4 Introduction
This paper investigates the effect of the share of female directors on financial firm performance.
An issue that is particularly important, since the equal distribution of gender among boards flourishes in public policy debates. As women face difficulties when trying to enter the boardroom, many countries (e.g. U.K., Spain, France, Belgium, the Netherlands, Sweden, Germany, Norway) are proposing legislation to increase the share of female directors. However, these policies are in most cases not mandatory. Companies are able to deviate from gender quota under the so-called ‘comply or explain’ principle. In contrast, since 2006, Norway maintains a quota system of at least 40% of each gender, with forced liquidation as sanction of no compliance. The introduction of this quota rule can be seen as an experiment whereby board composition has changed due to an exogenous factor and regardless of the needs of a company.
Therefore, this study analyses listed Norwegian companies.
Gender diversity has been studied extensively in the past decades. Women are found to have other qualities than men (e.g. Hengelsen, 1990; 2011; Hillman et al, 2002; Nielsen and Huse, 2010; Adams and Funk, 2009; Silverman, 2003; Andreoni and Vesterlund, 2001) and behaviour of board members seems to interact with board composition (e.g. Seierstad and Opahl, 2011; Adams and Ferreira, 2009; Adams and Funk, 2009; Huse, 2008; Erkut et al, 2008).
Moreover, board composition seems to be endogenously influenced (e.g. Gregoric et al, 2009;
Brammer et al, 2007; Carter et al, 2003; Carter et al, 2010; Adams and Ferreira, 2009), board size and firm size are positively related to the number of female directors. In addition, concerning the relation of the share of female directors and financial firm performance, the direction of causality is at times questioned (Ryan and Haslam, 2005; Adams and Ferreira, 2009; Bøhren and Størm, 2006; Farrel and Hersch, 2005; Carter et al, 2010). This leads to contradicting empirical results. Various researches find at first side a significant impact of the share of female directors on financial performance however, this gender effect appears to evade when other firm- and board characteristics are taken into account (Farrel and Hersch, 2005;
Gregoric et al, 2009; Adams and Ferreira, 2009; Carter et al, 2010). Since the Norwegian quota system is introduced, regardless of its influence on company performance, the implementation of the mandatory gender quota offers the opportunity to circumvent endogeneity problems moreover, it determines the direction of causality.
Considering the forced increase of female board members among Norwegian firms, this paper basically deals with three questions. First, does the share of female directors affect financial firm performance? Second, does the effect of the share of female directors on financial performance depend on the industry in which a company operates? Third, does the share of
5 female directors before the quota became mandate, influence financial firm performance after the introduction?
This study contributes to the growing literature on the Norwegian gender quota, as it includes, besides the percentage of female directors, the yearly change of the share of female directors as independent variable. This change reflects the effect of the high number of replacements, which is at least partly due to the quota system, and tokenism. Second, this study assumes a more complex link between gender diversity and financial performance and therefore focuses on differences among industries. In addition, no other research includes the percentage of female directors by the time the quota became mandate, as independent variable.
However, this is a substantial explanatory variable since this determines the extent to which a firm had to restructure its board during the two years of transition.
Annual stock returns and Market-to-Book values are used as forward-looking performance indicators. Subsequently, the accounting measure Return on Assets, which is partly backward looking, is used as a performance indicator for the short term. Furthermore, the estimations are corrected for board size, general fluctuations on the European market, firm size, and year effects. The original sample contains all Norwegian public limited liability companies (PLC’s), for the period 1999 to 2009. The unbalanced sample consists of 1080 unique firms and 5669 year-firm observations. Information about board size and the number of female directors is compiled by The Norwegian Business Register, Oslo. Financial data are gathered from Thompson Datastream. Since these data are only available for listed firms, the sample has shrunk to 223 unique firms and 1660 year-firm observations. All data are measured at the 31st of December of each calendar year.
Both, cross section and panel data estimation models are used in order to answer the first research question. Whereby, the cross sectional effect is analyzed as well as the effect of the share of female directors on financial firm performance over time. The results of the estimated regressions expose that female directors are more highly represented among less valuable firms. However, it cannot be determined that a larger share of female directors has a negative effect on the financial performance of a company over time. In addition, the yearly change of the share of female directors is used as independent variable. Outcomes determine a strong and robust negative effect on annual stock returns. As this effect reflects the disproportionate increase of the share of female directors under pressure of the quota system, it implies that Norwegian companies have suffered from the implementation of the gender quota. To answer the second question, the share of female directors is intermingled with industry dummies.
Female directors are found to be higher represented among non-manufacturing and more serve-oriented industries. Furthermore, the size of the effect of the share of female directors on
6 financial performance appears to differ on industry level, this implies that firms react different to the quota depending on their business specific environment. Lastly, this study finds that a larger share of female directors in 2005, leads to a smaller increase of female directors in the subsequent years. Moreover, companies without any female directors in 2005 experience a relative loss in the first year after the transition period, 2008. This relative loss may be due to the sudden replacement of 40% of the board under pressure of the government. Moreover, some companies may have had a hard job finding the right qualified women.
The coercive approach of the regulations on the equal gender distribution has led to a replacement of 24.2% of all board members of the Norwegian PLC’s within two years. The development of the board composition of all Norwegian PLC’s over the entire observation period is depicted in Figure 1. The main goal of the gender law was to employ all potential human resources by increasing the amount of women in boardrooms. Consequently, the law was proposed without the consent of business leaders and consideration of probable consequences on company performance. In 2003, the first quota law had passed the parliament:
by the 1st of July 2005, both sexes must be represented by at least 40%. However, this quota system was voluntary hence, in July 2005 not even 20% of all firms were in compliance.
