Paper prepared for The 4th European Conference on Politics and Gender 2015 June Uppsala University, Sweden

Transcription

1 Effect of Women s Policy Agencies on Equal Employment Policies in Europe Senem Ertan, Ph.D. Social Sciences University of Ankara, Turkey Christine Scheidegger, Ph.D. St. Gallen University, Switzerland Abstract Do women s policy agencies make a difference for gender equality in the work force? This disputed question of feminist literature is rarely examined from a cross-national perspective. Therefore, this paper analyses the relationship between variety of permanent state engagements in gender equality politics and equal employment policies in Europe. It builds on a four-point typology of women s policy agencies, which scrutinize executive, legislative, judiciary, and other types of state arenas and their effect on equal employment policies. The main aim of the paper is to explain the extent to which comprehensive gender equality polities predict better employment policies for women among the EU member states as well as among other states in Europe such as Iceland, Switzerland and Turkey. The theoretical framework combines state feminist and public policy literatures. This study undertakes a comparative analysis of the EU member states and other European states, in order to determine the effect of EU membership on equal employment policies, all else being equal. In addition, this study aims to differentiate the effect of various institutional forms of women s policies agencies on equal employment policies. We use Qualitative Comparative Analysis (QCA) as an approach and an analytical technique to do this Middle-N study. Findings suggest that there are different pathways to adoption of equal employment policies. Institutionally inferior gender equality polities go well together with better equal employment policies. Paper prepared for The 4th European Conference on Politics and Gender 2015 June Uppsala University, Sweden 1/25

2 1. Introduction Gender gaps in employment still remain as a big problem throughout European countries. European feminist policies therefore, aim to address the issues such as unemployment rates, pay levels, or other types of discrimination at work. The Eurofound 2013 Report on Women, Men and Working conditions in Europe demonstrate that discrimination by gender is obvious particularly in the areas of women s access to the labour market (Eurofound 2013). Moreover, varied employment patterns are very much different for women and men, women occupying part time, low paid, and low status jobs whereas men mostly gets better positions and conditions even though they are equally qualified. However, apparently not all European countries are similarly responsive to equal employment needs of women. For example; equal pay gap in Estonia is as high as 29.9 percent whereas in Poland it is 6.4 percent (Eurostat 2015). Giving all these, it is very necessary to study pathways to effective equal employment policies in Europe. Considering the decision making process of gender equality policy, state feminism literature particularly assigns a very important role on two key actors; women s movements and women s policy agencies. This paper particularly focuses on the effect of polity of women s policy agencies as it has been rarely studied from a cross-national perspective. The main purpose of this paper is to analyse the role of women s policy agencies in adopting equal employment policies in Europe. Do women s policy agencies make a difference for gender equality in the work force? This disputed question is rarely examined from a crossnational perspective for whole Europe. More specifically the paper answers the question: How related are European women s policy agencies with gender equal employment policies within and outside of the European Union? This paper will also scrutinize the effect of European Union on the adoption of equal employment policies because the geographical focus of this paper on Europe requires us to discover whether EU as a unique supranational power could positively affect the adoption of equal employment policies. There is a wide range literature in Europe focusing on women s equality at work and the government responses to improve women s employment status. The European Union (EU) particularly, has been widely mentioned as being one of the most important non-state actors in the process of the development of gender equality policies after 1970s (Macrae 2006). The European Union is an extraordinary non-state actor which has a policy making and enforcing power over its member states. Although there is a second branch of literature which argues that European Union legislation is not always implemented at the 2/25

3 national levels (Bagilhole and Byrne 2000; Gregory 1995). Thus, this paper is an attempt to contribute in these scholarly discussions as well. This paper runs a comparative analysis of the EU member states and other European states to determine the effect of EU membership on equal employment policies, all else being equal. In addition, this study also aims to discover whether the effect of women s policies agencies forms has an impact on equal employment policy adoption in Europe. By having these objectives in mind, this study applies Qualitative Comparative Analysis (QCA) to identify pathways to effective gender equal employment policies in a cross-sectional analysis of 41 European countries. The data was originally developed by utilizing a combination of two dataset created by the authors (Ertan 2014; Scheidegger 2014). The following second section of this paper reviews the existing literature and defines guiding hypothesis for this study. The third section explains the methodology and the data. The fourth section presents our analyses and based on those main findings we present in the fifth sectionour conclusions. 2. Literature Our two main variables are gender equality polity and equal employment policy, which will be defined in the following paragraphs Definitions of important concepts Gender equality polity: With gender equality polity we mean the specialised administrative entities within the state apparatus. Those so-called women s policy agencies are organisational entities with the duty to advance women s empowerment and promote gender equality. Examples of women s policy agencies are the Minister of Emancipation of the Netherlands, the Gender Equality Bureaux of Japan and the Equality Commission in the Parliament of Spain. All women s policy agencies in a particular country form together the gender equality polity of this country. The gender equality polity is the specialist organisational administrative structure of the overarching gender equality architecture of a state to implement gender sensitive policies. Gender equality architecture as conceptualised by Walby, Armstrong and Strid (2012: ) entails legal framework and their implementation in a context of intersectional inequalities, beside the women s policy agencies. The paper concentrates on state equality polities focused on gender. 3/25

