Women’s talents are currently being underutilized at decision-making levels, in particular at the top level. Change is necessary in both the political and corporate world to strengthen Europe’s competitiveness, combat the current economic crisis and create a sustainable future in which all talents are used to the full, and all voices are heard in decisions shaping Europe’s future.—European Commission’s Network to Promote Women in Decision-making in Politics and the Economy (June 2012).

It’s incredible that most boardrooms have one or two women in their boardroom. I am not sure you can leave it to us men to get the change to get that 50 percent representation… I think it needs to be forced on us by law.—Virgin Founder & CEO Richard Branson (May 9, 2013).

Introduction

As illustrated in the above quotes, women have failed to attain equal representation on corporate boards of directors, a concern which has attracted considerable practitioner, policy, and scholarly interest (Economist 2011a, b; Pande and Ford 2011; Torchia et al. 2011; Catalyst 2013; European Commission 2012). Across 67 countries, females comprise only 10.3 % of board directorships, with some of the lowest rates in Morocco (0 %), Japan (0.9 %), and Chile (2.4 %), and some of the highest rates in Norway (42 %), Sweden (28 %) and Finland (27.2 %), and France (22 %) (see Fig. 1). These overall low levels of board representation are surprising as women’s presence on corporate boards is often associated with firms’ higher returns on equity, operating profits, and share prices, as well as greater governance controls and accountability, and better recruitment and retention of women throughout the organization (Bilimoria 2006; Terjesen et al. 2009).

Fig. 1
figure 1

Percentage of female directors on corporate boards in 67 countries (2013). Notes These figures are not directly comparable as the number of firms, particularly large firms, varies country to country, and not all sources count in the same way. Sources Adams and Kirchmeier (2013), European Commission (2012), Gladman and Lamb (2012), and Authors’ own calculations

Full size image

Scholars have investigated a number of individual and firm influences on women’s promotion to board directorships, identifying resource dependency (Hillman et al. 2000), network ties (Westphal and Milton 2000; Arfken et al. 2004; Hillman et al. 2007), and other organizational and industry characteristics (Hillman et al. 2007; Ryan and Haslam 2007). Existing literature examines institutional factors impacting the pre-quota legislation percentages of women on boards, including the proportion of female senior managers, gender pay gap, history of female political representation (Terjesen and Singh 2008), and national economic and cultural environments (Grosvold and Brammer 2011; Adams and Kirchmeier 2013).

Extant research neglects the important role of political institutions, including ruling parties and government legislation, as well as mimetic isomorphism in the adopted governance practices across corporate boards of directors. This is surprising given the extensive worldwide debate and legislation regarding the establishment of gender quotas in boards (Kanter 2012; Reding 2012). Furthermore, once implemented, this legislation generates the most substantial change to the representation of women on boards—far greater than any individual, firm, industry, or country-level factor previously identified (Adams and Kirchmeier 2013). Gender quota legislation has two clear ethical aspects: first, in a pre-legislation environment, women may be underrepresented despite their equal competence; and second, in post-quota legislation, women may be named directors of publicly traded and/or state-owned enterprises even when they are not the most qualified candidates.

Across countries, enacted legislation takes a variety of forms but generally consists of a set gender quota (usually 33–50 %), time period (often 3–5 years), and penalties for non-compliance (e.g., in Spain, any board appointment that violates the quota is considered null; in Norway companies are dissolved). The Norwegian government was the first to establish a 40 % female quota in 2003, for compliance by 2006 for state-owned firms and 2008 for publicly traded firms. Spain established a 40 % female quota in 2007 for compliance by 2015, and only for publicly traded companies with more than 250 employees. Eight other countries/regions with recent quota legislation are Belgium, Finland, France, Iceland, Israel, Italy, Kenya, and Québec. Another 15 countries have included in their respective (often revised) country corporate governance codes the requirement to report gender diversity recruitment efforts and board gender/diversity composition (i.e., under the codes’ “comply or explain” principle). (See Table 1 for a summary of legislation on the 10 countries with board gender quotas and Table 2 for a summary of the 15 country corporate governance code requirements on female board requirements). There are proposals for gender quotas in publicly listed companies in Denmark, Ireland, the Netherlands, South Africa, and Sweden; voluntary targets are in place in several countries including Austria and Poland. In sum, while serious public policy and corporate debates are taking place in many countries, other countries such as Indonesia, Japan, and Mexico have had limited debates of gender quotas in the media, parliament, or other public forums.

Table 1 Countries with gender board quotas
Full size table
Table 2 Countries with codes of good governance that include board gender recommendations
Full size table

Gender quotas force firms to respond quickly to identify, develop, promote, and retain suitable female talent for the corporate board leadership structure. Sanctions range from “soft” penalties such as no consideration for public subsidies and state contracts (Spain) to forcing a non-complying firm to de-list from a particular country’s stock exchange and/or relocate the headquarters to another country (e.g., Norway) (Bøhren and Staubo 2013). While some countries have fallen short of enacting legislation on board gender quotas and sanctions, 15 have introduced explicit principles in their codes of corporate governance on how firms should seek to increase their board diversity. Even though these codes are non-binding, there can be strong normative pressures to comply with them given industry standards or stakeholder expectations. We return to this point in the section on path dependency factors triggering hard law on gender quotas as well as in our “Discussion” section.