Subsequently, in December 2005, the quota rule unexpectedly became mandate.
Figure 1: The figure presents the average amount of women, men and total members of the board of directors of all, listed and non-listed, Norwegian PLC's, for the period 1999 to 2009. The average number of female directors has risen from 0.45 to 2.46 during the observation period. The conversion of the graphs of men and women indicates that male board members are replaced by female board members. Data are from the Norwegian Business Register, Oslo. The sample contains 1080 unique firms and 5669 year- firm observations. Measures are taken at the 31st of December of each calendar year.
0 1 2 3 4 5 6 7
1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009
Average Number of Board Members
Year
Figure1: Members on the Board of Directors of All, Listed and Non- listed, Norwegian PLC's, from 1999 to 2009.
Women Men Total
7 This paper is organized as follows. The existing literature is reviewed in the first section.
The second section continues with the background setting. The third section, presents the research objective of this study. In the fourth section, the data and variables used in this research are being reported. Section five describes the methodology used. Subsequently, the results are presented and analysed in section six. Finally, section seven contains the overall conclusions, as well as limitations and suggestions for further research.
8 I. Literature Review
A. Diversity en Performance
Since education has become more accessible for a broader range of people, the effect of diversity on group performance appears to become visible. On the one hand, diversity leads to better strategic decision-making (Bantel, 1993, cited by Erhardt et al, 2003). Moreover, a greater knowledge base, creativity and innovation will bring a competitive advantage (Watson et al, 1993, cited by Erhardt et al, 2003). On the other hand, heterogonous teams are slower in their actions and responses, since they are more likely to disagree (Hambrick et al, 1996, cited by Erhardt et al, 2003). Diversity can be divided in observable and non-observable characteristics. Observable diversity includes characteristics as gender, race, ethnicity and age, non-observable diversity factors as personality, values, knowledge and affection (Erhardt et al, 2003). Performance can also be measured by various indicators. Creativity, board efficiency and strategic decision-making are less concrete measurable compared to for example financial performance. This paper focuses on gender diversity within the board of directors and financial performance of listed Norwegian firms in the setting of a mandatory quota system.
A.1 Characteristics and Behaviour
Previous investigations find evidence of personal characteristics influencing leadership style and strategic decisions (Betrand and Shoar, 2003; Malmandier et al, 2010, also highlighted in Matsa and Miller, 2011), and gender differences appearing in management styles, attitudes, professional experiences, interests and underlying values (Helgesen, 1990, cited by Seierstad and Opsahl, 2011; Hillman et al, 2002; Nielsen and Huse, 2010, Huse, 2008; Adams and Funk, 2011). Therefore, a greater share of female directors in male-dominated companies will contribute to new thinking and problem solving, which could result in higher productivity and a better working environment (Helgesen, 1990, cited by Seierstad and Opsahl, 2011). Adams and Funk (2009) find, when measuring values according to Schwartz (1992) that female directors are more benevolent and universally concerned, but less power oriented then men, they tend to have a more power-sharing style (Bradshaw et al, 1996). Moreover, they are less tradition- and security oriented, and more risk-loving than their male counterparts. Furthermore, women are found to be more long-term oriented (Silverman, 2003;), more stakeholder oriented, more likely to align with shareholders (Adams and Ferreira, 2009; Adams and Funk, 2011) and more altruistic (Andreoni and Vesterlund, 2001), also highlighted by Matsa and Miller (2011).
Expectations and stereotypes based on gender are present were minority groups exists. When both sexes are equally balanced, groups are likely to be formed due to other characteristics (Seierstad and Opsahl, 2011). In addition, Huse (2008) explores that group dynamics and the
9 working style of boards affected by women influence firm performance, rather than the share of women itself.
Huse and Solberg (2004) find the power game at the top as the most important element in understanding board behaviour. The contribution of female directors highly depends on their capability to make alliances with the most influential actors, to spend time on preparations being present on the most important decision making areas and to take leadership roles. When entering the board as a token, a woman is often not part of the ruling alliances. The actual contribution of a woman does not only depend on their characteristics but, also on the possibilities the board offers to women to make a difference (Nielsen and Huse, 2010). Adams and Ferreira (2009) state that women are tough monitors because they are more independent.
Moreover, they are better prepared and have higher attendance records than their male counterparts. In addition, male directors have fewer attendance problems when being part of a more gender-diverse board. This has a positive influence on the effectiveness of the board and the quality of decision-making (McInerney-Lacombe et al, 2008). On the other hand, women on top level of a predominantly male environment may alter their behaviour in a way gender differences will vanish (Adams and Funk, 2011), possibly to avoid being stereotyped (Huse, 2008). In this case a board will lose the benefits of diversity. Despite the female representation, these boards only contain masculine behaviour (Sherigan and Milgate, 2005, cited in Terjesen et al, 2009). The presence of an additional woman, changes the way men react to women, this helps women to be more vocal in meetings. However, the existence of three women within a boardroom is sufficient to tilt the group dynamic and to normalize women’s presence as leaders (Erkut et al, 2008). Consequently, the dependence of behaviour to board composition and the environment of the business makes the influence of board structure on performance inconsistent.
A.2 Female Directors and Firm Performance
Existing research concerning the influence of the share of female directors on financial firm performance provides mixed evidence. Table I provides an overview. Depending on the performance measures, control variables and estimation conditions, positive (Carter et al, 2003;
Smith et al, 2006), negative (Bøhren and Strøm, 2005; Adams and Ferreira, 2009) and in some cases even no significant relation (Farrel and Hersch, 2005; Gregoric et al, 2009 Ch 3; Carter et al, 2010) is found.
