Faculty of Economics and Business
MSc International Business and Management
Master Thesis
“Governance Regime and Gender Diversity in
Corporate Boards of Chinese Listed Firms”
By
Dania Ran Fang (S2042223)
Supervised by:
Kees van Veen
May, 2013
Brouwerstraat 5-7
9712 NA Groningen
! "!
“Governance Regime and Gender Diversity in Corporate
Boards of Chinese Listed Firms”
By Dania R. Fang*ABSTRACT Manuscript Type: Empirical
Research Question: To what extent have Chinese listed firms integrated gender diversity on their board of directors and how can the differences between firms be explained?
Research Findings: From the cross-sectional analysis of 283 Chinese listed firms sampled from Fortune China 500 of 2011, the results indicated that the type of ownership and the industrial orientation of the firms largely determine the gender diversity on the corporate boards. To be exact, state affiliated ownership is negatively related, while service-oriented industry is positively related to gender diversity.
Theoretical Implications: This study is the first to incorporate the concept of governance regime in the attempt to explain gender diversity in boardrooms in general, and gender diversity on the boards of Chinese listed firms in particular. Moreover, the relevance of other gender diversity organizational determinants that are proven to be significant in the Western literature, are discussed and tested in the context of the Chinese firms.
Policy Implications: This study contributes to the discussion of the women presence on corporate boards in transition economies, with emphasis on the importance of a nation’s governance regime. It becomes apparent that policy makers need to be aware of their actions and, especially, in maintaining and enforcing fair gender legislations in leading to the utilization of gender diversity advantages.
Keywords: Gender diversity, Board of directors, Supervisory board, Governance regime, Chinese listed firms.
*Acknowledgement: I would like to foremost thank professor Kees van Veen for his guidance and patience, for
! #! TABLE OF CONTENTS
LIST OF ABBREVIATION 4 CHAPTER 1 INTRODUCTION 5
CHAPTER 2 LITERATURE REVIEW & HYPOTHESES DEVELOPMENT 7
2.1 Governance Regime & Gender Diversity 7 2.1.1 Dominant Ideology 9
2.1.2 National Business System 13 2.1.3 Governance Practices 15
2.2 Firm Size, Industry, Network & Gender Diversity 20 2.2.1 Firm Size 20
2.2.2 Service Industry 21 2.2.3 Network Effect 22
CHAPTER 3 METHODOLOGY & DATA 23
3.1 Sample & Data Collection 23 3.2 Variables & Measures 24 3.3 Data Analysis 28
CHAPTER 4 DESCRIPTIVE RESULTS 29 CHAPTER 5 TESTING HYPOTHESES 32
5.1 Correlation Matrixes 32
5.2 One-way ANOVA Analyses 36 5.3 Regression Analyses 37
CHAPTER 6 CONCLUSION & DISCUSSION 40
6.1 Conclusion 40 6.2 Implications 42
6.3 Limitations & Further Research 43
! $! LIST OF ABBREVIATIONS
BoD - Board of Directors
CCGLC - Code of Corporate Governance for Listed Companies CLPRC - Company Law of the People’s Republic of China CPC - Communist Party of China
CSRC - China Securities Regulatory Commission MCPC - Municipal Committee of People’s Congress NBSC - National Bureau of Statistics of China ! PB - Personal Bureau
PBOC - People’s Bank of China POEs - Private-Owned-Enterprises
QFII - Qualified Foreign Institution Investor
SASAC - Assets Supervision and Administration Commission of the State Council SB – Supervisory Board
SCPRC - State Council of the People’s Republic of China SHSE – Shanghai Stock Exchange
! %! 1. INTRODUCTION
With the introduction of the Upper Echelon theory by Hambrick and Mason (1984), research into the effect of top management team (TMT) composition on organizational performance started to flourish. In the aftermath of large corporate governance scandals1 during the beginning of this decade, more attention has been fixated on the composition and the efficiency of the corporate boards of directors (BoD). Literature suggests that boards with a diversity (e.g gender, nationality, age) of directors can benefit from a broad range of information, perspectives, insights and resources (Carter et al., 2010; Hillman et al., 2000), which can improve the qualities of strategic decisions and monitoring functions (Marimuthu, 2009; Carter et al., 2003; Daily & Dalton, 2003).
Contributing to the diversity in the composition of corporate boards, the presence of female directors and their impact on the boards’ effectiveness has become one of the trending debates in corporate governance (Terjesen et al., 2009). The attention arises from a fairly low number of female directors all over the world (Grosvold & Brammer, 2011; Terjesen & Singh, 2008) and a number of benefits associated with appointing women as directors, such as the access to a fuller range of social and intellectual capital (Daily et al., 1999) and improvement of corporate reputation (Brammer et al., 2007). The benefits are likely to show in the performance of the firm, as some researches revealed a positive link between gender diversity on boards and superior financial performances (e.g. Carter et al., 2003; Erhardt et al., 2003).
As the social- and performance-related benefits of gender diversity on corporate boards take the front seat in past researches (Nielsen & Huse, 2010), less attention is given to the determinants of female presence on boards. Hillman et al. (2007, p.941) point out the importance this, stating that ‘studying individual characteristics and behaviors of female directors can yield insight into how specific women advance into the boardroom, but it cannot answer the question of why some firms have female directors and other do not’. As the authors attempt to distinguish organizational predictors of female representatives on boards, they conducted a panel study of 950 publicly traded U.S firms between 1990 and 2003 and found positive impact of organization !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
1 Such as in the case WorldCom, a relatively weak board of directors granted a risk-seeking CEO substantial amount of loans
! &! size, industry types and network effects (when a firm is linked to other firms with female directors) on the presence of women on BoD. A string of other research has also shown correlations between firm size, board size, industry types and female presence on boards (e.g. Singh et al. 2001; Hyland & Marcellino, 2002; Terjesen & Singh, 2008; Adams & Ferreira, 2009; Brammer et al., 2007; Sealy et al. 2007; Joy, 2008). Some comparative studies on organizational predictors have also revealed a positive impact of firm performance on the presence of women on boards, suggesting that a shortage of supply allows women to self select the firms, or that these firms are able to focus more on diversity goals (e.g. Singh et al., 2001; Farrell & Hersch, 2005). Other determinants mentioned by various research also includes ownership concentration (Kang et al., 2007), presence of institutional stakeholders (Hillman et al., 2002; Browder, 1995; Gillan & Starks, 2000; Singh, 2005), the average age of the board members (Carter et al., 2003) and the presence of employee-elected directors, foreign board directors and young directors (Gregoric et al. 2009).