Despite the significant size and global scope of the phenomena, there is limited research on gender quotas in the field of corporate governance and business ethics. Taken together with the significant embedded ethical issues, policy implementation, and the corporate responses involved, the question of gender quotas, including their antecedents, represents a substantial knowledge void. Our research asserts that the pattern of countries which have introduced gender quota legislation, regulation, or serious discussion is not random. In particular, we seek to examine: What factors lead governments to legislate gender quotas for corporate boards of directors?

We develop an integrated model of how corporate governance gender quota legislation is influenced by three institutional domains: (1) the female labor market and gendered welfare policies, (2) left-leaning government coalitions, and (3) a legacy of path-dependent gender equality initiatives in the public policy arena as well as in the corporate governance codes. In so doing, our article answers calls for examining the role of government and political institutions in shaping the corporate governance structures and strategies (Aguilera and Jackson 2003). We have also tried to incorporate scholarly recommendations to develop multi-level theories (Hitt et al. 2007), better understand the relationships between business and society (Jones 1983), build theory by combining theoretical lenses (Okhuysen and Bonardi 2011), and bring in new theoretical perspectives when examining board governance issues (Daily et al. 2003), especially related to institutions (Aguilera and Jackson 2003), public governance (Benz and Frey 2007), and policy (Terjesen et al. 2009). Furthermore, recent research highlights the need for systematic comparative research of empirical evidence (Grosvold and Brammer 2011; Terjesen and Singh 2008; Adams and Kirchmeier 2013), including of gender relations and welfare regimes (Orloff 1996) to supplement the mostly single nation studies (e.g., Huse et al. 2009). Thus, we develop three propositions supported by empirical evidence from comparative national case studies around the world.

Theoretical Background: Institutional Context

As organizations are embedded in institutional environments, organizational practices tend to be responses to or reflections of the regulations and structures of the larger environment (Hall and Soskice 2001; North 1990). The adoption of organizational practices and norms co-evolving with institutions might become institutionalized. Institutionalization is the “process by which a given set of units and a pattern of activities come to be normatively and cognitively held in place, and practically taken for granted as lawful” (Meyer et al. 1987, p. 13). Institutionalization entails a certain degree of internalization and cognitive belief in the practice which is quite distinct from the concept of decoupling practices. When existing norms or corporate practices get developed into an enforceable norm, the goal is that that normative practice gets institutionalized by coercive or isomorphic means, and the intention is to minimize decoupling or lack of full institutionalization. We identify three key institutional factors at the country level which explain a great deal why some countries have decided to enact formal legislation on gender quotas in boards. These are existing gender welfare policy to support women’s labor market participation, nature of the political coalition in power, and institutional policy legacies in the effort toward gender equality. We develop our arguments for each of them, offer empirical evidence from selected national cases, and provide a summary of how these three institutional factors apply to our sample of 25 countries (10 with quotas and 15 with code recommendations) in Table 3.

Table 3 Welfare support, political coalitions, and path-→dependent initiatives
Full size table

Gender Policy Issues and the Welfare State

When examining women’s abilities to attain equal representation in the highest echelon of the labor market, it is critical to understand the underlying institutional context. A key part of this institutional context is the nature of the welfare state which it is closely connected to gendered dimensions of maternity leave, childcare, and female labor force participation. Although there are multiple ways to assess welfare states (Van Voorhis 2002), we focus on policies and provisions that are most germane to the issue of gender opportunities and their labor market participation. National welfare states play a major role in determining women’s economic activities, labor market participation, and occupational opportunities (Orloff 1996; Mandel and Semyonov 2006). Indeed, a large body of evidence in political science indicates that countries with more progressive social policies, larger public sectors, and greater benefits in terms of family policies have higher participation of women in the labor market (e.g., Esping-Andersen 1999). Each country grants idiosyncratic welfare “family policy” provisions which are targeted to helping families with children, by providing care for young children (i.e., childcare provision, subsidies to access childcare, etc.), and policies to assist (mostly women) with the balance of work and family (i.e., maternity benefits of leave and pay). Two important assumptions motivating these programs are that women with satisfactory and affordable childcare options will be more likely to return to the workforce (Orloff 1996) and that maternity leave provisions “increase women’s attachment to paid work in the short term” (Gornick et al. 1997, p. 48). Countries vary extensively in terms of the scope and type of family welfare policies.

We expect that a country’s level of welfare provision in terms of family services may be related to the legislation and regulation of gender quotas in corporate boards. Our logic is the following. Women who are in the labor force and in managerial positions are likely to have the ability to also sit on boards, if provided with the opportunity—relative to women who are outside the labor force. Furthermore, countries with greater family policies, especially as related to maternity benefits, are more to be aligned with the logic of gender equality and therefore, more likely to initiate quota legislation and regulation policies to provide opportunities for women to serve on boards, given their overall country cultural and societal values. By contrast, countries with limited family policy provisions are unlikely to develop and enact gender quota policies. Below, we present the case studies of Norway and other Nordic countries, as well as the U.S. and Germany (see Table 3 for the highlights of gendered/family welfare state policies for our entire sample of 25 countries).

