1
The Effect of a Gender Quota on Executive
Compensation
Did increasing gender diversity on boards due to the quota
legislation of 2008 bring more incentive based pay in
Norwegian public companies?
MSc Finance Thesis
Supervisor: Dr. Ligterink
Student: Tomas Urbanek
Student number: 11147288
2
Statement of Originality
This document is written by Student Tomas Urbanek who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.
The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
3
Acknowledgements
I would like to thank my supervisor and professor Dr. Ligterink for giving me thorough feedback during my presentations and meetings. Additionally, I would like to thank Mr. Bjorn Witlox for facilitating data from Datastream for me.
4
Abstract
The study shows that forced gender quota implementation has produced uncertainty in aligning the interest of the principal and the agent. It shows that volatility has increased with more females on boards and that managements had to change their methods of compensation. The research supports past findings that additional females on board facilitate higher equity based pay and total pay to both directors and the CEO. Independent directors that were female had the biggest impact on facilitating equity based pay and total pay for directors, whilst constraining CEO’s total pay. This research is the first ever conducted research that looks solely at the effect of a forced gender quota on compensations of boards and CEOs in Norway.
5
Table of Contents
1. INTRODUCTION………. 6
2. ADDITIONAL INFORMATION………... 8
2.1 The Norwegian Quota……… 8
2.1.1 Limited Liability Companies in Norway………... 9
2.1.2 Compliance with Legislation……… 9
2.1.3 International Awareness of the Quota………. 9
2.2 Norwegian Governance……… 10
2.2.1 Norwegian Boards………. 10
2.2.2 Stock Based Compensation……….. 10
2.2.3 Legislature Changes and Effects on Stock Compensation……….. 11
3. LITERATURE REVIEW………... 11
3.1 Glass Ceiling & Pipeline Theory……….. 12
3.1.1 Gender Quota Theories………... 12
3.1.2 Gender Discrimination Theories……….. 13
3.1.3 Common Female Appointments on Boards………. 14
3.1.4 Female Independent Directors………. 14
3.1.5 Evidence to Theory……….... 15
3.2 Principal -Agent Theory………. 16
3.2.1 Women in Monitoring Committees……….. 17
3.3 Equity Based Theory……….. 17
4. HYPOTHESES……….. 22
5. METHODOLOGY……… 26
6. DATA AND DESCRIPTION OF STATISTICS………... 32
7. RESULTS……….. 37
8. ROBUSTNESS CHECKS………. 41
9. CONCLUSION……… 48
6
1. Introduction
Gender diversification of boards is a pertinent topic that is forward looking. Many governments in Europe such as Norway, Denmark, Iceland, Spain and France are already obliged and binding to a quota system of female employment on boards and many more such as Germany and the UK want to establish it in their country in the upcoming years. Prior research on gender quota and firm performance by Ahern & Dittmar (2012) show that the legislation did not benefit Norwegian public companies causing a period of uncertainty on the Norwegian stock market.
The main topic of this research is the Norwegian gender quota and its effects on compensation. This research aims to aid current and past efforts on looking at gender diversity of boards and its effects on the firm. The question that this research aims to answer is whether public companies behave differently in terms of equity compensation when a new law on gender diversity has been announced, that will essentially affect the structure of all public boards in the country. The research question of the thesis is: “Did increasing gender diversity on boards due to the quota
legislation of 2008 bring more incentive based pay in Norwegian public companies?”
The research will test the relationship between an increasing number of females on boards on the directors’ and CEO stock and total compensation. The methodology will take into account the top 50 Norwegian public firms according to current market capitalization. Top 50 companies were chosen as many companies delisted or enlisted in a different country to avoid the quota. Additionally, there is a tendency of wealthier firms to give higher compensation, thus 50 top firms are used to be able to gather relevant data more easily. Firstly, the research will show how equity based pay evolved through time taking into account the time frame from years 2000 to 2015. Then the effect of increased female participation on executive equity based pay will be
7 analyzed through OLS regressions. Several control variables have been used that are based on the methodology of Adams & Ferreira (2009). Directors and CEO data on compensation have been retrieved from BoardEx and annual reports, whilst financials to create controls and fixed effects have been obtained from Datastream. Additionally, annual reports were used if there is any missing information from Datastream and BoardEx.
Many companies did not fulfill the 40 % female quota on boards by the compliance date of 2008, and were sanctioned by the Norwegian government. The reasoning behind noncompliance of the law was the volatility in share prices that was expected to be caused from the forced gender quota. Many firms were sanctioned, delisted or transferred their companies abroad. Prendergast (2002a) has explored the way compensation is adjusted to risk and uncertainty. This paper contributes Prendergast (2002a&b) research as it explores a new kind of uncertainty on the market, which is forced gender diversification of boards.
Adams & Ferreira (2009) made a similar study on gender diversification and its effects on boards in the USA. Two of the effects that they test is the effect of gender diversification on fraction of equity based pay and total pay for both directors and CEO. They analyze equity based pay as it is an important type of compensation that is implemented in uncertain events as shares are more likely to provide more performance based incentives than a fixed salary. Another reason why they test equity is that most bonuses are paid in equity and not in cash, thus equity is an important compensation mechanism. This paper will contribute their study by looking at a differing effect of forced gender diversification compared to unforced and testing how it impacts the pay structure of directors and the CEO in a European setting instead of an American setting.
8 Ahern & Dittmar (2012) looked at the effect of the forced gender quota on firm performance in Norway. This paper will contribute to their findings on the role of women on boards and its effects on corporate governance and compensation. They also find that more women on boards bring higher equity based pay. Additionally, this paper contributes to their findings on the common placement of women on boards and the importance of independent directors.
The thesis is divided into nine sections to make it more coherent for the reader. Firstly, additional information on the quota and governance of Norway is presented in part 2. This is followed by section 3 which covers the literature review and gives the reader a theoretical framework of the research. Section 4 relates the framework to the thesis and presents the desirable hypothesis for the research. Section 5 outlines the methodology that will be used to get to the empirical findings of the research. Part 6 describes the data in more detail, including summary of statistics and other aggregate findings. Section 7 presents and discusses the main results. Part 8 covers different robustness checks of the results and Section 9 concludes the main findings of the research.
2. Additional Information
2.1 The Norwegian Quota
There was a lot of discussion in Norway to increase equality of sexes in workplaces in the early 2000’s. The idea of a quota in Norway dates back to 1981, when the Norwegian parliament tried to implement it unsuccessfully. It took about 20 years for the quota to come back into discussions in the Norwegian parliament. In 2003, the Norwegian Company Act was firstly passed voluntarily, which stipulated a 40% public board representation of women. This legislation was made mandatory in 2006 and full compliance of this law had to be fulfilled by all public companies boards by 2008.
9
2.1.1 Limited Liability Companies in Norway
Many CEO’s of Norwegian public companies wanted to be heard and made public announcements regarding the fact that quota will bring inexperienced female workers to their companies (Criscione, 2002). As a response of many companies that could not find the right talent many ASA (public) companies turned private. Ahern & Dittmar (2012) document an increased probability of an ASA company delisting in years 2003-2009. In fact, out of 565 ASA companies in 2003, only 345 ASA companies were still public in 2008.
