Gender Equality in Context

Brigitte Liebig Karin Gottschall Birgit Sauer (eds.)

# Gender Equality in Context

Policies and Practices in Switzerland

Barbara Budrich Publishers Opladen • Berlin • Toronto 2016 © 2018 This work is licensed under the Creative Commons Attribution-ShareAlike 4.0. (CC-BY-SA 4.0)

It permits use, duplication, adaptation, distribution and reproduction in any medium or format, as long as you share under the same license,

give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

To view a copy of this license, visit https://creativecommons.org/licenses/by-sa/4.0/

© 2018 Dieses Werk ist beim Verlag Barbara Budrich erschienen und steht unter der Creative Commons Lizenz Attribution-ShareAlike 4.0 International (CC BY-SA 4.0): https://creativecommons.org/licenses/by-sa/4.0/

Diese Lizenz erlaubt die Verbreitung, Speicherung, Vervielfältigung und Bearbeitung bei Verwendung der gleichen CC-BY-SA 4.0-Lizenz und unter Angabe der UrheberInnen, Rechte, Änderungen und verwendeten Lizenz.

This book is available as a free download from www.barbara-budrich.net (https://doi.org/10.3224/84740727). A paperback version is available at a charge. The page numbers of the open access edition correspond with the paperback edition.


Barbara Budrich Publishers Stauffenbergstr. 7. D-51379 Leverkusen Opladen, Germany

86 Delma Drive. Toronto, ON M8W 4P6 Canada www.barbara-budrich.net

A CIP catalogue record for this book is available from Die Deutsche Bibliothek (The German Library) (http://dnb.d-nb.de)

Jacket illustration by Bettina Lehfeldt, Kleinmachnow, Germany – www.lehfeldtgraphic.de Editing by Máiréad Collins, Belfast, UK Typesetting by Ulrike Weingärtner, Gründau, Germany

# **Content**

#### **1 Introduction**


#### **2 Gender equality policies and their implementation**


*Christine Zollinger & Thomas Widmer*


#### **5 Postscriptum**


# **Introduction**

## **Gender equality: Policies and practices in Switzerland**

*Brigitte Liebig, Karin Gottschall & Birgit Sauer*

In many European countries women's suffrage and processes of democratization have resulted in profound developments and transformations toward gender equality in the twentieth century. Due to second-wave feminist movements, gender inequality policies have dominated the agenda of many public debates in Europe within the last 30 years. Yet, while formal political equality, citizenship rights and quantitative representation of women in political decision making institutions and processes seem to be necessary preconditions, they do not guarantee gender equality. Political citizenship does not translate automatically into social and cultural citizenship nor into equal social, economic and cultural participation. Until today, the acquisition of equality norms and participation rights are paralleled with new forms of social inequality and gender-specific disparities. Belonging to the female or male sex is still one of the most influential and significant social distinctions today. The gender division of labor in particular still works as a powerful mechanism of gender inequality as it implies unequal distribution of unpaid care work, translates into a gender segregated labor market, and generates inequality of wages and unequal access to social security.

However, variations exist across Europe with respect to the organization of the labor market and welfare state provisions. Comparative feminist research in welfare states has shown that welfare state institutions created path dependencies in structuring gender regimes. Drawing on Esping-Andersen's (1990) typology of welfare states, feminist research classified Nordic countries such as Sweden, Finland and Norway as weak male breadwinner or "dual breadwinner models", France as "moderate breadwinner model", and Germany and Austria as "strong breadwinner model[s]". Switzerland may be assigned to the latter category (Lewis 1992; Ostner/Lewis 1995). Different degrees of familiarization and de-familiarization (Leitner 2003) impact on the opportunity of women to combine work and care obligations and to get access to higher positions in the labor market. Also, the type of welfare state regime corresponds with the integration of women in public and political life: political citizenship rights and social citizenship rights cannot be separated (Lister 1997).

#### *Policies and practices in Switzerland*

By international standards, gender equality policy in Switzerland has started rather late. A strongly limiting factor for any political measures and activities to promote gender equality has been the lack of women's right to vote, which was achieved only in 1971 (Ballmer-Cao/Michel 2009). The Swiss semi-direct democracy, which is shaped to a high degree by plebiscites and popular votes, provides a sharp contrast to most other European countries, where processes of opinion building and decision-making take place within the framework of representative democracy. The enforcement of women's right to vote was promoted by women's suffrage movements and initiatives of various groups deeply rooted in civil society.

A further important milestone was set in 1981, when the principle of gender equality was incorporated in Swiss law. Since then a constitutional article guarantees the formal equality of men and women and commits authorities and legislators to eliminate existing discrimination, especially in areas of education, work and family, and the right to equal pay for equal work. The Equality Act of 1996 provided a striking point in gender equality policy, since it allowed individual women, but also organizations, to take legal action against discrimination in the workplace. Besides national drivers such as the women's movement and organizations, the equality commitment of international organizations such as the UN, the Council of Europe, and the EU, have been important drivers for Swiss equality policies (Ballmer-Cao/Michel 2009). The ratification of conventions such as the CEDAW (Convention of the United Nations on the Elimination of all Discrimination against Women) or the ECHR (European Convention on Human Rights) mark important steps towards greater gender equality in Switzerland. At the same time the standards set at the 4th UN World Conference on Women in Beijing 1995 have served as a model for the federal council of Switzerland for conveying and achieving equality between women and men within different fields of society (FOGE 2014).

Within the last thirty years politics and administrations, educational institutions and companies have designed a large number of action plans, programs and measures for improving gender equality in Switzerland. On the national level as well as on cantonal and municipal levels, gender equality policies are characterized by a high level of institutionalization. Since 1988 the Federal Office for Equality between Women and Men promotes the implementation of gender equality in all spheres of life and the removal of all forms of direct or indirect discrimination. These efforts indeed have supported comprehensive improvements in gender equality in education, employment and family in a relative short period of time and helped Switzerland to catch up with other Western countries.

Nevertheless, in the European context, Switzerland conveys an ambivalent picture with respect to gender equality (FOS 2014, Liebig et al. 2015). While the gender gap regarding participation in secondary education has been considerably reduced, fewer women than men in Switzerland complete higher education, and gender stereotypes continue to affect choice of education, training, and careers to a large extent (FOS 2014). And while female labor market participation in Switzerland ranks at the top throughout Europe today, women have not been able to attain the same success as men in terms of job position and pay (FOS 2008). Hence, gender-specific segregation in education and the labor market continues in Switzerland, and is considerably more marked than in other European countries (OECD 2011, 2012). The promotion of women into technical and upper fields of the labor market is below expectations, and also the percentage of women in political and economic decision-making remains rather low. Also in the sphere of family, little has changed with respect to the sharing of responsibilities for housework and/or care: Within the liberal conception of the Swiss welfare state, family and fatherhood rank as a 'private matter' and until today a sound legal foundation for paternal leave does not exist (Liebig/Peitz 2014). Despite modern notions of division of labor between partners, even young couples continue to follow traditional family models. Switzerland ranks at the very last place of 21 high-income OECD countries with respect to the support of equal parenting (Fatherhood Institute 2010). Single parents still suffer poverty, and domestic violence is a persistent basic problem in intimate partner relations between men and women. Finally, while women have been actively struggling for gender equality and to some degree have been successful, the involvement of men in this movement on a large scale has proven difficult thus far.

Overall, it can be stated, that due to extensive political efforts Switzerland has made substantial progresses in the field of formal equality, but gender equality has not yet been reached. There is a great need for more knowledge on the persistence of gender inequalities and a deeper understanding of why efforts to promote gender equality have only limited effectiveness so far. Moreover, the question arises, by which means the often subtle processes of (re-)producing gender inequality can be influenced.

#### *Need for further research*

In response to the dissatisfaction with the status of gender equality as well as related initiatives and policies, various research groups have initiated activities that have been assessing gender equality activities and their framework conditions on an international level within the last years. Namely projects on policy framing and the institutional conditions of gender mainstreaming implementation (MAGEEQ, 1998‒2002)<sup>1</sup> , or studies on gender equality policies across the EU Member States (QUING, 2006‒2011)<sup>2</sup> , have been important milestones in order to assess current problems of gender equality policies and formulate recommendations and standards for policy making. The international comparisons yielded important information on effective and sustainable gender equality strategies and the relevance to gender equality of different policy regimes. At the same time, international comparative analyses of policy areas such as family or labor market policy and international and regional surveys (e. g. Eurostat, UNO, ILO, OECD) have been producing extensive statistical information.3

Cross-national approaches illustrate how important it is to reflect on the contextual variance of institutional forms and social practices of gender (in-) equality. Policy analysis has to situate the effects of gender equality policy within the context of the objectives of gender-related activities and programs, and it has also to critically reflect on political arrangements themselves. More than that, it has to consider that the implementation of legal provisions depend on the active application and active promotion by political actors and institutions in different parts of society.

That means, although the results and experiences of other countries are very important for Switzerland, they can only be conditionally transferred since gender equality policy and activities are shaped uniquely by its legal, economic, cultural, and historical conditions. This is especially true for the federalist system of Switzerland, where law can result in different forms of implementation at the level of Swiss cantons, or within different regional contexts. Up until now little has been known about the effectiveness of existing gender equality policies within the context of regional conditions and specific stakeholder environments, as well as in relation to processes involved in the elaboration and implementation of activities and programs on institutional levels. The spheres of education work, family, and social security are characterized by specific structural conditions, rationalities, and ways of functioning,

<sup>1</sup> http://www.mageeq.net/

<sup>2</sup> http://www.quing.eu/

<sup>3</sup> By becoming a part of the "Gender Statistics Database" of the United Nations Economic Commission for Europe (UNECE) Switzerland has taken the first step towards situating its gender equality condition in the international context.

whose role in the persistence of gender inequality and the mechanisms and limits of gender equality activities has been insufficiently studied.

Starting from the need for more knowledge and action in gender equality, the Federal Council mandated the Swiss National Science Foundation to conduct the National Research Program "Gender Equality" (NRP 60) in 2007. The NRP 60 promoted scientific projects, which generated new knowledge about the challenges and limits of gender equality activities in Switzerland to date, and highlighted the institutional mechanisms, discourses and everyday practices, by which forms of inequality between men and women are perpetuated. The articles in this book illustrate some of the most important findings from this program. They stimulate new research questions and perspectives and establish a basis for developing innovative objectives and measures of a sustainable gender equality policy for Switzerland.

#### *About the articles*

*Gesine Fuchs, Andrea Leitner* and *Sophie Rouault* present a comparative analysis of governance regimes for equal opportunities. Their comparison of federal administrations in Germany, Austria and Switzerland situates equal opportunity measures in Swiss governments in the context of the most similar cases of late comers with respect to gender equality policies. The article shows that the gender governance regimes in the three countries do not follow common nor specific patterns. Rather, each equality regime shows a variety of strengths and gaps. In comparison with the private sector the federal administrations in these countries provide much more favorable provisions for gender equality such as policies for career promotion, quota, training, coaching and mentoring. However, heterogeneity of these equality measures is especially high in the Swiss federal administration although gender equality is an integral part of the personnel strategy. A lack of planning and controlling of these measures seems to be one of the major gaps not only in Switzerland, but also in Germany and Austria.

*Eva Nadai* investigates the promises of Swiss new so-called activation strategies in the labor market and social policy with regard to low qualified unemployed women. Drawing on feminist insights on the ambivalence of labor market integration of women as a route to gender equality, the author employs a concept of autonomy referring to capabilities for work, care and voice in order to analyze the effects of four different women-only and gender mixed Swiss activation programs. Based on an ethnographic approach the author explores how welfare workers as street-level bureaucrats in different welfare state agencies frame the notions of work, care and the clients' responsibilities, and how these cultural frames shape their 'investment' practices. Correspondingly, the clients' aspirations and their experiences too are taken into account. Findings reveal that in the light of scarce support for care obligations and further training the gender equality potential of enhancement and enforcement to labor market participation for the specific group under study remains ambivalent.

*Daniela Gloor's* and *Hanna Meier's* contribution provides an introduction into Switzerland's development of policies against violence in marriage and relationships, and critically reflects on the status quo in the fields of public discourses, policies, and research. The aim of the article is to evaluate the adequacy of current measures and services from the point of view of those women who have been affected by partner violence, and to identify persistent obstacles and difficulties of the victims in addressing institutions for support. Narrative in-depth interviews with 40 women in German speaking Switzerland serve as empirical basis. The findings basically confirm the usefulness of developing policy measures for victimized women as well as current interventions for victim protection and support. However, significant gaps and problems in implementation remain: too often violence continues, and perpetrators remain unhindered. Violent men do not receive helpful institutional support. Further, institutional contact results not only in support, but in a burden for women, when they have to keep up close contact to agencies. Starting from the experiences of women, the authors suggest a revised model for the cycle of violence and intervention.

The article by *Andreas Balthasar* and *Franziska Müller* offers new insights into the outcome of Swiss social transfer and tax policy reforms. The authors ask if and how gender equality knowledge is an integral part of these policies which are important instruments to either reproduce gender inequalities on the labor market and in families or to overcome these inequalities by policy innovation. Starting from this question, the article scrutinizes policy-making in social transfer and tax policy by focussing on how often, by whom, and why knowledge about possible effects on the compatibility of work and family was considered. The authors conclude that policy-making based on gendered evidence is characterized by extreme variety and inconsistency. Nevertheless, the analysis shows that administration plays an important role in bringing in gender-sensitive knowledge into policy-processes. Therefore, an improvement of gender mainstreaming processes needs gendered training and information for public officials. Also, the authors show that equality experts and their knowledge often are excluded from these policy processes. This, they conclude, also prohibits gendered evidence-based policy-making.

*Christine Zollinger* and *Thomas Widmer* establish the argument of "bounded possibilities" for gender equality and social cohesion in Swiss municipalities. The authors analyze childcare policies and argue that there is a broad variety in terms of organisation, provision, and quality standards. The authors distinguish four types of childcare services: traditional, market-liberal, social integrative and the universal system. While the traditional system has low potential for gender equality and for social cohesion, the market-liberal model has high potential for gender equality but low for social cohesion. Social-integrative childcare on the other hand has low potential for gender equality but scores high for social cohesion. Universal childcare, finally, has high potential for both gender equality and social cohesion. Case studies in two municipalities show that these variations are rooted in different framing strategies of political actors as well as in cantonal regulations in policy processes on childcare policies. But also the users' choice has an impact on the childcare system in the two municipalities under scrutiny.

The article by *Gisela Hauss* is dedicated to the intersectionality of gender and further dimensions of social inequality in institutions of social assistance in Switzerland. It focusses on the (re-)production of inequalities in the context of precariousness and poverty, an area which is often out of sight when gender equality issues are concerned. The text highlights the particularity of the Swiss welfare state, where the 'activation' of the unemployed is embedded into the paradigm of social investment, and critically reflects on related political measures. The case studies start from the question how, at the level of professional practices of consulting, life situations of unemployed women as welfare recipients are consolidated or changed. Based on an ethnographic approach, the author reconstructs different 'configurations of inequality', as they result from interactions between consultant and welfare recipients in three labor-market integration programs. As it is illustrated here, the category of gender becomes not an 'omni-relevant' category, but develops relevance in interdependence with further categorizations such as motherhood, age or ethnicity, type of residence, with considerably different consequences for women.

*Kathrin Bertschy* addresses wage discrimination at career entry in Switzerland, starting with the assumption that gender differences in early career wages should no longer occur, since young women today match or surpass men in educational achievements. Based on longitudinal data from the Swiss 'Transition from Education to Employment (TREE) Survey', the analysis shows that equally qualified women with identical occupational training already at the beginning of their career start with lower wages than their male counterparts. This can be partly explained with lower remuneration for female-typed occupations and the fact that even when equally qualified women often choose specific work roles that coincide with lower compensation. However, early wage discrimination is not only due to persistent gender segregation in the Swiss labor market and education system but also to an unexplained wage differential amounting to about 7%. In a discussion of reasons the author highlights the mechanisms of statistical discrimination and gender biased promotion and networking.

*Michael Nollert* and *Sebastian Schief* focus on cantonal patterns of gender-specific time-inequalities in paid and unpaid work based on data from Swiss Labour Force surveys (SAKE). The analysis reveals significant patterns of time inequality by cantonal groupings, among which Geneva and Basel City show the strongest pattern of a more equal distribution of paid and unpaid work. At the opposite end, gender specialization was highest in traditional, rural cantons. The results also show that there are different but distinct institutional configurations associated with these cantonal groupings, among others traditionalism, welfare spending, and size of public sector. With respect to political initiatives for change the authors emphasize not only the need to change to working time policies and working time organization, especially for men but also to promote innovative measures strengthening gender time equality in traditionalist cantons. Last but not least the presented results for Switzerland point to the need for further research as the role of welfare regimes in shaping gender-specific time inequalities tends to be underrated.

To the background of new initiatives to expand child care *Christina Felfe, Susanne Stern* and *Rolf Iten* address the role of childcare institutions as a policy tool to enhance gender equality. Based on cantonal data, they, for the first time, map in detail the regional provision of childcare in Switzerland including both coverage rates for early childcare and after-school care and show that both vary substantially between cantons, but also within cantons due to municipalities discretion in the provision of childcare. Furthermore, based on a newly set up data set with different cantonal and communal data, the authors can demonstrate that cantonal political enforcement of after-school care provision is able shifting the supply of respective care slots to a substantial degree. Another eminent insight refers to parental labor market participation. The findings show that each newly created after-school care slot caused one more mother to work full-time opposed to not working or working parttime and one father to reduce his work-load from full-time to part-time work.

The article by *Julia Nentwich, Franziska Vogt* and *Wiebke Tennhoff* focuses on the (re-)production of gender (in-)equality in the field of nurseries. Based on perspectives from organization studies and educational research the authors focus on the 'gendered subtext' (Acker) of everyday interactions and spatial arrangements. The empirical analysis of video data, of observations and interviews within four Swiss nurseries reveals mainly two distinct patterns of organizational practices: While in two nurseries the demands of care, meals, hygiene and order seemed to dominate organizational practice and follow the traditional femininity of the ideal housewife, in the other two nurseries care activities are basically organized as opportunities for learning and seem to overcome gendered organizational practices of childcare. Only the latter context seems to allow organizational change in the direction of a less gendered organization. The authors discuss these results in terms of their consequences for the understanding of care and education and conclude that a de-gendering of nurseries is highly relevant for educational quality.

The NFP 60 received some media attention in the Swiss press. *Andrea Maihofer* analyzes these reactions to the program in order to assess gender equality and gender studies in Swiss public sphere. Besides reflecting on the results of the program she critically assesses media coverage in the light of recent movements against gender studies and gender mainstreaming, public voices, which she labels as "anti-genderism". Some media articles explicitly criticized gender studies as "non-scientific", and at the same time display cognitive and affective resistance towards gender studies but also against gender equality policies. Obviously the deconstructive turn of gender studies, i. e. the claim that "gender" is not based in nature and biology, raises critique as well as fear. Maihofer claims that gender equality and anti-discrimination policies seem to be far from being accepted and are, to the contrary, disputed and contested. Moreover, the article argues that these struggles over the meaning of gender equality – and the NFP 60 – point to fundamental societal disputes about the Swiss gender order. The NFP 60, hence, is a player in a gendered "politics of truth" (Foucault).

### **References**


equality/files/gender\_pay\_gap/ 130424\_final\_report\_role\_of\_men\_en.pdf (retrieved 2.3.2015).


# **2 Gender equality policies and their implementation**

# **Comparing governance regimes for equal opportunities: Federal administrations as employers in Switzerland, Germany and Austria**

*Gesine Fuchs, Andrea Leitner & Sophie Rouault*

## **1 Introduction**

Paid employment is a core field of action for all gender equality policies, since individual autonomy has gained ever more importance in current welfare state reforms and in the new paradigm of the social investment state (Morel, Palier, and Palme 2012). Yet, resilient inequalities subsist despite a wide range of gender policy measures. Recent research has shown that fragmentation and conflicts between transversal equality policies and related policy fields are responsible for only limited outcomes (See Prognos 2014, Schratzenstaller 2014, BMFSFJ 2011, Appelt 2009). We follow this stream of research and comparatively analyze the interplay of institutions and measures of corporate equal opportunity policies in one type of organization – the federal administration in Germany, Austria and Switzerland. Public administrations have several interesting features: they are a typical and important sector of employment. They are legally bound to promote gender equality, yet their respective policies are the subject of political interventions and conflicts. Thus public employers implicitly and explicitly act as model employers. Furthermore, their policies are a genuine organizational activity and we can thus analyze the interaction of different instruments. For a better future policy design in general it would be necessary to understand the policies' underlying dynamics. In this respect, the possibilities of effective equal opportunity policy is the focus of our analysis.

We define equal opportunity policies as all programs, measures and laws that aim at equal participation of men and women in paid work, especially in terms of access (education, working hours, childcare, parental/family leave) and quality (employment conditions, tasks assigned, remuneration, further education, career, dismissal or protection against sexual harassment, see Wahl 2005, 68; Bothfeld, Hübers, and Rouault 2010, 37).

Gender equality is a normative and multidimensional concept with legal, social, economic, cultural and symbolic components. Equal opportunity policies are transversal and affect numerous policy fields from social security to structural policies. They challenge principles and norms in the gender specific division of labor and care work, in the social recognition of occupations and in gender stereotypes. Thus we start our analysis from a regime perspective, which is better suited to look at interactions of policies than the isolated study of individual measures alone. Heather MacRae defines gender regimes as follows:

"Gender regime" refers to a set of norms, values, policies, principles, and laws that inform and influence gender relations in a given polity [...]. A gender regime is constructed and supported by a wide range of policy issues and influenced by various structures and agents, each of whom is in turn influenced by its own historical context and path. (MacRae 2006, 524–25)

In order to analyze institutions and measures of equal opportunity policies, we discern three kinds of 'embedded' regimes: the (national) *gender regime* is the most comprehensive concept that covers economic, social, symbolic and cultural norms, values and practices of gender relations. A *policy regime* is comprised of political regulations in a specific policy field relevant for gender relations like employment, social security or reproductive rights. Finally, a *governance regime* is composed of specific institutions and measures elaborated in a specific field of action. We are mainly concerned with these latter governance regimes.

In order to be effective, a governance regime must combine technically consistent and normatively coherent instruments (Howlett et al. 2009, 172; Bothfeld 2008, 13). Such policy instruments are generally based mainly on information (information and awareness raising campaigns, advice), authority (legal or self regulation), treasure (grants, taxes, financial incentives) or organization (e. g. provision of services, government reorganization; Howlett et al. 2009, 114–35). However, the design, combination and implementation of policy instruments have long been known as demanding and conflict-laden processes, where ideas, interests and institutional traditions interact (Peters 2002). Policy measures in given governance regimes are *consistent* if they do not contradict each other and aim at a similar behavior of target groups, e. g. the rise of female employment. Policy measures are *coherent*, if they are logically related and normatively appropriate with regard to current social values and realities (like different earner and care giver models). For example, recent research showed that generous parental leave arrangements and part-time work increase the employment rates of women but may deepen occupational segregation, since the share of women in top positions may decline and the gender pay gap increases (Hegewisch and Gornick 2011).

We assume three quality criteria for the *policy instruments*, namely: their scope, degree of obligation, and differentiation (Bothfeld, Hübers and Rouault 2010). Scope depicts if and how far measures are de facto accessible for all employees. The degree of obligation depends on (material and immaterial) sanctions in cases of non-compliance: rules must be clear, specific, measurable and verifiable in order to take effect. Differentiation means that measures are designed so as to prevent evasions by the target groups. Differentiated measures specify implementation procedures and entail the necessary resources to do so.

We differentiate between institutional and substantial policies. Institutional policies are rules of interaction for different groups of actors in an organization. They do not prescribe specific solutions. Substantial policies are concrete instruments like regulations, incentives, programs or information campaigns, which aim at the direct solution of a social problem and at an altered behavior of target groups. Substantial policies in our case address employees, institutional policies the management in an organization. The latter policies may be as controversial as substantial policies, since they always shift power relations of actors within the organization. We assume a considerable influence of institutional policies on the extent and implementation of substantial policies.

In analyzing equal opportunity policies, we specifically ask (1) if the institutional and substantial policy instruments meet the assumed quality criteria of scope, obligation and differentiation; (2) if the governance regimes are consistent and coherent, i. e. do not contradict each other, aim at the same target and are in line with social values and realities and (3) if any governance regime in the administrations better meet the two abovementioned kinds of criteria for effective equal opportunity policies.

In the next chapter we explain our case studies (2), and after a short description of federal administrations as employers (3), we analyze institutions (4) and measures (5) for gender equality and conclude (6) in what extent conditions for effective equal opportunity policies are met in the administrations.

## **2 Case studies of federal administrations**

Corporate equal opportunity policies of federal administrations are well suited for an analysis of a governance regime with its interactions of institutions and policy measures, since they apply a set of differentiated regulation principles namely in the fields of reconciliation and career development.<sup>1</sup> Furthermore, public administrations serve as a normative model for all employers on how to design corporate policies for gender equality (Henninger and Ostendorf 2005, 11–17). This unique configuration has hardly attracted scholarly attention so far (cf. Strachan 2011, 347) although these policies cover a substantial group of employees, are widespread and elaborated. Existing studies mostly cover the Anglo-American traditions (cf. Conley and Wright 2011, Conley and Page 2010, Conley and Page 2014), mainly provide quantitative data (OECD 2010) or have a sociological focus (e. g. Andresen et al. 2003). However, in our analysis of institutions and measures in federal administrations we can draw on a quite elaborate body of research on gender equality in private companies and firms (Krell, Ortlieb, and Sieben 2011, Funder 2014 and Jeanes 2011). Furthermore, a substantial number of comparative studies on national gender equality policies2 allow us to consider the influence of different gender policy regimes on our policy field.

Our comparison of German-speaking countries follows a most similar cases design (Lauth et al. 2009, 68–75), where cases share key features of the political context: Switzerland, Germany and Austria are federal democracies with similar bureaucratic traditions (see also chapter 3). The three countries share conservative ideas of good motherhood and traditions of the male breadwinner model. They also have in common recent modifications from this model: according to the modernized breadwinner model, men continue to work full-time, yet more and more women are (part-time) employed and remain the principal caregivers.

However, the countries differ in their welfare regimes with consequences for the logic of equal opportunity polices in general: In the *conservative welfare regimes* of Austria and Germany, support for families within expanding equal opportunities logic is quite strong. Extensive parental leaves should support reconciliation of work and family life. In the *liberal-conservative regime* of Switzerland<sup>3</sup> , childcare arrangements remain mostly a private matter,

<sup>1</sup> An analysis of federal administrations as gendered organizations remains well beyond the scope of this article. However, we look at this issue elsewhere (Fuchs et al. 2015, ch. 7, forthcoming).

<sup>2</sup> For Switzerland see Scheidegger 2008, Nentwich 2006; for Germany Lang 2007, Lang 2009, Rudolph 2009 and for Austria Appelt 2009, Lang and Sauer 2012. A rich comparative body of research stems form the Research Network on Gender and the State RNGS and from Quality in Gender+ Equality Policies QUING, see McBride and Mazur 2010 and Haussman and Sauer 2007.

<sup>3</sup> Comparative welfare state research predominantly classifies Switzerland into the conservative regime type (Bambra 2007, 328), but a closer looks reveals a peculiar mix of liberal and conservative elements, such as low level of decommodification and comprehensive subsidiarity, see Pfau-Effinger 2005, Häusermann and Zollinger 2014, 911–19.

with a minimal paid maternal leave in force only since 2005. Liberal traditions however gave way to a comparatively strong individualist anti-discrimination law with possible class action (in force since 1996, see Fuchs 2008). Austrian and German anti-discrimination provisions are weaker, although the European Union significantly influenced their development.<sup>4</sup>

It was not possible to study the entire administrations, so we analyzed smaller units, ministries and offices. These are the ministries of labor and social affairs and finance (in Germany and Austria) as well as the State Secretariat of Economic Affairs and the Federal Finance Administration in Switzerland, which are both part of larger departments of Economy and Finance respectively. The more comprehensive Swiss project included two more units.<sup>5</sup> Swiss offices and units have a significant amount of independence under the roof of a department (with only small central staff). Therefore we compare 'units' as the functional equivalent of the 'ministries' in the other countries. The selection was based on the assumption of the Swiss project: It was hypothesized that the proportion of women in leadership positions and the extent of gender-related political activities influence the implementation of equal opportunity policies; in the units selected, these factors occur in different combinations (Fuchs et al. 2015, ch. 2).

We collected our data in subsequent steps. First, we looked at the legal framework. We then analyzed policy documents like reports, statistics and gender equality plans for the administrations as a whole and for the units and ministries. We then conducted explorative interviews with external experts as well as 6 to 28 internal expert interviews with gender equality officers, line managers, HR experts and members of top management in the chosen ministerial units (Bogner et al. 2005) and coded them thematically. After having analyzed the individual case, we collaboratively compared the governance regime across cases.

## **3 Federal administrations as employers**

Comparative research on bureaucracy (Painter and Peters 2010, Kuhlmann and Wollmann 2013) places our cases in a "Germanic" or "Continental-European" tradition with a strong emphasis on the rule of law, legal regulations and a very high status of permanent civil servants. Cooperative federalism and

<sup>4</sup> For the different logics of compliance, see Falkner, Treib, and Hartlapp 2007 and Kodré and Müller 2003.

<sup>5</sup> The additional units are the Swiss Agency for Development and Cooperation DEZA and the Federal Office for Agriculture BLW.

the principle of subsidiarity lead to comparatively small federal bureaucracies, which focus on tasks related to sovereignty. Historically, federal administrations have been institutions with masculine traditions ("*Männerbund*", cf. Kreisky/Löffler 2009). Women have long been tolerated as salaried clerks only (Gottschall 1990, Kling 1999) mirrored in "celibacy rules" (women had to leave public service when they married until 1920 in Austria, until 1953 in Germany and in Switzerland until the 1960s). Still the women's share of lifelong civil servants in Germany and Austria is well below their share of salaried employees<sup>6</sup> .

All federal administrations under study have been considerably feminized since the late 1970s thanks to growing female education levels and an expansion of the share of salaried employees. In 2011, about 10% of the workforce in the three countries was employed in public administration, which is well below the OECD average of 15% and has been slightly diminishing since 2001 (cf. OECD 2013, 102). Main indicators of female employment are summarized in table 1.