One of the best examples of a large, institutionalized welfare state is Norway which has significantly expanded its childcare provision, early childhood education coverage, and maternity/parental leave duration since the mid-late 1980s (Gornick et al. 1997; Henderson and White 2004; OECD 2006). Norway is one of only a handful of countries to offer full wages for maternity/parental leave, since the late 1980s, while the neighboring Nordic countries of Sweden and Finland cut their benefits over the last two decades (from 70 to 80 % of the full wage, respectively) (Henderson and White 2004). The Norwegian childcare provision system is particularly supportive of women with professional careers: childcare services are extensively used by university-educated mothers (41 %) as compared to mothers with only secondary education (21 %) (OECD 2006). Taken together, Norwegian policies appear to enable women to participate actively in the labor market, although the country’s employment rate of mothers with children under three is the second lowest in the OECD, at 18 %, compared to an average of 52.4 % (OECD 2006). Furthermore, Norway counts with labor flexibility with working women; 66.8 % are full-time and almost a third (33.2 %) part-time. The other three continental Nordic countries (Finland, Denmark, and Sweden) share some of the world’s most extensive family policies, granted almost entirely by the state (some portions of Denmark’s provisions are employer-provided) and are among the first to initiate legislation or corporate governance code regulation for gender quotas. For example, although Sweden does not have a gender quota, there is a requirement in its corporate governance code to “strive for equal gender distribution” in the board. Sweden has one of the highest rates of female board representation worldwide. Teigen and Wängnerud (2009) argue that the radical Swedish gender equality discourse may lead to the assumption that the government does not need to introduce formal legal gender quotas.

By contrast, another set of countries has extremely limited parental leave and childcare provisions. For example, among the Organization for Economic Cooperation and Development (OECD) nations, the U.S. has one of the least generous set of policies related to maternity leave and childcare provision (OECD 2006). While 12 weeks of maternity leave is available to most working mothers through the Federal Maternity Leave Act (FMLA), there is significant variance in payment, and there is no set paternity leave. Furthermore, there exists limited childcare provision, particularly for children under the age of five. In the U.S., approximately 90 % of childcare is serviced by the private sector and is mostly paid for by parental contributions. The American female workforce tends to be full-time (81.8 %) rather than part-time (18.8 %) (OECD 2006). The U.S. corporate governance code which is one of the most underdeveloped and poorly institutionalized, counts with a principle to “disclose” whether diversity is a consideration when directors are named, and how the gender inclusion principle is implemented and to what extent it is effective. Other OECD nations with limited family provision policies and no gender board quota or recommendations in their governance codes include Mexico and Slovakia.

In the middle of the spectrum of family policies, Germany has low to moderate family leave policy support with just 14 weeks of maternity leave policies and limited childcare—capable of accommodating only 35 % of children below age 3 (Regierung Online 2011). Germany does not count with a quota legislation but since 2010 Germany’s code of good governance has included recommendations to seek to find “the appropriate degree of female representation” in the management and supervisory boards, as well as to disclose the implementation of such diversity efforts in the companies’ corporate governance annual reports. Taken together, evidence indicates that

Proposition 1

The greater the country’s family policy welfare provision for females in the labor market, the more likely the country is to establish gender quotas for boards of directors.

Left-Leaning Government Coalitions

A rich-comparative political science literature establishes partisan influence as an essential aspect of democracies, especially as the party composition of a government shapes public policy (Schmidt 1996). The foundation of this assumption is a “parties-do-matter” view which conceptualizes politics as a market where governments exchange policies for electoral support (Parsons 1959). Politicians want to achieve and maintain their prestige, power, and income which come with being in office (Downs 1957). Public opinion is more likely to be taken into account if its focus is domestic rather than foreign (Page and Shapiro 1983), although the general public’s influence is often not as powerful as well-organized businesses, employee, and professional groups (Schattschneider 1960). Societies vary in the extent to which the public supports the establishment of equal opportunities for females, including the development of gender quotas. Schmidt (1996) proposes various key elements which support the partisan hypothesis. Taken together, political parties choose policies compatible with office seeking and preferences of their electoral constituencies.

Although, the party systems can be classified variously, and many political issues are multi-dimensional, the political spectrum is most commonly defined by a left-to-right scale of party dualism and measured by social (liberal vs conservative) and economic (interventionist vs laissez-faire) policies. On this spectrum, parties’ positions can easily be constructed from their perspective on social and economic policy. For example, early work describes the right-wing party as concerned with the fight against inflation, while the left-wing party is more concerned about unemployment (Hibbs 1977). Left-wing governments increase the positive effects of shocks on aggregate social expenditure; right-wing governments undertake powerful cutbacks in replacement rates as reaction to structural alteration (Amable et al. 2006). Furthermore, a shift in the left–right party composition of government is generally associated with an adjustment in policy initiatives and legislation (Imbeau et al. 2001). The society’s self-conception of politics varies. In view of the welfare state regime or other historic experiences, political parties considered left in one country might actually pursue politics that are considered to be center or even right in other countries.

