The Impact of Board Gender Diversity on Corporate Social Responsibility
and Firm Risk: Evidence from the United Kingdom.
Msc International Business and Management
Master Thesis
Stephen Kiely S2507668
18/06/2014
3 Abstract
This thesis investigates the relationship between the board gender diversity, corporate social responsibility and firm risk amongst the largest FTSE 350 companies in the United Kingdom. Board gender diversity is found to be positively associated, although weakly, with increased corporate social responsibility performance in the case of both environmental and social CSR. Interestingly, evidence is found that supports a significant difference in CSR
performance between boards with one woman and boards with no women, but no increase in CSR performance with the further addition of women. This thesis also finds that this
4 Table of Contents Abstract 3 Table of Contents 4 Chapter 1 – Introduction 5 1.1 Research Question 8
Chapter 2 – Literature Review 14
2.1 Firm Risk 14
2.2 Corporate Social Responsibility 16
2.3 Board Gender Diversity 18
2.4 Women and Corporate Social Responsibility 20
2.5 Corporate Social Responsibility and Firm Risk 26
2.6 Women and Firm Risk 28
Chapter 3 – Methodology 30
3.1 Data Sources and Sample Selection 30
3.2 Variables 32
3.3 Analysis 36
Chapter 4 – Results 38
Chapter 5 – Analysis and Discussion 55
Chapter 6 – Conclusion 61
6.1 Managerial Implications 61
6.2 Limitations and Suggestions for Further Research 62
References 64
5 Introduction
In the light of the recent financial crisis and numerous other corporate scandals, it is ever more apparent that companies have been falling down in terms of their responsibility towards their shareholders and stakeholders. In recent years, companies have not only brought
enormous risk upon themselves, but on the economy as a whole. Questions are being asked regarding the oversight and governance of these companies, and whether boards are doing enough to ensure that companies are behaving as they should with regard to their wider societal responsibilities. It has been suggested that companies whose boards had more female representation, may have fared better through the economic crisis and afterwards as they were less exposed to the risk (Perlberg, 2012). However, the evidence on this is mixed, with some evidence that having more women on the board led banks to actually take part in more risky behavior leading up to the recent financial crisis (Berger et al., 2013), supported by research that people are more risk taking in the presence of the opposite sex (McAlvanah, 2009). This raises the interesting question of whether the presence of women on the board helps to lower the risk of companies and, if so, how does this occur given that their attitude to risk may be no different than that of male directors.
6
persistent female under representation and disproved common explanations at the time that related to lack of ambition, experience and commitment.
An important consideration is to what effect, if any, having women on the board of a
company might contribute to companies that have done so. While the effect of board gender diversity on financial performance has been under investigation for a long time now, it is only in recent years that research has begun into the impact of board gender diversity on the
corporate social responsibility performance of companies. In a US study, Bear et al. (2010) examined the effects of board gender diversity on firms’ corporate social responsibility ratings and in turn the influence of CSR on firm reputation. They found that board diversity had a significant positive relationship with a firm’s CSR ratings and that this in turn mediated a relationship with the institutional reputation of the firm. Bernardi et al. (2009) found that having a higher percentage of women on the board of directors of a Fortune’s 500 company had a positive association with being listed on the ‘World’s Most Ethical Companies’ list. An Australian study by Galbreath (2011) into the effect of board gender diversity on corporate sustainability showed gender diversity to positively affect social sustainability, but did not show a significant effect on environmental sustainability. These findings tend to suggest a link between board gender diversity and corporate social responsibility performance, which in itself, is interesting, but it is also useful to look at how this can materially benefit a company.
Previous research on the topic of firm risk supports the notion that corporate social
7
increased CSR performance and boards with higher female representation, this raises the question whether women board members might be lowering the risk of firms, and using indirectly CSR as a mechanism to do so.
This thesis is concerned with determining whether board gender diversity has a positive impact on the corporate social performance of companies and if this in turn lowers the risk to which the company is exposed. In order to determine if this is truly the case we first need to provide a theoretical and evidence based grounding for such an argument. We first have to set about determining whether the attitudes and viewpoints of women differ in regards to CSR from that of their male counterparts. This seems like an important prerequisite for an ultimate difference in the CSR performance of the board. However, there is also the possibility that when women are on the board of a company that the activities of the board change which could mean that the board is more active or that more open discussion is taking place, whereby ultimately better CSR decision-making is made. The next link in the argument is how a woman or a number of women influences the decision-making of a board when they are almost always a minority, as is especially the case in the UK. We provide a theoretical underpinning for how this may occur. This paper also looks at how the effect of women in the position of non-executive board member and executive board member may differ with regard to CSR performance, due to the differing nature of their responsibilities.
8 Research Question
The main research question of this thesis is whether the participation of women on the board of a company increases the corporate social responsibility performance of the company and whether this in turn reduces the risk of the company. The question follows the research by Bear et al. (2010) which demonstrated a similar link between board diversity, corporate social responsibility and corporate reputation in the healthcare industry in the United States. The proposition of this thesis is that the participation of women on boards may change the behavior of companies in the United Kingdom with regards to the wider societal
responsibilities of the companies, and that by doing so the company is at less risk of negative consequences and, ultimately, of damage to the company.
This research question is important because answering it will help to give a better understanding of the effect that board gender diversity may have on the corporate social responsibility performance of companies and the risk of these companies. It helps further the understanding of large companies in the United Kingdom and their decision-making
especially with regard to how they manage risk. Previous research has not considered board gender diversity and risk management in the same way as this thesis does, as this thesis proposes that CSR performance acts as a form of risk management and mediates a
relationship between board gender diversity and lower firm risk. Also, research in this field has not always differentiated between the particular aspects of CSR and so this research will give a more detailed picture of the effect of board gender diversity on CSR. Research of this type has not been carried out in the United Kingdom which is an important economy and has a different institutional structure and approach to corporate social responsibility and to corporate governance than in the United States, where most of the research has previously been carried out.