! '! Taking the limitations of the existing literature into account, this study will address gender diversity on corporate boards in twofold. Firstly, it has an empirical purpose of defining the level of gender diversity on the boards of Chinese listed firms. Secondly, this study aims to explain gender diversity by shedding light on the Chinese governance regime. As the first bid to find organizational determinants of gender diversity on the boards of Chinese firms, this study also incorporates different perspectives that are often used to explain gender diversity in the Western studies. It should be emphasized that the (statistical) information in China on subjects of corporate governance, recruitment processes, and employment in general and gender in specific is less available than in Western countries. Although it is improving, information remains scattered and has yet to be acknowledged by more reliable sources. This study pieces together the available information from various sources in order to understand the whole picture, although many gaps remain.
The research question of this thesis is formulated as follows:
To what extent have the Chinese listed firms integrated gender diversity on their board of directors and how can the differences between firms be explained?
This paper proceeds with a literature review and the development of the hypotheses. The first section of that chapter focuses on defining the Chinese governance regime and how it affects gender diversity on the boards of Chinese firms. This is done through addressing three features of governance regime: dominant ideology, national business system and governance practices. The second part of that chapter deals with organizational determinants that are suggested by the study of Hillman et al., (2007): firm size, industry and network effect. Chapter three elaborates on the methodology and the data used in this study. Chapter four and five describe the descriptive results and the results of the regression analyses, respectively. This paper ends with chapter six, which consists of the conclusion, implications, limitations and suggestions for further research.
2. LITERATURE REVIEW & HYPOTHESES DEVELOPMENT
2.1 Governance Regime & Gender Diversity
! (! structural context that creates different opportunities for different gender groups (Cooke, 2011). The main factors of the structural context that play important roles in determining the gender diversity on boards are embedded in the specific governance regime that a company is subjected to (Van Veen & Marsman, 2008; van Veen & Elbersten, 2008). In defining a country’s governance regime, a framework that captures three interrelated levels of governance regime developed by Maclean et al. (2006) is used (see Figure 1). The detailed features of governance regime touch on issues such as corporate governance code and practice, business structures and ownerships, and dominant ideology about elites, which defines the institutional rules that influence the recruitment processes of directors in direct and indirect ways (Van Veen & Marsman, 2008). Practices such as who decides on quality criteria and who influences the final selection ultimately determine who will be seated on a board and as result; the accessibility of boards to women is affected. For the purpose of defining the Chinese governance regime and its influence on gender diversity on corporate boards, this study will follow the three elements, starting with the least visible element: dominant ideology.
! )! 2.1.1 Dominant Ideology
Dominant ideologies are the underlying ideas, beliefs, values and assumptions on which governance rules and practices are drawn (Maclean, et al., 2006). The culturally defined ideologies play important roles in explaining gender diversity on boards because they constitute the ways a society views gender equality, and therefore affect practices and behaviors that express opportunities and barriers for women in the managerial labor market (Cooke, 2011). This is true for China, as the traditional Chinese society was founded on strong patriarchic principles (Lee et al, 2004). When the Communist Party of China (CPC) came into power in 1949, gender roles in the public domain changed and women were encouraged to take part in education and the labor force. Since the economic reform in 1978, investment in child-care facilities and policies made more education, employment opportunities and resources available for women to progress in the labor market (Jiang, 2000; Yukongdi & Benson, 2005). Such new opportunities for women lead to the development of their entrepreneurship and managerial careers (Cooke, 2005). Women emancipation explains the increasing proportion of women participating in the workforce; however, the Chinese women seem to still be bounded to patriarchic traditional views that undermine their opportunities of excelling to the higher end of the managerial hierarchy (Aaltio & Huang, 2007).
In order to explain the current status of gender diversity on the Chinese corporate boards, the following sections will examine how dominant ideologies have influenced the managerial development of female directors in two aspects: the education and career paths. These two factors play important roles in the recruitment and the functioning of female directors.
Education. Dominant ideologies influence the educational career of female directors in two
areas: the level and the background of education. These areas ultimately determine the fields in which they will excel. In China, women were excluded from education until 1949 when the CPC liberated women and encouraged gender equality in education and workforce (Lee et al, 2004). Since then, the number of female students attending college and university in China has continued to rise steadily2. Women’s educational level in China is often used to explain why the !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
! *+! proportion of women in management is much lower than that of men (Cooke, 2003). However, with the improvement of educational level in the newer generation of women, the inferior education levels will lose some justification for the lack of women on the management level, where university qualifications are usually the requirement for the job entry (Cooke, 2003). Nevertheless, concerning the current female directors who likely have completed their education in the 1980s, the argument may still stand on why fewer women serve on corporate boards today. Although there are no official documents that could provide information on the study choices made by female students in the mentioned period, the cultural and social aspects of the Chinese society could give a good indication of the popular studies chosen by the female students. During the 1980s, the female students were among the first/second generation of college and university graduates. The social acceptance and perception of women in the public domain were still relatively patriarchic and traditional at the time. Viewed as to serve secondary roles in society, women were encouraged to aim for ‘feminine’ occupations that were characterized bysecurity, less demanding job nature, flexible working schedules, child-care facilities and family-oriented sick leave policies (Chow & Ngo, 2011). These characteristics usually lead to occupations such as teaching in kindergarten, primary and middle schools, nursing, secretarial and financial/accounting related jobs, and other welfare related jobs. Thus, they were likely to choose studies that could be related to those fields. Men, on the other hand, were portrayed to fulfill advanced positions in society, such as in the area of science, politics and business. Based on this assumption, it seems that women are more likely to work and, if the level of education and other circumstances permit, excel to the upper echelon in educational, financial, and welfare related fields.
Career Paths. Due to the centrally planned economy system before 1978, virtually all college
and university graduates were allocated to their jobs in State Owned Enterprises (SOEs) by the government (Chow & Ngo, 2011). Once they were allocated to their jobs, the government kept records of superior performances of individuals through assessments and on-job trainings to further promote or re-allocate and train those who had shown greater talent or expertise3. As a !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 2000, they were 41%, 36.1% and 24% and in 2006, 48.06%, 46.36% and 33.87% obtained their bachelor’s, master’s and doctorial degree, respectively, were women (Liu, 2008). !
3 For example, students who had educational backgrounds in finance could be allocated to banks to fill positions as accountants.
! **! more market-oriented economy emerged after 1978, graduates were allowed to find their own jobs in the job market. More employment opportunities had also become available in the private sector (Chow & Ngo, 2011).