The weakness of the evidence appears to be caused by endogeneity problems (Carter, et al, 2003, Adams and Ferreira, 2009; Gregoric et al, 2009; Brammer et al, 2007; Carter et al, 2010). It is hard to determine whether the share of women itself affects firm performance or
10
Table I: Overview of studies investigating the relationship between women on the board en firm performance.
Authors What is researched Sample & Time Period Performance Measure
Direction of causality/
Endogeneity Main findings/ Comments on quota
Erkut et al (2008) The effect of the presence of one, two, three or more women.
56 women with top management functions
The presence of more than one woman, changes the way men react to women.
Three women in a boardroom normalize women's presence as leaders.
Carter et al (2003) Diversity, women and minorities and firm value.
Fortune 1000 Firms;
1998-2002 Tobin's Q
The proportion of women and minorities increases with firm size and board size, but decreases as the number of insiders increases.
Positive relationship between the percentage of women or minorities directors on the board and firm value.
Smith et al (2006) Gender diversity and firm performance.
2500 Danish Firms;
1993-2001
Gross Value Added &Ordinary Profit /Net Turnover;
Ordinary result &
Net result after Tax/Net Assets
Board diversity affects firm performance.
The share of women has a positive effect on firm performance, even after controlling for firm characteristics, mainly driven by the qualifications of women.
However, insignificant after controlling for unobserved firm specific factors. Quota:
" It is important to attract, recruit and increase the number of women who are qualified to be selected into boards of directors. The implementation of gender quotas might otherwise have a negative effect on firm performance."
Bøhren & Strøm (2005)
The proportion of women directors and firm performance.
Listed Non-Financial Norwegian firms; 1989- 2002
Tobin's Q
Causation runs in both directions, board mechanism determine each other.
A higher level of gender diversity is negatively and significant related to firm performance.
Adems & Ferreira (2009)
The impact of female directors on board input and firm outcomes;
Endogeneity Issues.
S&P 500; S&P MidCaps;
S&P SmallCaps; 1996- 2003
Tobin's Q; ROA
Most evidence is not robust when the endogeneity is addressed.
Women have higher attendance records, men have higher attendance records in a more diverse board. Negative effect of gender diversity on firm performance.
Quota: "Our results suggest that mandating gender quotas for directors can reduce firm value for well-governed firms where additional monitoring is
counterproductive".
Farrell & Hersch (2005)
The extent to which gender impacts the selection of a director to serve on the board.
Fortune 500 and
Service 500; 1990-1999 Stock price
Corporations respond to either internal or external calls for diversity.
No significant abnormal returns are found on the announcement of women added to the board. The demand for women directors is not based on performance.
Gregoric et al (2009)
Gender diversity and company economic performance.
Around 500 publicly listed Nordic companies. 2001-2007
Tobin's Q; ROA;
Growth
Board diversity in endogenous influenced.
No positive effects, lack of significance. However, significant positive correlation.
Quota: "Firms follow a goal of 'minimum' diversity, different aspects of diversity tend to substitute each other. Excessive outside pressure may crowd-out any internal recognition of the benefits of female directors."
Carter et al (2010)
The number of women directors/
ethnic minority directors and financial performance.
Fortune 500 Firms;
1998-2002 Tobin's Q; ROA
Board diversity and financial firm performance appear to be endogenous. Do not find evidence of causation.
Positive significant relation when performance is measured by ROA. No relation when performance is measured by Tobin's Q. Quota: "Decisions concerning the appointment of women to corporate boards should be based on criteria other than future financial performance."
Ryan & Haslam (2005)
Whether women are appointed under more precarious and risky conditions.
100 companies London
Stock Exchange Stock price
Women are appointed under more precarious business conditions.
More women are appointed when a firm experiences bad performance.
11 that board composition is endogenously influenced by firm characteristics. Board size (e.g.
Gregoric et al, 2009; Adams and Ferreira, 2009; Brammer et al, 2007) and firm size (Carter et al, 2003; Hillman et al, 2007) are positively related to the number of female directors. Adams and Ferreira (2003) cited by Farrell and Hersch (2005) explicate the later by higher visibility, larger companies would have a larger demand for diversity since they are more in the public eye.
Moreover, women may be more attracted to boards of larger companies and thereby more easily recruited. Furthermore, women are more likely to be found in boards that are linked to other boards that have female directors (Hillman et al, 2007).
Additionally, some researchers shed light on the direction of the causal relationship of board diversity and firm performance. The historical performance, future expectations and the climate on the financial markets may influence recruitment policies for women. Ryan and Haslam (2005) find that during an overall stock market decline, companies that appoint female directors are more likely to have experienced a consistently bad performance in the preceding months, compared to firms that appoint men. Due to a continuing pattern of poor performance, the positions of women were associated with a higher risk of failure. Farrell and Hersch (2005) also point to reversed causality however, they state that the demand for female representation allows women to self-select better performing firms. Adams and Ferreira (2009) show a reversion from a negative to a positive effect when including fixed effects. More specific, in cross section, firms with a lower value have more women on the board however, firm value raises over time when more women are hired. This implicates that the larger share of female directors improves the financial performance of a firm. Conversely, this is not robust when endogeneity is addressed. Moreover, Smith et al (2006) determine that female directors are affecting firm value, not the opposite. In contrast, Carter et al, (2010) do not find evidence for this direction of causation. Finally, Gregoric et al (2009, p. 55) state that reversed causality, in the case that better performing companies select more women, would still reflect the believe that a more diverse board will contribute to company performance, since the shareholders who chose new directors, also determine the market value. The introduction of the mandatory quota system has caused a forced increase of women on the board table whereby the direction of causality is determined. This gives the opportunity of investigate the relation between gender diversity and financial performance in an extraordinary setting.