The conceptual model for this research thesis follows the line of the main research question. There are three main concepts in the model; board gender diversity, corporate social
responsibility and firm risk. There is evidence for board gender diversity being associated with higher corporate social responsibility performance and this in turn being associated with a lower level of firm risk; we expect to see corporate social responsibility mediate a
9
positive relationship as research by Bear et al. (2010) and Galbreath (2011) supports such a link. Adams and Funk (2012) indicate that more gender diverse boards may be more likely to embrace stakeholder interests to a greater extent, as female board members were found to be more benevolent and universally concerned than their male counterparts. The next element of the model is the relationship between corporate social responsibility and firm risk which we expect to be negative, as both theoretical (Waddock & Graves, 1997) and empirical evidence (Orlitzky & Benjamin, 2001) on the topic supports. The last element of the model is the relationship between board gender diversity and firm risk which because of past research we expect to a negative relationship. However, there is evidence that firms with higher female representation on the board have lower levels of risk (Wilson & Altanlar, 2009), but there is also evidence that the women in these positions are not more risk-averse in financial
decision-making than their male counterparts (Adams & Funk, 2012). This is why we propose that improved CSR performance may actually be responsible for a relationship between board gender diversity and firm risk through mediation.
Figure 1: Conceptual Model
There are a number of questions that have to be approached in the course of answering the main research question in order to determine if there are indeed relationships present. As part of the research question, in this thesis we investigate whether or not the participation of women on the board of a company increases some aspects of the CSR rating and not others. The reason for asking this sub-question is that previous research on sustainability (Galbreath, 2011) has shown gender diversity to positively affect social sustainability but not
10
understanding of the question. The overall CSR rating of a company is made up of a number of different elements such as social, environmental, corporate governance and economic sustainability. This paper distinguishes between the various elements that make up the total CSR score, in order to determine which particular parts, if any, women influence positively. Many previous papers have not done this, and doing so will ultimately help us to better understand the effect that board gender diversity has on CSR ratings. This thesis proposes a number of reasons why the participation of women on boards may lead to enhanced CSR performance such as women board members having appreciation for different issues, or attitude towards certain topics, than their male counterparts. There is also the possibility that while the participation of women may influence a certain aspect of CSR, this aspect may not have a significant role in the reduction of a company’s risk, and so it is important to locate and be clear on where the link may lie for this reason also.
Another sub-question that is investigated in this thesis relates to the effect that a woman in the position of Chief Executive Officer (CEO) might have on the CSR performance of a company. This is carried out to investigate if when a woman is in the more powerful position of CEO, there may also be a relationship present with the CSR performance of company that differs from when the CEO is a man. This is an important question to investigate; however, we note that the selection procedure for the position of CEO differs greatly from that of a non-executive board member. It is a very competitive process to become CEO of a large company and the person must pass through various steps in what is a difficult career
progression. This selection process may mean that by the stage it comes to the selection of a CEO, the men and women in the running for the position may be quite similar in terms of ability and mindset, due to self-selection for that career path in the first place and also due to the selection process of the employer.
11
some time ago meaning that the CEO is no longer the person who founded the company anyway, at least not for the female CEOs in the sample.
The positions of CEO and non-executive board member also differ greatly as the job of the CEO is to lead the management of the company whereas the role of the non-executive board member is to monitor and oversee management on behalf of shareholders. Because of this the same theoretical underpinnings do not apply to both positions in terms of their possible influence on CSR performance. However, if the attitude of women is shown to be different to men in terms of their views on CSR then there is still the potential that this could apply and lead to different CSR performance. If there is a relationship between CEO gender and CSR performance then we will continue to investigate whether a similar relationship is in place with the risk of a company, as if there is a relationship between gender diversity and risk we might expect to see this evolve when a woman is in the top job. As an extension of this question we also investigate whether a similar relationship may exist when a women is in the position of Chief Financial Officer. In this sample, the Chief Financial Officers are members of the board of directors and hold the position of executive director on the board. Given the significance of the role of CFO in the running of a company, it is important to determine whether a woman CFO might be associated with higher CSR performance or lower firm risk. As the CFO has quite a strong influence in the financial decision-making of a firm, it would be reasonable to expect that if different risk preferences really exist among men and women with regards to financial decision-making; we may see some evidence of it here.
12
their male counterparts, as there are a number of reasons why this may not necessarily be the case. There has been much research into the fundamental differences between men and women. However, there are a number of reasons also that we might expect the difference between women and their male counterparts to shrink as they get into top management positions. Niederle and Vesterlund (2008), show that women often try to avoid competitive environments such as top management positions. Therefore, it is possible that women who do pursue leadership positions might more be similar to the men who do so, than the average man and woman. Therefore, there is reason to believe that a certain amount of self-selection occurs in regard to the women that make it to top management positions i.e. those women who are not interested being on boards or are not suited to it, may not end up in that position. Those women that do end up with a position on the board of directors may generally be more similar to their male counterparts on the board as they have come through the same
environment and reacted to similar incentives. Women who are in a predominantly male environment may also adapt their behavior so that the expected gender differences in attitudes and behavior may diminish.
A question that also needs to be addressed is whether female participation on the board of a company leads directly to a reduction in the risk of a company. For this to be the case then the women who are on boards would most likely have to be more risk-averse than their male counterparts. The evidence on this subject conflicts as even though women are thought to be more risk-averse in financial decision-making (Powell & Ansic, 1997), in the case of average men and women; research has not found the same among risk attitudes of women and men whose job involves financial decision-making (Atkinson et al., 2003). This thesis addresses the issue that if the women on boards are not actually more risk-averse in financial decision-making than their male counterparts, as some research suggests, then how they might they be responsible for lowering firm risk.
The next question that is to be investigated is whether an increase in CSR performance acts as risk management and reduces the risk of a company in the United Kingdom. This
13
Another point that is addressed in the literature review is the way in which a woman, or a number of women, who are almost always the minority on boards, may actually go about influencing board decision-making. They first have to be able to affect decision-making, if they want to change the way that the company acts, whether that be in a more responsible direction or not. This issue will be examined through the use of a number of theoretical frameworks. This thesis also looks at the effect that the number of women on the board has, as quite a bit has been written about the possibility of tokenism with regard to companies having just one woman on the board and also with regard to there being a critical mass, or certain of number of women on the board, that when reached changes the performance of the company. Tokenism suggests that a person is in a position with no real power to influence decision-making and is essentially just there as a token. There is an investigation of whether a company with one woman on the board has a significantly higher rating than one with no woman on the board. This is done to determine if tokenism is at play, as we might expect to see to no significant difference in performance between all-male boards and boards with one woman on them. The critical mass theory (Kanter, 1977) provides an explanation of what might happen as the number of women increases on the board of a company. Critical mass theory supports the idea that where once there are enough women present they are able to work together to further issues important to them, or else they are no longer solely viewed as a representative of their gender and their opinions are taken more on their merits. Scholars have argued that this happens once a critical mass of female participation has been reached, previous research suggesting that this is the case when three women are on the board (Joeck et al., 2013), this can leads to a difference in the performance of a company. This might also be the case with CSR performance and so we examine the particular breakdown of when the effect on CSR might take place and in turn, how this might impact the risk level of the firm.
