Female board members in competitive environments

Female Board Members in Competitive Environments

Abstract

In 2012 the Dutch government proposed an advised quota of 30% female board members for Dutch stock listed companies to be reached in 2015. Using the female board index and annual EBITDA’s of Dutch stock listed companies this paper tries to reflect upon the impact of that decision and finds out that actually the opposites holds true. This paper also compose different competitive environments as a possible explanations for less female participation in board using the FSQCA method together with the average legislative period of board members and their annual average awarded bonuses.

C.A.F. Lammers 10019464

19/08/2016 Carsten Gelhard Master Thesis

Statement of Originality

This document is written by Student Coen Lammers 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.

For Nadine & Silvia,

Proving the point that men would be totally lost without proper female guidance

Table of conduct

1. Introduction 5

2. Theoretical framework 7

2.1 The Glass Ceiling effect 7

2.2 Strategic Human Resource Management 8 2.3 The advantages of a diverse workforce 10

2.4 The Quota 12

2.5 Maternity leave and Evaluation forms 12

2.6 Competitive Environments 14 2.7 Hypotheses 16 3. Methodology 18 3.1 Sample 18 3.2 EBITDA 20 3.3 Competition 21 4. Results 26 4.1 Sample 26 4.2 Reliability 27

4.3 Student’t-test and OLS regression 28

4.4 Testing the hypotheses 30

5. Discussion

5.1 Interpreting results 34

5.2 Contribution to existing literature 39

5.3 Managerial implication 40

5.4 Limitations and Future research 41

6. Conclusion 42

7. References 43

1.Introduction

“If you judge a fish by his ability how to climb a tree he will always appear to be stupid” It is a common English saying, that teaches the importance to respect people who are different, because they might have different skillsets, which should and cannot be compared to each other. The Dutch have a likeminded sentence, which is commonly used by high school teachers to highlight the difference in the abilities of certain students. This sentence is as followed: “boys are better in math than girls, while girls are mostly better in languages than boys”.

Though it is true that people have different skillsets, this saying does not hold ground in Dutch society nowadays. Though boys still are more interested in mathematics and science (Buser et al., 2014), female students excel in every course compared to their male counterparts (except for gymnastics) in high school (CBS, 2016) and even the percentage of women that graduate from universities each year is far beyond the amount of men for the last couple decades (Reuben et al, 2013). However, be that as it may, this transaction has not yet been made to businesses where the male/female ratio in most business environments is still leaning heavily to the masculine side (Management Scope, 2015).

The underrepresentation of females in management boards or the underpayment of females while having basically the same job description has been a well investigated problem of the last four decades in management research. This phenomena is commonly described as “the Glass Ceiling effect” (for instance in Lyness and Thompson, 1997) as there seems to be a hidden obstacle that prevents women of entering management boards on high firms all over the world.

While before it seemed that it was mostly an image problem of countries and companies that somehow prevented women from entering the highest paying jobs, or the jobs with the most status (Lyness and Thompson, 1997; Adams & Fereira, 2007) and which was associated mostly with gender discrimination, recent investigation suggest that there might be evidence

that companies also would perform better if companies were to appoint more females to their boards (Shen et al, 2009; Hoogendoorn, 2011)

Proof of this was for instance found when the Norwegian government obliged Norwegian stock market listed companies to have a male/female ratio of at least 0.4 in there management board and supervisory boards combined in 2003. When this ‘quota’ (as the obliged male/female ratio is more commonly known) was introduced, not only did the logical effect happen that more females were appointed to management functions to avoid harsh penalties, but after accounting for their years of experience Norwegian stock market companies seemed to perform better than before the quota was introduced (Sol, 2011)

Following these results the Dutch government declared to its own stock listed market companies at the end of 2012 that they should strive for a quota of 30% females board members in every company in 2015 (Dekker, 2015). The Dutch quota did differ from the Norwegian as it was an advised quota instead of a obligatory rule. Since the year 2015 is officially over and the annual results of all Dutch stock market listed companies have been released, it seems fitting for this paper to reflect on how this verdict of the Dutch government has been integrated in the Dutch stock market and if there is evidence that companies are indeed performing better than they were in 2012.

The structure of this paper will be as followed. Chapter 2 will give a brief overview of the main topics in the glass ceiling discussion and will focus more on the quota of 2015 and its presumed effect on company performance. Chapter 2 will also introduce a possible factor why so less females can be found in the management board of stock listed companies namely: the competitive environment. Chapter 3 will explain the different measures used in the research to give a clear insight on the results in chapter 4. These results will be reflected upon and explained in chapter 5 together with managerial implications and proposed future research.

2. Theoretical Framework

This chapter will provide a literature review of the topics that are most associated with the lack of female directors in top management positions. The differences between male and female employees have been fiercely debated the last four decades and many research has been done in different areas to solve pieces of the puzzle. This thesis also only tries to solve a piece of the puzzle and it is therefore of the uttermost importance to understand which part exactly. The other issues of the gender inequality in business will be touched briefly upon, but the main subjects of chapter 2 will be focused on the seemingly inability of female employees to reach top management positions. This phenomena will be explained through the concept of the glass ceiling, strategic human resource management and competition.

2.1 The Glass Ceiling effect

As is described above there are many issues regarding the seemingly inequality between men and female. However, as Cotter (2001) states it: “Not all gender or racial inequalities need to be defined as glass ceilings”. The glass ceiling has been a favourite colourful phrase that describes what we already mean by gender or racial inequality (Cotter, 2001) and is used by different academics for different phenomena. Albrecht (2003) for instance uses the glass ceiling metaphor not only for the occupational segregation between men and women, but also uses the term freely when talking about women being systematically paid less than men while they are in the same job occupations (Albrecht, 2003). Though this paper is under the

impression that equal payment of male and female employees is an important issue, a whole thesis could be addressed to this issue alone and therefor it is not included in this paper.

Since the glass ceiling is a metaphor, a lot of different explanations can be given who all slightly differ from each other. This paper will use the description of Baxter & Wright (2000) as its main explanation of the glass ceiling being “the disappropriate amount of

females in top management”. A phenomenon whereby women do quite well in the labour market up to a point where there seems to be an effective limit on their prospects (Baxter and Wight, 2000; Albrecht, 2003)

In their paper Baxter & Wright (2000) find evidence that women experience more difficulties than men to be promoted in the hierarchical ladder, but that these obstacles become relatively bigger the higher levels they reach. A phenomenon that is also reported in the paper of Lyness & Thompson (1997) who also find evidence that higher executive female employees report more obstacles than at the lower level.