B.1 Possible Effects of the Quota regarding Theory
There are several scenarios concerning the effect of the mandatory gender quota on firm performance. At first, financial performance might be affected negatively due to a forced soar of female board members. Baysinger and Butler (1985) find evidence suggesting that a public
12 policy prescription in favour of any particular board composition reform would be inappropriate in an economy where firms are different and continuously changing. In addition, if firms choose their board composition in line with their characteristics (Coles et al, 2008), which determine the costs and benefits of the board’s monitoring and advisory roles (Linck et al, 2008), a regulatory framework could be in contrast with the maximization of firm value (Adams and Ferreira, 2009; Nygaard, 2011). Moreover, firms may need to select second best board members, due to a shortage of qualified women (Farrel and Hersch, 2005).
Concerning the second scenario, firms with suboptimal governance are expected to benefit from a more diverse board. According to the Resource dependency theory1 (Pfeffer and Salancik, 1978, highlighted by Carter et al, 2010; Gregoric et al, 2009 p. 58), “a more diverse board will provide more valuable resources, which should produce better firm performance”.
Moreover, these firms might structure their board in order to maximize the private benefits of the management, known as ‘captured boards’ (Bebchuk and Fried, 2005). By raising the number of female directors, the share of independent directors will soar (Nygaard, 2011; Carter et al, 2003), and firm value is expected to rise because agency costs are reduced (Ahern and Dittmar, 2011). Concluding, firms in which more diverse perspectives and non-traditional approaches or monitoring align with the interest of shareholders, might benefit from the quota.
Thirdly, if there is no distinction between qualified female directors and other competent directors, board diversity does not influence a firm’s governance structure and financial performance. The desirability of equal distribution would only be a public policy issue (Carter et al, 2010). In addition, if boards are merely ‘window-dressing’, regulations on board composition will have no effect on firm value either (Westphal, 1998; Helland and Sykuta, 2004 cited in Ahern and Dittmar, 2011).
When it was determined that a board containing 40% women would increase the financial performance of a firm, a quota system would be redundant to raise the share of female directors. Moreover, the influence of female directors does not have to be the same among different firms. According to the Contingency theory (Fiedler, 1967, highlighted in Carter et al, 2010), firms may react in a different way to the gender quota, depending on business specific features. The level of information asymmetry (Nygaard, 2011) and the trade-off between monitoring and advisory (Helland and Sykuta, 2004; Linck et al, 2008; Adams and Ferreira, 2009) determine the influence of the forced recruitment of women. Furthermore, women are
1 The Resource Dependency Theory advocates the role of the board members in providing an important channel of information, facilitating access and strengthening the commitment to important firm constituencies and resources. (Pfeffer and Salancik, 1978)
13 more likely to make a positive contribution to firms that stress the importance of strategic and Corporate Social Responsibility control tasks rather than operational, financial and behavioural control tasks (Huse, 2008). Moreover, Francoeur et al (2008) state that firms operating in complex environments benefit from a high proportion of female officers.
In addition, empirical evidence suggests that type of industry is significant in explaining the representation of women on corporate boards (Betrand and Hallock, 2001). Firms operating in industries with a greater number of female employees are more likely to have women on their boards (Hillman et al, 2007). Furthermore, female directors are more likely to occupy top management positions in non-manufacturing, and service oriented industries (Harrigan, 1981 cited by Farrell and Hersch, 2005; Goodman et al, 2003; Brammer et al, 2007). Due to their interpersonal skills they have a competitive advantage regarding intangible products (Goodman et al, 2003) moreover, these firms have a larger pool of women available for board positions (Gregoric et al, 2009 p.62). In contrast, manufacturing companies may not maintain enough women in the ‘pipeline’ (Burke, 1997, cited in Brammer et al, 2007) hence, they cannot fulfil the board positions with qualified women who know the industry well enough and posses a valuable network. According to research, firms that do not include women to their board for own consideration, suffer from the mandate gender quota (Dittmar and Ahren, 2011; Nygaard, 2011; Matsa and Miller, 2011).
B.3 The Effect of the Quota on Firm Performance
Since the quota was introduced - independent of the effect or possible results on firm performance - it can be determined that the increase of women exogenously is increased to almost 40%. The subsequent investigations, present results on the exogenous effect of the Norwegian gender quota on firm performance.
Firstly, Ahern and Dittmar (2011) conducted an event study on the stock price reaction to the initial announcement of the (voluntary) quota in February 2002. They found a negative stock return of -3.54% for firms without any female directors compared to -0.02% for firms with at least one female director. Moreover, firms with at least one woman on their board in 2002, had a higher industry-adjusted Tobin’s Q compared to firms without any female directors.
Furthermore, they find evidence for a negative relation between Tobin’s Q and the forced change of female representation on the board. These results are compared to firms in countries that do not have to comply with a mandatory gender quota and, significant differences are observed. Their results suggest that the constraints imposed by the law had a large negative effect on firm value, commensurate with the massive organization of corporate boards imposed
14 by the gender quota. Their findings are in line with the statement that companies construct their boards to the needs of their business.