In summary, the problem that this thesis sets out to solve is to develop a greater
14 Chapter 2 – Literature Review
This literature review starts first by discussing in more detail the three major concepts of the paper; firm risk, corporate social responsibility and board gender diversity, and then we continue by explaining how they relate to one another in the theory. The review of the relevant literature then leads to the formulation of the paper’s hypotheses.
2.1 Firm Risk
Truly successful economic performance shows itself in two ways, both in high returns and in low risk. Risk is one of the most important concepts in the world of business. Past events such as the recent economic crisis and the dot-com crash of the early 2000s before that have shown that financial performance is not the only measure of a business that matters and that underlying risk is ignored at the peril of businesses, investors and society as a whole. There are no situations where risk is completely absent and as such there is always a trade-off between risk and return. Risk has been defined in many different ways, often depending on the context; however, at its most basic it is the probability of variance in an expected
outcome, in this context this variance equates to financial loss (Oxfordictionaries.com, 2014). It is commonly seen as a negative consequence with the focus usually on the downside, and this has been observed by the finding that the idea of economic loss fits more closely with manager’s perspective of risk rather than just variance itself (Chiles & McMackin, 1996). This is an important distinction as it affects the decision on which particular type of risk we focus in this thesis. Bettis (1983) wrote that managing this risk lies at the heart of competitive strategy. Many other theorists have proposed that the management of business risk is central to organizational evolution, which determines which organizations will survive and grow and which will decline and ultimately die (Child, 1972; Summer, 1980). There are a number of different types of risks spoken about in business and these risks present themselves in a number of different ways.
15
true picture of the risks of a company (Brilloff, 1972; 1981). The second method of measuring risk is to use market measures, such as volatility, which focus on market
performance and are more proximate to the effects of CSR performance and therefore show a more marked response than accounting measures (Orlitzky & Benjamin, 2001). These types of measure are less susceptible to different accounting procedures and represent investor’s evaluations of the ability of a firm to generate future earnings (McGuire et al., 1988). However, there are also obstacles with measures based on the stock market, as according to Ullmann (1985), the use of market measures suggests that investor’s valuation of a firm’s performance is indeed a correct and proper performance measure (McGuire et al., 1988). Indeed, it is also true that no one risk measure can capture all possible risk that a company is exposed to.
16 2.2 Corporate Social Responsibility
Corporate social responsibility (CSR) is an extensive concept in the field of business. The implications of CSR and its components have grown steadily throughout the past decades (Carroll & Shabana, 2010), and it has become an integral part of business practice over the last decade or so. Many companies, and almost all large ones, now include a section
dedicated to their CSR performance in their annual reports and on their websites, or produce separate CSR sustainability reports altogether. Initiatives such as the Global Reporting Initiative (GRI) have become the norm with 4000 companies in 60 countries, including many of the world’s largest companies, now following the guidelines outlined for sustainability reporting or triple bottom line reporting (Database.globalreporting.org, 2014). This illustrates the importance to which companies now attach to such activities.
CSR also encompasses many different issues, from charitable giving to reduction of carbon emissions to childcare for employees. Obviously, these are quite disparate topics but all come under the umbrella of corporate social responsibility. Part of the reason for this is that there are many definitions for CSR. There is no single dominant or agreed upon definition for corporate social responsibility among companies and so each company differs in how they approach corporate social responsibility, with some focusing on certain aspects more than others. According to the European Commission (Europa.eu, 2011) corporate social responsibility is:
“the responsibility of enterprises for their impacts on society. CSR concerns actions by companies over and above their legal obligations towards society and the
environment. Certain regulatory measures create an environment more conducive to enterprises voluntarily meeting their social responsibilities.”
While the World Bank (Ward, 2004) defines corporate social responsibility as:
“the commitment of business to contribute to sustainable economic development, working with employees, their families, the local community and society at large to improve quality of life in ways that are both good for business and good for
17
Carroll (1991) noted that while companies were required to meet their economic and legal responsibilities, ethical and philanthropic activities were expected from companies as members of society. It is on this basis that the notion of corporate social responsibility are built and of good corporate citizenship, that companies are citizens in the society they operate and that they should act according to the norms and rules of said society. Scholars have argued that there are a number of ways in which CSR can have a positive impact on a company’s performance. Research has suggested that it leads to better access to valuable resources (Cochran & Wood, 1984; Waddock & Graves, 1997) and helps firms to attract and retain higher quality employees than they would otherwise (Greening & Turban, 2000). There is also the potential to allow for the better marketing of products and services, as consumers may have a preference for purchasing more sustainable products and services to which companies can appeal (Moskowitz, 1972). CSR can also have a positive impact by creating unforeseen opportunities for the company (Fombrum et al., 2000) and by contributing towards gaining social legitimacy (Hawn et al., 2011). Also socially responsible firms are less at risk from negative rare events as they are more aware of negative social and
environmental externalities. Many firms embrace CSR, and are ultimately successful by doing so, because it gives them a competitive business advantage over firms that do not, through new innovations, customer engagement, cost savings and brand differentiation (Carroll & Shabana, 2010; Smith, 2011). This competitive advantage extends into many functions of the firm especially its relationship with its stakeholders. By signaling that they care about their stakeholders and wider society as a whole, companies are able to gain business and to improve their reputation.
However, research rooted in neoclassical economics has argued that CSR unnecessarily raises a firm’s costs, putting the firm in a position of competitive disadvantage in comparison to its competitors (Friedman, 1970; Aupperle et al., 1985; McWilliams & Siegel, 1997; Jensen, 2002). The contention from this point of view is that resources are being taken away from the core focus of the company and that management is in fact destroying shareholder value. Friedman (1970) especially argued that it was wrong for management to give to charity using company funds and that should they wish to give money to charity they should use their own.
It is true that engaging in CSR activities often does involve accruing extra costs to the
18
meta-analysis on the topic of CSR and financial performance found an overall small but positive relationship, which indicated that CSR at least does not harm the financial
performance of firms and that they should therefore not be afraid to partake in these activities to some degree at least (Margolis et al., 2009).
2.3 Board Gender Diversity
Board gender diversity is a topic of much discussion and research (Adams & Ferreira, 2009; Francouer et al., 2008). The under representation of women on the boards and in the top management of companies is of concern to many governments.