*Table 1: Main indicators of female employment in federal administration and overall economy*

*Sources: Management data OECD 2013, 123; other data Fuchs et al. 2015, ch. 3.*

<sup>6</sup> The status of civil servants has been abolished in Switzerland in 2000.

In the last decade, female employment increased modestly but steadily by about 3% in all cases. Yet, the considerably different levels of female employment, ranging from 25% in Germany to 41% in Austria, are tied to the varying structures of federal administrations and are not indicators for policy success: in Germany, nearly 60% of employees work for the federal armed forces (*Bundeswehr*). In Austria, a proportion of the teachers are federal employees and constitute 30% of the public employment workforce. Employment in federal administrations remains horizontally and vertically segregated with a lower percentage of women in higher positions. Again, our three cases are well below the OECD averages. Yet, our country data show a modest rise of women managers in the last decade. As federal administrations have more elaborated and transparent pay schemes than the private sector, their gender pay gaps are much smaller than on the respective national averages, namely 12% compared to 18% in Switzerland, 15% compared to 24% in Austria and 1% compared to 22% in Germany (Fuchs et al. 2015, ch. 3). Finally, the federal administrations have a gendered structure of part-time employment, which is well below national averages for men and women. This mirrors the male tradition of these organizations. Part-time work has increased over the years for both genders, but the female dominance persists.

## **4 Comparing institutions for gender equality**

All federal administrations are legally obliged to actively promote gender equality. Institutional regulations are detailed in specific laws in Austria and Germany, while in Switzerland, the Law on Federal Employment is a framework legislation in the tradition of New Public Management, which leaves specific gender equality arrangements at the discretion of departments and administrative units. Yet, the general Gender Equality Act binds public employers also. Federal Employment laws in Germany and Austria contain differentiated rules concerning various forms of leaves, part-time or telework, whereas Swiss federal employees have less legal entitlements which are regulated in ordinances and decrees. In order to attain gender equality, the legal regulations in Germany and Austria focus on the promotion of women (eg. via quotas) and reconciliation measures, while Swiss regulations stand in the liberal tradition of anti-discrimination with a comparatively stronger Gender Equality Act (see Fuchs 2013). Such differences also exist when it comes to gender equality officers and gender equality plans, essential instruments of institutional policies.

#### *4.1 Equality officers: status, role, and coordination*

Since the early 1990s, the positions of gender equality officers have become increasingly institutionalized. German and Austrian gender equality laws for the civil service have provided them with a special status and varying participatory rights concerning the personnel policies of their respective department. The Swiss federal employment law bestows departmental and administrative entities responsibility for the draft and implementation of gender equality policies. This led to very heterogeneous and 'situational' institutionalization: they either established a centralized staff position, set up equality representatives or integrated equality issues into personnel commissions.

German equality officers are elected every four years by female employees, while Austrian and Swiss officers are nominated by their head of department or unit. This gives the German officers legitimacy as representatives of female employees, while their counterparts are rather representatives of the employer. Generally, equality officers assume a dual role: on the one hand, they offer individual guidance on anti-discrimination and reconciliation issues to employees, on the other hand they have an advisory role towards their human resource department or towards any line manager. This ranges from an occasional consultative role in the Swiss case to a mandatory participation in all gender relevant personnel decisions in the German case. Equality officers participate in recruitment or promotion procedures with an advisory role (routinely in Germany and Austria, only rarely in the Swiss Foreign Office). Officially, German and Austrian officers are 'watchdogs' in these procedures as regards quota regulations. Their participation in strategic decisions of personnel policy planning is rather consultative. Swiss gender equality officers can give advice on figures, measures and aims for the Federal Personnel Strategy. German and Austrian officers are consulted on a more regular basis. Nevertheless, the practical influence of Austrian officers strongly depends on their hierarchical position. German equality officers have to be informed extensively and at an early stage about such decisions and their opinion has to be adopted or counter-argued within a precise deadline. This broad participation in the German case has been secured by federal court judgments, which hint to the contentious character of these rights. A binding and differentiated regulation of consultation and participation rights of equality officers is more likely to result in the routinized consideration of gender equality issues of personnel policy, yet it does not guarantee it. Table 2 summarizes the differing degrees of obligation as well as scope and differentiation as our three assumed quality criteria.

Equality delegates differ in their area of responsibility. German officers are responsible for gender issues only, their Austrian counterparts should promote equal treatment case of ethnicity, religion/ideology, age and sexual orientation. Gender equality in Switzerland is part of the 'diversity management' portfolio along with multilingualism and the promotion of handicapped employees. The integration in the personnel policy gives gender issues only limited institutional visibility.

German and Austrian equality officers are free from any instructions in their position, while Swiss officers act in a hierarchical context. In German and Austrian contexts, the number of gender equality officers are pre-determined according to the number of employees of a department, but without fixed time resources for the Austrian Gender Equality Officers. Swiss equality officers are working on a strongly reduced part-time basis (with an exception at the Foreign Office) and they are from time to time supported by a group of voluntary colleagues or by the personnel commission.

The coordination and networking of equality officers is crucial for the exchange of information and strategies. It also helps to build a shared professional identity. In Austria, the Minister of Women's Affairs heads a large working group gathering equality officers from all ministerial departments and other federal organizations that meets regularly. Less institutionalized in Germany, this networking activity was merely acknowledged in 1991 by the law but the corresponding inter-ministerial body (IMA) was provided with a solid legal status only in 2015. In the Swiss federal service, this coordination used to take place in a so-called Forum before being abolished in 2012, reinstitutionalization attempts being underway. Interviews suggest that attempts for a stronger institutionalization in Germany and Switzerland failed, because such a co-ordination was perceived as a threat to the autonomy of ministerial units in dealing with gender issues.

The professionalization of gender equality policies is still fragile. Whereas German equality officers see the regular development of their career guaranteed by the law, an official acknowledgment of the specific competencies acquired in this office is not to be found in any of the civil services under study, which hints at the limited legitimacy of the function.


#### *Table 2: Gender equality officers in comparison*

*\* Federal Department of Foreign Affairs*

Undoubtedly, interventions of gender equality officers can be more efficient and targeted in the implementation of equal opportunity measures, when the law comprehensively pre-defines their status, gives them sufficient resources, competences as well as information and exactly defined participation rights and when they are networked with other colleagues and gender equality institutions. The degree of obligation of gender equality officers seems to be highest in the German case, where their legitimacy is strengthened through their election and where they have broad competencies to participate in personnel decision making. These competencies are also quite differentiated and the limitation on gender issues helps to specialize and to focus resources. The Swiss case underlines that the institutional integration of gender equality issues in personnel policymaking leads to their weak institutional visibility; moreover, it makes policies dependent exclusively on the management – or a strong personality of the officers. The actual scope of the institution gender equality officers remains limited mostly to their own organization; only the Austrian working group provides substantial time and space of exchange and discussion. In all three cases, however, the status of equality officers remains nevertheless fragile: as exchange and networking remain (legally) inadequate, the development of gender equality officers as a profession is hardly possible.

#### *4.2 Equality plans: strategic instruments with limited effectiveness*

In all our cases, the departments have a *legal obligation* to periodically establish equality plans. Such plans have a double role of planning and controlling. They can be a strategic planning tool for the entire administration and should also help the departments to plan, monitor and evaluate the implementation of gender equality measures. Ideally this would lead to mainstreaming gender into all relevant personnel decisions. In general, the human resource departments are responsible for the plans. Drafting, reporting and participation of gender equality officers differ widely across our three case studies:

The *Austrian* Women's Promotion Plans (*Frauenförderpläne*) of every department cover a period of six years. They are published in the Federal Law Gazette and respective statistical data should be updated biannually. They include mostly qualitative measures, whose focus is however strongly oriented towards the fulfilment of legally binding quotas set for the whole federal civil service. A working group of all equality officers in a given department is endowed with the drafting of the plan, while the decision on the included priorities and measures is in the hand of the head of ministry; equality officers moreover monitor the implementation and evaluation of such plans. Regularly, comparative reports for the entire administration are published.

In *Switzerland*, the Federal Office of Personnel drafts every four years the personnel strategy, which has to be adopted by the government. This program defines "target ranges" for women, linguistic groups and handicapped people, i. e. it has mostly quantitative aims. Departments and offices may define other, more qualitative goals. All departments produce annual progress reports (*Fortschrittsberichte*), which cover gender equality issues as well as multilingualism and the integration of handicapped employees. These internal, non-public reports include quantified objectives, planned measures and recommendations. Gender equality officers act mostly in a consultative role and draft the progress reports of their department. The Federal Office of Personnel (EPA) publishes an annual report.

The *German* legislation prescribes that gender equality plans (*Gleichstellungspläne*) of federal ministries entail a statistical status quo, quantified objectives and deadlines and a list of corresponding policy measures; they are due every four years, with an interim report every two years to measure the attainment of set objectives and to justify possibly their non attainment. These reports are internal working documents and not publicly accessible. Equality officers are systematically consulted on the content of the plan of their department and can comment, amend, or veto it; moreover, they are legally entitled to complain about the missing, delayed, or defective implementation of the plan. However, the plans are very heterogeneous and implementation varies: in 2009, half of the departments did not have an equality plan. Establishment and quality of these plans depend thus on the assertiveness of gender equality officers and on gender awareness of the HR units.

Equality plans in the three federal civil services appear quite *differentiated* and have a (legally) *binding* character. But no formal sanctions are foreseen in case of non-execution or non-attainment. In Austria, the publication of ministerial plans nevertheless activates a mild version of negative incentives through 'shaming and blaming'. Unfortunately, this reduces the efforts to publish very ambitious plans.

In all cases, an obligation to regular *reporting* exists and a departmental unit coordinates this process. Controlling and evaluation procedures remain relatively weak, as they are limited to a formal control of quantitative objective attainment in all three cases. Although technically possible, a systematic comparison across departments is carried out in none of our cases. One reason is that this is regarded as infringing upon the autonomy of ministerial units. The definition of women's quotas in *Austria* and target ranges in Switzerland serve as important points of reference. On the other hand, quotas in Germany are hardly relevant, as they are considered as un-implementable. Moreover, benchmarking differs: Austrian reports are published and contain only a few, but comparable indicators; *Swiss* reports focus also on quantitative indicators, and the trade union *Verband des Personals der Öffentlichen Dienste* (VPOD) publishes a 'shadow report' with figures on women's share in management, part-time in management, the gender distribution of performance evaluations as well as bonuses7 The integration of gender issues in the centralized controlling of HR management in Switzerland enables a quantitative

<sup>7</sup> See www.vpodbern.ch/Berufe/Bundespersonal.htm (13.05.2015)

evaluation of equality plans in a regular and systematic manner, but without much qualitative content; The *German* administration lacks a targeted, standardized and analytical reporting, as plans are not published and 'progress reports' (*Erfahrungsberichte*) present only aggregated data. .

In sum, none of the administrations fully exploit the potential of gender equality plans. In order to take effect, plans should contain three elements: (1) a (statistical) status quo on gender equality in each department, (2) a list of (quantified) objectives and of measures aimed at reducing remaining inequalities with a precise timeline and (3) possibly an obligation to justify the non-attainment of objectives previously set. Finally, they should have a high degree of differentiation as regards a list of indicators to be surveyed and common guidelines. Only recently the Austrian Ministry of Women's Affairs has provided a detailed guideline on how to draft and implement such plans as strategic instruments. Additionally, gender equality officers should participate possibly in the plans' development and in the compilation of the data. Finally, a centralized controlling of their adequate implementation across the whole federal administration is necessary to prevent evasion or lip-service. Systematic reporting and comparison could make successful strategies and problems visible and thus bring about coordination and adjustment. There are points to start with, like the annual Swiss reports and 'shadow reports'; in Austria, the straightforward comparisons of several indicators could be expanded by analyses that take into account specific conditions in the different ministries.

## **5 Comparing measures for gender equality**

Non-institutional or substantial measures for equal opportunities in the three cases are concentrated on the reconciliation of work and family life for parents and on the career development of women, mainly of highly qualified employees. Other target groups are addressed only with single measures or passive equal opportunity policies like anti-discrimination rules or gender sensitive language.

### *5.1 Reconciliation of work and family: between part-time employment and parental leave*

In all three federal administrations most experts we interviewed agree that reconciliation policies are a central pillar of equality policies. This is not surprising at all, since they do not provoke distribution conflicts – as it can be the case by the promotion of female careers, and since in principle they address both women and men. However, these measures in fact mainly address women due to the still vivid traditional gender specific division of labor.

*Part-time work* and *teleworking* are rather new developments and their share is still lower than in the private economy (see table 1). The possibility of part-time work for civil servants was introduced in Germany only in 1969, in Austria in 1979 and in Switzerland in 1986. In all three cases part-time work and teleworking represent the lynchpin for enabling continuing careers by women with small children. Schemes supporting the right to part-time work and teleworking for parents of young children weaken the traditional principle of permanent availability and the culture of presenteeism of managers. This is especially visible in Switzerland and Germany where part-time work increases also in higher positions

In Switzerland, the increase of part-time work is mainly the result of changes in norms and values. Until 2013, the possibility of part-time work was mentioned in very general way only in legal documents. Nevertheless, part-time work increased and seemingly created only few conflicts between employees and management. Since then, parents after maternity leave have the right to reduce their hours to a minimum of 60%. In Austria and Germany, employees and civil servants are legally entitled to part-time work, but there are still differences for civil servants, who cannot reduce their working time to less than 50% or who cannot work in part-time longer than 15 years in both countries. The proliferation of part-time also poses challenges: it is demanding for teams (coordination, team spirit) and may lead to an overburdening of part-time workers, because often the reduction of work time is not compensated for. There are also hardly rules for employees who wish to return to full-time work. However, the available data for Switzerland indicates that part-time is mainly used with high working shares of about 70% of fulltime obligations. Finally, the promotion of part-time work perpetuates the gender-specific division of labor and thus it may negatively impact on female careers: in Germany, full-timers are promoted 2.3 times more often than part-timers (Fuchs et al. 2015 ch. 5).

Another important instrument for reconciliation in federal service are *parental leaves*, which follow the traditions of the national policy regimes. In *Austria*, parental leave regulations for the public sector exceed the provisions for the private sector and allow paid parental leaves for up to 30 months. Also, the highly differentiated models of parental leave are supplemented by a 'daddy month' for public administration.<sup>8</sup> In Germany, the share of men among parental leavers is about 12% – slightly under the share of men in the private

<sup>8</sup> The *"Papamonat"* is a special arrangement for public administration entitling fathers with a paid one-month leave after childbirth.

sector. In Austria the share of men is estimated to be higher than in the private sector. Mirroring liberal welfare traditions, Swiss federal administration has no paid parental leave after 16 weeks of paid maternity leave, although fathers in federal administration get two weeks paid leave.

Part-time work and long parental leaves show their effect in discontinued employment and impact negatively on career development. Austrian gender equality plans especially address this problem and provide support for reintegration after leave is taken, promoted through individual counselling and coaching, organized meetings or information during parental leave as well as training opportunities for returners. Similarly the German equality law foresees such measures. Explicit inclusion of men in parental leave schemes and the expansion of public childcare facilities should foster female employment but they do not reframe the ideology of the male breadwinner/female caregiver model. With incentives for long career breaks on the one hand and expensive supporting programs for returners on the other, this policy profile results in a stop and go-policy for women's employment (Leitner et al. 2013; Appelt 2009). However, in Switzerland continuous employment, probably in part-time, emerges as a new *Leitbild*, due to the lack of parental leave arrangements. Unfortunately, this is not a satisfactory alternative, since in both models of reconciliation – continued employment or re-insertion after parental leave – the main share of early childhood care remains women's work. Parental leaves and part-time thus remain very ambiguous instruments for gender equality. *Childcare* is a precondition for continued employment of parents. In Germany this seems not to pose a problem since the administration is mostly centred in cities and a statutory entitlement for a place in nurseries is in force since 2013. In Austria, experts appreciated high-quality childcare in employees' day care centres; yet all-day childcare is in short supply even in Vienna. The Swiss liberal-conservative regime defines childcare still as a private responsibility. Some federal departments used to offer their own childcare facilities. More importantly, high private childcare costs are subsidized for parents with monthly gross household incomes up to 20,000 Swiss francs. There are concerns that this subsidizing scheme has trap/threshold effects and thus may pose an inconsistent incentive, namely to reduce working hours. Still, this instrument goes beyond the otherwise scarce promotion of reconciliation.

Reconciliation measures are especially differentiated in Austria, to a slightly lesser degree in Germany. The promotion of reintegration and "daddy months" reduce the potential burden of mothers, however, without changing the gendered division of labor. In contrast, the few Swiss measures are hardly differentiated, which also reduces the scope of reconciliation strategies – they remain a private matter that has to be negotiated by couples. The development of part-time work shows that its proliferation can be the result of legal entitlements as well as changes in values and norms.

#### *5.2 Women's career development: between individual promotion and quotas*

Gender quotas in all occupational fields and at all hierarchical levels of federal civil services are the main instruments to stimulate the promotion of female careers. These quota regulations are explicitly anchored in gender equality law in Germany and Austria. Swiss 'target ranges' setting the proportion of women to be reached in the different salary categories are laid down in the Personnel Strategy. Female career promotion mainly addresses higher qualified women since male dominance persists especially in high positions.

In Germany and Austria gender quota regulations for recruitment and promotion take the form of so-called "performance-based quotas" (*qualifikationsabhängige Entscheidungsquote*) that allow a preferential treatment of women in hiring and promotion procedures under restrictive conditions – there must be a gender underrepresentation in the occupational sector considered and the candidates in competition for the post offer should have the "same qualification".<sup>9</sup> Austrian quota regulations with standardized procedures have led to more transparent and professional recruitments. Yet, the concept of "same qualification" is often too vague to be efficiently applied. As observed in Germany, "sameness" in qualification, i.e., in aptitude, capability and technical achievement is evaluated along ever more refined criteria, so "same qualification" is hardly ever stated, depriving quota regulations of any effect (cf. BMFSFJ 2006: 34; Bednarz-Braun 2000: 139–142; Papier and Heidebach 2014). Furthermore, their implementation lacks resources, predefined (interim) goals and evaluation procedures. Although gender quota regulations in the Austrian and German federal civil services are developing limited effects in individual recruitment or promotion cases, yet their symbolic value ensures indirect effects in process regulations, support for equality officers and last but not least for the implementation of sustainable measures to promote women's careers. In Switzerland, the target ranges of the federal personnel strategy for the proportion of women in different hierarchical levels serve as incentive for female promotion, especially because they are accompanied by reporting obligations. In Austria regular publications on the participation of women in different departments, functions and salary categories are objective indicators for a success.

<sup>9</sup> Preferential treatment must not be applied if it would lead to unreasonable hardship for the candidate of the over-represented sex (so-called "hardship clause").

As a consequence, training and individual support of women should increase the number of potential applicants for higher positions. In Austria this is promoted through a set of different measures like privileged invitations of women for further education, special courses for women, coaching, mentoring and supervision and also gender-sensitive teaching materials. Similarly, the Swiss administration offers coaching, mentoring and 'potential analysis'. While training is focused on higher qualified women in Austria and Switzerland, the German federal administration offers a broader access to tailor-made training schemes but specific measures for women are rare and often *ad personam*. Whether the increasing qualification levels of women and the shortage of skilled workers will automatically improve the situation of women is a topic of discussion especially in Germany. Undoubtedly, promotion structures, respectively the public employment law with lifelong careers within the organization are an obstacle to gender equality.

To sum up, the implementation of measures for career promotion of women – like gender quotas, principles for recruiting processes, performance assessments, training measures and individual support – is very heterogeneous across departments and also depends on organizational cultures. Austrian and German quota regulations have a higher degree of obligation, but they lack differentiation of goals and clear definitions of 'qualification'. Swiss target ranges are less obligatory, yet profit from differentiated planning and evaluation. Substantial policies in the field of career promotion can have an immediate effect on gender imbalances, particularly in higher positions. Thus they are more likely to produce distribution conflicts than reconciliation measures. Hence, the former are far more contentious than the latter – and have a lesser degree of obligation. However, recruitment and promotion regulations put reconciliation measures back on the agenda. Reconciliation and career promotion are two complementary sets of instruments for a comprehensive and effective gender equality strategy.

## **6 Conclusion**

Our analysis shows that none of the three governance regimes has a constant weak or strong combination of criteria for efficient gender equality policies. Rather, our cases have different strong points and gaps. In general, federal administrations design more and more favorable provisions for gender equality, well above the legal standards set for the private economy. As steady modest increases in gender equality show, they set model regulations for reconciliation, recruitment and female career development.

Yet, actual policy measures and the use of institutional means differ strongly between administrative units and ministries in a country. Heterogeneity is especially high in Switzerland and is hardly related to objective reasons like the tasks, structure and duties of the various units. To the contrary, this heterogeneity occurs even when policies are closely regulated, but insufficiently controlled like in Germany and Austria. Ministerial autonomy increases the probability that regulations are adapted according to differing organizational cultures and individual experiences of managers. Diverging implementation appears arbitrary and incidental. Gender equality policies, then, may be seen not as a routinized and legitimate state activity.

At the same time, institutional policies became increasingly differentiated and more and more coordinated with general personnel policies – this concerns the competences of Austrian and German gender equality officers as well as the gender equality plans and reports. But a systematic monitoring and a standardized benchmarking do not exist in any of our cases.

Substantial policies for career promotion exist, but are not always implemented thoroughly, or they produce conflicts with reconciliation policies. The use and non-use of quota regulations show that a re-definition of 'qualification' and 'performance' is crucial. A broad range of training, coaching and mentoring exists to strengthen the pool of qualified female employees, but units and departments have a wide discretion and these instruments also depend on the organizational culture or the persons in management positions of a given units.

All three governance regimes for corporate gender equality in federal administrations are normatively coherent with the modernized breadwinner model: part-time work and reconciliation measures allow for a continuous female employment in federal service. More precisely, they are also coherent with their different national gender regimes, and differences especially in the degree of legal obligations correspond to the more liberal or more conservative state traditions. The Swiss liberal profile is mirrored in the emerging norm of continuous employment in federal administration. In Austria and Germany the national expanding opportunity profile for women matches the extended parental leave and reinsertion schemes as well as quota provisions. On the other hand, there is no consistent and binding implementation of effective career promotion. Long-standing part-time careers of women and their slow progress in management positions correspond at least partly with the modernized breadwinner models, where women still are the flexible reserve. This will not lead to comprehensive gender equality.

Altogether, gender equality policies are well dovetailed with personnel policies of units and departments, especially in Switzerland, where gender equality is an integral part of the personnel strategy. A high degree of consistency however would also mean to critically examine (and re-design) apparently gender neutral regulations for recruiting, performance evaluation and promotion. This does not take place systematically in any administration, but awareness of this problem is on the rise.

The main reason for insufficient gender equality regulations in all three federal administrations seems to be a lack of specific monitoring and controlling instruments. The traditional implementation autonomy of ministries and administrative units outplays the weak coordination of gender equality policies on the level of the federal administration. Gender controlling, then, is limited either to quantitative features, or reports facts only on an aggregate level (like in German gender equality reports). If indicators are disaggregated, they are mostly limited to the participation of women without contextualized interpretations of the differences (like in Austria). Finally, indicators may be mainstreamed to such a degree that gender equality becomes a rather unimportant feature among others and is hardly visible (like in Switzerland).

Probably the most important result concerns the significance of institutional policies especially in planning and controlling. It is important to develop differentiated monitoring instruments that provide refined indicators, clear objectives, specific measures and procedures that are elaborated enough not to be circumvented. Equally necessary are sufficient human and financial resources. However, a further differentiation of institutions and measures alone will not bring about gender equality in administrations. A public debate about equality issues and objectives could promote gender equality as a legitimate and commonly shared value also in public administrations.

Further research on corporate gender equality policies in civil services should evaluate, firstly, how regulations impact on every-day practices, and secondly, which chances and limitations prevail in these gendered organizations. International comparative research should compare and analyze the effects of the two reconciliation profiles – continuous employment versus re-insertion – on the quality of employment and female career development. Access to more detailed data than those available to us in this study would be necessary, including data on part-time employment models, duration of leaves and individual careers.

#### **References**

Andresen, Sünne, Irene Dölling, and Christoph Kimmerle. 2003. *Verwaltungsmodernisierung als soziale Praxis: Geschlechter-Wissen und Organisationsverständnis von Reformakteuren.* Opladen: Leske + Budrich.


## **Whose welfare – whose autonomy? Welfare, work and care in social investment practice**

*Eva Nadai*

## **1 Introduction: Women's work – choices and constraints**

In much of feminist theory, labor market participation is regarded as indispensable for women's autonomy and emancipation. Employment provides (partial) economic independence from a male partner, thus allowing for a more equal power balance (Hobson 1990), as well as access to the public sphere. Wage work is associated with social contacts, social recognition and self-fulfilment – in short, "employment is not just a 'good thing', it is crucial to the welfare of the adult" (Lewis/Giullari 2005: 97). However, as critics argue, "hegemonic welfare state feminism" (Ostner 2004: 45) tends to generalize the value of labor force participation. For many women the normative model of autonomy through continuous employment is neither accessible nor desirable, and there are cultural and class differences with respect to the meaning and relative importance of work and care and to unequal access to employment that is actually rewarding in terms of pay and self-fulfilment. Moreover, as Nancy Fraser (2009: 110) notes, feminism's high esteem of wage labor devalues unpaid care work, and by supplying "a good part of the romance that invests flexible capitalism with a higher meaning and a moral point" unwittingly legitimises neoliberalism. Empirically, women have indeed entered the labor force in increasing numbers, and their employment has become more continuous (BFS 2008). However, in some countries with a high female employment rate – notably in Switzerland – this increase is mostly due to part-time work, and in many countries there are still clear gender differences regarding employment patterns (ibid.). While these facts are undisputed, the interpretation is not. The question is whether women's limited labor-force participation reflects their "true" *preferences* or is merely a consequence of *structural and cul-* *tural barriers to continuous full-time employment* according to the male model, and, to what extent it hampers women's chances for economic independence (Allmendinger/Leuze/Blanck 2008; Hakim 2000; Jones 2012).

There is another uncanny affinity between feminism and mainstream policy, namely between the feminist emphasis on women's labor market participation and the post-welfare state's focus on universal inclusion into the labor market. Within the framework of the *social investment paradigm* state expenditures are conceptualized as investments in human capital to generate productive workers for competitive labor markets. A strong labor force is, moreover, regarded as necessary to ensure the long-term financing of social security systems (Esping-Andersen 2002). The issue of individual autonomy qua wage work espoused by feminists is here framed as "personal responsibility", thus as a duty of the individual towards society. Women figure prominently in the social investment strategy: as underutilized labor force potential, as contributors to family income, and to social security systems, and, moreover, they are addressed in their reproductive capacity as mothers and carers of children as the future workforce. They are caught between the "farewell to maternalism" (Orloff 2006) and a "new maternalism" (Jenson 2015). On the one hand, social policy is pushing for the "adult worker model" that expects labor market inclusion from every adult person, irrespective of gender and care obligations (Daly 2011; Lewis 2001). This entails reconciliation policies to facilitate all women's employment (Stratigaki 2004), as well as activation policies for the special target group of social transfer recipients (Betzelt/ Bothfeld 2011; Rudolph/Niekant 2007). On the other hand, women are functionalised as mothers and secondary earners, while gender equality issues like the gender pay gap or discrimination in the labor market are relegated to the background (Jenson 2015).

In this context matters of choice and constraints are again different for *unemployed women,* as welfare beneficiaries are often *forced into the labor market*, albeit as secondary earners only (Bothfeld/Betzelt 2011a; Jaehrling 2009). Regarding the situation in the U.S., Dodson (2007: 262) observes a veritable "ban on motherhood": the radical enforcement of welfare-to-work policy actually prevents poor mothers from caring for their families (see also Solomon 2008; Weigt 2006). The situation is, of course, acerbated for single mothers, who are especially vulnerable to the gradual abolishment of maternalist exemptions from work obligations (Kull/Riedmüller 2007; Wright 2011). Yet, even under conditions of forced labor market participation, there is some empirical evidence that these women still appreciate work and activation programmes, because it helps them escape the stigma of welfare dependency, bolsters their self-esteem and, for some, offers a chance for developing new skills and capacities (Nadai et al. 2013; Solomon 2008; Weigt 2006). However, reconciling work and care is even more demanding for welfare recipients and low income mothers who very often have to cope with a variety of problems (poor health, children with special needs, making ends meet etc.); thus work requirements add significantly more stress (Dodson 2007; Solomon 2008; Weigt 2006).

This chapter examines the *contradictory consequences of social investment policy for women's autonomy* with the example of *unskilled unemployed women* receiving either unemployment or social assistance benefits in Switzerland.<sup>1</sup> For them the promise of autonomy qua wage work seems elusive, yet they are forced to seek a job. While Switzerland has implemented activation principles in the social security system since the mid 1990s2 , the social investment paradigm has been adopted in a "light" version at best (Bonoli 2010). Investments are limited to promoting the employability of social transfer recipients whereas the more encompassing vision of preventive long-term investments in the human capital of the whole population has not really gained ground. Investments in the unemployed consist of individual and collective work and educational programmes. Participation in these programmes is mandatory, as is active job searching; non-compliance is sanctioned by curtailment or cancellation of benefits. The mandatory character of activation is obviously in stark contrast to the idea of autonomy. Yet, investments in women's employability might offer them the chance to gain autonomy by getting a job with an income. Given the limited cultural, economic and social capital of these women and institutional restrictions, the question is to what extent activation is conducive to women's welfare – does it improve their economic and social situation and enhance their autonomy?

Answering such a question requires, first, a discussion of the concept of autonomy and its operationalization (section 2). As Pfau-Effinger (2005: 8) contends, cultural conceptions of wage work and of who should work are key elements of welfare arrangements. These normative frames become manifest in concrete practices within street-level bureaucracies, which, because of the inherently discretionary nature of human services work, operate as mediators of social policy (Brodkin 2011). Thus, I will analyze next how welfare workers frame the notions of work, care and their clients' responsibilities, and how these cultural frames shape their "investment" practices (section 3). The concept of autonomy implies the possibility of choice − it is therefore important

<sup>1</sup> Entitlement for unemployment benefits is based on insurance contributions during at least 12 months within the two-year period before unemployment. Benefits are paid for a maximum of 200 up to 520 days depending on length of prior contribution, age and child support obligations. Social assistance is means-tested and eligibility depends on the financial situation of the whole household.

