Searching for Women on Boards: An Analysis from the Supply and Demand Perspective

Manuscript Type. Review. Research Question/Issue. This paper seeks to provide a systematic review of the multidisciplinary theoretical approaches to women on boards in order to understand the factors that hinder and facilitate the access of women to boards, to show the instruments that can be used to promote women to senior corporate positions, and to outline a research agenda suggesting gaps that still need to be filled. Research Finding/Results. Women's access to boards appears to be fragmented in research silos from a variety of areas, lacking a comprehensive view that provides instruments to overcome the barriers hindering the access of women to corporate boards. More in particular, this paper has found very little scientific analysis to understand what instruments can be the most efficient in eliminating barriers for women to reach boardrooms given different cultural environments. Theoretical Implications. This paper aims to create a comprehensive framework for understanding the presence of women on boards and for indicating existing gaps to be filled by new research in the future. This framework will help future researchers in analyzing specific instruments and to measure their efficiency in eliminating gender imbalance. Depending on the approach taken for research, the theoretical backgrounds used vary. While on the supply side the predominant theories are gender role theory, gender self‐schema, and work–family conflict, the demand side is based on gender discrimination, human and social capital theory, resource dependence theory, and institutional environment theory. Practical Implications. This research provides suggestions to typify causes and provide nuanced policy tools to promote women into leadership positions. Future lines of research are proposed to fill the gaps in understanding female representation in top management positions.


INTRODUCTION
I n recent decades, research on women on boards (WoB) has emerged and has resulted in a prolific amount of literature. Researchers have analyzed this issue from different angles, especially through the business case for diversity (Bilimoria, 2000;Dezsö & Ross, 2012;Terjesen, Sealy, & Singh, 2009).
Beyond the business case, contingency theorists advocate moral arguments over research results to promote and improve gender diversity (Gregory-Smith, Main, & O'Reilly, 2014;. Although previous research has helped to place the issue of WoB on the agenda for both practitioners and policy makers, knowledge is still fragmented without exploring endogenous causes. Understanding all the benefits of a diverse board is essential in creating the most efficient tools to achieve equality. Therefore, our research goal is to shed light on the reasons why barriers preventing gender equality at the board level exist and to understand how to tackle such barriers and gauge their level of success. We approach this question through a systematic review of the multidisciplinary approaches of women on boards that involves taking major reviews of literature prior to 2009 (Machold, Huse, Hansen, & Brogi, 2013;Terjesen et al., 2009) and updating them with more recent academic findings. In order to frame our analysis for the barriers confronting women in reaching board level, we have considered board directors as a labor market that could be analyzed from the perspective of supply and demand (Withers, Hillman, & Cannella, 2012). We focus on this perspective to shed light on the causes of inequality for WoB and to help analyze the efficacy of instruments that can be used to overcome the identified issues.
In the realm of WoB, the supply and demand perspective has been mentioned in the literature to explain some empirical outcomes, such as the gender gap in pay (Pucheta-Martínez & Bel-Oms, 2015) or differences in firm performance (Martín-Ugedo & Mínguez-Vera, 2014). Nevertheless, studies that have used some strategy to try to isolate demand and supply factors for the underrepresentation of female directors are more scarce (e.g., Farrell & Hersch, 2005;Gupta & Raman, 2014;Mateos de Cabo, Gimeno, & Escot, 2011).
Supply and demand-side distinctions are essential in diagnosing the causes for the underrepresentation of WoB. On the demand side, companies' decisions, potential for discrimination, and glass ceilings (Powell & Butterfield, 1997) imply that obstacles for WoB are most formidable at the top of the corporate hierarchy. In contrast, a supply-side analysis holds that gender imbalance is ultimately due to female considerations and constraints, such as different values, personal considerations about the family, and career decisions.
In general, there is a lack of research that links the efficiency of a solution instrument to the deconstruction of identified barriers. Supply and demand-side structures support the existing literature by building a holistic framework to understand the gaps. This framework provides researchers with a more comprehensive structure that can identify barriers preventing women from reaching the board as well as potential solution instruments and their relative efficacy. It also informs policymakers in creating appropriate public policy instruments to tackle gender equality in leadership positions. For instance, if a mandatory quota is above the supply of female directors, the outcome will be worse than voluntary quotas (Labelle, Francoeur, & Lakhal, 2015).
The use of the supply and demand-side perspectives also has important prescriptive implications. Understanding why companies increase the number of WoB is critical to solving this problem; the mechanisms used to promote female representation in the pool of board-member candidates are not necessarily the same as those aimed at eliminating or reducing potential bias in the nomination process.
The paper is organized as follows. We first explain the methodology used to select articles and previous literature reviews. In the next two sections, we analyze literature that explores existing barriers for women from both the supply and demand perspectives. The fourth section links previously identified barriers with recent research that identifies instruments to increase the number of WoB from both the supply and demand sides and their efficacy in achieving parity on boards. Finally, we analyze the identified content and methodological gaps in the existing literature and suggest future areas of research.