There are two arguments surrounding female representation on the boards of companies, the ethical argument and economic argument. The ethical dimension of board gender diversity is not really up for debate in this thesis, except to say that it is quite immoral and illegal to discriminate against one gender over another. On the other hand, some argue quotas that oblige companies to have a certain percentage of women on their board are in fact
discriminatory and may lead to the appointment of people who are not the best candidate for the position.
However, the focus of this thesis is rather on the effect that board gender diversity might have on companies as opposed to weighing ethical and economic dimensions of the argument. If an economic argument were to settle on the side of all-male boards having superior
performance, then it would very likely still be a short-sighted position to take to exclude half the population from consideration as potential board appointees, aside from the fact that it would be unethical and illegal.
19
i.e. the CEO and CFO, and the non-executive directors whose job it is to monitor the
management. Entrusting the company into the hands of the managers can lead to opportunism which can damage the company and so the board is charged with monitoring the management by the shareholders. If the board is complacent then they may fail in their accountability to shareholders and stakeholders and allow management to act in its own self-interest. The executive directors clearly aren’t charged with monitoring themselves as that would still leave the potential for the agency problem. According to Terjesen et al. (2009), a common assumption in agency theory is that outside directors are more likely to act independently than their inside director counterparts and will therefore act as good monitors for
shareholders’ interest. In one sense women board members may be more likely to be the outsider in this case as they are unlikely to be part of the existing “old boy’s club”.
20
An important theory in the field of the board gender diversity is critical mass theory. Critical mass theory proposes that until a certain a threshold or critical mass of a minority in a group is reached, then the minority individual may only be regarded as a token figure and not substantially influence group decision-making processes (Kanter, 1977). It puts forth that where there are only one or two women on a board, the dominant contingent (i.e. males) has control over the group and its culture and women board members are not treated as
individuals but as representatives for their “category”. However, when a certain number of minority members is reached they may ally together and influence the culture of the group. It is important to note that Kanter (1977) did not postulate distinct female traits would lead women and minorities to perform those jobs differently than similarly qualified males once a critical mass was reached.
In a widely publicized report, Kramer et al. (2006) explored critical mass theory in interviews and discussions with fifty women directors, from which they concluded that “having three or more women on a board can create a critical mass where women are no longer seen as outsiders and are able to influence the content and process of board discussions more substantially”. In contrast to Kanter (1977), they proposed that a critical mass of women brings distinctly feminine traits to the boardroom including a tendency towards being more collaborative, more active in asking questions and raising different issues. A study of the link between board gender diversity and the financial performance of German firms found that the relationship followed a U shape (Joecks et al., 2013). The performance of companies with gender diverse boards fell short of those companies with all male boards, until a “magic number” of three women was reached, whereupon these companies’ performances surpassed uniformly male boards, supporting critical mass theory in the boardroom. It is wise then to investigate whether there is a critical mass of woman board members that, when reached, affects corporate social responsibility performance. This is done in an effort to better understand the dynamics of the relationships under investigation in the thesis.
2.4 Women Board Members and Corporate Social Responsibility
21
governance and economic sustainability. This thesis will look at these individual measures to discover which might be associated with female participation on boards and to determine the overall effect that board gender diversity has on CSR.
A study of Fortune 500 companies found that those companies with more women on the board were more likely to appear on the “World’s Most Ethical Company” list (Bernardi et al., 2009). Furthermore, McCann and Wheeler (2011) found that female board participation was associated with higher corporate responsibility performance in a study of FTSE 100 companies. These findings indicate a positive relationship between the number of women on the board and the CSR performance of a company. According to Hillman et al. (2002), women on boards are more likely than men to be support specialists and community influencers, and that because of this they may sensitize the board to CSR initiatives and provide perspectives helpful in addressing CSR issues. Women who are appointed to boards often have different backgrounds and experience to their male counterparts and so they may also bring with them different resources and insights (Zelechowski & Bilimoria, 2004). It is a possibility that women board members could be appointed to boards for the purposes of dealing with CSR issues if they are indeed community specialists and so this might go some way to explaining why there may a relationship between board gender diversity and CSR rating. Research by Gul et al. (2011) suggests that boards with more women on them could benefit from improved quality of discussions and greater oversight of firm’s disclosures and reports. This would indicate that more attention was given to CSR and so greater results could be expected from companies with more female representation.
While one woman on the board may make a positive impact, there is also the possibility that tokenism is taking place and that because the female director is in the minority her opinions and contributions are not taken seriously. This paper investigates the effect on CSR
performance as the number of women increases to discover if tokenism is taking place, and if there is a critical mass of females on the board e.g. three, that makes a significant difference (Joecks et al., 2013). The studies that have previously been carried out in this area indicate that there is a positive relationship between female board participation and corporate social responsibility performance and so we expect to find similar results in this regard.
22
Hypothesis 1b: There is a critical mass of women board members that is associated with increased CSR performance.
The effect that female board members have on the social responsibility of companies has been investigated by a number of authors. This is a very important aspect of CSR as it is one which customers and stakeholders of the company may be most sensitive to. Research carried out by Adams and Funk (2012) indicates that more gender diverse boards may embrace stakeholder interests to a greater extent, as female board members were found to be more benevolent and universally concerned than their male counterparts.
A study of the board members’ gender on corporate social responsiveness orientation in S&P firms carried out by Ibrahim and Angelidis (1994) indicated that women were more oriented towards discretionary elements of corporate responsibility than men (who were more
concerned with economic performance). There were, however, no significant gender differences with regard to legal and ethical dimensions of corporate social responsibility. Bernardi and Threadgill (2011) found that companies with women on the board were more likely have a formal employee volunteer program, sponsor charities and to have stronger relationships with surrounding communities. However, this could be because there tend to be more women on the boards of larger companies which also happen to be those with the resources to engage in these kinds of activities and to appoint board members for the purpose of dealing with these issues. In a study of board gender diversity and corporate philanthropy, Williams (2003) finds that female board members are more sensitive to corporate
responsibility activities. This may influence the way that boards approach their internal and external responsibilities. The evidence from previous studies therefore seems to indicate that there is a positive association between female board participation and social performance.
Hypothesis 1c: The number of female members on a firm's board is positively related to the firm's Social CSR rating.
23
environmental aspects of CSR. Interestingly, they found that having more than three women on the board had a positive influence on environmental CSR. This could indicate that a critical mass is reached at this point which allows women board members to enact the changes to company decision-making that they desire. The impact of women top managers and directors on corporate environmental performance was investigated in a study of 500 large US firms, which found that that firms that incorporated women in their top management team and board of directors exhibited superior environmental performance, with the impact being greater for the board (Kimball et al., 2012).