For our purposes, we are interested in analyzing how different left–right political inclinations might influence gender corporate policies. In particular, we propose that left-wing parties are more likely to enact legislation of gender board quotas, and we offer three supporting arguments. First, there is a long-standing literature in political science, based on Rueschemeyer et al. (1992) power resources theory, claiming that left-wing party governments tend to be more egalitarian and to also have a greater impact in terms of both policy differences as well as distributional outcomes. This account argues that legislative seats controlled by left-oriented political parties and the strength of the unions are determinants in explaining why some countries have larger social spending and more egalitarian income distribution than others.

A second argument refers to the political opportunity to introduce gender issues in the political agenda and debates of established political parties. On the one hand, in the 1980s and 1990s, some left-wing political parties presented “new values” in their political agendas such as environmental concerns, gender equality, and more recently immigration (Inglehart 1997). These new values are more likely to emerge in affluent countries with organized, corporatist structures of policy making. Here, we see an opportunity to introduce corporate diversity issues in the context of broader societal gender equality. On the other hand, mixed market economies which do not fall naturally into the Hall and Soskice’s (2001) dichotomy of liberal market economies or coordinated market economies, such as Italy, Norway, and Spain, are less dependent on partisan classic demands from labor and management. Left-wing-oriented political parties in these mixed capitalist countries have a greater ability (and possibly greater incentives) to propose and legislate beyond traditional political issues such as employment rights and focus on gender rights (Molina and Rhodes 2007).

A third argument suggesting that left-wing governmental political coalitions might be more inclined to legislate on gender quotas in boards draws on the literature on “equality of opportunity” versus “equality of results,” respectively, associated with left-wing and right-wing parties and public opinion (Borre and Scarbrough 1995). Although, the equality of opportunity is quite salient in the U.S., for instance, by seeking to assure that men and women have equal opportunity to a given job, there are strong arguments drawing on country values and public opinion fostering initiatives and social policies toward equality results (Brooks and Manza 2007; van Oorschot et al. 2008). Gender quotas are a clear example of the logic of “equality of results”—in part, because it is thought that society by itself will not reach it such outcomes.

Turning to the empirical evidence, we investigate the role of the political coalition in the government office at the time that the gender quota legislation was passed in parliament or equivalent legislative body. When categorizing different parties on the left–right spectrum, Conservative and Christian Democratic parties are considered right, whereas Socialist and Green parties are on the left side of the spectrum (excluding extremist parties at both ends). We discuss the country cases of Norway, the Netherlands, and Spain, and the rest of the country cases are summarized in Table 3.

Norway constitutes a special and interesting case of how the process of gender legislation was introduced. Labor party women championed the initiative and developed an alliance with the feminists among the Christian democrats (Sørensen 2011). The proposal was first put forward by the Minister for Gender Equality, Ms. Valgerd Svarstad Haugland of the Christian Democratic Party. Another important actor was Minister of Trade and Industry, Mr. Ansgar Gabrielsen of the Conservative Party who surprised everyone, including his own party, by launching a gender quota proposal in February 2002. This joint action of Labor party women and Gabrielsen eroded oppositional attacks; only Progress Party delegates voted against the reform (Storvik and Teigen 2010). The government in power at the time when the law was passed in 2002 was a minority coalition consisting of the Conservative Party (Høyre [Right in Norwegian]), the Christian Democrats (Kristelig Folkeparti), and the Liberals (Venstre [Left in Norwegian]) (Government Administration Services 2011).

Another country example which fits our left-wing political government logic in the context of soft law is the Netherlands where the issue of gender appointments in the board was introduced in its revised code of good governance in December 10, 2008, by a government which had been in office for 2 years and composed of left-wing parties: Christian Democratic Alliance (Christen-Democratisch Appèl), the Labour Party (Partij van de Arbeid), and the Christian Union (ChristenUnie). Their coalition agreement had already consented to encourage employers to appoint more women in senior positions (Government of the Netherlands 2007). Whereas the Christian Democratic Party (center-right) and the Labour Party (center-left) can be easily classified on the left–right spectrum, the case is more difficult for the Christian Union (representing an example of varying classification depending on distinct issues). This party is considered conservative in social policy issues such as abortion and euthanasia but center-left in regard to economic matters such as welfare state and environment.

Spain fits our arguments quite well, having passed gender quota legislation in 2007. Since 2004, Spain has had a minority government led by the Socialist Workers Party (Partido Socialista Obrero Español, PSOE) under the leadership of President Rodríguez Zapatero. During his first term in office, when the quota legislation was voted in, there was support from the United Left Party (Izquierda Unida) and the Republican Left of Catalonia (Esquerra Republicana de Catalunya). All three parties are on the left side of the political spectrum and are pro quota, supporting our contention. Based on the above, we suggest

Proposition 2

Countries with left-leaning governments are more likely to establish gender quotas for boards of directors.