4 There are various institutional forms of women s policy agencies (McBride Stetson and Mazur 1995: 311; McBride and Mazur 2010: 48, 52 57; Berkovitch 1999: ; Testolin 2001: 9 25; Goetz 2004: 2 3; Sawer 2004: 2 7; Squires 2007: 34; Jahan 2008, 2010). i Countries vary in the numbers and categories of employed women s policy agencies. There is no default women s policy agency employed by all countries, even thought that most categories exist in all world regions. Within one state apparatus countries employ a number of different women s policy agencies (Scheidegger 2014: ). There is no geographical concentration of countries with similar women s policy agencies. Instead countries with similar combinations of women s policy agencies have little else in common. For instance the institutionally most comprehensive cases are Spain, the Philippines, Costa Rica, Ireland, El Salvador, Brazil, Colombia, Peru, Greece and Panama. That means in other words that usually neighbours have very different women s policy agencies, which applies for European countries as well (Scheidegger 2014: ). The Scandinavian countries have developed gender equality polities, but none of them has a most comprehensives gender equality polity Equal employment policy: In literature, the concept of equal employment policy has been widely used as a very broad phenomenon including all types of policies which directly or indirectly improves women s position at work. These include; wage policies, maternity or breast feeding policies, sexual harassment policies or childcare policies. Mazur (2002) divides this broad phenomenon into two groups calling the first group reconciliation policies and the second equal employment policies. According to Mazur (2002), reconciliation policies is defined as policies that target how women and men deal with the double burden of work and family (p. 32) and includes such policies as maternal leave, breastfeeding or childcare policies. Equal employment on the other hand includes any policies that seeks to promote feminist aims in paid labour (Mazur, 2002: 32). Thus, we define equal employment policies in terms of Mazur s framing of equal employment policy and we do not include reconciliation policies in this study. Gender equality polity and equal employment policy are studied in state feminist literature. Therefore, the following part of this section will focus on the main arguments of state feminism. 2.2 State feminism literature and reflections on employment policies State feminism is a concept which covers not only policies but also politics. It encompasses a dynamic relationship between feminist movements and feminization of politics and takes the 4/25

5 attentions to institutionalization of feminist interests in the state. State feminism has been described differently by scholars. For example, Stetson and Mazur (1995) define it as activities of government structures that are formally charged with furthering women s status and rights (pp. 1 2). On the other hand, Lovenduski (2009) defines it as the advocacy of women s movements demands inside the state (p. 4). Krook (2005) describes it as attempts by state actors to improve women s status through public policy (p. 6). The common point of all these definitions is that state feminism requires a state based action in order to improve women s status in the society. These actions can be done in the state apparatus through alliances between femocrats, who are feminists male or female working inside different levels of the state institutions. Despite variety of state institutions, in state feminism, state actors are seen as willing and able to initiate changes in existing patterns of gender inequality (Krook 2005: 6). Hernes defines state feminism as feminism from above in the form of gender equality and social policies and the feminization of welfare state relevant professions (1987: 153). On the other hand, Bochorst and Siim define it as feminisation of political elite through integration of women into the political parties. (2008: 210). In short, scholars mostly emphasize the role of state institutions in achieving progress in gender equality policies in a way responsive to feminist movements. The theory of state feminism has been developed on two main pillars. The first pillar leans on the role of state agencies that are established in order to support and stimulate women s rights. These state based women s policy agencies may be widely divergent being the Women s Bureau in the United Stated the oldest women s policy agency (WPA) in the world. Other examples of WPAs are; commissions, various offices, departments, secretaries, advisory councils, ministries etc. These state agencies play a crucial role in bridging the demands of women s activist and the state apparatus (Lovenduski 2009). However, women s policy agencies have not always been effective ( Stetson and Mazur 1995). For example, in Italy, Equal Status and Equal Opportunity National Commission had a limited impact on policy making since party membership was still a criterion for the appointment of members and party ideologies deeply affected framework of policies (Guadagnini 1995). In Poland, the Liga Kobiet (Polish Women s League), which was a quasigovernmental organization, had a very limited influence in policy making first of all because Church s activism was very powerful in the country and second, because soviet-type regime was dominated by patriarchal state interests (Robinson 1995). Thus, institutional capacity that is determined by different factors such as the scope and type of organization whether political or administrative or its proximity to political power, the resources that it has been given, is 5/25