2.1.2 Compliance with Legislation
The quota was voluntarily passed in 2003 by the Norwegian parliament. However, since that period by 2005 there was only a 17% representation of women on boards. Due to non-compliance many sanctions were introduced. When the law came to full non-compliance in 2008, 77 out 345 public companies did not comply with law. Most of these firms were contacted by the Ministry of Trade and Industry to inform them that they have three weeks to comply; otherwise they were subject to a forced dissolution. In 2008, the law came to full effect and by mid-2008 the average share of women on public boards was around 40%. Ahern & Dittmar (2012) document an increase of women on boards of private firms by 3% whilst in public firms the increase was 37% for the years 2003-2009. Thus, it can be seen that this was an exogenous shock that took time for public boards to adjust.
2.1.3 International Awareness of the Quota
Many non-profit organizations and economic development institutions such as the UN have stressed women equality for a long period of time. The EU Parliament is another organization that stresses on gender equality at work places. These efforts have not been wasted. In fact
10 according to the Catalyst report of 2015, since 2003 the parliaments of Denmark, Finland, Iceland, Italy, The Netherlands, Israel, Kenya, Spain, Canada, India, Belgium, Australia and Spain have complied to a similar quota on public boards as in Norway. With current efforts of German, Czech and the UK parliaments, quotas are expected to come in the near future. Hence, it is an issue that not only affects some countries, but will eventually be a worldwide phenomenon. Governments and firms could use this research to see if forced quotas work efficiently on public boards of companies. Giving insights on compensation pay structure and its change over time due to quota legislation in force.
2.2 Norwegian Governance
2.2.1 Norwegian Boards
According to Ahern & Dittmar (2012) boards in Norway are not very different from the ones that are present in the USA. They find that the number of independent board members has a higher representation in a Norwegian board than a US board. The average size of a board in the USA is 7 members; however Norway had an average board size of only up to 6 members (Yang et. al, 2008).
During the quota the shareholders got to elect which women would go on boards. However, traditionally in Norway 1/3 of the board can be elected by the employees within the firm itself. Hence, it could be possible that shareholders elected more experienced women who also had a higher probability of working on other boards at the same time. This is due to the exogenous shock on board structure and the fact that there was not enough time and search costs to search for better candidates.
11 In Northern Europe, stock compensation has since 1990’s become a popular trend (Oxelheim & Randoy, 2008). Comparing total director and CEO pay of USA with Scandinavian countries Oxelheim & Randoy (2008) find that in 1999 USA director and CEO total pay was more than 40 times greater than in Sweden and Norway. Some of the reasons for compensations to be lower in Northern Europe are: less demanding work tasks, transparency, public disclosure of compensation, influence of union representations and the structure of the economy itself.
2.2.3 Legislature Changes and Effects on Stock Compensation
Oxelheim & Randoy (2008) give evidence of an effect of legislation on stock compensation in Norway. When new tax law came into effect for equity based compensation, the number of stock options in use fell by 18,600 from an initial 19,600 options from years 1995-1998. Suggesting that equity based pay is more sensitive to tax policy changes. In 2003, there was an accounting regulation that took into effect called the IFRS-2. This legislation was supposed to make stock based pay more transparent. Melis & Carta (2009) record that this new regulation has made equity based pay more favorable for managements due to transparency reasons. This regulation gave certain benefits to executives such as the fact that stocks that were unvested were more easily facilitated back after leaving and entering a new company due to transparency reasons.
3. Literature Review
The literature review is divided into 3 sections: Glass Ceiling & Pipeline Theory, Principal Agent Theory and Equity Based Theory. Each of these sections have different purposes within the framework context. The Glass Ceiling & Pipeline Theory tries to discuss different gender discrimination theories, the most common function of a female director and past evidence of
12 effect of gender diversity on firm performance. This section is created to show why in many cases it was not in the best interest of companies to hire women on boards and explains how it could have created volatility and uncertainty in companies’ performance in Norway. The Principal Agent Theory section looks into how difficult it was to find experienced female talent in Norway after the implementation of the quota. This section also discusses the monitoring effect of having more women on board and its reflection on governance and compensation. The Equity Based Theory section tries to give the reader insight into reasoning behind equity based pay as an incentive and retention mechanism applying it to the quota example. All three sections aim to clarify the reader with three important facts. Firstly, that it might have not been in the best interest for public companies to hire women during the quota. Secondly, that the quota was actually inefficient in the short run and caused a lot of uncertainty. Thirdly, that equity based pay should increase due to this uncertain event in order to align interests of the company with the agent, and to be able to measure performance better.
3.1 Glass Ceiling & Pipeline Theory
The “glass ceiling” or “pipeline theory” states that there is a barrier to which women can progress in their career due to their gender difference (Morrison, 1987). The quota implemented by the Norwegian parliament can provide the initial steps that facilitate women to break this barrier, as women do not tend to have networks with the “old boys club” that can help them climb up their career in terms of position within the corporation. Thus, quotas should be able to overcome any prejudices in a business setting by creating more exposure for female professionals.
13 Coate & Loury (1993) argue that if high quality talent cannot be found, gender quotas may be an inefficient mechanism that shifts focus of women in investing less in their careers as the qualification level to become one of the members of the board is lowered. Essentially, creating a board appointed of less experienced women, with women having minimal say in board decisions and minimal opportunities to have better networks. This claim is supported by Becker (1971) and his “Human Capital Theory,” which states that employees usually make rational decisions in terms of investing in their future (education and work experience). Hence, if women built up less human capital then their career progression options would be constrained.
3.1.2 Gender Discrimination Theories
There are a few theories that look into gender decimation in a business setting. Kanter (1977) proposed the “social certainty theory” which states that there is a desire for decision makers to appoint people on boards who they prefer to work with. This effect is called the “similar to me effect”, implying that employers like to hire people that are similar to themselves. Strober (1984) concluded that some boards believe in patriarchy and like to keep women in a dependent status. This status gives aid to the male counterpart’s career opportunities. Larwood, Swazajkowski & Rose (1988) write about the “rational bias theory” stating that there tends to be an intentional bias in managerial decisions, where managers are not interested in making any effort to decrease gender discrimination and want to act in their own best interest. Motowildo et. al (1986) believes that gender discrimination can be unintentional, due to the common stereotype of having a male executive on board with some qualifications. Thus, many managers can be influenced by a characteristic such as gender in granting a position, which is unrelated to the skills necessary for the job.
14
3.1.3 Common Female Appointments on Boards
Evidence from past research on gender diversity in boards and its effect on firm performance give some general information on what positions women tend to be appointed on a board of a company. Adams & Ferreira (2009) find out in their research on gender diversity in boardrooms in the USA, that women are more likely to be appointed independent directors than men. Specifically, 87% of women on boards from years 1998 to 2003 were independent directors, whilst 6.4% were women inside directors.
However, independent directors sometimes struggle to bring value on firms. Myers & Majluf (1984) find that due to information asymmetry there is an informational advantage of insiders over the independent directors and hence an increase of these directors would decrease firm value. Duchin et. al (2010) relates exogenous legal effects of increased outside directors due to the Sarbanes Oxley 2002 agreement. His findings suggest that information asymmetry has to be reduced in order to keep up with the increased presence of outside directors and still maintain a positive Tobin’s Q.
3.1.4 Female Independent Directors
Seierstad & Opsahl (2010) find through empirical analysis that the quota in Norway increased the number of women in the positions of independent outside directors. However, they also found that the quota caused more experienced women to work on several boards at a time. Thus, findings collected by other researchers such as Ahern & Dittmar (2012) about the significant percentage increase of women employed on boards in Norway due to the quota, might be ambiguous. As evidence from Seierstad & Opsahl (2010) suggest that more experienced women were often appointed to work on boards of several public companies. The quota led to an
15 increase of the maximum number of board seats per outside director from four to eight board seats. Thus, a percentage increase of females on boards might be accounted by an outside female director who is working on several boards at a time and not a proportional increase of females to males in all public boards.
