GENDER DIVERSITY WITHIN BOARDS OF DUTCH SMES: HOW MUCH DIFFERENCE DOES DIFFERENCE MAKE?

GENDER DIVERSITY WITHIN BOARDS OF DUTCH SMES:

HOW MUCH DIFFERENCE DOES DIFFERENCE MAKE?

Master Thesis, MscBA, specialization Small Business & Entrepreneurship

University of Groningen, Faculty of Economics and Business

August 12, 2014

Marije Boersma

Student number: 1934279

Oude Arnhemseweg 23HS

3702 BA Zeist

+31 (0)642447632

marijeboersma@live.nl

Supervisor: Dr. M.J. Brand

Co-assessor: Dr. K. van Veen

1

GENDER DIVERSITY WITHIN BOARDS OF DUTCH SMES:

HOW MUCH DIFFERENCE DOES DIFFERENCE MAKE?

ABSTRACT

Over the past years, there has been increasing attention to the influence of gender diversity within boards. There have been two arguments for increasing diversity within boards, also known as the normative and the business case for diversity. The normative case is based on equality of individuals and the business case expects a positive effect on firm performance. The relationship between gender diversity and firm performance is known as the direct effect of gender diversity. An indirect effect may also exist, by the influence of a board’s process and performance. The main goal of this paper was to present a complete picture of the direct and indirect effects of gender diversity within the board of Dutch SMEs and to identify the antecedents of gender diversity. It was expected that gender

diversity within the boards of SMEs positively influences the firm’s performance. It was also expected that gender diversity positively influences the process and performance of a board. The process and performance was assumed to be a mediator within the relation between gender diversity and firm performance. No significant results were found. Therefore, it can be concluded that ‘difference’ (gender diversity) does not make any significant ‘difference’ (firm performance). Therefore, the decisions of managers to appoint women on boards should be based on criteria other than firm performance.

Keywords: SMEs, supervisory board, gender diversity, process and performance, monitoring role, firm performance, the Netherlands

2

TABLE OF CONTENT

2.4 Conceptual model and research hypotheses ... 12

3. METHODOLOGY ... 15

3.1 Data collection methods ... 15

3.2 Analysis plan ... 20

4. RESULTS ... 21

4.1 Descriptive statistics and correlations ... 21

4.2 Hypothesis testing ... 21

5. DISCUSSION AND CONCLUSIONS ... 26

5.1 Conclusions ... 26

5.2 Implications and directions for research ... 28

5.3 Implications for practice ... 30

REFERENCES ... 31

APPENDICES ... 34

I. Overview of the constructs and their sources ... 34

II. Summary of extra regression analyses ... 36

3

1. INTRODUCTION

People differ in a lot of observable and unobservable ways (Grosvold et al, 2007). This variety in gender, age, ethnicity, education, experience and knowledge is also known as diversity. Studies have investigated the direct and indirect effects of diversity on group processes and performance and have found positive and negative outcomes (Williams and O’Reilly, 1998). These outcomes include increases in creativity, but also poorer communication and a decrease in flexibility. One group within an organization is the board of directors. The board of directors is important, because it is considered to be the bridge between the shareholders and the management of a company (Van den Berghe and Levrau, 2004). One of the most significant issues faced by managers, directors and shareholders is the gender, racial, and cultural composition of the board of directors (Carter et al., 2003).

Over the past years, there has been increasing attention to the influence of gender diversity within boards (Rhode and Packel, 2010). Poor female representation on corporate boards of directors is a global phenomenon and the importance of improving the gender balance of corporate boards is increasingly recognized across the world (World Development Report 2012, Terjesen and Singh, 2008). The widespread recognition of difficulties involved in improving the gender balance of corporate boards has prompted the introduction of a number of initiatives such as those that support the educational development of women leaders, mentoring activities within organisations, the generation of increased media attention, more sustained political pressure and in some countries, legislative compulsion (Grosvold et al, 2007). Two examples of companies taking such initiatives are Mc Kinsey and Price Waterhouse Coopers. Mc Kinsey organized a three-day workshop in Paris in May 2014 for the Next Generation Female Leaders: women from Europe and the Middle East. Price Waterhouse Coopers organized a special day in Amsterdam for talented Dutch female students to stimulate them to apply for a job at Price Waterhouse Coopers.

4

This research focuses on Dutch Small and Medium-sized Enterprises (SMEs). SMEs in Europe play an important role, because they provide about 75 million jobs and represent 99% of all enterprises. In the Netherlands, SMEs comprise 99,6% of all enterprises and employ 68% of the labour force (OECD, 2012). According to the definition of the European Commission (2003), SMEs are enterprises that employ 0-249 persons, have an annual turnover not exceeding 50 million euro and/or an annual balance sheet total not exceeding 43 million euro.

The business environment of Dutch SMEs is getting more and more complex, unpredictable and unstable and the CEO’s decision-making process may be mainly intuitive and based on overconfidence (Van Gils, 2005). To have a board is considered to be important, since a board of directors can be regarded as a mechanism that regulates relationships between the firm's internal and external stakeholders. Board members can also add knowledge, resources and networks. A well functioning board of directors could strengthen the strategic decision-making process and also increase the company results on the long term. (Van Gils, 2005). Therefore, understanding of the board of directors in SMEs is has become increasingly important in the field of management (Gabrielsson and Winlund, 2000). There are indications that board members of SMEs fulfill other roles than board members of large firms, because of their different ownership and control structures (Neville, 2011).

This paper conceptually and empirically addresses gender diversity, the process and performance of the board and firm performance. Four hypotheses are developed about the direct and indirect effects of gender diversity. These hypotheses will be tested, based on data from a survey among CEOs. The main goal of this paper is to present a complete picture of the direct and indirect effects of gender diversity within the board of Dutch SMEs, by combining and building on the current literature about gender diversity, the process and performance of the board and firm performance. A sub goal is to identify the antecedents of gender diversity. The new combination of constructs is a contribution to the literature.

5

2.

THEORY

In this chapter the three main constructs of this research will be further explained: diversity, board performance and firm performance. The first section will give an overview of the current literature about diversity. Diversity is categorized into three sub sections: diversity in general, gender diversity and (gender) diversity within boards. The second section is about firm performance with sub sections about the definition and the empirical evidence. In the third section, the performance of the supervisory board will be discussed. This chapter closes with the research hypotheses and the conceptual models in which all constructs are combined.

2.1 Diversity

Diversity in general

Diversity can be referred to situations in which the actors of interest are not alike with respect to some attribute(s) or to variety or a point or respect in which things differ (Jackson et al., 1993). Within the existing empirical work several definitions of human diversity have been used. An important distinction is made between demographic and cognitive dimensions of diversity (Grosvold et al, 2007). Demographic diversity covers the observable characteristics of diversity, such as gender, age and ethnicity (Grosvold et al, 2007). Cognitive diversity is non-observable and includes education, experience and knowledge (Solakoglu, 2013). One reason for differentiating between observable and nonobservable types of diversity is that when differences between people are visible, they are particularly likely to evoke responses that are due directly to biases, prejudices, or stereotypes (Jackson et al., 1993). Many psychologists, economists, sociologists, anthropologists, communication and education researchers and organizational scholars have conducted laboratory and field studies examing the effects of diversity on group performance (Williams and O’Reilly, 1998). They found that diversity in work related groups appears to be a double-edged sword by increasing the opportunity for creativity and improving performance (the value-in-diversity hypothesis) as well as the likelihood that group members will be dissatisfied and fail to identify with the group (Jackson, et al., 1993; Williams and O’Reilly, 1998).