Baxter & Wright (2000) test this inequality also in three different countries in the world (the United States of America, Sweden and Australia) and find supporting evidence for this in both Sweden and Australia (the results of America were not significant enough to draw conclusions). This shows that the glass ceiling problem is not only a West European, but is rooted all over the world and although yearly incremental progression is made (Meyerson & Fletcher, 2000) the report of for instance Lückerath-Rover (2015) show that there still a long way to go.

However to understand why the Dutch government and companies aim to have more female board members, one must first understand why it might be important to have females in the board of directors. This can be explained through the theories of strategic human resources management.

2.2 Strategic Human Resource Management

The fist papers to link strategic human resource management to the glass ceiling effect were the papers of Erhart et al. and Dreher, both published in 2003. These papers explain that strategic human resource management can not only help to understand the social phenomena of the glass ceiling, but also why it is important to break through this ceiling (metaphorically speaking).

The last couple of decades human resources management has become more and more important for companies. Human resources management has changed from being only a utility to satisfying regulatory agencies (Perry-Smit, 2000) to ‘the’ means of gaining competitive advantage (Pfeffer, 1994). This advantage can be achieved by integrating strategy into human resources management creating the term: strategic human resource management, which was first introduced by Beer et al. (1994), which states that by using your human resources effectively you can gain a sustainable competitive advantage.

A lot of authors have investigated strategic human resource management ever since, searching for benefits of integrating role behaviour of human resource management and gaining competitive advantage through strategy. While some of these researches viewed strategic human resources management as a combination of an external source (strategy) and an internal source (human resources management) (Arthur,1994; Mcduffy, 1996) others argued that the most significant competitive advantage can come from effectively utilizing your human resources (Quest, 1987). Cappelli and Singh (1992) support the idea of Quest as they say that “competitive advantage arises from firm-specific valuable human resources assets, which are difficult to imitate” (Cappelli and Singh, 1992, p.186)

So what are these inimitable human resource assets? Shen et al. (2009), together with (Dreher, 2003 and Erhart and al., 2003) argue that it are people and, to be more specific, diverse people. Diversity leads to improved quality of management decisions and can raise companies overall effectiveness, efficiency and profitability. Though Van der Walt & Ingley (2003) describe that diversity can manifest itself in many way such as religion, sexual orientation, social status or ethnicity, the most common and oldest diversity issue still remains to this day gender (Shen et al, 2009).

2.3 The advantages of a diverse workforce

Shen et al. (2009) argue that integrating diversity in your workforce can lead to a competitive advantage and in this they are not alone. It is important to highlight the importance of this competitive advantage because as Brown et al. (2002) state: “if a diverse workforce does not lead to improved performance, then the question of who sits in the board of the company or how this board operates has no practical value. The appointment of women in the board than only has a symbolic value (Brown et al., 2002)

Hoogendoorn et al. (2011) show that when teams are composed of both men and women they perform better, make better decisions and have higher profit, then the teams that are composed of only men or only women. These heterogeneous team as they are called, show that a mixed workforce leads to overall better results compared to the homogeneous teams (teams composed of only men or only women). This same phenomena is seen in the results of the paper of Apesteguia (2012), were groups composed of only women have an overall lower score than teams of only men, but that when the teams are mixed these scores are higher than the homogenous teams. The best scoring teams are those composed of 2 men and 1 female, a ratio of 0.33% which is remarkably close to the quota the Dutch government is aiming for.

Adams and Fereira (2007) explain why heterogeneous groups perform better in their paper because men and women “keep each other in check” through different working methods. Although they argue that this effect is mainly in companies with lesser shareholder rights.

This let to recent research from for instance Van Esch and de Groot (2015) who compared high gender diverse groups with low gender diverse groups which had comparable, equally weighted portfolio’s. Though the results were not significant, a positive connection was shown between portfolios of companies with higher board diversity who outperformed the portfolio’s with lower board diversity when looking at stock prices. (Van Esch & de Groot, 2015).

Lückerath-Rover (2011) also shares these data and adds to this that companies with no female board members, can also signal that a company is not willingly to invest in choosing the best candidates for a job (through for instance a lack of good evaluation) and therefor those companies who do are rewarded with better results (more on this in chapter 2.4).

This statement can be supported with the facts that were given in chapter 1 that at high school and at the University the amount of females and their accomplishments (grades, extracurricular activities) far exceeds the amount of male graduates for the last decades (e.g. Reuben, 2013; CBS, 2016). So it is clear that it is not a lack of skill or intelligence, but still females are rarely found in the board of executives, as only 7 of the 100 most influential Dutch managers are feline (Management Scope, 2014).

Given the evidence above it is clear why the Dutch Government is so focused on promoting more female managers in the top management as there are clear indications in the theory that companies who have more females board members have overall better results. However not all research is unified on this. Mainly on the short term company results sometime drop. Ryan and Haslam (2005) for instance found that appointing a women to the board instead of a men had negative consequences for a company on the short run. The performance of companies who appointed a women were significantly lower during the first five months in comparison to companies that appointed a man (Ryan and Haslam, 2005).

Pande & Ford (2011 also agree on this saying that evidence of introducing a quota in a board is more mixed. They find that female entry is mostly correlated with the practices of how management solves problems, but that this change negatively influences short term profit. Though research seems to be divided on this issue regarding the short term over the long term it seems they are quite harmonised.

2.4 The quota

Following the evidence Hoogendoorn et al. (2011), Apesteguia (2012) and Van Esch & De Groot (2015) it is understandable that the Dutch government suggest a quota that encourages companies to have at least 30% of their management board be females and that it was not created out of thin air. The Dutch government followed by example with this quota in 2012, copying a very effective approach that Norway introduced to its stock market listed companies in 2003 (Dekker, 2015). The Norwegian government made the mandatory rule that companies who were enlisted at the Norwegian stock market had to have a male/female ratio of at least 40% in their boards.

This decision did not only succeed in getting more females in top management positions, but when taking in account the years of experience these females had and their overall age compared to men, the female board members performed better than their male counterparts. This is also in line with Reuben’s (2013) observation that women perform overall better than men when they are in the same age category (students in this research).