Path Dependence of Gender Equality Initiatives

In this section, we argue that gender board quotas are also determined by institutional path dependencies and spillover effects. In particular, we propose that gender board policies are both conditioned by existing gendered public policies due to path dependency (inertia) and the political legislative decision to move from soft law (non-binding regulation within codes of good governance principles) to hard law (statutory impositions with penalties for violation). We discuss each in turn.

Path dependence describes the “causal relevance of preceding stages in a temporal sequence” (Pierson 2000, p. 252). At an early stage, different policy paths are equally possible, and there is a wide choice of potential social outcomes. Once a certain path has been chosen, future decisions are significantly influenced by the previous decisions (Greener 2005), and it is often impossible to reverse (Pierson 2000), and if so, only with some minor deviation and in an incremental manner (Greener 2005). The costs of reversing the initial path can be high (Levi 1997). In economic terms, (a) the relative benefits of a current activity increase over time when compared with other possible options and (b) exit costs rise (Pierson 2000). In the case of legislation, a ruling government is likely to lose credibility when introducing legislation which is inconsistent with previous laws. Political institutions and policies are especially change resistant, because they are designed to be difficult to overthrow, as policy-making bodies might want to bind themselves and their successors to achieve “credible commitments” (Pierson 2000). Crouch (2001) supports this by claiming that political and social actors have difficulties departing from patterns set by their predecessors. The concept of path dependency in an international context opposes convergence theory which claims that countries can adapt to market pressures and external factors in different ways. These differences are derived from institutional lock-ins which are resistant to outside pressures. In the case of labor relations, it is evident when distinct dimensions co-vary in their organizational level, diversity increases or remains stable (Traxler et al. 2001). Social scientists have used path dependency theory especially in comparative politics (Pierson 2000), for example, to analyze the development of European party systems in different country members (Lipset and Rokkan 1967; see Kay 2005).

The first path dependency that we explore is that originating from gender quotas in the policy arena. There exists a robust and rigorous literature on the origins and diffusion of gender quotas from the public domain to the private one (Englestad and Teigen 2012; Fagan et al. 2012). We see gender quotas on boards of directors as a prolongation on the existing path of gender quotas on political parties and legislative bodies which started in Western Europe, mostly among a select number of socialist and social-democratic parties, and later extended to green parties and some conservative parties (Krook 2010; Dahlerup 2006). Our logic is the following. Assuming that a higher representation of female parliamentarians leads to more women-friendly policies, one would expect that board quotas are more likely to be found in countries with gender quotas for public elections, or in countries where female representation is high. Thus, a long history of universal suffrage should be related to higher female board representation and therefore more women-friendly policies.

We find evidence that some types of path dependency predict legislation for board gender quotas. Freidenvall et al. (2006) argue that the Nordic countries (Denmark, Finland, Iceland, Norway, and Sweden) achieved the top of the European ranking in terms of women’s representation in parliament due to structural and socio-economic factors. In Nordic countries, women’s empowerment is viewed as an incremental process where gender equality develops gradually and is strongly embedded in the country’s ideology. In Spain, the Netherlands, and Belgium, female representation in parliament is likewise higher than 35 % (European Commission 2012), and gender quotas in politics have either been introduced or are being strongly considered. At the other end of the scale, Greece and Cyprus are among the countries with the fewest women parliamentarians (European Commission 2013), and neither have gender board quotas.

With respect to gender equality laws as a path-dependent precursor to board quotas, Norway is a classic case of a history of previous government decisions which paved the way for gender quota legislation. Although the oldest European constitution (still in effect and dating back to 1814) did not contain many individual rights, it did include the notion of equality. Norway emphasized equality between men and women and was one of the first countries to grant women suffrage, in 1913. Norway’s first major effort to prohibit gender discrimination was the introduction of The Gender Equality Act in 1978 which obligated public authorities and private sector organizations to promote gender equality and account for it on a yearly basis (Ostensen Noss 2006).

In the Spanish case, gender equality initiatives date to the 1960s, although the Institute of the Women (Instituto de la Mujer: IM) was only founded in 1983. By the mid-1990s, two equality plans to promote women had been established, placing gender in business on the public agenda. As the IM was not conceived to develop national gender policies, its main task was to influence other government institutions. Although no actual legislation was passed through IM, the organization paved the way for quota legislation through promoting research on gender issues, providing information, and serving as the contact point for women who had been discriminated (Valiente 1997; González Menéndez and Martínez González 2012). Taken together, these country cases illustrate that the introduction of gender quotas is not random—but rather a consequence of a long-term path-dependent processes related to gender equality.