METHODOLOGY: SYSTEMATIC LITERATURE REVIEW
In order to understand the latest evolution of the analysis around women on boards in academic literature, a systematic literature review has been adopted (Campopiano, unpublished;Newbert, 2007). The process has employed the selection of articles from recent academic journals, specifically published in English peer academic journals since 2009, when the latest comprehensive review on the field by Terjesen et al. (2009) was published.
The selection criterion included peer-reviewed journal articles from January 1965 to May 2015. The first round of article selection was done on Business Source Complete, EconLit, and Behavioral Sciences Collection databases (EBSCO) using the keywords "female directors," "women on boards," and "women directors" in the title or the abstract. This initial search generated 187 articles. We then looked specifically for the main journals treating the topic of women on boards from the perspective of barriers and instruments to solve them. The journals meeting the criteria were Academy of Management Journal, Corporate Governance: An International Review, Harvard Business Review, Journal of Business Ethics, Journal of Management, Journal of Management Studies, Scandinavian Journal of Management and Strategic Management Journal. This second phase brought in 12 more articles relevant to the field. In order to ensure the relevance of the articles, all articles were read completely, checking for substantive relevance by identifying the discussion related to women on boards. Additionally, we included eight articles from sociology and psychology that had been cited in the sampled articles and were relevant to the research objectives.
The increasing participation in this discussion is spearheaded by the journals that have targeted this topic since 2009. These outlets are mainly the Journal of Business Ethics (13 articles), Corporate Governance: An International Review (7) and Strategic Management Journal (7) followed by British Journal of Management (2) and Journal of Management and Governance (3).

BARRIERS TO GENDER EQUALITY ON BOARDS
The underrepresentation of WoB may have supply and demand explanations. On the one hand, supply-side effects such as women's values and attitudes, expected gender roles, or family conflicts, can all result in a relatively limited pool of qualified female candidates for board positions (Bygren & Gähler, 2012;Gregory-Smith et al., 2014). On the other hand, demand-side effects on the side of corporations, such as discrimination, can hinder the progression of women up the corporate ladder.

Supply-Side Barriers
Following Pande and Ford (2011) and Terjesen et al. (2009), we divide supply-side barriers into three different groups: gender differences in values and attitudes, identification with gender role expectations, and work family conflict.
Gender Differences in Values and Attitudes. Potential differences between the genders in terms of values and attitudes can result in motivational differences between men and women to reach top leadership positions (Eagly, 2005). Women have been proven to be generally less hard achievers, less power-oriented (Adams & Funk, 2012) and less powerhungry than men (Schuh, Hernandez-Bark, Van Quaquebeke, Hossiep, Frieg, & Van Dick, 2014). Women sometimes even demonstrate more conservative behavior in boards and managerial decision making (Baixauli-Soler, Belda-Ruiz, & Sanchez-Marin, 2015).
Research that identifies basic differences among gender values and attitudes is mainly grounded in social constructionist theories and individuals' socialization (Weyer, 2007). This can even go further in social constructs, considering that gender differences cannot be assessed by the analysis of the current women managers, as feminine stereotypes might have already been denied. Somehow, these women select a management carrier and show features in terms of needs, values, and leadership roles, analogous to those of men who chase managerial careers (Powell, 1990).
Identification with Gender Role Expectations. Women might self-identify with expected cultural gender roles, creating potentially new, internal barriers to leadership positions. In this process, some females would not attempt to go for top management positions as they conflict with their personal self-image (Eddleston, Veiga, & Powell, 2006;Korman, 1970). It may be that individuals who perceive themselves as more masculine are more likely to aspire to senior management positions than those who identify less with masculinity (Powell & Butterfield, 2013). The theoretical basis for this stream of research is often related to gender self-schemas that are internalized during childhood through gender socialization processes (Greenwald, 1980).
A particular form of identification for women leaders is the stereotype threat, the fear that one's behavior may conform to an existing stereotype and thus may negatively impact performance (Steele & Aronson, 1995). This threat leads, paradoxically, to the confirmation of the stereotype by diminishing performance and lessening motivation to succeed, generating vulnerability and anxiety in female leaders (Hoyt, Johnson, Murphy, & Skinnell, 2010).
Work-family Conflict. Women's commitment to family responsibilities, often labeled as work-family conflict, is probably the most commonly identified barrier preventing women from reaching leadership positions (Greenhaus & Beutell, 1985;Newell, 1993;Wirth, 1998). Women tend to devote more hours than men to family activities yet the same number of hours to work (Eby, Casper, Lockwood, Bordeaux, & Brinley, 2005) and this leads to unequal career opportunities (Straub, 2007). Nevertheless, the WoB literature is inconclusive in confirming that family life is a career barrier for women. Contrary to what is typically expected, some women have been found to not experience work-family conflict, but rather a higher positive spillover than men (Powell & Greenhaus, 2010). On the other hand, fathers have found greater chances of promotion, whereas women's opportunities remain unaffected by motherhood (Bygren & Gähler, 2012).