Gender diversity

6

But it is a bit more complicated when gender diversity is included: the relationship between gender diversity and group process is dependent upon the proportion of the first and second group and not only the fact that there is group heterogeneity (Kanter, 1997). This expectation is based on research that has shown that as the proportion of individuals in a group who passes a particular characteristic grows smaller, those who posses this characteristic will become increasingly aware of their social identity (Williams and O’Reilly, 1998). Therefore, Kanter (1977) identified four different types of groups, based on proportionality effects, as shown in table 1. Each category has its own interaction processes. According to Kanter (1977), the skewed groups are most problematic. In tilted or balanced groups, the diversity of members will allow for productive discussions, which positively affects group performance (Konrad and Kramer, 2006). These groups indicate the existence of a critical mass.

TABLE 1

Kanter’s different groups Uniform

groups

All members share the same (visible) characteristics. With respect to gender diversity, all members (100%) are female or all members are male.

Skewed groups

One dominant type controls the few and therefore controls the groups and its culture. The few are called ‘tokens’ and are not treated as individuals, but as representatives for their category. With respect to gender diversity, a male dominated skewed group consist of a maximum of 20% women.

Tilted groups

A less extreme distribution: minority members can ally and influence the culture of the group. This category would consist of 20-40% women.

Balanced groups

In this group, majority and minority turn into potential subgroups where gender differences become less and less important. The focus is on different abilities and skills of men and women. A balanced group consists of 40-60% women.

Diversity within boards

Composition of the board of directors is perhaps the most widely studied variable in governance research, but there is little consensus on its effect (O’Reilly and Main, 2012). Diversity within boards refers to any kind of difference between members (Nekhili and Gatfaoui, 2012). Growing academic and practitioner literature has highlighted the homogeneity of corporate boards and suggested that this raises significant ethical, political and economic issues (Grosvold et al, 2007). The ethical and economic issues are also known as the normative and business case for board diversity. These two cases will be further explained now.

7

illustrates social responsibility and equal opportunity standards and it is a signal of good corporate governance (Nekhili and Gatfaoui, 2012; Brammer, 2009; O’Reilly and Main, 2012). It may also increase corporate image and corporate reputation, because female board members promote greater responsiveness to the needs of female customers (Brammer, 2009). Another reason for adding females, according to the normative case, is that those females can provide a valuable form of legitimacy in the eyes of potential and current employees and they also symbolise career possibilities to prospective recruits (Hillman et al., 2007).Dale-Olsen et al. (2013) found that increasing women’s representation on boards may also have positive long-term effects on women’s opportunities and willingness to seek out other high-ranked positions in the labor market, which may have important and wider-ranging second-round effects on gender equality. Brammer et al. (2009) share this view, by stating that increased board diversity may lead to improvements in workforce motivation and loyalty.

In many countries there has been a pressure for governance reforms that foster gender diversity in the boardroom, based on the normative arguments (Joecks et al., 2012). Norway was the forerunner of gender equality, by introducing laws in 2002 that required that women should comprise 40 percent of the boards of directors by January 1, 2008 (Dale-Olsen et al., 2013). Spain followed this example and enacted a law prescribing a 40 percent quota of female board members in 2015(Joecks et al., 2012). France adopted a law in 2011 which established that by 2016 40% of executive board members of large companies should be women (Nekhili and Gatfaoui, 2012). The Netherlands also introduced a quota of 30% female board members for large firms, but this is not a strict rule. Companies who do not meet this quota, have to give an explanation in their annual report. A recent publication in the Netherlands (Lückerath-Rovers, 2013) showed that at present only 18% of the supervisory board members of large firms are women.

The business case for board diversity assumes that diversity within a board has positive economic effects. This perspective assumes that talents and competencies are not evenly distributed across demographic groups and therefore, boards should have diverse members (Grosvold et al., 2007). These diverse members can stimulate creativity and innovation, increase marketplace understanding, add valuable new perspectives, improve outcomes, problem solving and the decision-making processes, which may finally improve firm performance (Rhode and Packel, 2010; O’Reilly and Main, 2012; Nekhili and Gatfaoui, 2012; Solakoglu, 2013). Directors with diverse skills, experiences and backgrounds are more likely to raise questions that add to, rather than simply echo, the voice of management. As a result, they discourage groupthink and reduce malpractice (Brammer et al., 2009; Nekhili and Gatfaoui, 2012). Gender diversity also leads to broader talent pools (Adams and Ferreira, 2009; Solakoglu, 2013).

8

that females have different experience sets, beliefs and perspectives and therefore, females have the potential to link organizations to different constituencies than males. Second, Nekhili and Gatfaoui (2012) found that males and females exhibit different moral reasoning, handle ethical decision-making differently and such differences can help to improve and to promote more productive discussions. Third, female directors raise issues that pertain to multiple stakeholders by asking difficult questions about tough issues and use their interpersonal skills to positively influence board processes. Fourth, female directors are frequently felt to bring a consumer or community orientation to the board that is particularly valuable in certain industries and service businesses (Fryxell and Lerner, 1989). Fifth, female directors could also be more independent, because they are not in the old boys’ network. Sixth, female directors might provide useful role models for female middle managers, audit the results of hiring and promotion policies with respect to women or better represent the firm to a constituency (Harrigan, 1981). Seventh, female directors might play an important role in enhancing corporate effectiveness by inspiring, motivating and mentoring female employees. Eighth, female directors are used to being more patient and open minded, which makes communication easier and therefore helps diffuse information from the board to investors (Nekhili and Gatfaoui, 2012).

2.2 Firm performance

As already mentioned, the business case provides arguments that gender diversity has a direct, positive economic effect for a firm. In the following section, the definition of firm performance will be given and an overview of the empirical evidence regarding the business case will be discussed.

Definition

A board normally exists to serve the company’s goals and to increase the performance of the company. Firm performance can be seen as the total of three components: financial performance, operational performance and organizational effectiveness (Venkatraman and Ramanujam, 1986). Financial performance reflects the fulfillment of a firm’s economic goals and this measure has been dominant in many papers (e.g.: Adams and Ferreira 2009; Nekhili and Gatfaoui, 2012; Dale-Olsen et al. 2013 and Joecks et al., 2012). In these papers, market and accounting based measures like growth, sales, profit, return on investment and stock-market returns are used. The second component, operational performance, emphasizes non-financial measures like market share, product quality and other measures of (technological) efficiency. The last component, organizational effectiveness, covers the organizational goals and the influence of stakeholders, but most studies restrict their focus to the first two definitions of firm performance (Venkatraman and Ramanujam, 1986).