<sup>2</sup> In the unemployment insurance activation principles were introduced earlier than in social assistance.

to take into account what the clients want and whether they are supported to realize their aspirations (section 4). The paper draws on empirical data collected in an ethnographic research project on social investment and activation practices in Switzerland (Nadai/Hauss/Canonica 2013). The study explored the gender equality potential of these practices by examining to what extent and in what ways welfare state agencies "invest" in unskilled unemployed women. We analyzed the institutional regulations and cultural frames underlying the practices as well as the interlocking strategies of both welfare workers and clients. More details on methodology and data will be presented below (section 2).

## **2 Theoretical framework, methods and data**

Autonomy is a contested, yet central concept in feminist thought, and it has also been proposed as a major goal of social policy (Bothfeld/Betzelt 2011b). For the purposes of this paper, there will not be the opportunity to present this debate in detail. Mainstream philosophical concepts have been criticized by feminists for a number of reasons, most importantly for positing an atomistic, asocial, abstract self, which exists outside of social relationships and strives for an illusory absolute autonomy (Friedmann 2003: 81‒97; Mackenzie/ Stoljar 2000). In response, feminists have developed more socially bound notions, which acknowledge that autonomy is a matter of degree. Friedman (2003: 14) conceives of autonomy as "choices and actions (…) that mirror wants or values that an acting person has reflectively reaffirmed and that are important to her". Choices can be considered autonomous if they come about relatively unimpeded by coercion and manipulation. Furthermore, autonomy requires certain causal conditions, such as having significant alternative options and the personal ability of having and understanding values, making choices and pursuing one's commitments in the face of obstacles. Importantly, Friedman argues for a content-neutral notion of autonomy – in order to be autonomous, a persons' choice itself does not need to endorse the value of autonomy. For instance, if women choose to affirm traditional gender norms (such as caring for the family rather than striving for economic independence qua wage work) this is still considered autonomous behavior if the choice "accorded with and issued from their deeper wants and commitments" (ibid.: 24).

By recognising options and individual faculties as preconditions for autonomy Friedman's concept is close to the *capability approach (CA)*, which puts the real freedom to lead a life according to one's own values centre stage. Whereas Friedman defines abstract conditions for autonomy, the CA is more concerned with concrete social conditions that enable genuine choices (Sen 2000; Nussbaum 2011). "Capabilities" are conceptualized as a space of options for acting and being that is more encompassing than the "functionings", i. e., the options a person actually realizes. Thus, the freedom to choose is an inbuilt feature of the notion of capability. It is dependent on access to material resources, as well as social conversion factors (such as norms, institutions, structures) and individual characteristics (such as physical, emotional and cognitive faculties), which are necessary to make use of material resources.

Both Friedman (2003) and Nussbaum (2011) stress the importance of personal capacities that enable individuals to exercise autonomous choices. Yet, neither discusses in detail how people acquire this capacity. Needless to say, formal education is considered a paramount conversion factor in the CA-literature. However, the competency to act autonomously cannot be reduced to formal education. According to Sedmak (2011) "fundamental capabilities" such as the competence for self-reflection, for developing alternatives to the status quo, for engaging in one's live, for making judgments and decisions cannot be "distributed (…) like bread and water" (ibid.: 36). Instead they are dynamic, bound to the subject and acquired in long-term biographical processes. They have to be *actively cultivated*, lest they waste away.

As a broad normative framework the CA does not give a precise operationalization of what material level or social conversion factors would constitute sufficient conditions for capabilities (Robeyns 2005). *With respect to activation policy* three important capabilities have been suggested: the *capabilities for work, for care and for voice* (Bonvin 2009; Dean et al. 2005; Lewis/ Giullari 2005). Activation policy has been much criticized for forcing the unemployed into programmes and jobs irrespective of their own preferences or of employment quality. In contrast, CA-proponents insist on people's right to choose work that they consider intrinsically and/or extrinsically satisfying: jobs that offer some self-fulfilment and material security. The capability for care implies the right to care for others without incurring disadvantages. In the context of activation, this means that also poor and unemployed parents should be able to choose caring for their families over wage work. Finally, granting welfare beneficiaries the capability for voice means allowing and enabling participation in decisions on activation goals and measures. All three capabilities rest on acceptable "exit options", i. e. on unconditional sufficient material support.

The notions of capabilities for work, care and voice are still rather abstract. This poses the methodological challenge of how to identify empirically, first, people's "genuine" wants, second, the broad array of social and individual conversion factors that enable or obstruct these choices and, third, how these factors are related to social investment policy. Dorothy Smith's "institutional ethnography"-approach provides a promising framework to meet this challenge, because it systematically links subjective experiences with social structures and discourses. Institutional ethnography treats so-called macrosocial phenomena like social structures or discourses not as abstractions but as "doings": as constituted by the observable coordination of action in concrete relations (Smith 2005: 57‒60). Research therefore starts from "actual activities of actual people" (ibid: 54). Yet, such particular standpoints and experiences are not described as an end in itself but serve as "entry point into forms of social organization which shape local settings but originate outside of them" (Grahame 1998: 352). These "ruling relations" (Smith 2005) are textually mediated, thus objectified, systems of knowledge and social organization that transcend local contexts and "generalize and universalize across multiple local settings" (ibid: 42). Texts coordinate action within a particular setting and link it to social relations, which temporally and spatially extend the observed context (ibid: 103). Thus, analyzing texts as "documents in action" (Prior 2008) is crucial to tracking institutional relations that are not directly observable. So, our study focused the interplay of institutional regulations of the unemployment insurance and social assistance with the actual practices of welfare workers, which in turn are shaped by these textually mediated regulations, but also by their organizational and professional beliefs and their negotiations with clients. On the part of the unemployed, we analyzed how their educational, professional and familial biographies and their current situation enabled or limited the way they engaged with the "investments" they were offered or refused.

Smith (2005) herself understands institutional ethnography as a theoretical perspective, not a particular method. In our study, we actually used an ethnographic approach based on participant observation, interviews and document analysis. The observational sites comprized six welfare state agencies: a regional placement office of the unemployment insurance, a social services agency operating within the regime of means-tested public assistance – these institutions decide on investments, i. e. which clients get which measures. As sites where activation measures are applied we chose two women-only programmes and two mixed-sex work programmes. In addition to informal conversations as part of observations, a total of 46 semi-structured formal interviews were conducted: 16 with welfare workers, 22 with clients (16 women, 6 men), 6 with experts and 2 with employers who offer internships as a form of work programme (Nadai et al. 2013: 6‒10). The data were analyzed with grounded theory coding techniques (Strauss/Corbin 1990) and membership categorization analysis (Lepper 2000).

**3 Work as obligation or capability – welfare workers' perspective**

*"Do you want a job to feel good and have a hobby or to be able to pay your bills?"* This question of an unemployment counsellor to her client captures well the dominant notion of work, namely of *work as breadwinning* and as the *moral obligation* to take responsibility for one's own living. Clients who insist on finding fulfilling work or who regard unemployment as an opportunity for a major occupational reorientation are regarded as being *"choosy"* and *"unrealistic"* and as capitalizing on society for advancing their own careers. This is the shared view of work in all six fields of our study. With respect to gender the notion of work as breadwinning has indeed been generalized to women too, including mothers. For example, in social assistance the age limit of the youngest child legitimizing the mother's economic inactivity has gradually been lowered to three years.<sup>3</sup> And, even for single mothers of small children the guidelines of the social services agency in our study postulates re-entry into the labor market "as soon as possible". On the other hand, activation practice must, to a certain extent, take into account clients' personal preferences. As research drawing on Foucault's governmentality approach has demonstrated, activation operates via people's subjectivities and aims at transforming these subjectivities, namely at forming an active entrepreneurial self (Dean 1995). Therefore, clients are also encouraged to explore their occupational aspirations and to develop strategies to realize them (Maeder/Nadai 2009). Work, then, is also cast as a *capability* or *means to self-fulfilment* having an intrinsic value − again for men and women alike. In our study we did not find the idea that employment may be less important for women due to their alternative role in the family (but see Rudolph 2007).

One of the women-only integration programmes gave this meaning of work a distinctly gendered twist. Work was explicitly regarded as a *means of women's empowerment*, namely as a precondition of economic and social independence and a way to realize hidden potentials. As the programme manager put it: *"It's incredible, how they flourish, while they're in the programme."* We did not encounter this argument with regard to women in the other fields of our study, however, where women's access to the labor market did not seem remarkable. A substantial percentage of this programme's clientele are *migrant* women who have never been employed in Switzerland. They are depicted as

<sup>3</sup> In Switzerland social assistance is legally regulated on the cantonal level and administered on the communal level, so there is considerable local variation in practice. The Swiss Conference of Social Welfare provides guidelines to harmonise social assistance practices, which, however, are not binding. The age limit of three years is stipulated in these guidelines.

#### *Eva Nadai*

coming from oppressing "patriarchal cultures", where they were confined to the home, were not used to move freely in public or see themselves as subjects with entitlements and were denied educational and occupational opportunities (for similar stereotyping see Kirk/Suvarierol 2014). Thus, beyond empowering them as women, access to the labor market is also regarded as a *mechanism of social and cultural assimilation*. By leaving home and setting out to work these migrant women can become like liberated Western women (Hauss in this volume; Hauss/Nadai 2009). Labor market participation as a way of cultural assimilation to Swiss society was a common topic in all our research fields. Ethnicity as "otherness" is primarily seen as a major impediment to employment; assimilation, therefore, as a necessary precondition for gaining access to the labor market, and at the same time as a result of labor market participation. Especially for male immigrants, work is conceptualized as the double social duty of breadwinning and assimilation, whereas for women assimilation qua wage work is also regarded as a vehicle for emancipation beneficial for the women themselves. Discourses of social investments, integration of migrants, and gender intertwine, thus leading to *"ethnogendering" the function of work*.

Compared to wage work, *care* still is a *highly and explicitly gendered* issue, namely a female concern. It is primarily seen as a handicap by restricting women's flexibility and availability for the labor market. To put it in the blunt wording of an unemployment counsellor: *"With a child they have a millstone around their neck."* Inversely, the same counsellor said: *"Fathers are not really fathers, fathers are always the workhorses."* We also observed a certain mistrust of mothers who declare themselves ready to work 'too much, too soon', i. e. who want a full time or substantial part-time job despite having small children. The welfare workers suspect that these mothers do not really want to work, but *"hold their hands out, while the husband works."* Again, migrant women were distrusted more than Swiss women. Yet, apart from the issue of clients' real availability for the labor market, the unemployment insurance and its work programmes are institutionally indifferent to matters of care.

In contrast, for social assistance care is an ambivalent issue: on the one hand the cases of means-tested welfare comprize households, not individuals. Therefore, the combined earning-capacity of the whole family comes under scrutiny. So, mothers are regarded as potential breadwinners too, and care is a secondary concern. Consequently, welfare workers are ready, *"to shake a patriarchal family model, in which the woman stays at home and the man goes out to work"*. Yet, degendering is pursued only to a certain degree. Social workers do not urge a role reversal, not even when the wife clearly has a better chance to find a job or earn more. Rather, they aim for a "one-and-ahalf-breadwinner-model" at most, with the father working full-time and the mother part-time (see also Keller/Modak/Messant-Laurent 2013; Streuli/ Kutzner 2005: 315). On the other hand, the social investment paradigm regards the family as an important site of human capital production, hence its focus on early education and the welfare of children (Jenson 2015). The guidelines of the social services agency in our study state that children need the presence of a primary carer – *"generally the mother or the father. Parents of small children should be allowed to weigh the demands of the labour market and childcare".* Despite the gender-neutral wording, welfare workers regard the mother as primary carer. Also for single parents, social workers favor parttime employment. Yet, institutional day-care is sometimes seen as more suited to the needs of the children, especially for migrant children. "Defamilializing" these children by placing them in professional childcare serves a similar assimilating function as labor market participation is supposed to do for adults: day-care helps them learn the language and to get acquainted with the values and norms of the dominant culture.

The cultural frames of work, care and responsibility used by welfare workers are gendered and ethnicized in complex ways. Analytically, the process of social investments entails three distinct steps: selecting investment objects (clients), selecting investments (work or educational measures) and applying these measures. These selections are based on the categorization of clients, which is guided by the frames described above. With an eye on the "profitability" of investments welfare workers sort clients according to their availability for the labor market, the marketability of their work capacity and their behavior in order to allocate measures (Nadai et al. 2013). These often conflicting categorizations are contextual and situated accomplishments (Lepper 2000), which result from the intersection of personal characteristics like gender, age, ethnicity, family status, education, health as well as of the institutional regulations of social security regimes, migration regimes and societal norms (Nadai in press). Most importantly the unemployment insurance treats clients structurally as individuals, while social assistance addresses them as members of a household. In the former, women are penalized for non-standard employment patterns, which diminishes their entitlements (Stutz/Knupfer 2012), but once they are eligible for unemployment benefits they are not disadvantaged regarding the allocation of investments in their employability. Statistical data show that women do not receive less measures in general, nor less of the successful measures (Nadai et al. 2013: 14 ff.). This corresponds to the cultural degendering of work in this institution.<sup>4</sup> In contrast, within the regime of so-

<sup>4</sup> However, since the unemployment insurance provides a wider range of measures, especially educational measures that are lacking in social assistance, women's restricted access to unemployment benefits diminishes overall "investments" in them (Nadai et al. 2013, see also Jaehrling 2009 for a similar institutional effect in Germany).

cial assistance women – especially migrants – are clearly underrepresented in activation programmes. Here it seems that social workers invest more in men as primary breadwinners of the household and regard women as carers and secondary earners, hence as less "profitable" investment objects.<sup>5</sup> Moreover, migrants' availability and marketability is hampered by the restrictive regulation of work permits and by the non-recognition of foreign educational credentials.

## **4 Work and care as promises and burdens – clients' perspective**

*"I like this programme very much. It makes women feel more productive, more like real women, you know, not just like mothers."* With these words a single mother on welfare commented on the work programme she had been assigned to. Her statement reflects both the feminist discourse on the emancipatory value of work and the social investment paradigms' ideal of the adult worker. "Real" women are productive outside the family; they are more than "just" mothers. Our research confirms once more the findings of other qualitative studies that the majority of unemployment insurance and welfare beneficiaries holds on to the dominant work ethic and wants to work (Bescherer/Röbenack/Schierhorn 2008; Ludwig-Mayerhofer/Behrend/Sondermann 2008; Schallberger/Wyer 2010; Solomon 2008). None of our interviewees said she or he preferred a life without employment at all, but the mothers did not want to work full-time. For all interviewees wage work serves the function of social integration – without work they feel excluded. Almost everyone mentioned the function of work as a source of income, hence economic independence. Most respondents feel ashamed and stigmatized for being dependent on any kind of welfare benefits, be it social assistance or social insurance benefits. Finally, a minority of our sample attaches importance to work as a vehicle for self-fulfilment.

Lisbeth Jausli's dream job, for example, always was nursing.6 But she never gets the opportunity for formal occupational training, marries young and is

<sup>5</sup> The underrepresentation of women may also be due to welfare workers' informal rationing of services (Brodkin 2011) because of the scarcity of activation programmes in social assistance. According to unpublished data of the Swiss Federal Statistical Office only 2% of clients participate in work integration programmes. The lack of programme places may increase the selectivity of investments in this institution compared to the unemployment insurance.

<sup>6</sup> The names of clients are pseudonyms.

a stay-at-home mother for roughly 15 years. After divorcing her first husband and losing custody of her children, she has a nervous breakdown and is on welfare for two years. When she is again capable to work, she is sent to the unemployment office. She mentions her professional dream, but the counsellor tells her, it is more important to quickly find a job now and eventually pursue her dream later. So Lisbeth takes a job as an unskilled warehouse clerk, which she loses after 10 years due to restructuring. She registers again at the unemployment office. At the same time she decides this is the moment to make her old dream come true and enrols for a Red Cross-class as a nursing aide. Because she has done so without consulting her counsellor the course fee is not paid for by the unemployment insurance. Rather, the counsellor tells her to feel proud for making it on her own. Moreover, Lisbeth is not allowed to look for nursing aide jobs only, but on the contrary is assigned to apply for unskilled factory jobs. However, the counsellor soon places her in a work programme, which in turn finds her an internship in a nursing home for dementia patients. During that time she also leaves her violent second husband, who tries to control her economically by blocking her bank account. Her story has a happy ending: her internship is turned into a regular 80%-job as a nursing aide, which pays enough for her to move into an own apartment and file for divorce.

For Lisbeth Jausli wage work is indeed a *gateway for emancipation*: she makes an autonomous choice and attains the capability of work in terms of becoming economically self-reliant and having a fulfilling job. Like for her, for roughly a third of the unemployed, we interviewed for our study, work means more than making a living or a nice change from domestic life; it is *essential for their identity.* They develop their own long-term plans for their occupational future and are ready to go to great lengths to realize their goals (for other cases see Nadai et al. 2013: 32 ff., Nadai/Canonica 2014: 359 ff.). However, the capability for work is not readily granted by the welfare state. Often the clients have to *persevere against welfare bureaucracies*, because institutional regulations prioritise quick labor market re-entry, prevent substantial investments in education for unskilled people and because welfare workers legalistically stick to bureaucratic rules. Clients are, thus, *refused resources and the capability of voice*. Lisbeth could not assert herself when she was unemployed for the first time, because she lacked the financial means and was only just recovering from a major personal crisis, thus probably unable to stand up to a powerful welfare bureaucracy. Her second husband accuses her of taking advantage of him, when she forgoes full unemployment benefits by registering as part-time unemployed to be able to attend her nursing aide class. Of course, she vehemently rejects this blame, but there is a kernel of truth to it: her chance to finally realize her dream partly rests on having a partner supporting her, since

#### *Eva Nadai*

her minimal unemployment benefits are most likely beneath the poverty line. Likewise, her later part-time job as a nursing aide provides a sufficient *income* for herself, but *could not provide for a family*. The capability for work and the chance to act autonomously, thus, are also dependent on the social conversion factor of specific family constellations. Unlike Lisbeth some of the other work-oriented clients we interviewed were (slightly) better equipped with cultural capital and occupational experiences as individual preconditions for autonomous choices. Yet, as single mothers and/or migrants they found they had to overcome even more obstacles such as having to provide and care for a family or fighting for work permits.

For the majority of our respondents wage work is not associated with self-fulfilment – it is, in the words of Jetmire Rexhep, *"just normal, that's part of life"*. To which she adds: *"What would I do at home?"* Jetmire is a divorced mother of two children who are now 10 and 12 years old. After the divorce she is briefly doing shift work as a temporary worker. When her son has to repeat class at school she wants to reduce her hours, but her employer declines. She quits, is now on welfare and participates in a work programme where she has been waitressing in the canteen for a year when we meet her. She immigrated as a teenager and finished her compulsory schooling in Switzerland, but never obtained formal occupational training. She also has some work experience as a sales clerk, but doesn't want to work in this field anymore. Instead, she'd prefer a canteen job because of the regular hours. In fact, work content doesn't really matter: *"What I want is just a job."* What matters is a job that is part-time (60% at most) and doesn't interfere with her children's schedule. In particular, she wants Wednesday afternoons off, when her children are not in school and Saturdays to accompany her son to his football games. In other words, Jetmire puts her family first and restricts her availability for the labor market accordingly. Like her all the mothers in our sample try to set limits in order to be able to uphold their own standards of good care – including the work-oriented women discussed above. Their choices, thus, follow the *rationality of a moral economy oriented to the perceived welfare of their families* (Dodson 2007) rather than the labor market logic of the ideal worker existing only for the job. And, in the Swiss welfare arrangement, they are still allowed to do so. Jetmire, for example, is not required to look for a full-time job, although her teenage children go to school and after-school day-care. While welfare beneficiaries don't have the unconditional capability for care – they cannot choose to be full-time carers when their children are beyond a given age limit – they are, to a certain degree, still granted maternalist exemptions, albeit at the price of prolonged poverty.<sup>7</sup>

<sup>7</sup> Of course, both the mothers' preferences for part-time work and the welfare workers allowances for them are also influenced by the lack of affordable childcare in Switzerland.

At some point Jetmire's social worker suggests formal training as a nursing aide. Jetmire declines, again because of the irregular hours in that line of work. However, there seem to be other reasons too, namely her lack of belief in her own abilities. For instance, in her current occupational field she prefers a canteen because "*in a normal restaurant with many different menus … I don't know if I could learn that. Here, we only have two menus per day, that's a lot easier."* Jetmire is not the only woman in our sample who is *hampered by self-doubts and low self-esteem*. We found a group of clients who have been discouraged all their lives by their families, teachers, partners, and as a result are not used to reflect on their needs and don't dare making plans (Nadai et al. 2013: 30 ff.). To put it differently, they weren't able to develop *fundamental capabilities*. Therefore, like Jetmire they sometimes act ambivalently with respect to possible investments in them. They need a lot of encouragement to be able to take up the rare opportunities they are offered, e. g. to go back to school and to get formal education. While especially the two women-only programmes of our study tried to promote autonomy competency and supported their clients' educational aspirations (Nadai et al. 2013: 20 ff.; Nadai in press), research on activation shows, that this is the exception rather than the rule (e. g. Schallberger/Wyer 2010). In our sample, roughly half of the women were actively encouraged to realize professional plans. This finding is partly due to the nature of the selected institutions, for one of the women-only programmes caters to young unskilled mothers with the explicit goal to assist them in finding an apprenticeship post. However, among the supported women were also clients of the other institutions. Unlike Jetmire, some of the women actually overcame their insecurities and started *"making small steps"*, as one of them said, to a more autonomous life by developing and pursuing their own plans.

Thus, the meaning of work influences how the unemployed integrate activation measures into their lives, i. e., to what extent they are actually capable and willing of making use of possible investments in them. This is not to say that individual "motivation" is the crucial factor, which is how welfare workers tend to see it. Rather, how the women frame and weigh work and care, and how much effort they are willing to invest in their occupational future is dependent on their educational, occupational and familial biographies and on the present social and individual conversion factors that delimit their options.

## **5 Conclusions**

Women's equal access to wage work is a preeminent concern of feminism, since an own income and participation in the public sphere are seen as cru-

#### *Eva Nadai*

cial conditions for emancipation. Feminism's valuation of paid work converges with the social investment paradigm's emphasis on universal labor force inclusion. Social investment policy promotes the adult worker-model and it promises investments in human capital acquisition to generate productive, self-reliant citizens. Those outside the labor market, namely social transfer recipients, are to be activated by means of investments in their employability, so they, too, can find their way back to work. Theoretically, the social investment paradigm has the potential to foster women's autonomy by acknowledging them as workers and by investing in their capacities. This chapter focuses the choices of unemployed women with little "human capital". Usually, the success of activation is assessed with respect to labor market entry and welfare rolls exit rates respectively. From a gender and capabilities perspective this view is too limited. Rather, the relevant question is whether unskilled unemployed women gain *autonomy:* in the narrow sense of financial independence qua employment and in the broader sense of leading their lives according to their own values. This entails the freedom to choose between wage work and care (or any combination of the two), to get a say in activation goals and measures as well as adequate support to develop autonomy competency (Friedman 2003) and the necessary skills to find rewarding work. These capabilities for work, care and voice (Bonvin 2009; Dean et al. 2005) depend on a *complex set of social and individual conditions* at the intersection of welfare state, labor market and family. Exploring social investment policy on the ground – the actual practices of different welfare state agencies – reveals inconsistencies and mixed effects regarding women's autonomy.

Since the respondents of our study were, by definition, unemployed we cannot draw any definite conclusions as to their chances of gaining *financial independence* through employment. Five women found a job or an apprenticeship position, another two worked as interns with relatively good prospects of getting an apprenticeship position later. For the other 15 clients in our sample there was no solution in sight yet.<sup>8</sup> All of the jobs were in female dominated occupations (nursing, day-care, cleaning, hairdressing, clerical work) as were the aspirations of the other, yet "unsuccessful" women. Typically wage levels in these jobs are rather low, even more so for unskilled workers.<sup>9</sup> Since most of the women want to work part-time, they will very likely not attain complete economic independence but will either depend on a partner or on welfare (or on both). Swiss and international research lends credibility to this argument. For instance, Aeppli (2010: 8) found that 18% of his sample of single moth-

<sup>8</sup> We have only sketchy data on the overall success rates of the programmes in our study. They vary between 13% and roughly 30% of clients who find a job (Nadai et al. 2013).

<sup>9</sup> Moreover, although the women also chose them for their supposed family-friendliness (i. e. part-time options), these jobs actually often involve irregular hours.

ers still needed social assistance after labor market entry, and 61% said, their standard of living was not better or even worse after leaving welfare (ibid.: 17; also Guggisberg et al. 2012: 20). Labor market re-entry after welfare tends to lead to precarious employment and continued dependence on the state or on a male partner (Corcoran et al. 2000; BFS 2009; Bothfeld/Betzelt 2011b).

Regarding the *capability of work* welfare bureaucracies indeed exert pressure to work and to accept any kind of job regardless of personal preferences and the quality of jobs. On the other hand, welfare workers play on the ideal of work as self-fulfilment to motivate clients for self-improvement. Thus, activation practice hovers between encouraging and supporting clients' professional aspirations and cooling them out (Maeder/Nadai 2009). Likewise welfare workers act ambivalently with regard to gender. Primarily, the unemployed are addressed as individuals in a (partly) degendered way, namely as adult workers who are obliged to earn their own living. And, for both genders, work holds the promise of self-fulfilment, albeit in a very limited way. But as soon as issues of *care* are at stake the individual is framed as part of a collective, namely the family. Especially social assistance as an institution addressing the household as a whole oscillates between treating the family as a financial unit, both parents therefore as earners, and as a normalizing institution and production site for human capital, which requires the presence of a primary carer. Most often it is still the mothers, who are seen as carers, so they are then regarded as a secondary investment object.

The women we interviewed basically shared the adult worker norm and wanted to work. However, the relevant institutions do not provide sufficient support to facilitate labor market entry for disadvantaged women. In particular, *institutional regulations prevent substantial educational investments* in those who have no formal training in the first place, because both the unemployment insurance and welfare adopt the less-eligibility principle. Social transfer recipients must not be "privileged" through publicly funded investments, while the working population has to pay for their further education. Therefore educational measures must not lift them above their former educational and occupational status (Nadai, in press). So, unskilled unemployed women are disadvantaged qua class (lack of education) rather than qua gender. In contrast to German research on the gendered allocation of activation measures (e. g. Bothfeld/Betzelt 2011a; Jaehrling 2009; Rudolph 2007) we did not find that women were generally disadvantaged in the individualising unemployment insurance. But they definitely were underrepresented in the programmes of social assistance. In this household-centred institution women are clearly treated as secondary earners within a modernized breadwinner model. Without formal occupational education, however, the women's chances of attaining economic independence are slim, especially in unskilled jobs in typically female occupations.

The unemployed women in our study did not have the full *capability for care*. However, the Swiss welfare arrangement is a far cry from a "ban on motherhood" (Dodson 2007). Rather, it is based on the one-and-a-half-earnermodel (Stutz/Knupfer 2012: 8) and *maternalist compromizes*: welfare workers tolerate or promote part-time work of mothers, even if this means prolonging welfare spells. For, in the social investment paradigm women's function as carers and educators is regarded as essential for the long-term production of human capital. Therefore, the needs of children come first (Jenson 2015). This is in line with some women's wants, as Jetmire Rexhep's case has illustrated. Even under the restrictions of activation, women for whom the family is just as or more important than work still have some autonomy with respect to balancing time for work and care. In fact, the pressure for full-time work some respondents felt stemmed from rigid time structures in the labor market and in apprenticeships and, in one case, from migration authorities.10 Of course, low levels of material support from the welfare state and low wages in unskilled jobs exert pressure too. However, the maternalist compromizes of welfare workers can also stifle the ambitions of more work-oriented women. We observed cases where women were held back rather than supported on account of being the mothers of small children. Moreover, the women's options regarding the capabilities for work and care were also dependent on *familial support*. Most of the interviewed lone mothers had very little or no support by the fathers of their children, hence they were limited in their flexibility in the labor market. On the other hand, some women could count on their relatives regarding care and material support.

The case of Lisbeth Jausli demonstrates the limited *capability for voice* granted to clients. She was not the only one denied the autonomy to decide which goals to pursue or which programmes to participate in. But Jetmire Rexhep's example points to another condition of the capability for voice, namely that asserting oneself requires the competency to develop personal commitments and pursuing them in the face of obstacles (Friedman 2003). Some of the women in our sample had been discouraged and deprived of "voice" early on in their lives, and it was their families and teachers that blocked their capabilities in this way (Nadai et al. 2013). For them activation programmes turned out a chance to cultivate fundamental capabilities, to find out what they might be able to do and to get some assistance to realize aspirations. Even if this kind of support is exceptional and will not immediately change the social and

<sup>10</sup> For migrants being on welfare can lead to losing their residence permit. The woman in question therefore had to work more hours than she wanted in order to leave welfare and keep her permit (see Nadai et al. 2013: 32).

economic situation of these women, it may be a small step to more personal autonomy. More fundamental improvements would require substantial educational investments, gender equality in the labor market in terms of working conditions and pay, and reconciliation policies acting on time, money and care services on the household and the collective level (Lewis/Giullari 2005: 97). Above all, social policy must allow a pluralism of emancipation conceptions instead of enforcing a uniform adult worker model for all.

## **References**


Clarissa/Niekant, Renate (eds.): Hartz IV. Zwischenbilanz und Perspektiven. Münster: Westfälisches Dampfboot, pp. 110‒134.

Rudolph, Clarissa/Niekant, Renate (eds.): Hartz IV. Zwischenbilanz und Perspektiven. Münster: Westfälisches Dampfboot.

Schallberger, Peter/Wyer, Bettina (2010): Praxis der Aktivierung: eine Untersuchung von Programmen zur vorübergehenden Beschäftigung. Konstanz: UVK.

Sedmak, Clemens (2011): Fähigkeiten und Fundamentalfähigkeiten. In: Sedmak, Clemens/Babic, Bernard/Bauer, Reinhold/Posch, Christian (eds.): Der Capability-Approach in sozialwissenschaftlichen Kontexten. Überlegungen zur Anschlussfähigkeit eines entwicklungspolitischen Konzepts. Wiesbaden: VS Verlag für Sozialwissenschaften, pp. 29‒52.

Sen, Amartya K. (2000): Development as freedom. New York: Knopf.


Weigt, Jill (2006): Compromises to carework: the social organization of mother's experiences in the low-wage labor market after welfare reform. In: Social Problems, 53, 3, pp. 332‒351.