Another important regulation legacy is the debate and corporate policies on the issue of board diversity which is most prominently exercised through the codes of good governance. Codes are defined as soft law, because they are non-binding, yet firms under their jurisprudence have the legal obligation and normative pressure to either comply or publicly explain why they do not comply. These justifications can be economically or politically driven and tend to be accepted. Although codes are non-binding, they have become critical tools to diffuse effective governance practices cross-nationally as well as to coerce companies to comply with and internalize their recommendations (Aguilera and Cuervo-Cazurra 2004, 2009). As discussed in Aguilera et al. (2013), these codes are also important landscapes to test governance policies. For example, the Kodak Code in Germany tested its recommendation to disclose compensation and once German companies got accustomed to the practice then the German Parliament passed a mandatory reform law on compensation disclosure.

Similarly, in the terrain of board composition, at least 15 codes have an explicit recommendation on gender diversity and its desired goals, although they range from a recommendation to make an effort to recruit and retain female directors to specific goals for gender ratios. In particular, Nordic countries were among the first to initiate gender quota regulative efforts either with hard law or through the codes of good governance code. For example, Finland has a quota of 40 % for state-owned enterprises, and a “comply or explain” principle on gender diversity board composition in their code. Interestingly, the recently amended UK Corporate governance code by the Financial Reporting Council (FRC), which was heavily influenced by Davies’ (2013) investigation and subsequent report on the obstacles to the appointment of women directors, requires listed companies in the London Stock Exchange to develop a policy concerning boardroom diversity. The revised code and the FRC recommend and track that FTSE 100 companies aim for a minimum of 25 % female board member representation by 2015. Yet, Davies (2011, p. 2) warns in his report that if the voluntary approach does not achieve a significant change “government must reserve the right to introduce more prescriptive alternatives.” We expect that countries which have these governance practices, debates, and reporting initiatives in terms of gendering the board will be more likely to take the step of formally legislating quotas:

Proposition 3

Countries with a legacy of initiatives to achieve gender equality are more likely to establish gender quotas for boards of directors.

Discussion

This article has developed an institutional theory approach to examine factors related to board gender quota legislation. We investigated aspects of the institutional environment on the establishment of gender quotas on corporate boards. Our findings suggest several key implications for current debates in theory, practice, policy, and future research which we address below.

Implications for Theory

First, from a theoretical perspective, we present evidence that government and political institutions play a major role in shaping the development of corporate governance regulation specific to gender equality. The enactment of this gendered corporate policy is facilitated and embedded within a particular institutional environment. We find the strongest support for welfare states that promote gendered dimensions of work (i.e., maternity leave and childcare), left-leaning government coalitions, and a legacy of path-dependent initiatives from public sector and corporate domain toward gender equality. Countries with two or three components of this political system model (i.e., Norway, Spain) are most likely to quickly implement a sweeping set of gender quota laws with enforced penalties for non-compliance. By contrast, countries with only one institutional component generally have limited progress of gender quotas. For example, most African countries have neither welfare provision nor a history of path-dependent initiatives, but varying political regimes, and no quotas or gender diversity recommendations in their codes of good governance when they exist. Taken together, this suggests considerable promise for incorporating theory from political science into corporate governance research, a previously neglected area. Furthermore, the diversity of political systems illustrates the importance of examining a comprehensive set of political institutions.

Second, our conceptual arguments also touch on the important nuances that exist between hard law (statutory legislation) and soft law (codes of conduct) (Aguilera et al. 2012). It is conceptually relevant to acknowledge that while soft law is materialized in our context through “the comply or explain” principle in the codes of good governance, this might lead to decoupling practices. That is, firms might be publicly recognized for making efforts toward endorsing a given practice when in fact, they do not fully internalize these efforts. Conversely, hard law, which is typically accompanied by sanctions for non-compliance, entails that whether firms agree or not with a given practice, they are obliged to follow it. Hard law also involves the creation of enforcing agencies. In this article, we have examined the precursors of a country deciding to pass in Parliament (or equivalent) a hard law on the diversity of the board composition which is a significant policy step from incorporating gender diversity recommendation as part of a code of good governance which companies are encouraged to follow. We are able to show that certain characteristics in the political and policy environment must exist in order to enact corporate gender laws.

Implications for Practice and Policy

Our findings also highlight critical implications for practice and policy. First, we identify specific-tipping points in terms of support from an elite or celebrity government or business leaders. For example, the most recent country to initiate gender quotas, the United Arab Emirates, was due to support from Prime Minister Sheikh Mohammed bin Rashid Al Maktoum who tweeted “We have also made a decision to make the representation of women, in all the boards of directors of companies & gov. entities, compulsory” (Al Maktoum in Peacock 2012). In Norway, support came equally from female (e.g., Valgerd Svarstad Haugland of the Christian Democratic Party) and male leaders (e.g., Ansgar Gabrielsen, Minister of Trade and Industry) (Sørensen 2011). Among Norwegian business leaders, early support was granted from powerful women such as Elisabeth Grieg (co-owner of the Grieg Group, and board member of many companies) and Anne Kathrine Slundgård (board member in a number of large Norwegian companies). Furthermore, the majority of 28 female business leaders with board positions expressed a positive opinion about quotas (Dagens Næringsliv 2002), more broadly endorsing gendered policies. This suggests that the “Queen Bee syndrome” of older women in powerful positions who may deliberately hold back initiatives for other, particularly younger, women may not apply in this context (Terjesen and Singh 2008).