Dominant ideologies manifest into several barriers that explain why women tend to cluster at the lower end of the managerial hierarchy. First, women lack motivation and resources to pursue career advancement due to the lack of social expectation and tolerance for it (Hakim, 1996), which is influenced by the traditional view of women’s role in family household. It is difficult for highly educated or established women to find husbands as most Chinese men are not willing to marry a woman who is better educated than they are or have higher incomes (Bowen, 2003). Many couples in China decide that the husband’s career should take precedence, while the women are supportive of their husband’s career and share their success (Korabik, 1993; Zhang et al., 2001; Cooke, 2005). Family responsibilities such as chores, child and elderly care also weaken women’s ability to pursue career advancement (Hewlett et al., 2011). Family is an integral part of women’s identities and is socially unacceptable for a woman to refuse taking on family demands. A clear career gap in career advancement between male and female of the same age group suggest that women’s career advancement is interrupted by the struggle of low salaries that reduce their abilities to afford domestic help and the burdens of housework that prevent them from spending adequate effort and time on their career advancement (Zhang et al., 2001).
! *"! presents particular difficulties for women in China because influential individuals are predominantly men. Such ‘mentor-protégé’ or any close working relationships between a man and a woman can be problematic from a social norm point of view (Zhang et al., 2001).
Thirdly, dominant ideologies have manifested barriers for women’s managerial progression through employment legislations and practices, especially in the government owned sector (Ebrahimi et al., 2001). In general, the legal and constitutional recognition of gender equality has not been followed by public campaigns or legal monitoring mechanisms to ensure the equality in practice (Cooke, 2003). Furthermore, the employment legislation itself contains gender bias views (Borchard, 1995). For one, the age related selection policy4 indicates below the age of 35 as a suitable age group for management trainings and promotions; an age period in which women can easily miss out due to child bearing and caring responsibilities. Secondly, the early retirement policies5 for women to retire five years earlier than men, at the age of 55, affect the duration of managerial positions held by women. Although further documents6 have been issued to prolong the working life of female elites to the retirement age of 60, the opportunity to do so largely depends on their employer.
Further discriminatory employment practices suggest that senior positions in government affiliated institutions and businesses are mostly granted to former government-affiliated officials as rewards for their party member loyalty and for their familiarity and compatibility with the state owned sector (Korabik, 1993). These officials are individuals who usually have served in the army, which was a typically "male bastion" in China (Dalsimer, 1985), as for example Generals, pilots, or technicians. After returning from the army, they have received governmental trainings and educations, and were able to secure jobs as senior directors or managers in state owned units. Through this they obtain their intensive management and industry expertise that help them to expand into the corporate world. Women are less fortunate to obtain such senior positions !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
4 In 1990s, the state employer implemented an age related policy for management training and development. This policy indicates
that talents below the age of 35 years old shall be selected for management training. Potential candidates above the age of 35 shall not be considered for their first promotion. Individuals in junior ranks of management shall not be nominated for further
promotion once they are above the age of 40 for women and 45 for men.
5 Stated in ‘The Ordinance of National Civil Servants for the Present’ established by the Ministry of Labor and Social Security,
female employees, in general, retire five years earlier than their male colleagues in the same occupations: age 50 for blue-collar and 55 for white collar. In addition to this, women are also allowed to take early retirements (5 years earlier): to make rooms for the fast rising younger talents and also for personal desires such as caring for grandchild and elderly parents.
6 The state (Ministry of Personnel) further amended the retirement regulations, in order to retain and utilize expertises more
! *#! through government affiliation because they are likely to remain in the welfare sectors after serving in the army as nurses. Another trait of these former government officials is their credibility, which is achieved by a lengthy period of observation by local government. Such a process for women is more likely to be interrupted by changing their place of residence upon marriage, thus preventing them from accelerating to higher ranks (Judd, 1990). Due to the number of career advancement barriers presented in the state-affiliated sector, many ambitious women prefer and excel in the private sector where it is less bureaucratic and offers more opportunities for advancements (Hewlett et al., 2011).
2.1.2 National Business System
An important aspect of the national business system that concerns gender diversity on the corporate boards of Chinese listed firms is the ownership structure. Different recruitment rules and practices are applied to different type of ownerships (Chen, 2012), and thus, affect the accessibility of the boards for women differently. The co-existence of different ownerships in China began with the Economic Reform in 1978, which involved the corporatization of state-owned enterprises (SOEs) and eventually the privatization of SOEs in the late 1980s (Firth et al., 2002). Many of these firms are listed on the Shanghai (SHSE) or Shenzhen (SZSE) stock exchanges, which were established in 1990s. During the privatization of the SOEs, which are now further divided into wholly state-owned, state holding and state participation companies7 (Holz, 2003), the private sector grew (Qian & Wu, 2003). In 2001, 94% of the listed companies in China were estimated to still have state interests (Xue, 2001).
A typical listed Chinese firm usually has three groups of shareholders: the state8, legal persons9 and domestic individual investors, each accounting for about 30% percent of total shares outstanding (Liu, 2006). With the installment of the Qualified Foreign Institution Investor (QFII) by China Securities Regulatory Commission (CSRC) and the People’ Bank of China (PBOC),
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
7 From a state shareholding point of view, this means that the State Participation firms are subsidiaries of State Holding
companies, and the State Holding companies are the subsidiaries of the Wholly State-Owned firms, with declining government ownership at each hierarchy.
8 The state shares are shares obtained by the government institutions or departments representing the central government when
they invest capital into stock corporations or acquired through legal procedures.
9 The state-owned legal person shares are shares obtained by state-owned legal persons, government affiliated institutions, or
! *$! which provisioned foreign capital to invest in A-shares10 and other Renminbi-denominated securities subject to the grant of investment quotas, foreign shares began to appear and increase among listed firms (Holz, 2003). However, many listed firms do not have foreign shares. Next to foreign shares, there are also employee shares, which is a unique feature of the Chinese stockholding system. The employees of the company collectively own the shares. Most listed firms do not have employee shares and they typically account for a very small fraction of total shares outstanding when they exist. When present, foreign and employee shares combined on average make up less than 10 percent of total shares outstanding (Holz, 2003). Furthermore, institutional investors and shares such as mutual funds11 are relatively new to China (Wei, 2007). In 2007, there are 35 fund corporations estimated. The percentage of shares that can be held by fund corporations in a listed firm is strictly restricted by the prescriptions of the CSRC. Other institutional investors, such as pension funds and insurance companies, are not allowed to directly invest in the stock markets.
Based on the aforementioned characteristics, the ownership of Chinese listed firms seems to be heavily concentrated in the hands of large block shareholders, with 2/3 to be state-affiliated (Nolan, 2004; Liu, 2006). This structure of ownership distribution can diminish women’s opportunities to be appointed as directors on corporate boards because women seem to have several barriers to claim senior positions in state-affiliated institutions and businesses (see chapter 2.1.1, section Career Paths). Moreover, the concentrated ownership condition in emerging markets are not as influenced by capital markets as are the dispersed ownerships in Anglo-Saxon countries (Weimer & Pape, 1999). The increase in ownership concentration is likely to lighten the issue of moral hazard, which can weaken the influence of capital markets analysts, minority shareholders of the firm (Weimer & Pape, 1999; Sabherwal & Smith, 2008) and social pressures (Kang et al., 2007) concern for gender diversity on companies’ boards. Overall, the effect of concentrated ownerships does not seem to favor gender diversity; therefore, Chinese listed firms with dispersed ownerships are expected to have a higher level of gender diversity on their corporate boards than firms with concentrated ownerships.