6 important in effectiveness and efficiency of women s policy machineries (Stetson 2003b; Lovenduski 2009). Accordingly, state feminists define four types of WPAs classified according to their capacity to frame policy debates and their support for women s movement goals. In this typology, when the agency asserts the demands of women s movement and effectively genders the policy debate, then it is classified as an insider. When agency advocates for the movement goals but is not able to gender the policy debate, it is called marginal. When the agency is not an advocate of women s movement but is able to gender the policy debate, it is classified as a non-feminist. And lastly, if the agency is neither able to gender the policy debate nor to support the women s movement goals, then it is called symbolic (Lovenduski 2009). According to this classification, a WPA might be insider for one policy issue whereas might be a symbolic one for another issue because the perspective of WPA might change from one issue to the other. In short, even though many women s policies have been created in last decades, it would not be correct to assume that all these agencies strongly advocate and act for adoption of gender equality policies. However, creation of women s policy agencies might still symbolize government s willingness to institutionalize gender equality. Therefore, one can still argue that the existence and effectiveness of such institutions demonstrate the government s willingness to deal with gender equality problems. For this reason, we expect a positive relationship between women s policy agencies and equal employment policy. However, not each institutional form of women s policy agency has the same connection to labour market policy. There are a number of institutional forms of women s policy agencies with a potential strong effect on labour market and therefore on equal labour market policies. We call them the strong women s policy agencies related to labour market. The hypothesis therefore is: H1: Existence of labour market related women s policy agency positively affects adoption of equal employment policies. In the executive branch of government the strong labour market related women s policy agencies are the Entity in Top Level Government, the E1, and the Ministry, the E2. While the Entity in a Ministry (E3) and the Coordinating Body (E4) have less power over equal employment policy due to their institutional remoteness and weak leverage. In the legislative branch of government the Legislative Committee, the L2, may have a direct impact on labour legislation. While the other women s policy agencies in the parliament, such as the Competence Centre of the Legislative Assembly (L1) and the Legislative Caucus (L3) are target more on internal functioning of the parliament than on particular legislation. The third group of potentially strong labour market related women s policy agencies are the Human 6/25

7 Rights Body (J1), the Equality Authority (J2) and the Monitoring Body (O3). All three have an intervening character. The Working Group in the Judiciary, the J3, and the Commission, the O3, are less powerful and therefore excluded from the analysis. The strong labour market related women s policy agencies can be grouped into executive, legislative and intervening into labour market policy in Europe. As a second and complementary pillar, women s rights movements and civil society create an unavoidable pressure on states, particularly in industrial democracies (Htun and Weldon 2010; Weldon 2002; Lovenduski 2009). Celis et. al. (2008) argue that women s substantive representation does not only appear in elected political bodies or other state agencies but also in different levels of political forums one of which is the women s movements. The characteristics of women s movements determine to the extent to which these movements will be successful in influencing the policy-makers. Autonomy and the strength (Weldon 2002), the development level of the movement, its closeness to the left, priority of the issue on the movement agenda, and cohesiveness of the movement organizations (Lovenduski 2009) are some important characteristics of women s movements effectiveness. Weldon (2002), in her empirical analysis of representation of women in policymaking of 36 democratic countries concluded that the existence of strong and autonomous women s movements accompanied by effective women s policy agencies increases the likelihood of successful policymaking on women s rights. Similarly, Lovenduski (2009) in her edited volume of State Feminism and Representation demonstrates that state agencies by their selves are not adequate sources of legislation, which is responsive to women s demands. Therefore, creation of these agencies should be supplemented by cohesive women s movements for a more responsive substantive representation. However, in any case the success of women s groups highly depends on the special and effective linkages between WPAs, feminists in the government and women s movements. At the same time, however, different groups in women s movements may represent different interests of women. Thus, one cannot assume unique women s interest and homogenous women s movement but various groups might prefer to negotiate each other and unify for a collective objective. Vargas and Wieringa (1998) differentiate between two types of strategies that different groups in women s movements apply to achieve their goals. Solidarity or sisterhood that refers to creation of strategic alliances between different groups of feminist organizations for a common aim is the first strategy. The second strategy is that women s groups try to get autonomy from the male dominated structures of the society and they focus on bottom up awareness raise of individual women. In any case, the way how 7/25