However, in a study of Australian companies Galbreath (2011) found no significant relationship between female directors and environmental sustainability. This is in line with the majority of these types of studies, where board member diversity has not had a significant relationship with environmental sustainability performance, although it has not been
investigated sufficiently. One reason that female board members might not have a strong influence over environmental issues is that male board members are more likely to have degrees and backgrounds in more technical disciplines, including engineering and science, than female board members (Singh et al., 2008). This may lead the board to weigh their opinions more strongly as issues dealing with the environment such as climate change and carbon emissions reduction tend to be associated with hard science, technology and engineering. The board may believe that those board members with these backgrounds are more qualified to give advice on the subject of the environment and the company’s options for dealing with environmental issues.
The evidence that female board participation influences environmental CSR performance therefore is not very conclusive, and so we propose that there is no significant relationship between the two.
Hypothesis 1d: The number of female members on a firm's board is not associated with the firm's Environmental CSR rating.
24
more gender diverse boards appear to be tougher monitors than those without women. They found that female directors have better attendance records at board meetings than their male counterparts and that they sit on more monitoring-related committees than male directors. Male directors were also found to have fewer attendance problems the more diverse the board was. A Canadian study conducted by Brown et al. (2002) found that boards with three or more women performed significantly better in terms of corporate governance than all male boards, explained that this was the result of improved risk management and monitoring. Stephenson (2004) also argued that firms with more female directors tended to have improved risk management and were more likely to employ non-financial performance measures, such as innovation and social responsibility. However, there is an alternative to the previous point of view. Finkelstein and Mooney (2003) found that demographic diversity was not shown to improve governance performance in Standard & Poor’s (S&P) 500 firms. Siciliano (1996) also discovered no evidence for improved board operating efficiency which suggested that diversity did not influence the board’s monitoring role.
Hypothesis 1e: The number of female members on a firm's board is associated with the firm's Governance CSR rating.
25
their study of 2,500 Danish firms. However, Marinova et al. (2010) found no such positive effects in their study using evidence from 186 listed firms from the Netherlands and Denmark which indicated that board gender diversity had no effect on firm performance. While the findings seem to indicate that there could be a relationship between board gender diversity and financial performance, the measure used by the Thomson Reuters Asset 4 database is an aggregate of many different financial measures and is unlikely to reveal any telling results due to this. It is however included for the sake of transparency as it used in the calculation of the overall CSR score.
Hypothesis 1f: The number of female members on a firm’s board is not associated with the firm’s Economic sustainability CSR rating.
Although the previous evidence supports a link between female board participation and corporate social responsibility performance, we propose that these reasons will not extend to the situation where a woman is the Chief Executive Officer of a company. There has not been a lot of research carried out regarding this particular link, although Manner (2010) did carry out a study about the influence that CEO characteristics may have on corporate social performance. It found that having a bachelor’s degree in the humanities, having a breadth of career experience and being a woman had a positive influence on social performance.
However, we propose that the same explanations for the link between female board members and social responsibility, such as a more benevolent and universally concerned attitude (Adams & Funk, 2012) and a greater propensity for charitable giving (Ibrahim & Angelidis, 1994; Soares et al, 2011), may not apply in this case. These factors are positive influences on CSR performance; however we may not expect to see a positive association between a female CEO and a firm’s CSR rating, as the position of CEO and non-executive board member differ quite a lot in the responsibilities involved. There is a very different selection process for the positions and little evidence that a female CEO after going through this process of selection will be significantly different than a male CEO. We also investigate the presence of a link between a female Chief Financial Officer and a company’s CSR performance, however again we do not expect to see a significant relationship between the gender of the CFO and CSR performance. The theoretical basis for the influence that non-executive board members might have on CSR performance do not apply well to the positions of CEO or CFO, as their
26
association between the gender of either the CEO or the CFO and the CSR performance of a company.
Hypothesis 2a: A female CEO is not associated with a firm's CSR rating.
Hypothesis 2b: A female CFO is not associated with a firm’s CSR rating.
Corporate Social Responsibility and Firm Risk
As mentioned earlier in the text, truly successful business performance shows itself in two ways, both in high returns and in low risk, and if corporate socially responsible behavior by firms were to heighten this risk then firms may not see the positive benefits to their financial performance from such corporate socially responsible engagement (Orlitzky & Benjamin, 2001; Orlitzky et al. 2003). Corporate socially responsible behavior is expected to decrease firm risk from the perspective of both the instrumental stakeholder theory (Donaldson & Preston, 1995; Jones, 1995) and good management theory (Waddock & Graves, 1997) because firms that disregard their stakeholders concerns e.g. pollution, are at a higher risk of negative actions such as litigation and damage to their reputation and brand. By addressing stakeholder concerns, corporate social responsibility can act as risk management of social and environmental issues that could potentially harm the company and reducing the probability of a crisis or a negative rare event.
27
The CSR performance variables are an aggregate score calculated depending on the strengths and concerns of the firms related to CSR issues. Lankoski (2009) found the impact of social performance to be more positive for reducing negative externalities (e.g. reducing concerns), than for issues which generate positive externalities (e.g. having strengths). This conforms to the stakeholder theory framework, as she argues that avoiding or reducing concerns are a priority for stakeholders, while having CSR strengths are only secondary in the expectations of shareholders. It also supports the notion that CSR acts as risk management in this way but possibly only when the CSR itself is related to more risky activities. A study by Luo and Bhattacharya (2009) showed that higher corporate social responsibility performance lowers undesirable firm-idiosyncratic risk, which is the risk that is specific to a company and not the market as a whole.
In the past CSR has been considered by many managers and financial analysts as an activity that actually increases risk because it was believed to take resources out of the main function of the business without providing a benefit to the company. However, Kytle and Ruggie (2005) contend that CSR is related to corporate risk management in two ways, first by providing information about what the risks to the company are, and second by offering the company effective means to respond to those risks. According to Kytle and Ruggie (2005), corporate socially responsible behavior can be used by firms to manage social risk and
therefore firms that are better able to manage stakeholder relationships this way are less likely to be engaged in the court of public opinion by their stakeholders where they have less
control and there is greater potential for reputational damage. Firms that proactively anticipate and address environmental and stakeholder concerns are likely to be in a better position to decrease the variability of their business returns (Orlitzky & Benjamin, 2001).
28
and future prospects to be riskier than that of a high-CSR company (Waddock & Graves, 1997 ).