Demand-Side Barriers
Gender Discrimination. Gender discrimination can be defined as a prejudice or bias based on gender (Becker, 1957). We can identify different types of potential discrimination affecting women in top positions. First, gender can be used to proxy unmeasured specific and differential group characteristics. This might lead to judgments based on average group characteristics rather than individuals, resulting in statistical discrimination (Phelps, 1972). This concept is closely related to mistake-based discrimination, the systematic underestimation of women's skills (Wolfers, 2006). Along the same lines, tastebased discrimination or a preference for male leaders is often ingrained in cultural and social conventions that associate corporate leadership with masculinity (Heilman, 2001;Pande & Ford, 2011). Such cultural norms ascribe socialized characteristics to men and women, shaping expectations on what constitutes appropriate behavior (Eagly, 1987) and can result in a biased promotional system (Hoobler, Wayne, & Lemmon, 2009). Finally, implicit discrimination (Bertrand, Chugh, & Mullainathan, 2005) refers to biases that people may never consciously acknowledge.
In the WoB literature, there is support for gender bias in the appointment of female directors. This has been proven recently through the case study of UK listed companies (Gregory-Smith et al., 2014) and Spanish corporations (Mateos de Cabo et al., 2011).
One particular type of discrimination that explains the low numbers of WoB is tokenism (Kanter, 1977). Tokenism posits that when the presence of different types of persons within a given work group are reduced, those minority members became symbols or "tokens," and they are viewed as representatives of their social category rather than as individuals. Therefore, boards where female representation is zero or low are more likely to appoint a female director (Gregory-Smith et al., 2014). However, WoB do not always comply with the pressures of conformity, and in these cases, tokenism's visibility mechanism does not affect them. Female ratios are positively related with WoB's perceptions to information sharing, social interaction, and influence, giving credence to some kind of isolation and role entrapment mechanism (Elstad & Ladegard, 2012).
Social identity theory (Tajfel, 1972) attempts to explain why some agents, men in this case, tend to be appointed to senior corporate positions. The similarity and identification among members of groups create a division between in-and outgroup members. This implies that in-group members would receive better evaluations, creating a barrier for the out-group individuals to join these networks (Terjesen et al., 2009). Furthermore, group dynamics may hinder the influence of WoB as they may be considered different from the rest of the members, i.e. as out-group members (Carter, D'Souza, Simkins, & Simpson, 2010;Westphal & Milton, 2000;Zhu, Shen, & Hillman, 2014). Additionally, social identity theory addresses the way identity influences interactions among individuals from different groups. In this realm, Chen, Crossland, and Huang (2014) argue that higher levels of WoB influence intra-board social psychological processes increasing decision-making comprehensiveness, which in turn results in more exhaustive evaluations of major strategic proposals. Indeed, within the context of mergers and acquisitions of S&P 1,500 firms, greater female board representation was negatively associated with both overall firm acquisitiveness and target acquisition size.
Finally, the glass cliff (Ryan & Haslam, 2007) is described as the tendency to prefer women for senior positions for organizations in crisis (Haslam, Ryan, Kulich, Trojanowski, & Atkins, 2010). Cook and Glass (2014b) found that women are more likely than men to become the promoted CEO of weakly performing firms and that when firm performance declines during the tenure of women CEOs, they are likely to be replaced by white men. Even in extreme scenarios, when a woman manager asserts directive authority, these implicit biases can make others struggle with it, through hostility and resistance (Eagly & Karau, 2002). In order to escape this criticism, female leaders need to be perceived as effective, but also show strength and sensitivity, while only strength is needed for male leaders (Johnson, Murphy, Zewdie, & Reichard, 2008).