Study by M. Fich & Shivdasani (2006) shows how firms with busy outside directors tend to exhibit lower levels of corporate governance and a reduced Tobin’s Q. Additionally, departures of such directors created positive abnormal returns. Each additional dictatorship that an outside director held created a negative abnormal return for the original companies in which they held their seats from the beginning. Additionally, if the recorded percentage increases of gender diversity on boards take into account a large fraction of female busy independent directors, the quota itself might have not been efficient to give all female professionals the chance to gain a seat on the board.
3.1.5 Evidence to Theory
The future uncertainty and potential volatility in earnings due to the gender quota, created incentives for Norwegian public firms to avoid the quota and the sanctions related to non-compliance. Many firms were sanctioned, some delisted and a few incorporated in other countries, mainly in the UK. Ahern & Dittmar (2012) find that throughout the time period of 2001 to 2009, the number of public companies in Norway has decreased by 70% and private limited companies experienced an increase of more than 30 %. They also found that there was negative relationship between the number of new private listed firms and gender representation on public boards. Thus, this is implying that many companies have rather gone private than having to commit to the quota.
16 Ahern & Dittmar (2012) in their study show a detrimental effect on firm’s profitability in Norway due to the quota. They show that gender diversity on boards came at a high cost of lower stock returns and lower Tobin’s Q. Their main reasoning is the fact that many firms had to hire less experienced young women on boards.
Matsa & Miller (2011) in their study on the effects of Norwegian quota found out that costs have increased and operating income declined in many public companies due to the quota. Especially, the costs that were aggregated during the compliance period from 2006 to 2009 have increased the overall costs to firms substantially. However, they also find that the amount of laid off employees during the quota was much less than others’ research suggests. This could be because of the fact that there was no need to hire more females, but to hire females as independent directors serving on multiple boards of different companies
3.2 Principal -Agent Theory
The standard theory in Corporate Governance would state that an agent (manager) should act in fiduciary duty of the principal (shareholders). The shareholders are mainly focused on value maximization and the agents (managers) should try to achieve targets to maximize value for the firm. However, there is informational asymmetry between the principal and the agent. Hence, it is not always clear whether the agent will act in the best interest of the principal.
Looking at the perspective of value maximization; past empirical evidence found by Ahern & Dittmar (2012) suggests that quotas do not add much value to the company in the short term and long run in terms of stock prices and firm value. Thus, to avoid these effects it would be in the best interest of the whole company not to appoint females on boards as this creates costs to the firms.
17 Additionally, it must be taken into account that the pool of experienced women to choose from was very limited. Looking at the studies of Ahern & Dittmar (2012); they show how the quota led to a decreased Tobin’s Q over the period since the announcement in 2003 to 2009 with main explaining factor being uncertainty and inability to find the best talent. Principals incentivize aims by aligning the interests of principal with the agent through compensation. Stock based pay is an aligning mechanism that aligns the goals of the principal to the agent efficiently and makes him/her more engaged in the company. As according to Jensen & Meckling (1976) equity based pay is one of the key mechanisms that can aid to the reduction of principal agent problem.
3.2.1 Women in Monitoring Committees
Adams & Ferreira (2009) find that women had a higher probability of gaining a position in monitoring committees, and much lower chance of getting a position in compensation committee. This is also proved by evidence from Norway. Nygaard (2011) shows in his study that woman created tougher monitors on boards reflecting higher caution to stock prices and efforts of managers. Adams & Ferreira (2009) find that an increase in gender diversity on US boards increases the sensitivity of CEO turnover to poor stock performance. Indicating that a CEO is more monitored on a gender diverse board and hence more sensitive to the outcomes of the market. Empirical evidence from Adams & Ferreira (2009) findings show that equity based pay has increased in more gender diverse boards Thus; it could be possible that it would be in the best interest of the shareholders and the agent that executives get compensated through equity. As this form of compensation can help align the interests of the two parties and the agent will care more about the value he will create for the principal (Besanko, et. al 2010)
18 Holmstrom & Milgrom (1987) predicted in the principal-agent theory that there is a negative trade-off between risk and incentives in a compensation contract. They argue that increasing pay for performance incentives creates greater empowerment of employees, but increases uncertainty over their compensation. It has been predicted that higher output risk within a company would lead to lower pay for performance to a risk-averse agent. This is due to the fact that with more uncertainty in earnings, pay for performance becomes more expensive to compensate the workers for the additional risk of not meeting their performance target in the future in a current uncertain environment.
However, many empirical studies have proved this trade off to be wrong empirically, showing substantial evidence that pay for performance is positively linked with uncertainty or volatility. Prendergast (2002a) tries to put together all empirical evidence from previous literature on this trade off phenomenon. He finds empirical evidence in favor of a positive relation between uncertainty and pay for performance incentives. Prendergast (2002a) explores the impact on executives, sharecroppers, franchises and sales force workers. His findings show that with future uncertainty in earnings and the more complex the businesses operations are, the more likely there is an incentive contract implemented. Since, the principal may not have the best idea of what targets and efforts to achieve within the unstable future prospect due to the expected volatility. In more stable environments agents are simply delegated the tasks that they should do, whilst in uncertainty offering incentive agency contracts provides a better measure of performance to base pay on. Directors on boards, in his research prove to be the most common setting for this practice. The reasoning he suggests behind this contradiction to theory is that in an uncertain environment it is much difficult to give tasks and award inputs of agents. Thus, in an uncertain environment he finds that many contracts change to incentive pay for performance based on
19 overall output. This form of agency contract reflects the benefits and costs, and facilitates a suitable behavior for the management to cope with the future uncertain events. Since delegating tasks in an uncertain environment and controlling the inputs of agents and awarding them accordingly has proved to be difficult. Adams & Ferreira (2009) empirical results on evaluating the effect of an increased gender diversity of US boards on director compensation prove Prendergast (2002a) prediction. Volatility also increases equity based pay with more women on boards. Adams & Ferreira (2009) compare equity based pay and total pay for both directors and CEOs. They find that equity compensation increases with more women on boards. The fundamental reason why equity based pay is used as an incentive mechanism is that shares and options provide more performance based incentives than a fixed salary. Another fundamental reason to why equity based pay is important in this thesis is the fact that directors do not usually receive performance based bonuses paid in cash but in equity. Thus, equity based pay is used to examine the effect of a gender quota on compensation as a variable that is used to incentivize employees and retain in uncertain situations.
Becher et. al (2005) findings show that equity based pay for directors increases their incentives to perform and increases the value for shareholders of the firm. This type of compensation is made to motivate directors and to constrain any opportunism from managers. Brick et. al (2006) also find that equity based pay is important for future investment and firm performance, finding that increased director pay tends to be positively related to future performance. An example of improved future performance due to equity based pay is the banking industry where Becher et. al (2005) find an increase in accounting earnings with more equity based pay.
The standard principal agent theory would show that aligning the interests of principal with agent is important to bringing appropriate corporate governance within a firm. It would seem
20 appropriate that with more equity given to agents, this should align the incentives of the firm with the agent. Since, the agent owns a higher share of equity it would be in his/her best interest for the company to perform well on the stock exchange. However, tendency in the past was that upper management owned a very little stake of equity of the company. Recently, this trend has changed and many agents get a higher equity stake to align the interests of the principal with their interests (Franzetti, 2011). Equity would be favorable if there is an assurance that the agent will not divert from interests of the company. However, equity in terms of restricted stock (unvested equity) compared to stock options can be in terms of long term incentives and long term retention planning a highly effective tool (Hall & Murphy, 2003). This form of equity based pay has become very popular in the last decade as it assures that agents try to maximize shareholder value and also empowers, incentivizes and retains management for better firm performance.