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
10 A-shares are common stock issued by Chinese firms that are listed on mainland stock exchanges: SHSE and SZSE. These
shares are issued and traded in Renminbi.
11 Mutual funds are a type of investment made a collective investment vehicle which gathers money from many investors to
! *%! 2.1.3 Governance Practices12
Governance practices constitute the most visible layer of governance regime. The recruitment rules, regulations and practices determine the accessibility of boards to women by defining who decides on the quality criteria and who influences the final selection. With the purpose of explaining the recruitment practices of board members and how they affect gender diversity, a standard structure of the Chinese board is briefly explained first.
FIGURE 2: The Standard Board Structure of Chinese Listed Firms
The Board of Directors The Supervisory Board
!!!!!!
!
Chinese Board Structure. With the increase of non-state-owned investments, the concept of a
corporate governance system was introduced to the Chinese listed firms under the Company Law of the People’s Republic of China (CLPRC) in 1993. The law required all listed firms to launch a BoD and a supervisory board (SB) (see Figure 2). Although this set up resembles the board structure of a two-tier system that includes an executive board and a SB (Hall & Soskice, 2001; Wei, 2007), it differs radically from a typical two-tier board because the Chinese SB does not possess any voting rights (Xiao et al., 2004). Due to the ineffective governance role of the !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
12 All information concerning the Chinese BoD structure, the recruitment of directors and recruitment channels derives from
interviews conducted by Chen (2012), unless stated otherwise. The interviewees among the conducted interviews included a former SASAC official and three directors from a SOE, a state holding company and a state participation company, respectively. The transcript of these interviews can be viewed in the Appendix A and B of her thesis “Corporate Governance, Interlocking Directorates and Regulatory Influence of Chinese Listed Companies” (p.68-93).!
! *&! Chinese SB, the Chinese corporate governance system has been also argued to resemble to the Anglo-Saxon’s version of a one-tier board system (Tam, 2002).
The board of directors consists of three segments: 1) the executive board, 2) the non-executive board and 3) the independent board (Chen, 2012). According to CLPRC, only in cases of limited liability companies that are established by two or more SOEs or other state-owned investors, should the executive board includes 1/3 employee representatives. The non-executive board members have their primary occupation outside of the focal firm and serves as part time. In addition, they are also allowed to hold shares in the focal company. The independent board consists of outside directors with expertise in law, finance or accounting. They differ from non-executive directors, as they cannot hold shares in the focal company.
The supervisory board, which is described as a parallel organ of BoD (Cooke, 2011), consists of inside and outside supervisors. The segment of inside supervisors is required, by CLPRC, to include at least 1/3 employee supervisors. Furthermore, outside supervisors are included in order to enhance the independent monitoring functions performed by SB. During regular board meetings, supervisory directors are able to attend, but without the ability to vote.
Recruitment Practices in SOEs. According to the information captured in the interviews
conducted by Chen (2012), SOEs rely on governmental institutions for the nomination, appointment and removal of the directors and tend to explore internal managerial labor market for candidates, while private-owned-enterprises (POEs) do not. These differences imply different level of board accessibility for women, thus resulting in different level of gender diversity between the boards of SOEs and POEs.
! *'!
which consists of academics from research centers, universities and institutions, as well as government officials, and directors and managers of SOEs. Typically, former13 government officials are elected to occupy the seats of the chair of the BoD and the chair of the SB due to their qualifications, credibility, familiarity, and compatibility with the state owned sector. The likelihood that those government officials are women is small (see chapter 2.1.1, section Career
Paths).
FIGURE 3: The Structure of Governmental Institutions and Businesses
Assets Supervision and Administration Commission of the State Council14 (SASAC) is responsible for filling in the rest of the seats on the BoD and the SB in SOEs, with the need of final approval from the local government. However, it is not their job to appoint the employee representatives in both the BoD and the SB. The main purpose of the non-executive directors in SOEs is to monitor the executive board and protect the interest of the government, therefore the non-executive seats are usually occupied by government representatives who are usually current !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
13 After being appointed as a full time board member, they will be officially detached from their government ranks. Nevertheless,
they all remain under the government administration of SASAC.
14 SASAC, which was established in 2003, took on the role of claiming legal liabilities and property rights on behalf of the state
! *(! serving SASAC officers. The elected SASAC officers are likely to be men because women face several barriers in accomplishing senior ranks in governmental institutions and bureaus (see chapter 2.1.1, section Career Paths). The likelihood of a low gender diversity level on the executive and non-executive boards of SOEs seems to be created by the problematic career advancement opportunities for women are in governmental institutions, businesses and bureaus. The CCGLC states that in order to fill the positions as independent directors and outside supervisory directors, individuals are obliged to have expertise in the fields of management, industry, finance, accounting or law. The SASAC, which is in charge of appointing independent directors in SOEs, will look into the PB to recruit qualified individuals. It is expected that women have bigger chance to be appointed as independent directors and outside supervisors due to their professional expertise, than as executive and non-executive directors. Peterson and Philpot (2007) suggest that men and women have different roles in corporate boards, with women less likely to be seated on key committees. While women are less likely to be on executive committees, gender becomes a non-significant factor in the likelihood of being on the nomination, compensation, finance, or audit committees. According to CCGLC, at least one independent director who is selected from the audit committee shall be an accounting professional. As women are estimated to excel, especially in the financial and accounting fields (see chapter 2.1.1, section Education), women have reasonable chances to be appointed as independent directors and outside supervisors.
The Workers Congress of the SOEs is responsible for electing the employee representatives in executive board and inside supervisory board. However, their appointment as directors must be permitted and approved by the local SASAC. The chance of women filling the seats of employee representatives is likely because a relatively large proportion of trade union officials are women15. In China, the union’s role has traditionally been seen as a welfare role for which women are considered better suited than men (Cooke, 2011). The possibility of the presence of female employee representatives in SB is more plausible than in executive board, considering not all executive boards are required to include employee representatives.
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
15 Women workers representation issues in China are relatively homogeneous, particularly given the fact that only one union is
! *)! Overall, the level of gender diversity on the boards of SOEs is expected to be low, with the female directors clustered in the SB. Cooke (2003) states that the government pressure for affirmative action against gender inequality has encouraged governmental organizations to put more women in managerial ranks. However, this has also led to other issues such as giving women the titles without authority. Since SBs have no voting rights in the Chinese corporate governance system, there is a sense of tokenism attached to appointing women as directors.