8 women s movement groups interact with the state is a crucial issue. Women s groups might actively oppose dealings with the state or they can be more conciliatory in terms of articulating their demands in a way that the state can respond (Vargas and Wieringa 1998). In this context, there are four possibilities how a state might interact with women s movements. Lovenduski (2009) states that when the states accept women s movement; and the policy change that was made responds the movement s demands, it is called a case of dual response. In the second case, which she calls co-optation, state accepts the involvement of women s movement in the policy process but it does not respond the demands of it when the policy is formulized. Thirdly, pre-emption is the case when the state does not allow the actors of women s movement to intervene into the policy process even though it produces responsive policies. And lastly, the case of no response occurs when the state neither allows women s movement into the process nor produces women s friendly legislation demanded by the movement. Looking to the state feminist arguments on the importance of strong women s movement we expect a positive relationship between the strength of women s movement and gender equality policies. H2: The stronger the women s movement, the better are the equal employment policies. A weakness of the state feminism literature is its narrow focus on welfare states. So far many of the state feminist literature included only democratic welfare states making the theory of state feminism only applicable for that sample of countries. However, WPA and civil society contexts also exist in semi-democratic countries and in the developing world such as in Turkey, proving to achieve some progress in the legislation for gender equality. Therefore, this study also aims to test the arguments of state feminism literature for a sample of countries that is are not democratic, post-industrialized countries. 2.3 Effect of European Union on equal employment policy Another factor with a potential impact on equal employment policies in Europe is the European Community, respectively today the European Union. The European Union makes policies through four main institutions: 1) The Commission 2) The Council of Ministers 3) The Court of Justice 4) The European Parliament (Buckley and Anderson 1988). The Commission plays the role of independent advisor for the Council of Ministers and do not mostly initiates legislation but put sanctions on them. The Council has the power to enact regulations, directives, decisions or recommendations and opinions. From these features, 8/25

9 regulations and directives are the most powerful means of policy as both of them are legally binding. They differ from each other because regulations have a direct legal force whereas the implementation of directives is more flexible as it depends on the legal system of member states. On the other hand, decisions or recommendations and opinions do not have any mandatory power as they only aim to give advice to the member states. Turning to the European Parliament, its role can be described as debating and giving consent to the bills initiated by the Commission. Furthermore, it urges EU institutions to take action on related issues as well as carries out detailed researches through specialized committees (Buckley and Anderson 1988: 7 10). The Treaty of Rome was signed in 1957 and it included the article 119 as a first commitment of the European Community to gender equality issues that bound member states to adopt principle of equal pay of equal work (Buckley and Anderson 1988). The literature on gender equality policy in European Union agrees that economic considerations about imbalanced competition in the created economic community are behind the inclusion of this article and not gender equality as an aim (Ellina 2003; Sumer 2009; Rosilli 2000). Thus even though more social policies aiming to improve women s status has been adopted later on, they have been mostly focused on women s labour market participation such as women s rights at work or women s double duty of work and family. While during 1960s and 1970s the focus of the EU was more on women s access to employment, training and education, more recently sexual harassment regulations, childcare policies and maternity, paternity or parental leave policies has been issue to European policy initiatives (Macrae 2006; Sumer 2009; Rosilli 2000). Against this background of this expansion of equal employment politics within the EU we expect a positive effect of EU membership on equal employment policies. The impact of the EU has not always been very clear cut. Even though the EU legislation has given women more legal rights, member states have not always had incentives to implement those laws. For example, European Community directives on equal opportunity and reconciliation policies have had a limited impact in terms of actual policy outcomes in the United Kingdom (Bagilhole and Byrne 2000). Similarly, member states have had a modest respond to EU legislation on sexual harassment till mid-1990s because national laws have not been good enough to support the effective procedures for implementing those laws (Gregory 1995). Despite the limitations and problems of implementation, there is evidence that European Union policy on gender equality had a positive impact on member states (Hoskyns 1988). 9/25

10 Thus in the EU member states the analysis of gender policy cannot be limited either to the national or to the EU level but must be understood within the broader context of multilevel and multiscalar process (Castree 2000), where overlapping influences have helped establish a set of gender rules and norms to challenge the exclusive dominance of national gender regimes (Macrae 2006: 523). This together with the singular character of the European Union as a non-state institution with enforcement power affecting domestic policy making of member states. Therefore, we hypothesized the effect of EU as following; H3a: The countries who are member to the EU are expected to show better equal employment policies than Non EU members. The causal linkages are the EU s mandatory power on its member states and the integration into a common European market. EU countries vary in general (economic wealth, governance) and in particular in employment policies and gender equality polity. Long-term EU members had much longer time to adapt to a common European market than newer EU members, which struggled in the last twenty years with the transformation from socialist to capitalist systems. The long and thorny history of labour market regulation of the EU in regard to gender equality (cf. RC study on equal pay legislation) and the efforts to harmonise gender equality politics with directives (Kriszan, Skjeie and Squires 2014: 54 55) makes it likely that equal employment policies are more developed among long-term EU members than among short-term or not EU members. The regulative integration may have a time effect component. Accession needs time to translate into domestic implementation and candidate countries are less integrated into the common market. The causal link in this strand of literature are policy learning and continued cooperation within and beyond institutionalized forums of the European Community and today the European Union. The often similar regulative frameworks of other countries around the EU member is partly a spill-over effect born out of the economic necessity of economically small and open countries (Katzenstein 1985) to have similar access to the common market of the EC/EU. H3b: The countries who are long-term member of the EU are expected to show better equal employment policies than Non-EU members, EU candidate countries and short-term members of the EU. 10/25