However not all introductions of quota were successful as the introduction of a quota in India actually decreased the short term profit of companies (Pande & Ford, 2011) and when the quota was removed so did the female directors, indicating that a quota does not only put women at the top but also keeps them there, which is a good thing according to Hoogendoorn et al (2011) or Apesteguia (2012)

2.5 Maternity leave and Evaluation forms

Although overwhelming amounts of research point out that there is a good possibility that appointing more female board members might lead to higher performance of a company this has not yet been translated to the workplace as of 2015. Females are still underrepresented in the top level of almost every company in the world (Lückerath-Rovers, 2015; Management

Scope 2014) and the problem might not be with the indifference of companies to try and find suitable female employees, but maybe the lack of choice.

Together with the glass ceiling problem, a lot of research has been done to try and find the crux of the problem why female do not reach top management functions, because when an obstacle is defined, companies can start working towards a solution. Issues that are proposed are for instance that evaluations forms, to decide if one would be a good manager, which might be unintentionally biased because they would advocate character traits that a more commonly associated with men like being assertive and risk takers (Nieva, 1980; Sol, 2011). Also men tend to boast more about their performances then females do overall and therefor might get more attention for achievements that are equal to those of a less upfront females (Reuben, 2013).

Another possible reason for the low amount of female directors that comes up is the work/family balance (Lyness & Tompson, 1997). Females tend to have more difficulties regaining the same authority and position after they leave for a certain period, because of for instance maternity leave. (Lyness & Tompson, 1997; Dreher, 2003) This obstacle is mostly felt in low and middle management, but could explain why lesser female middle managers are promoted to top management.

Both these issues are very interesting and have some valid points, however in the time in which this paper had to be written, it was not possible to find multiple companies that wanted to release their evaluations forms from which conclusions could be drawn. The maternity leave issue idem ditto is a very good possibility. However since this paper is specifically looking at Dutch stock market listed companies who all have the same maternity leave, as is stated in the Dutch law, this paper had no different countries to compare to. For the hypotheses the results of the companies need to be very comparable to give a reliable outcome (more on this in chapter 3.1). This can only be done if all companies are from the same country. Therefore this paper

chose not to include the maternity leave discussion since it would temper with the reliability and comparability of the results of stock listed companies if other countries were included A last possible explanation for the lack of female directors was hinted at by Reuben et al. (2013). Reuben et al. (2013) showed that upcoming female students tend to choose different educational courses than the male students. If men and women choose different subjects then it is not surprisingly to assume that they also end up in different professions. However a possible reason why females might choose different subjects might have something to do with them being able to flourish and excel in different professions then most men strive to. This led this paper to the following possible moderator that has effect on female employment in top management: Competitive environments.

2.6 Competitive Environments.

It has been hypothesized before that gender differences in both labour market and education might be explained through the amount of competitiveness men and women prefer in their jobs (Bertrand, 2011; Buser, 2012). Bertrand et al. (2014) find that having a prestige and competitive element results in the academic tracks of math and science can account for 20% of the gender gap. Buser et al. (2012) are also intrigued why women shy away while having better grades (see chapter 2.3), but explain it through the choice of profiles in high school. Despite higher grades and equal capabilities in math girls choose less prestigious profiles than boys which the authors links to the avoidance of competition. (although in this study in only account for 15% compared to Bertrand et al. 20%)

An indication that this might be also the case in work environments can be found in the research of Gneezy et al., (2013). Men tend to be twice as competitive as women in patriarchal societies (which is an accurate description of the work environment as it is predominantly dominated by men) and excel in these kind of environments. Niederle & Vesterlund (2007) also

find evidence for this. In their research they introduce a competitive element to a group who previously had no incentive to compete against each other. In this investigation a tournament like setting is created were men and women compete to get a bonus awarded to them by completing certain task. First they have to solve tasks and getting paid just for completing the task. Women and men both perform equally well. Then a competitive tournaments scene is created in which participants compete to get a good output, but also against each other (Niederle & Vesterlund, 2007). Although again no evidence is found that one gender performs better than the other, Niederle & Vesterlund (2007) find that about 75% of the men prefer the tournament like payment, while compared to their feminine counterparts, only 35% of the women choose this option as the preferred one. In this they are not alone List (2010) has the same findings that men significantly respond more to jobs with relative performance.

Part of this is explained by Beckman & Menkloff (1998) & Faccio et al. (2014). In Beckmans and Menkloff’s research they find that female fund managers from for instance the United States of America and Germany shy away from competition, because they are more risk averse, which is in line with other gender related studies. Faccio, Marchica & Mura (2014) find evidence for this in their paper in the form of female CEO’s having lower leverage, less volatile earnings and higher chances of survival compared to otherwise similar company who are run by a male CEO.

Following this logic it could be highly possible that females might enjoy a less competitive workplace than men and that this could be translated in the amount of applications for jobs in these areas or interest in management positions. This possibility was first introduced by Lückerath-Rover (2010) who divided the stock listed companies of the Netherlands in different business environments. The industries were as followed: oil & gas, basic materials, industrial, consumer goods, healthcare, consumer service, telecom, financials and technology. She found that just as her previous research not a single industry was able to reach the quota of

30% (although in all fairness the telecom companies together managed to almost reach it with 27,3%) and that especially healthcare and oil & gasses scored astoundingly low (below 5% female directors).

The business environment however is mostly looking at competitions from an external perspective, as it is focused on the field of work a company is in, instead of looking at the company itself. Although Lückerath-Rovers (2010) showed that there is a difference in the amount of appointments in different industries, companies themselves could greatly differ in the amount of internal competition in the company itself among co-workers. Because List (2010) describes competitive environment as an competitive climb up a company its economic ladder (which is internal), this paper will look at internal competition.

However, as can be read in Bertrand (2011) paper, evidence is difficult to find since this is a rather new research agenda and finding reliable, good measures of competitiveness and their outcomes are difficult to construct let alone to find. This paper will however try to construct the measures for a competitive environment from the key components described above (risk taking and internal competition) to form a competitive environment, which can be read in chapter 3.3

2.7 Hypothesises

Following from the six sub chapters above. This paper will focus on three different subjects regarding the glass ceiling effect. It will investigate if a balanced board of directors

(consisting of both females and male) performs better then heterogeneous boards as is assumed in multiple paper such as Hoogendoorn et al. (2011) and Apesteguia (2012). Therefore the first hypothesis will be

H1: Companies with female board members perform better then companies without female board members

Secondly this paper will investigate if it is possible to conclude that an increase in the amount of female board members means that there is also an increase in the overall results of a company. This hypothesis differs from H1, since the second hypothesis does not look at the performance of companies that have female board members in a given year, but tries to find if there is evidence that the Dutch stock market listed companies that appointed female directors perform better after the introduction of the Dutch quota in 2012 than those who have not. If this is true than it would suggest that the quota is not only effective to solve the image problem of the glass ceiling, but also efficient for company results as an increase to more female employees would be correlated to an increase in company performance. Therefor the second hypotheses will be:

H2: an increase in female board members leads to higher company performance

The last and final hypothesis of this paper will investigate if the lack of female mangers in the management board can be (partly) explained through the amount of internal competitiveness in a company and if there is a difference in the amount of female employees among different competitive environments. Following the paper of for instance List (2010), Niederle & Vestlund (2007) and Beckman & Menkloff (1998) this is most likely the case and thus the third hypothesis will be:

3. Methodology 3.1 Sample

3.1.1. Selecting The Companies

The Norwegian quota of 40% feminine board members was obliged for all the stock market listed companies in Norway in 2003. The Netherlands adopted the same rules in 2012 although with a bit lower male/female ratio of 30%. For this reason this paper will look only at companies who have been listed as Dutch stock market listed companies since the quota is only applied to these companies (as was also explained in chapter 2.5).

Another reason to only use companies from the Dutch stock market is because the same rules apply to all these companies. Rules and regulations about for instance bonus structurers are different for every country and can make up a difference of around 33% (Lückerath-Rovers, 2015) (more about this in Chapter 3.3 competitive environments). A comparison between different companies will be the fairest if all companies are compelled to follow the same rules, providing more reliable results.

The last reason to only use Dutch stock market listed companies is because their annual reports have to be transparent for all Dutch citizens and have to be deposited around April. Thereby presenting the most contemporary and reliable results, since they are obtained from their original source, fact checked and composed by well-known accountancy firms.

For the sample, this paper first looked at all the companies who were listed in as Dutch stock market companies in 2012, since this was the year that the quota was introduced in the Netherlands. In 2012 the amount of companies listed at the Dutch stock market was 96.

The aim of the Dutch quota was to have 30% females managers in 2015. Therefore, when looking at results, this paper will compare the 2012 stock listed companies with those that are Dutch stock market listed in 2015. Another reason for this is that the economic crisis had a major effect on most company results in 2013. Making a comparison with this year would be

unfair, because most company results were worse than 2012 though this was mainly due to an external reason. The amount of Dutch stock market listed companies in 2015 was 84 companies.

However not all these 84 companies were already in the Dutch stock market in 2012. Just as the stock market is fluctuating so are the companies who are in it. A couple of the companies who were Dutch stock listed in 2012 had sincere losses due to the economic crisis. This resulted in the bankruptcy of certain companies. Other companies just moved their business to another country or left the Dutch stock market for different reasons (DE masterblenders). Similar to the stock listed companies in 2012, the 2015 stock listed companies also did not all already appear on the Dutch stock market in 2012. Either being brand new entries such as Flowtraders, or coming over from another country and only recently registering as Dutch stock market listed company in the years between 2012-2014.

Since eventually a comparison has to be made between the results of a company in 2012 and 2015 (hypothesis 2) this paper excluded all companies who did not appear on both the 2012 and the 2015. This resulted in a sample size of 69 which can be found in appendix A.

3.1.2. Board Members

The Netherlands work with a two-tier board model. This means that Dutch stock market companies have two separate functioning boards which are called the supervisory board and the executive board (Lückerath-Rover, 2007) This paper will look at the two boards combined as one, because the Dutch quota does not have different ratio for both supervisory board members (non-executive directors) and the executive board (executive directors) (Lückerath-Rovers, 2015) and thus looks at the jointly when determining if they uphold the 30% feminine quota. To calculate the board members this paper used the Dutch Female Board Index of both 2012 and 2015 and calculated the amount of board members, the amount of female board members and the male/female ratio of both 2012 and 2015. The abbreviation of the Female

Board Index will not be used because this paper wants to prevent any confusion and suggestion that it got its results from the American Federal Bureau of Investment which is more commonly associated with the FBI abbreviation. The results of the female board index can be found in appendix A.

3.2 EBITDA

To calculate the performance of Dutch stock market listed companies a lot of different measures have been used in the last decade. Van Esch & de Groot (2015) look at for instance stock price and return on equity. Lückerath-Rover (2011) also uses return on equity as well as EBIT and return of stock. However return on stock and return on equity are two measures that measure profitability. Though profitability is an indication of performance it is not possible to say they are completely alike. Return on equity is the net income of a company divided by the shareholders equity. Although dividend is paid to the common stockholders after net income is calculated and thereby does not influence net income, the same cannot be said about preferred stock which is paid before net income. Mrs. Lückerath-Rover (2011) states in her article that there is a difference in the paid-out dividends of companies with female directors compared to male directors. By not paying out dividends the same, the return on equity can be influenced and therefor it is not only a difference in attitude, but also a difference in outcome performance since handing out lesser stock by female directors leads to better return on equity.

A much fairer comparison which is briefly touched upon in the article of Lückerath-Rover (2011) in a simplified matter (EBIT) is the EBITDA. EBITDA is short for “earnings before interest taxes depreciation and amortization”. The reason these numbers are chosen instead of just profit is because EBITDA’s give a better insight in the performance of a company. Profit can be highly influenced by a company through for instance having high depreciation value in a certain year. This not only effects the amount of taxes a company has to

pay in a certain year, but could also influence how profitable a company is perceived that year. Since this paper is in not necessary interested in the profitability of a company, but is mainly focused on how Dutch stock market listed companies perform, the EBITDA’s give a much more realistic view of a company its performance in given years. From personal experience the author of this paper also knows that EBITDA’s are also most commonly used to valuate companies when composing the annual Forbes ‘100 or the Quote 500 list.

The EBITDA’s have been collected from the annual year reports of companies themselves from 2011 to 2015. Though most companies disclose EBITDA there is no legal requirement to do so according to the general accepted accounting principle (McClure, 2014). In this few cases the EBITDA was obtained from the information in the financial statements of a company itself (operating profit + depreciation of that year). In two occasions the annual results of 2015 were not yet deposited. These companies (Bever and Nesdap) were removed from the Dutch stock market listed companies this paper investigates, since these results are essential for calculating the performance of a company for this research. This left 67 companies remaining for the final sample.