Hall and Murphy (2003) argue that stock options are used to attract, retain, motivate and recognize subordinates to the firm. Drobetz et. al (2004) give three valid arguments why stock compensation is important. Firstly, equity options attract highly skilled workers in a competitive environment who see a profitable future of the company. Executives will more likely remain in a firm if he/she has unvested stock options. More specifically, non-tradable stock options are the most important type that retains executives in the firm. This equity based pay is paid in accordance to a performance metric and has certain vesting periods that create incentives for executives to stay in the firm. Lastly, this form of compensation is most favorable for firms as it is relatively an inexpensive form of compensation.
Oyer (2004) states that unvested equity could serve as an incentive tool to retain executives. However, this incentive tool could vary depending on the performance of a firm. He also states
21 that is mainly used in more profitable firms, as they try harder to retain their successful managerial talent. Jochem, Ladika & Sautner (2015) show how un-vested equity is a highly effective tool for managerial retention. Their results suggest that a 1% standard deviation decline in un-vested stock options leads to an increase of CEO departure and director departure by more than 50%. This supports the study of Adams & Ferreira (2009) which empirically claim that the number of stock based compensation has increased with an increase in gender diversity on US boards. Yermack (2002) claims that stock based pay varies with current competition on the executive labor market. Indicating that with more competition and uncertainty, increased stock based pay is present to attract but also to retain experienced valuable executives.
Relating equity based pay with common female placements on boards (independent director, monitoring committee). Hoskisson et. al (2009) argue that increased monitoring on boards facilitate a higher pay for performance. They find that monitoring leads to an increase in incentive pay, whilst excessive pay contributes to excessive monitoring. This is supported by the fact that when monitoring is intensified, this induces the manager to a shift in risk preferences, which have to be compensated. Prendergast (2002b) also supports this view and states that in an uncertain setting, monitoring is not the most efficient as it very difficult to evaluate performance. Due to the possibility of mistakes in investigations and evaluations of monitoring committees in uncertain situations, employees are motivated and delegated better through high incentive pay mechanisms. Thus, he offers the idea of a pay for performance mechanism based on incentives to reach the goals of the company and reduce the uncertainty in monitoring. Rosenstein & Wyatt (1997) find that increasing equity based compensation of directors also increases the CEO turnover to stock performance. This could explain the phenomenon behind Ahern & Dittmar (2012) findings that show that CEO turnover has increased in Norway with a decreased firm
22 performance as many women were hired as independent directors and inside monitoring committees. Ryan & Wiggins (2004) also find that with more independent directors on the board more incentive pay is used as a form of compensation. They argue that with an excess of independent directors more bargaining power over CEOs is present. This leads to an increase in pay for performance to provide better alignment of interests within the company.
With previous literature showing vast evidence supporting the claim that equity based pay should increase during higher uncertainty, this thesis will aim to contribute to previous findings. With the expectation of a positive effect between volatility and equity based pay, this thesis will try to examine whether a forced gender quota will facilitate the same results as in Prendergast’s (2002a) uncertainty setting and whether it will complement the results of Adams & Ferreira (2009) that found a similar effect from empirical results from the USA.
4. Hypotheses
The framework presented above has hopefully induced the reader into the effect that is to be examined in this research. A short summary of main points are presented followed by the desired hypotheses for the thesis. Ahern & Dittmar (2012) proved that women created difficulties to companies in both search costs and short term performance. Many women after the legislation worked on more boards than one; hence it is ambiguous whether women really represented a 40% representation on boards of all public companies due to the forced gender quota implementation in Norway. Due to information asymmetry between new comers (independent female directors) and inside directors, due to the difference in level of experience, due to the creation of a tougher monitoring system and substantial search costs all come to conclude that it
23 might have not been in the best interest for Norwegian public companies to be hiring women in such a forced way as the mandate obliged them to do. The thesis has also introduced the reader to the concept of using equity based pay in more uncertain situations. Reasoning that equity based pay has been used as an equity incentive pay mechanism in companies with uncertain future prospects in the past. The literature review also explains that equity based pay creates more alignment of shareholder and principal goals.
Utilizing these insights from past findings on equity based pay, gender theories and outcomes of the quota; the thesis builds upon four hypotheses. These hypotheses are based on the uncertainty that is created due to an event such as the forced gender quota and its impact on compensation of the board and the CEO. The hypotheses are related to the hypotheses presented by Adams & Ferreira (2009) with the difference of using Norwegian companies instead.
Hypothesis 1: Having more women on board will cause fraction of equity based pay from total pay to increase for directors
𝐹𝑟𝑎𝑐𝑡𝑖𝑜𝑛𝐸𝑞𝑢𝑖𝑡𝑦𝐵𝑎𝑠𝑒𝑑𝑃𝑎𝑦𝑖,𝑡
= 𝛼𝑖 + ⅄𝑡 + 𝛽1𝑅𝑂𝐴𝑖,𝑡+ 𝛽2𝐵𝑜𝑎𝑟𝑑𝑠𝑖𝑧𝑒𝑖,𝑡+ 𝛽3𝐹𝑟𝑎𝑐𝑡𝑖𝑜𝑛𝐹𝑒𝑚𝑎𝑙𝑒𝑠𝑖,𝑡
+ 𝛽4𝐵𝑜𝑎𝑟𝑑𝐼𝑛𝑑𝑒𝑝𝑒𝑛𝑑𝑒𝑛𝑐𝑒𝑖,𝑡+ 𝛽5𝑇𝑜𝑏𝑖𝑛′𝑠𝑄𝑖,𝑡+ 𝛽7𝑉𝑜𝑙𝑎𝑡𝑖𝑙𝑖𝑡𝑦𝑖,𝑡+ 𝛽8𝐿𝑜𝑔𝑆𝑎𝑙𝑒𝑠𝑖,𝑡
+ 𝛽9𝐵𝑢𝑠𝑖𝑛𝑒𝑠𝑠𝑠𝑒𝑔𝑚𝑒𝑛𝑡𝑠𝑖,𝑡+ 𝜀𝑖,𝑡
Prendergast (2002a) states that if firms with better governance give out more equity based pay, all theories centralized by the standard principal agent problem would predict that total pay should also increase due to the compensation for extra risk of uncertainty. Hence, the second hypothesis aims to explore whether total pay has also increased.