Secondly, Nygaard (2011) performs an event study around the 9th of December 2005, the day at which the mandatory quota, with forced liquidation as a sanction of no compliance, was announced. Different results regarding the level of information asymmetry and the share of female board members before the announcement are found. A positive significant relation is found for firms with low information asymmetry and few female board members. This result indicates that the effect of the female representation law depends on the needs of a company.
Since women are seen as independent board members, an increase of women in the boardroom might be beneficial for companies that profit from a higher level of monitoring.
However, according to Edmans (2011), cited by Matsa and Miller (2011), the results of both event studies should be questioned. He states that stock markets undervalue intangible benefits of governance strategies. In addition, the market perception of the capacity of female directors leading a company may be distorted since female directors were rare before the announcement of the gender quota, this makes the expectations unfounded (Wolfers, 2006, highlighted by Matsa and Miller, 2011).
Thirdly, Matsa and Miller (2011) execute a triple- difference method and compare the financial performance of Norwegian firms, which have to comply with the quota, to Norwegian firms that are not affected by the gender quota and firms elsewhere in Scandinavia. They find a relative decline (-4%) in corporate profitability among firms that needed to comply with the quota. The inclusion of the control variables board size, the average number of other board seats occupied, the average board member age and the share of new directors has minimum effect on the estimates, this is line with the assumption of the exogenous increase. Additional to other research, they explored the causes of this decline. They explain the decrease by increased labour costs from fewer layoffs and a relative increase of employment. They add an alternative explanation, stating that companies were somewhat dysfunctional during the transition period.
Their results are consistent with changes in board composition affecting corporate governance and strategy, and with prior research suggesting that female managers may be more stakeholder (Adams and Funk, 2011) and long term oriented (Silverman, 2003) than their male counterparts. Since the effects of the quota have showed up during credit crisis (2008), the effects of the global recession around 2001 are studied. The lack of an effect during this earlier crisis supports their interpretation of changes in profit and employment reflecting the impact of the quota.
15 When taking the share of female directors before the announcements of the introduction of the gender quota into account, the negative effects were stronger among firms that were required to add the most women to their boards (Ahern and Dittmar, 2011; Nygaard, 2011;
Matsa and Miller, 2011). The short period (two years) in which companies had to balance their gender composition might have influenced their firm performance in a negative way.
Table II: Overview of studies investigating the relationship between women on the board en firm performance in the context of the Norwegian Gender Quota.
Authors What is researched Sample &Time Period
Performance
Measure Main findings Cause
Seierstad &
Opsahl (2011)
The gender bias, the emerge and sex of prominent directors, and directors' social capital.
384 Norwegian Public Listed Companies; May 2002- August 2009
The share of women on boards has not increased above the mandate 40%. The senior boards positions remain restricted to men. The quota created a small elite of women directors.
Companies may be not persuaded by the utility arguments of the gender quota. Temporal shortage of skilled and qualified women.
Ahren &
Dittmar (2011)
The relationship of firm value and board characteristics.
165 publicly listed Norwegian firms;
2001-2007
Stock price;
Industry adjusted Tobin's Q.
Negative impact of mandated board changes on firm value; Firms without women in 2002 higher negative mean return.
Less capable boards, the quota led to younger and less experienced boards, increases in leverage and acquisitions and deterioration in operating performance.
Nygaard (2011)
The impact of a forced increase in gender diversity on firm performance.
All board member in Norwegian PLC's, listed and non-listed; 1999- 2009
Stock price; ROA
Firms with a low level of
information asymmetry experience a positive effect of the forced increase of gender diversity. Strong negative correlation between the conversion decision of non-listed firms and the share of female directors during 2006 and 2007.
Increase of independent directors is not in line with needs of all companies.
Masta & Miller (2011)
Impact of quota on management style and firm performance;
affected firms, compared to unaffected firms.
Panel of Scandinavian companies; 1999- 2009
Operating profits/
Total Assets
Firms affected by the quota undertook fewer workforce reductions than comparison firms, increasing relative labor costs and employment levels and reducing short-term profits.
Differences is corporate strategy, female managers are more stakeholder and long term oriented.
16 II. Background Setting
A. Gender Equality
The Scandinavian countries are known for their successful policies and efforts to include women in their labour market. Norway was ranked third in the United Nations Gender-related Development Index of 2008 (Hausman et al, 2008). The Norwegian Gender Equality Act of 1978 stimulated the creation of a women-friendly social-democratic welfare society. Additionally, in 1981 the first law on gender composition of publicly appointed boards, councils and committees was amended (The Norwegian Government, 2008). Norway maintains a strong corporate governance structure (La Porta et al, 1998) moreover, the stress on the importance of equality and the coercive approaches (Burke and Vinnicombe, 2008) make it not surprising that Norway was the first country in the world that came with regulations on the gender composition on boards of companies.
A.1 The Intention
The main purpose of regulation on board composition was to recruit more women into boardrooms. The government claimed cultural aspects and traditional ideologies for keeping women out of these positions. Gender equality, a fairer society and the distribution of power were important arguments. Furthermore, breaching ‘the Old Boys Network’ in order to employ all the potential human resources available was imperative. Advantage should be taken of the talent and competence of women since they attain education to a higher extent than men (Seierstad and Opsahl, 2011; Ministry of Children, Family and Equality, 2003, cited in Nygaard, 2011). When business cases result in evidence of board diversity having a positive influence on financial firm performance, the Norwegian government affirms that the quota would increase the bottom line (Seierstad and Opsahl, 2011). However, the quota was proposed without the consent of arguments concerning competitiveness and financial performance of enterprises and furthermore, the law was implemented regardless to the opinion of business leaders, who state that there was a lack of qualified female directors (Criscione, 2002, cited in Nygaard, 2011).