Biased Perceptions of What Women May Bring to the Board. Thorough research finds generalized biased perceptions toward female directors' capabilities, expertise, resources, and networking capacity. These biases manifest into additional barriers for women in leadership. Board selectors usually assume that women lack the adequate expertise or knowledge, i.e., human capital (Becker, 1964;Ragins, Townsend, & Mattis, 1998). Hence, WoB face the stereotype of being underprepared and less effective than their male counterparts (Mensi-Klarbach, 2014;Nielsen & Huse, 2010). Nevertheless, certain literature has a different view, suggesting that rather than less human capital, women have unconventional backgrounds and, therefore, merely have different types of human capital (Hillman, Cannella, & Harris, 2002;Ruigrok, Peck, & Tacheva, 2007). In terms of experience, lack of or reduced board experience is often seen as the main barrier for women (Groysberg & Bell, 2013) when looking for independent directorships. However, it appears that women are more likely to have alternative experiences as directors, for instance, on the boards of NGOs or small firms, but less in top managerial positions. These women tend to compensate their reduced board experience with formal education, leading to higher numbers of women with MBA degrees and international experience (Singh, Terjesen, & Vinnicombe, 2008). In Canada, for instance, women appointed to all-male Canadian boards have a specialized skill, either by being insiders or being specialists (Dunn, 2012).
Another common assumption, based on social capital theory (Coleman, 1988;Loury, 1977) resides on the connections that candidates bring to boards, from inside and outside the firm (Kim & Cannella, 2008). Women's traditional reduced access to networks has been identified as one important problem in accessing leadership positions (Ibarra, 1992;Kanter, 1977;McGuire, 2002;Ragins et al., 1998). One solution is the creation of women's professional networks, but such communities yield fewer leadership opportunities, provide less visibility, and generate less recognition and endorsement (Ely, Ibarra, & Kolb, 2011).
The capacity of influence has also been widely analyzed in the WoB literature. Ingratiation, the interpersonal influence behavior that enhances one's interpersonal attractiveness to become more likeable to specific others (Kumar & Beyerlein, 1991) can be a key element for the potential demand of WoB. The deferential and submissive quality of ingratiation can be of especial importance in environments where personal trust has been and still is essential, as in top management teams (Kanter, 1977). Directors' ingratiatory behaviors toward colleagues could yield board appointments at other firms. This behavior implies relatively refined forms of adulation and conformity. In some cases, women could benefit from more sophisticated tactics that are difficult to be interpreted as manipulative when indulging in ingratiatory behavior (Stern & Westphal, 2010).
Resource dependency theory (Pfeffer & Salancik, 1978) is also often mentioned as a theoretical framework to explore general perceptions about the resources women can bring to a directorship. This theory relies on the idea that female resources could provide critical benefits to the firm that may provide them with greater opportunity to attain top positions. Hill, Upadhyay, and Beekun (2015) found that female CEOs receive higher compensation and are less likely to exit the position than white males. Nevertheless, these perceptions also serve as barriers to women's aspirations, linking the resource dependence role of directors with committees' assignments and gender. Therefore, women are more prone to be appointed to public affairs committees and less to executive committees (Peterson & Philpot, 2007). Having women in charge of audit committee chairs reduces audit fees for corporations, as women seem to improve the effectiveness of internal control activities (Ittonen, Miettinen, & Vähämaa, 2010).
Institutional Environment. Some institutions and their potential rigidities in the external environmental may produce structural barriers for women on corporate boards (Cook & Glass, 2014a;North, 1990). Countries with higher numbers of WoB also show more women in senior management positions, smaller gender pay gaps, and shorter periods of women's political representation . Additionally, institutional elements, such as female presence in the labor market, welfare state attitude toward gender, and the presence of left parties in government, influence gender equality policies (Terjesen, Aguilera, & Lorenz, 2015). Accordingly, cultural and legal institutional systems seem to have a highly significant effect on board diversity (Grosvold & Brammer, 2011) as well as the main actors in charge of the national public policies for WoB (Seierstad, Warner-Søderholm, Torchia, & Huse, 2015).