3.3. Competition

Until now this paper has only used quantitative data to calculate its it measures. Both the EBITDA’s and the amount of female board members in Dutch stock listed markets are hard number. However to determine the competitiveness of a business environment a combination of qualitative and quantitative research is used. As is explained in chapter 2.6, the amount of competitions and risk aversion are key components of a competitive work environment, but real measures do not yet exist. To calculate the competitive environment this paper uses a couple of tools to arrive at its conclusion. Two of these components are the average legislative period of board members and the percentage of the bonus that is awarded to top management and will

be more elaborated in the section bellow. However, to determine how competitiveness certain business environments are, this paper will also use qualitative categorisation in the form of FSQCA to divide the Dutch stock market listed companies in different groups. This will be explained chapter 3.3.3.

3.3.1 Average legislative period

The average legislative period gives an indication of how competitive the work environment is, where quick switches of board members indicate a more dynamic environment then long legislative periods. The average legislative period of board members is calculated from the annual year reports from the Dutch stock market listed companies. In these reports it is stated when board members were appointed. Board members leave a company for three reasons, due to resignation, end of terms (which is clearly stated in every annual report) and unfortunate and untimely demise. The amount of years the board members worked for their respective company is than divided by the amount of board members to calculate the average legislative period of board members. The average legislative period can be seen in Appendix C.

3.3.2 Bonuses

Niederle & Vesterlund (2007) partly explained the under representation of women in high pay jobs due to competitive payment. Price (2009) replicated this investigation and although he did not come to the exact same conclusion he did replicate the results that men and women prefer different compensation schemes. In business it is possible to see the fluctuating compensation in the form of awarded amount of bonuses compared to the regular salary. Bonuses can indicate how fluctuating earnings can be. It can therefore be associate with risk aversion, because a bad decision can lead to lower earnings if your salary is composed of a high amount of bonus which

some might not prefer to the amount of uncertainty. If Niederle & Vesterlund (2007) where right then less females would prefer such a form of compensation.

The bonuses are calculated with the data of the VEB (vereniging voor effectenbezitters), which is the Dutch Union for Stockowners that annual presents the results of companies and the salaries of all the Dutch stock market listed management boards. On the site of the VEB it is possible to find how the salaries of individual management board members are composed. The income of an individual member of the board can be composed of: regular salary, shares, bonus, retirement savings, share options and in some cases a not specified remainder (which was never more than a couple of percentage). In the annual year reports of individual stock market listed companies can be read that different companies have different approaches to awarding bonuses to their board members. From these reports it became clear that shares, share options and bonuses are all seen as variable income and thereby bonuses. To calculate the average amount of bonuses awarded to board members, all these variable awards were added up and transformed into a percentage of the total income of a board member that year. All board known bonus percentages were calculated and added up and then divided by the amount of gathered bonuses to compose the average awarded bonus. The average awarded bonuses can be seen in Appendix C.

3.3.3 FSQCA

Apart from average awarded bonus and the average legislative period of board members this paper also use the FSQCA method, which is short for Fuzzy Set Qualitative Comparative Analysis. This method allows a researcher to look at quantitative data and transform them in such a matter that they can be qualitative interpretable. Most of the time using qualitative data means that only a handful of cases can be viewed at the same time, but this method allows to gather interpretable data from many different sources of the same time.

Both with the Crisp Set Qualitative Comparative Analysis and with the Fuzzy Set Qualitative Comparative Analysis a researcher looks at data from a company (amount of employees working there, average salary, industry, strategy) and has to grade this data with a grade between 0 to 1. The difference with Crisp Set and Fuzzy Set is in its use. Crisp can only give either the grade 0 or the grade 1 while the Fuzzy Set can also give grades intermediate, sort of like a Likert Scale. This makes for a more accurate distribution of data as a researcher does not have to look at data like it is only black and white, but can also implement multiple greyish tint to his heart’s desire.

The FSQCA is appropriate to use in this paper as it mostly commonly used with social phenomena’s (which a competitive environment would certainly qualify for). It is not possible to determine if a company is competitive if it scores above the sample mean nor if it is not if it is below the average mean. Ultimately it depends on qualitative identifying distinctive groups with substantive knowledge and is therefore seen as the middle way between qualitative and quantitative measurement (Crilly, 2013)

In this paper the five-fuzzy set of FSQCA is used to categorise the Dutch stock listed companies into 5 different competitive environments. The reason for 5 categories is because the 5 point Likert scale is the most widely used scale, used in management research and surveys and therefor this paper also chose to make a distinction between 5 different options instead of a simple yes or no or a continuous number. The categories are 0 (almost no competitive environment), 0.25 (fairly limited competitive environment), 0.5 (moderate competitive environment), 0.75 (high competitive environment) and 1 (extremely competitive environment).

Using FSQCA is a qualitative way of dividing Dutch stock market listed companies into categories and thus leaves room for interpretation. However this is preferred to just multiplying the average legislative period with the average awarded bonuses to form the competitive environment of a company, because companies might have the same awarded bonus systems and average legislative period, but still might differ immensely form each other on other parts. In the paper of Lückerath-Rovers (2015) it is described that female board members have an average legislative period of 6,2 years. This was also taking into account next to both legislative period and average awarded bonuses to compute the different competitive environments with the FSQCA. In Appendix C an overview is provided of how many women where promoted to the management board/supervisory board in the last 7 years (since 6,2 is rounded up to 7) and how many of them resigned/ were fired. This impacted companies score on the FSQCA. AFC Ajax for instance has a fairly low average awarded bonus for instance (which would indicate a low competitive environments), its average legislative period on the other hand is 5 years, which is quite quick and would contrastingly indicate a more competitive environment. AFC Ajax eventually was put in the 0,5 category because in these past 7 years only one female was ever appointed to board member, but who resigned within a stunning seven months! This paper thought this to be an indication that there had to be some indication of the competitive environment indeed and therefore placed AFC Ajax in the 0,5 category while Hollands Colours had for instance the exact same average awarded bonuses, but was placed in the 0 category. The amount of female board members appointed in the last 7 years and the amounts of resignations of female board members can both be found in appendix C. The number between the clasps are the years women have worked at a company. Bellow 4 years is considered very competitive, around 6 average and above 8 is considered not very competitive.

4.Results 4.1 Sample

This paper tries to discover the impact the Dutch advised quota has had on the hiring of women for Dutch stock market listed companies and the performances of these companies. However a couple of things need to be taken into account when looking at this sample.

As is already described in chapter 3.1, the Dutch stock market consisted of 96 companies in 2012, but this is different from the year 2015 were there are just 84 companies. Of these 84 companies fifteen of them are new entries in the years 2013-2015, which logically cannot be used in a paper that compares results from 2012 to 2015 onwards. These companies were therefore removed from the list leaving a 69 remaining.