The next reason is that investors and lenders now make greater use of CSR screens and therefore companies with low CSR performance may find their access to market capital restricted which leads to greater financial risk (McGuire et al., 1988). A paper by Jo and Na (2011) investigating the relationship between CSR engagement and firm risk in controversial industry sectors e.g. tobacco, alcohol, gambling, sought to find out whether firms in these industries were able to lower the risk of their companies through socially responsible
behavior even though they were taking part in industries which were viewed to be detrimental to people, the environment and to society. Their research uncovered that there was a negative between CSR and firm risk in these controversial industries. There was also some evidence in one sample of a positive relationship between CSR engagement and firm risk which possibly supported the view that companies were attempting to window-dress their activities or
legitimize questionable business activities without trying to reduce risk. Godfrey et al. (2009) found that CSR activities carried out by the firm aimed at a firm’s secondary stakeholders and society at large actually provided an “insurance-like” benefit but that participation in technical CSR which was aimed at trading partners carried no such benefit. Given the support from previous research, we propose that there is a negative association between the CSR rating of a company and the risk of the firm.
Hypothesis 3: There is a negative relationship between the CSR rating of a company and the risk of the company.
Women Board Members and Firm Risk
29
However, most of the studies related to risk attitudes have been done with workers and the general population. As we have previously shown, there are a number of reasons why we might expect the women who are appointed to boards to not be typical of the general population. A study by Adams and Funk (2012) into the boards of Swedish publicly-listed companies showed that female directors were less traditional and security oriented as well as more willing to take higher risks than their male counterparts. Another study carried out by Atkinson et al. (2003) compared the performance and investment behavior of male and female fixed-income mutual fund managers and found no significant difference between males and females terms of performance, risk management or other fund characteristics. This tends to support the argument that having more women on the board might not lead to more risk-averse decision making. However, the proposition of this thesis is not necessarily that women board members are more or less risk-averse in financial decision-making than their male counterparts. Instead, we propose that through a positive impact on firm corporate social responsibility and in turn its relationship with firm risk which we explained previously, board gender diversity might ultimately help to lower the level of risk of the company.
30 Chapter 3 – Methodology
As mentioned in the previous chapters the main aim of this research is to examine the relationship between gender diversity on the boards of FTSE 350 companies, the largest companies on the London Stock Exchange, and corporate social responsibility performance and its eventual impact on firm risk. This study is based on a quantitative analysis of board composition data from FTSE 350 companies, their corporate social responsibility scores and firm risk measures. The analysis contained in this thesis is of the period of the year 2012 which was the most recent year for which there was complete data regarding CSR performance for the majority of the companies in the sample.
3.1 Data Sources and Sample Selection
The data for this research is obtained from a number of different sources. This thesis uses external secondary data sources, meaning that the information was collected by others and was not collected first hand by this researcher. The advantage of using this type of data source is that it would have not been feasible to collect this data for such a large sample using primary methods. The disadvantage of using secondary data is that there is a reliance on the original collection of data having been done properly. In the case of this thesis, only reliable and transparent data sources are used, and so the best efforts are made to ensure that the findings of the thesis are valid and meet the highest standards of reliability.
This thesis makes use of the Thomson Reuters Asset 4 database for the corporate social responsibility performance variables. There is much variation in the CSR data sources and variables that are used in research of this kind; however, this database contains
environmental, social and governance (ESG) ratings which are the standard categories of CSR. It also contains a rating for economic sustainability which we also investigate for the sake of transparency as it contributes to the overall CSR score. The Thomson Reuters Asset 4 database deals specifically with company CSR performance data and contains 4000 of the world’s largest companies, which includes the largest 328 publicly listed companies
operating in the United Kingdom. The CSR information in the database is based on over 250 key performance indicators (KPIs) and over 750 individual data points, which are organized into 4 pillars of CSR performance; environmental, social, corporate governance and
31
120 experienced analysts that collect and standardize the data ensuring its accuracy and comparability”(Thomsonreuters.com, 2014).
CSR performance data was not available for all the UK companies in the database for the period in question and therefore these companies were removed from the sample. The companies that are included in the sample are listed in Appendix 1 at the end of the thesis. This thesis also makes use of the Cranfield School of Management Female FTSE Report, which is produced annually in conjunction with the Barclay’s Group and with the support of the UK government. The Female FTSE Report examines the progress of women in the boardroom in the United Kingdom boardrooms and examines where targets are not being met. The data on the genders of the Chief Executive Officers and Chief Financial Officers in the sample of companies was retrieved from these reports and annual reports were consulted where any questions arose in order to ensure accuracy.
In order to determine the true relationship between the independent and dependent variables under investigation we need to account for other factors which could impact the CSR performance or firm risk of the companies in the sample. We first determined which other factors i.e. control variables, could influence our dependent variables. The company data required for the control variables e.g. market capitalization, used in the linear regression analyses were then obtained from the Thomson Reuters Worldscope database. The stock return data which was used to calculate the firm risk variables was obtained was also
retrieved from this database. This database contains fundamental financial data on public and private companies from around the world. Thomson Reuters Worldscope is “the premier fundamentals database for deep international company coverage and history” and “offersthe broadest company coverage, representing 99% of the global market capitalization”
32 3.2 Variables
Independent Variables
Gender Diversity Absolute: This variable is represented by the number of women on the board of the company. It is lagged by one year under the assumption that the women on a company’s board must first be there a certain period of time in order to have an influence on the decision making and performance of the company.
Gender Diversity Percentage: This variable is the number of women on the board expressed as a percentage of the total number of board members of a company. It is also lagged by one year under the assumption that the women on a board must first be there a certain period of time in order to have an influence on the decision making and performance of the company.
CEO Gender: This variable is a dummy variable which assigns a company 1 if the Chief Executive Officer of the company is a woman and 0 if the Chief Executive Officer is a man. For the purpose of the study a company is only assigned 1 if the woman who is CEO has been in place for at least a year prior to the measurement of CSR performance. This is done under the assumption that they need have been in the position for a period of time to have an influence on decision making and the performance of the company.
CFO Gender: This variable is a dummy variable which assigns a company 1 if the Chief Financial Officer of the company is a woman and 0 if the Chief Financial Officer is a man. For the purpose of the study a company is only assigned 1 if they woman who is CFO has been in place for at least a year prior to the measurement of CSR performance. Again this is done under the assumption that they need have been in the position for a period of time in order to have an influence on performance.