INSTRUMENTS THAT INCREASE THE NUMBER OF WOMEN ON BOARDS Supply Side
The reduced numbers of female candidates in the pipeline to boards is one of the most highly cited reasons for the implementation of public policies promoting WoB. Although the literature does not normally link barriers with solution instruments, research that analyzes instruments for female leadership in business often assumes particular barriers that women face. Gender differences in values and attitudes might make women less visible than men for leadership promotions. To tackle this problem, research on the topic has already offered instruments to help increase women's visibility and other talent pools (Mateos de Cabo et al., 2011). Among these tools are the presence of role models and mentoring or training programs available to women with professional aspirations. The final set of instruments focuses on policies to reduce work-family conflicts. Policies to achieve a more equal distribution of domestic responsibilities between men and women (e.g., equal and non-transferable paternity leave), have been by far the most widely analyzed solutions in the literature.
Aspirations and Visibility. Visibility encourages and motivates women to attain leadership positions (Pande & Ford, 2011). These initiatives include candidate databases, such as Europe's Global Board Ready Women. The utility and efficacy of such databases were discovered more than a decade ago (Zelechowski & Bilimoria, 2004) but have not been the focus of academic research. Another way to increase the public profile of female candidates is through visibility in the media, although women still seem to be underrepresented across this medium (De Anca & Gabaldon, 2014a;Mateos de Cabo, Gimeno, Martínez, & López, 2014).
Role Models and Mentorship. Role models can strongly influence career development (Gibson, 2003(Gibson, , 2004. Individuals identify themselves with role models by looking for similarities among a large pool of numerous positive examples (Gibson, 2004; with whom they may interact in different moments in time. Motivation to attain positions of power can be enhanced by female role models (Waldman, Galvin, & Walumbwa, 2013) in top positions; they particularly motivate women and help guide their individual development (Gibson, 2003(Gibson, , 2004 in many cases through admiration and idealization (Kelan & Mah, 2014). Professional identification is key for women in business as it helps them to discover role models they could emulate (Sealy & Singh, 2010).
The lack of female role models is often cited as a reason for the low numbers of women in managerial positions (Sealy & Singh, 2010). People usually look for role models similar in gender or race (Eriksson-Zetterquist, 2008;Kelan & Mah, 2014;Sealy & Singh, 2010). Women frequently find themselves having to look externally because of the dearth of role models in their close professional settings (Singh et al., 2006).
Mentors could also be considered eventual role models, but relationships in this case are more rooted in both parties' interactions, usually more limited and longer term (Durbin & Tomlinson, 2014). However, role modeling can be formalized through mentorship or even sponsorship, offering women extra guidance in their professional track.
When men mentor women, it is believed they offer less psychosocial support. When women mentor other women, facilitating integration into the firm's culture might become more difficult, as these women might be less well integrated themselves (Groysberg, 2008) and, in some cases, even considered for fewer positions and receive less mentoring (McDonald & Westphal, 2013). Female mentors usually provide more personal and emotional support, career development guidance, and role modeling identification than men (Fowler & O'Gorman, 2005), and thus more psychosocial help than male mentors (Okurame, 2007). Although senior women are more likely than other women to have mentors, they are still less likely when compared to their male colleagues (Groysberg, 2008). However, any extra help provided by sponsorship to female candidates has been considered as the best way to break the glass ceiling (Hewlett, Peraino, Sherbin, & Sumberg, 2010).
Work-Family Policies. Conciliation seems to be an issue for all working women, but it has a tangible effect on the number of candidates for WoB. Work-family policies are structured as organizational policies and practices to help employees in controlling their work hours and workload (Kossek, Baltes, & Matthews, 2011). Some of the most relevant work-family arrangements by employers are leave to take care of dependants (Den Dulk, Groeneveld, Ollier-Malaterre, & Valcour, 2013) and flexible working hours (Allen, Johnson, Kiburz, & Shockley, 2013). There is general agreement that providing flexible work measures benefits women's work-life structures (O'Neil, Hopkins, & Bilimoria, 2008). Women with university degrees and postgraduate qualifications also must contend with work-family conflicts, sometimes choosing to opt out of the workforce (Leslie & Manchester, 2011), reducing of the number of candidates for WoB. However, taking advantage of these policies is frequently incompatible with climbing up to management positions (Drew & Murtagh, 2005). Career consequences of an employee's decision to use family-friendly policies can be negative (Manchester, Leslie, & Kramer, 2010). In this sense, everyone who uses parental leave policies would be perceived as uncommitted and, therefore, unlikely to be considered for promotions (Leslie & Manchester, 2011). The main issue here is that work-family practices are framed as women's issues, perpetuating the assumption of work-family conflict as an obstacle faced only by women (Leslie & Manchester, 2011). In fact, work-family conflict is cited more frequently for women than for men as the reason behind career withdrawal (Moe & Shandy, 2010).
Given that gender equality programs are intended to encourage and support women in their careers, women are more likely to support these programs while men might perceive them as a threat to their careers (Van den Brink & Stobbe, 2014). The idea that work-family policies are targeted only at women is why these initiatives are looked down upon (Kossek et al., 2011); consequently, men using work-family arrangements might be seen as violating traditional genderbased roles (Leslie & Manchester, 2011).
However, resistance toward equality initiatives is not limited to men; some women perceive little necessity for such changes, especially if beneficiaries of equality programs are confronted with unwarranted doubts about their qualifications (Van den Brink & Stobbe, 2014). The main problem with these benefits is that they must be supported by workplace culture. Can these policies be used without backlash? Users of such programs must also find ways to avoid cultural stigmatization and being seen by one's employer as less committed or performing less effectively (Kossek et al., 2011).