Our research indicates that agencies with the agenda of the discussion and possible implementation of board gender quotas may be most successful when targeted at countries with greater welfare state provisions, a left-leaning political coalition in office, and a longer history of gender equality initiatives in political and corporate spheres. Furthermore, initial efforts may be best invested in other related gender equality areas or in supporting left-leaning political parties to set the stage for future gender based board quota legislation.

Future Trends and Debates

We expect an increase in the debates around legislation for women’s presence on corporate boards in at least five domains. First, in Europe and the U.S., women outnumber men as university graduates, including professional degrees in business, engineering, and law, and the employment rates of young women exceed their older counterparts (Bureau of Labor Statistics 2012; European Commission 2012; Eurostat 2012). As women join the labor force in increasing numbers, they may also be expected to be ambitious in terms of reaching the highest echelons. We already observe some signs of this demographic trend. For example, there are several non-government agencies regularly reporting on the presence and development of women on the boards of directors, e.g., the Amsterdam-based European Professional Women’s Network (EPWN). Moreover, many of these efforts are going global, e.g., the UK’s Cranfield Female FTSE Index runs initiatives in the Middle East, Hong Kong, and India; and Catalyst extends outside its U.S. and Canada bases to Europe and India. We also find government bodies involved. In the UK, the government has commissioned initial and follow-up reports by Davies (2011, 2013). At the transnational/regional level, Pollack and Hafner-Burton (2000) note that transnational bodies such as the European Union are shaping country policy. The latest transnational EU initiative in the area of gender quotas, the 2012 EU Directive for Gender Balanced Boards, is a great example.

Second, an increasingly salient area of research is why some countries chose to regulate with hard law, while others stay with soft law. There is emerging research in the global governance field mostly triggered by social responsibility initiatives (such as the United Nations Global Reporting Initiative, or the Kimberly Process) as well as industry quality norms (such as ISOs). This research shows two findings. First, while some countries work quite effectively with socially agreed norms and regulations, others require hard law. Aguilera et al. (2006) address this sharp distinction between two Anglo-Saxon countries in the context of governance regulation. They find that although the U.S. and the UK share the characteristics of common law and similar financial and labor market institutions, when it comes to governance regulation, the UK is a lot more soft law-oriented (starting with the 1992 Cadbury Report of corporate governance), while the U.S. functions with hard laws such as the 2001 Sarbanes–Oxley Act. The argument is that once a norm is endorsed by all parties, it is not necessary to have the stick; the social peer pressure is sufficient to enforce it. This is in a way also the argument that we have unveiled in our discussion with Nordic countries when they are questioned why they do not follow Norway. For instance, Swedish policy makers are closely watching how Norway’s hard law on gender quotas evolves, but they feel that their society does not need to have this immense legal pressure on firms.

The second finding is that in order to avoid institutional arbitrage by having companies engage in certain practices abroad, because regulation is not there or not enforced, there is a movement referred to as global governance in which corporations and stakeholders chose to sign a global code of conduct. Even though these codes are generally hard to enforce, there is pressure from competitors to watch each other and assure that all firms comply.

A third key trend that we have uncovered, albeit varying, is the role of media attention. For example, in Norway, Storvik and Teigen (2010) document that once the Norwegian gender quota was passed, there was limited media coverage of gender quotas. This suggests that while consensus in public opinion and media coverage may have a major punctual impact, once the issue is resolved, or the demands have been served, it disappears from the public agenda and is replaced by more pressing issues. Public interest may also flag when there is “diversity fatigue” attributable to the stall in making progress in increasing women on boards (Branson 2012). Elaborating on this discussion would expand the growing literature on the media’s influence on corporate governance (e.g., Bednar 2012).

Finally, our research has identified a major catalyzing force in Norway, with mimetic isomorphism to other countries which have sought to emulate the Norwegian gender quota policy in order to gain legitimacy for their gender equality policy efforts. This idea of mimetic force is consistent with prior work in corporate governance and other areas of management. For example, the UK’s 1992 Cadbury Report identified a number of recommendations to mitigate corporate governance failures. The report recommendations were subsequently adopted in the entire industrialized world and most emerging markets (Aguilera and Cuervo-Cazurra 2004, 2009). Another example in the corporate social responsibility domain is Norway’s Sovereign Wealth Fund stipulations for investment which have subsequently been adopted by other institutions seeking to invest in socially responsible organization (Vasudeva 2013) and to gain such legitimacy.

Limitations

We acknowledge several limitations of our study which should be considered in the future as gender regulation evolves, and we can more systematically study more cases, and we have more data about its debates and negotiations. First, we provide a detailed examination of only a handful of countries—particularly those with comprehensive legislation, some regulation, or serious discussion of gender quota issues. Although our sampling choice was deliberate to provide deep insights into the respective contexts, other countries, particularly in Africa, Asia, and Latin America, should be studied in detail, especially with the help of indigenous scholars in these countries who can offer full perspectives on the national context.