Of these 69 companies that did appear on the Dutch stock market from 2012 to 2015 two companies (Bever and Nesdap) did not yet deposit their annual report at the time of writing this paper. Therefor these companies were also removed from the list. The company Makintosh went bankrupt in 2015 and therefor was also removed since its results would be too much of an outlier.

Of the 66 companies remaining, the VEB did not have up to date numbers on how awarded bonuses were computed for 4 Dutch stock market companies (for instance Royal ten Cate). Since the average awarded bonus is a key element in this paper for computing the competitive environment these stock market listed companies were also removed from the sample leaving a total and final number of 62 companies in this sample remaining for the hypotheses to be tested on.

4.2 Reliability

The reliability of these measurements are quite high, since most of the numbers are taking directly from the annual reports of the Dutch stock marketed companies themselves. The earnings before interest, taxes, depreciation and amortization (EBITDA) are mostly presented in the annual reports of a stock market listed company and otherwise were calculated using the numbers presented in annual reports or provided on the site Company.info, which is a site that provides large excel sheets which includes all kinds of data of companies such as profit, annual results and EBITDA’s.

The legislative period of board members is also reliable since it is taken directly from the source as the retirement and appointments of all board members (both supervisory and executive) is in the annual reports of the company itself. The composition of the average awarded bonuses is taken from the VEB which is a widely known expert on the Dutch stock market, as it is the Vereniging van Effect Bezitters which freely translated means Union of Stock Owners. Therefor the results of how these bonuses are awarded are deemed reliable by this paper.

The appointment of females in the management board from 2012 to 2015 is taken directly from the Female Board Index from Lückerath-Rovers’papers from 2012 and 2015. To compose the amount of resignations and appointment of females in the last 7 years the Female Board Index papers of Lückerath-Rovers from 2008 to 2015 were used as well as the annual reports of the Dutch stock market companies.

4.3. Student ‘t-test & OLS regression

For both the first and second hypothesis the Student’t-test and the ordinary least square regression will be used to investigate if there is any evidence of significant results. The Student’t-test is a test commonly used in social and management research, that assesses whether the means of two different groups are statistically different from each other (Trochim, 2006). The Student’ t-test is especially useful when you compare two groups to find out if their means are significantly different from each other. (Trochim, 2006) Since this is the case in both hypothesis 1 and 2 this paper finds the usage of the T-test justified.

This paper also makes use of OLS regression analyses (ordinary least square) for both the all three hypotheses. The Student’t-test gives insight if there is a difference between certain groups, but a regression analysis does more than that as it also investigates the relationships between these variables. Usually, it is used to ascertain the causal effect of one variable upon another (Skyses, 1992). Data is assembled on the underlying variables of interest and then regression is employed to estimate the quantitative effect of the casual variables upon the variable that they influence (Skyses, 1992). Just as with the t-test, the OLS regression analysis typically assesses the “statistical significance” of the estimated relationship, that is, the degree of confidence that the true relationship is close to the estimated relationship. (Skyses, 1992). However, before we can conduct an ordinary least-square regression or the Student ‘t-test have to meet a certain requirement.

The t-test and OLS regression are commonly applied when a dataset follows the normal distribution. Mordkoff (2000) explains that it can be assumed that all populations from which samples are taken are normally distributed at least when large samples are used from above 30 and that this assumption in that case is relatively uncontroversial. Since the sample of this paper is 62 companies, the requirement of N>30 is met. Also when drawing a graph it can be seen from the distribution that it looks normally distributed (see figure 1)

Figure 1

The other assumption for the ordinary least-square regression, which also needs to be established, is that all the variable are single response variables as the overall mean of a scale is needed to explore the relation between these two variables (Hutcheson & Moutinho, 2011). Since both female employees and the EBITDA are both composed of a single variable this requirement is also met.

The amount of companies does not exceed N = 100 in this paper, which is normally preferred when using quantitative data as a higher N most of the times makes a test more reliable just by including more cases. Therefor this paper uses a p-value of 0.1 (instead of the more custom 0.05) as the maximum outcome to justify and determine whether outcomes are significant or not.

4.4. Testing the hypotheses

The first hypothesis tests if there is significant evidence to assume that companies with female board members performed better then companies without female board members. To test this, the EBITDA’s of the years 2014 and 2015 were used, since these are the most recent data available. The two groups which were compared with each other in the t-test and OLS

0 0,05 0,1 0,15 0,2 0,25 0,3 0,35 0,4 0 10 20 30 40 50 60 70

Grafiektitel

regression were composed as followed: if a stock market company had no female directors in 2015 (having a 0 in table female 2015, appendix A) it would be placed in the group of Dutch stock market companies without females in the board. If a stock market company had females in their board, regardless of the amount it would be placed in the group: stock market company with female board members. The difference (growth/loss) in EBITDA between 2014 and 2015 is used as the measure for performance as it the most fair way to determine performance based on individual years (see also chapter 3.2.). The reason a percentage is chosen instead of the absolute difference is because the Dutch stock market listed companies differ greatly in their respective annual earnings. If the biggest firm with a couple of women in its management board would produce a higher EBITDA then all the other Dutch stock market listed companies then this would greatly influence the results, while the smallest EBITDA producing companies would be totally overlooked. A percentage therefor is the most fair way to look at the firms performance.

After running the Student T-test the p-value given is 0.3567. Since 0.3567 is way higher than the 0.1 value this paper justified as significant there is no evidence to assume that companies with female board members perform better then companies without female board members. When looking at the results of the regression analyses it can also be concluded that companies do not perform better when they have female board members. In fact the coefficient of the amount of females in boards is negative (-0.7724) (Standard error 0.5213) and it gives a p-value of 0.0662. This is lower than 0.1 and therefore significant.

H1: Companies with female board members perform better then companies without female board members. Not Supported

Following these outcomes we cannot conclude that companies who have female board members perform better than companies who have not and from the OLS regression we can conclude that in fact companies who have less female board members perform better, though in all fairness the R^2 is very low (0.03419). (More regression results can be view in appendix D)

For the second hypothesis, again two groups are formed to compare with the Student ‘t-test and the OLS regression. This second hypothesis partly tests the impact of the advised Dutch quota from 2012 to 2015. For this test the division is done as followed: when a stock market listed company has had an increase in total female board members it was placed in the group “more female board members” if the amount of female board members was decreased or was the same in the years 2012 and 2015 the stock market company was placed in the group “no more female board members”. The reason this paper looked at total amount of appointed females instead of an increase in percentage of female board members is, because an increase in the amount of board members could decrease the amount of male/female ratio while a company could have had an increase in its appointed female board members next to an increase in total directors overall. Therefor the total amount was used.

