and performance quality is measured in the collection process. Not only the annual report, but also information of the company’s website, NGO websites, Stock Exchange Filings, CSR reports and news sources are taken into account. Both algorithmic and human processes are used to make sure the CSR quality is measured (Thomson Reuters, 2019). This process is shown in figure 7, Appendix II.

Figure 8 in appendix II shows how the Thomson Reuters database comes to an ESG combined score, the dependent variable used in this study to measure the CSR performance. As shown, the ESG combined score consists of the Thomson Reuters ESG score minus the ESG controversies which can negatively affect the CSR of the firm. The ESG score is based on 178 comparable and relevant fields to power the overall company assessment and scoring process. These are next grouped in 10 categories, which are defined in table 1 on the next page.

The combination of these 10 categories is weighted proportionately to the count of measures within each category, formulating the three pillar scores: the Environmental, Social and Governmental groups, the ESG (Thomson Reuters, 2019). The ESG controversies score is deducted to get the overall ESG s+core, because these controversies impact the CSR of the firm negatively. This exists for example when a scandal occurs, like lawsuits, ongoing legislation disputes or fines (Thomson Reuters, 2019). The TR ESG combined score is next calculated based on the percentile ranks scoring methodology, leading to a score between 0 and 1.

(22)

21 Table 1 ESG categories and their definitions (Thomson Reuters, 2019)

Pillar Scores TR ESG Score Definition

Environmental Resource Use Score The capacity to reduce the use of materials, energy or water.

Emissions Score The commitment and effectiveness towards reducing environmental emissions in the production and operational process

Innovation Score Capacity to reduce the environmental costs and burdens for its customers, creating new market opportunities through new environmental technologies and processes or eco-designed products

Social Workforce Score Effectiveness towards job satisfaction, a healthy and safe workplace, maintaining diversity and equal opportunities and development opportunities for its workforce

Human Rights Score Effectiveness towards respecting the fundamental human rights conventions

Community Score Commitment towards being a good citizen, protecting public health and respecting business ethics

Product Responsibility Score

Capacity to produce quality goods and services integrating the customer’s health and safety, integrity and data privacy.

Governance Management Score Commitment and effectiveness towards following best practice corporate governance principles

Shareholder Score Effectiveness towards equal treatment of shareholders and the use of antitakeover devices

CSR Strategy Score Practices to communicate that it integrates the economic (financial), social and environmental dimensions into its day-to-day decision-making processes.

(23)

22 3.1.2 Independent variables

To measure the impact of the board’s gender diversity on CSR, the main independent variable is the percentage of female directors to total directors in the board (Majeed, Aziz & Saleem, 2015; Rahman, Zain& Al-Haj, 2011; Mahmood et al. 2018). This score is provided in the database of Eikon, based on the following formula:

𝐵𝑜𝑎𝑟𝑑 𝑔𝑒𝑛𝑑𝑒𝑟 𝑑𝑖𝑣𝑒𝑟𝑠𝑖𝑡𝑦 = 𝑛𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝑤𝑜𝑚𝑒𝑛 𝑖𝑛 𝑡ℎ𝑒 𝑏𝑜𝑎𝑟𝑑

𝑡𝑜𝑡𝑎𝑙 𝑛𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝑏𝑜𝑎𝑟𝑑 𝑚𝑒𝑚𝑏𝑒𝑟𝑠∗ 100

To test if the gender quota affects the relationship between the board’s gender diversity and the CSR, a gender quota dummy is added. Countries with a gender quota get the value 1, countries without the gender quota receive the value 0. These gender quotas are only considered when they are binding, so for countries with a legislative gender quota. These are France, Italy, Germany and Belgium (European Commission, 2018). The independent variable is named BGD, and QUOTADUM.

3.1.3 Control variables

Control variables are used to control for the possibility that the firm’s CSR performance is due to external factors. These control factors are chosen because prior literature examined the factors to have an effect on CSR performance or disclosure in particular. At first there are some other board characteristics considered to influence the CSR, like board size (Said, Zainuddig& Haron, 2009), board independency (Mahmood et al. 2018) and board age (Handajani et al., 2014).

The board size is measured as the total number of directors on board, following Said, Zainuddig and Haron (2009). A higher board size is examined to positively affect the CSR disclosure, because there are more different members, bringing in more experience. Wang and Kelan (2013) also show a positive effect of the board size, focusing on the CSR overall activities. They state that a larger board has wider exchange of ideas and experience, leading to better involvement in CSR activities (Dyduch and Krasodomska, 2017). This variable is called BOARDSZ.

The board independency is also expected to influence the CSR performance. Following Cheng and Courtenay (2006), boards with a majority of independent directors have significantly higher levels of voluntary disclosure than firms with balanced boards. They state that independent directors aim for more transparency regarding sustainability, and thus should increase the CSR disclosure because they control the management more effectively. This is only based on CSR disclosure, but as stated by Soana (2011) the measurement of CSR disclosure presupposes that social disclosure is a good proxy of the overall CSR performance. Therefore, we can use the same control variables as used for the CSR reporting variable, implying they also have an effect on the CSR performance. Independence is calculated by dividing the total number of independent directors by the total number of directors in the board. The independence variable is called BOARDIND.