Recruitment Practices in POEs. Corporate boards of POEs are not affected by the recruitment
practices that are conformed to SOEs, and therefore, are expected to have a higher level of gender diversity. The recruitment process involves the nomination of new directors by the nomination committee, which usually consists of executive and independent directors, and the final appointment of new directors in the general shareholders’ meeting by one-share-one vote. The same process is carried out in state holding and state participation companies, although the central/local government and the SASAC are still responsible for a number of directors and supervisors according to the percentage of state ownership.
Unlike SOEs, private firms rely on their network and the use of public recruitment agencies that targets the mass public to recruit new board members, which eliminate several government-affiliated barriers for women to enter the boardroom. Being more market-oriented than SOEs, more attention is place on the merit of the candidates, which could produces a diverse group of nominees. Furthermore, growth opportunities for women within POEs are more liberal and less burdened by regulations than in SOEs (see 2.1.1, section Career Paths), which means that women have more opportunities to be promoted into the executive board. In addition, private firms, such as family owned, may apply nepotism and appoint entrusted family members such as spouse, children and close relatives to assume key managerial positions (Ralston et al., 2006), which could result more female presence on their boards.
Based on the aforementioned discussion on the effects of governance regime on gender diversity, the following three hypotheses are formulated:
! "+! Hypothesis 1b: The executive and non-executive boards of SOEs have a lower level of gender diversity than the executive and non-executive boards of POEs.
Hypothesis 2: Firms with dispersed ownerships have a higher level of gender diversity on their boards than firms with concentrated ownerships.
2.2 Firm Size, Industry, Network & Gender Diversity
Besides the effects of the governance regime on the gender diversity of corporate boards, there are other alternative organizational factors that can influence the presence of female directors. Following Hillman et al. (2007), the impact of organizational size, industry type and network effects on the likelihood of female presence on the boards will be addressed. The authors used the resource dependence perspective to argue how directors’ gender matches and benefits the needs of organizations. Although the development of these factors are based on developed countries, they will be incorporated in this study to examine their possible affects on gender diversity in the Chinese corporate context. It is important to include these factors in this study because they can provide alternative explanation for the female representation on BoD and eliminate spurious relationship between gender diversity and ownership structures.
2.2.1 Firm Size
Past research has found the significant contribution of firm size to the representation of women on board (Adams & Ferreira, 2009; Hillman et al., 2007). It has been argued that larger firms are often under the most scrutiny by the various stakeholders, as well as the public and therefore face greater pressure to correspond to these expectations of gender diversity. However, this is different in China, as there appears to be no external pressure for firms to consider gender diversity in boardrooms, in order to achieve organizational legitimacy. CCGLC has also not addressed gender as a desirable background for director candidates. Furthermore, governmental promotion of gender equality has not been followed by public campaigns for the furtherance of equality in practice and no legal monitoring mechanisms are executed to ensure the implementation of such legislation in practice (Cooke, 2003).
resource-! "*! based view (e.g. Burke, 1999). It suggests that larger firms are more complex and therefore require a larger set of human capitals on the boards to help with monitoring and better decision-making. Based on this, the following hypothesis is formulated:
Hypothesis 3: Larger firms have a higher level of gender diversity on their boards than smaller firms.
2.2.2 Service Industry
In the western literature, numerous studies present correlations between particular industries and the number of women directors in the firm (e.g. Hyland & Marcellino, 2002; Brammer et al., 2007; Hillman et al., 2007; Sealy et al., 2007; Adams & Ferreira, 2009). Referring to resource dependency theory, Hillman et al. (2007) argued that industries with a larger female employment base are likely to increase the value of having female directors because it can provide a valuable form of legitimacy in the eyes of potential and current employees. Thus, higher gender diversity on the boards is likely to be found in those industries with a large amount of female employees. Furthermore, Brammer et al. (2007) suggested that the correlations between certain industries and presence of female directors could be explained by the close proximity to the final consumers of the firm’s products or services. Due to the diversity among the consumers of products and services, having female directors may be particularly valuable to those companies (Adams & Ferreira, 2009).
! ""! total number of Chinese consumers (Lee et al., 2004). They have tremendous influence on consumption and purchase decisions as they take charge in family buying process, as well as purchase more personal products. Their participation as consumers of products and services could increase the value of having female directors on the boards of service oriented companies. Based on these arguments, the following hypothesis is formulated:
Hypothesis 4: Firms in services have a higher level of gender diversity on their boards than firms in manufacturing.
2.2.3 Network Effect
Hillman et al., (2007, p.945) stated that “other firms are an important source of uncertainty in the external environment of a firm and, as a result, firms form links with each other to reduce this uncertainty and obtain information, communication, and resources.” Such links between firms are formally formed through their director’s home company affiliation and other directors. The authors argued that interlocking directorates could convey gender diversity on focal company’s board and provide additional information and links to other female directors. Because better-networked firms have superior access to scarce information and resources such as qualified director candidates, female directors tend to be dispersed throughout an inter-organizational network.
! "#! the firm’s ownership type, which influences the recruitment process and determines the accessibility of the board to women. As the government and the SASAC are responsible for the appointment of the directors in SOEs, the inter-organizational network may play a lesser role in SOEs than in POEs. In conclusion, the following hypotheses are formulated:
Hypothesis 5a: Firms with outside female director links have a higher level of gender diversity on their boards than firms with no outside female director links.
Hypothesis 5b: POEs with outside female director links have a higher level of gender diversity on their boards than SOEs with outside female director links.
These discussed organizational characteristics will be included in the analysis. Doing so will help to determine whether these factors play a significant role in explaining the differences in gender diversity between firms, and whether the expected governance regime effects discussed in hypotheses 1a, 1b, and 2 are confounded by them.
3. METHODOLOGY & DATA
3.1 Sample & Data Collection
The sample for this study is drawn from the list of Fortune China 500 in 2011. This list consists of Chinese firms that are listed in Shanghai, Shenzhen, Hong Kong, New York, Singapore, London, and Tokyo, and are ranked according to their revenues. In total, their revenues account for 47% of the country’s GDP in 2011. Nevertheless, it is worth noting that this list does not represent the biggest companies in China. Prior to settling for Fortune China 500, Chinese Top 500 ranked by China Enterprise Confederation and China Enterprise Directors Association was considered as it captures a more accurate picture of the big companies in China. However, we quickly came to the realization that data accessibility was going to be a big issue. Seeing Fortune China 500 represents many subsidiaries of listed firms on Top 500 and has greater availability of data, Fortune China 500 was opted. Within the Fortune China 500, 5.8% of the companies are on the list of Chinese Top 500.