To compare the performance of companies this paper uses the difference in EBITDA from 2011 and 2012 and the difference in EBITDA from 2014 and 2015. The reason for this is because the Dutch advised quota was introduced by the Dutch government in November 2012. Therefor the difference in EBITDA from 2011 and 2012 is the last glance in the performance of a company before the quota was introduced. The reason the EBITDA difference of 2014 and 2015 was chosen is because most companies show abnormal annual results in 2013 due to the economic crisis. Therefor the difference in EBITDA in 2012-2013 would be tremendously low and in 2013-2014 extremely high. An increase in women in 2013 would therefor most likely be correlated with plummeting results. However this paper did not think that blaming the economic crisis on women would be entirely fair

After running the t-test for the second hypotheses the p-value that was given was 0.0905. This p-value is lower than the earlier established justified p-value of 0.1 to be significant. This means that there is significant evidence that there is indeed a difference in the results of a company which hires more women than a company which does not. The OLS regression shows also a p-value below 0.1 (0.083). However just as with the coefficient in hypothesis 1 the coefficient is negative (-1.0201) (standard error 0.8093). Therefor hypothesis 2:

H2: An increase in female board members will lead to higher results for a company.

H2 Is also not supported. However again the R^2 is very low with 0.025. (More results can be viewed in Appendix D)

For the last hypotheses of this paper, females have been divided into 5 different competitive environments using the FSQCA method (3.3.3 for more information) ranging from 0 to 1. These groups do not contain at least N > 30 and when looking at the distribution of women among these categories (figure 2) it can be seen that it does not have a normal distribution. 0 5 10 15 20 25 30 0 0,3 0,5 0,7 1 N u m b er o f fe m ales Competitiveness

Distribution of females

It can be seen that the amount of women who work in less competitive environments (0 and 0,3) is higher than the total amount of females working in high competitive environments (0,7 and 1), but this is mainly because the amount of very competitive stock market companies is only 7. To test if there is a significant difference between the means of these different competitive environments, this paper uses the Kruskal-Wallis method. The Kruskal- Wallis method is used when looking at ordinal, independent data in more than two categories, when normal distribution cannot be assumed. Since the competitive environments are ordinal (0,5 is more competitive than 0.3 which is more competitive than 0 etc.) and independent from each other (Stock listed companies who are placed category 0.7 do not appear in other competitive environment, as is the same for every other competitive environment) it is possible to use the Kruskal-Wallis test to test if a specific group is significantly different from the others. The Kruskal-Wallis test does not tell you which group is significant different from another. So if the Kruskal-Wallis test finds a significant difference a Mann-Withney-Wilcoxon test is needed to find out which specifically. However after running the test in SPSS the outcome is that the distribution of females is the same across all categories with a p-value of 0.103. (Spss output can be found in appendix E) This is higher than the 0.1 and therfor there is no evidence that supports hypothesis 3

When using OLS regression (although this is technically not correct since the data is not normal distributed) a p-value of 0.6063 is found which is also higher than 0.1.

H3: Females are less likely to work for companies who have high levels of internal competitiveness. H3 is not supported.

5. Discussion

5.1 Interpreting the results

5.1.1 Female influence on company performance

Although an overwhelming amount of authors proposed a possibility between appointing female board members and an increase in company results, this paper did not find any evidence that divers boards would perform better and that appointing female employees to the board of directors would have a positive impact on company results.

Contrary to common believe, the results in chapter 4 for hypothesis 1 and 2 even suggest that in fact the direct opposite holds true for stock listed companies. The coefficients of both hypothesis 1 and 2 are negative (See appendix D). Meaning that for every extra female board member the results (% EBITDA increase) will be negative. However it is important to take this with a little grain of salt. First of all the R^2 are very low. Both of them are below 0.1 which is considered pretty weak. R^2 explains how much variance in female employment can explain the differences in EBITDA and since this is fairly weak, it is not possible to conclude that women alone are responsible for the lower EBITDA of stock listed companies. In fact the R^2 highlight that there are most likely many other factors that influence the EBITDA score apart from female mangers.

Secondly the p-values of both hypothesis 1 and 2 were low, meaning that there is a high possibility of a connection between female board members and % EBITDA decrease. However this is only because this paper allows for a significant p-value of 0.1 due to the fact that the amount of Dutch stock market listed companies that where active in 2012 until 2015 does not exceed N=100. If more companies were added to the listed (which sadly was impossible to do, since there are no more Dutch stock listed companies) the outcomes might have been different when the justified p-value would drop (to for instance a more common used 0.05) and thereby

making the now significant results of the Student ‘t-test of hypothesis 2 and the significant results of the OLS regression of both hypothesis 1 and 2 not significant anymore.

However assuming that the results of chapter 4 actually do indicate a drop in performance for companies there are a couple papers who can explain this difference with the literature papers that predicted the contrary rise in performance.

Two of these papers have been named in the literature review in chapter 2.3. These are the papers of Ryan & Haslam (2005) and Pande & Ford (2011). Both suggested that a company would only be negatively influenced in profit on the short term, because for instance shareholders would be warry of female management and the ‘new course’ the company would pursue, but also that men tended to feel more at ease at the position and made better (profit increasing) decisions than their female counterparts. However both paper suggested that this was only for a short term and that differences faded after a period of 5 months. This paper could be an indication that this period actually is longer then the proposed 5 months and is in fact a couple of years or permanent.

Another possible explanation for the different results presented in chapter 4, in comparison with the views of the papers presented in chapter 2 can be found in Joecks & Vetter (2013). Joecks & Vetter, (2013) find that there is a certain U shape when exploring the link between gender diversity and performance. This means that the appointment of women actually decreases profit until a certain female/male ratio is achieved. They call this ratio the “critical mass”. The critical mass is supposed to be around 30% in their view and only after this ratio has been reached will companies actually benefit from appointing more females to the board. This statement is however a bit contradicting in the paper of for instances Apesteguia (2007) who say that companies will function best if teams are composed of around 30%. However the

“critical mass” idea and its U-shaped form could indicate that a certain amount of female directors has to be reached until it will start having effect. However where this critical mass lies exactly, research is still divided (as is illustrated with the example above). This is a very interesting thought, however since the amount of companies that reached the 30% ratio is remarkably low (only 8 out of the original 84) this paper was not able to look into this more thorough since the outcome of this group would be highly questionable due to its low N. (More on this in chapter 5.4)

Adams (2007), while being an advocate of more female board members, also suggests that mandating certain amounts gender quota in boardrooms may not increase board effectiveness on average and might even reduce it (p.3). This has some high resemblance with Joecks & Vetter (2013), since the 30% quota does not seem to be very effective for companies (only more than 30% would result in better performances, while others argue that it would be until 30%), but it also indicates that there might be a problem with the quota itself.