Empirical evidence

9

governance. As mentioned earlier, the empirical evidence still remains mixed. Most research has been done within large companies by using financial data from the annual reports. Lückerath-Rovers (2013) found a positive, linear relation between gender diversity in the board and ROI of large companies in the Netherlands.

An interesting relation between gender diversity in the board and firm performance has been found by Joecks et al. (2012). They found a positive effect on firm performance only after a ‘critical mass’ (about 30 percent women in the board) has been reached. Before this percentage, diversity negatively affected firm performance, so the relationship follows a U-shape. These negative outcomes could be caused by the following reasons. The more dissimilar the directors are, the more they could disagree and the more conflict there could be on the board (Adams and Ferreira, 2009). Second, decision making in a diverse board takes more time and therefore, firm performance may decrease in sectors that require a quick response to market shocks (Solakoglu, 2013). These arguments match the critical mass theory of Kanter (1977).

2.3 Board performance

In the following sections, the performance of the supervisory board is discussed. The actual functioning of the board is explained in the first section, by clarifying its main roles and the influence of (gender) diversity on these roles. In the second section, board performance will be discussed. The last section explains why this research focuses on SMEs.

Board roles

The board of directors needs to fulfil a two-fold role: on the one hand, boards are expected to control and monitor the company and on the other hand, they need to be involved in strategy (Van den Berghe and Levrau, 2004). These reflect the two main functional roles of a board that can add value to a firm: the monitoring and the service role (van den Heuvel et al., 2006).

The monitoring role, also known as control role, is based on the agency theory. Within this theory, conflicts of interest occur among shareholders and managers, and the board of directors is designed to monitor the behaviour of managers by ensuring that they operate in the interest of shareholders (Bertoni et al., 2014). This phenomenon is also known as the principal-agent problem (Van den Heuvel et al., 2006).

10

connections (Van den Berghe and Levrau,2004). The network role is based on the Resource Dependency Theory. An important aspect of this theory is that organizations are seen as open systems, dependent upon external entities for survival (Pfeffer, 1972). This theory regards corporate boards as an essential link between the company and its environment and the external resources. This link is necessary for good corporate performance, because it helps to reduce dependency between the organization and external contingencies, lowers transaction costs, diminishes uncertainty for the firm and ultimately aid in the survival of the firm (Hillman and Dalziel, 2003; Lückerath-Rovers, 2013).

The advisory role is based on the RBV, which is about the resources of a firm. These resources include all assets, capabilities, organizational processes, firm attributes, information, knowledge, etc. controlled by a firm that enables the firm to conceive and implement strategies that improves its efficiency and effectiveness (Daft, 1983). According to Barney (1991), there are three categories of firm resources: physical, human and organizational capital resources. The human capital resources are most important when considering board roles and they include the training, experience, judgment, intelligence, relationships and insights of individual managers and workers in a firm (Barney, 1991).

The influence of gender diversity on board roles

According to Joecks et al. (2012) male and female directors differ in a whole range of aspects: females are more risk averse than males, they are less aggressive in their choice of strategy, and more likely to invest in a sustainable way. Because of these differences, male and female board members may fulfill other board roles. Researchers have indeed found that female directors are tougher monitors than male directors and that the presence of women directors strengthens boards in their capacity to reduce agency conflicts (Adams and Ferreira, 2009; Nekhili and Gatfaoui, 2012).

Board role performance

CEOs have their own views on what constitutes a good board; therefore the opinion of the CEO is often used to measure board performance (Van den Berghe and Levrau, 2004). Van den Heuvel et al. (2006) investigated the importance and intensity of board roles according to CEOs. Higher effort of the board members, leads to a higher intensity of a certain board role (Huse, 2007). A higher intensity of a certain role will increase the effectiveness and satisfaction and may ultimately lead to higher firm performance. Nagel (2009) built on the work of Van den Heuvel et al. (2006), by measuring the importance and intensity again and thereby adding the CEO’s satisfaction. The reason for including satisfaction was based on the difference between efficient and effective performance. Doing something right (efficient), can be measured with role importance and role performance, but doing the right thing (effective), can only be measured by satisfaction.

11

actively during the board meetings (Gabrielsson and Winlund, 2000). Paper boards consist of passive board members and do not add much value to the firm.

So far, the normative case and the business case for diversity have been discussed. The business case assumes that gender diversity within a board increases firm performance. Firm performance is further explained and the empirical evidence has been presented. After that, the board roles and board performance has been discussed. Traditionally, the focus of researchers was on the direct relationship between board characteristics (gender diversity) and firm performance, without looking at the board roles (Dale-Olsen et al., 2013; Joecks et al., 2012; Adams and Ferreira, 2009). This so-called black box approach has not given consistent and conclusive results and therefore, researchers started looking for possible mediators. Zahra and Pearce (1989) developed an integrative model in which four attributes (composition, characteristics, structure and process) influence the board roles and these board roles lead to corporate financial performance. Forbes and Milliken (1999) also included board roles in their model of board processes and their impacts on board effectiveness. To conclude, it seems that the relation between board diversity and firm performance can be direct, or indirect via the intensity and performance of the board roles.

Focus on SMEs

12

2.4 Conceptual model and research hypotheses

Building on the literature review presented above on the effects of gender diversity, the following conceptual model is proposed:

FIGURE 1

Figure 1 presents a graphical depiction of the expected relations between the three main constructs. It is expected that gender diversity has a direct effect on the firm performance (H1). But this focus may be too narrow and therefore, this research will also investigate the indirect effects of gender diversity, via the process and performance of the supervisory board (H2 en H4). Process of the supervisory board includes the intensity of the board roles. Performance of the supervisory board covers the satisfaction per role and the overall satisfaction. If gender diversity within the board has an influence on the intensity and satisfaction of the board roles, then ‘process and performance of the supervisory board’ may have a mediating role (H3). The relations between the several constructs will be further explained in the following section.

Direct effect

As pointed out earlier, the business case theory states that gender diversity will increase firm performance. This theoy is not based on the process, but only focuses on the outcome. Therefore, the following hypothesis has been developed:

H1A: Gender diversity within boards of SMEs is positively related to their firm performance.

Given that the empirical evidence on this linear relation remains mixed, the following hypothesis is based on the ‘critical mass’ theory of Kanter (1977) and Joecks et al. (2012). Numbers seem to make a difference and therefore, hypothesis 2B is formulated as follows:

13

Indirect effect

As discussed in the previous sections, it was found that there are differences between female and male board members. In line with the existing theory about females being more likely to join monitoring committees, it is expected that the intensity of the monitoring role will be higher with higher female representation. It is also expected that gender diversity increases the satisfaction of the CEO with the monitoring role: the higher the role intensity, the higher the effectiveness and the more satisfied the CEO will be. Having a gender majority (less than 70%) is considered to strengthen these relationships, based on the critical mass theory. A critical mass is needed to take advantage of the diversity of a board. Therefore, the following hypotheses are developed:

H2A: Gender diversity within boards of SMEs is positively related to the intensity of the monitoring role

H2B: Gender majority less than 70% within boards of SMEs causes a relatively high intensity of the monitoring role.