Demand Side
Demand side policies are designed to raise the number of WoB by affecting the behavior of company-hiring directors. The most commonly used tools to fight any kind of discrimination, conscious or unconscious, are affirmative action policies, promoting the presence of the most unrepresented gender, women in the majority of cases. Affirmative action can be differentiated into "equality of outcomes" and "equality of opportunity" (Seierstad & Opsahl, 2011). Equality of outcome refers to hard measures against gender inequality to increase the number of WoB (quotas, targets, or earmarking). Equality of opportunities looks instead to stimulate demand rather than forcing it: soft law initiatives, corporate governance codes, and the "comply or explain" principle (Nielsen & Tvarnø, 2012), or even voluntary quotas or raising awareness (Pande & Ford, 2011).
Legislative gender quotas have been applied to different arenas in the political, social, and economic domains of life (Meier, 2014). The fact is that all over the world, barriers for women remain, and only countries with mandatory quotas have been able to get close to gender equality in the boardroom. Moreover, quotas can provide a way for women to sidestep discrimination (Pande & Ford, 2011).
Quotas are meant to disrupt structural barriers and create endogenous instruments to sustain female recruitment and presence beyond specific numbers. Critical mass theory (Kanter, 1977) shows that when a certain level is reached, the subgroup's degree of influence grows. Some studies have tested this critical mass theory, exploring women's contributions to corporate boards of directors. Achieving this critical mass might affect innovation (Torchia, Calabro, & Huse, 2011), quality of corporate social responsibility (CSR) reporting (Fernandez-Feijoo, Romero, & Ruiz, 2012), or CSR perceptions (Post, Rahman, & Rubow, 2011). Furthermore, women's perceived influence increases when they are more highly represented on the board (Elstad & Ladegard, 2012).
As not all countries have implemented gender quotas for corporate boards, simulations have been run to test their potential effects. Kogut, Colomer, and Belinky (2014) used a simulation to understand what kind and what level of mandatory quotas would be needed in the United States to significantly increase gender equality on boards. Their results suggest that quotas improve women's clustering and connectivity, thus benefitting women's positions. However, without this improvement, women's positions worsen. In general terms, gender quotas force firms to find, "identify, develop, promote, and retain suitable female talent" for their boards (Terjesen et al., 2015: 235). France has also approved a gender quota recently, as the presence of women was much lower than men, but there were similar candidates in terms of education and expertise (Dang, Bender, & Scotto, 2014). Mateos de Cabo et al. (2011) found that the presence of women on Spanish boards also made the appointment of additional women more likely.
However, gender quotas can come with drawbacks, especially if they are analyzed beyond the business case (Seierstad, 2015). They might create more gender diverse boards, but shareholders might suffer if previous male directors are replaced by less competent women just because of the gender quota requirement and other corporate legislation (Ahern & Dittmar, 2012). Due to the reduced number of women in the pipeline, the appointment of less experienced directors may result in poorer performance; women may invest less in their careers given their relatively easier career path, and eventually lead owners and shareholders to reject policies imposed on their boards based just on gender (Pande & Ford, 2011). Some of the effects, initially unexpected, are the creation of a group of women holding multiple seats on boards -"the golden skirts" (Seierstad & Opsahl, 2011).
Gender quotas could also affect companies in unexpected ways. In Norway, one of the negative effects of gender quotas was that some firms ended up with the right organizational form but the wrong board, as women on average were younger and had less board experience (Bøhren & Staubo, 2013). Going even further, some firms exposed to the gender balance law in Norway chose to exit into an organizational form not regulated under the gender quota, citing mandatory regulation on gender balance in the boardroom as lowering the firm's value (Bøhren & Staubo, 2013). Additionally, some chairmen were less satisfied with female board members not contributing positively to perceptions of the board's decisions (Brunzell & Liljeblom, 2014).
Alternatives or complements to mandatory quotas are soft quotas or voluntary efforts to meet quotas, which serve as other examples of instruments for equality of opportunity. In a recent initiative, Australia introduced a soft regulatory approach through recommendations regarding gender diversity policies, creating strong external pressures to conform. This voluntary period approach has been successful in creating more gender diverse boards in Australia (Chapple & Humphrey, 2014), even creating high expectations for the country's future economic development (Galbreath, 2011). In Spain (Lucas-Pérez, Mínguez-Vera, Baixauli-Soler, Martín-Ugedo, & Sánchez-Marín, 2015) and the Netherlands (Lückerath-Rovers, 2013), business cases linking effectiveness of boards to gender diversity have also proven to be effective. Extreme examples have been seen in transitional economies, usually with underdeveloped corporate governance systems and no regulations on gender diversity (Nguyen, Locke, & Reddy, 2014); nevertheless, some positive effects from the business-based argument have been felt (Abdullah, Ismail, & Nachum, 2015), bringing talent into candidate pipelines (Tatli, Vassilopoulou, & Özbilgin, 2012). Meanwhile, in New Zealand, the number of WoB has remained stagnant and may require mandated quotas in order to move forward (McGregor, 2014).
However, in most countries, voluntary approaches to gender equality on boards go hand in hand with corporate governance codes (Seierstad & Opsahl, 2011). Usually, private companies respond to the warning of potential legislated gender quotas by including references to the need for diversity on their boards in their corporate governance reports. The effect of these codes relies on peer pressure among corporations and pressure from stakeholders and media, as corporate governance codes usually do not imply penalties for non-compliance.
High levels of transparency create higher pressure on companies to comply, and ultimately to cope with any kind of biases in selection or treatment. For instance, nomination committees influence gender diversity in boards (Kaczmarek, Kimino, & Pye, 2012). The presence of women can have direct effects on corporate reputation (Bernardi, Bosco, & Columb, 2009;De Anca & Gabaldon, 2014b) as well as CSR actions (Bear, Rahman, & Post, 2010;Boulouta, 2013), and corporate sustainability (Galbreath, 2011). Furthermore, media can have a great impact in breaking stereotypes and influencing decisions. In this regard, journalistic concerns, for example, can lead to less social discrimination in top positions (Park & Westphal, 2013).
Finally, national context idiosyncrasies can play a very important role in establishing roots for the equal presence of women in top managerial positions. In this sense, understanding different national realities from the structural-institutional side (Terjesen et al., 2015) and the different actors involved in launching different national public policies (Seierstad, 2015) can determine the outcome in terms of female presence. Socio-cultural contexts, such as gender parity (Post & Byron, 2015) or Hofstede's cultural dimensions (Carrasco, Francoeur, Labelle, Laffarga, & Ruiz-Barbadillo, 2015) illustrate tolerance for inequality and can be critical for the demands of WoB.