Second, our research examines institutional factors at the national institutional level. Our conceptual framework could be extended to a city or province level. For example, in Canada, the Québec case illustrates that there may be specific initiatives in municipalities or provinces which are distinct from the overall country level. Other examples include the German cities of Berlin and Nuremberg which have both initiated efforts to increase gender representation on municipal-owned companies (and subsidiaries) to 50 and 40 %, respectively (Pande and Ford 2012), in contrast to the rest of Germany. Investigating institutions at the regional level would extend recent work in economic geography.

Future Research Directions

In closing, we would like to suggest seven promising directions for future research. First, scholars could consider the evolving nature of the issue of gender and diversity more generally as corporate governance codes get revised and updated. The UK latest code revision in May 2013 shows a clear trend toward making board diversity issues not only more explicit but also more accountable. Prior work in comparative corporate governance (e.g., Aguilera and Jackson 2003) could be extended longitudinally to consider the overall corporate board model (e.g., Anglo-American), number of board tiers, role of scandals, prior history of reforms; and possible extensions to other organizational types (e.g., small and medium-sized and new companies which traditionally have disparate levels of female representation) (Smith et al. 2006). It may be that gender quotas evolve simultaneously or lag or lead other corporate governance codes. Furthermore, there is the possibility that some quota proposals and code revisions may be revoked in the future, for various reasons, ranging from gender standards which have been met or a falling out of favor with the ruling political coalition.

Second, further research could include a broader institutional perspective to examine the role of business institutions such as bonus/pay caps, and tax incentives. Alternatively we could adopt an institutional work perspective and explore how corporations and stakeholders such as women’s consumer associations shape gender quotas.

Third, our findings uncover considerable variation in the gender quota (e.g., 33–50 %), time period (e.g., immediately although with no sanctions to 8 years), organizations (e.g., state-owned enterprises and/or public traded firms, especially large in terms of employees and revenues), and sanctions. It would be interesting to follow and study the policy debates and negotiations with the interested parties involved in the process and better understand the stakeholder–government negotiations which led to the final legislation.

Four, future research could explore what factors might help countries to pass this legislation once firms have experimented with it. This would be a case where bottom-up development, experimentation, and adjustment of corporate practices get transplanted to the policy level. In fact, in the course of our research, we identified numerous cases of corporate firm support and exposure to firm-level gendered policies. For example, Deutsche Telekom (a DAX-30 German company) established an internal gender quota goal of 30 % female senior and middle management (Clark 2010). Most recently, Japanese retailer Aeon sets a goal of 30 % female managers by 2016 and 50 % by 2020, up from the current share of approximately 10 % (Japan Times 2013).

At an institutional level, California retirement institutions have taken action to ensure diversity on the boards of the companies that they invest in (Trautman 2012). These cases illustrate the possibility in the future that change may come from bottom-up, through corporations. Another recent example, aimed a different set of players in the corporate sphere, is the May 2013 UK resolution requiring that executive search firms must have a voluntary code of conduct in terms of reporting gender in the search process (Davies 2013). This requirement will certainly make search firms more aware of diversity efforts.

Five most existing quotas affect only publicly traded and state-owned enterprises, yet the vast majority or firms in many economies are privately held, and often entrepreneurial and family-owned. Given that certain institutional factors are particularly important in entrepreneurial environments (Terjesen et al. 2013), these relationships should be investigated. For example, recent work highlights that welfare support is significantly different within countries in terms of provisions for the self-employed versus those employees of the government and private sector. Furthermore, family enterprises naturally tend to have more gender diversity due to the presence of wives, daughters, and other members of the family. These family, private, and entrepreneurial firms deserve unique attention.

Six, a growing body of research discusses the post-quota ramifications for firms, including de-listing, going private, or registering in another country (Ahern and Dittmar 2012) and the promotion of “golden skirts”— women who hold a number of directorships (Huse 2011) or figurehead directors promoted due to their celebrity status (Branson 2012). This phenomenon should be investigated empirically and in consideration of national institutional structures. This line of enquiry could be further extended to multiple levels: boardroom dynamics, firm financial, and corporate social performance implications of various sets of quotas, individual post-board careers of men and women who left their directorships, and national levels of public opinion about female leadership and gender equality.

Finally, gender board quotas are only one type of quota. There may be quotas for other visible diversity traits (e.g., ethnic, age, educational background, professional background, racial minorities, etc.) which should be examined in the future, especially in countries with relatively heterogeneous work populations. We suspect that the conceptual model that we have proposed would nicely fit into other diversity dimensions.

Conclusion

This research is, to our knowledge, the first to examine the institutional factors associated with gender quota legislation for boards of directors. Our comparative analysis incorporates an institutional perspective. Taken together, countries with greater welfare to support women’s participation in the labor market, left-leaning partisan government coalitions, and a prior history of gender equality initiatives are most likely to establish gender quota legislation for boards. These findings significantly extend our understanding of government politics on business strategy and structure and provide a gendered framework for a discussion of comparable corporate governance.