H2C: Gender diversity within boards of SMEs is positively related to the satisfaction of the monitoring role.

H2D: Gender majority less than 70% within boards of SMEs causes a relatively high satisfaction of the monitoring role.

The bulk of research has focused on the direct associations between board attributes and company performance, ignoring the indirect path, through board roles (Zarah and Pearce, 1989). To my best knowledge, previous studies did not yet include gender diversity within the relation between the process and performance of the board and the firms’ performance. Building on the theory of Zarah and Pearce (1989) it is expected that the performance of the board is the indirect path between gender diversity and firm performance. Therefore, the following hypothesis is developed:

H3: Overall satisfaction of the board roles within SMEs is a full mediator of the relation between gender diversity and firm performance.

This research continues on the work of Nagel (2009), who investigated the intensity, importance and satisfaction of board roles of Dutch SMEs. He found that the advice role was most intensively executed and also added most value, followed by the monitoring and the network role. Therefore, the following hypothesis is developed:

14 2.4.4. Antecedents of gender diversity

In line with the theory of Zahra and Pearce (1989), it is expected that there may be additional constructs influencing the gender diversity of a supervisory board. These constructs can be divided in three categories: individual, firm and context level, as shown in figure 2.

The CEO is important, because the CEO drives the diversity management process and supports the benefits of gender-diverse boardrooms (Nekhili and Gatfaoui, 2013). The focus on the director’ characteristics is not new, it was already mentioned in the work of Zahra and Pearce (1989). The ‘individual’ construct include the gender, age, children and attitude of the CEO. O’Reilly and Main (2012) indicated that firms are more likely to appoint a woman to their board if the firm has a female as CEO, so the gender of the CEO plays an important role. The appointment of women may also be determined by the desire of the CEO to work with a well-diversified board (Nekhili and Gatfaoui, 2013). This desire is dependent on the benefits that a CEO expects from gender diversity and in this research; this desire is called ‘Attitude CEO’. The last individual aspect is the possibility of the CEO having children. As far as known, this aspect has not yet been investigated in the current literature about gender diversity. It is expected that a CEO with children may have a different attitude about gender diversity, especially about working women. Dutch people are often quite old-fashioned in their attitudes. From a traditional view, it can be expected that a male CEO with children is negative about women in board rooms, as those women should take care of their children. For female, it may be the other way around.

FIGURE 2

The influence of antecedents on gender diversity

At firm level, the percentage of female employees plays a role. There is evidence that firms with more female employees are more likely to have women directors (O’Reilly and Main, 2012). According to Lückerath-Rovers (2013), more female employees at all levels of a company will probably lead to more women in senior positions and ultimately, the board.

15

(O’Reilly and Main, 2012). Internal and external parties may cause this pressure. For example, the institutional pressure of the quota of 30 percent female board members for large firms may also have an influence on SMEs. Pressure from investors, through shareholder proposals, also encourages a firm to increase gender diversity within their board (Nekhili and Gatfaoui, 2012). In a large firm, the ownership and management is often separated, in contrary to a small firm. Therefore, the CEO of a small firm has more influence on the nomination of female board members than a CEO of a large firm.

3. METHODOLOGY

This chapter discusses the method of this research. A quantitive approach is used to investigate the direct and indirect effects of gender diversity in the supervisory board. First, the sample and collection procedure are clarified, including a justification of the chosen method (survey). Second, the definition and measurement of the constructs are explained. This chapter ends with an analysis plan. Within this chapter, validity and reliability issues are also addressed.

3.1 Data collection methods

3.1.1. Sample and collection procedure

The data used in this study was gathered via primary (survey) and secondary (Orbis) sources. To be able to send the surveys, the companies were obtained from the Orbis database on May 12th 2014. The initial sample included 705 companies located in the Netherlands, employing fewer than 250 persons, with a supervisory board, a maximum annual turnover of 50 million euro and an annual balance sheet total with a maximum of 43 million. These requirements were based on the definition of an SME, according to the European Commission (2003). The telephone number and website address of most companies were included in the database. By visiting each website, an emailadress of the firm was obtained. The companies who did not have an emailadress on their website were called to ask for their emailadress. Each company received an email with an invitation to participate in this research. 96 Companies were deleted from the sample without being mailed, because of the following reasons: not existing anymore(6), no contact information(10), double mentioned in Orbis/belonging to the same parent company(11), no answer via phone and no website/e-mailadress availabe(8), no supervisory board(6), not an SME(21), not a Dutch firm(30) and not willing to participate(4).

16

is largely based on previously used; well-validated measures (see Appendix I for constructs and sources). Most questions were copied, translated and rewritten in Dutch, because the CEOs were all Dutch. Before conducting the research, several experts1 have been asked to check and test the survey. The final survey has been developed after some minor modifications. The survey was sent to attention of the CEO, consisted of mainly closed questions and took 5-10 minutes to complete. All communication with the companies has been on working days within working-hours, to reduce the possibility of biases caused by different situations or timing.

The survey has been sent to 609 companies by email in May and June of 2014 and was followed up by phone calls. Qualtrics Survey Software has been used to mail the surveys to the CEOs and to gather the responses. 26 Companies replied on the email and were deleted of the sample, because they did not have a supervisory board(8) or they employed more than 250 persons(18). Therefore, the final sample included 583 companies. After sending two reminders via email, the total number of respondents was 97 (16,6% response rate), of which 70 contained sufficient data to be used in this study’s analysis (12,0% final response rate). This response rate is comparable witch earlier research within SMEs, using surveys: Van den Heuvel et.al (2006), 9,2% and Nagel (2009), 10,2%. According to Bourque and Fielder (2003), response rates in the range of 10-20% are common for online surveys. An important element in the success or failure of a survey is timing. The non-response rate of this research may be due to the Holidays. The period in which the survey was sent was covering Ascension Day and Pentecost. During phone calls and responses via email, this reason was often mentioned. The possible significant differences between the early and late respondents have been statistically checked with the Independent T-test to increase the generalizability of this research. No significant differences have been found. By using a high level of standardization in the data collection processes (e.g.: telephone script, sending exactly the same e-mails and reminders at the same time), process variation and possible reliability biases have been reduced.