A.2 The Introduction
The first public hearing on a possible gender quota for private company boards, in October 1999, was derived from the gender equality act from 1978. Several extensions were suggested, however the following new law was proposed: ‘Boards of all listed firms must be represented by both genders. For boards with 4 or more members, each gender should be represented by at least 25%’. In 2001 a new proposal suggested to incorporate the quota into law and with a higher target: 40%. March 2002, the government announced that work towards
17 a law-proposal was continued. Furthermore, private companies were requested to increase the representation of women voluntarily. In 2003, the government together with the business community initiated different programs to surge the share of female directors. During the same year, the law proposal had past the parliament. The law contained a representation of 40% for both sexes and covered government owned companies and all publicly limited liability companies (PLC’s). These companies had to be voluntarily in compliance by the 1st of July 2005 however, there were no consequences by then. When in July 2005 only 15,5% of the firms were in compliance, the public debate centred on whether the quota should be binding by law and what sanctions should be put in place. At the 9th of December, the government abruptly decided to mandate the quota with forced liquidation as sanction of non-compliance. This announcement was experienced as a shock through the whole country, particularly since the Prime Minister one week earlier only had spoken about fines as possible sanctions (Nygaard, 2011). The new Section 6-11a2 of the Norwegian Public Limited Liability Company Act entered into force on the 1st of January 2006, giving all PLC’s two years to recruit sufficient female directors.
The law covers all Norwegian PLC’s registered at the Norwegian Business Register and with their principal in place of business in Norway. The quota rule does not apply to limited liability companies (LTD’s), since these companies mainly consist of small family enterprises where owners are personally represented on the board (Hoel, 2008). After a few last warnings, all PLC’s were in compliance by April 2008. In the period between 2003 and 2009, the share of female board members of PLC’s has increased by 24%, this was only 2% to 5% for LTD’s (Ahern and Dittmar, 2011). This difference indicates the exogenous increase of female directors within PLC’s.
A.3 Effects regarding the Intention of the Quota
The implied intention of the introduction of the gender quota was to create a more gender balanced power-distribution of the influential board positions and thereby to generate a larger pool of women acting as directors. Due to the law, the share of female board members has increased form 16% in 2005 to around 40% in 2008. Seierstad and Opsahl (2011) find however, that the share of female board members has not further increased above the minimum
2§ 6-11a: Demands regarding the representation of both sexes on the Board
(1) On the board of a public limited company both sexes shall be represented as follows:
1. If the board of directors has two or three members, both genders shall be represented.
2. If the board of directors has four or five members, each gender shall be represented by at least two members.
3. If the board of directors has six to eight members, each gender shall be represented by at least three members.
4. If the board of directors has nine members, each gender shall be represented by at least four members, if the board of directors has more than nine members, each gender shall be represented by at least 40 percent.
5. The provisions of paragraphs 1 to 4 shall apply to the election of deputies.
18 of 40%. Additionally, the most senior positions remain restricted to men, as the share of companies with a woman chair has remained stable during the transition period. Also Farrell and Hersch (2005) find evidence of firms operating in a manner consistent with tokenism, women appointed for affirmative action reasons may lead to a perception of women as less valuable board members (Huse and Solberg, 2006). On the other hand, Niederle et al (2008) find that gender-based affirmative actions stimulate highly skilled women to participate. In addition, both, the maximum number of directorships and the number of prominent directors has increased, mainly by women Seierstad and Opsahl (2011). This development of prominent directors also contradicts with the goal of equality. Therefore, these findings might indicate that companies are not persuaded by the utility arguments of the gender quota and that a forced increase of women representing board seats might have created a temporal shortage of skilled and qualified women available to take up such positions.
A.4. Pragmatic Reactions in Responds to the Gender Quota
The implementation of the mandatory gender quota represented a major and sudden intervention in the selection of directors. Due to the impact on the governance of companies, pragmatic solutions to evade the law are expected among PLC’s.
There is evidence to suggest that companies build constructions in such a way that subsidiaries can escape from the gender law (Reiersen and Sjåfjell, 2010). Moreover, it is expected that women will be added to boards even if no other directors are resigning as the introduction of the quota makes board diversity a goal of all PLC’s (Farrel and Hersch, 2005).
Also the apparition of the ‘Golden skirt’ - women who are serving on multiple boards - can be interpreted as a signal from companies searching for ad hoc solutions. From 2005 to 2008 the maximum amount of directorships held by a single female director increased from 4 to 8 (Seierstad and Opsahl, 2011; Ahern and Dittmar, 2011), 75 female elites occupy one-fourth of the 868 board positions held by women (Hoel, 2008). In addition, the share of foreign board members in Norway is larger compared to other Nordic countries. A peaked is observed in 2006 as new selected women in order to comply with the quota are more likely to be foreign (Gregoric et al, 2009).
Furthermore, Ahern and Dittmar (2011) and Nygaard (2011) find a significantly different distribution of PLC’s and LTD’s between 2001 and 2009. In 2009, the number of PLC’s had decreased 30% since 2001, the number of LTD’s, which are not affected by the quota, had increased more than 30% during the same period. Hoel (2008) states that throughout 2007, 79 companies decided to register from PLC into LTD. Ahren and Dittmar (2011) find a significant negative relationship between the probability of delisting after the introduction of the
19 mandatory quota and the share of female board members in 2002. By classifying the reasons of delisting into five categories and controlling for firm size, risky investments and industry effects, they find ‘Acquisition by a private or foreign company’ as the most common reason.