11 Therefore, it is crucial to have EU members and Non EU members in the sample as well as candidate countries and potential candidate countries to represent the different levels of integration. 3. Research design, methods and data In this study, QCA analysis will be applied as a most appropriate method for a cross-national Middle-N study. The analysis includes 41 countries which are geographically from European continent. We used QCA as an approach (Ragin 2014: 81) to identify asymmetric causation (Ragin 2014: 87 88). To identify different pathways of countries to reach certain equal employment policies QCA is well equipped as it assumes causal complexity and provides tools to detect multiple pathways leading to a certain outcome (equifinality) (Ragin 2014: 84). The labour market related strong categories of women s policy agencies are the main independent variables. The categories of women s policy agencies are measured over the classification of Scheidegger (2014a: ). The measurement gives the situation for the years The EU membership is an intervening variable measured for the year Equal employment policies are the dependent variable, which is measured according to the index developed by Ertan (2014). 3.1 Operationalization of variables Dependent variable: The dependent variable of this paper was measured by using a ten-point original equal employment scale of Ertan (2014). The scale take into consideration both legal regulations and government s policies and programs to improve women s position in employment. The first component consists of a sum of equally weighted scores (1 point for each) for six questions on legal employment regulations. The data was composed by the World Bank Woman, Business and the Law 2012 Database with the following questions: 1) Are there laws or constitutional provisions mandating equal pay for equal work? 2) Are there laws mandating nondiscrimination in hiring practices on the basis of gender? 3) Can women work the same night hours as men? 4) Can women work in the same industries as men? 5) Do the statutory retirement ages for men and women in the private sector are different? 6) Are there laws protecting employees from sexual harassment in the work place? 11/25

12 The second component measures government actions such as government programs, specific women s employment action plans, vocational training, which address to improve women s status in employment. The data is mainly gathered from CEDAW ii State Reports and WomanStats Database. Scale points are arranged in the following way; 0) There are no or weak policies considering the principle of equal treatment, equal pay and sexual harassment for women in work place. There may be study restrictions about certain areas of employment. Government has no or weak programs to train women for nontraditional employment. 1) There are some policies referring to the principle of equal treatment, equal pay and sexual harassment for women in work place. There are no study restrictions for women. Government has some programs to train specifically women for nontraditional employment. 2) There are specific and comprehensive policies referring to the principle of equal treatment, equal pay and sexual harassment for women in work place. There are no study restrictions for women. Government has programs to train specifically women for nontraditional employment. Those two components were taken as added sum of the both scores ending up with a ten-point scale. For the QCA analysis this composite scale was recoded into a fuzzy set (fs) variable equal employment score (fs) reaching from 0.0 to 1.0. A high equal employment score (1.0) stands for a fully developed gender equal labour market policy. The crisp set variable (dichotomous), where 1 = effective equal employment policy (between 0.6 and 1.0) and less effective equal employment policy (0.0 till 0.4) is a cruder measurement of the same variable. For coding details see Table 3 in the appendix. 3.2 Independent variables: Gender equality polity as the main independent variable of this paper was taken from Scheidegger (2014, 2015), which encompasses the most topical and most comprehensive data available for women s policy agencies and gender equality polities. The original data set contains all women s policy agencies in 151 countries for the years and comes from a survey among UN member states conducted by the United Nations Division for the Advancement of Women (UN DAW) iii. 12/25

13 For the QCA analysis the gender equality polity was recoded to strong labour market related women s policy agencies. The strong labour market related women s policy agencies in the executive branch of government, E1 and E2, were taken together to a strong labour market related women s policy agencies in the executive, the E1E2. Countries with either an E1 or an E2 are coded as 1 = present and 0 = for absent. The intervening strong labour market related women s policy agencies J1, J2 and O3 were recoded similarly (for coding details see appendix Table 3). The second set of independent variables concerns around the effect of the European Union. We used two coding according to the two hypotheses H3a and H3b; EU membership and EU legacy. EU membership is coded as 1 = EU member in 2015 and 0 = Non EU member, candidate countries and potential candidate countries. EU legacy is coded in two ways. In the dichotomous variable EU legacy (cs) the 1 stands for EU member in the 1990s or before and 0 represents all Non EU members, the candidate and potential candidate countries and countries with EU accession in the 2000s. The fussy set coding, the EU legacy (fs) was coded 0.0 for Not EU member, over 0.2 for candidate countries and potential candidate countries, to 0.6 for countries with EU accession in the 2000s, 0.8 for EU accession in the 1990s and 1.0 for EU membership before the 1990s. The gap between 0.2 of (potential) candidate country and 0.6 of effective accession in the 2000s mirrors the long way between those two distinct positions. The fussy set variable represents various levels of integration into EU. The third independent variable, women s movements is operationalized by calculating the number of women s groups divided by squared root of the country s population. In this way we neutralized the possible effect of population size on the number of women s organizations because more populated countries are expected to have higher numbers of organizations. The data on the number of women s groups was gathered from the Global List of Women s Organizations available at: The data on population was gathered from the World Atlas, accessed on March 23, 2012 from: 4. Analyses and main findings This section is designed in three parts. The first part reviews the effect of EU membership on gender equal employment policies. Second, pathways to effective equal employment policies are discussed. On the contrary the third part discusses pathways to less effective equal employment policies. 13/25