24 𝑇𝑜𝑡𝑎𝑙𝐷𝑖𝑟𝑒𝑐𝑡𝑜𝑟𝑝𝑎𝑦𝑖,𝑡
= 𝛼𝑖 + ⅄𝑡 + 𝛽1𝑅𝑂𝐴𝑖,𝑡+ 𝛽2𝐵𝑜𝑎𝑟𝑑𝑠𝑖𝑧𝑒𝑖,𝑡+ 𝛽3𝐹𝑟𝑎𝑐𝑡𝑖𝑜𝑛𝐹𝑒𝑚𝑎𝑙𝑒𝑠𝑖,𝑡
+ 𝛽4𝐵𝑜𝑎𝑟𝑑𝐼𝑛𝑑𝑒𝑝𝑒𝑛𝑑𝑒𝑛𝑐𝑒𝑖,𝑡+ 𝛽5𝑇𝑜𝑏𝑖𝑛′𝑠𝑄𝑖,𝑡+ 𝛽7𝑉𝑜𝑙𝑎𝑡𝑖𝑙𝑖𝑡𝑦𝑖,𝑡+ 𝛽8𝐿𝑜𝑔𝑆𝑎𝑙𝑒𝑠𝑖,𝑡
+ 𝛽9𝐵𝑢𝑠𝑖𝑛𝑒𝑠𝑠𝑠𝑒𝑔𝑚𝑒𝑛𝑡𝑠𝑖,𝑡+ 𝜀𝑖,𝑡
To examine the effect from both top management and at executive CEO level, the same framework applies to CEO compensation and thus will be explored through similar hypotheses to hypothesis 1 and 2. Since, Adams & Ferreira (2009) found that CEO turnover and equity based pay has increased with more women on boards.
Hypothesis 3: Having more women on board will increase the fraction of equity based pay from total pay for the CEO.
𝐹𝑟𝑎𝑐𝑡𝑖𝑜𝑛𝐸𝑞𝑢𝑖𝑡𝑦𝐵𝑎𝑠𝑒𝑑𝑃𝑎𝑦𝑖,𝑡
= 𝛼𝑖 + ⅄𝑡 + 𝛽1𝑅𝑂𝐴𝑖,𝑡+ 𝛽2𝐵𝑜𝑎𝑟𝑑𝑠𝑖𝑧𝑒𝑖,𝑡+ 𝛽3𝐹𝑟𝑎𝑐𝑡𝑖𝑜𝑛𝐹𝑒𝑚𝑎𝑙𝑒𝑠𝑖,𝑡
+ 𝛽4𝐵𝑜𝑎𝑟𝑑𝐼𝑛𝑑𝑒𝑝𝑒𝑛𝑑𝑒𝑛𝑐𝑒𝑖,𝑡+ 𝛽5𝑇𝑜𝑏𝑖𝑛′𝑠𝑄𝑖,𝑡+ 𝛽7𝑉𝑜𝑙𝑎𝑡𝑖𝑙𝑖𝑡𝑦𝑖,𝑡+ 𝛽8𝐿𝑜𝑔𝑆𝑎𝑙𝑒𝑠𝑖,𝑡
+ 𝛽9𝐵𝑢𝑠𝑖𝑛𝑒𝑠𝑠𝑠𝑒𝑔𝑚𝑒𝑛𝑡𝑠𝑖,𝑡+ 𝛽9𝐶𝐸𝑂𝐴𝑔𝑒𝑖,𝑡+ 𝛽10𝐶𝐸𝑂𝑡𝑒𝑛𝑢𝑟𝑒𝑖,𝑡 + 𝜀𝑖,𝑡
Hypothesis 4: Having more women on board will increase CEO total pay
𝐶𝐸𝑂𝑡𝑜𝑡𝑎𝑙𝑝𝑎𝑦𝑖,𝑡
= 𝛼𝑖 + ⅄𝑡 + 𝛽1𝑅𝑂𝐴𝑖,𝑡+ 𝛽2𝐵𝑜𝑎𝑟𝑑𝑠𝑖𝑧𝑒𝑖,𝑡+ 𝛽3𝐹𝑟𝑎𝑐𝑡𝑖𝑜𝑛𝐹𝑒𝑚𝑎𝑙𝑒𝑠𝑖,𝑡
+ 𝛽4𝐵𝑜𝑎𝑟𝑑𝐼𝑛𝑑𝑒𝑝𝑒𝑛𝑑𝑒𝑛𝑐𝑒𝑖,𝑡+ 𝛽5𝑇𝑜𝑏𝑖𝑛′𝑠𝑄𝑖,𝑡+ 𝛽7𝑉𝑜𝑙𝑎𝑡𝑖𝑙𝑖𝑡𝑦𝑖,𝑡+ 𝛽8𝐿𝑜𝑔𝑆𝑎𝑙𝑒𝑠𝑖,𝑡 + 𝛽9𝐵𝑢𝑠𝑖𝑛𝑒𝑠𝑠𝑠𝑒𝑔𝑚𝑒𝑛𝑡𝑠𝑖,𝑡+ 𝛽9𝐶𝐸𝑂𝐴𝑔𝑒𝑖,𝑡+ 𝛽10𝐶𝐸𝑂𝑡𝑒𝑛𝑢𝑟𝑒𝑖,𝑡 + 𝜀𝑖,𝑡
25 The variables CEO / Chairman, CEO gender have not been included from the original OLS regression presented by Adams & Ferreira (2009) due to the fact that in Norwegian public board structure it is not common to have a chairman and CEO at the same time. CEO gender dummy is excluded from the regression as almost all of the CEO’s from the listed companies where male.
Another difference in this research in comparison with Adams & Ferreira (2009) is the fact that this paper does not use the log transformation of dependent variables in its regressions. However, Adams & Ferreira state that the results they conducted were similar even if they did not make this specification on the dependent variable. Hence, this paper uses dependent variables that are not a log transformation.
Possible endogeneity problems can arise because of some unobservable firm characteristic that could affect the equity based pay and total pay of both directors and the CEO. Omitted variables that could affect the selection of female directors and compensation choices may entail correlations that are spurious. This research takes into account that changing corporate culture could affect the results of the regressions. Hence, this thesis includes firm fixed effects to address any time invariant characteristic of a firm. An alternative endogeneity problem could be multicollinearity between the main explaining variables according to past evidence, which are fraction of females on boards and number of independent directors. If all findings presented in literature review are correct and that women were mainly hired as independent directors in Norway, then these two variables could be correlated. To address this problem the research omits the fraction of females on boards to test the effect of just fraction of independent directors in the robustness section. Additionally, an interaction term is created as product of a fraction of females on boards and fraction of independent directors that is compared to the omitted regression in part 8 of the thesis.
26
5. Methodology
5.1 Relevant Data and Sources
Replicating the study of Adams & Ferreira (2009) but on a Norwegian setting the thesis contains the same independent and dependent variables as in their paper. Utilizing cross sectional data with fraction of equity based pay directors, fraction of equity based pay CEO, total pay of directors, total pay of CEO as the key dependent variables. ROA, Tobin’s q, number of business segments, log of Sales, board size, volatility, CEO age, CEO tenure are variables needed as control variables to get results in line with study of Adams & Ferreira (2009).
Firm fixed effects and industry effects are included by incorporating dummies into the regressions based on firm codes and industry codes respectively. These controls are used to limit time invariant characteristics and cross industry characteristics that can possibly change the results. The time period that is relevant for this research is 2000 to 2015. This time period is used to effectively see if there were any changes in equity based compensation before and after the quota went fully into effect in Norway.
27 Table 1: Selected Public Norwegian Companies
Due to the fact that many ASA firms that were public turned to AS private companies, this research will take into account only the top 50 public companies according to current market capitalization to have a complete data sets for all the years. This is in accordance to theory as bigger firms should also be able to easier facilitate equity based compensation for its directors and CEO, due to the bigger source of financing. Additionally, the larger market capitalization firms will be the ones where there will be the highest competition to get a job, and thus equity based pay is more likely to be present.