! "$! with strictly A-shares were included in the sample. This means that all companies with B shares, H shares or any other forms of foreign shares or securities supplements were excluded from the dataset. Furthermore, companies that have withdrawn from A-shares at the end of the year 2011 were eliminated. Ultimately, the dataset consists of cross sectional data on 283 Chinese listed companies in 2011, with the identification of 4,198 board members in total. These companies spread across 30 provinces, which covers all regions of China.
The information were manually collected from company annual reports of 2011, provided by Chinese Financial website http://finance.sina.com.cn/ and http://www.eastmoney.com/. Additional data on board members were collected from online resources such as
http://finance.china.com.cn, http://finance.sina.com.cn/, http://search.10jqka.com.cn/ and
http://baidu.com.
3.2 Variables & Measures
Gender diversity. Gender diversity is the dependent variable in this study. It is operationalized as the percentage of female directors present in the BoD and SB combined. In the case of hypothesis
1b where only the gender diversity of the executive and non-executive segment of the board16 is
concerned, gender diversity will be operationalized as a percentage of female directors present in the executive and non-executive segment amongst the total number of directors in the executive and non-executive segment of the board.
In order to identify director gender, all consulted company annual reports included pictures of the individual directors next to their function descriptions. Thus, gender ambiguous names proposed no issue in this case. In the dataset, the gender male was coded as 0, and the gender female was coded as 1.
SOEs and POEs (State Affiliated Ownership). For each listed Chinese firms in the database, the
weight of different types of shares in their total share capital were enclosed in their annual reports. From there, it was possible to identify the character of the owners as well as their ownership percentages. In order to distinguish which firms are SOEs and which are POEs, the
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
16 The executive and non-executive directors are put into one joint segment of the board in this study because Annual Reports do
! "%!
percentage of ownerships classified as direct state-owned shares17 and/or state-owned legal person share18 were accumulated to establish the total state affiliated ownership within a firm. As results, total state affiliated ownership within 227 firms ranged from 0.13% to 90.80% with the mean of 44.82%. The remaining 56 firms out of the total 283 firms have 0.00% state affiliated ownership.
Branding SOEs and POEs according to the total state affiliated ownership percentage proposes challenges and confusion. Ideally, SOEs and POEs could be distinguished by using a cutoff at 33.33% of state affiliated ownership. According to the Company law19, it is suggested that a
shareholder is a significant shareholder when it has a share that exceed 1/3 of the total share in the company. This means that firms with state affiliated ownership exceeding 33.33% could be considered to be a SOE. However, according to standard defined by the National Statistic Bureau of China (NSBC), a firm is only considered to be a SOE if it has state control that translates to having more than 50% of the total shares. Such considered SOE falls under the category of wholly state owned enterprise. This raises the question of what is considered to be a POE. As the results of governmental ownership decentralization, the majority of the modern Chinese firms are former SOEs that are likely to endure lingering characteristics of SOEs. Thus using a cutoff of 0% on state affiliated ownership to determine whether a firm is a POE would eliminate any effect of state affiliate ownership, but at the same time overestimate the shares held by state affiliated owners. Inevitably, inserting dummy variables based on the mentioned cutoffs have limitations. Therefore, the operationalization of state affiliated ownership will stay as a percentage, instead of relying on dummy values.
Concentrated Ownership. Ownership concentration is measured by the percentage of biggest
share owned by a shareholder among all substantial shareholders. Relating back to the same theory proposed in measuring State Affiliated Ownership in the last paragraph, firms with biggest share owned by one substantial shareholder that exceeds 33.33% of total share could be !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
17 State-owned shares refers to shares held by governmental agencies or authorized institutions on behalf of the State.
18 Legal person shares refer to shares of a joint stock company owned by another company or institution with a legal person status.
The legal person shares can be indirectly held by the State if the shareholders are State-owned companies.!
19 Article 104 of Company Law states: When a shareholder attends the shareholders' assembly, he shall have one voting right for
! "&! considered to have a concentrated ownership. However, to be consistent with the measurement of
State Affiliated Ownership, percentage is applied in the measurement of ownership concentration and no dummy variable is assigned.
Firm Size. Firm size is measured as the logarithm of total assets and the number of employees.
Many studies have used assets and the number of employees as a measure of firms size, sales is also among the common metrics (Hillman et al., 2007). To avoid false image of the firm size that can be created by one single variable, hence both variables are incorporated in the study.
Service Industry. Earlier studies often use The Standard Industrial Classification (SIC) code
system to differentiate between service-oriented firms (less than 2000 and greater than 3999) and non-service-oriented firms (mainly manufacturing-oriented firms with SIC code from 2000 to 3999) (e.g. Richard et al., 2007). As result, ‘Transportation, Communications, Electric, Gas and Sanitary Services’, ‘Wholesale Trade’, ‘Retail Trade’, ‘Finance, Insurance and Real Estate’ and ‘Services’ industries are usually fall under the category of service-oriented, while ‘Agriculture, Forestry and Fishing’, ‘Mining’, ‘Construction’ and Manufacturing’ are non service-oriented. However, the Chinese classification standardized by the National Bureau of Statistics of China (NBSC) is coded differently. By matching the nature of the industries with previous industries categorized according to the SIC code method, I was able to organize the 12 existing industries in the dataset as follows: ‘Transport, Storage and Post’, ‘Finance and Insurance’, ‘Wholesale and Retail’, ‘Reduction and distribution of electricity, gas and water’, ‘Information Transmission, Computer Service and Software’, ‘Accommodation and Restaurants’, ‘Real Estate’ and ‘Radio, Film and Television’ are categorized as service-oriented industries and are represented by a dummy variable of 1 in the dataset. The remaining industries ‘Manufacturing’, ‘Mining’, ‘Construction’ and ‘Agriculture, Forestry, Animal husbandry and Fishing’ are categorized as non service-oriented and are coded by dummy variable of 0. With the dataset of total of 283 firms, 33.57% of them are service-oriented firms.
Network Effect (1). The network effects hypothesis is tested by measuring the outside female
! "'! 283 companies in the dataset, 159 firms have directors who also serve on other boards. Among these firms, the average outside female director links is 4.25, ranging from 0 links to 20 links.
Network Effect (2). Due to the fact that the network effect variable, which is duplicated from the
study of Hillman et al. (2007), does not take into consideration of the simultaneous influence of ownership types on gender diversity in China, an extra variable will be added to the regression analyses to test the interactive effect of ownership types and network on gender diversity. This variable is operationalized by multiplying the percentage of private ownership with the count of outside female director links.