33 Mediator Variables
Overall CSR score: This variable represents the overall corporate social responsibility performance of a company. Each company is assigned a score between 0 and 100. It is determined as the mean value of the four indicators of corporate social responsibility performance i.e. social, environmental, corporate governance and economic sustainability. Overall CSR Score =
= Overall CSR Score = Social Performance Score
= Environmental Performance Score = Corporate Governance Performance Score = Economic Performance Score
CSR performance acts as a mediator variable when we are determining gender diversity’s relationship with firm risk. The overall CSR score is used as the mediator variable in this research as it represents all aspects of corporate social responsibility; however, we investigate the relationship between board gender diversity and the individual aspects of CSR in order to better understand where the affect may lie.
Environmental performance: This is a sub-score of the overall CSR performance score. It
measures a firm’s resource wastage, emissions reduction and product innovation. A score of between 0 and 100 is attributed to each company.
Social performance: This is a sub-score of the overall CSR performance score. It measures a
firm’s performance in the areas of training and development, health and safety, product responsibility, community awareness, human rights behaviour, employment quality, and commitment to diversity. A score of between 0 and 100 is attributed to each company.
Corporate Governance performance: This is a sub-score of the overall CSR performance
score. It measures a firm’s performance in the areas of board structure, compensation policy, board functions, shareholder rights, and vision and strategy. A score of between 0 and 100 is attributed to each company.
Economic performance: This is a sub-score of the overall CSR performance score. It
34
company’s overall financial health. A score of between 0 and 100 attributed to each company.
Dependent Variables
This paper uses two different measures of risk which are calculated for the relevant period using data from the Thomson Reuters Worldscope database.
Total Risk: The first risk measure that we compute is the firm’s total risk, which is standard practice in the literature. The total risk variable is a measure of stock return volatility, which is a function of both upside and downside variations (i.e., downside losses and upside gains). This variable is measured as the annualized standard deviation from the weekly stock returns over the length of the previous year.
Annualized Standard Deviation =
= the ith observation
= the mean return
= the number of observations
= the number of periods in a year (e.g., 12 for monthly data)
Downside Deviation: The second risk measure that we develop is the firm’s downside risk or annualized downside deviation, which focuses specifically on those stock returns that fall below a minimum threshold or that do not meet a minimum acceptable rate of return (MAR). This measure is different from normal standard deviation as it only focuses on downside risk. Downside risk is of greater concern to management and investors than positive divergence, as neither usually views an appreciating stock price as a problem. The minimum acceptable return (MAR) is set at zero in our model as our interest is solely on capturing negative variance; therefore capital preservation is sufficient rather than setting an arbitrary rate of return. This variable is measured as the annualized deviation from the weekly stock returns, which were negative, over the length of the previous year.
Annualized Downside Deviation =
= returns for the ith observation where performance is below the MAR = the minimum acceptable rate of return
= the number of observations
35 Control Variables
In our multivariate analysis we include control variables that previous research finds have an effect on corporate social responsibility performance and on firm risk.
Board Size: The size of the board is included as a control variable as the number of women on the board is relative to the total number of board members, and as larger boards may have more resources, which may have an influence on CSR performance.
Firm Size: Firm size is measured as the natural logarithm of the market value of common equity, or market capitalization, at the end of the fiscal year end prior to the measurement date of the CSR performance and risk measures. Prior research indicates the use of the natural logarithm of market capitalization to be the best course of action (Fama & French, 1992; Brennan et al., 1998). This is done because there is quite a large difference in the market values of the firms in the sample and it better creates a linear relationship between the independent and dependent variables. Firm size is expected to be positively related to CSR performance, as larger firms may have more resources to dedicate to CSR activities compared to smaller firms, and to be negatively related to a firm’s risk.
Book-to-Market ratio: The book-to-market ratio is measured as the ratio of book to market value of value of common equity at the end of the fiscal year prior to the measurement of risk. It has been used as a proxy for risk and is expected to be positively related to a firm’s risk.
Return on Assets: This indicator is used as a proxy for firm performance. Return on assets (ROA) is calculated by dividing a company's annual earnings by its total assets, and gives an idea of how efficient management is at using a company’s assets to generate earnings. It is denoted with a percentage value.
36
which are Oil & Gas, Basic Materials, Industrials, Consumer Goods, Healthcare, Consumer Services, Telecommunications, Utilities, Financials and Technology.
3.3 Analysis
In order to test the relationships outlined in the conceptual model, this research follows the mediation procedure outlined by Baron and Kenny (1986). However, we also alter the mediation procedure following the recommendations for the criteria needed in establishing mediation, namely that there need not be a significant direct zero-order effect between the independent and dependent variables to establish mediation (Zhao et al., 2010). Baron and Kenny (1986) proposed a four step approach in which several regression analyses are conducted and the significance of the coefficients are examined at each step. First, the relationship between the independent variables (board gender diversity) and the dependent variable (firm risk), a regression analysis is carried out to confirm that the independent variable (board gender diversity) is a significant predictor of the dependent variable (firm risk). The second step is to confirm through regression that the independent variable (board gender diversity) is a significant predictor of the mediator variable (CSR performance), as if the mediator is not associated with the independent value then it is not possible for it to mediate anything. Thirdly, the relationship between the mediator variable (CSR performance) and the dependent variable (firm risk) is determined using regression analysis.
Figure 2 - Conceptual Model
37
Kenny (1986), if these relationships are non-significant then mediation is not possible. Finally, a multiple linear regression analysis is carried out on the relationship between the mediator variable (CSR performance) and the dependent variable (firm risk) and mediation is supported if the effect remains significant after controlling for the independent variable (board gender diversity). If the independent variable (board gender diversity) is no longer significant in the analysis then the findings support full mediation. However, if both the independent and the mediator variable are still significant then the findings support partial mediation, which indicates that the mediator variable accounts for some of the relationship between the independent variable and the dependent variable. Zhao et al. (2010) point out that there need not be a direct relationship between the independent and dependent variable in order for mediation to take place and that there can still be mediation in this case. Zhao et al. (2010) argue that to establish mediation, all that matters is that the indirect effect is
significant. This can occur when the independent variable is significantly related to the mediator variable and the mediator is in turn is related to the dependent variable; however there is likely an omitted mediator in the direct path relationship.