86.1% of these firms had not any women on their board in 2002. Furthermore, they find that firms are more likely to delist if they have a younger board with less CEO experience. Nygaard (2011) investigates whether converting firms systematically had less female directors during the period 1999 to 2008, after controlling for firm characteristics no correlation between the conversion decision and the share of women on the board of listed companies is found.
However a strong negative correlation is found for non-listed firms in the years 2006 and 2007.
A reason for this difference might be the extra effort it takes a listed company to convert to LTD.
A listed firm has to concentrate its ownership and delist in order to be able to change its status (Matsa & Miller, 2011). These observations provide indirect support for the argument that a regulatory framework on gender distribution is in contrast with the maximization of firm value (Baysinger and Butler, 1985; Adams and Ferreira, 2009; Nygaard, 2011).
20 III. Research Objective
Empirical findings in the existing literature regarding the influence of the share of female directors on financial firm performance still vary, one reason for these mixed results is the endogeneity problem. The introduction of the mandate gender quota in Norway, has created a natural experiment whereby all PLC’s had to increase their share of female directors to 40%.
Hence, the purpose of this study is to examine the effect of female directors on financial performance by circumventing the endogeneity problem as the direction of causality is assumed to be clear. The first research question is:
1. Has the share of female directors an effect on the financial firm performance of listed Norwegian companies? And if so, is this relation positive or negative?
In addition, the existing literature hardly provide research, Nygaard (2011) is an exception, of the exogenous gender regulation regarding business specific environments.
Brammer et al (2007) states that a close proximity to final customers plays an important role in shaping board diversity. Moreover, research concerning female directors among different industries, finds that the representation of women is higher in non-manufacturing and more service-oriented industries (Harrigan, 1981 cited in Farrell and Hersch, 2005; Goodman et al, 2003; De Anca, 2008). According to the Contingency theory and Resource dependency theory it is expected that the effect of the share of female directors on financial performance depend on business specific characteristics. Subsequently, the second research question is:
2. Does the size of the effect of the share of female directors on financial performance depend on the industry in which a company operates?
Lastly, this study analyzes the impact of the share of female directors in 2005. The pre- quota share of female directors determines the change a company had to make in order to comply with the mandatory gender quota. Moreover, the pre-quota share of female directors may reflect the needs of a company (Coles et al, 2003; Linck et al, 2008). Therefore, it can be expected that the share of female directors in 2005 is positively related to financial firm performance in the years after the quota. Continuing on the research of Ahern and Dittmar (2011) who find a positive relation between the percentage of women in 2002 and Tobin’s Q in the years 2007, 2008 and 2009, this paper addresses the following research question:
3. Does the share of female directors before the quota became mandate, affect the financial firm performance after the introduction?
21 IV. Data and Descriptives
This section gives an outline of the data used for this research, furthermore the sample will be described and a summary of the statistics will be given.
A. Construction of the Database
This paper investigates the relation between the share of female directors and financial firm performance in the setting of the mandatory gender quota implemented by the Norwegian government. Only Norwegian companies that have or had the status to comply with the quota are contained in this research, the so called public limited liability companies (PLC’s). The development over time is investigated cross-sectional, therefore the used sample contains panel data.
Data about the board composition of the Norwegian PLC’s is compiled by the Norwegian Business Register (NBR), Oslo. The dataset reports the number of female directors and the total number of members that were seated on the board of directors of all listed and non-listed Norwegian PLC’s. Measurements were taken at the 31st of December of each calendar year for the period 1999 to 2009. This period is chosen since the legislative process of the introduction of the quota has taken place in this period and also the first two years after the final date at which all PLC’s had to be in compliance with the new law, are included. This dataset is unbalanced, counts 1080 unique firm and 5669 year-firm observations. The fact that the dataset is unbalanced can party be explained by firms that has changed their status after the introduction of the quota (Ahern and Dittmar, 2011; Nygaard, 2011) though, it is not a priori clear whether this is a direct effect of the quota. Table A.II (Appendix) shows a substantial decrease of PLC’s during the observation period.
In order to gather financial information of the firms from Thomson Reuters DataStream, a financial numerical database, company names are matched to ISIN-codes3. Out of 258 companies that were matched to an ISIN-code, DataStream provided financial information for 223 of them. These are all companies that are listed or have been listed in the past. By eliminating the non-listed companies without ISIN-code, the sample is still unbalanced, contains 223 unique firms and 1660 year-firm observations. In line with the findings of Nygaard (2011) less listed PLC’s has changed their status after the introduction of the gender quota. (Table III).
3Company names were matched to ISIN-codes at www.ISIN-codes.com and at the Orbis database. The identification codes of the NBR are used to detect companies that had changed their names in the past. Names changes are checked with the
‘List Changes of the Oslo Børs’ and the Orbis database.For companies that have acquired another, observations of the largest firms, were used for the years before the acquisition. Size is based on board size. Data was deleted for five small companies that were acquired.
22 B. Variables
The most important data for this study are obviously the percentages of female directors on the board of Norwegian firms. As stated before the NBR provided the number of female directors and the total number of directors per firm per year. To make changes on the board composition more clear, the following extra variables are created: the number of male directors, the share (in percentage) of female directors, the yearly changes in the share of female directors, the distance to the quota (40% women) in percentage and the percentage of firms that are in compliance, i.e Men, Percentage of Women, Δ Percentage Women, Distance to Compliance and Percentage in Compliance. Table A.I explains how these variables are generated.