14 4.1 Effect of the EU on equal employment policy The simple fact of being an EU member has less effect on equal employment policies than the hypothesis H3a expects. From the 29 countries with a high score in equal employment policies 25 are members of the EU and four are not. Consequently, there is a significant Pearson s correlation of 0.52 between EU membership and equal gender employment policy fs (0.99 significance, two-sided). However, the countries without an EU membership and a high equal employment policy score demonstrate that EU membership is not the only way to reach a developed equal employment policy. Therefore the hypothesis H3a is partly refuted. Equal employment policy quality is affected by EU membership, but the membership alone is not the sole explanation. The picture is different when the effect of EU legacy on equal employment policy is considered. The QCA analysis of EU legacy cs and equal employment policy cs shows that there is limited diversity in the data (see Table 1) One possible configuration has no cases; being a long-term EU member and having a low score in gender equal labour policy is absent. Or in other words none of the long-term EU members has a low score in gender equal labour policy. Thus, this finding suggests that long-term EU membership significantly rules out bad equal employment policies, respectively improves adoption of good equal employment policies. The legacy of the EU membership seems to work as a base line against poor equal employment policy. Table 1: Distribution of cases and limited diversity EU membership legacy Serbia, Montenegro, Turkey, Bosnia and Herzegovina, Macedonia, Slovenia, No cases Equal Employment Poland, Belarus, Ukraine, Czech Republic, Russia, and Moldova Score 1 Iceland, Albania, Austria, Cyprus, Spain, Ireland, Luxembourg, Germany, Bulgaria, Switzerland, Slovakia, Malta, Belgium, Netherlands, Sweden, Denmark, Romania, Croatia, Lithuania, Estonia, United Kingdom, Finland, France, Greece, Norway, and Latvia Portugal, Hungary, and Italy 14/25

15 Table 2: Values of different variables. Country E1 or E2 L2 J1 or J2 or O3 EU legacy equal employment score term Spain & 2 Ireland & 2 Luxembourg & 2 Germany & 2 Greece & 2 Portugal & 2 Slovakia & 2 Hungary & 2 Czech Republic * & 2* Poland * & 2* Bosnia & Herzegovina Turkey Ukraine Belgium Montenegro Denmark & 2 Netherlands & 2 Italy & 2 Sweden & 2 Austria & 2 Latvia & 2 Norway n. c. United Kingdom Albania * Finland Cyprus Croatia Romania Serbia Macedonia Belarus Russia France Bulgaria Malta Lithuania Estonia Slovenia * Iceland n. c. Moldova Switzerland n. c. Note: * = contrasting case; n. c. = case not covered by solution term. 15/25

16 This absence in the cross tabulation explains the rather high value of Pearson s correlation of 0.49 (0.99 significance, two-sided) between EU legacy cs and equal employment policy cs, despite the number of cases without EU long-term legacy and a high score in equal labour policy. The correlation is with 0.67 (0.99 significance, two-sided) even stronger between EU legacy fs and equal employment policy fs. Overall, countries without EU membership or rather short history in EU vary in their equal employment record. There are three short-term EU countries (Czech Republic, Poland and Slovenia) with a rather low equal employment score (0.4). While long-term EU members have a rather strong equal employment score. As the cross tabulation and Table 2 indicate there are many countries outside EU or with a rather young EU membership with a strong equal employment score (e.g. Iceland, Malta and Croatia). This pattern supports hypothesis H3b. In sum, there is an effect of the European Union on strong equal employment policy. The effect is stronger for long-term EU members than for short-term EU members. 4.2 Pathways to effective gender equal employment policies Due to this pattern the EU legacy and not the weaker EU membership became part of QCA analysis. In the following it explains together with the strong labour market related gender equality polity the existence of developed gender equal employment policy. That is: EU-legacy + E1E2+L2+J1J2O3 > good equal employment policy. The solution term of QCA is: EU legacy * E1E2 + EU legacy * J1J2O3 + E1E2 * not L2 * not J1J2O equal employment policy 1. solution term 2. solution term 3. solution term outcome The positive sign separates three separate solution terms. Each of those solution terms leads alone to the outcome of a good equal employment policy. Each solution term represents a pathway to a good equal employment policy. Those three pathways are equifinal in this sense. A country can be covered by several solution terms. This is indicated in the last column of Table 2. For instance Spain is covered by the first and the second solution term and the United Kingdom by the first and the third solution term. As it is visible in this column almost all cases with a good equal employment policy of 0.6 to 1.0 are covered by one of the three solution terms. Not covered are the three cases Norway, Iceland and Switzerland. All three are 16/25