FUTURE DEVELOPMENTS
This research attempts to understand the factors that hinder or facilitate women's access to boards and, for that purpose, we have conducted a literature review on issues and constructs that explain the reasons behind this inequality at board level, using a theoretical framework of supply and demand. Results indicate that, while barriers have been the object of academic research, instruments' efficiency is gauged through case study analyses, making it difficult to draw general conclusions. Results also illustrate that there is more scientific research on the demand side than on the supply side with the exception of the focus on work-family issues.
The paper proposes a more holistic approach using the theoretical framework of supply and demand. Linking supply and demand is complex from an empirical point of view, due to the great difficulty involved in both matching processes (Withers et al., 2012). However, the structure helps to identify existing gaps and thus is vital for building a more comprehensive framework for future research. Therefore our recommendation is to conduct scientific research that can combine supply and demand in the different areas identified. In order to do this, the research should be conducted: (a) differentiating supply and demand to determine in which cases the reasons behind inequality on boards are due to supply or demand factors; and (b) linking barriers with instruments to analyze the effectiveness of those instruments on the identified barriers.
An interdisciplinary approach, based on the supply and demand-side framework can help to move the existing literature forward by identifying the most appropriate methodology for measuring and selecting the appropriate linkages among problems and policies, bringing together different methods, to help policymakers in the design of policies to achieve gender parity on boards.
In addition to the general recommendation, and in more specific terms through the review, it has identified that certain concrete areas are still not sufficiently covered by research and they would benefit from further research. The following sections state those in more detail.