The total sample included 70 companies, employing on average 85 employees. The age of the firms ranged from 4 to 169 years, with an average of 43 years. Less than a quarter (21,4%) of the companies is a family firm in which the family owns 51-99 percent of the shares (26,7%) or 100 percent (73,3%). Most companies were in the maturity stage (55,7%), followed by the growth stage (30,0%), decline (12,9%) and start-up (1,4%). Most companies (55,1%) were active in the business to business branche. Having a supervisory board was obligatory for most companies (60,3%). Most CEOs were male (81,8%). The companies had 187 board members and 25 of them were female, which corresponds to 13,37% female board members. This is higher compared to the research of (Boersma et. al, 2013), they found that 11,7% of the board members within Dutch SMEs were female. The average board consisted

1

17

of 3 members, with 0(62,1%), 1(32,8%) or 2(5,2%) female board members. Five years ago, most boards included the same number of female board members (75,9%), but 19% included less female board members. 20,7% Of the CEO’s expected that their boards will have more female members in 2019, against 50% keeping the same number, 3,4% would have a lower number and 25,9% did not know the answer. 40,8% Of the CEOs indicated that they currently focus on increasing the gender diversity within their company. The frequency of board meetings ranged from 1 to 10 times a year, with an average of 4 times. The CEO also indicated the ‘most used board role’: the monitoring role was used most often (74,5%), followed by the advisory (23,6%) and the network role (1,8%).

3.1.2. Definition and measurement of variables 3.1.2.1. Independent variables

Gender diversity within the supervisory board. The independent variable of interest in this

study was gender diversity. To identify the gender of the directors, the CEO was asked to give the number of women within the firm’s supervisory board. The board size, defined as the total number of supervisory board members, was also asked. Having these two scores, it became possible to calculate the percentage of women in the board as a new variable. A dynamic aspect was also included, by asking the CEO the number of female board members at three different times: at present, in 2009 and in 2019, measured via four possible outcomes (1:equal, 2:less, 3:more, 4:don’t know). By doing so, increases and decreases could be identified. The use of a multi-period measure allows better control of changes in diversity and increases the reliability (Lückerath-Rovers, 2013). Based on the research of Joecks et al. (2012), this research used a dummy for the majority of female board members(0:less than 30 percent female board members, 1:30 percent or more female board members).

Antecedents of gender diversity. The antecedents were measured in different ways. The

individual aspects of the CEO included: gender(0:male, 1:female), age(measured in years), children (0:no, 1:yes) and attitude. Attitude was measured by three statements based on the article of O’Reilly and Main (2012) about diversity in general and about board diversity. The CEO had to indicate to which level he/she agrees on a 5 point Likert scale from 1:totally disagree to 5:totally agree. Examples of the attitude items are: ‘gender diversity within my supervisory board is important’ and ‘gender diversity within my supervisory board positively influences the reputation of my firm’. A reliability analysis resulted in an acceptable Cronbachs Alpha score of ,84.

18 3.1.2.2 Dependent variable

Firm performance. This study uses objective and subjective measures for firm performance

(general and financial), by using primary and secondary data. Since primary and secondary data both have their limitations, this research combines them to get a more complete view. Lumpkin and Dess (1996) state that research should include multiple performance measures, because focusing on a single measure or a narrow range of performance may result in misleading descriptive and normative outcomes. Combining different measurement instruments from different sources is called triangulation and this increases the reliability and validity of this research. The mixed evidence (see section 2.2) for the relation between gender diversity and financial performance was another reason to include different measures of firm performance.

The primary data was gained via the survey. A difference is made between the satisfaction of the performance of the firm and the performance compared to competitors. The survey included four questions about own firm performance(FirmPerf.CEOown). One question was about the overall satisfaction of the success of the own company and the other questions were more specific questions on satisfaction with growth in revenue and profit (Gundlach, 1998). Examples of these satisfaction items are: ‘overall, I would rate my company as successful’ and ‘I am satisfied with my growth in revenue and profit’. All these questions were measured on a 5 point Likert scale (1:totally disagree, 5:totally agree with a defined neutral point). A reliability analysis resulted in an acceptable Cronbachs Alpha score of ,74. The CEO also had to evaluate the firm’s turnover, profit, market share, customer satisfaction and customer retention compared to its competitors (FirmPerf.CEOcompetitors) via five questions. According to Rust (1999), performance evaluations are more meaningful when they are assessed comparatively and that these evaluations are generally quite consistent with secondary, external published performance information as well as objective performance data internal to the organization. The CEO was instructed to evaluate their firm’s performance related to its competitors on a 5-point scale that ranged from ‘much worse’ to ‘much better’. A reliability analysis resulted in an acceptable Cronbachs Alpha score of ,71.

19 3.1.2.2 Mediating variables

Process and performance. Process includes the intensity of the board roles. Performance

includes the importance and satisfaction of each board role and the overall CEO satisfaction. These mediating variables were measured by statements. The first category of statements was about the intensity, importance and satisfaction of the separate board roles. These three aspects were measured on a 5 point Likert scales: intensity (1:not at all, 5:to a high degree), importance (1:very unimportant, 5:very important) and satisfaction (1:very unsatisfied, 5:very satisfied). Overall satisfaction was calculated by taking the mean score of the satisfaction of the three board roles. A reliability analysis resulted in an acceptable Cronbachs Alpha score of ,81.

The survey also included questions about the most and least used board role and which board roles was of most and least value according to the CEO. A question about the number of supervisory board meetings a year was also included in the survey, to filter out possible ‘paper boards’.

3.1.2.2 Control variables

Focusing on the links between gender diversity in the board and firm performance, there is a need to neutralize potential disturbing factors, which may have an effect on the quality and strength of these relationships. These factors are called control variables and these variables help to increase the accuracy, robustness and reliability. This research includes five control variables: firm industry, firm size, firm age, family firm and life cycle.

Firm Industry. The industry context plays an important role in shaping firm decisions

regarding board diversity (Grosvold et.al, 2007). The survey included a question about the main buyer of the products/services (0:business to business, 1:business to consumer).

Firm Size is measured by the number of employees, via a question in the survey. It is expected

that the larger the firm is, the more women in their supervisory board, because larger firms are more visible to the public and larger firms often have larger boards (Lückerath-Rovers, 2013). The natural logarithm of the number of employees has been calculated and used.

Firm Age. This third control variable is also often used in the current literature. The CEO had

to indicate the firm’s age (in years).

Family Firm. The survey also included a question about the company being a family firm(0:

20

Life Cycle. The CEO was asked to indicate in which stage of development the firm currently is

(1:start, 2:growth, 3:maturity, 4:decline). Fryxell and Lerner (1989) found that older, mature industries have favourable representation of women on boards. A dummy is developed, by combining start and growth (0) and maturity and decline(1).

To get an idea about the CEO’s opinion about the relation between gender diversity within the supervisory board and firm performance (hypothesis 1), two statements about this subject were also included in the survey: ‘gender diversity within a supervisory board causes a better financial outcome’ and ‘gender diversity within my supervisory board causes a better financial outcome’. The scales were anchored with 5 point Likert-type response categories of strongly disagree(1) to strongly agree(5). These outcomes will be discussed in more detail in the conclusions section.

3.2 Analysis plan

This research uses means, standard deviations, correlation and regression analyses and the independent t-test to examine the descriptive statistics and to test the hypotheses. Several items that are used to operationalize the variables were tested with Cronbachs Alpha to ensure the reliability and consistency of the measure. They all scored sufficient: higher than 0,70.

21

4. RESULTS

This chapter includes the findings of this research and consists of two sections. The first section includes the descriptive statistics and correlations. In the second section, the results of testing the four hypotheses and the antecedents are discussed.