17 not EU members yet, but they show a good equal employment policy. Beside the high coverage of the solution term there is another measurement to evaluate the quality of the QCA solution: the consistency. The consistency says how many cases covered by the solution term have the assumed outcome. As the last column says that there are only four cases which have the favourable conditions but have a low equal employment score. Those are contrasting cases. The first pathways (EU legacy * E1E equal employment policy) states that longterm EU membership (EU legacy between 0.6 and 1.0) together with strong labour market related women s policy agencies in the executive branch of government (E1 or E2 = 1) result in a high equal employment score (between 0.6 and 1.0). Hence, long-term EU membership and strong executive women s policy agencies promote a developed equal employment policy. This solution applies to the countries: Spain, Ireland, Luxembourg, Germany, Greece, Portugal, Slovakia, Hungary, Czech Republic (contrasting case), Poland (contrasting case), Denmark, Netherlands, Italy, Sweden, Austria, Latvia, and United Kingdom. Most of those countries have a rather good (0.8) or a good equal employment policy (1.0). Czech Republic and Poland are two contrasting cases with EU legacy and strong labour related women s policy agencies in the executive branch of government, but a rather bad score in equal employment policy (0.4). The second pathway (EU legacy * J1J2O equal employment policy) states that a long-term EU membership (EU legacy between 0.6 and 1.0) together with strong intervening labour market related women s policy agencies (J1, J2 or O3 = 1) result in a high equal employment score (between 0.6 and 1.0). In other words, long-term EU membership and strong intervening women s policy agencies promote a developed equal employment policy. This solution applies to even more cases than the first one: Spain, Ireland, Luxembourg, Germany, Greece, Portugal, Slovakia, Hungary, Czech Republic (contrasting case), Poland (contrasting case), Denmark, Netherlands, Italy, Sweden, Austria, Latvia, Finland, Cyprus, Croatia, Romania, France, Bulgaria, Malta, Lithuania, Estonia, and Slovenia. It contains as well the two contrasting cases Czech Republic and Poland. The third pathway (E1E2 * not L2 * not J1J2O equal employment policy) states that strong labour market related women s policy agencies in the executive branch of government (E1 or E2 = 1) and the absence of strong legislative and intervening labour market related women s policies result in a high equal employment score (between 0.6 and 1.0). In other words, with and without a long-term EU membership strong labour market 17/25

18 related women s policy agencies in the executive branch of government promote a developed equal employment policy. This solution applies to United Kingdom and Albania. This solution contains no contrasting cases and both cases have no equal employment policy score below 0.8. Those three pathways show that long-term EU membership is beneficial for equal employment policy in combination with strong intervening or executive women s policy agencies. In the absence of a long-term EU membership the absence of strong intervening and legislative women s policy agencies together with strong labour market related women s policy agencies in the executive branch of government brings a good equal employment policy as well. The three cases not covered by the solution terms can be considered as three other pathways to a developed equal employment policy without EU membership. Norway has strong executive and intervening women s policy agencies. Switzerland has no strong labour related women s policy at all and Iceland shares it strong intervening women s policy agencies with a number of other countries (France, Bulgaria, Malta, Lithuania, Estonia, Slovenia, and Moldova). However, Moldova as a Non EU member has a considerable lower equal employment policy score (0.2) and Slovenia as a short-term EU member has a smaller equal employment policy score (0.4) than Iceland (1.0). 5. Conclusions The main aim of this paper was to discover whether strength of women s policy agencies play a crucial role on the adoption of equal employment policies in Europe. Secondly, this paper scrutinized whether and how EU membership has an impact on successful equal employment policies. There are two main pathways to an effective equal employment policy and four less travelled pathways. The first pathway underlies the relevance of the Entity in Top Level Government (E1) and the Ministry (E2) as two strong labour market related women s policy agency forms, which interact with a long-term membership in the EU towards a strong equal employment policy The second pathway stresses the relevance of strong intervening women s policy agencies, such as Human Rights Body (J1), Equality Authority (J2) and Monitoring Body (O3), together with a with a long-term membership in the EU to reach a strong equal employment policy. The third pathway to good equal employment policy highlights again the relevance of strong executive women s policy agencies (E1E2), but emphasis the absence of 18/25

19 other strong labour labour market related forms of women s policy agencies in order to reach good equal employment policies. The other three solutions apply each for a single Non EU country with a high equal employment score. Another important finding of this paper is that the EU membership per se does not play the most important role in the adoption of effective equal employment policies. It is rather the long-term effect of the EU membership coming together with strong gender equality polities leading to good employment policies. This means that countries which are member of the EU before the 2000s were more successful to improve women s employment status whereas later EU members and other European countries where not always that much successful. 19/25

20 Notes I For identification of equivalents institutional forms see Scheidegger (2014: 85, 87). II CEDAW stands for Convention for the Elimination of all Discrimination of against Women. It s a major human rights treaty. III Today part of UN WOMEN. 20/25