The variables gathered from the DataStream database, present information on the industry, market value, EBITDA, annual stock returns, number of employees, total assets, capital expenditures and sales. A description of these variables is given in Table A.I. Financial data are converted to U.S. Dollars since companies report in Euro’s, British Pounds, U.S. Dollars and Norwegian Kroner. Similar to the data on board composition, measures of the financial data are taken at the 31st of December of each calendar year, or the closest day available. Since the share of female directors is expected to be unequally distributed among industries (e.g. Harrigan, 1981; Goodman et al. (2003); Singh and Vinnicombe, 2003; Hillman et al, 2007), different sectors are separated. The classification of the industries is based on Level 2 of the ICB of DataStream, this divides the total market into ten industries. Since there was only one firm with the industry classification ‘Telecom’, and two firms with the classification ‘Utilities’, these company were merged respectively to the industry groups ‘Technology’ and ‘Basic materials’.
The subsequent industries are distinguished in eight categories: Oil & Gas; Basic Materials;
Industrials; Technology; Consumer Goods; Healthcare; Consumer Service and Financials.
Financial firms are included in the sample for completeness but the results hold if they are omitted. Descriptions of the industry classifications are given in Table A.I. Table A.III summarizes the number of unique firms and the year-firm observations each industry category encloses.
B.1 Independent variables
The independent variable in this study is the share of female directors, Percentage Women. This is in line with prior research on the influence of women on the board on the financial performance of a company (e.g. Carter et al, 2003; Smith et al, 2005; Nygaard, 2011;
Dittmar and Ahern, 2011). Additionally, the yearly changes in the share of female directors, Δ Percentage Women, will serve as independent variable, in a number of robustness checks. The
23 yearly changes in the share of female directors will stress the peaks of the increase of female directors, which are assumed to be partly the effect of the forced board restructurings. Hence, it is expected that women selected as a token are included in this increase. Therefore, the effect of yearly changes of female directors on financial performance will reflect the effect of the mandatory gender quota. Subsequently, industry dummies are intermingled with the percentage of female directors in order to consider the effect on industry level. Finally, in order to examine whether the share of female directors before the quota became mandate has an effect on financial firm performance after the introduction, the share of female directors in 2005 is taken as independent variable.
B.2 Dependent variables
As dependent variable, three different measures of financial performance are included.
(1) Annual Stock Return, (2) Market-to-Book value and (3) Return on Assets.
Values of Annual Stock Return are created by taking the natural logarithm of the yearly stock returns. Secondly, the variable Market-to-Book value is generated by dividing the market value by total assets. This variable has a very skewed distribution, for this reason the logarithm is used in the regression analyses. Both, Stock Market Returns and Market-to Book value are indicators for performance on the long term. The third performance indicator, Return on Assets (ROA) is computed by dividing EBITDA by total assets. EBITDA is net income with interest, taxes, depreciation and amortization of goodwill added back to it. This is a good indicator to compare profitability between companies and industries because it eliminates the effects of financing and accounting. Also this variable has a very skewed distribution. To use this variable in the regression analyses, the cubic roots of the absolute numbers are computed while the sign (negative or positive) is preserved hence, more observations are maintained. ROA is a performance indicator for the short term, partly backward looking, since this is an accounting measure.
B.3 Control variables
According to the existing literature (e.g. Carter, 2003; Hillman et al, 2007; Gregoric et al, 2009), variables representing information about the size of a company are employed as control variables in the regression analyses, i.e. number of employees, capital expenditures and sales.
Logarithms of these variables are used since the distribution of these variables is skewed.
Furthermore, Board Size, is included since prior research presents evidence that board size has a significant influence on firm performance (e.g. Carter, 2003; Brammer et al, 2007) and correlates with the number of female board members (e.g. Carter et al, 2003; Farrel and Hersch, 2005; Gegoric et al, 2009). Moreover, the logarithm of the yearly European market return is
24
0%
5%
10%
15%
20%
25%
30%
35%
40%
Average Percentage of Women
Year
Figure 2: Percentage of Female Directors of All Norwegian Listed and Non- listed PLC's, from 1999 to 2009.
All Norwegian PLC's
Listed Norwegian PLC's
included, in order to compensate for general trends. Finally, year dummies are included in some cases in order to check for robustness.
C. Summary Statistics
This part provides more insight concerning the datasets used. At first, the general statistics of the entire dataset including all Norwegian PLC’s are reported. Secondly, more detailed statistics are presented of the dataset with only listed PLC’s. Finally, differences between the industry categories are described.
The original sample contains 1080 unique Norwegian firms that were at least one year registered as PLC in the period 1999 to 2009. The average number of female directors has risen from 0.45 in 1999 to 2.46 in 2009. While in 1999 only 5.1% (in 2000 even no more than 4.9%) of the board of directors was female, in 2008 the average percentage of female directors was 39.0%. By the time of the official announcement of the mandatory quota, December 2005, 15.8% of all Norwegian PLC board members were female. This can be derived from Figure 2.
Thus, 24.2% of all board members needed to be replaced within two years. Only 15.6% of the PLC’s were in compliance with the gender quota of 40 % in 2005 (Table A.II provides more details on the average board characteristics of all Norwegian PLC’s per year).
Figure 2: This figure presents the average percentage of female directors of all Norwegian PLC’s and the listed Norwegian PLC's, for the period 1999 to 2009. The left cylinder represents the share of all Norwegian PLC's. The right cylinder presents only the listed PLC’s. The increase of women accelerates in 2005 and the percentage of women stabilizes in 2008. Although the representation of women was higher among listed companies, due to the mandatory quota system these differences have disappeared. Data are from the Norwegian Business Register, measures are taken at 31st December of each calendar year.