4.1 Descriptive statistics and correlations

The means, standard deviations, correlations and coefficients of Cronbachs Alpha of the study variables are presented in Table 2.Regarding the control variables, FirmAge was negatively correlated with GenderMajorityBoard and positively correlated to FamilyFirm. As expected, the intensity of the monitoring role (IntensityMonitoring) correlated positively with the satisfaction of the monitoring role (SatisfactionMonitoring). The two subjective measures of performance (FirmPerf.CEOcompetitors and FirmPerf.CEOown) correlated positively with each other. The subjective measures did not correlate with the objective measures. Table 3 shows the means, standard deviations, correlations and coefficients of Cronbachs Alpha of the antecedents. FirmIndustry was positively correlated to the percentage of females in a company (Ant.Firm.PercentageFemale).

4.2 Hypothesis testing

Independent T-tests have been performed to check the relation between the control variables and gender diversity. Two control variables were significant. First, FirmAge was significant (t(56)=2,66, p=,01). This means that the age of the firms without female board members(M: 51,33, SD:49,37) differs from the age of the firms with female board members(M: 22,36, SD: 16,06). So, older firms have less female board members. Second, FamilyFirm was also significant (t(56)=-2,86, p=,01). This means that family firms without female board members (M:1,72, SD:,45) differ from family firms with female board members (M:2,00, SD:,00).

TABLE 2

Descriptive statistics and correlations for study variables

Variables M SD 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 1.FirmIndustry ,33 ,48 2.FirmSize 4,00 1,16 -,13 3.FirmAge 43,14 41,26 -,32 ,21 4.FamilyFirm ,21 ,41 -,09 ,31* ,37** 5.LifeCycle ,69 ,47 ,03 -,03 ,27* ,13 6.SatisfactionMonitoring 3,93 ,77 ,25 ,16 ,04 ,05 -,07 7.SatisfactionOverall 3,61 ,70 ,24 ,05 ,10 ,18 -,05 ,83** 8.IntensityMonitoring 3,96 1,04 -,04 ,16 ,14 -,17 -,06 ,46** ,32* 9.IntensityOverall 3,27 ,75 -,17 ,20 ,14 -,02 -,15 ,54** ,56** ,67** 10.FirmPerf.CEOown 3,71 ,73 ,34* ,01 -,28* -,06 -,23 ,12 ,17 -,08 ,11 (,74) 11.FirmPerf.CEOcompetitors 3,57 ,55 ,10 ,21 ,03 ,07 -,16 ,01 ,03 -,14 -,07 ,44** (,71) 12.Firm.Perf.ROA -2,77 30,00 ,15 ,18 ,03 -,05 ,13 ,00 ,00 -,13 -,14 ,16 ,01 13.Firm.Perf.EBIT 399,56 1975,09 -,03 -,08 -,12 -,06 ,07 ,17 ,18 ,12 ,10 ,06 ,07 ,29* 14.PercentageFemaleBoard ,14^ ,20 ,19 ,10 -,34* -,32* ,06 ,03 -,10 -,02 -,21 -,04 ,04 ,12 ,12 15.GenderMajorityBoard ,28 ,45 ,22 ,10 -,34** -,28* -,02 ,11 -,03 -,06 -,25 -,02 ,11 ,11 ,07 ,90** Note: Cronbachs Alpha coefficients are on the diagonal in parentheses. N=6,6 *p < ,05, **p <,.01, ^ = fraction

TABLE 3

Descriptive statistics and correlations for antecedents

23

The third regression was significant (90% confidence interval), R2=,29, F(7,38)==2,18, p=,06, but PercentageFemaleBoard or GenderMajorityBoard was not significant. The control variables FirmIndustry and Lifecycle mainly caused the significant outcome of this regression. Therefore, gender diversity and gender majority less than 70% do not affect the CEO’s perception of their own firm performance. The fourth and last regression was not significant, R2=,12, F(7,38)=,74, p=,64. Gender diversity and gender majority less than 70% do not affect the CEO’s perception of the firms’ performance compared to their competitors. These results provide no support for hypothesis 1A and 1B.

TABLE 4

Summary of regression analysis for variables predicting firm performance Dependent variables

Predictor ROA EBIT CEOown CEOcompetitors

Stand Beta t p Stand. Beta t p Stand. Beta t p Stand. Beta t p Constant n.s. 1,14 ,26 n.s. ,83 ,41 n.s. 9,63 ,00 n.s. 8,88 ,00 1.FirmIndustry ,02 ,14 ,89 -,11 -,61 ,55 ,31 2,02 ,05 ,15 ,87 ,38 2.FirmSize -,14 -,82 ,42 -,11 -,63 ,53 -,04 -,29 ,77 ,25 1,48 ,15 3.FirmAge -,12 -,64 ,53 -,10 -,56 ,58 -,26 -1,58 ,12 ,14 ,78 ,44 4.FamilyFirm -,14 -,83 ,41 -,02 -,13 ,90 ,09 ,61 ,55 ,03 ,15 ,88 5. LifeCycle ,00 ,02 ,99 ,08 ,48 ,64 -,28 -1,92 ,06 -,10 -,60 ,56 6.PercentageFemaleBoard ,18 ,47 ,64 ,32 ,82 ,42 ,09 ,27 ,79 ,00 ,01 1,00 7.GenderMajorityBoard -,15 -,39 ,70 -,25 -,64 ,52 -,21 -,62 ,54 ,11 ,29 ,77

Coefficients predictor gender diversity

,10 ,07 ,29 ,12

F ,58 ,38 2,18 ,74

p ,77 ,91 ,06 ,64

Hypothesis 2A stated that gender diversity within boards of SMEs is positively related to the intensity of the monitoring role. Hypothesis 2B stated that gender majority less than 70% within boards of SMEs causes a relatively high intensity of the monitoring role. To test whether gender diversity was positively related to intensity of the monitoring role, a regression analysis was carried out. FirmIndustry, FirmSize, FirmAge, FamilyFirm and LifeCycle were used as control variables. The regression was not significant, R2=,22, F(7,36)=1,46, p=,21. Gender diversity did not affect the intensity of the monitoring role, i.e. hypothesis 2A is not supported. Gender majority less than 70% did not cause a relatively high intensity of the monitoring role and therefore, hypothesis 2B is also not supported.

24

Hypothesis 3 predicted that overall satisfaction of the board roles within SMEs would be a full mediator of the relationship between gender diversity and firm performance. To test this hypothesis, the method of Baron and Kenny (1986) is used. Since the first step, starting with a regression analysis between the independent variable (PercentageFemaleBoard) and the proposed mediating variable (SatisfactionOverall), was not significant (R2=,01, F(1,53)=,51, p=,48), the other steps of the method were not necessary. Therefore, there was no support for hypothesis 3: overall satisfaction is not a mediator of the relationship between gender diversity and firm performance.