5.2 Statistic Method
Utilizing the data, two way line graphs will be constructed to see the evolvement of both director and CEO equity based pay and total pay. Then the research will try to estimate a relationship between increased fraction of females on boards and executive compensation. OLS is the method that will be used to get estimates on the effects between the variables. The method uses panel
FRONTLINE LTD FARSTAD SHIPPING ASA STOLT NIELSEN ODFJELL ASA
PROSAFE SE ARENDALS FOSSEKOMP. SUBSEA 7 S.A. NRC GROUP ASA ABG SUNDAL COLLIER ORKLA ASA SCHIBSTED ASA DNO ASA
EKORNES ASA NORSKE SKOGINDUST. KONGSBERG GRUPPEN GYLDENDAL ASA TTS GROUP ASA HAFSLUND ASA INCUS INVESTOR ATEA ASA STOREBRAND ASA NTS ASA
MARINE HARVEST ASA NORSK HYDRO ASA NORDIC SEMICONDUCTOR SKIENS AKTIEMOLLE DATA RESPONS ASA OLAV THON EIENDOMSS. I.M. SKAUGEN ASA TOMRA SYSTEMS ASA AF GRUPPEN ASA VEIDEKKE ASA TGS-NOPEC GEOPHYSIC PHOTOCURE ASA KITRON ASA TELENOR GROUP SOLSTAD OFFSHORE ASA DOF ASA BYGGMA ASA STATOIL ASA
FRED. OLSEN ENERGY PETROLEUM GEO-SVCS BIONOR PHARMA ASA GC RIEBER SHIPPING BONHEUR ASA WILH WILHELMSEN BORGESTAD ASA SPAREBANK 1 SR BANK TIDE ASA CONTEXTVISION AB
28 data balanced regressions with controls for each industry and firm. Above including dummies as controls for firm fixed effects and industry effects, year dummies will be included as well to fully replicate the study of Adams & Ferreira (2009) Year dummies are included to exclude the possible aggregate trends of the collected data and to be able to facilitate a test for a more causal relationship between the variables.
5.3 Performance & Control Variables
The research design, takes into effect control variables and industry dummies for any endogeneity problem that might occur. The method is a replication of Adams & Ferreira (2009) method where they also use OLS regressions to test how gender diversity on boards increases CEO and director stock based pay and total pay. Due to the fact that this research is concentrated on Norway, there might be differing endogeneity problems that might occur later at the stage of collecting results through OLS. The study by Adams & Ferreira (2009) has conducted firm fixed effects and industry firm fixed effects separately into different regressions for each dependent variable. This research will combine these control variables in all regressions due to the fact that the results this paper obtained by combining were nearly the same as when they were conducted separately, due to the relatively smaller size of the sample that is used in this paper.
5.3.1 Testing Director Equity Based and Total Pay
Table 2: Definition of variables testing directors’ fraction of equity based pay and total pay
Dependent Variables Fraction of Equiity Based Pay Total Pay Directors
Independent Variables Description Data Source Predicted Sign Predicted Sign
ROA Annual net income/ total assets Datastream (+) (+)
Log of Sales Annual log of revenues in Norwegian Krone (NOK) Datastream (+) (+)
Tobin's Q The market value of equity plus asset book value minus the book value of equity Datastream (+) (+)
Volatility The standard deviation of monthly stock returns of a firm over the 15 year period Datastream (+) (+)
Fraction of Females The fraction of females in a board Annual Reports (+) (+)
Fraction of Independent Directors The fraction of independent directors on the board Annual Reports (+) (+)
Board Size The number of directors on the board Annual Reports (+) (+)
29 Table 2 depicts the dependent and independent variables used in the research, the source of the independent variables and its predicted sign according to previous research.
ROA
The ratio is calculated using the firm’s net income and dividing by the book value of assets. We would expect a positive coefficient since it would imply that with a higher ROA we would expect more available sources to pay for equity based pay and total pay.
Tobin’s Q
This is a proxy for performance of a firm depending on the market. The market value of a firm will be calculated as the market value of equity plus asset book value minus the book value of equity. The expectation would be that with an increase in Tobin’s Q there would be higher equity based compensation. However, from previous studies we know that Tobin’s Q is expected to decrease for many firms due to increased gender diversity. Due to volatility in Tobin’s Q it would also be predicted that equity based pay should also increase (Prendergast, 2002).
Volatility
This metric measures the riskiness of the firm in terms of stock prices. It will be calculated as the standard deviation of monthly stock returns over the 15 year period that is measured. Volatility would also be positively related to equity based pay (Prendergast 2002a&b). From previous studies on Norwegian quota we would expect stock prices to be more volatile with more women on boards. Hence, aggregate volatility should have a positive and significant impact on equity based pay overall.
30 Amount of segments the firm has will be retrieved from annual reports. This variable serves purpose for firm complexity. The idea is that with more business segments the firm has more revenue streams to be able to afford to give equity based pay to its executives. Hence, an increase in business segments would be expected to have a positive effect on equity based pay.
Log of Sales
This is the amount revenues the firm has. The revenues will be subject to a log function to see the increases in sales instead of sales amounts itself. It is used in this research as a proxy for firm size. The effect of this coefficient is expected to be positive and significant. Since, bigger firms with higher revenues would imply more available financing for executive equity based pay. The variable is a log variable as it gives clearer and better results according to Adams & Ferreira (2009). Using log variable the thesis has also got more significant results when the log of sales was included instead of sales.
Fraction of Female Directors
This is essentially the ratio of females to men. This variable should show a positive and significant effect on director equity based pay.
Fraction of Independent Directors
The ratio of independent directors to incumbent directors, measures how much independence the board has. This variable is expected to be correlated with equity based pay. This is supported by the fact that women tend to be hired in Norway mainly as independent directors. Thus, with more women acting as independent directors implies more uncertainty.
31 This is essentially the number of members on the board. The expected effect of this coefficient would be positive. As an increase in board size would mean facilitating extra pay for one extra director and having available financing to do so.
5.3.2 Testing CEO Equity Based and Total Pay
Table 3: Definition of variables explaining CEO’s fraction of equity based pay and total pay
To test CEO fraction of equity based pay and total pay we use the same control variables as seen in Table 3. However, CEO age and CEO tenure are new control variables that are added into the OLS regression. They are added to change the specification to test the relationship between the increase of women on boards and change in CEO fraction of equity based pay and total pay.
CEO Age
According to past results from Adams & Ferreira (2009), an increase in CEO age would decrease the amount of equity based pay and total CEO pay with more women on board. Thus, we would expect a similar effect in our research.
CEO tenure
The expectation from study by Adams & Ferreira (2009) would be that an increase in CEO tenure, with more women on board would decrease CEO equity based pay and total pay.
Dependent Variables Fraction of Equity
Based Pay Total Pay CEO
Independent Variables Description Data Source Predicted Sign Predicted Sign
ROA Annual net income/ total assets Datastream (+) (+)
Log of Sales Annual log of revenues in Norwegian Krone (NOK) Datastream (+) (+)
Tobin's Q The market value of equity plus asset book value minus the book value of equity Datastream (+) (+)
Volatility The standard deviation of monthly stock returns of a firm over the 15 year period Datastream (+) (+)
Fraction of Females The fraction of females in a board Annual Reports (+) (+)
Fraction of Independent Directors The fraction of independent directors on the board Annual Reports (+) (+)
Board Size The number of directors on the board Annual Reports (+) (+)
Number of Business Segments Firm complexity and diversity in revenue streams BoardEx/ Annual Reports (-) (-)
CEO age The age of the CEO BoardEx/ Annual Reports (-) (-)
32
6. Data and Description of Statistics
Data was compiled from annual reports and databases using top 50 public companies of the Norwegian stock exchange according to market capitalization. Variables such as ROA, Tobin’s Q, Sales and monthly stock prices were retrieved from Datastream. Other variables such as annual number of shares dedicated to total directors, the total annual pay of directors, total annual amount of shares issued to CEO, total annual pay for CEO, directors’ equity based pay, CEO’s equity based pay, directors’ fraction of equity based pay and CEO’s fraction of equity based pay were retrieved and computed from annual reports that can be obtained from the Oslo Stock Exchange website or individually through companies’ websites. CEO age, CEO gender were retrieved from BoardEx. CEO tenure that was not able to be found on BoardEx was obtained from annual reports. Data that was missing or needed clarification where double checked with annual reports. Phone calls were made to companies in case reports needed more clarification. All total pay amounts and equity based pay amounts are in Norwegian Krone (NOK).