Board Size. Previous findings yield that larger boards have influence on the number of female directors on board (e.g. Hyland & Marcellino, 2002; Brammer et al., 2007). Hence, the board size is used as a control variable. All listed Chinese firms are conformed to establish a two-tier board. Board size is operationalized as the total number of directors sitting on the BoD (consists of executives, non-executives and independent directors) and SB (consists of inside and outside supervisors) of a firm. In the dataset, the average board size is 15 members, with the maximum size of 26 and minimum size of 8 persons.
Firm Performance. Past studies have reported that women tend to serve on board of better
! "(! Firm Age and IPO Age. The age of a company is added as a control variable because it is
believed to have influence on a firm’s performance (Ericson & Pakes, 1995). In turn, it could have indirect influences on corporate governance as well as female directors attendance on boards. Firm age is calculated by two variables. First, firm age refers to the years (until 2011) since a company is founded. Second, IPO age refers to the years (until 2011) since a company is listed.
Regulated Industry. The regulatory strength of different industries is worth controlling for,
especially in case of Chinese firms. According to Chen et al. (2005), companies that operate in energy, public utilities, finance and telecommunications industries are more regulated than other industries due to their strategic importance to the Chinese national economy. They are protected and nurtured by a comprehensive set of industrial policies installed by the Chinese government. Due to their sensitivity to the state, the firms operating in these particular industries are likely to be under the ownership and control of the state, which will influences the accessibility of the boards to women. This variable is operationalized by categorizing the 12 industries identified in the dataset as follows: ‘Reduction and distribution of electricity, gas and water’, ‘Finance and Insurance’, ‘Information Transmission, Computer Service and Software’ and ‘Radio, Film and Television’ are categorized as regulated industries and are represented by dummy variable 1 in the dataset. The remaining industries are given the dummy variable of 0. According to this categorization, 35 firms of the current dataset are operating in regulated industry.
3.3 Data Analysis
! ")! As for hypothesis 1b, which discusses different population groups (namely only the executive and non-executive segment of the board) than other hypotheses, analyses of variance (ANOVA) were executed on the firm based and individual based dataset.
4. DESCRIPTIVE RESULTS
To start with, it is interesting to see how female directors are distributed through out the firms. The total board seats generated from 283 firms are 4,198 in which female directors occupy 507 (12.08%) of these. Looking closely at how they are distributed, 18.02% of the firms have no female directors on board. The rest is distributed as follows; 31.10% firms have one female director, 23.32% firms have two, 16.61% firms have three, 5.30% firms have four, 3.53% firms have five, and 2.12% firms have six female directors on board. It seems that the Chinese firms have integrated gender diversity to a considerable extent. The 18.02% of which firms have no female directors, is lower than the findings of Hillman et al. (2007) study on 950 publicly traded U.S firms, which yield 44.21% (n=950) of sampled firms had no female directors. It is also lower than the general 25% of European companies that have no women on the board (Heidrick & Struggles, 2011). However, looking closely at the proportion of women on board, China with its 12.08% appears to be lower than Germany (13%), The Netherlands (15%), Denmark (18%), Finland (25%), Sweden (29%) and Norway (33%). While all these countries have a relatively lower percentage of firms with no female directors: Germany (14%), The Netherlands (24%), Denmark (0%), Finland (4%), Sweden (4%) and Norway (10%).
! #+! is interesting to look at the high concentrated ownership in combination with of low percentage of female directors and high state affiliated ownerships, as it could possibly imply a correlation between the three factors.
TABLE 1
Descriptive Results of the Chinese Listed Firms Firms (Mean, N= 283)
0% Cutoff 33.33% Cutoff 50% Cutoff
Variables
SOEs POEs SOEs POEs SOEs POEs
Percentage of women 11.70 14.41 10.47 14.65 9.42 13.88
State affiliated ownership 44.82 0.00 55.40 8.76 63.12 19.68 Concentrated ownership 42.72 40.35 42.52 41.87 44.25 41.05 Firm size_Employees 14,238 12,805 15,281 12,000 17,697 11,713
Firm size_log Assets 10.30 10.06 10.32 10.61 10.38 10.18
Service-oriented industry 0.36 0.32 0.32 0.39 0.30 0.38 Network Effect (1) 2.52 1.82 2.2 2.64 2.10 2.55 Board size 15.30 12.96 15.54 13.85 15.79 14.26 Firm age 14.57 14.34 13.74 15.62 12.65 15.64 IPO age 11.26 8.32 10.78 10.54 9.76 11.23 Tobin’ Q 0.82 1.12 0.90 0.85 0.94 0.84 ROA 3.92 5.50 3.92 5.50 4.59 4.02 ROE 0.08 0.12 0.07 0.12 0.10 0.08 Regulated industry 0.14 0.05 0.14 0.10 0.14 0.11
! #*!
Other variables in table 1 have shown a rather inconsistent mean throughout different cutoffs. Looking at outside female director link, at 0% cutoff, it seems that SOEs have interlocks which leads to more outside female director connections. This is opposite of what is expected because SOEs have a lower gender diversity than POEs. However, this balance completely changes at 50% cutoff when POEs appear to have more outside female director links. Nevertheless, no concrete correlation can be concluded at this stage. Firm age, IPO age and different measures of performance seem to be among other variables that are inconsistent through out different cutoffs. However, their differences are not substantial.
The Chinese governance regime does not only lead to different gender diversity between different types of firms, but it also suggests difference in gender diversity between different board segments within firms due to various talent pools and selection procedures. The data show that out of the total 1,318 supervisory board members, 236 (17.92%) are female directors (see Table 2). This percentage shows that most female directors serve on supervisory boards, independent board comes in at second with 12.93% and executive and non-executive board with the lowest female director presence at 7.34%. This aligns with what is expected as the executive and non-executive segment of the board poses as the most challenging for women to enter.
TABLE 2
Percentage of Female Directors per Board Segment per Type of Firm
Firms (Percentage of female directors and total numbers)
0% Cutoff 33.33% Cutoff 50% Cutoff
Board Segments
SOEs POEs Total SOEs POEs Total SOEs POEs Total
Executives and non-ex. Directors 6.41 (1482) 11.55 (329) 7.34 (1811) 5.25 (1086) 10.48 (725) 7.34 (1811) 4.84 (703) 8.94 (1108) 7.34 (1811) Independent Directors 13.24 (878) 11.52 (191) 12.93 (1069) 12.60 (643) 13.38 (426) 12.93 (1069) 12.44 (426) 13.22 (643) 12.93 (1069) Supervisory Directors 17.18 (1112) 21.84 (206) 17.92 (1318) 14.85 (835) 23.19 (483) 17.92 (1318) 12.84 (545) 21.47 (773) 17.92 (1318) Total 11.58 (3472) 14.77 (726) 12.08 (4198) 10.22 (2564) 14.99 (1634) 12.08 (4198) 9.38 (1674) 13.87 (2524) 12.08 (4198)
! #"! seems that the POEs in this category, with strictly no state affiliated ownership, have a similar amount of female directors present in both executive and non-executive segment and independent segment of the board. However, these patterns do not continue at 33.33% and 50% cutoffs as in both cases, the gender diversity on the independent segment of the board in SOEs decreases and become lower than POEs. Also, differences in gender diversity between executive and non-executive segment and independent segment of POEs become clearer.