Hypothesis 4 stated that the overall satisfaction of the board roles within SMs is positively related to firm performance. Four regression analyses were carried out to test whether satisfaction was positively related to firm performance (ROA, EBIT, CEOown, and CEOcompetitors). FirmIndustry, FirmSize, FirmAge, FamilyFirm and LifeCycle served as control variables. As shown in table 5, the first regression was not significant, R2=-,15, F(6,37)=1,12, p=,37 and the second regression was not also significant, R2=,12, F(6,37)=,83, p=,55. Therefore, overall satisfaction does not affect the firms’ EBIT and ROA. The third regression was significant, R2=,31, F(6,37)=2,78, p=,03, but SatisfactionOverall itself was not significant. The control variables FirmIndustry and Lifecycle mainly caused the significant outcome of this regression. Therefore, satisfaction does not affect the CEO’s perception of their own firm performance. The fourth and last regression was not significant, R2=,14, F=(6,37)1,00, p=,44. Satisfaction does not affect the CEO’s perception of the firms’ performance compared to their competitors. Hypothesis 4 is not supported.

TABLE 5

Summary of regression analysis for variables predicting firm performance Dependent variables

Predictor ROA EBIT CEOown CEOcompetitors

Stand Beta t p Stand. Beta t p Stand. Beta t p Stand. Beta t p Constant n.s. ,57 ,57 n.s. -,65 ,52 n.s. 6,07 ,00 n.s. 6,53 ,00 1.FirmIndustry ,08 ,45 ,66 -,19 -1,1 ,29 ,29 1,84 ,07 ,24 1,37 ,18 2.FirmSize -,16 -,98 ,33 -,11 -,66 ,51 -,08 -,54 ,60 ,27 1,64 ,11 3.FirmAge -,16 -,90 ,38 -,21 -1,2 ,25 -,25 -1,55 ,13 ,15 ,81 ,42 4.FamilyFirm -,21 -1,25 ,22 -,10 -,62 ,54 ,10 ,65 ,52 ,00 ,02 ,99 5. LifeCycle -,06 -,39 ,70 ,11 ,65 ,52 -,31 -2,15 ,04 -,15 -,92 ,36 6. SatisfactionOverall ,10 ,62 ,54 ,30 1,77 ,08 ,08 ,55 ,59 -,11 -,65 ,52

Coefficients predictor gender diversity

,15 ,12 ,31 ,14

F 1,12 ,83 2,78 1,00

p ,37 ,55 ,03 ,44

25

Table 6 shows the outcomes of the regression testing the relationship between the antecedents and the percentage of female board members. This regression was significant (90% confidence interval), R2=,39, F(11,32)=1,84, p=,09. Therefore, the antecedents of gender diversity clearly affect the percentage of female board members. Ant.Ind.Gender and Ant.Ind.Children seem to play an important role in this regression.

TABLE 6

Summary of regression analysis for variables predicting gender diversity Dependent variable Predictor PercentageFemaleBoard Stand.Beta t p Constant n.s. ,51 ,61 FirmIndustry ,14 ,84 ,41 FirmSize ,14 ,85 ,40 FirmAge -,27 -1,56 ,13 FamilyFirm -,20 -1,21 ,24 LifeCycle ,08 ,49 ,63 Ant.Ind.Gender -,31 -2,06 ,05 Ant.Ind.Age ,04 ,25 ,80 Ant.Ind.Children -,28 -1,79 ,08 Ant.Ind.Attitude -,07 -,43 ,67 Ant.Firm.PercentageFemale ,01 ,07 ,95 Ant.Context.Pressure ,19 1,20 ,24

Coefficient predictor gender diversity

F p

,39 1,84 ,09

26

5. DISCUSSION AND CONCLUSIONS

This last chapter covers the conclusions of this research by answering the research question in the first section. The second section discusses the implications and directions for (further) research and also includes the limitations. The last section discusses the practical implications of this research.

5.1 Conclusions

The goal of this study was to present a complete picture of the direct and indirect effects of gender diversity in the boards of SMEs and to identify the causes of gender diversity. Results show that gender diversity within the supervisory board is not related to firm performance, which means that the proposed direct effect is not found. A gender majority less than 70% neither has a positive influence on the performance of the firm. These outcomes are in line with results of other researchers (Farrell and Hersch, 2005; Hillman et al., 2007; O’Reilly and Main, 2012 and Dale-Olsen et al., 2013).

Looking at the indirect effects of gender diversity, it was proposed that gender diversity would be positively related to the intensity of the monitoring role, but no evidence was found for this relationship. This is not in line with the outcomes of Nekhili and Gatfaoui (2012) and Adams and Ferreira (2009), who stated that female directors are known to be tougher monitors. Gender diversity neither did affect the satisfaction of the monitoring role and having a gender majority less than 70% did not make a difference. This is not in line with the ‘critical mass’ theory of Kanter (1977) and Joecks et al. (2012).

It was proposed that the overall satisfaction of the CEO would be a full mediator of the relation between gender diversity and firm performance. Findings showed that this was not the case. This may indicate that female board members do not fulfil other roles than their male colleagues, but it can also be caused by the fact that 62,1% of the companies did not have any female board member, which made it harder to find significant effects of gender diversity.

Results show that satisfaction of the board roles is not significantly related to firm performance. This is not in line with the theory of Nagel (2009), who found that a higher intensity causes a higher added value.

Finally, it was expected that antecedents may influence gender diversity and this relationship was found to be positive and significant (90% confidence interval).Two control variables were important in this research:GenderCEO and ChildrenCEO, they both have a negative influence on gender diversity. So, having a female CEO or/and a CEO with children, decreases the chance of having a gender diverse board.

27

LifeCycle means that firms in the maturity and decline phase are less satisfied with their own financial performance (FirmPerf.CEOown).

To ‘check’ the results of this research, the direct relation was also asked to all CEOs (with and without female board members). Their opinions varied: 30,9% of the CEOs thought that gender diversity within a supervisory board would indeed increase firm performance, 36,4% did not agree and 32,7% gave a neutral answer. When looking at the relationship between gender diversity within their own boards and increasing firm performance, the CEOs were more sceptical: 20,4% agreed, but 42,6% did not agree and 37% answered neutral. These opinions match the outcomes of the regression tests: because no significant relationship was found.

The total number of board members within the Dutch SMEs was 187, with 25(13,7%) of them being a female. Lückerath-Rovers (2013) found that 18% of the supervisory board members of large Dutch firms are women, so SMEs have less female board members. This can be explained by the fact that the Dutch legislation about the gender quota is only obliged for large firms. But despite this, CEOs of the SMEs indicated that they are currently focussing on increasing gender diversity within their firm. 20 Percent of them felt pressure to increase gender diversity and they also indicated that they want to increase the number of female board members, which indicates that gender diversity is indeed an important and contemporary issue, but that the consequences of gender diversity are still ambiguous.