The focus on the shattering of the glass ceiling might be too much focused on bringing diversity to the boardroom. As is briefly touched upon in chapter 2, there are a couple of indications that evaluation forms unintentionally advocate male characteristics to be good determinant to decide if one is a good manager or not. Burgess and Tharenou (2002), find for instance that women are more philanthropically driven compared to the companies preferred economically driven which is highly present with the male director, which should be aimed for by female who want to be appointed to board of directors. Other things Burgess and Tharenou (2002) find is the amount of maternity leave (also briefly discussed in chapter 2.5) and having a proactive approach to placing themselves in higher positions compared, which are both also more male characteristics than female. It seems that women with male associated characteristics (females in male disguise to put it bluntly) reach the top more easily than there more feminine counterparts. If a quota is introduced the logical follow up is that more females will be promoted

to the board of directors. However this could mean that they also bring a different management style with them (because different characteristics lead to different approaches) that takes time to get accustomed to. This together with their years of experience could be an indication that it takes some time before a company will see results.

This idea is also supported by Sol who explains that female managers perform better when there has been accounted for their years of experience. However this also indicates that if only the total performance is observed there might be a loss, when the moderation variable of years of experience is left out and just overall results are observed. It indicates, just as Pande & Forde (2013) describe, that companies might lose profit on the short term, but might profit in the end.

This is trend can also be seen in the paper of Ahern & Dittmar (2011) who describe the total performance of the company as the exogenous change of the corporate boards and who find that the imposed quota constraint companies and caused a significant drop in the stock prices of the stock listed companies, because the quota led to younger and less experienced boards. Ryan & Haslam (2005) call this concept the glass cliff (as an extension of the glass ceiling metaphor) meaning that a company can hurt itself when only looking at gender when appointing new board members. Ahern’s & Dittmars (2011) results also showed a decrease in operating performance (which is used to calculate EBITDA) and an increase in leverage and acquisitions of Norwegian stock listed companies , but it could well be that these results will be different when the women have aged and are on the exact same level as men. It could well be that a time period of only 7 years (2004-2011) was not enough to see the long lasting effects of appointing more female board members. As would also be the case with this paper its 5 years (2011-2015).

5.1.2 the competitive environments

The p-value of 0.106 is above the justified 0.1 and therefor this paper did not find any significant differences that indicate that women prefer less competitive environments above hyper/very competitive environments. This does not mean that hypothesis 3 (women prefer less competitive environments above competitive environments) is not true, but only that this paper was not able to find supporting evidence.

As was stated in chapter 2.7 the competitive environment is a fairly new research area and therefor it might be that the parameters of average legislative period and average awarded bonus, combined with the FSQCA to create multiple competitive environments were not sufficient enough alone to compute the environments.

Other possible variables, for instance stress among employees and peer evaluation, could prove very valuable additions to composing the competitive environment. Multiple authors (Menge & Sutherland, 1987; Neil, 2011) have associated stress with competition as one (competition) is most likely the cause of the other (stress). Competing against your own colleagues could increase your stress level due to the fact that you take on more workload to stick out. This is stated in Cohen (1994) who states that an increase in workload will also lead to an increase in stress. If employees at certain stock listed companies would indicate high amount of stress this could be an indication that it is caused due to an competitive environment in combination with the other parameters.

Another possible variable to include is the existence of peer evaluation within a stock listed company. This is proposed by Alagna (2012) who finds that when men are put in competitive systems they tend to uphold higher standards for performance than feminine groups and therefor rate there fellow employees lower. The existence of a peer evaluation by itself does indicate a more competitive environment, because an employee is being kept in check not only

by its manager, but also by it fellow employees as he is graded by them. This could lead to quicker turnover of the management board (Defond & Park, 1999), since performance evaluation by peers can improve board of directors abilities to identify unfit CEO’s which in turn also combines perfectly with the average legislative period of a board member and therefor seems to be also a good indicator for a competitive environment.

A last reason why this paper did not find any evidence that female directors prefer less competitive environments than men, might have something to do with the proposed male characteristic of the currently board seat holding female members (Burgess and Tharenou , 2002). If these women have characteristics that are very similar to men then they might be completely unaffected by the amount of competitiveness of a certain environment or maybe even prefer a more variable salary and flexible legislative period. A thorough research in the middle management and below of stock listed companies could give different outcomes since less percentage of the female employees might possess these characteristics and a possible barrier might become visible that way.

5.2 Contribution to existing literature

This paper contributed in three ways to the already existing literature in the glass ceiling discussion. First of all it shows that the appointment of female employees does not automatically mean an increase in firm performance. In fact if conclusions have to be drawn on this paper alone, it seems that the direct opposite holds true. Secondly it proposes a different way of evaluating companies using the EBITDA instead of profit or the return on stock since it is a much more reliable performance indicator as it does not necessarily measures the stock market listed companies their profitability, but a more overall view and is less influenced by for instance preferred shareholders (Lückeratz- Rover, 2011).

The third and final contribution of this paper is the attempt to create different competitive environments. There is a lot of room to enhance these five different competitive environments of businesses, but this paper could serve as a base upon which other research in this area can build.

5.3 Managerial implications

As is stated also in chapter 5.1 managers should not automatically fire every women in the management board after reading this article. The p-values may be significant, but the R^2 are relatively low meaning that very little variance can be explained trough the amount of female employment and the decrease in earnings. However a good take away from this article is that there is certain proof of the Glass Cliff effect described by Ryan & Hamel (2005) who say that looking just at gender when appointing someone for a board might not lead to positive effects and maybe even negative results if no other values are also taking into account.

The Dutch government should however take notice of this paper. Companies only seem to hurt themselves with forcefully appointing women to management positions. This might be because almost all of the Dutch stock market listed companies are below the critical mass of 30% (Joecks and Vedder, 2013). The Dutch government should decide whether it wants to abandon this soft push towards more female board members (in the form of the advised quota) or go for it and turn the advised quota into a mandatory minimum (and preferably even increase the ratio of 30%). It will solve at least the image problem and within a couple of years it might be better to reflect if it also increased overall results as was more elaborately explained in section 5.1.1.