Supply Side
The review has found abundant literature demonstrating concrete gender differences in values and attitudes, identification with gender role expectations and work-family conflicts (see Table 1 for more detail). The review, however, has also found some important gaps. First, no conclusive evidence exists regarding specific attitudes and behaviors of female directors, such as risk aversion; given that this preconception is one of the main obstacles preventing women from reaching leadership positions, it is crucial that it should be properly understood and thoroughly researched. Controlled experiments across two distinct societies, such as the one by Gneezy, Leonard, and List (2009), could play a pivotal role when it comes to testing gender differences in values and attitudes.
Previous studies on potential biases in the selection process of directors only observe the characteristics of successfully appointed candidates (Farrell & Hersch, 2005;Gregory-Smith et al., 2014). Therefore, there is need of an empirical set-up where director candidates as well as the recruiters and final decision makers, could be observed to identify whether the low percentage of female directors is actually rooted in the searching and hiring processes. Future research should look for new methodologies, such as simulations, laboratory experiments, and event studies to help uncover knowledge that may be difficult to obtain through traditional secondary data (Finkelstein, Hambrick, & Cannella, 2009;Joshi, Liao, & Roh, 2011).

Supply-Side Instruments
In the case of supply policies, we see a general scarcity of research linking these solution initiatives with actual improvement in the presence of WoB. Promising future research, as indicated by Withers et al. (2012), could be the integration of economic and social approaches.
The small numbers of women in top positions, as well as the time lag needed for the effectiveness of policy initiatives, has made it difficult to analyze the role of career instruments such as mentoring or sponsorship, as well as the family-work implications for these women. For instance, we need to know whether the inclusion in databases of female candidates increases the likelihood of these candidates to be nominated to a board, or whether mentoring or sponsorship create a real difference in women's career climbs to the top.
We also encourage future research to investigate the impact of different organizational career structures. Organizations that accept and support women will clearly have a competitive edge in keeping their most talented employees (O'Neil et al., 2008).

Demand Side
There is solid literature that identifies the barriers on the demand side in different areas including: gender discrimination, biased perceptions of women's attitudes due to lack of human capital, lack of social capital, and ingratiation (see Table 2 for more detail).
There are several types of discrimination in labor markets; indeed, the characteristics of these different types of discrimination are usually mixed. To prescribe the right instruments to increase WoB, we must first identify the type of gender discrimination that occurs. Other forms of gender discrimination can be based on sociological theories such as status construction theory (Berger, Ridgeway, & Zelditch, 2002) or social dominance theory (Sidanius & Pratto, 2004). Researching the effects of these forms of discrimination on WoB would lead to greater insight into developing effective instruments to increase WoB.
If bias exists at the board level, it would be very useful to propose a model to try to understand the internal psychological/organizational mechanism through which these processes take place. In this sense, structural equation modeling can be very useful to measure and relate constructs and variables than could be part of the game.

Demand-Side Instruments
When it comes to demand instruments, research is still expanding, looking more comprehensively at its applications to the real world. The implementation of quotas or pseudo-quotas at national level can force companies to comply but not necessarily believe. For this reason, analysis of voluntary positive gender policies versus mandatory approaches is needed to understand their potential effects and counter-effects in increasing gender equality (Hillman, Gibson (2003Gibson ( , 2004 Women identification to role expectations  Greenhaus and Beutell (1985) Work and family policies Drew and Murtagh (2005) Content Newell (1993) Wirth (1998) (Continues)   (2002) Terjesen et al.
Methodology instruments: Ryan and Haslam (2007)     . Furthermore, we need to understand how companies interpret and apply these regulations on a day-today basis. Comparisons between countries implementing the same gender quotas will allow researchers to understand where barriers to gender equality originate and to tackle them more efficiently (Hillman, 2015). Finally, the golden skirt phenomenon has been used to criticize the implementation of quota policies. Contrary to what has been found in Norway (Seierstad & Opsahl, 2011), in Italy quotas are not associated with more female members on multiple boards (Profeta, Aliberti, Casarico, S'Amico, & Puccio, 2014). This discrepancy shows us that we must look for methodologies to help us test this assertion. If it is true, then we should ask which part of the phenomenon is due to demand-side (glass ceiling) and what part is due to supplyside factors such as a reduced pool of eligible women.

Cross-Cultural Research
The WoB literature would benefit from a more cross-cultural perspective, analyzing whether the gender gap on boards is due to supply or demand factors and how this varies across cultures. In analyzing the cultural context, it is important to acknowledge that the same instruments may have varying levels of success in different cultural contexts. For instance, Farrell and Hersch (2005) found that the presence of WoB was an obstacle for the appointment of new female directors in the US, but Mateos de Cabo et al. (2011) found the opposite effect in Spain.
Starting from a general approximation, several papers have looked at WoB in different national and cultural environments (Ahern & Dittmar, 2012;Bianco, Ciavarella, & Signoretti, 2015;Mateos de Cabo et al., 2011). However, results do not always point in the same direction. Therefore, more research is needed in cross-or multi-cultural samples, following Terjesen et al. (2015); Seierstad andOpsahl (2011), or Mateos de Cabo, Gimeno, andNieto (2012), or in other new national contexts such as emerging markets (Abdullah et al., 2015).

Contribution to the Literature
By framing the issue around supply and demand-side factors throughout a systematic review of recent research, we have garnered new insights, loops and gaps in the actual research, with a comprehensive holistic framework centered on the issue of WoB. This research framework can be used in future research and can fill in existing gaps in the literature, collectively enabling policymakers to correctly diagnose the issue of inequality and provide efficient instruments to correct existing disparities.