(24)

23 The board age is also taken into account, taking the average age of the board as variable. Following Handajani et al. (2014), it is expected that the older the board’s directors are, the higher the extent of corporate social responsibility. Older directors should be able to encourage the implementation of policies and strategies for CSR more than younger directors (Handajani et al. 2014). The variable is named BOARDAGE.

Other control variables based on internal and external factors of the firms are the company size (Terjesen, Couto and Fransisco, 2016), financial leverage (Dyduch and Krasodomska, 2017), profitability (Gamerschlag, Möller and Verbeeten, 2011) and the industry profile (Dyduch and Krasodomska, 2017).

The company size is based on total assets (Terjesen et al. 2016), stating that the higher the total assets, the larger the company is. The larger the company seems to be, the more it engages in CSR activities to ensure that the public sees the organization as legitimate, following from the legitimacy theory. Also, larger companies are more exposed to public scrutiny, so more likely to spend more on CSR because of its higher visibility (Barnea and Rubin, 2010). The variable is called SIZE.

As also will be stated in section 4.2.1., the variable based on total assets has to be converted to the natural logarithm (Zhang, 2019). This is done to control for the non-normality, in case of the total assets some firms will have very large total assets, for example in the finance industry. By converting the total asset-variable to its natural logarithm, these so-called outliers are omitted and the distribution of the data is normalized (Zhang, 2019).

The financial leverage is measured using the ratio of debt to total assets (Dyduch and Krasodomska, 2017). Following the agency theory, more highly leveraged firms disclose more voluntary information to reduce the agency costs and in that way the cost of capital. However, in the research by Dyduch and Krasodomska (2017) is also examined that there might be a negative link between the degree of leverage and CSR disclosure. Because a low degree of leverage ensures that creditor will exert less pressure to constrain the manager’s discretion over CSR activities (Purushothaman et al. 2000). As already stated for the BOARDIND factor, the financial leverage is expected to affect the CSR performance in the same way as the CSR disclosure (Soana, 2011). So the financial leverage might influence the CSR-performance, however it is not sure in which direction. The variable is called LEV. The third company specific variable considered is the firm’s profitability, measured by the return on assets. This is calculated by dividing the company’s net income by the total assets of the firm (Dyduch and Krasodomska, 2017). Following Gamerschlag, Möller and Verbeeten (2011), it is expected that profitable firms face higher social constraints and public exposure than less profitable firms. In this way, profitable companies may have to be more active on CSR performance, because they are operating within the norms of society (Gamerschlag, Möller and Verbeeten, 2011). The variable is called PRFT.

(25)

24 At last, the effect is also controlled for the industry profile, because the industry in which the firm is active might influence the level of CSR activities. Examining the effect on CSR disclosure, companies classified as consumer and energy supplying industries provide more CSR information than other firms (Gamerschlag et al. 2011). This might be due to the fact that firms in different industries face different forms of public visibility. According to Gamerschlag et al. (2011), consumer industry companies are more exposed to public position, so they have a stronger incentive to diminish political costs through CSR disclosure.

On the other side, it is stated that the service sector is less exposed to public position, because companies in the software, insurance or technology industry should have a lower environmental impact (Gamerschlag et al., 2011). Overall, a company being in a different industry can differently implement CSR reporting practices, and therefore act different on CSR performance.

This effect is also examined by Haniffa and Cooke (2005), stating that ‘the influence of industry type of CSD practice depends on how critical the effects of their economic activities impacts on society’. They found that two industrial sectors disclosed significantly more than other sectors; the trading& service- and the construction& property- sector (Haniffa and Cooke, 2005). This was partly in line with the expectation that only the consumer-oriented industries exhibit more social disclosure. Following Cowen, Ferreri and Parker (1987), they state that companies in this industry want to enhance their corporate image among market consumers, which can increase the amount of sales.

Both papers use a different categorization of the industry type. Gamerschlag et al. (2011) divide the companies over 18 different industries. For Haniffa and Cooke (2005), only five different industries are compared, using the Kuala Lumpur Stock Exchange classification. Depending on the chosen database, in this paper the Standard Industrial Classification is used, provided by Eikon (Thomson Reuters, 2015). Overall, to control for the industry effect, a dummy variable is used for the different industries. The variable is called PROFILE. An overview of the dependent-, independent- and the control variables is given in table 2 on the next page.

(26)

25 Table 2: Summary of all variables included in this research

Symbol Full name Measurement

Independent variables

BGD Board’s gender diversity The percentage of women in the board of directors QUOTADUM Gender quota dummy Dummy:

- 0 country without legislative gender quota - Countries with a legislative gender quota

Control variables: board characteristisch

BOARDSZ Board size Total number of directors on board

BOARDIND Board Independence Number of independent directors / total number of directors in the board

BOARDAGE Board age Average age of the board members

Control variables: Internal/external factors

SIZE Company size Total assets, in US dollars *1000 LEV Financial leverage Debt/total assets, in US dollars *1000 PRFT Profitability Return on assets, in US dollars *1000 PROFILE Industry profile Dummy by using the SIC codes

Effect of board’s gender diversity on Corporate Social Responsibility performance– comparison within Europe