As discussed in chapter 2, there were several reasons to suppose that boards of small and large firms differ. But there were also contradicting indications about the intensity of the board roles. On the one hand, it was expected that the intensity of the monitoring role was lower, because one person is often the owner and manager of an SME and the fact that many SMEs are family firms and therefore, they do not want much interference from the board on control tasks. On the other hand, there were signals that the intensity of the advice role was higher, since owners of SMEs may be entrepreneurs with relatively little management experience. These arguments were all based on the intensity of the roles, but not combined with gender diversity. This research only focussed on the monitoring role, because research done in large firms did combine gender diversity with the intensity of the board roles and found that female board members seem to be tougher monitors. This may again be caused by the differences between males and females. Since the outcomes of this research did not find any significant differences, it might be that males and females (within SMEs) do not differ that much. The stereotypes may not hold for the board members of SMEs.

28

not significant2. In search for a possible explanation of the non-significant outcomes, an additional regression has been done between gender diversity and the importance of the different board roles. These outcomes were interesting. First, a regression has been done between gender diversity and the importance of the monitoring role. The regression did not have any significant outcomes.3 Second, a regression between gender diversity and the importance of the advisory role has been done. This regression was significant, but only because MajorityGenderBoard was significant (see Table 8, Appendix II). Third, a regression between gender diversity and the importance of the network role has been done, and a significant outcome has been found for MajorityGenderBoard (see Table 9, Appendix II). These outcomes show that gender majority negatively affects the importance of the network role. So, having a more gender diverse board lowers the importance of the network role. This may be caused by the diversity itself: having different board members with different backgrounds may automatically generate more ‘networks’ and thereby making the network role less important.

5.2 Implications and directions for research

This research suffers from a number of limitations that could serve as a basis for future work. The first limitation is the small sample size, which may affect the generalizability and reliability of this research. All possible SMEs have been contacted. A higher sample size may be reached by using a different way of data gathering. Sending the survey by post to all companies may increase the response rate. Another option is to involve the Dutch Government in the research and let them collect the data.

The second limitation is the fact that only Dutch companies are included in the analysis. This may lower the generalizability to other countries, because they may have other legislation influencing the (gender diversity within) boards. Therefore, it is suggested to repeat this study in other countries, to check the external validity of this research. By using data from other countries, the sample size will automatically increase. This makes it possible to use all the statistical and analytical tools, to divide the sample into categories (f.e. categories of Kanter (1977) or industry groups) and increase the change of finding significant relations. Next to that, cross-country analyses can be done, to identify possible differences between countries.

The third limitation covers the (short) time period. This research just shows the situation at one moment in time, although one measure (number of female board members) includes several periods. Further studies may concentrate on a longitudinal data, covering a longer time period. It would also be interesting to repeat this research within a few years, to see the possible differences. The reason for not finding significant

2

Gender diversity > intensity advisory role (R2=,27, F(7,36)=1,89, p=,10, but advisory was not significant) Gender diversity > intensity network role (R2=,27, F(7,36)=1,91, p=,10, but network was not significant) Gender diversity > satisfaction advisory role (R2=,24, F(7,36)=1,62, p=,16)

Gender diversity > satisfaction network role (R2=,18, F(7,36)=1,10, p=,39) 3

29

effects now may be influenced by the current period. It was only a year ago that the legislation for gender diversity within boards was implemented for large firms. CEOs of the SMEs indicated that they feel pressure and are focusing on increasing gender diversity. The outcomes may differ if the research will be done a few years from now, when the number of female board members has been increased.

The fourth limitation is about the fact that the subjective and objective financial measures did not correlate with each other. The objective measures (EBIT and ROA) may not have been good indicators of the relative firm’s performance, since the sample included companies of very different sizes, who where active in many different sectors.

The last limitation is about reverse causality. Existing literature mentions this possibility: better performing firms tend to have more female board members (Joecks et. al, 2012). It is also possible that women self-select the better performing firms and that better performing firms are able to focus more on diversity goals (Farell and Hersch, 2005). A longitudinal study will provide the opportunity to measure these influences.

This research used the perspective of the CEO. Although this is often an important person within a company, it might also be interesting to study the perspective of the female board members itself. What are their motives to become a board member? Are they facing any problems? What are their experiences? Further studies can dive into these details, with a more qualitative method to gain a deeper understanding.

Although ChildrenCEO was never measured before, it turned out to be an important control variable in this research: having children negatively influenced gender diversity. Future research could further investigate the role of this variable.

Many CEOs (60,3%) indicated that having a supervisory board is obligatory for their company. This is an interesting fact, since most SMEs are not obliged to have a supervisory board according to the Dutch law. It could be that having a supervisory board is a requirement of certain shareholders. Further research could focus on the CEOs interpretation of ‘being obliged’ to have a supervisory board.

30

5.3 Implications for practice

31

REFERENCES

Adams, R.B. & Ferreira, D. (2009). Women in the boardroom and their impact on governance and performance. Journal of Financial Economics. 94, 291-309

Barney, J.B. (1991) Firm resources and sustained competitive advantage. Journal of Management, 17, 99-120

Baron, R. M., & Kenny, D. A. (1986). The moderator–mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of Personality and Social

Psychology, 51, 1173-1182

Berghe, L.A.A. van den & Levrau, A. (2004). Evaluating boards of directors: what constitutes a good corporate board? Corporate Governance, 12 (4), 461-478

Bertoni, F., Meoli, M. & Vismara, S. (2014). Board Independence, Ownership Structure and the Valuation of IPOs in Continental Europe. Corporate Governance: An International Review, 22(2), 116–131 Bhagat, S. & Bolton, B. (2008). Corporate governance and firm performance. Journal of Corporate

Finance, 14, 257-273

Boersma, M., Martina, D. & Edzes, M. (2013). The Female Board Index in Dutch SMEs 2013 (not published)

Bourque, L.B., & Fielder, E.P. (2003). How to Conduct Self-Administered Surveys. London: Sage Publications

Brammer, S., Millington, A., & Pavelin, S. (2009). Corporate reputations and women on the board.

British Journal of Management, 20, 17-29

Carter, D.A., Simkins, B.J. & Simpson, W.G. (2003). Corporate governance, board diversity and firm value. The Financial Review, 38, 33-53

Daft, R. (1983). Symbols in organizations: A dual content framework for analysis. Organizational

Symbolism. Greenwich, CT: JAI Press, 199-206

Dale – Olsen, H., Schone, P. & Verner, M. (2013) Diversity among Norwegian boards of directors: Does a quota for women improve firm performance? Feminist economics, 19 (4), 110-135

Emans, B., 2004. Interviewing: theory, techniques and training – Designing the Interview Guide. The Netherlands: Noordhoff Uitgevers

Farrell, K.A. & Hersch, P.L. (2005). Additions to corporate boards: the effect of gender. Journal of

Corporate Finance, 11 (1-2), 85-106.

Financing SMEs and Entrepreneurs (2012) An OECD Scoreboard, OECD

Forbes, D. P. & Milliken, F. J. (1999). Cognition and corporate governance: understanding boards of directors as strategic decision making groups, Academy of Management Review, 24, 489 – 505. Fryxell, G.E. & Lerner, L.D. (1989) Contrasting corporate profiles: Women and minority