Wright, Sharon (2011): Relinquishing rights? The impact of activation on citizenship for lone parents in the UK. In: Betzelt, Sigrid/Bothfeld, Silke (eds.): Activation and labour market reforms in Europe. Basingstoke: Palgrave MacMillan, pp. 59‒78.

# **Violence against women – an indicator of gender equality?!**

*Daniela Gloor & Hanna Meier*

## **1 Introduction**

Regarding developments in the field of 'violence against women' distinct changes in social perception and responses to the problem can be identified. In Switzerland – as in other countries – women's movement raised the topic and initiated the social debate in the late seventies. The subject at that time was addressed explicitly in the light of gender inequality and discussed as a reflection of discrimination and oppression. Intimate partner violence was situated within the culturally established uneven gender relations and gendered power hierarchies, and the aim was to change the situation on an individual as well as on the societal level. Innovative answers included the setting up of women's shelters: help for self-help was a central approach and specialized services for women who suffered violence and their children were established by NGOs. The focus of the activities was on victim's empowerment and the expansion of their agency – to enlarge their options for action and their opportunities for change.

In Switzerland the introduction of the federal Victim Assistance Act (Opferhilfegesetz) in 1993 marked for the first time a broader recognition of the problem, beyond the activities of women-centred NGOs and services. From the mid-1990s initiatives of women's shelters, women's support services/victim assistance agencies and Gender Equality Offices targeted further sensitization of statutory services, promoting the need for co-operation and interdisciplinary co-ordinated actions in responding to intimate partner violence. The first intervention projects and multi-agency approaches against domestic violence were set up following these initial developments. The efforts involved primarily the police and justice system, the health sector, guardianship authorities, and men's counselling services. Changes concerning legal frameworks on federal and cantonal level were initiated from 2000 in order to address the problem of violence in relationships (heterosexual and homosexual). These comprise of making violence in marriage and relationships a public offence to be prosecuted by the state (2004), protection measures against domestic violence under civil law, Art. 28b ZGB, (2007) and, since 2003, barring orders for domestic violence on cantonal level of police laws. St. Gallen and Appenzell Ausserrhoden were the first cantons that introduced in 2003 the possibility of eviction orders against the offender.

Today violence in marriage and relationships is no longer a socially acceptable behavior and the institutional system for help and intervention has changed accordingly. Much more institutions take responsibility for the problem and intervention measures and support have increased as well as the cooperation between the agencies. However, alongside the transition it can be observed that 'old ways die hard'. In spite of significant changes two key features of the problem 'violence in marriage and relationships' have remained.

Firstly, with the expansion of institutions that are involved we observed the tendency for the problem to become more and more detached from the context of gender relations. Today it is the term 'domestic violence' that prevails in social and institutional discourse, yet leaves the gender dimension of the problem unnamed. Thus, 'violence against women' is virtually relegated into the taboo zone.1 Although research results show that women are for the main part the most affected by this sort of violence (systematic abusive behavior and coercive control in relationships) the currently dominant denomination doesn't reflect this any longer; in politics as well as in various practice fields the gendered nature of violence in intimate relationships is contested time and again.<sup>2</sup>

Secondly, little has changed in regards to the fact that violence against women is a widespread problem by which many women are affected. Up to now neither research nor the practitioners in the field indicate any reduction of experience with this type of violence.

Against this background of change and persistence in the social perception of the problem the article deals with the question of whether, and to what extent, today's legal context and the provisions of the support system meet

<sup>1</sup> This sort of linguistic withdrawal of the gendered character of the problem is less observable in other countries, for instance the United Kingdom or Austria, where terminology more often includes gender: 'violence against women' or *'Gewalt gegen Frauen'*. Concerning language use and its implications for framing the problem of intimate partner violence see Klein (2013) and Gloor, Meier (2013).

<sup>2</sup> Men also experience violence in relationships. As research demonstrates, the situation of affected men and that of affected women require differentiated examination. For an overview see: Eidgenössisches Büro für die Gleichstellung von Frau und Mann (2012). For the scientific discussion on the experciences of violence by men and by women see.: Kimmel (2002); Gloor, Meier (2003); Schröttle (2010).

the needs of women who experience violence from their partner. The contribution draws on the findings of qualitative research conducted in Switzerland in 2011/12 and focuses upon the subjective view and experiences of women who got in contact with the intervention system because of their partner's violent behavior. The subject of this examination is how today's intervention system works according to those who use it. The article extends the social policy discourse by including the user's perspective, a perspective hitherto given little attention to. By way of introduction, we provide an overview of the state of policy development in the field, present the central intervention goals in the case of partner violence and reflect on the state of research on violence against women. Key findings concerning the women's views and experiences will be presented and discussed in the results section. Reflections on implications of our findings complete the contribution.

## **2 State of the art: policy and research in the field of violence against women**

#### *2.1 Policy area "gender based violence"*

International organizations of which Switzerland is a participatory member clearly express the state's obligation to eliminate violence against women. With the "Declaration on the Elimination of Violence against Women" in 1993 (resolution 48/104) the United Nations defined violence against women and intimate partner violence explicitly as a human rights issue. Thus states are obligated "to exercise due diligence in preventing, protecting and punishing such violations" (Hagemann-White 2008: 149). The resolution complements the aims of the Convention on the Elimination of All Forms of Discrimination against Women (CEDAW). The CEDAW-Committee, which monitors the implementation of the international human rights treaty, states that the issue of "violence against women" makes up part of the Convention on the Elimination of All Forms of Discrimination against Women. Therefore taking action against gender-based violence is to be understood as an important contribution to gender equality: "The definition of discrimination includes gender-based violence, that is, violence that is directed against a woman because she is a woman or that affects women disproportionately. It includes acts that inflict physical, mental or sexual harm or suffering, threats of such acts, coercion and other deprivations of liberty". (CEDAW-Committee, 1992) Swiss government ratified the CEDAW-Convention in 1997.

The Council of Europe has also engaged in the topic of "Violence against women" for many years. The Council's "Convention on preventing and combating violence against women" (Istanbul Convention) is based on the premise that such violence cannot be eradicated unless efforts are made to achieve greater equality between women and men (Council of Europe, 2011a and 2011b). The Istanbul Convention entered into force on August 1st 2014. Switzerland signed up to it on the 11th of September 2013, but it is not yet been ratified by the Swiss government.<sup>3</sup>

Governmental efforts concerning the implementation of gender equality politics in the field of domestic violence in Switzerland include a ministerial report (Bericht des Bundesrates, 2012)<sup>4</sup> which focuses on legal measures and institutional interventions. In the practical work of intervening in partner violence, numerous different institutions, agencies and professionals participate. Across these entities a consensus has been established as to the goals and overall direction of institutional intervention. At present three fundamental goals guide institutional action towards partner violence (domestic violence), with the intention to respond appropriately and improve the situation of victims:


In principle, all actors/agencies who are confronted with issues of partner violence are required to work towards these goals. The same goals underlie legislation on domestic violence. However, agencies and institutions act under different institutional and legal constraints and their engagement with partner violence has developed along different trajectories. Therefore, within the intervention framework individual agencies and offices take on different roles. Usually, one of the three fundamental goals is at the forefront of their practical work. Table 1 shows the current situation showing that institutional responsibilities cover the three goals rather unevenly.

<sup>3</sup> Missing in Switzerland is among other things a national helpline for victims as well as for relatives and friends. Very few European countries (Montenegro, Romania, Switzerland) do not provide this important service (http://www.coe.int/t/dghl/standardsetting/convention-violence/helplines\_en.asp).

<sup>4</sup> The report is based on the overview study "violence in relationships" (Egger, Schär Moser, 2008).


*Table 1 Primary institutional goals for intervention in partner violence*

#### *2.2 Research on violence against women*

In the US as well as in Europe a large research body concerning extent, conditions and forms of gender-based violence as well as its consequences and impact arose within the last 30 years. An overview of the development of this research field can be gathered from Dobash, Dobash (1992), Godenzi (1993) or Jasinski, Williams (1998). Recent developments and new focal areas are discussed in Hester (2004), Stark (2007) and Hagemann-White et al. (2008).

The premises of national and international changes in state policy combating violence against women and the role of civil society and the (different) concepts of gender equality in these processes constitute a further area of a more recent research interest (see Verloo 2007; Hagemann-White 2008; Htun, Weldon 2012; Krizan, Popa 2014).

Multiple national prevalence studies were conducted in Europe in the nineties and the first decade of 2000. The studies are informative about the extent and manifestations of the problem (synopsis in: Martinez et al. 2005). In Switzerland representative data was gathered in 1994 and 2003 (Gillioz, De Puy, Ducret 1997; Killias, Simonin, De Puy 2005). A joint prevalence study on violence against women was recently conducted at EU level including all 28 member states (European Union Agency for Fundamental Rights, FRA, 2014). While, due to methodological differences of the studies, the rates of women affected by partner violence do vary considerably, the extent of the problem always proves to be high: lifetime prevalence of intimate partner violence against women varies between 10% and 40% (Martinez et al. 2005).

Based on the practice and experiences of shelters and women's counselling so-called intervention projects or multi-agency cooperation approaches were established throughout Europe from the 1990s onward (Harwin, Hague, Malos 1999; Gloor et al. 2000; WiBIG 2004). Furthermore, domestic violence was increasingly perceived as a legal issue and constituted as a subject of criminal law (see Büchler 1998; Humphreys, Carter et al. 2006; Gloor, Meier 2009). Research turned its attention to professional practice and the activities of statutory and private agents in addressing violence in relationships by examining gaps and potentialities of institutional cooperation in achieving consistent intervention and optimized support. Studies and evaluations in this field reveal how the different agents define the problem, how and according to which guidelines and concepts they're taking action, and which measures of intervention may or should be adopted at an increasing rate (see: Seith 2003; Hester, Westmarland 2005; Hanmer et al. 2006). These studies dominantly focus on the following agencies: police, prosecution and courts as well as social services and child protection. Recently, the first comparative studies concerning protection orders and their impact have been conducted within the EU area (Aa et al., 2015). Also the health care system and its specific tasks in approaching the problem have become an object of research (e. g. Hagemann-White, Bohne 2003; Gloor, Meier, 2004).

We may sum up that today studies on institutional change in addressing the problem as well as evaluations of specific measures play an important role (Council of Europe 2006; Hagemann-White, Kelly, Römkens 2010). The perspective of those who are affected by partner violence yet remains to a large extent a blank space. In Switzerland there are no studies scrutinizing the viewpoint and concerns of the affected women with respect to the changes that have taken place in the support and counselling system. A fact which must be recognized as a significant gap. The question was hardly ever posed as to how women who suffer or have suffered violence from their partner experience the new measures; how they themselves perceive the implemented changes in procedures, support and intervention.

## **3 About the project**

As noted, in recent years very few studies concentrated on the women's perspective (Hague et al., 2003; Helfferich, Kavemann 2004; Abrahams 2010). It can be said, that the 'customer's point of view' has a weak presence in the actual discourse on support as well as in research. The position of affected women clearly contrasts with the situation of other stakeholder groups: In the health and social welfare sector so-called 'service user groups' today play an active public role within the institutional discourse; for example service user groups and involvement of people with disabilities, people with experiences in psychiatry or people living with HIV/AIDS. Comparable forms of inclusion or participatory structures are not existent in the field of violence against women.

Thus, considering the goals of intervention and support in responding to violence against women – that is to take victims seriously, understanding and empowering them in their specific situation – it is essential to investigate whether the intention of empowerment actually is met.<sup>5</sup> The decisive question is: Are today's measures and assistance, from the point of view of those who experience partner violence, adequate to provide the intended 'quality assistance' and appropriate to address their problems? And what are the obstacles and difficulties that these women still have to face? It is to these open questions the presented study seeks to contribute answers.

The survey this article is based on concentrated on the experiences of women and what they experience today when they have to contact institutions for support in the case of partner violence. Violence in relationships is the most frequent form of domestic violence, with a proportion of 80% to 100% women constituting the dominant victim group within institutional support.<sup>6</sup> Against this background the research focused on female survivors of partner violence.

In 2011 and 2012 the research team conducted narrative in-depth interviews with 40 women. The interviews took place in four German speaking Swiss cantons – Basel-Stadt, Luzern, St. Gallen and Zurich – and were arranged

<sup>5</sup> Coy and Kelly (2011) describe the concept of empowerment as facilitating and enlarging women's "space for action" which implies reinforcement of agency and expanding possibilities for making choices. The issue of 'women's agency' – in general and in the context of partner violence – gained increasing theoretical and empirical attention of contemporary gender research; see f. i. the research network 'Gender and Agency': genderandagency.univie.ac.at/schwerpunkte/ [date of access: 27. April 2015].

<sup>6</sup> The cases of domestic violence known to the police involve for more than 80% cases of violence in relationships (see Bundesamt für Statistik 2012: 13 f.). According to this Federal police data the aggrieved party is in 81,1% female and in 18,9% male (ibid: 21; own calculation).

with support and logistical help from police, women/victim's aid agencies and women's shelters. These agencies set up contacts between researchers and interviewees. Interviews lasted from one to three hours and were transcribed verbatim; they constitute the primary empirical evidence of the study. The qualitative analysis focused on women's views of their contacts with different agencies and institutions. The analyses, guided by the principles of grounded theory methodology, foreground women's subjective perspectives and viewpoints. The goal was to understand how victimized women experience their situation, which measures, services, and approaches they find supportive and helpful, and which they find stressful and difficult. The results presented below are based entirely on the 40 narratives and reflect the main issues of the qualitative reconstruction of women's views and experiences.

## **4 Key findings and conclusions**

#### *4.1 Women's view of today's intervention goals*

When women talk about their histories and experiences with partner violence they themselves raise the three fundamental intervention goals (see section 2.1) without being asked about them. The narratives show that the intuitional goals fully align with the wishes and needs of the women. First and foremost for women is the need to end the partner's violence and control. Violence needs to end and women no longer want to suffer from threats, assaults and humiliations.

#### *Ending violence*

From the perspective of affected women the goal of ending violence means, for instance, "to live in peace and quiet". It also means not being afraid anymore and not living with constant uncertainty and insecurity. The need to end violence clearly goes beyond the immediate crisis. It means not only the end to an immediate threat, humiliation or dangerous situation but a *sustainable* life without violence. Women's choice of words illustrates this: "being able to live my life", "be left in peace", to have "just a normal life". These are recurring phrases when women articulate their wish for violence to end.

For several women ending violence also meant that the violence should not be passed on to their children. They want to prevent their children from becoming perpetrators or victims. Ending violence for them also means disrupting the cycle of violence for the future of the children.

Eventually, they'll grow up, get married, have a family, and I just didn't want, that these things would leave their traces on them. (Int 17, SI-Abs 41)

That THEY [children] find a way, later in life, nonetheless, DESPITE EVERYTHING … to enter into a relationship and one WITHOUT beatings. Because it's two BOYS and that is ALWAYS at the back of my head. (Int 8, Abs 117; higher case letters: subject's emphazis)

#### *Protecting and supporting victims*

Women made clear the expectation that institutions should support them through their ordeal. This means first and foremost that agencies and experts should acknowledge and take seriously the circumstances and victimization experiences that women share with them. This is particularly important if women have no noticeable physical injuries "to show for it".

This expectation may seem self-evident but in the perspective of the women it is decisive. When agencies or professionals do not fully take in what women tell them, interaction with them turns into the opposite and becomes disempowering. If this fundamental aspect of interaction is short-changed, serious consequences ensue as narratives reveal: the response of the agencies and their concrete intervention measures will not be properly tailored to a woman's actual circumstances; they will not contribute to safety but may pose additional dangers. Not least, in such situations women are not referred to specialized services or referred too late.

For the women in this study, 'protection and support' means that they are accorded the proper space to share their experiences and are taken seriously. From their perspective this is the essential and necessary foundation on which further, appropriate measures can be built so that institutional interventions actually are supportive and effective.

#### *Holding perpetrators accountable, offering support for behavior change*

Holding the perpetrator accountable for his behavior is a topic that came up repeatedly in the study. It became clear that from the women's perspectives there is a wide range of ways in which a perpetrator might take responsibility for his behavior. For some women it means criminal prosecution and an explicit sanctioning by the justice system. For these women official punishment of the perpetrator through the legal system is the called-for societal response. There ought to be a substantive discussion of the violent behavior; the man's behavior should be named and highlighted as wrong. And something should be set into motion that will change his behavior. Criminal proceedings and sanctions should get the man to confront his actions and solve the problems he creates.

Other women do not mention criminal proceedings, but nonetheless they too articulate the need for the man to take responsibility. They speak about the need for help, that is, that he should seek and receive help. What matters from these perspectives is that the (ex-)partner gets to the point where he admits his behavior as problematic, takes it seriously and starts to change. These women quite often state that the man might need a little "encouragement" on this path and be forced along for his own good. Targeted help and support is in these views a meaningful way of holding men accountable.

#### *Conclusion*

The three basic institutional aims align with the central needs of victimized women. In terms of implementation this is a useful and encouraging finding: women experience institutional action as suited to their needs and expectations if intervention actors focus on ending violence, protecting victims' safety, offering appropriate and empowering support, and setting limits to perpetrators, holding them accountable and offering support for change.

#### *4.2 Women's experiences regarding the achievement of intervention goals*

Analyses further focused on the intervention framework and questioned to what extent the three fundamental goals of institutional intervention are, in women's perspective and experience, actually reached?<sup>7</sup>

#### *Ending violence*

The women's testimonies draw attention to fundamental problems with measures that are supposed to end violence. Some women reported that after an institutional intervention the partner's violence ended. Often, however, women say that whatever measure was taken had no effect: measures are insufficient for effectively ending violence.

A sobering realization emerging from the conversations is that the success of measures to a large extent depends on the violent man's willingness to cooperate. Interventions and orders of responsible agencies are effective so long as the man "plays along". If the man decides that the orders and directives against him are none of his business and that he will ignore them, the measures remain ineffective – neither does the violence end nor does the perpetrator suffer any further consequences for his defiance.

<sup>7</sup> Findings about women's interactions with the individual agencies listed in Table 1 are detailed in the full report in the respective chapters on institutions (Gloor, Meier 2014).

The experiences of victimized women demonstrate vividly that a man's refusal to cooperate rarely has consequences. Certainly, barring orders and injunctions are imposed; however, when the man does not "voluntarily" adhere to them but continues to contact, harass, endanger or control his (ex-) partner, such actions usually have no further consequences for him. There is no additional institutional action that would reinforce the measures and end the violence.

In such situations the woman's problems intensify: she is exposed to more violence and it is *her* responsibility to do something about it. She needs to report the breach of barring orders or injunctions to the authorities and get in touch with relevant agencies and institutions – when the man *continues* to be violent, the woman needs to *continuously* report. At this point, any positive contact between woman and police turns into the opposite; her repeated reports do not endear her to police. In addition, she experiences their intervention as ineffective because little further action follows aside from scolding the perpetrator- "you must not do this!" From the perspective of the woman such warning is more of a validation for the violent man than a sanction: The man actually gets the message that he can do whatever he wants with impunity.

#### *Protecting and supporting victims*

The analysis of the interviews shows that the changes of the past years with regard to the goal of victim protection and support – appropriate interaction and advice – have been effective. Many women report good interactions with various agencies and institutions, which they found helpful and supportive. For many of these contacts the interviewees described interactions in which they felt acknowledged and taken seriously.

Police certainly rank among the institutions which have engaged intensely with domestic violence over the past years and where much has changed. The analysis shows that this is mirrored in women's experiences. Nowadays victimized women frequently report very good and helpful initial contacts with police, in which their situation is taken seriously and help, further information and support are promised and offered.

Furthermore, of great significant is tailored, professional support from specialized victim support agencies and women's shelters. For a large number of women it is these agencies that offer decisive help and support. Women experience as a particular turning point having access to information that helps them sort through their options, and access to the wider support system, including potential or actual legal intervention. It is significant for women to have practical support and help with decision-making as they contact further agencies, institutions and professionals. Last but not least, for many women it is important to have the space to talk about their experiences and circumstances, to reflect on these in the presence of an experienced and empathetic listener, and to sound out their own wishes and capabilities. Victimized women experience all of these elements as profoundly validating; they contribute to a woman's ability to (again or for the first time) recognize and expand her space for action.

The goal to protect and support victims is also relevant in interactions with other institutions and agencies. Here, however, it becomes clear that interventions are somewhat haphazard: they depend on individual staff members' knowledge and awareness and are not integrated into institutional policy. Surprisingly common are gaps in terms of referrals to specialized support (victim's aid, women's shelter) and connecting victims to the full range of available services. Instead, women frequently hear rash advice such as to press charges, which can embroil them in a drawn-out, dire process.

#### *Holding perpetrators accountable, offering support for behavior change*

The findings show that the third intervention goal at present hardly comes to fruition. The conversations with the women rarely suggested that intervention in domestic violence includes holding the perpetrator accountable.

Aside from barring orders and injunctions there is hardly any institutional engagement with the perpetrator in which he is held to account. Particularly noteworthy is the finding that neither victim nor perpetrator are made aware of the fact that violence in marriage and intimate relationships is now an offence prosecuted by the state. Nor do women's experiences with prosecution and legal proceedings suggest that this new legal status affects the way in which cases of domestic violence are processed.

Similarly, specialized counselling for violent men is largely underutilized. The opportunity and benefit of referring violent men to agencies or programs that offer education against violence and specialized individual counselling are rarely considered. Instead, the perpetrators remain largely outside the view of institutions; they neither have to fear sanctions nor are offered problem-oriented support and help with change.

#### *Conclusion*

Institutional intervention in partner violence has borne mixed results. Whereas numerous efforts and successes can be noted with regard to protection and support for victims, with regard to the other two goals systematic implementation has not yet taken place. All too often violence against the female partner continues and is not ended, and efforts to hold the perpetrator accountable, whether through criminal legal measures or counselling, are the exception rather than the rule.

## *4.3 Implications of today's intervention practice*

#### *Facing a mountain of work*

The narratives of victimized women draw attention to an issue which up to now has received too little consideration in the intervention framework. For women who experience partner violence and subsequently engage with the institutional intervention and support system this step means opportunities for help and change. Yet at the same time, for most women this also means, in addition to coping with the violence, a mountain of work involved in engaging with these institutions. It is telling to merely list with how many institutions women stay in contact with:


The interviews show that one-time contacts with different agencies are an exception. Most of the time there are multiple contacts, sometimes within a short period of time, in other cases over months or years. The partner's violence generates considerable "institution work" for victims which they suddenly have to manage as well, these include:


In part these are interactions women choose themselves; in part women find that their changed circumstances force them to get in touch with certain institutions, and institutions also get in touch with them. As a consequence, victimized women need to engage with a multitude of institutions and professionals. This contrasts markedly with the finding mentioned above that institutional engagement with the violent man, be it through sanctions or counselling, is largely missing. While victims (have to) engage with multiple agencies, the perpetrator is mostly left on its own. Institutions appear to prefer to be in touch with the woman (for instance, in her role as mother), while the man disappears from view. Hence, in the conversations the women express the additional stress caused by institution work. This is expressed in metaphors and figures of speech such as having to "bend over backwards", "endless maelstrom", and "I can't do this anymore". Only for contact made with women/ victim's aid agencies do women occasionally mention time and energy savings.

Further "institution work" and stress is created when the man 'badmouths' the woman to an institution or uses an agency to get in touch with her again; we call this "instrumentalising institutions". Examples include the man reporting the woman to police or prosecutor, filing charges against her in civil court or with child protection agencies (e. g., about divorce, visitation rights/ child contact, child support payments or requests for money; accusations that the woman abuse the child, etc.), accusations with the migration office, for instance, because of fake marriage. In such cases the women have to appear before institutions and agencies to defend themselves and may even need to engage a lawyer.

#### *The intervention phase viewed by the women*

According to the interviewees intervention reality is much more complex than conventionally presumed by most of the intervening agencies. Women do not experience a linear pathway from intervention to normalcy. Initial interventions lead to subsequent agency contacts and for the woman a mountain of institution work piles up. If the violence, as happens frequently, does not end despite institutional intervention, for her "institution work" becomes even more intense and time-consuming. The experiences of the interviewed women reveal that today's intervention system entails a remarkably increasing workload for the victimized women.<sup>8</sup>

The circumstances of our sample illuminate that a differentiated understanding of 'the intervention phase' is required. Police, victim support and women's shelters gave us addresses of interviewees only when a woman's case was considered closed. The intention was to interview women who lived in a new normalcy, i. e. one free of violence. However, the interviews made increasingly clear that this was the case only for a few of the 40 interviewees.

New normalcy is exemplified in some of responses when women were asked what they wished for their future. One would assume that violence *and*

<sup>8</sup> The findings also indicate – indirectly – how much work domestic violence creates for institutions and, in turn, raises the costs for society at large; see Eidgenössisches Büro für die Gleichstellung von Frau und Mann (2013).

follow-up problems were a thing of the past when women expressed these wishes:

Vacation, vacations! (Int 34)

That everything stays as it is. (Int 30)

That my little house will eventually be ready. (Int 26)

However, such comments were rare. Very few women made carefree and forward-looking comments that implied a newly gained ordinary everyday life – an arrival in normalcy.

In contrast, many answers indicated that the women were desperately hoping to arrive at a new normalcy and reconquer a normal life. What does "normal" mean? They wish for a life without violence, without uncertainty and without fear, and they want to be able to trust the feeling of having finally overcome the consequences of victimization. For the women this would also mean that contacts with intervention agencies (finally) have come to an end. Yet, many responses show that women are still in the midst of the intervention phase coping with institutions and follow-up problems at the time of the interview. In the responses to the question about the future the longing for normalcy and an "ordinary" life expressed like this:

I don't have many wishes, wishes I do not have many – my wish still is: I just want to live in peace and contentment [laughs]. I don't want a million and nothing. […] So many … so many people find it crazy that we [woman and child] have to move again after only five months … It had to come to this … I see this clearly, and … Yes, I look forward to, I am so looking forward to my … normal life [laughs] – only an ordinary life. (Int 5, SI-Abs 130)

I just want to live a normal life. […] Just a normal life. Go to work, and then come back and make dinner... [fighting tears] […] These are my hopes for the future. Just... a normal life. (Int 18: SI-Abs 241)

#### *Conclusion*

Based on the experiences of the interviewed women we suggest a revized model for the cycle of violence and intervention which extends today's predominant notion of "violence", "intervention", and "return to normalcy". The intervention phase needs to be understood in a more comprehensive and multifaceted way. Therefore, the intervention phase is split into two phases (see Table 2): a phase of acute crisis intervention (phase III) and in a phase of managing institution work and violence-related follow-up problems (phase IV).


*Table 2 Partner violence and intervention: an expanded model*

In many cases women experience multiple cycles of phase III and IV. A direct step from crisis intervention (phase III) into normalcy (phase V) is seldom. Considering this fact and giving increased attention to a differentiated understanding of intervention by taking more into account the implications and significance of phase IV the issue of taking pressure off the victimized women becomes a relevant, new element of intervention goals.

Coping with partner violence requires considerable work and effort from victimized women. Intervention typically begins in response to an acute crisis (phase III) after which a long and stressful phase ensues in which women are constantly confronted with a multitude of institutional contacts and follow-up problems (phase IV). Phase IV lasts much longer than is usually assumed, an issue which has not received proper attention in public/institutional perspectives. The primary focus is mostly on crisis intervention measures (phase III) with the assumption that these will lead to normalcy (phase V). Yet, findings show that reality is more complex, crises may last longer than expected and intervention extends beyond crisis. This result needs to be integrated into institutional and policy understandings of how to intervene in partner violence.

#### *4.4 Interventions do not stop violence*

#### *Ending institutional contacts – no guarantee to the end of violence*

As noted, the effectiveness of measures to end violence largely depends on the perpetrator's willingness to go along with them. This observation acquires additional significance with the finding that even at the time of the interviews several women were still threatened. These women had worked with multiple agencies, which in turn had intervened and had then considered the cases "closed". Nonetheless, the men continued their violent behavior. Therefore, the women were still living under dangerous and constraining circumstances.

One in four interviewees (11 out of 40 women) described how in her daily life she still experiences various forms of abuse – controlling behavior, terrorizing text messages, threats, verbal abuse, targeted harassment, bad-mouthing, stalking – or needs to be prepared for physical violence. Child contact also often provides a context in which the man seizes the opportunity to further threaten and abuse the woman. After the interviews were completed we learned of one life-threatening attack against one of the 40 women.

Furthermore, 8 of the 40 women were suffering from constant anxiety and the fear that "it will start up again", even though living in relative peace at the time of the interviews. They could not rule out that the man would become "active" again; for instance, if he returns from abroad where he went under cover; when a court or divorce hearing is scheduled; if she enters into a new intimate relationship; or simply if she accidentally crosses paths with him.

In sum, at the time of the interviews almost half of the women still felt endangered, either because violent actions were continuing or because further violence had to be expected.

#### *Additional victims*

Results show that the interviewees were not the only partners who experienced violence from the respective man. In about a third of the cases (12 out of 40 interviews) additional female partners had been identified who also had experienced violence from the same man. This includes relationships the man had before and after his marriage or relationship to the woman in the interview. In two cases the same man was violent against four different women (including the interviewee). The analysis furthermore shows that multiple children from the different relationships are affected.

A new female partner can be an "efficient solution" for a victimized woman. The interviewees experienced that the existence of a new female partner significantly and effectively improved their own situation. From the outside, this "solution" appears decidedly undesirable. It is alarming to realize that violence does not end even though one particular woman is no longer victimized – instead a different woman suffers violence from the same man.

#### *Conclusion*

The findings show that the end of institutional contact is not a reliable indicator that the man's violence against the woman has ended. The current system of regulations, measures, and implementation effectively ends neither threats against the woman nor violence against the woman.

The finding of sequential violence against different female partners emphasizes the significance of and need for effective intervention in terms of "ending violence" and "holding the perpetrator accountable and offering help". The existence of repeat violence across multiple partners indicates that violent men do not receive helpful institutional support. As a consequence they carry their attitudes and behavior from one relationship to the next. This finding points to the urgent need of an institutional engagement with perpetrators of domestic violence.

## **5 Discussion and closing remarks**

The findings of the study confirm the usefulness of the goals that state, society and policy-makers have formulated for intervention in partner violence. The intervention network can indeed help and support victimized women, and appropriate forms of help and support are empowering. Institutional intervention actually aimed at victim protection and support which offers tailored advice expand victimized women's space for action, and support pathways out of the violence. Institutional intervention in domestic violence proves to be an effective instrument to promote gender equality.