21 References Alter, K. J. and Vargas, J. (2000). Explaining Variation in the Use of European Litigation Strategies, European Community Law and British Gender Equality Policy, Comparative Political Studies, 33(4): Auer, P. (2002). Changing Labour Markets in Europe: The Role of Institutions and Policies. Bagilhole, B., & Byrne, P. (2000). From Hard to Soft Law and from Equality to Reconciliation in the United Kingdom. In L. Hantrais (Ed.), Gendered Policies in Europe: Reconciling Employment and Family Law. London: Palgrave MacMillan. Bochorst, A., & Siim, B. (2008). Woman-friendly policies and state feminism: Theorizing Scandinavian gender equality. Feminist Theory, 9, Buckley, M., & Anderson, M. (1988). Introduction: Problems, Policies and Politics. In M. Buckley & M. Anderson (Eds.), Women, Equality and Europe (pp. 1-19). London: Macmillan. Castree, N. (2000). Geographic Scale and Grass-Roots Internationalism: The Liverpool Dock Dispute, Economic Geography, 73(3), M. Buckley & M. Anderson (Eds.), Women, Equality and Europe. London: Macmillan. Celis, K., Childs, S., Kantola, J., & Krook, M. L. (2008). Rethinking Women's Substantive Representation. Representation, 44(2), Daeren, L. (2001). The Gender Perspective in Economic and Labour Policies State of the Art in Latin America and the Caribbean, Women and Development Unit ECLAC-GTZ project Institutionalization of gender policies within ECLAC and sectoral ministries Santiago de Chile. Ellina, C. A. (2003). Promoting Women's Rights: The Politics of Gender in the European Union. New York: Routledge. Ertan, S. (2014). How to Study Gender Equality Policy Cross-Nationally? Aggregate or Disaggregate Gender Equality Policy Indices?, Social Indicators Research, (DOI: /s ). Eurofound (2013), Women, men and working conditions in Europe, Publications Office of the European Union, Luxembourg. Eurostat, (2015). Women earned on average 16% less than men in 2013 in the EU, News Release, retrieved on May 25, 2015 from: Gardiner, F. (1997). Introduction: welfare and sex equality policy regimes. In F. Gardiner (Ed.), Sex Equality Policy in Western Europe (pp. 1-24). London: Routledge. Gregory, J. (1995). Sexual Harassment: The impact of EU Law in the Member States. In M. Rossilli (Ed.), Gender Policies in the European Union (pp ). New York: Peter Lang. 21/25

22 Guadagnini, M. (1995). The Latecomers: Italy's Equal Status and Equal Opportunity Agencies. In D. M. Stetson & A. G. Mazur (Eds.), Comparative State Feminism. Thousand Oaks, California: Sage Publications. Hernes, H. (1987). Welfare State and Woman Power: Essays in State Feminism. London: Norwegian University Press. Hoskyns, C. (1988). Give Us Equal Pay and We'll Open Our Own Doors:A Study of the Impact in the Federal Republic of Germany and the Republic of Ireland of the European Community's Policy on Women's Rights. In M. Buckley & M. Anderson (Eds.), Women, Equality and Europe (pp ). London: Macmillan. Hoskyns, C. (1996). Integrating Gender: Women, Law and Politics in the European Union. London: Verso. Htun, M., & Weldon, S. L. (2010). When do Governments Promote Women's Rights? A Framework for the Comparative Analysis of Sex Equality Policy. Perspectives on Politics, 8(1), Kantola, J. (2010). Gender and the European Union, New York: Palgrave Mac Millan. Krook, M. L. (2005). Quota Laws for Women in Politics: A New Type of State Feminism? Paper presented at the European Consortium for Political Research, Joint Sessions of Workshops,, Granada, Spain. Lewis, Jane (2006). Work/family reconciliation, equal opportunities and social policies: the interpretation of policy trajectories at the EU level and the meaning of gender equality, Journal of European Public Policy, 13(3): Loutfi, M. F. (2001). Women, Gender and Work, Geneva: International Labour Office. Lovenduski, J. (2009). Introduction: State Feminism and the Political Representation of Women. In J. Lovenduski (Ed.), State Feminism and Political Representation (pp. 1-19). New York: Cambridge University Press. Macrae, H. (2006). Rescaling Gender Relations: The Influence of European Directives on the German Gender Regime. Social Politics, 13(4), Mazey, S. (1998). The European Union and women's rights: from the Europeanization of national agendas to the nationalization of a European agenda? Journal of European Public Policy, 5(1): Pascall, G. and Lewis, J. (2004). Emerging Gender Regimes and Policies for Gender Equality in a Wider Europe, Journal of Social Policy, 33 (3), Ratner, R. S. (1980). The Policy and Problem: Overview of Seven Countries. In R. S. Ratner (Ed.), Equal Employment Policy for Women (pp. 1-53). Philedelphia: Temple University Press. Rees, T. (1998). Mainstreaming Equality in the European Union. Education, Training and Labor Market Policies. Chapter 3- Conceptualizing Equal Opportunities. 22/25