33 Table 4: Description of Statistics
As can be seen from the Table 1 of Descriptive Statistics, the mean of fraction of females on boards is 0.30, which is under the regulated amount of 40% females on boards. This variable has a range of minimum at 0.00 and maximum at 0.60. Thus, there were instances in time were these selected public companies had much less than 40% of women on their boards. The mean for the amount of independent directors is indicated at 0.24 showing that there is some relationship between females on boards and the number of independent directors. The range for this variable varies between 0.10 and 0.30. The mean for the amount of shares issued annually to the CEO is 18,868 numbers of shares and this is in a range from 6,158 shares to 25,985. Similarly for the amount of shares issued to the whole board; the mean is at 6,294 annually and ranges from 2053 to 8,750. This indicates that there were large differences in management practices of issuing shares throughout the 15 year old period. The total pay for directors and CEO also show a large
Variables Observations Mean SD Minimum Maximum
ROA 900 4.74 3.48 -1.10 11.63
Log of Sales 900 16.70 0.23 16.26 16.99 Tobin's Q 900 1.67 0.40 1.16 2.71 Volatility 900 0.19 0.27 0.05 1.94 Fraction of Females 900 0.30 0.12 0.00 0.60 Fraction of Independent Directors 900 0.24 0.09 0.10 0.30 Board Size 900 6.00 0.06 5.00 7.00 Number of Business Segments 900 4.81 1.39 2.00 13.00 Total Directors' number of shares 900 6294 2114 2053 8750 Total CEO's number of shares 900 18868 6360 6158 25989 Total Pay Directors 900 4632693 1011670 2888924 5961502 Total Pay CEO 900 2164214 582135 1100505 2970400 Equity Based Pay Directors 900 340895 124416 122278 495067 Equity Based Pay CEO 900 1022684 373247 366835 1485200 Fraction Equity Based Pay Directors 900 0.30 0.05 0.15 0.33 Fraction Equity Based Pay CEO 900 0.50 0.07 0.33 0.60 CEO age 900 47.00 1.90 30.00 62.00 CEO tenure 900 3.42 0.78 2.00 6.00
* Total Directors' number of shares is the total amount of shares issued for all directors within the whole board *Total CEO's number of shares is the total amount of shares issued to the CEO
*Total Pay Directors is the total amount of pay in Norwegian Krone (NOK) including fixed salary, equity and other benefits (pension plan) *Total Pay CEO is the total amount of pay including fixed salary, equity and other benefits (pension plan)
*Equity Based Pay Directors is the value of total shares issued to the board of directors in Norwegian Krone (NOK) *Equity Based Pay CEO is the value of total shares issued to the CEO in Norwegian Krone (NOK)
34 range with a high standard deviation from the mean. Thus, the total pay and total shares issued to whole board and CEO have experienced some volatility. The mean for fraction of equity based pay for the whole board is 0.30 with a range from 0.15 to 0.33. Fraction of equity based pay for the CEO has a mean of 0.50 with a range from 0.33 to 0.60. These statistics show that there was a movement in the value of equity based pay in relation to total pay for both the whole board and the CEO. To comment on volatility the mean is at 19% with a standard deviation of 27% whilst the minimum has a value of 5% and maximum of 194 %. Hence, indicating that there were some bigger changes that impacted volatility of stocks throughout the time period.
Figure 1: Total Directors’ Number of Shares vs Total CEO’s number of shares
Figure 1 depicts that the total amount of shares issued to the whole board of directors and CEO have been increasing since 2002. This could be because of the IFRS-2 accounting regulation on equity issues that was in discussion in late 2002, which created greater transparency in equity
2000 3000 4000 5000 6000 7000 8000 9000 T o ta l D ire c to rs ' N u m b e r o f S h a re s 6000 8000 10000 12000 14000 16000 18000 20000 22000 24000 26000 T o ta l C EO 's N u m b e r o f Sh a re s 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 Year
35 based pay. It is also the period under which politicians already were deciding on whether they should implement a quota system on Norwegian boards of public companies. There is a sharp increase in number of shares issued to both the whole board and CEO following 2003 the voluntary compliance of the regulation, 2006 when the regulation was made mandatory and 2008 where companies had to make full compliance of the regulation. Following up to 2008 where global recession has impacted the companies and the number of shares issued to both directors and CEO’s decreases in the following years. Figure 1 complements the study of Ahern & Dittmar (2012) and Adams & Ferreira (2009) which also document an increase in equity based pay with more women on boards.
Figure 2: Total Pay Directors vs Total Pay CEO
Figure 2 depicts the total pay to whole board of directors and the CEO; it shows an increasing trend of total pay to both CEO and directors from the period 2000 to the period 2008. This
3 .0 e + 0 6 3 .5 e + 0 6 4 .0 e + 0 6 4 .5 e + 0 6 5 .0 e + 0 6 5 .5 e + 0 6 6 .0 e + 0 6 T o ta l Pa y D ire c to rs 1 .0 e + 0 6 1 .5 e + 0 6 2 .0 e + 0 6 2 .5 e + 0 6 3 .0 e + 0 6 T o ta l Pa y C EO 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 Year
36 complements the study of Adams & Ferreira (2009) that saw a similar phenomenon when testing whether more women on boards facilitate higher total pay for CEO’s. This also supports Prendergast (2002b), who predicted that total pay should also increase with equity based pay due to the compensation for extra risk of uncertainty.