However, the study also points to significant gaps and problems in the intervention system. On one hand, the findings call for further refinements in victim protection and support. From the victimized women's perspective the common notion of agencies' interventions which presumes a linear pathway from intervention to normalcy does not correspond to their reality; on the contrary, initial interventions lead to countless, persistent agency communication women have to cope with. Today the partner's violence generates for the victimized woman a mountain of "institution work". A more differentiated understanding of 'the intervention phase' is required which extends beyond intervention in acute crisis situations and integrates the follow-up phase in which women have to manage a pile of work in communication with a vast number of agencies. On the other hand, findings reveal considerable challenges the current intervention system faces in ending violence and holding perpetrators accountable. Results indicate that the perpetrator remains comparatively undisturbed and unchallenged by the agencies. The lack of institutional actions concerning the perpetrator and directly involving him may be considered a prime reason that violence in marriage and relationships persists a widespread problem of which many women are affected.

The experiences of victimized women indicate that in the areas 'ending violence' and 'holding perpetrators accountable' further policy efforts and institutional developments are needed, and innovative research should support this process. Concerning the area 'protecting and supporting victims' taking pressure off the victimized women should constitute a new element of this intervention goal.

In our recommendations, government agencies as well as decision-makers at various agencies and institutions (policy level), professionals in practice fields, the cantonal coordination and intervention offices against domestic violence, and the federal office on domestic violence are called upon for further development of intervention in partner violence and for a consistent implementation of existing and new measures. It is important to overcome the narrow understanding of "domestic violence equals physical violence". Instead, partner violence (domestic violence) needs to be recognized as a multifaceted behavior pattern that includes a variety of violent actions. Institutions need an accurate conceptualization of the issue in order to respond appropriately to the experiences and circumstances of victims.

From the perspective of research on violence in marriage and relationships it would be welcomed if the Istanbul Convention was ratified and its principles implemented. The Istanbul Convention offers guidance for politics and administration on how to further develop intervention in violence against women – and thus to strengthen gender equality.

## **References**


work 6. Available at: www.cahrv.uni-osnabrueck.de/reddot/190.htm [Date of access: 13 January 2014].


# **Gender equality and evidence-based policy making: Experiences from social transfer and tax policy reforms1**

*Andreas Balthasar & Franziska Müller*

### **1 Introduction: Problem description**

One of the aims of gender equality policy in both Switzerland and Europe is to promote the compatibility of working and family life. Women and men are to have the same opportunities at home and at work (Inauen/Maisenbacher 2011). This requires, among other things, a political framework supporting women who want to work and men who want to look after the children. Two policy areas which can discriminate in this regard are tax and social transfer policy. These two policies specifically affect the extent to which it is possible and worthwhile for one or both parents to work fewer or more hours. An analysis of the OECD from 2004 shows that substantial gender inequality in OECD countries is due to misdirected incentives in tax and social transfer legislation (OECD 2004).

This problem is particularly acute in Switzerland as these policies are defined at cantonal level. In the last decade, several studies have exposed the diverse aspects of the cantonal tax and social transfer systems that slow down women's employment in Switzerland. Empirical findings have shown that, depending on the place of residence, level of income and number of children it may not make sense financially for both partners to work (e. g. Knupfer/ Bieri 2007; Bertschy et al. 2012; Bonoli et al. 2010; Bütler 2007; 2009; Ehrler et al. 2012; Ott et al. 2010; Schwegler et al. 2011; 2012; Bühler 2004; Stutz/

<sup>1</sup> This paper is based on the Swiss National Science Foundation project "Gender mainstreaming in the welfare state – the case of tax and social transfer policies in Swiss cantons. Obstacles to evidence-based policy-making in a gender-relevant policy field" (Reference number 406040-129073/1). The authors would like to thank those who managed the NFP 60 plus Joachim Blatter and Ariane De Rocchi for their support.

Knupfer 2012). This is in spite of the fact that favoring a traditional family model with a single male breadwinner and a single female in charge of the home over the dual breadwinner or dual earner model, is not only bad for gender equality but also is also economically detrimental. There is no doubt that the Swiss economy will require a male and female workforce in the future due to demographic change. Cantonal tax and social transfer policies which accept or even support current gender inequality, for example by favoring the single earner model, are thus also economically harmful.

It is thus understandable that policies promoting gender equality draw attention to policy areas that do not explicitly advance gender equality but can have discriminatory effects on women or men, such as tax and social transfer policy. This approach is called gender mainstreaming: "Gender mainstreaming is the (re)organization, improvement, development and evaluation of policy processes, so that a gender equality perspective is incorporated in all policies at all levels and at all stages, by the actors normally involved in policymaking" (Council of Europe 1998: 15). Bothfeld argues that the success of gender mainstreaming measures is strongly related to political learning and therefore depends on two conditions: the learning aptitude of the actors involved and their access to gender-relevant knowledge (Bothfeld 2005).

The question of whether the involved actors have access to genderrelevant knowledge is the starting point of this chapter. It focuses primarily on how often, by whom and why knowledge about possible effects on the compatibility of work and family life was considered within Swiss cantonal legislative processes for tax and social transfer policy. The aim of this chapter is to draw conclusions on how to raise the profile of gender equality within tax and social transfer policy in the future by drawing on past experience and by taking gender knowledge into account.

The chapter is structured as follows: the first section addresses the conceptual and empirical basis of the study. It includes a definition of how we understand the terms "gender equality", "gender-sensitive" and "evidencebased". In the second section, we present details of the key outcomes of the research project. The focus is mainly on how often evidence-based and gendersensitive knowledge was applied to legislative processes, on the institutional players who initiated the application of this knowledge and possible explanations as to why the knowledge was considered as part of these legislative processes. In section three we draw conclusions based on the empirical findings. We discuss ways to improve the availability of evidence-based and gendersensitive knowledge for tax and social transfer political processes, so that gender equality can be given more prominence in these areas in the future. Finally, we discuss some theoretical and methodological questions in order to highlight perspectives for further research.

## **2 Conceptual and empirical basis of the study**

#### *2.1 The concept of gender equality*

In Switzerland, family issues have long been perceived as a private matter; therefore, they took a long time to join the political agenda. In consequence, Switzerland is a "laggard" in comparison to other countries as regards the modernization of family policies (Mosimann/Giger 2008: 228, Häusermann/ Kübler 2010: 164). In such a context, the provision of family-external childcare facilities and the financial stimulation of their use represents not only a "liberal" but also a "transformative" approach to gender equity (Fraser 1994; see also Blatter et al. 2015). It not only allows freedom of choice for women and men to balance their work and family life according to their individual preferences, but it also contributes systematically to the transformation of family practices towards the so-called "dual breadwinner or dual caretaker model". This family model allows both men and women to participate in the labor market and share family duties, especially caring for children and the elderly (Pfau-Effinger 1998).

A social transfer and tax system which is neutral vis-à-vis any form of sharing family duties is clearly a most important precondition for allowing women to participate in the job market and men to stay at home for carework, which in turn has positive effects on all the normative criteria that Nancy Fraser (1994) postulates as necessary to achieve gender equity: for example, it reduces the risk of poverty for women and the risk of exploitation within the household; it contributes not only to income equality, but also to leisure-time equality and equality of respect.

In such a context, focusing on social transfer and tax systems which act against gender equality seems important to us. If we ask if social transfer and tax policy reforms take gender-sensitive information into account in the process of policymaking, we understand this as information about gender-relevant effects in the broadest sense. In practice, this normally includes information on disincentives to starting or expanding gainful employment (especially for women with children). However, it often also includes information on incentives for both partners to work (e. g. through financial relief for families through extra-familial childcare).

#### *2.2 The concept of evidence-based policy-making*

The concept of evidence-based policymaking has gained importance at an international level over the past few decades. The British Labour Party was the first to advocate the significance of a sound scientific basis on which to shape policy: "What matters is what works!" became the slogan for the desire to move from an ideology to an evidence-based approach to policymaking (Nutley et al. 2007: 10). During the last decade, this approach has reached most parts of the European Union (Solesbury 2001). Research on evidence-based policymaking explicitly deals with the relationship between knowledge, political learning and policymaking. Several forms of evidence can be relevant to policy makers: expert knowledge; published research; existing statistics; stakeholder consultations; previous policy evaluations; outcomes of consultations; costs of policy options; output from economic and statistical modelling. The availability and validity of data are indispensable to evidence-based policymaking.

Evidence-based policymaking has a tradition stretching back to well before the 1990s (Balthasar/Müller 2014). In the 1950s, the American political scientist Harold Lasswell (1956) wanted to initiate a multi-disciplinary and normatively-oriented policy science providing relevant knowledge for and about political decision-making processes (Jann/Wegrich 2003). However, in 1986 David Collingridge and Colin Reeve relativized this ideal in their book *Science Speaks to Power: The Role of Experts in Policy Making*. They showed that science couldn't meet the demand from politics for reliable data for shaping policy, as the aim of science is not first and foremost to provide reliable information but organized scepticism, doubt and falsification. Jürgen Habermas (1979) followed a similar path to Collingridge and Reeve by examining the relationship between science and politics. He postulated a pragmatic model of cooperation, overriding the hierarchical relationship between science and politics and having decisions made through an interactive process between decision-makers and experts (Linder 1991).

The problematisation of the relationship between science and politics by authors such as Habermas, Collingridge and Reeve shows that the concept of evidence-based policymaking has its opponents. The criticisms are mainly based on three arguments: *First*, the concept of evidence-based policymaking is obviously based on instrumental rationality. Critics argue that evidence-based policymaking should be seen primarily as a deviation from norm- and interest-based justifications of political decisions and that it is a myth that tries to mask increasing uncertainties and ambiguities (Rüb/ Strassheim 2012). *Second*, it is argued that the concept of evidence-based policymaking is based on the optimistic assumption that the quality of political decisions can be improved through knowledge. Frey and Ledermann contend: "On the one hand it presumes that the quality of policies, their effectiveness and efficiency can be measured in an objective and conclusive way. On the other hand it assumes that policymaking can be rationalized. This optimism has been the starting point of many critics of EBP [evidence-based policymaking] and gave reason to the allegation of naivety" (Frey/Ledermann 2010: 3). *Third*, it is alleged that the concept of evidence-based policymaking doesn't take sufficient account of the fact that knowledge is often used as a strategic resource to retain and expand power (Parsons 2002). For politicians and policymakers, power is a necessary condition for the achievement of their policy objectives. Post-structuralist and constructivist approaches have contextualised the notion of "evidence-based policy making" (Gibbons et al. 1994, Van der Knaap 1995). Rather than regarding science-based information as indisputable knowledge, it is viewed as a collection of arguments, which can be debated, accepted and disputed. Several studies show that the interdependence of power and knowledge can be conceived as a continuous discourse (Radaelli 1995). Therefore, some authors prefer to speak of "evidence influenced" or even just "evidence aware" instead of "evidence-based" policy, in order to reflect a more realistic view of what can be achieved (Nutley et al. 2003: 2).

Proponents of evidence-based policymaking don't deny that political decisions are made on the basis of political debate and power structures. The role of evidence, however, is primarily seen as limiting the scope for purely ideological argumentation (Howlett 2009; Nutley et al. 2007; Solesbury 2001). A properly functioning democracy needs arguments that are not only opinion-based but also evidence-based (Nutley et al. 2003). Moreover, the question as to whether and how knowledge is used within political decision-making processes is also relevant when knowledge is used as a strategic resource to extend power or implement a political agenda.

There is no uniform understanding of evidence-based policymaking in literature (Frey/Ledermann 2010). While some authors qualify knowledge as evidence-based only when it originates from randomized control trials (Davies 2004), others also recognize alternative methodological approaches, such as quantitative and qualitative case studies, public statistics or expert knowledge. We understand evidence as any kind of systematically generated information used to bolster diagnostic, evaluative and prognostic claimmaking in public argumentation. In contrast to strong relativists, we would argue that the more the information is based on systematic and transparent methods of data collection and data analysis, the more powerful the evidence in the policy process should be. Furthermore, for analytical and practical reasons, we focus on the use of evidence produced by academics or experts within and outside public administration. The result is a definition that includes empirical knowledge based on scientifically collated data, as well as expert knowledge based either on theoretical considerations or practical experience (see Nutley et al. 2007).

#### *2.3 Empirical basis of the study*

This chapter is based on a comprehensive survey of all cantonal legislative projects involving social transfer and taxation policy presented to the respective parliaments between January 2008 and September 2011 for further discussion. All legislative amendments affecting tax legislation for direct taxation of natural persons and all changes in economic social aid, family allowances, alimony advances, premium reduction for obligatory health insurance, subsidies for extra-familial childcare and grants. During the period under review, there were 76 cantonal gender-relevant legislative amendments, of which 60 were subjected to more in-depth study as part of the quantitative analysis.<sup>2</sup> This involved gathering documents on the legislative processes and structured interviews with the responsible officials from the cantonal administration.

Of the 76 legislative amendments in question, 41 were in taxation and 35 in social transfer policy. Many other gender-sensitive social transfer amendments made during the period under review were not included. This happened for a number of reasons, for example because they were not regulated in a law but in an ordinance, or because they were submitted to parliament for discussion prior to the start of the period under review. On the whole it can be summarized that there have been many opportunities to sufficiently address the issue of equality in tax and social transfer policy during the period under observation. On average, there has been one bill per year in the areas of taxation and social transfer per canton, which could have been used to act against any discriminatory factors.

The analyzed tax law revisions have mainly served to increase the deductions for children and the costs of childcare. There was debate in some cantons as to whether the deduction should only apply to extra-familial childcare or be extended to a couple's own childcare. Tax legislation amendments designed to reduce the general burden on the families with low or medium income were also considered gender-sensitive, as the single earner model is normally particularly lucrative in these income areas. The abolishment of the so-called "marriage burden", i. e. introducing full separation of income to remove a disincentive to work, was also considered gender-sensitive.

Regarding the changes in social transfer, there are many changes in children's and family allowances. They mostly took the form of amendments to cantonal laws based on the federal law governing family allowances, which came into force on 1 January 2009. There were also many legal revisions for extra-familial childcare, which were mainly driven by the strong growth in de-

<sup>2</sup> Due to resource constraints it was only possible to review one change in tax law per canton, even when in some instances a number of thematically relevant legal amendments were made.

mand for childcare outside school hours. These legislative processes were, for the most part, subjected to intense debate with the issue of family and career taking centre stage.

## **3 How often, by whom and why is evidence based and gender sensitive knowledge applied?**

Let us now turn to the results of our research. The main emphasis lies on the frequency of access to evidence-based and gender-sensitive knowledge in the legislative processes in question, the institutional players which initiated the incorporation of evidence-based and gender-sensitive knowledge and possible explanations as to why evidence-based and gender-sensitive knowledge is incorporated in cantonal legislative processes governing tax and social transfer policy.

### *3.1 Application frequency of evidence-based and gender-sensitive knowledge*

The collection and interpretation of the results are classified according to four levels and several sub-groups of evidence-based knowledge. The involvement of knowledge is applied systematically on all levels; however, the degree of objectivity lessens level by level. The first two levels comprize empirical knowledge either gathered externally by scientific means (Level 1) or gathered and evaluated internally (Level 2). Levels 3 and 4 systematically include expert knowledge, either from the consultation of experts and professionals from the administration (Level 3) or the consultation of interest groups (Level 4).

Table 1 shows the types and frequencies of evidence included in the 60 Swiss cantonal legislative processes which were studied in detail.


*Table 1: Level and frequency of included evidence*

*Source: authors' own work.*

The results of the survey indicate the regular inclusion of evidence-based knowledge in legislative processes. This applies under the condition that the meaning of evidence is broadly defined and, in addition to scientifically-based knowledge, the practical knowledge of experts and representatives of interest groups is also considered evidence-based:

In all 60 legislative processes examined in detail, internal cantonal evaluations and analyses took place. In 44 of the 60 cases, there was an exchange of experiences between different cantons. In 7 of 60 cases, surveys or analyses were commissioned externally to universities or private research companies. In conclusion, it can be ascertained that the cantons often base their policymaking on evidence, however, only in exceptional cases directly on scientific reports commissioned to universities or private research companies.

The number of cases in which evidence-based and gender-sensitive knowledge was included in the legislative process, is significantly smaller:

In 39 of the 60 legislative processes examined in detail, there was a deliberate focus on gender-equality aspects. In 17 of the 60 cases, there was an exchange of experiences between different cantons on gender-related aspects. In 11 of the 60 cases, gender-sensitive knowledge was used from co-reporting and consultation procedures. In 6 of the 60 cases, surveys or analyses were commissioned externally to focus on gender-equality aspects.

As regards the cases in which evidence-based and gender-sensitive knowledge was included in the legislative process, it can be seen that there are no relevant differences between tax and social transfer policy concerning the access to the first two forms of evidence (empirical knowledge produced by academics and empirical knowledge produced by the administration). On the other hand, for legislation involving tax other cantonal administrative offices were contacted in 56 percent of cases, compared to 35 percent for social transfer policy. This contrasts with incorporating expert knowledge from interest groups, which occurred in 39 percent of legislative processes in social transfer policy and only 13 percent for tax policy.

It is worthy to note that equality experts (equal opportunities officers) were explicitly included in only twelve of the 60 cases reviewed and gender-related information was only discussed in eight cases. In specific instances, the equality experts joined the in-house committees or working groups of the administration in which the review was initially discussed. This leads us to the overall conclusion that expert knowledge on equality was only rarely included in legislative processes for tax and social transfer policy.

#### *3.2 Influence of institutional actors on the inclusion of evidence-based and gender-sensitive knowledge*

#### *3.2.1 The impulse to include evidence-based and gender-sensitive knowledge*

The questionnaires conducted as part of the project allow us to acquire insights into those institutional actors which included evidence-based, gender-sensitive knowledge. We divided the actors into the following categories: administration, executive and parliament. Table 2 shows what type of information was introduced and by which actor. Only information pertaining to equality-related knowledge was considered.


*Table 2: Impetus to including equality-related information*

#### *Source: authors' own work.*

*\* The n of 85 refers to the number of equality-related information generated during the 60 legislative processes. It is different from the n of 96 in Table 1 (second column: evidence-based and relevant to gender equality), as the issue of the impulse for consulting experts was not differentiated for each type of expert but was instead applied once for all experts (3. A-D).*

The procurement of equality-related information was initiated on 94 occasions during the 60 legislative processes analyzed, in the form of research, analysis or consultations (see Types 1‒4). In 74 percent of the cases, the impetus came from the administration with the executive accounting for 22 percent. Only three of the 94 triggers originated with the parliament. Minor differences are apparent in regarding the different types of information obtained. The main impetus came from the administration for all four types. However, the executive did include knowledge from interest groups relatively frequently.

The total of 94 information procurement triggers in Table 2 shows that the same type of gender-relevant knowledge was initiated by different actors simultaneously within a legislative process. For example, there were seven legislative processes where the administration and the executive suggested commissioning the administration to conduct analyses. Not once did the trigger come only from the executive, instead it came much more often from the executive and administration combined. This means that the administration plays a most important role concerning the decision of including evidence in policy making.

#### *3.2.2 Timing of the inclusion of evidence-based and gender-sensitive knowledge*

Our investigations also reveal at which stage of the political process the impetus was given for including evidence-based, gender-sensitive knowledge. We distinguished between the following stages:


Table 3 presents the results. Again, only information procurement involving equality-related knowledge is taken into account.


*Table 3: Type of gender-related evidence, by stage*

*Source: author's own work.* 

*\* The n of 96 refers to the number of equality-related information generated during the 60 legislative processes (see also Table 1 second column: evidence-based and relevant to gender equality).*

The results in Table 3 show that gender-relevant evidence was commissioned, generated internally or obtained through consultation 119 times during the 60 legislative processes. The figure of 119 was arrived at because the same type of information procurement extended through different stages of a process. For example, an administrative department made calculations about the financial effects on the Canton and the families in the impulse and pre-parliamentary stages.

In about two thirds of the cases, the evidence was included in the pre-parliamentary stage, but only in less than a third of cases in the impulse stage. It is interesting that the procurement of information through research and analysis commissioned externally was used the other way round: two thirds of studies were commissioned during the impulse stage and one third during the pre-parliamentary stage.

For the second and third type of information procurement (inclusion of empirical knowledge from the administration, inclusion of expert knowledge from the administration or science) it is often the case that the information search is released only during the pre-parliamentary stage. By contrast this is rarely the case during the impulse stage. In other words, types of information procurement are often established during the impulse stage, which extend into the next pre-parliamentary stage. As an example, an interest group or head of unit specialised in family and equality affairs is involved at the start of the legislative process and is then consulted throughout the process.

In conclusion, the impetus to include evidence-based and gender-related knowledge came from the administration in most instances. It originated a lot less from the executive. The lack of impetus from parliament is also striking. However, this may be due to the fact that evidence-based knowledge had already been processed during the pre-parliamentary stage and the members of parliament were then able to use it. As a result, they didn't need to initiate its development. For those legislative processes where evidence-based and gender-sensitive information has been produced, the impetus mainly came during the impulse or pre-parliamentary stage. Furthermore, it is often the case that types of information procurement are established, which then continue into the next pre-parliamentary stage. For that reason, the starting phase is crucial concerning the decision to include evidence in policy making.

#### *3.3 Explaining the differences in the inclusion of evidence-based, gender-sensitive knowledge*

The research project also analyzed the use of evidence-based and gender-sensitive knowledge. One question was whether the varying frequency of inclusion of such knowledge is attributable to structural differences between cantons or procedural differences in relation to specific features of the legislative projects.

#### *3.3.1 The influence of structural inter-cantonal differences*

A quantitative approach was taken to the analysis of the influence of canton-specific structural factors. The approach is based on the use of evidence-based and gender-related information in cantonal tax and social transfer policy being contingent on two groups of influencing factors. On the one hand are those factors fundamentally in favor of incorporating evidence in the legislative process. On the other hand are factors which tend to favor recourse to gender-sensitive information. To this end, data on 13 variables was taken for the 60 identified legislative processes (e. g. population size of the canton, relative size of the cantonal administration, a conservative majority in the executive branch, the proportion of women in the legislative and executive branches). The selection of the variables was theory-based. The relevant data is taken from available statistics (e. g. population size or party composition of cantonal governments) or from expert interviews with administrative personnel responsible for tax system and social transfer policies in the 26 Swiss cantons. The data evaluation was initially descriptive then bi- and multivariate.<sup>3</sup>

This analysis shows that the legal or strategic embedding of the compatibility of family and employment in the constitution or in a cantonal model has a positive effect in terms of demand for evidence-based and gender-related information. This finding is supported by many publications on gender mainstreaming. These sources emphasize that the explicit formulation of political will at the highest level is fundamental to the inclusion of gender-specific considerations for a cross-sectoral issue (see Malli 2005). Furthermore, the quantitative analysis revealed a connection between staff numbers and the procuring of evidence-based, gender-sensitive information: the more employees there are in public administration per 100 inhabitants, the more likely it is that evidence-based and gender-related information will be available for the legislative process. This finding also corresponds to previous research in other areas of policy. Larger administrative offices have more resources and opportunities to commission and monitor studies and conduct their own factbased investigations (Rieder/Widmer 2007). Finally there was a link between party membership, the gender of department heads and the availability of evidence-based, gender-related knowledge: members of left-of-centre parties tend to identify more with issues relating to gender equality, female department heads also refer more to evidence-based, gender-sensitive information than their male counterparts. The hypothesis that the presence of women in positions of political power is conducive to successful gender mainstreaming has been confirmed (Mosimann/Giger 2008). However, the importance of

<sup>3</sup> The methodological basis and the results of the quantitative analysis are presented in detail in Balthasar/Müller 2014.

women being represented in legislative authorities for gender-sensitive policymaking is a debated question with no definite answers (Kulawik 2005).

#### *3.3.2 The influence of procedural differences between the models*

The question of the extent to which the inclusion of gender-sensitive evidence was influenced by specific features of the legislative projects was examined on the basis of four detailed case studies. Detailed internet research was carried out to reconstruct the legislative processes. All the documents relevant to the legislative process were viewed (legal texts, communications, documents on the approval procedure, motions, information from the cantonal authorities). Information which wasn't accessible online was evaluated on site in the cantonal archives. Occasionally press releases and reports in the printed media were also considered and personal interviews with administrative personnel responsible for tax system and social transfer policies were held.<sup>4</sup>

The case studies were organized on a theoretical basis. The results showed that political processes that follow the linear phase model of the rational policy cycle theory are conducive to the application of evidence-based and gender-sensitive knowledge (DeLeon 1999). The example of the law passed on extra-familial childcare facilities in Canton Fribourg (2011) illustrates this point. At the start of the legislative process, there were no obvious coalitions in favor of extra-familial care. On the other hand, there was a fundamental normative consensus that tax and social transfer policy should not hinder gender equality. The question was not whether the motion to improve extra-familial childcare should be passed, instead it was how – i. e. using which operational means – to tackle the issue that dominated the discussion. The fact that initially there were no solutions forthcoming may well have contributed to the receptiveness to evidence-based knowledge. This openness enabled the administrative authority to compile a wide range of knowledge. Evidence was used to demonstrate deficits, which were taken as a basis on which to formulate efficient measures.

The example of Canton Fribourg also shows how evidence-based, gender-sensitive knowledge is favored when a legislative process is broadly supported. At the end of the 1990s, a first committee was founded as part of a commission to promote a comprehensive family policy, the remit of this committee was to review the requirement for equality-oriented policies within a broad political development programme and to define the basis for a cantonal family policy. After the cantonal government decided on a total revision of the law on childcare facilities for children of pre-school age, it set up a sec-

<sup>4</sup> The theoretical and methodological basis and the results of the case studies are presented in detail in Balthasar et al. 2012 and Blatter et al. 2015.

ond committee in 2007 to formulate a draft law. This committee also enjoyed broad acceptance with political parties and interest groups. The committee decided to commission two external studies to bring in some expert support for their draft law. The conclusions of these studies confirmed the cantonal government's position in improving extrafamilial childcare in Canton Fribourg and were included in the submission to the parliament.

In situations where contrasting value judgments come into conflict, gender-related evidence plays less of a role in the legislative process. The case studies of cantons Uri, Aargau and, to a certain extent, Nidwalden illustrate this point and correspond to findings from earlier studies (Freiburghaus/ Zimmermann 1985; Frey 2012). With reference to the political science approach of the Advocacy Coalition Framework (Sabatier/Weible 2007) and the Multiple-Stream-Approach (Zahariadis 2007), it was shown that, even when there are strongly conflicting opinions, evidence-based facts can still come into play. One example of that is the tax law revision in Canton Uri in 2011. In this instance, important coalitions with opposing viewpoints were involved and they brought their different gender-political agendas into the process. This came to the fore through the discussion about providing relief to families with children through tax deductions not only for extra-family childcare, but for family childcare too. The proponents of such a deduction argued that a tax deduction just for extra-family childcare discriminates the traditional family model. The Governing Council opposed this argument. It provided evidence with evaluations by the tax authorities that such a deduction would instead penalize dual-earner families. In addition it ordered a judicial report by a professor in tax law (Simonek 2008). This report stressed that a tax deduction for parents who look after their own children would infringe the tax harmonization law and would therefore be unconstitutional. The weight of the evidence in these instances was mainly influenced by the reputation of the parties involved. By using evidence from a professor in tax law, the Federal Supreme Court and Federal Council, the Governing Council of Canton Uri managed to convince those members of parliament who were undecided, thus preventing measures that would have been detrimental to gender equality. This strategy worked in Canton Uri, although many of the members of parliament did not share the views of the Governing Council.

The importance of the "right" choice of evidence in a context of strongly differing views, is also illustrated by the example of the revision of the social assistance and prevention act in Canton Aargau (rejected in 2012). Put under pressure by a parliamentary request of the Liberal Party, the government started to revise the existing law on social security and prevention, which included various elements to support extrafamilial child care provisions. To support the direction of the revision, the government asked a private company to calculate the potential costs of the new law for the municipalities. In this instance the evidence ultimately proved counterproductive, since it primarily provided ammunition for the conservative coalition. The progressive coalition did not succeed in providing convincing facts to support its demands for uniform legal framework conditions for extrafamilial childcare. The evidence provided actually strengthened the opposing case as it focused on the costs instead of the benefits of extra-family childcare.

## **4 Conclusions for an evidence-based gender equality policy**

The quantitative analysis and the case studies' outcomes show that the influences involved in improving the availability of evidence-based and gender-sensitive knowledge in legislative processes are extremely varied and numerous. As a result, the scope for drawing general conclusions for equality policy from the project work is limited.

However, there is clearly a wealth of potential opportunities to draw attention to equality in the areas of tax policy and social transfer. For the years under review, there were 76 cantonal legislative amendments identified relating to equality. There were also many other relevant amendments to ordinances and regulations that were not included in the study.

The research indicates the important role of the administration as an initiator in involving evidence-based and gender-sensitive knowledge. In most cases the administration acted as the trigger for applying that knowledge. One key result of the study is that only a quarter of all people who work in public administration and are responsible for tax and social transfer legislation evaluate gender-sensitive information, e. g. in literature, when preparing draft laws. It is therefore necessary to improve access to evidence-based, gender-related information (Müller/Balthasar 2014). This could be done through publications or websites, for example, which raise awareness and target specific groups. It could also be done by sensitizing specific target groups to the false incentives often inherent in tax and social transfer laws and informing them about the corresponding publications.

One government authority from the case studies demonstrated an important condition for incorporating evidence-based, gender-sensitive evidence: this authority doesn't only plan the operational implementation of issues but actively shapes them in the legislative process at a strategic level. The foundation for including evidence-based and gender-related information is set in the impulse stage of legislative processes. That is where types of information procurement are established, which are subsequently maintained throughout the pre-parliamentary stage. Shaping means, for example, setting an agenda where the responsible parties, based on their own professional competence, preemptively and objectively demonstrate causes and requirements of a political issue and formulate various proposed solutions. In the next phase of policy formulation, the priority is then to demonstrate the most efficient and effective proposal and point out the corresponding concrete equality-sensitive evidence in the debate.