5. TESTING HYPOTHESES
5.1 Correlation Matrixes
Before regressions were performed, correlation matrix was carried out. In Table 3, the correlations of all variables can be observed. Hypothesis 1a relates to the level of gender diversity on a firm’s board to the character of the firm that is whether they are SOEs or POEs. The level of gender diversity is represented as the percentage of women on the boards on the 283 companies. To characterize the type of companies, the percentage of state affiliated ownership is used. The correlation between this gender diversity level and state affiliated ownership is significant (r = -.232, p = .000, n = 283), although it is weak in strength. This relation implies that the more state affiliated ownership is present in a firm, the lesser female directors presence there will be. In retrospect, women have more access to the boards of POEs than the board of SOEs. Therefore, hypothesis 1a is accepted.
! ##! tend to be more concentrated in firms that have more state affiliated ownerships, which confirmed what has been expected.
Hypothesis 3 relates to the level of gender diversity with the size of the firm. The size of the firms has been operationalizd in two ways: by the total number of employees and by natural logarithm of total assets. The theoretical discussion has lead to the assumption that larger firms have higher gender diversity. The results in the correlation matrix show other wise; both number of employees (r = -.051, p = .197, n =283) and natural logarithm of total assets (r = -.048, p = .212, n =283) as measurements for firm size seems to be non significant. Thus, hypothesis 3 is not supported.
Hypothesis 4, which links the level of gender diversity with service oriented firms, shows a significant correlation (r = .177, p = .001, n = 283). This finding aligns with previous Western based studies that found higher gender diversity among the boards of service oriented firms (e.g. Brammer et al., 2007; Adams & Ferreira, 2009). Although hypothesis 4 is supported, the relationship is weak.
Hypothesis 5a relates the level of gender diversity to the outside female director links, which tests the network effect. The correlation between these two variables appears to be positive, but it is not significant (r = .012, p = .419, n = 283). Therefore, hypothesis 5 is not supported. Further revealed in correlation matrix, only two of the control variables were found to be associated with gender diversity on boards. These were firm age (r = .192, p = .001, n =283) and IPO age (r = .121, p = .021, n =283), although both relationships are weak.
! "#! TABLE 3: Correlation Matrix of All Variables (Firm based)
Gender diversity
State-affiliated Ownership
Board Size Concentrat ed Ownership Firm Size Employee Firm Size Log. Assets Firm Age
IPO Age Tobin’s Q
ROE ROA Service Oriented Industry Regulate d Industry Network Effect (1) Gender diversity Correlation
Sig. (1-tailed) 1 -.232** .000 -.058 .166 -.208** .000 -.051 .197 -.048 .212 .192** .001 .121* .021 .049 .208 -.059 .161 .039 .257 .177** .001 .039 .255 .012 .419 State-affiliated Ownership Correlation Sig. (1-tailed) -.232** .000 1 .254** .000 .424** .000 .130* .015 .236** .000 -.244** .000 -.021 .363 -.016 .395 -.011 .426 -.014 .406 -.024 .343 .102* .043 -.038 .265 Board Size Correlation
Sig. (1-tailed) -.058 .166 .254** .000 1 -.042 .241 .121* .021 .461** .000 .001 .495 -.056 .175 -.089 .068 .045 .224 -.022 .356 .022 .354 .336** .000 .274** .000 Concentrated Ownership Correlation Sig. (1-tailed) -.208** .000 .424** .000 -.042 .241 1 .074 .108 .048 .209 -.433** .000 -.307** .000 .076 .101 .002 .485 .085 .077 -.149** .006 -.120* .022 -.053 .188 Firm Size Employee Correlation
Sig. (1-tailed) -.051 .197 .130* .015 .121* .021 .074 .108 1 .436** .000 -.189** .001 -.100* .046 -.039 .255 .026 .329 .100* .047 -.121* .021 -.051 .196 .089 .068 Firm Size Log. Assets Correlation
Sig. (1-tailed) -.048 .212 .236** .000 .461** .000 .048 .209 .436** .000 1 -.037 .268 -.073 .110 -.275** .000 .031 .304 -.072 .115 .168** .002 .377** .000 .346** .000 Firm Age Correlation
Sig. (1-tailed) .192** .001 -.244** .000 .001 .495 -.433** .000 -.189** .001 -.037 .268 1 .625** .000 -.188** .001 .009 .439 -.170** .002 .169** .002 .109* .034 .032 .294 IPO Age Correlation
Sig. (1-tailed) .121* .021 -.021** .363 -.056 .175 -.307** .000 -.100* .046 -.073 .110 .625** .000 1 -.174** .002 -.020 .370 -.122* .020 -.025 .339 -.026 .330 -.081 .088 Tobin’s Q Correlation Sig. (1-tailed) .049 .208 -.016 .395 -.089 .068 .076 .101 -.039 .255 -.275** .000 -.188** .001 -.174** .002 1 .108* .036 .667** .000 -.131* .014 -.135* .012 -.141** .009 ROE Correlation Sig. (1-tailed) -.059 .161 -.011 .426 .045 .224 .002 .485 .026 .329 .031 .304 .009 .439 -.020 .370 .108* .036 1 .198** .000 .012 .419 .001 .493 .038 .263 ROA Correlation Sig. (1-tailed) .039 .257 -.014 .406 -.022 .356 .085 .077 .100* .047 -.072 .115 -.170** .002 -.122* .020 .667** .000 .198** .000 1 -.099* .048 -.112* .030 -.041 .246 Service Oriented
Industry Correlation Sig. (1-tailed) .177** .001 -.024* .343 .022 .354 -.149** .006 -.121* .021 .168** .002 .169** .002 -.025 .339 -.131* .014 .012 .419 -.099* .048 1 .478** .000 .094 .058 Regulated Industry Correlation
Sig. (1-tailed) .039 .255 .102* .043 .336** .000 -.120* .022 -.051 .196 .377** .000 .109* .034 -.026 .330 -.135* .012 .001 .493 -.112* .030 .478** .000 1 .245** .000 Network Effect (1) Correlation
Sig. (1-tailed) .012 .419 -.038 .265 .274** .000 -.053 .188 .089 .068 .346** .000 .032 .294 -.081 .088 -.141** .009 .038 .263 -.041 .246 .094 .058 .245** .000 1 N= (283)