Figure 3: Cross Correlation Matrix
Figure 3 presents the cross correlation matrix of all independent and dependent variables. The figure depicts some interesting results. Volatility is quite strongly correlated with total directors’ number of shares, total CEO’s number of shares, total pay directors, total pay CEO fraction of females, equity based pay for both directors and CEO and fraction of equity based pay for both directors and CEO. This finding is shown also in Adams & Ferreira (2009) paper, where it is explained that more females on boards create volatility through which incentive pay increases as well to monitor and manage the pay scheme more efficiently. Total Directors’ number of shares is closely correlated with total CEO’s number of shares, total pay directors, and total pay CEO, ROA, Tobin’s Q, fraction of females, fraction of independent directors, equity based pay for both directors and CEO, fraction of equity based pay for both directors and CEO and log of sales. Similar results are present for total CEO’s number of shares and total CEO’s number of shares;
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 1. Volatility 1.00
2. Board Size 0.01 1.00 3. Total Directors' number of shares 0.92 0.03 1.00 4. Total CEO's number of shares 0.89 0.01 0.69 1.00 5. Total Pay Directors 0.80 0.16 0.88 0.87 1.00 6. Total Pay CEO -0.57 0.15 0.92 0.91 0.94 1.00 7. ROA 0.23 0.03 0.50 0.49 0.22 0.27 1.00 8. Tobin's Q 0.60 0.01 0.84 0.11 0.21 0.16 0.49 1.00 9. Fraction of Females 0.77 0.02 0.93 0.92 0.94 -0.45 0.26 0.03 1.00 10. Number of Business Segments 0.10 0.02 0.35 0.13 -0.28 -0.29 0.51 0.05 0.30 1.00 11. Fraction of Independent Directors 0.23 0.01 0.92 0.60 0.89 -0.55 0.51 0.22 0.98 0.16 1.00 12. Log of Sales 0.21 0.03 0.67 0.68 0.77 0.75 0.26 0.01 0.96 0.20 0.90 1.00 13. Equity Based Pay Directors 0.90 0.05 0.99 0.83 0.95 0.81 0.48 0.80 0.90 0.23 0.77 0.78 1.00 14. Equity Based Pay CEO 0.95 0.03 0.89 0.95 0.86 0.96 0.51 0.12 0.91 0.18 0.76 0.73 0.83 1.00 15. Fraction Equity Based Pay Directors 0.89 0.04 0.98 0.81 0.98 0.82 0.54 0.80 0.89 0.14 0.79 0.69 0.98 0.80 1.00 16. Fraction Equity Based Pay CEO 0.93 0.04 0.85 0.93 0.81 0.96 0.44 0.14 0.92 0.15 0.80 0.65 0.90 0.99 0.81 1.00 17. CEO tenure 0.91 0.02 0.75 -0.19 0.55 0.17 0.49 0.18 0.45 0.44 0.67 0.60 0.13 -0.20 0.22 -0.23 1.00 18. CEO age 0.01 0.01 0.68 0.18 0.40 -0.40 0.75 0.33 0.38 0.09 0.62 0.45 0.11 0.30 0.18 0.62 0.68 1.00
37 however it is interesting that it is less correlated with Tobin’s Q and negatively correlated with fraction of independent directors. This complements the study of Ahern & Dittmar (2012) that independent directors bring better monitoring and thus restrict the amount of pay for performance for the CEO. Fraction of females is highly correlated with total directors’ number of shares, total CEO’s number of shares, total pay directors, total pay CEO, equity based pay for both directors and CEO, fraction of equity based pay for both directors, CEO and fraction of independent directors. This complements the study of Adams & Ferreira (2009) which made a similar analysis in their paper. It is important to note that fraction of females is highly correlated with the fraction of independent directors, suggesting that in many cases females on boards in Norway were independent directors as also documented in Seierstad & Opsahl (2010) research. The total Pay for both CEO and directors indicate a positive correlation with fraction of females on boards, fraction of independent directors, equity based pay for both directors and CEO, fraction of equity based pay for both directors and CEO, and log of sales as documented in Adams & Ferreira (2009) paper, where they conducted a very similar empirical analysis investigating the relationship between more females on boards and total pay for both CEO and board of directors.
7. Results
The results presented are OLS regressions performed on STATA software. Total pay of both directors and CEO’s are in Norwegian Krone (NOK) and the number of stocks issued is the actual amount of stocks issued. All regressions had both year dummies, firm dummies, and
38 industry dummies included. All coefficient result have their t-statistics in brackets and an asterisk shows the level of their significance (*10%, **5%, ***1%).
Figure 4: OLS Regression Testing Directors’ Compensation
Figure 4 represents the OLS regressions that aim to test Hypothesis 1 and Hypothesis 2. As we can see from the regression coefficients, having more females on boards increases the fraction of equity based pay to directors as a form of incentive pay and also increases the total pay of directors. This has also been the case in the study performed by Adams & Ferreira (2009) where they found the same effect when they tested gender diversity of boards and its impact on directors’ equity based pay. Volatility also has positive and significant coefficient results both on incentive equity based pay and total pay of directors, supporting the study made by Prendergast (2002b). He showed that with more volatility incentive pay in the form of equity tends to
Independent Variables Dependent Variables Fraction of Equity
Based Pay Directors' Total Pay
ROA 0.0018* 18687.90*** [6.30] [11.91] Log of Sales 0.0555** 212320*** [7.77] [78.96] Tobin's Q 0.0134** 115845.90** [4.12] [6.78] Volatility 0.2420*** 173265.21** [7.7] [3.4] Fraction of Females 0.1174*** 199555** [5.18] [16.85]
Fraction of Independent Directors 0.1588** 77893.20
[5.29] [12.37]
Board Size 0.0049 19857.78
[0.41] [0.32]
Number of Business Segments 0.0034 -2896.07
[4.12] [6.43]
*Total Directors' number of shares is the total amount of shares issued for all directors within the whole board
*Total Pay Directors is the total amount of pay to the board in Norwegian Krone (NOK) including fixed salary, equity and other benefits *Equity Based Pay Directors is the value of total shares issued to the board of directors in Norwegian Krone (NOK)
39 increase in order to form an efficient pay structure based on outcomes instead of inputs in uncertain circumstances. Fraction of independent directors also has a positive and significant effect on the fraction of equity based pay and the total pay for directors. This can be due to the fact presented by Seierstad & Opsahl (2010), which found evidence that woman on boards in Norway, were mainly appointed as independent directors. Thus, due to more uncertainty in hiring talent that would be experienced enough for the boards, equity and total pay have increased to compensate for the risk and to create an efficient incentive pay scheme.
The regression output presented in Figure 4 shows a positive relationship between ROA, board size and fraction of equity based pay. The results presented in Figure 4 suggest, that with an increased board size there is more equity based pay to directors. This can be because of the necessity of compensating an extra member on the board. ROA has a positive and significant coefficient as an increase in the earning quality in assets would imply more capabilities of a firm to be able to compensate directors more. The regression results support hypothesis 1 and 2 and show a positive significant effect of having more women on boards and total pay and equity based pay. Thus, supporting the main research of the thesis and being in line with prior literature and empirical results.
40 Figure 5: OLS Regression Testing CEO’s Compensation
Figure 5 shows the linear regression results testing Hypothesis 3 and 4. The results presented depict a positive and significant relationship between fraction of females on board and CEO equity based pay and total pay. It is interesting to see that an increase in volatility increases the fraction of equity based pay to the CEO but decreases the CEO’s total pay. This can be explained by the Ahern & Dittmar (2012) paper that finds a positive relationship between more females on boards and increased monitoring on boards in Norway. They also find that with increased
Independent Variables Dependent Variables Fraction of
Equity Based Pay
CEO's Total Pay
ROA 0.0059*** 83555.13** [7.32] [4.27] Log of Sales 0.1832** 153789*** [13.36] [27.44] Tobin's Q 0.0444 297165.1** [16.21] [13.22] Volatility 0.3101*** -1255.765 [0.30] [0.10] Fraction of Females 0.2963*** -25230** [21.10] [20.43]
Fraction of Independent Directors 0.3548** -147417**
[20.75] [8.05]
Board Size 0.0045 19857.78
[0.92] [7.34]
Number of Business Segments 0.0122 -72369.87
[21.10] [10.97]
CEO age 0.0048 -53574.37**
[11.34] [12.21]
CEO tenure -0.0306* 67795.24**
[21.66] [6.01]
*Total CEO's number of shares is the total amount of shares issued to the CEO
*Total Pay CEO is the total amount of pay including fixed salary, equity and other benefits (pension plan) in Norwegian Krone (NOK) *Equity Based Pay CEO is the value of total shares issued to the CEO in Norwegian Krone (NOK)