Another insight from the empirical research is that the notion of evidence should not be restricted to systematically generated knowledge about the effectiveness and efficiency of specific policy measures. This type of evidence can play productive roles in law-making processes, but it seems that they are used only under specific circumstances. In consequence, those who promote or study evidence-based policymaking should take into account other kinds of evidence, which play different roles in the policy process. This may be evidence-based information gathered and evaluated internally, expert knowledge or knowledge of consulted interest groups. If there is a plurality of policy goals, which compete for normative dominance or cognitive recognition, systematically generated knowledge serves other functions, such as proving the legitimacy of disputed policy goals and the appropriateness of proposed policy measures or generating attention for specific policy goals and connecting policy measures to goals in other policy fields (Blatter et al. 2015).

It is also worth noting that equality experts were included relatively rarely in the relevant commissions or working groups. Their systematic inclusion in preparing legislation could result in equality-sensitive evidence receiving more attention when revising tax and social transfer laws. For that to happen, equality experts need to try to get more involved in tax and social transfer policy, while those responsible within government would need to be more active in including them in the process. A useful instrument to estimate the potential of exerting influence could be policy monitoring (Smith 2001). That means analyzing tax and social transfer policy systematically with a view to pending revisions of laws and ordinances to include equality issues at an early stage and be proactive in adding them to the ongoing discussions.

## **5 Reflections and need for research**

The analyses pose a number of theoretical and methodological questions; they also leave some research questions unanswered.

This chapter is based on concepts of "gender-sensitive" and "evidence-based policymaking", which are very much open to question. We expressed a number of significant theoretical reservations in the first section. On that basis, the chapter takes the position that the state can and should contribute to shaping tax and social transfer policy in such a way as to promote gender equality. We consider a social transfer and tax policy in line with the "dual breadwinner/dual earner" model (Pfau-Effinger 1998; Pfau-Effinger 2000) to resemble gender equality, as it is conducive to an equal distribution of childcare and gainful employment. We are aware that this is a restricted perspective, which doesn't take account of the fact that the division of gainful employment and family structure between men and women is heavily influenced by cultural factors (Pfau-Effinger 2000). Furthermore, a more open interpretation of the term "gender-sensitive", which doesn't focus solely on the compatibility of family and working life, would also help to pave the way for further policies which favor gender equality. There are also good reasons not to restrict the concept of evidence-based policymaking excessively to a narrow rational-instrumental perspective. Evidence is not only applied to the political process to show that a measure is a practical and efficient means of reaching a political goal. The application of systematically acquired knowledge is just as important in consolidating certain problem definitions or in documenting and/or questioning the legality and constitutionality of certain measures (Blatter et al. 2015).

These theoretical reservations about the rationale of the study are complemented by methodological challenges arising from the analysis of the role of evidence in political processes. Firstly, it is difficult to distinguish between evidence and non-evidence-based lines of argument. Secondly, it is not easy to determine empirically whether evidence has been used in the research process. Thirdly, it often proved difficult to reconstruct legislative processes retrospectively: these processes normally last for a number of years and their pre-parliamentary progress is often poorly documented. As a result, the people now responsible often were not sure whether or how empirical or expert knowledge had been used in past years for a specific legislative project.

The study also leaves important research questions unanswered. First and foremost would be the question of how much influence evidence-based and gender-related knowledge had on the content of legislative processes. Some initial indications were nevertheless acquired through case studies. Some showed that the use of evidence in itself does not in any way ensure the adoption of gender-related measures. Systematically acquired knowledge can also be used to achieve the opposite outcome (Canton Nidwalden). Other case studies, though, suggest that the inclusion of systematic knowledge at the right juncture can support gender-sensitive measures (as in Canton Fribourg for example) and that the failure of gender-sensitive policies can also be attributed to insufficient use of evidence (Canton Aargau) (Blatter et al. 2015). This underlines the importance of carrying out further research on the extent to which evidence helps to shape public policy.

## **References**


http://www.egalite.ch/uploads/pdf/QuandLeTravailCoute/QuandLeTravail-Coute-3CantonsAll.pdf


to evidence-based policy making in a gender-relevant policy field. Literature review. Luzern: University of Lucerne, Institute for Political Science and Interface Institute for Policy Studies.


# **3 Interdependencies between gender and other categories of social inequality**

# **Varieties of childcare policies in Swiss municipalities: Bounded possibilities for gender-equality and social cohesion**

*Christine Zollinger & Thomas Widmer*

## **1 Introduction**

Until the 1990s existing policy instruments in Swiss family policy, such as financial transfers for families with children, supported a traditional male-breadwinner model (Häusermann 2006, Häusermann/Zollinger 2014). In the last decade, Swiss family policy adapted towards 'new' social needs of women to reconcile work and care responsibilities (Bonoli 2005, Taylor-Gooby 2005). Political claims for childcare policies have been included in the political agendas at federal, cantonal and municipal levels.

Since the beginning of the millennium, several municipalities have institutionalized publicly supported childcare services (Stadelmann-Steffen/Oehrli 2013). However, the regulations for non-parental<sup>1</sup> childcare (policy-designs) differ among municipalities and among what is on offer. They differ not only in terms of fee structure but also in terms of regulations concerning the degree of professionalization, i. e. the existence of underlying pedagogical concepts and of regulations concerning the education of care workers. Moreover, some municipalities have set up day-care centers, whereas others provide childcare by (third-party) families (Tagesfamilien).

Recent insights on family policy development in Switzerland from Häusermann and Kübler (2010) have shown that the emergence of a new frame, namely family policy as labor market policy, has led to ambiguous agreements between employers, market-liberal and left-wing parties supporting the institutionalization of childcare services at the federal level. However, it remains unclear whether the variety of childcare services at the municipal

<sup>1</sup> Parental childcare includes childcare by other family members such as grandparents.

level – where market-liberal and right-conservative parties dominate many governments – can be explained by the concept of frames as proposed by Häusermann and Kübler (2010).

Moreover, as a typical social investment policy, childcare policies are discussed as having a high potential for more equality in societies (Esping-Andersen et al. 2002, Morel et al. 2012). First, by the activation of female human capital for the labor market they can contribute to gender-equality in employment. Second, by investing in the education of pre-school children these services can contribute to social cohesion between people with different socio-economic backgrounds. Theoretically, the implications of childcare policies on the two dimensions of equality are of particular interest, given that the literature has long agreed that we need a better understanding of the trade-offs between distributive and socio-cultural goals of new social policies (Fraser/Honneth 2003).

Against this backdrop, we are interested in two interrelated questions. First, we investigate why there can be observed such a broad variety of childcare services at the municipal level. Second, we analyze the potential consequences of such a broad variation in childcare provision for both gender- equality and social cohesion. Herewith we join two questions obviously connected but rarely analyzed in a common framework including policy formulation and policy consequences.

Our argument goes as follows: To answer the first research question, we hypothesize that the existence of four frames in the policy space of childcare policy making – a gender-equality, a labor-market, a male-breadwinner frame and a social-integration frame – have an influence on the specific type of childcare service provided in a municipality. To answer the second research question, we hypothesize that the four ideal typical models of childcare services have different potentials for gender-equality and for social cohesion. Hereby we subsume socio-economic dimensions such as income and education under the label social cohesion, with the exception of gender dealt with separately under the heading of gender-equality.

The hypotheses will be investigated by a causal process analysis applied in a comparative qualitative case study design. The focus is on the specific childcare-arrangements institutionalized in two most similar suburban Swiss municipalities in the canton of Berne.

In the chapter, we will first present the theoretical background, before we discuss design and methodology. We will proceed with the presentation of the empirical findings from the case studies, before these findings will be contrasted with the theoretical expectations. In the concluding chapter, we will answer the research questions and dicuss our findings, with the article concluding with some implications for the praxis and future research.

## **2 Theoretical framework and hypotheses**

This chapter first presents a theoretical framework explaining the relationship between the dynamics of the political decision-making process and the varieties of childcare services, before a conceptualization of the expected outcomes of those services follows.

#### *2.1 A theory explaining the varieties of childcare services*

#### *State of the art*

The expansion of childcare policies in Switzerland since the mid-1990s, as well as other West-European countries, constitutes a major policy change. It involves an important change both in the underlying conceptions of social risks to be addressed, as well as in the pursued role of women and mothers in society (Fleckenstein 2011: 5445, Häusermann/Kübler 2010: 166). Therefore, the literature has acknowledged that traditional theoretical factors for explaining welfare state development, such as power resources or structural explanations, have to be supplemented with the concept of ideas and frames in order to explain family policy change (Fleckenstein 2011, Häusermann/ Kübler 2010, Kübler 2007).

#### *Theoretical framework and hypothesis*

Häusermann and Kübler (2010) show that the emergence of a labor-market frame in the field of family policy on the federal level, has led to a new winning coalition between liberal actors and actors from the left supporting the provision of non parental childcare. However, this new winning coalition can be denoted as 'ambiguous agreement', because the actors in this coalition are pursuing different policy objectives (Häusermann/Kübler 2010: 184‒185, Palier 2005). Liberal actors see in the public provision of day care services a possibility to activate high skilled female human capital into labor market charged with a danger for a gap of high skilled human capital. Left-wing parties at the same time focus on achieving gender-equality in employment with non-parental childcare. The authors argue that "family policy development must be explained by a focus on underlying policy frames (…) that organize thought and behaviour of policy actors who coalesce to produce majorities in decision-making opportunities" (Häusermann/Kübler 2010: 163). *Framing* means here the ability of actors to influence how a social problem has to be interpreted (Benford/Snow 2000: 216). Frames as interpretations of a given political measure clarify its aim (Hall 2010: 171‒172). Policy frames emphasize specific aspects of a measure while fading out others. However, in contested politics distinct frames put forward by different actors compete perpetually with each other (Tarrow 1994: 123).

Hence, the authors have shown, that diverse policy frames in family policy-making are a potential for coalitions between distinct actors. Moreover, we argue that the frame put forward in a decision-making process, has an influence on the design of the resulting childcare policies. Put differently, the varieties of childcare services can be explained by different policy frames in policy-formulation. We hypothesize that actors in decision-making processes make choices about policy designs based on their understandings of the challenges and problems they face (Jenson 2009: 449).

*Hypothesis 1: Different frames which are mobilized to build a winning coalition in the decision-making process explain diverse types of childcare services.* 

#### *Conceptualization of the policy space of childcare policy-making*

Family policy is undergoing an important shift concerning the abolishment of the logic of a transfer-oriented poverty policy, which was the predominant aim of non parental childcare in a traditional male-breadwinner state, and the rise of a new concept: the social investment concept. In contrast to social transfer policies, which aim to protect families concerned with poverty, the focus of social investment policies is on the redistribution of opportunities (Esping-Andersen et al. 2002, Jenson 2010: 77). The social investment approach follows a preventive logic and is oriented towards the medium- and long-term future. "(…) it should avoid 'spending to insure against misfortune' but be willing to make 'investments' that will increase capabilities" (Jenson 2010: 77). Poverty prevention in the social investment logic implies an investment in 'new' social risks of women and low skilled persons.

We argue that in the context of this fundamental shift, decision-making processes in the field of childcare policy are concerned with a multiplication of frames (Jenson 2010). First, we argue that according to the social investment logic, the policy space of childcare policy-making is multidimensional (Bonoli 1997, Bonoli/Natali 2012). Policies for non-parental childcare concern investments in children and women. Second, according to Jenson (2010: 73‒74), social investment as an ambiguous concept provides a common-sense meaning which is open to multiple interpretations and can penetrate and link numerous policy communities.

Hence, non-parental childcare pursues different goals at the same time: Perceived as an investment in women, it can first be framed as a genderequality policy (Häusermann/Kübler 2010: 173). According its potential for de-familialization, childcare services substitute the female caregiving function in male breadwinner systems (Fraser 1994, Leitner 2003, Orloff 1993). From this perspective, childcare services contribute to gender-equality in employment, since they allow women to pursue career advancement. Second, the investment in women can be framed as a labor-market policy. By the activation of female human capital it can supply labor markets with high skilled women (Häusermann/Kübler 2010: 173). Moreover, perceived as an investment in children, childcare policies can contribute to the social integration of children from different socio-economic and socio-cultural backgrounds and can support social cohesion (Esping-Andersen et al. 2002). Hence, it can thirdly be framed as a social-integration policy. Moreover, in the realm of a male-breadwinner model, which was dominant in Swiss family policy until the 1990s, non-parental childcare can be seen as a means to prevent poverty of those women concerned with the loss of a male breadwinner (Binder et al. 2003). Consequently, non-parental childcaree can fourthly be framed as a male-breadwinner policy. Therefore, the policy space of childcare policy-making is conceptualized as presented in Figure 1.

#### *Hypothetical relations between frames and policy designs*

Based on the theoretical argument of Häusermann and Kübler (2010), we argue that these four frames in the space of childcare policy-making create an ideational opportunity structure for new coalitions supporting childcare policies. More concrete, we suppose that depending on the frame mobilized by political actors in a decision-making process, we can expect four different ideal typical models of non-parental childcare services (see Table 1).


*Table 1: Hypotheses about frames and types of childcare services*

The *universal model of childcare services* is most probably designed as a public provided day-care center, which is characterized by an income related fee structure. This model allows universal access for families of all socio-economic strata and enables the integration of children with different socio-economic and cultural background, since it is designed to support the activation of women and an investment in children. Regulations concerning the pedagogical education of care workers complement this gender-equality oriented model. A public visible, highly professionalized offer containing an income related fee structure displays a typical example of a universal childcare service.

The *market-liberal model of childcare services* relies mainly on the (neo-) liberal idea of minimal state intervention and a limited investment in social welfare. Similar to the universal model of childcare this model supports the activation of women for the labor market, whereas promoting social cohesion by an investment in the early education of children is not on the agenda. Hence, this model comes without an income related fee structure. Services are provided by private organisations, such as profit-oriented companies or by non-profit associations. Following the market logic of demand and supply, no regulations concerning a pedagogical care-concept neither for pedagogical education standard of the employees exist.

Social cohesion among children from different socio-economic and cultural backgrounds is the main focus of the *model of social-integrative childcare*. By an income related fee structure, it mainly attracts children from the lower income strata and with migration background. This model does not intend to activate women for the labor market and is strongly oriented at traditional gender roles. Services are provided by mothers most preferably in their own households with a low degree of professionality in terms of pedagogical education standards for care workers.

The *model of traditional childcare* is designed neither to make women available for the labor market nor as an investment in children and is characterized by an offer provided by mothers in their home. Similar to the model of social-integrative childcare, care work is not seen as a profession, which has to be paid for. Hence, the offer does not include a pedagogical concept and is characterized by a low degree of professionalization in regards to regulations according pedagogical education of care workers. In contrast to the social-integrative model an additional investment in childcare services is excluded and the offer is rarely visible in the public space and targeted to a limited group of single mothers.

## *2.2 Conceptualization of the outcomes of childcare policies*

#### *State of the art*

Recent studies about the outcomes of childcare policies in Switzerland show two things. First, Schlanser (2011) and Abrassart and Bonoli (2014) show, that day-care centers are used more frequently by families from higher than from lower income strata. Second, Stadelmann-Steffen (2007) argues that for the Swiss cantons there is a correlation between the existence of public day-care centers and women's employment rates. This relation is stronger for women from high and middle-income strata than for women from low-income strata. Hence, public day-care centers affect mostly women with high and middle incomes. However, both studies investigate only the outcome of day-care centers and only on the level of the user. Evidence about the outcomes of other types of services and about the outcomes on the providing side are lacking so far.

#### *Theoretical framework and hypothesis*

Ten years after the introduction of the dimension of gender into the welfare state conception (Fraser 1994, Orloff 1993) efforts to include both gender and class into the conceptualization of welfare states appear important, "(…) since most feminist research on welfare states have privileged gender at the expense of class" (Sainsbury 2008: 106). Demands for 'recognition of difference' for example of groups mobilized under the banner of gender have replaced class interest as source of political mobilization (Fraser 1995: 68). Fraser and Honneth (2003) assume though that thinking about contemporary social justice requires both dimensions – redistribution and recognition.

As mentioned earlier, as a social investment policy, childcare policies can potentially contribute to two dimensions of equality. First, as a reconciliation policy, those services can contribute to more gender-equality in employment. Second, as an investment in early education of children they can contribute to the redistribution of opportunities from the better off to the children from families from lower income strata (Esping-Andersen et al. 2002, Jenson 2009: 446, 2010: 61‒66).

Based on these arguments we hypothesize that all four different ideal typical models of childcare policy have according to their design a specific potential for gender-equality and for social cohesion (see Table 2):

*Hypothesis 2: Diverse types of childcare services have different potentials for gender-equality and for social cohesion.* 

*Table 2: Hypothetical potential of the four ideal types of childcare services for gender-equality and social cohesion*


### *2.3 Operationalization of the outcomes*

To obtain a comprehensive understanding of the trade-offs of different ideal types of childcare services in terms of gender-equality and social cohesion, we assessed the social outcomes of those services on two levels, namely on the user *and* on the provider side. According to the social investment paradigm, social security should be obtained by commodification of previous labor market outsiders (women). Hence, we argue that only if we analyze both, whether women can be activated for labor market and at the same time whether they find good jobs, we can assess the potential of those policies for gender-equality. This comprehensive view captures the outcomes of childcare services along the two dimensions and between provider and user of those services (see Figure 2).

#### *Outcome on the level of the user*

As a social investment policy, which promotes more equality of opportunities the aim of new social spending on the level of the user can be seen as twofold. First it can offer labor market opportunities for parents (mainly mothers). Second, it can redistribute opportunities by investing in the future of children. Consequentially, it is argued here that the ideal types of childcare services can be seen as effective, if they support the integration of women into the labor market (dimension of gender-equality) and if the services are used equally by children from different income classes (dimension of vertical stratification).

#### *Outcome on the level of the provider*

To conceptualize the gendering outcomes of new social policies on the level of the provider, we follow the argumentation of Lister (1994). The parallel dimension of de-commodification (Esping-Andersen 1990) proposed by Lister is de-familialization, which she defines as "the degree to which individual adults can uphold a socially acceptable standard of living, independently of family relationships, either through paid work or social security provision" (1994: 37). Consequently, from a provider perspective an ideal type of childcare service will be classified as having a high potential for gender-equality, if an employment in a service provides enough income to maintain an independent household for the care workers (dimension of gender-equality). Moreover, if the employment structure for persons in childcare supply allows for the integration of care-workers from different socio-economic backgrounds, the ideal type of childcare service can be classified as having a high potential for social cohesion.

## **3 Case selection and methods**

The contribution focuses on policies for childcare services in Swiss municipalities. According to the subsidiary principle, municipalities do not only have to implement regulations determined by the cantonal level, but they possess their own scope of action (right to act and right to decide). The municipality holds the main authority to design public policies for day-care services. Hence, a care-arrangement in a Swiss municipality represents a case for the analysis. A care-arrangement consists of all public policies for non parental childcare in a select municipality. The focus lies on public offers for pre-school childcare (0 to four years of age).The comparative qualitative case study design (Yin 2009) involves two suburban municipalities in the canton of Berne. The goal is to work out what conditions can explain the varieties of policy design in two most similar municipalities (Gerring 2007). In the most-similar case selection strategy a range of independent variables with a potential influence on family policy design are held constant (see Table 3).


*Table 3: Criteria and indicators held constant for case selection* 

Two municipalities in the canton of Berne have been selected: In 2012, 9'700 persons were living in Münchenbuchsee and 3'988 in Pieterlen. During the time of investigation, Pieterlen was governed by a majority of members of the liberal party FDP (3 seats FDP, 1 SVP and SP). Münchenbuchsee was governed by a majority of members of the right conservative party SVP (4 seats SVP, 3 SP, 1 EVP and Greens). The share of families with children below seven years (around 12%) as well as the share of single parents (around 5%) lies in both municipalities around the average of Switzerland (BFS 2000). The unemployement rate in both municipalities was around the Swiss average of 1.8% and the share of people in need of social assistance was 4.8% in Münchenbuchsee and 6.1% in Pieterlen (BFS 2000).

For the case studies we conducted 14 expert interviews with representatives from the government, from the administration, from childcare providers and from political parties (Kvale 2007). Through expert interviews and document analysis we collected data about coalition formation. For the within case analysis the method of process tracing was employed (George/Bennett 2005, Gerring 2007). Outcome data were collected with a written survey. Data about the income-related use of different services and data about the income distribution came from cantonal sources.

## **4 Findings**

The chapter first presents the framework for childcare policies in the canton of Berne. A description of the decision-making process for each case with a discussion of the frames used in the processes follows. Finally, the outcomes of a given type of childcare service are presented.

#### *4.1 Framework for childcare policies in the canton of Berne*

Since 2005, the canton of Berne defines the regulations for non-parental childcare offers in a decree (Verordnung über die Angebote zur sozialen Integration, ASIV). This decree contains guidelines concerning quality standards and an income related fee structure, which is applied equally to all municipalities and all kinds of childcare provision. The minimal fee for care per day and kid is 6.60 CHF.

Moreover, by the cantonal system of equalization of financial burdens, the canton pays the municipalities a fixed amount (Normkosten) for each place in an offer for non-parental childcare approved by the municipality and authorized by the canton. Whereas care workers in a day-care center have to have a tertiary education, daily mothers have to attend an 18 hours introduction only. Moreover, daily mothers provide care in their own home, whereas daycare centers are in need of specified rooms that are suitable for children in pre school age (see Table 4).


*Table 4: Cantonal regulations for childcare services in the canton of Berne*

#### *4.2 Case study Münchenbuchsee*

#### *Decision-making process*

The local green party in 1998 initiated a popular initiative asking for a public day-care center. However, the issue of the popular initiative was coupled by significant opposition by the municipal council even in the preparatory stage of the ballot (Interview statement). Strong opposition came particularly from the SVP. The people rejected the initiative at the ballot of June 7th 1998.

Two women launched an association for a private day-care center, which opened in autumn 2001 and attracted – although costs are high – an impressive demand with a waiting list three months after project start. In May 14th 2004 the people of Münchenbuchsee approved in a public vote a budget of more than 300'000 CHF for twelve public endowed places in the privately provided day-care center (Interview statement). The cantonal system of equalization of financial burdens covers the expenses for the twelve places, the municipality provides a deficit guarantee of 50'000 CHF.

#### *Patterns of childcare policy-making*

A coalition of SP, the Greens and FDP supported the day-care center in 1998 and 2004 (Botschaft Grosser Gemeinderat 1998, 2004). However, these parties supported the public funding of a day-care center with reference to different frames. On the one hand, members of the Green party as well as of the SP supported the bill with reference to a gender-equality frame and – to a lesser degree ‒ with reference to a social-integration frame. Members of the liberal party on the other hand supported the bill with reference to a labor-market frame. The national-conservative party SVP, refused the bill with reference to a male-breadwinner frame very strongly in the year 1998. In the year 2004, at least some members of the SVP joined the coalition of SP, green party and FDP with reference to a labor-market frame.

The initiative for the institutionalization of a day-care center in 1998 was launched by the female president of the Green party, mainly with reference to a gender-equality frame. For her it was important, that the next generation would not face the same reconciliation problems as she did. However, beneath the gender-equality frame, a social-integration frame was present as well. The initiative supporters wanted to create a high quality day-care center.

Interestingly, interview statements show that the social-integration frame was not actively put forward in the decision-making process. Members of the Greens and the SP speculated that such a framing does not convince the SVP as well as the FDP. A majority of the FDP in the parliament supported the initiative during the whole period with reference to a labor-market frame. According to an interview statement, they saw in the integration of high skilled female human capital into labor markets a possibility to ensure economic growth. Interestingly, the creation of the day-care center was not perceived as an extension of the welfare state. Non-parental childcare activates female human capital for the labor market, what in turn should bring a return of investment for employees and – in the form of taxes – for the municipality. In parliament a majority of the FDP supported the initiative but only under one condition: The day-care center has to be provided privately and the amount by the municipality should be not higher than 30'000 CHF (Botschaft des Grossen Gemeinderates 1998, Debatte Grosser Gemeinderat 2004: 14). Different interview partners have pointed out that a private organization is much cheaper for a municipality: Private associations, in which the organization and administration of the day-care center is made by voluntary workers, bear the financial risk of the provision of a day-care center.

Opposition against the initiative came from the SVP. Different interview statements point out that those actors refused the creation of a day-care center with reference to a male-breadwinner frame. They denied any public demand for such an offer. According to them, it was not the duty of the state to organize and provide non-parental childcare, since this can and has to be provided by the families on their own (Interview statement).

Interestingly, the SVP, even though they refused a day-care center until 2004, did not challenge at any point the existing offer of family day care. According to an interview statement, it was the position of the SVP that offers for non-parental childcare should be provided by housewifes, which do that for a low wage in their leisure time. The offer provided by daily mothers was according to them adequate for the existing demand for childcare (Botschaft des Grossen Gemeinderates 1998: 4‒5). With reference to a male-breadwinner frame, they refused the need for an expanded offer for childcare until 2004. Then, at least some members of the SVP supported a day-care center with reference to a gender-equality as well as a labor-market frame (Debatte Grosser Gemeinderat 2004: 14‒5).

The decision-making process in Münchenbuchsee shows that different actors put forward different frames at different points in time. The multiplication of frames led to a specific care arrangement consisting of a day-care center and an offer of family day care. The FDP supports a day-care center with reference to a labor-market frame, supported only an offer provided privately. Representatives of SP and the Greens supported the day-care center with reference to a gender-equality frame. For the SVP in turn the existing family day care offer was supported because it was seen as the most adequate form for childcare. The existence of an ambigous agreement between a gender-equality und labor-market frame on the one hand and the existence of a male-breadwinner frame on the other hand led to a care-arrangement containing a day-care center which can be labelled as universal childcare in combination with a family day care offer, which can be labelled as a traditional childcare service.

#### *Type of childcare service*

From 2004, the care-arrangement in Münchenbuchsee comprises a combination of a day-care center on the one hand and family day care on the other. Caused by its high degree of professionalization in terms of education level of care workers and pedagogical concept, the day-care center is categorized as a universal model. Although set up with an income related fee structure whereas it has a bias toward a market-liberal model, since a private actor has to provide it. On the other hand, there exists a privately provided offer by families. This offer is categorized as a traditional child-care service, caused by its low level of professionalization. Because of the income related fee structure it has a bias towards a social integrative model of childcare.

#### *Outcome of the type of childcare service*

We will first present our findings on the level of the users of the services, before we proceed with the findings for the outcomes on the level of the provider. Concerning the potential of the childcare services for gender-equality on the level of the user, the data show that both offers are used by around 11% of the family households with children below seven years. Hence, both offers support female employment and have therefore a high potential for genderequality.

Table 5 presenting the findings on the proportion of children using a service from different income quintiles and shows the potential of the care-arrangement for social cohesion on the level of the users. In both types of childcare children from low-income strata are underrepresented. However, children from low-income strata use the traditional childcare twice as often as the universal childcare. At the same time, children from high-income strata use the universal childcare to a much higher degree (28%) than their fair share of using (4%). Additionally children from high-income strata use universal childcare nearly four times more often than traditional childcare. To summarize, whereas children from families from middle and high-income strata are overrepresented in the universal childcare, children from families from low-income families are overrepresented in the traditional childcare service. Both offers have a limited potential for social cohesion.



*Source: Gesundheits- und Fürsorgedirektion des Kantons Bern*

To assess the potential of the type of childcare service for gender-equality on the level of the provider, we investigated whether the care workers gain enough income to maintain an independent household. Whereas all ten employers in the universal childcare service earn more than 3400 CHF a month (Interview statement), a care worker in the traditional childcare service earns between 320 and 480 Swiss francs per month if one extrapolates to full-time employment. The wage of a woman employed by the association is 6.30 Swiss Francs (CHF) per hour and child and the average degree of employment is 30‒40 percent. This is far from enough to maintain an independent household in Switzerland. Working in poverty is defined in Switzerland by a salary of less than 2450 CHF for a full time employment. Consequently, the traditional childcare service is categorized as gendering whereas the universal childcare model is categorized as de-gendering.

To assess the potential for social cohesion of the offers on the level of the provider, we investigated the socio-economic and socio-cultural background of the care workers employed in an offer. As illustrated in Table 6, the share of employees with migration background is very low in both offers. However, since in the universal childcare service only women from middle-income strata are employed it cannot be very helpful in obtaining social cohesion. We see only a very slightly tendency of the traditional childcare service to employ women from lower income strata and with migration background. Hence, both offers have a rather low potential for social cohesion.

#### *Table 6: Outcome on the level of the provider of the service: vertical redistribution, Münchenbuchsee (2011)*


*\*Low income strata was defined as an income of less than 41'000 CHF per year, high income strata as an income of more than 100'000 CHF per year.* 

#### *4.3 Case study Pieterlen*

#### *Decision-making process*

At the beginning of the year 2001, the Social Service and the Tutelage and Welfare Authority set a request to the Municipal Council, to establish a project group to establish an offer for childcare. The municipal council, headed by a majority of members of the liberal party FDP and members of the party SVP, doubted the existing demand for childcare, but they accepted to establish a project group in April 2001. It had the task to assess the demand for public childcare and to establish a concept for a non-parental childcare supply. At December 21st 2001 the cantonal Health and Welfare Directorate granted permission for ten places in a day-care center funded by the canton (276'000 CHF). At the municipal assembly of June 4th 2002 the people of Pieterlen supported the establishment of a public day-care center with ten places and the payment of 70'000 CHF as a deficit guarantee by the municipality.

#### *Patterns of childcare policy-making*

Members of the liberal party FDP as well as members of the social democrats (SP) put forward the creation of a childcare offer in Pieterlen with reference to a social-integration frame. Actors from the right-wing conservative party SVP refused the demand at least in the beginning with reference to a male-breadwinner frame. However, as it will be shown, the social-integration frame had an influence on the model of childcare, which was chosen, whereas the male-breadwinner frame did not.

The project group, headed by a male municipal councilor from the liberal party FDP, had the task to establish a concept for non-parental childcare. Different actors reported that not only the other members of the liberal party FDP but as well the conservative actors from the SVP had to be convinced of the idea to institutionalize non-parental childcare supply.

According to interview statements, the reference to a social-integration frame was most important to convince a majority of the municipal council and even more important to convince the people of Pieterlen to support a public day-care center. The leader of the project group played an important role in putting forward the idea of social cohesion and integration as an aim of non-parental childcare. He referred to the integration not only of children from low socio-economic strata but as well of the integration of children with different cultural backgrounds. According to his position, the social integration of children into the society should be achieved with a professional public offer ensuring early education.

Interestingly, this social-integration frame was put forward equally by all interviewees. They stated that Pieterlen was concerned with a high share of unemployed, single mothers and social welfare clients. Their children and children from families with migration backgrounds lead, according to this view, to problems when they entered school in terms of social behavior and language. Persons, whom we interviewed felt challenged by social pressure. Obviously, the best solution to this problem was seen in the establishment of a public day-care center.