All the data in the thesis is prepared and analyzed using IBM SPSS statistical analysis
software. The analysis uses two different statistical tests to test the hypotheses outlined in the thesis. The first test is the Pearson product moment correlation which is used to determine the degree to which quantitative variables are linearly related in a sample. It indicates precisely the degree of relationship and measures the degree of correspondence between different variables. The second type of test used in the thesis is a multiple linear regression analysis which is a statistical technique that makes used of a number of explanatory variables in order to predict the outcome of a response variable. Multiple linear regressions allow one to model the relationship between the explanatory and response variables. This is particularly
38 Chapter 4 - Results
In this section we present the results of the tests carried out to investigate the hypotheses of this thesis. Table 1 shows the descriptive statistics about the gender diversity, board size and CSR performance of the 300 UK companies in the sample. The average number of women on boards is 1.01 women and the maximum number of women on any of the boards is 4 women. In terms of percentage, the maximum share of the board that women make up is 50 percent which is a perfectly gender diverse board, however the average percentage of women on boards equates to only 10.32 percent, which is very low. The mean board size of the companies is 9.25 board members, with the smallest board consisting of just 5 board
members and the largest consisting of 17 board members. As is clear from the table, the CSR performance of firms in the sample varies greatly. In the case of overall CSR performance the average score is 69.225, but the results do range from between 4.08 and 96.65, indicating vastly different focus and attention to CSR by companies in the sample. The individual elements of CSR follow a similar pattern to the overall CSR score as can be seen in the table. Figure 3 represents the minimum, maximum and average CSR scores of the companies in the sample. In Figure 4 we can see the number of women on the boards of the companies in the sample. 35 percent of the companies in the sample have no women on the boards of the companies which is quite a high proportion, while 39 percent of the companies have one woman representative the board. After that the number of companies with more than one woman declines in comparison and there is only a relatively small group, only 8 percent, of companies with three or more women on the board.
Table 1 - Descriptive Statistics
N Minimum Maximum Mean
39
Figure 3 – Minimum, Maximum and Average CSR ratings
Figure 4 – Number of women on FTSE 350 boards
Table 2 – Pearson Correlation
Board Size Firm Size ROA
Market to Book Board Size 1 Firm Size 0.545** 1 ROA -0.071 0.194** 1 Market to Book -0.057 -0.297** -0.288** 1
**. Correlation is significant at the 0.01 level (2 tailed). n=300 *. Correlation is significant at the 0.05 level (2 tailed).
40
Next the correlations between the control variables are investigated. The selected variables are board size, firm size (as measured by market capitalization), return on assets and the company’s market to book ratio for the year ending 2011. These correlations are performed to ensure the prevention of mulitcollinearity at later stages. The results of the Pearson
correlation show that board size and firm size are quite correlated (.545) which is expected as the larger a company is the more likely it will have more board members. There is also some correlation between firm size and return on assets (.194) and the market to book ratio (-.297), which indicates that larger firms are performing better and are at less risk than smaller firms. Finally, the results show a negative correlation between return on assets and market to book ratio (-.288) which demonstrates that better performing firms are associated with having less risk. The highest correlation is between board size and firm size (.545); however, as it is not close to 1 we are not concerned about multicollinearity in this case. The results for the control variables are as expected, and so we continue with their use for controlling in the later steps. Other control variables that are used are industry dummies to control for differences between different industries. For the sake of clarity and brevity we do not show the results for the individual industry dummies throughout the results, but instead we indicate when we have controlled for industry.
Table 3 – Pearson Correlation
Women Absolute Women Percent. Board
Size Firm Size ROA
Overall
CSR Environ. Social Govern. Economic Total Risk Downside Risk Women Absolute 1 Women Percentage 0.940** 1 Board Size 0.485** 0.241** 1 Firm Size 0.394** 0.250** 0.545** 1 ROA 0.033 0.071 -0.071 0.194** 1 Overall CSR 0.316** 0.247** 0.307** 0.505** -0.008 1 Environmental 0.309** 0.228** 0.346** 0.428** -0.023 0.848** 1 Social 0.322** 0.271** 0.246** 0.491** 0.031 0.869** 0.690** 1 Governance 0.175** 0.150** 0.187** 0.270** 0.012 0.652** 0.462** 0.511** 1 Economic 0.255** 0.183** 0.268** 0.475** -0.050 0.814** 0.540** 0.611** 0.422** 1 Total Risk -0.153** -0.115* -0.179** -0.400** -0.174** -0.261** -0.269** -0.167** -0.119* -0.231** 1 Downside Risk -0.195** -0.174** -0.147* -0.265** -0.099 -0.245** -0.275** -0.135* -0.127* -0.205** 0.887** 1
**. Correlation is significant at the 0.01 level (2 tailed). n=300
41
42
analysis and carry out linear regression analyses in order to determine the true relationship between board gender diversity and the CSR performance variables.
The two risk measures, total risk and downside risk, are strongly correlated (.887) at the 0.01 level, which is to be expected as they are both derived from the same data and downside risk is a measure of the negative fluctuations while total risk is the volatility of stock returns in both positive and negative directions. The results in the table also indicate a negative and significant correlation between board gender diversity and both risk measures. In the case of the number of women on the board and total risk we see a negative significant correlation (-.153). This is also found to be the case with the number of women on the board and
43 Table 4 – Multiple Linear Regression
Dependent variable Total Risk Downside Risk
Variables Model 1 Model 2 Model 3 Model 4
Beta Sig. Beta Sig. Beta Sig. Beta Sig.
Women Absolute -0.046 0.440 0.051 0.407 -0.133 0.042 -0.003 0.962 Board Size 0.024 0.721 0.049 0.462 0.025 0.731 0.055 0.444 Firm Size -0.306 0.000 -0.395 0.000 -0.191 0.009 -0.305 0.000 ROA -0.030 0.579 -0.069 0.196 -0.027 0.656 -0.074 0.194 Market to Book 0.281 0.000 0.256 0.000 0.104 0.089 0.071 0.234
Industry dummies No Yes No Yes
Adjusted R squared 0.224 0.000 0.276 0.000 0.077 0.000 0.174 0.000 n=300
Following the method prescribed by Baron and Kenny (1986) for the investigation of the mediation effect, we first look at the relationship between the independent and dependent variable. However, the Pearson correlations do seem to indicate a lack of relationship
between the variables. In the first model, board gender diversity is not significantly related to total risk, similarly neither are the size of the board or the prior firm performance as
measured by return on assets. The only significant predictors are the firm size (-.306) which is negatively and significant related at the 0.01 level, and the market to book ratio of the company (.281) which has a positive and significant relationship with total risk. In the second model, we control for industry and board gender diversity is still not significant. Again only firm size and market to book ratio are significantly related to total risk at (-.395) and (.256) respectively. The adjusted R squared of the model indicates that the variables used in the regression analysis account for 27.6% of the variation in the total risk of a company in the sample.