Hence, at the municipal assembly a majority of the people voted for the establishment of a public day-care center to take care of children from different socio-economic strata whose parents have to work (Botschaft Gemeindeversammlung Pieterlen 2002). By a highly professionalized offer, pedagogical educated care workers should follow a pedagogical concept in order to enhance social, emotional and congitive capabilities of the children (Botschaft Gemeinderversammlung Pieterlen 2002).

The data shows that actors from the liberal party FDP put forward the establishment of a day-care center with reference to a labor-market frame. The discussion was more about the economic situation in which women have to work to gain enough income for a family (Interview statement). However, this frame was not center stage in the decision-making process. Astonishingly, the gender-equality frame was never mentioned in the political discussion. Moreover, interview statements indicate that the argument of gender-equality was not helpful for obtaining support.

From the beginning, non-parental childcare was framed according to interview statements as being mainly a solution to a social and an integration problem. Actors from the liberal party FDP supported by representatives from left parties suggested this frame. Although members of the conservative party SVP used in the beginning a male-breadwinner frame, the social-integration frame succeeded in establishing a majority coalition, which supported the institutionalization of a publicly provided day-care center.

#### *Type of childcare service*

In 2012, the care-arrangement in Pieterlen consists of a public provided daycare center. Most importantly, the municipality of Pieterlen regulated that the day-care center has to be organized and provided by the municipality with an income related fee structure. There is common sense among interviewees that the day-care center in Pieterlen follows a high quality pedagogical concept above the cantonal requirements. Therefore, this offer fits to the universal childcare model. However, because of its strong focus on pedagogical education of children and its outstanding pedagogical concept it has a bias toward a social-integration model of childcare.

#### *Outcome of the type of childcare service*

We will first present our findings on the level of the users of the services, before we proceed with the findings for the outcomes on the level of the provider. Concerning the potential of the type of childcare for gender-equality on the level of the user, the data show that the offer is used by 27.2% of the family households with children below seven years. Hence, female employment is supported to a high degree and it can thus be assessed as having a high potential for gender-equality. Table 7 presents the findings for the proportion of children from different income quintiles (social cohesion) using the service. The data show that children from all income strata use the universal childcare service equally.


*Table 7: Outcome on the level of the user of the service: vertical redistribution, Pieterlen (2011)*

*Source: Health and Social Welfare Directorate, Canton of Berne*

We now turn to the findings on the provider side. According to interview statements all six employees with tertiary education have an annual wage of more than 41'000 CHF. They have on average a workload of 73% and a salary of at least 2500 CHF supporting an independent household. As a consequence, the universal childcare service is categorized as de-gendering.

To assess the potential of the offer for social cohesion on the provider side, we investigated the socio-economic and socio-cultural background of the care workers. As it is shown in Table 8, women from low income strata and women with migration background are employed in the universal childcare service. Hence, we observe a potential for social cohesion.

*Table 8: Outcome on the level of the provider of the service: vertical redistribution, Pieterlen (2011)*


## **5 Comparison**

#### *5.1 Explanations for distinct childcare services*

As findings indicate, different frames used in the process of policy-formulation to build a winning coalition shape the variety of resulting childcare services. First, we found that two very similar municipalities, Münchenbuchsee and Pieterlen, have etsablished different types of childcare services (see Table 9). Second, the data has shown that the frames shaping the decision-makingprocesses had consequences for the chosen type of childcare service.



In Münchenbuchsee we found a majority coalition supporting a universal childcare model, which formed around a gender-equality and a labor-market frame. At the same time, the existence of a strong male-breadwinner frame enhanced the development of a traditional childcare service, however with an income related fee structure. Both developments are expected by the hypothesis. In Pieterlen, we found that a strong social-integration frame led to the establishment of a universal childcare service. This development contradicts the postulated hypothesis.

Hence, our theoretical expected relationship between a certain frame and an ideal typical model of a childcare service was only partially supported by the data. This can be caused by the fact, that empirically there does not exist one ideal typical model of childcare. As it has been shown by our data, a universal childcare model can either have a bias toward a social-integrative model or a market-liberal model of childcare. In our cases, the income related fee structure was given by cantonal regulations. Hence, even though a strong male-breadwinner frame existed in Münchenbuchsee, it did not lead to an ideal typical form of a traditional childcare service.

Our findings supported the hypothesis that winning coalitions form around different types of childcare services that allow specific combinations of different childcare policy frames. In order to build majority coalitions for certain types of childcare services, actors made use of different frames in the process of policy-formulation.

#### *5.2 Outcomes of different types of childcare services*

Different types of childcare services have different potentials for genderequality and social cohesion. All services had a high potential for the activation of women for labor markets and hence, a high potential for gender-equality on the user side. The services provided non-parental childcare from 22% (Münchenbuchsee) up to 27% (Pieterlen) of all families with children below seven years. However, the findings have shown that different types of childcare services have different potentials for gender-equality on the provision side. Whereas employees in a universal childcare offer earn enough to maintain an independent household, the wage of care workes in a traditional childcare offer is too low to support an independent household.

Interestingly, the universal childcare model present in both cases did not had the same potential for social cohesion in both municipalities. Whereas this potential was high in the case of Pieterlen, the potential was low in Münchenbuchsee. We found that the co-existence of a universal childcare service and a traditional childcare service led to social segregation in the use of those services. Whereas the universal childcare service was mainly used by middle and high-income strata, the traditional childcare service was more often used by low-income families. Most interestingly this is despite families having to pay the exact same amount for care in both offers. According to interview statements, the reason for these different patterns in the use of those services lies in the higher timely flexibility of family day care.

To summarize, our findings indicate the existence of trade-offs not only between gender-equality and social cohesion but as well between the outcomes of the two dimensions on different levels. For example, concerning gender-equality, we found that a service can enhance female labor market inclusion on the level of the user, but at the same time provide precarious working conditions for care workers on the provider side. A service can enhance gender equality both on the level of user and provider, as the case for the universal childcare service in Münchenbuchsee, but at the same time be problematic in terms of social cohesion at the level of the user.

## **6 Conclusion**

Since the beginning of the millennium, Swiss municipalities have established new social policies for childcare that ease conflicts between paid work and family life. However, the design of public policies for non-parental childcare varies not only in terms of the organization of its provision, but also in regards to quality standards. Childcare services, as a typical example of a social investment policy, can contribute to more equality in the society. However, the potential of those policies for gender equality and social cohesion is all but clear.

Against this backdrop, the interest of our study was twofold. First, we investigated how the variety of childcare services in Swiss municipalities can be explained. Second, we examined what potential the diverse types of childcare services have for gender-equality and social cohesion.

The main results showed that the two municipalities under investigation have established different types of childcare services. Münchenbuchsee has institutionalized a care-arrangement consisting of a combination of a traditional and a universal childcare service. Pieterlen has institutionalized a universal childcare service only. Our data indicates that different policy frames in the process of policy-formulation shape partly the resulting variety of childcare services. As proposed by Häusermann and Kübler (2010), framing effects emerged as important variables explaining different outputs of decision-making processes. However, cantonal regulations play an important role in how those frames translate into a specific type of childcare. Thus, future reseach should carefully retrace the process of coalition formation, and flesh out how institutions influence the relationship between frames and output.

To obtain a comprehensive understanding of the trade-offs between different types of childcare services in terms of gender-equality and social cohesion, we assessed the social outcomes on two levels, namely on the user *and* on the provider side. Most surprisingly, the data show that parents from different income classes use different types of childcare services differently, although the costs are similar for both offers due to public subsidies. Whereas parents from low-income classes use the traditional childcare service provided by families more frequently, parents with middle and high income use universal childcare services supplied by day-care centers more frequently. These offers differ sharply in their degree of professionalization in terms of requirements for pedagogical conception and educational standard for care workers. Hence, parents from low-income classes can barely profit from high quality childcare services provided by day-care centers. They use the offer with lower quality standards provided by families, which is more flexible in terms of opening hours. Moreover, the results show that the offer of childcare services supplied by families produces precarious, low-paid jobs, mostly for women. Thus, on the provider side childcare policies can have *a social rebound effect* for these female employees. On the other hand, we have shown that the offers are used by an amount of around 25% of all families with children below seven years in both municipalities and can thus contribute to gender-equality on the level of the user of those services.

In accordance with other authors we found that the way in which states respond to the childcare gaps in post-industrial societies is highly relevant for gender-equality and for social cohesion (Jenson/Sineau 2003, Morel 2007). Hence, it is argued that in the future the design of policy measures for gender-equality, namely those for enhancing the reconciliation of care and work, should be carefully crafted. Otherwise, there is a danger, that the adaptation of the welfare state to 'new' social demands of women produces suboptimal outcomes in terms of social cohesion and even gender-equality. Therefore, these effects have to be analyzed more broadly based on additional cases in varying contexts.

For the current debate on promoting a higher degree of labor-market integration of shigh-skilled women it is important to recognize the following two points: First, the social-liberal winning coalition, nowadays often responsible for an extension of non-parental childcare, mostly favors childcare policy designs oriented towards a privileged clientele. Second, depending on specific design, childcare provision can reduce social cohesion and gender-equality by providing high quality services for high-income groups and precarious employments for women.

### **References**


# **Configurating the positions of unemployed women: Professional practices in social welfare**

*Gisela Hauss* 

## **1 Introduction**

*"A family with a child: Well, I'd look at things and multiply everything times seven because the children obviously get no more than a packet of crisps, and not even then will there be enough to make ends meet."* This statement reflects the dilemma of a social worker advising a mother about to take up employment in the low-wage sector and whose dependency on social services is about to cease. *"Not even then will there be enough"*: This comment indicates the significant time burden and psychological strain suffered by the working poor balancing employment and care responsibilities; it encapsulates what it means to live in poverty after welfare dependency ends, a predicament in which even children do not have enough.

Switzerland is said to be an affluent country with a high level of employment. Not everyone, however, is able to enjoy Switzerland's high living standards. Approximately every seventh person in the country lives under precarious conditions. Women are at greater risk of poverty than men in Switzerland. On average, they are less well qualified, earn less, and often have lowly-paid part-time jobs (Federal Office of Statistics 2010). Despite pursuing gainful employment, an above-average number of women are unable to secure their livelihood. In Switzerland, as in Germany, a growing sectoral segregation of the labor market can be observed. Women are assigned to certain activities, for instance, social services, care duties, and interpersonal tasks in the services sector. Many pursue unqualified or rather highly qualified, yet lowly-valued, underesteemed activities (Nickel 2007).

Switzerland's approach to poverty is not fundamentally different from that of other Western European countries. In Switzerland, as elsewhere, fewer

#### *Gisela Hauss*

social welfare payments (direct transfers) have been since the 1990s. Labor market integration has instead become the normative target objective with the state investing in its citizens' "employability" (Dahme/Wohlfahrt 2008, Jenson 2009). The welfare state is transforming into an activating, investing, or even post-welfare state. Activation is aimed at persons not actively participating in the primary labor market (anymore). Forcing the unemployed into the labor market is embedded in the paradigm of social investment. In an investment concept driven by economic rationales, public funds are to be used to develop the productive potential of individuals. Resources are to be used purposefully and selectively, most of all where they promise the greatest possible impact. Welfare benefits are paid not only to individuals who are already productive or need to be productive, but also to those who might become productive (in future). Remedying current problems, and alleviating distress, becomes secondary (Jenson 2009: 405), in comparison to other Western European countries that are increasing their social investments, for instance, in child development, in combatting poverty, in care structures with educational aims, and in basic training, Bonoli (2010) observes that Switzerland is practicing a "light" version of social investment policy, focused on intensifying labor-market access.

This paper discusses the poverty action taken by the Swiss state welfare at the level of concrete interactions. Analysis is framed by a gender- and intersectionality perspective. Of particular interest is the positioning of unemployed women in the context of welfare measures, a field of investigation situated at the interface between several dimensions of inequality. The central question is how, at the level of the interactions occurring in welfare programs and counselling, diverse life situations (for instance, poverty, gender, and care duties) are broken open and reconstituted, consolidated or changed. Gender, as becomes apparent, is not an "omni-relevant," unambiguous category. It is instead interdependent and multiply overlaid (Heintz/Nadai 1998). This paper shows how in the context of precarity and poverty manifold inequalities are reproduced, re-established, or balanced out in the implementation of social policy measures.

## **2 Poverty, precarisation, and gender**

Current poverty research observes a dynamization of poverty, which manifests itself as a temporalisation and destructuring of poverty. Poverty, thus, is no longer evident chiefly as a lasting marginalization of the lower social strata. Instead, it largely constitutes the risk of merely temporary episodes of poverty as part of individualized life courses (for a critical view, see Groh-Samberg 2009; see also Kessel/Klein/Landhäusser 2002; Kronauer 2002, Huster/ Boeckh/Mooge-Grothjan 2008). Poverty situations are discussed in social, that is, relational terms, as multi-tiered social conditions of participation and exclusion (Castel 2008, Paugam 2008).

Following Robert Castel (2008/1995), "precarisation" denotes the divisions and boundaries also running through the sectors of gainful employment. Castel distinguishes various "social positions," which his analysis describes in terms of "zones" of "integration," "vulnerability," and "uncoupling" (Castel 2008: 13). He considers the zone of social vulnerability an "unstable intermediate zone, in which people combine a precarious work situation with fragile support from their immediate surroundings" (13). Precarious working conditions in this zone can be characterized as temporary. They lack guaranteed social security and institutionally secured chances of participation. Such situations are fixed-term, associated with permanent employment uncertainty, and thus render long-term planning impossible (Kraemer 2007, Dörre 2009). The problem of precarity has also been discussed in social welfare over the last ten years. Both the relevant studies and welfare statistics confirm that whereas a considerable number of welfare recipients are gainfully employed, their income is insufficient to secure their livelihood (Kutzner et al. 2004).

Gender research (Aulenbacher 2009) has engaged with precarious working conditions for much longer than poverty research. Its analysis of society considers the flexible and marginal employment of women since the rise of industrialization, and has criticized the androcentric bias of traditional sociology. Thus, Aulenbacher (Ibid.) points out the lack of references to gender arrangements and family situations in current research on precarisation. She argues that social integration, vulnerability, and uncoupling depend on "how and through which gendered scopes of action people are integrated into various areas of society and forms of work" (2009: 77).

Winkler (2010) observes a narrowing in the current debate on precarisation. First, precarity is delimited from normal (Fordist) working conditions, which are implicitly male. Second, discussion focuses on gainful employment and blinds out the sphere of reproduction. Third, categories such as age, gender, qualifications, and family circumstances are excluded from structural analysis (167–169). Winkler further points out that categories such as ethnicity, nationality, and religion complicate access to the labor market, precisely because they are interlinked with gender, and thus promote structural tendencies toward precarisation. Also, a discriminatory residency status, unaccepted prescriptions regarding cultural and religious behavior, and lacking access to the labor and housing markets can precipitate or worsen precarious situations (169). Seen thus, the zones distinguished by Castel refer to multiple spheres of inequality, in which actors are differently positioned. In "complex social locations" (Landry 2006, XVI), or in "contradictory locations" (Anthias 2005: 37), superiority and subordination emerge from within the context of simultaneously effective structural inequalities (Kubisch 2008: 48).

Welfare employment programs involve persons reliant on public assistance for very different reasons. The paths on which they reach such programs and leave them again are manifold. Common to all, however, is their status as welfare recipients. As sociologist Georg Simmel has argued, from the moment people receive support they enter "a circle characterized by poverty" (Simmel 1995: 553). In these terms, this circle, as a site of welfare, is located outside society. And yet, "this outside is ‒ briefly put ‒ only a particular form of the inside" (546). According to Simmel, the "welfare relationship," which protects the status quo of society, the locations of poverty are tied back to society. Thus, the configurations of the reciprocal relations between the poor and the rest of society move to the centre of attention. Of sociological interest in this regard are the socially institutionalized forms of treating poverty (Simmel 1995). The "welfare relationship" lends contours to the social representation of poverty and influences the subjective experience of those affected by poverty (Paugam 2008). Focusing on the "welfare relationship," this paper starts from gender theory and intersectionality to consider the governmental action taken within the context of precarity. It addresses the interrelations between different categories and establishes gender as an important structural category.

## **3 Research perspective**

Empirical research conducted from an intersectional perspective is a desideratum. Discussing the US-American and German-speaking debates on intersectionality respectively, McCall (2005) and Knapp (2005) note that whereas the concept is understood as a paradigm, in actual fact only scant discussion exists about how intersectional analyses might be conceived and carried out (Kubisch 2008: 326). Helma Lutz and Sonja Kubisch have presented the first attempts at intersectional analysis. Lutz (2001) performed a secondary analysis of a Dutch vocational school class and revealed the lines of difference relevant in the group discussions investigated. Kubisch further advanced intersectional analysis by elaborating forms of stability in the construction of social difference (habitual construction of social difference) that exceed situational "doing difference" (Kubisch 2008: 312).

The research perspective underpinning this paper lays no claim to a multiple-level analysis in terms of the intersectionality paradigm (Degele/

Winkler 2007). Instead, it is guided by the category of "interdependence," a figure of thought introduced by Walgenbach (2007). This category transcends not only the notion of intersecting constant axes, but also metaphors of interlacing, interweaving, and criss-crossing (Crenshaw 1995: 358, 359, Klinger 2003, Klinger/Knapp 2005; for a critical viewpoint, see Walgenbach 2007, 2012). Whereas conceptions of interferences tend to assume that the category of gender has a "genuine core," which "coalesces," "conjoins," or "interlocks" with other categories, and only changes thereby, the concept of "interdependent categories" conceives gender as integrally related with other categories. Walgenbach assumes that it is barely possible to analyze gender without referring to other categories. This connection, however, does not define what happens between two categories, but rather the reciprocally dependent categories themselves. These categories are considered to possess an *inherently* heterogeneous structure (Walgenbach 2007, 2012). Similarly oriented are the model of the configuration of power relations (Gutiérrez Rodríguez 1996: 170) and the model of "different, simultaneously effective relations of superiority and subordination, discrimination and dominance," proposed by Birgit Rommensbacher already in the 1990s (1995). These models keep analysis open for fluid forms of the superimposition of social differences and inequalities. Also, they understand categories as inherently interdependent and thus depart from the notion that the category of gender has a "genuine core" (Walgenbach 2012: 20).

Based on these figures of thought, this paper analyzes the interactive production of configurations in the field of social welfare or, in Simmel's terms, of the "welfare relationship." In social welfare as an institution, the establishment of differences causes the unequal distribution of resources. This, in turn, affects the opportunities in life of welfare recipients, who are perceived and positioned differently in relation to one another (Schwinn 2007: 273). The present study considers this "institutional substructure" (272). It focuses on the relations between categorizations and the allocation of tangible and intangible resources such as counselling time and the calculation of benefits, which, in turn, affects access to education, employment, and income. Analysis follows Nadai's suggestion (2014: 31, 32) to employ Sacks's membership categorization analyses (MCA) in the field of social security (Sacks 1972, cited in Nadai 2014: 31). Accordingly, categories are grouped into "collections." As normative attributions, these collections include entire clusters of attributions, behavioral expectations, and indications of the selection of appropriate measures (Leppert 2000: 34, cited in Nadai 2014: 32). Analysis is aimed at interactive processes related reciprocally to the respective context (Hestling/ Eglin 1997: 27, cited in Nadai, Ibid.), and in which attention focuses on how unemployed persons present themselves (33, 34).

#### *Gisela Hauss*

This paper adopts an ethnographic approach. It analyzes everyday practice in three (social welfare) labor-market integration programs. Assuming that "frontline workers" are important actors in producing social structures, the present study focuses on "street level bureaucracies" (Lipsky 1980). In terms of "institutional ethnography" (Smith 2005), analysis concentrates on "the world of everyday experience" (Smith 2005: 43). In "everyday worlds," as sociologist Dorothy Smith observes (32), people engage in everyday interaction. On this level, social policy developments become observable in "issues, concerns, or problems that are real for people and that are situated in their relationship to an institutional order" (Ibid.). Here, in the microsociology of social institutions, the "ruling relations," which far transcend the investigated field, can be studied (Ibid.). Smith assumes that social orders should not be understood in abstract terms. Instead, they constitute themselves in the concrete relations produced through coordinated action. Consequently, social orders can be determined by describing concrete actions in a specific context.

Based on these methodological considerations, analysis below reconstructs the everyday interactions, techniques and strategies, and the structures for action and interpretation, in three different (social welfare) labor-market integration programs. Analyzing the categorization processes at work in these programs points to the prevailing social order and reveals the influential implications for the allocation of resources and measures. The results presented here are based on the analysis of observation protocols, on interviews with program staff, experts, and clients, and on internal and cross-organizational documents. A total of 54 observational units and 46 interviews were conducted. Data collection and data evaluation were based on Grounded Theory (Strauss and Corbin 1990). The situational and contextualized categorization processes were, as described, analyzed from the perspective of "membership categorization analyses" (Sacks).

## **4 Configurations of inequality: "Doing difference" in labor-market integration programs**

This paper explores three (social welfare) labor-market integration programs.<sup>1</sup> Each program is part of the "social security regime" (Knuth 2009) and tasked with carrying out labor-market measures. The first program, *Artigiana*, is a labor-market integration program for women; the second, *Inizia*, is aimed at young, unqualified mothers; the third, *Viadukt*, is gender-integrated

<sup>1</sup> For a discussion of a social services unit see Hauss (2014c)

and places clients in temporary jobs in the primary labor market. <sup>2</sup> All three programs are situated within the sphere of inequality defined by poverty and precarity. Of less interest here are what the programs offer officially, but instead how they address their clients and thus assign them to categories of inequality.

#### *4.1 Configuration: Single-parenting female migrant with impaired health*

Artigiana employs about 20 women in a workshop, in a shop selling the products manufactured in the workshop, and in a refectory. Usually, the women spend six to twelve months on the program, with extended stays possible. Most of the women assigned by social services to Artiginia are single mothers obliged to reintegrate themselves into the labor market.

The program comprises no formal education, but offers ad-hoc workplace training. It also provides language training, work-related instruction, and job application assistance. Artigiana is highly flexible: It allows participants to structure their working hours. The minimum workload is 20%, which can be increased incrementally. Whereas the group is homogenous by gender, participants' ages vary. Even though most participants have children, motherhood is not a formal entry requirement. In principle, childless women are also accepted. One staff member described the typical Artigiana participant as follows: *"She's a migrant, she's got children, and she's a single mother. She doesn't speak perfect German. And she may well suffer from mental and physical impairments."* Another staff member assumes that participants are *"single parents and need more support because they have multiple deprivation."*

The configuration "Single-parenting female migrant with impaired health" is associated with ambivalent attributions. First, it problematizes the reconcilability of reproduction and production. Second, it formulates an emancipatory impetus: Besides creating employability (that is, the capacity to pursue gainful employment), it creates scope for self-perception, for critically reflecting on one's self-concept, and for enhancing one's autonomy.

Addressing the issue of reconcilability within the context of migration makes the programme a place where traditional gender and generational roles are meant to be prised open or at least expanded.

The norm established for the socialization of women is work, which, by implication, affects the image (and assessment) of carework and housekeeping (Becker-Schmid 2003: 13). In the eyes of Artigiana staff, care work and housekeeping do not serve the integration of female welfare recipients into

<sup>2</sup> For an extended discussion, see (2014a, b/2015) .

#### *Gisela Hauss*

society. Instead, staff members assume that the dependencies and allegiances resulting from childcare makes mothers and children less independent and hence impedes socialization. Thus, the time spent by mothers at home is considered a phase in which, despite being housewives and childminders, they have in actual fact done nothing, nor been *"actively involved in the working world."* Caregiving and housekeeping are described as *"the being-at-home rut." "Their daily routine revolves round the child,"* as one staff member says, adding *"and that must slowly change."*

Poverty and migration make the lacking recognition of care work even more evident. On the scale "modern versus backward," ethnic attributions identify migrant women as backward, as embedded in the traditional models of work and family life prevailing in the authoritarian structures of patriarchal societies (Leiprecht/Lutz 2003, 2010). One staff member reports: *"For most of these women, life in Switzerland is quite different to back home. Here, we have small families, so they must learn to put their children in day care. They feel much more guilty than Swiss women about giving their children away. Especially if they come from Turkey or from regions with strong family structures."* Associated with this perspective is that these women are expected to change their behavior, to free themselves to some extent, and to rehearse new roles in the gender relationship. Another staff member explained how important taking up employment is for *"detaching oneself a bit from home."* The possibility of extending program participation accounts for this process taking a lot of time. Most women, says one staff member, need more than half a year to re-organize themselves, their *"children, household, and work."* But this, she adds, is an important prerequisite for making the women *"fit for the primary labor market."*

Artigiana staff allege that women with a migration background delay the process of individualization for ethno-specific reasons. Having grown up in authoritarian structures, these women have never had the opportunity to consider their interests and to discover what they are good at. These women appear to have advanced little down the path toward the emancipation demanded by second-wave feminism as a modern form of socialization.

Based on this notion of an ethno-specific delay of individualization, Artigiana has an emancipatory design. One staff member describes that the overarching objective of the program is not to expand skills, but *"horizons and skills."* Staff emphasize the need to create an encouraging learning atmosphere, in which participants are supposed to feel at ease. They are encouraged to try out new things and offered the necessary supervision. They are helped to take small steps toward self-managing their work. Artigiana aspires to assign the women tasks enabling them to discover and develop their own potential. Thus, they are given gender-atypical tasks. Programme staff observe that successfully performing manual tasks, operating machines, or hanging up lamps gives the women self-confidence and a certain independence in their everyday domestic life. Staff see the women "blossom" and begin to "glow" no sooner have they entered the programme.

The normative attributions associated with the configuration "Single-parenting female migrant with impaired health" are ambivalent. The figure of the woman with manifold dependencies and somewhat limited autonomy and reflective skills is the starting point for the compensatory support and supervision provided to enable women thus challenged to manage the biographical transition from care work and housekeeping to employment. The program offers migrant women the opportunity to "make a detour on their way into the labor market (Hauss/Nadai 2009). At the same time, the configuration of social inequalities is reinforced. These women are attributed otherness. Based on this attribution, they are adjusted to the norms of gender order and the division of labor asserting themselves in social policy. The program assumes the function of retraining full-time mothers as women capable of organizing themselves between employment and care duties so that they can somehow cope with the labor structures prevailing in the low-wage sector. The program thus promotes the twofold integration of women into the social fabric (Becker-Schmidt 2003). They continue to be responsible for unpaid housework and caregiving. These tasks, however, are devalued and supposed to be reduced to a minimum. Consequently, the pressure to participate in working life remains dominant in these women's lives. At the very moment they complete the programme to enter the labor market, discriminations based on their social situation catch up with them: Although this dimension of inequality inheres in the programme, it is barely made explicit.

#### *4.2 Configuration: Young women who get pregnant too early*

Inizia is a welfare program aimed at young, unqualified mothers aged between 16 and 25. In the year studied, there were twenty participants. The programme accepts women who have at least completed compulsory schooling and have a good knowledge of German. External childcare is provided for the duration of the programme. Most participants are single parents who are not involved in a steady relationship. Inizia aims to guide participants toward an apprenticeship. Its spatial-temporal structures are highly standardized and oriented toward a model of classroom learning consisting of five half days per week. The curriculum includes mathematics, career selection and advice, job application coaching, and communication skills. Other modules deal with child education, coping with personal and psychological obstacles, and sexual health. In a certain sense, the heavily regimented structures are subverted or made flexible through lively staff-participants exchanges conducted via e-mail, text messages, and Facebook.

The configuration "Young women who get pregnant too early" resonates with specific normative attributions. "Early pregnancy" refers to the deferral of a time pattern considered ideal for the life-course, namely, first establishing oneself as a stable wage earner and then assuming a secure role within a family. So whereas these women have not yet made the transition to economic self-sufficiency, they have already assumed the social role (and status) of a mother (Hurrelmann/Quenzel 2012).

Inizia participants are too young to be mothers, and too old to enter an apprenticeship. The configuration addressed by the programme refers to "anachronistic transitions" to adult status (Hauss 2014a ); as such, it is "irritating" (Nadai 2014). For Inizia, the young age of its participants is an action-guiding category. In terms of the social investment paradigm, resources invested in young persons and their training promise a high economic return. Thus, addressing these women as unemployed, poorly qualified young females strongly shapes the program and its everyday circumstances. Associated therewith is a series of attributions subsumed by program staff under the term *"multiple deprivation."* Staff speak of lacking occupational training, educational deficits, difficult family situations, fragmentary social networks, psychological burdens, traumatizing experiences, or *"migration background."*  The participants' young age is associated with behavioral attributions, for instance, that the young women are *"unstable."* When it comes to searching for employment, participants must be self-motivated, or else *"things simply topple over"* Staff never know whether things will work out because *"something can crop up (...) things are very fragile."* There are modules aimed specifically at young women, addressing participants' self-concept. Participants are encouraged to explore their inner psychic world, health, and sexuality. Painting therapy, for instance, helps them to express their feelings and wishes. Programme sessions are devoted to developing (stable) conceptions of self, identity, and self-esteem. Participants also work on *"stumbling blocks,"* that is, events and experiences in their past that thwart their search for employment.

Motherhood, which diverges from a normal life course, is assessed ambivalently, and time and again moves into the background in the context of youth-specific attributions. Nevertheless, having and raising a child is considered an "achievement" and a step into adulthood. For the women, the child represents a *"caesura:"* They often realize that *"things can't go on like this"* and begin taking responsibility.

At the same time, the configuration of "the young woman who got pregnant too early" implies a behavioral expectation: namely, that these women are too young and unable to raise their children in the best possible way. This expectation is action-relevant: Once a week, Inizia offers a morning infanteducation session, which the young mothers attend with their children. Sessions address the too strong dependency and emotional ties between mother and child. The young mothers are encouraged to leave their children in day care and to disentangle themselves from dependency, so as not to jeopardize their apprenticeship. The emotional deficits potentially arising from deferring important psychosocial needs, such as spending time with one's child, are not considered by the programme. These women have the opportunity *"to stand their ground in the working world,"* but they must bear the burden of doing so on their own.

The configuration of the young mother within the context of poverty and precarity is a new and inconsistent phenomenon within the social investment paradigm adopted in Swiss welfare. The interdependent category does not seem grounded in long-standing, proven practice, but is instead constituted under new conditions as needed. The social investment paradigm concretises itself in the construction of young women as young adults, whose education and employability pay dividends. Their parenting skills are developed as far as necessary, whereby particular emphasis is placed on detachment. The configuration of the young poverty-stricken mother provides detours for rehearsing reflective skills, for working on one's self-concept, and for taking intermediate steps toward detachment. The objective is a new independence, which is supposed to enable the transition to employment. Ultimately, however, the women are responsible for organising their lives between care duties, housekeeping, occupational training, and employment. They must cope with this task in poverty and, as Castel puts it, in "zones of vulnerability."

#### *4.3 Configuration: Refugee status and far removed from the Swiss labor market*

Viaduct is a welfare program closely linked to Jobcast, an unemployment program. Program participants (sixty persons in 2011) include refugees from war and crisis zones who are in Switzerland for the first time as well as delinquent young adults who have not yet gained a foothold in the Swiss labor market. Viaduct staff seek to place participants in the primary labor market and accompany the integration process by offering counselling sessions. Internships are conceived as unpaid work placements. Jobless persons in such situations are not meant to attend an integration program merely as a "*standard procedure*." To enter the labor market, they must first "*sort things out*." Thus, an envisaged six-month work placement may be extended several times. Employment measures take place in the labor market. In line with the concept of "detours into the labor market" (Hauss/Nadai 2009), the concept underlying the Viaduct program can be described as "a detour *through* the labor market." Integration efforts, then, take place not in a special program, but in a pool of jobs located especially (but not exclusively) in the non-profit sector. Viaduct is a gender-integrated program.

At the time of this study, most Viaduct participants were refugees. This was the result of a cantonal remit of 2008 to provide integration opportunities no longer merely for recognized refugees, but also for refugees or persons granted temporary admission to Switzerland. Refugee social services programs can delegate this remit to social welfare programs. Persons who have recently entered Switzerland on humanitarian grounds are a discriminated group, whose employment rate is lower compared to that of other migrants (Liebig et al. 2012). Structural prerequisites complicate access to the labor market (lacking integration programs, requirement-heavy granting of work permits, restrictive naturalization policy, inadequate language development measures, complicated recognition of foreign qualifications, few transitional opportunities, employment discrimination) (see Ibid., pp. 10–13). Of interest below are the categorization processes at work among persons who have recently entered Switzerland for humanitarian reasons. The same question applies as raised above about the other programs: To what extent is gender blanked out as a category or does it become relevant?

Viaduct staff categorize participants in terms of the specific values and attributes assigned to the domestic labor market, toward which migrants are positioned as alien. In the entrance-, appraisal-, and exit interviews studied, staff construe the Swiss labor market as supposedly homogeneous and as inwardly cohesive, and has been so since time immemorial. Following Elias und Scotson (1993), the figuration of inner cohesion is used by "the established" to produce superiority, and thereby to demarcate themselves from migrants. What forms is a "we-image" and a "we-ideal" (Elias 1986: 44), in terms of which new arrivals are "outsiders," who do not form a coherent entity and are assigned the status of "*those alien to work*." The more often the applicable rules are repeated to represent the Swiss labor market as coherent, the more evidently marked the outsider position of migrants becomes. Thus, program counsellors emphasize the connection between punctuality and reliability. One staff member listed the points to be observed by a trainee who had already taken all his days off: *"Always be punctual, no missing hours, notify employer in case of illness, no days off, no travelling at short notice, no holidays, fully completed further training."* 

Second, Viaduct staff emphasize that trainees should both perceive and perform their work-related duties and responsibilities loyally and independently. One operations manager described this aspect as follows: It is typical in Switzerland to see the work that needs doing, and then to do it. Employees think of the company and enjoy working. She points to the end of the table. If any rubbish were lying there, she remarks, someone would remove it without waiting to be ordered to do so. In Switzerland, people also like to work even if their boss is not watching. That is the ways things are in Switzerland, she says. Motivated work, in the eyes of this social worker, is a sign of loyalty toward the company. Sitting around before work begins, and waiting for work to begin officially, although there is a lot to do, is interpreted as working to rule, as a lack of motivation. This contrasts with another statement made by this social worker: *"I'd you to prove that you're a responsible coworker, that you're loyal toward the company."* 

Third, Viaduct staff describe the Swiss labor market as a place where employees are supposed to communicate. Thus, the manager of a nursing home figuratively described a trainee cleaner as a *"box full of surprises."* To all outward appearances, everything seemed fine. The man was friendly, quiet, unobtrusive, and never complained. But she had no idea how he actually felt, whether he enjoyed his work. Concretizing the situation, the social worker informed the trainee that although he was not expected to wear his heart on his sleeve, he could by all means tell his employer what he liked or disliked, and that he could address any problems and questions to his superior. The *"silent man"* was encouraged to open up, to come out of his shell. He was not being asked to speak volumes, but to speak from time to time, so that he would not be forgotten. The attribution of foreignness in the face of a labor market characterized as communicative is reinforced by the lacking language skills of recently arrived migrants.

In the Viaduct program, participants' refugee status at first seems to be gender-neutral. Its task, to place jobless persons in the labor market, makes the discursively heightened distance to the labor market a key issue in counselling sessions and work placements. The fact that the interdependence between refugee status on the one hand, and gender, ethnicity, and age on the other, can complicate labor-market access in structural terms, is not addressed explicitly in the program (for a critical view, see Winkler 2010). Gender becomes relevant in particular when women's alienness toward the labor market is doubled as it were. The work done by men in their countries or origin is barely recognized as a resource. Experiences in the home country are referred to only vaguely or with reservations in entrance and appraisal interviews: *"They all say they are mechanics,"* says one staff member, but obviously *"they don't have a clue."* Another comments that even after the interview he did not know whether the participant had worked as a *"tiler"* back home. The configuration "female refugee" means that women are positioned entirely outside the labor market and economic independence. In an exit interview with a young woman after a work placement that had left her without a follow-up and without further prospects, the responsible social worker confirmed that Mrs T. had never worked before, and that at least now she had work experience. The staff member's use of the category "*no previous work experience*" blinds out possible work experience in the country of origin and takes for granted that this program participant had so far been active solely inside the family. On the level of action, this conceptualization corresponds with the non-recognition of diplomas and higher professional qualifications earned in the home country.

## **5 Conclusion: Women in social welfare. Configurations of inequality**

Two of the programmes investigated here fly under the banner of the advancement of women, specifically of enhancing the qualification opportunities of their female participants. They are aimed at women's emancipation, that is, at young single mothers attaining professional qualifications. These target objectives create the space for learning processes and personality development for the duration of the programme. At the same time, they presuppose that the women's development toward emancipation and their transition to adulthood are categorized *a priori* as "incomplete." As immigrant women, their individualization suffers from shortfalls. As young women lacking any initial training, they have not yet reached adulthood, because they have not yet completed the partial transition to economic independence.

This configuration of single mothers is the starting point for a large number of so-called remedial programs. Using creative methods from social work, adult education and the women's movement, these programs strive to promote personality development, self-assurance, and interpersonal skills. Depending on the programme, the primary objective is to enable participants to become emancipated women or to take a step hitherto missing into adulthood. The programs see themselves as a moratorium, as an authorized period of delay, and allow detours. But the envisaged objective ‒ labor-market integration ‒ remains the same. For staff members, the creative learning processes offered by the respective program function in particular as a prerequisite for their clients' employability on the primary labor market. Ultimately, pedagogical concepts, and those promoting the advancement of women, are geared toward the economic objective of removing participants off the welfare roll as fast as possible. Although the aim to support women in gaining autonomy and educational opportunities is not abandoned, in the context of such programs it is declared an intermediate step and made to serve the overriding aim of integration into professional life.

Viaduct transfers the function of integration onto the labor market. Employment is considered a *"step into the outside world,"* into society. Whereas this may happen in single cases, in most of the situations observed in this study although participants filled their places as members of the labor force, they were attributed as "alien to work" and excluded from the social contexts in the country of immigration.

None of the three programs investigated acknowledges care work as work, as a type of work capable of effecting social integration. The skills gained through care work are not assessed as productive in terms of employability. Managing life with a child is considered an achievement, but the labor market considers the emotional dependencies and allegiances resulting from childcare an obstacle. This impediment needs to be overcome. For this reason, labor-market integration programs for women strive to educate participants to distance themselves to a certain extent from their caregiving duties. For instance, they must learn to place their children in day care facilities. Thus conceived, the required "detachment" from social welfare presupposes the mother's "detachment" for her child. This marks an interesting reversal of the youth-specific issue of the need to separate from one's parents. As Daly (2011) suggests, the aim is to overcome "old dependencies," or to have already surmounted them, so as to meet the prevailing labor-market constraints on flexibility as far as possible.

Here, significant differences to conceptions of the role of the mother in the social "zone of integration" (Castel 2000) become evident. In the field of precarity, the strong emphasis placed on a constant mother–child relationship in stable employer-employee relations becomes a mother-child relationship that jeopardizes not only the mother's independence but also the child's.

Viaduct addresses women with a refugee status in particular in their role as persons placed in the primary labor market. Labor integration is conceived in gender-neutral terms. Women are responsible for caregiving and housekeeping themselves. The reconcilability of care work and employment is only of marginal interest in the program.

The starting point for the measures adopted by women-only labor-market programs is the figure of the unemployed single mother. Perceptions of the interdependence between categories in this configuration vary from programme to programme. Inizia perceives gender and motherhood in particular as interdependent with participants' young age whereas Artigiana conceives an interdependence with participants' migration background and health impairments. These different configurations are associated with powerful expectations about behavior and their implications for action in the respective programs. In social welfare, notably, the configuration "mother who got pregnant early" is perceived in contradictory ways and associated with ambivalences. The interdependence between participants' young age and motherhood is rarely addressed at the same time in the program. Gutiérrez Rodriguez (1996) assumes that the category "woman" cannot be perceived on its own, but must instead be understood as a differentiated category, such as "female worker, peasant, migrant." Thus, "each gender configuration has its own historical and social specificity" (170). The specificity of young women in social welfare seems fragile, just as the history of how the state deals with young impoverished women is ambivalent. Again and again, concrete action blinds out this theoretically analyzable interdependence.

Strikingly, poverty is a dimension largely absent from client categorizations. Program participants are welfare recipients and thus live on the margins of subsistence. Many have debts or must be very careful with money. Their low-income backgrounds heighten the discrimination suffered by these women, in particular if they are single parents and do not yet understand the language of their country of immigration.

In most cases, they have access only to low-paying employment. The function of the programs is to accommodate these women to the conditions of badly-paid jobs barely providing enough to secure a livelihood.

Whereas women-only programs enable their participants to take intermediate steps and accompany their transitions, mixed programs leave this task to the labor market. The programs invest in women's employability, in a social stratum in which job seeking and the division between care and employment leaves little room for alternatives and choices.

In the "complex social locations" (Anthias 2005: 37) in which unemployed women are positioned in the programs, the dimensions of inequality are overlaid one upon the other. Precarisation processes and the assertion of a social-investment paradigm intensify these dynamics and contradictions. Thus, institutional categorizations along the lines of motherhood, age, ethnicity, type of residence permit, and language skills produce or worsen the unequal assessment of tangible and intangible resources within the programs. In the three programs studied here, gender is not an "omni-relevant" and unequivocal category (Heinz/Nadai 1998). Instead, it stands within diverse contextualizations; and, as Gutiérrez Rodriguez (1996) has observed, the category of gender becomes relevant in different configurations. Thereby, the figure of heterogeneously conceived interdependent categories (Walgenbach 2007/2012) becomes apparent not merely in its reciprocal interdependence, but beyond that in contradictory and ambivalent contexts, which, moreover, are not addressed at the same time.

### **References**


*Kategorie: Geschlecht. Empirische Analysen und feministische Theorien.* Opladen, pp. 163‒190.


Translation: Mark Kyburz

# **4 Gender equality in education, work and family**

## **Wage discrimination at career entry in Switzerland: Reasons and implications**

*Kathrin Bertschy*

## **1 Introduction**

In recent decades, it has been apparent that the educational and occupational pathways of women and men have been converging. Women, in particular, have benefited from the expansion of education and, with respect to the achievement of general education certificates, have advanced to the same level men (Blossfeld 1985; Hradil 1999; Haeberlin et al. 2004, 32; SKBF 2010, 113). While the educational attainment gap is closing, the gender gap in the workplace stubbornly persists: with all else remaining equal, it will take another 81 years to close the workplace gender gap completely, as concluded by the newest Global Gender Gap Report (WEF 2014). The higher educational participation of women is neither represented in higher professional positions nor in wages corresponding to their qualifications (WEF 2014, Eccles 2005). In OECD (Organisation for Economic Cooperation and Development) countries, women earn 16% less on average than men in similar full-time jobs. In Switzerland, this difference averages as high as 19% (OECD 2012). Gender based wage discrimination in Switzerland is well documented for the entire workforce. According to previous analyses, "objective" factors, such as personal qualifications, seniority, work experience or (occupational) activity account for 10‒12% of the wage differential between men and women. The other 8‒10% remain statistically inexplicable (Sousa-Poza 2004; Dini 2010; Strub and Stocker 2010).

Economic theories (Altonji and Blank 1999) focused mainly on productivity differences related to human-capital accumulation as the main source of gender gaps in wages. Given that young women now match or surpass men in terms of educational achievements, gender differences in early career wages should no longer occur. The previously mentioned studies based on cross-section data are not able to explain at which time in work history inexplicable wage differences arise or occur. Whether these are a result of different careers (e. g. productivity) that might not be recognizable in the data, or if they already exist at the point of entry into the labor market and thus are not a result of differences in productivity, but rather of discrimination, remains unexplained.

The Swiss longitudinal TREE Data (Transition from Education to Employment) presents a database for Switzerland that contains more detailed information on individual abilities, activities, and preferences and facilitated an analysis of gender wage differences that exist at career entry. Based on these data, this article addresses the following questions: Do women face lower wages at career entry? Are salaries for women and men influenced by formal qualifications or value orientation, and if yes, how? What is the impact of the distinct gender segregation in education and in the workforce? Does income represent a factor for the subsequent division of work within couple households?

The article is divided into three sections. The first part discusses the driving forces for income based on the existing literature (educational system, preferences, and economic theories). The second section explains the TREE data set and the empirical method applied. In the last part, the main results and conclusions are presented.

## **2 Previous literature**

#### *Occupational preferences, education system, and gendered school-to-work transitions*

The factors driving gender differences in the labor market are usually broadly categorized into three forces, which might be interconnected: preferences, productivity, and discrimination. Starting with preferences for career opportunities, many authors mention high gender segregation in the labor market as one of the main factors contributing to wage differences in Switzerland. In countries like France or the U.S., general education programs at the upper secondary level are common, while in German-speaking countries, vocational training at the upper secondary level is most frequent. Education focused on developing practical, technical, or occupational skills and institutional opportunities for occupational reorientation are rare. According to OECD (2009, Table c1.4) data, the vocational education rate at the upper secondary level is 71% in Austria, 65% in Switzerland and 57% in Germany. In contrast, this share is significantly lower in France and Spain (43‒44%). The significance of the Swiss education system in the reproduction of gender segregation might correlate to the central role of vocational programs at the upper secondary level, which pressures two-thirds of all school graduates into an institutionally narrow career path during adolescence. Young woman choose within a smaller field of occupations than men and are strongly inspired by gender-specific occupational fields (Leemann and Keck 2005). Since girls limit themselves to typical female apprenticeships, these "preferences" lead to jobs with lower salaries in the workplace (Palamidis and Schwarze 1989:121; Granato and Schittenhelm 2001; Imdorf 2005: 263).

During the transition from education to the labor market, this concentration further increases (Buchmann and Sacchi 1998; Müller and Shavit 1998) as a consequence of men abandoning professions dominated by women (Leemann and Keck 2005: 146). Gender specific (horizontal) segregation is particularly high for Switzerland in an international context and markedly so in male-dominated professions, especially in technical, manual professions in the manufacturing area (Charles 2005: 32ff).

Apart from the preferences in choosing a specific education – which is highly guided by the perception of jobs as male or female – the educational system is another important factor contributing to career segregation in Switzerland.

Imdorf et al. (2014) showed that in regions (cantons) of Switzerland with higher rates of vocational education at the upper secondary level (which ranged between 50 and 80%); gendered job transitions for men (but not for woman) are more likely. They concluded that the strong linkage between (gender-typical) educational experiences and (gender-segregated) jobs in the Swiss labor market correlates to rare opportunities for switching from a gendered program to a non-typical (e. g., mixed) one in the case of interest and/ or need.

In conclusion, preferences for apprenticeship, occupation, and participation in society play an important role in determining paths for education and career. However, it is important to recognize that these "preferences" are not the result of free choice, but, rather, are guided and/or controlled by mechanisms (values influenced by family, role models, childhood and adolescence, society, etc.) and are therefore, not self-chosen.

Preferences and productivity are linked in economic theories. Economical explanatory approaches focus mainly on productivity differences related to human-capital (education and qualification) to explain gender differences in wages. According to human capital theory (Becker 1964; Mincer 1974), differences in earnings can be explained through the difference in higher accumulation of human capital. Higher wages for men should thus be attributed to the higher qualification of these men. Reduced investment by women in their qualifications or human capital compared to their male counterparts is ex-

#### *Kathrin Bertschy*

plained with the theory that women anticipate future familial obligations and possible interruptions to their career timelines, making long-term training appear less fruitful to them (Henneberger and Sousa-Poza 1999). According to this explanation, when making choices in the labor market, women prefer possibilities that allow them to combine both work and family life (cf. Becker et al. 1995) while men tend to be more focused on continuity in work experience and invest proportionately more in their qualifications. The gender-specific segregation of the job market, as well as related wage differentials, can be explained, according to this theory, as a result of the different economic preferences of women and men.

Economic productivity theories might explain differences in wages and working hours, but they do not provide an explanation for wage gaps, which imply that women of similar qualification and productivity are not hired and employed under the same conditions as men. Given that young women now match or surpass the educational achievements of men, gender differences in early career wages should not occur. How, then, can such differences be explained?

In contrast, discrimination theory (Becker 1971) explains wage discrimination as a result of negative preferences and prejudices against female employment. According to the discrimination model, gender earning differentials may be attributed to direct discrimination against women by employers, employees, and customers. Employers with a "taste for discrimination" against women will hire fewer than the profit-maximizing number of women. Furthermore, the model predicts that men are paid above, and women below, their marginal product (Arrow 1973). This implies that discriminating employers earn lower profits than non-discriminators. However, in a competitive market, discrimination is costly and restricts the employer's scale and profitability. Hence, Becker (1971), Arrow (1973), among others, argue that under strong product market competition, firms may not be able to afford discrimination and will, therefore, behave in a more egalitarian fashion.

Theories of statistical discrimination (Phelps 1972; Arrow 1973; Spence 1973) utilize stereotypes, such as when employers base decisions on the average characteristics of the group to which employees belong. Wages are not related to productivity when evaluating employees, but instead on groupspecific and cultural characteristics to estimate the potential for work interruption. Gender may provide information on job commitment since women, *on average*, have higher turnover rates. This theory leads to general discrimination against women, insofar as discrimination against all women is implemented in anticipation of future work interruptions (Phelps 1972).

A widespread view concerning financial decision-making is that women are more risk-averse than men: It is argued, that women, for example, have lower preferences for variable pay systems – although this would pay off given their skills – and prefer fixed pay systems with lower salaries instead. A consequence of this stereotype is statistical discrimination, which diminishes the success of women in labor market. Whether or not risk-aversion differences between the sexes is a reality remains highly controversial in the literature. According to Schubert et al. (1999), under controlled economic conditions, women do not generally make less risky financial choices than men do. Le et al. (2011) showed that differences in risk aversion have an effect on wage differential, but explain only a small share ‒ 12% ‒ of wage differential.

Salary, which represents only one factor for the existing gender gap, is determined inter alia by the preferences of employer and employees, productivity and/or risk aversion regarding fixed versus variable wage. Just as those determinants affect the wage level (hourly wage), salary can play an important role in perpetuating gender inequalities.

Earnings can determine further gender inequalities, such as the distribution of labor between the sexes. Relative resource theories predict that an individual's working hours are affected by how their earnings compare to their partner's. Lundberg & Polak's (1993) separate-spheres-approach says that husband-wife households bargain over resource distribution, with each spouse's bargaining power determined by his or her market income. Becker's (1995) theory of specialization, or household production model, says that distribution of labor and family work among couples is affected by the hourly wages of both partners. According to Becker's (1991) theory of marital specialization, a couple can efficiently resolve the issue of household and work organization by having one partner specialize in paid work and the other focus on unpaid work. In the past, biological factors and women's disadvantages in the labor market, combined with less investment in education, meant that it was invariably efficient for women to be the ones to specialize in unpaid work. However, female education now outstrips male education. Thus, women have an increasing potential to earn more than their partners. Relative resource theories predict that if a female takes on the breadwinner role, her male partner would then specialise in household work and work fewer hours of paid work, a view that expresses a variant of the theory of specialization. Earnings determine gender roles in the labor market and at home, suggesting men will work fewer hours if their partner is the main or an equal earner (see Kanji 2013).

If discriminatory wage differences exist in an early stage of professional life and before starting a family, wage discrimination, together with the tax and benefit system for working parents (OECD 2012), reinforces gender inequalities in the labor market. Thus, wage inequalities have further implications on the persistence of gender inequalities that not only determine the next generation's preferences in educational pathways, but also in choosing careers and the resulting future earning potential.

## **3 Data, descriptive statistic, method**

#### *Data*

We used data from the Swiss Youth Panel and TREE (Transitions from Education to Employment http://tree.unibas.ch) to analyze career entry wages, wage discrimination, and salary increases within the first years on the job for individuals starting careers.

TREE is a longitudinal study based on a sample of over 6'000 young people who participated in the Programme for International Student Assessment (PISA) survey (BFS 2002; OECD 2002) for the year 2000 and left compulsory school the same year. Thereafter, until 2007, those adolescents were interviewed annually within the long-term study of TREE concerning the transition from school to profession. In 2010, a final survey focused on their occupational career took place. The TREE data is statistically representative of the gender, regional distribution, and continued education of all graduates in 2000. Within the conducted comparison of the entry wages of young women and men, we could control for a multitude of potentially wage-relevant factors including the specific education and competency of the adolescents (according to their PISA test results, their final grade(s) of apprenticeships, or education at a Gymnasium), their exact job specifications (occupation and tasks), the number of people employed by their companies, and even their value orientation and the socio-economic background of their parents.

The analyses were done with a sub-sample of 1'603 job beginners who hold secondary or tertiary education certificates and started working in fullor part-time jobs (at least 50%) between the years 2004 and 2010.

#### *Descriptive statistics*

Table 1 shows the variables used in this analysis and the descriptive statistics (for the definitions of the variables, see table A1 in Bertschy et al. (2014)). The salaries are based on data for monthly or hourly wages. The gross monthly wages are standardized to a 40-hour work week by using the data on the regular working hours per week.


*Table 1: descriptive statistics (weighted sample)*

*Source: Bertschy et al. (2014, Table 1: 287)*

#### *Kathrin Bertschy*

The average standardized gross monthly wages are 3'753 CHF per month for women and 4'058 CHF per month for men. The analysis of labor market entry and the wages of career starters was carried out by using the data of the subsequent next-but-one longitudinal survey, i. e. one to two years after acquiring the education certificate on secondary level II or the tertiary level. At that point, the young adults had a maximum of two years of work experience. On average, they had been working for seven months.

A major advantage of the data set is the very detailed registration of the skills and education of the young adults. The accumulated human capital can be reproduced by the PISA Reading Literacy Score, the final grades at the conclusion of apprenticeships or education (on a scale of 4 to 6 for successfully concluded education), and the (highest) completed level of education. At career entry, around 75% of the sample had completed an apprenticeship, 16% a vocational baccalaureate (Berufsmatur) or Matura (the latter, for the time being, without further education), almost 5% a degree at a university (or a University of Applied Sciences), and nearly 4% a degree at a Technical College (Fachschule) or higher vocational education (höhere Berufsprüfung, Tertiär B).

The distribution of the occupations is similar to the one found in the "Statistik der beruflichen Grundbildung" (BFS 2007). At first glance, the differences between men and women seem to be small; women more often have a Matura or "Tertiär-B-degree," whereas men more often completed a vocational baccalaureate (Berufsmatur). However, the women have better formal qualifications. Even though there are no big differences regarding final grades, on average, women have considerably higher Reading Literacy competences than their male classmates. The average difference of 32 points corresponds to almost half of one of the five levels of competences for reading literacy (BFS 2002: 24–25).

The distinct separation of gender within the (vocational) education becomes apparent when looking at the trained occupations. Individual educations, e.g. electrician or medical assistant, are clearly or even exclusively one-sided. Commercial apprenticeships represent the highest share: 22% of the sample have a degree as clerk. The 20 most common professional trainings and the 5 most common tertiary educations as well as the respective residual categories are integrated into the regression as dummy variables.

The distinct separation of gender within (vocational) education becomes apparent when looking at trained occupations. Individual educations, such as electrician or medical assistant, are clearly or even exclusively one-sided. Commercial apprenticeships represent the highest share: 22% of the sample have a degree as commercial clerk (kaufmännische/r Angestellte/r). The 20 most common professional trainings and the 5 most common tertiary educations, as well as the respective residual categories, are integrated into the regression as dummy variables.

The jobs of the young adults in the sample have been classified using the International Standard Classification of Occupations (ISCO). For the analysis, we used an aggregation of 24 categories. Within the sample, the job data does not necessarily have to match the education data. Thus, graduates from commercial apprenticeships can be classified in different jobs and ISCO categories accordingly. For example, they could work in company or project management, as clerks at a bank or insurance company, or in administrative positions. They could also work in an entirely different field after having left their original occupation. The aggregated ISCO categories allow a comparison of different challenging job categories.

Furthermore, the data gives information on firm size and shows if a person woks in a leadership position (14% of the sample) or part-time (women [18%] more often than men [5%]). Additionally, the separation of genders in the jobs can be observed by looking at the share of women within a specific job category.

The young adults were asked about their value orientation. For example, they were asked if it is important for them to reach the following goals in the future: "Earning a lot of money, a good salary," "Having a job with good career opportunities," "Having a job that feels like you are doing something meaningful," "Getting married or cohabiting in a committed relationship," or "Having children." The categories for answering ranged from 1 (completely unimportant) to 4 (very important). From the resulting items, we generated different indices by using the principal component analysis. The indices represent intrinsic (i. e. "Having a job that feels like you are doing something meaningful") and extrinsic (i. e. "Earning a lot of money, a good salary") motivations for work as well as the value orientation regarding partnership or family.

#### *Method*

We estimate an extended Mincer earnings function (Mincer 1974) for the analysis of wage differentials. According to the Mincer earnings function, potential wage differentials can be explained by differences in human capital. Besides the principal explanatory factors, such as experience and education, an extended equation involves further variables of influence, particularly socio-demographic traits of the persons and variables related to employment. The measurement of the gender wage gap for entry-level female professionals is done with counterfactual decomposition. This method was developed by Blinder and Oaxaca (both 1973) and presently qualifies as the standard method for measuring the gender wage gap (Jann 2008: 149). The wage equation for men is used as the reference wage equation (meaning, the equation "without" discrimination).


Firm size 10-49 0.041

Firm size 50-249 0.029

Firm size > 250 Leadership position

Proportion of woman by occupation

Work-time percentage (Parttime 50% –90%,

Reference:Fulltime)

Year of analysis *(a)*

Constant Anzahl (N) 1603

R^2 adjusted

AIC

 \*\*\*

0.083 \*\*\*

0.003 0.000 0.053 9.601

0.549 -1841.5

Estimated average wage

Decomposition

Men [CHF]

Women [CHF]

Difference

Explained Unexplained

[%]

5.99%

(0.019)

[%]

1.83%

(0.017)

[%]

7.82%

(0.012)

7.82% 0.50% 7.33%

(0.020)

(0.019)

(0.011)

3677

(26.926)

3976

(36.778)

3976 3677

(25.425)

(35.725)

9.903 \*\*\*

1603 0.57 -1923.6

982 0.545 -1168.4

982 0.597 -1266.6 Training/educational pathways

controlled

(Standard Error)

621 549 -717.2

621

0.569


Training/educational pathways and

professional activities controlled

(Standard Error)

10.402 \*\*\*

10.132 \*\*\*

7.721 \*\*\*

8.553 \*\*\*

0.051 \*\*

0.051 \*\*

0.043 \*\*

0.126

0.132 \*\*

0.034 \*\*

0.027 0.072 \*\*\*


0.023 0.017 0.075 \*\*\*


0.011 0.013 0.078 \*\*\*


0.068 \*\*\*

0.055 \*

0.110 \*\*\*

0.011 0.000

0.059 \*\*

0.048

0.096 \*\*\*

0.003

0.001


**Total**

Training/

Training/

Training/

Training/

Training/

Training/

educational

pathways and

**professional** 

**activities** controlled

educational

pathways

controlled

educational

pathways and

**professional** 

**activities**

controlled

educational

pathways

controlled

educational

pathways and

**professional** 

**activities**

controlled

[1]

Gender (Woman)

Experience Experience (square)

School type (Upper)

PISA Reading Literacy Score

Grade

Highest education degree (Reference = Vocational education VET)

General or Vocational Baccalaureate

Higher vocational education, Technical

University or Applied Science degree

Most frequent 20 vocational tracks and 5 tertiary education *(a)*

Age -0.013

Age (square) 0.000

Migration background

Partnership Socio-economic status

Value orientation: Intrinsic -0.010

Value orientation: Extrinsic

Value orientation: Familiy/relationship

Occupational activities (ISCO-categories)

Firm size (Reference < 10)

 *(a)*

 \*\*

 \*\* 0.048 \*\*

0.017 0.000

0.012 0.000


0.000 \*\*\*

0.033 \*

0.018 0.000 -0.011 0.012 0.000


0.000 \*

0.047 \*\*

0.010 0.001 0.011 0.009 -0.005



0.000 0.000 0.047 0.031 -0.001 -0.024 \*

0.004 0.008


0.000

0.036

0.041



0.010

0.009

College

0.008 0.314 \*\*\*

0.366 \*\*\*

0.260 \*\*\*

0.283 \*\*\*

0.340 \*\*\*

0.395 \*\*\*

0.273 \*\*\*

0.277 \*\*\*

0.297 \*\*\*

0.357 \*\*\*

0.273 \*\*\*

0.311 \*\*\*

0.006

0.008


0.010

0.017


0.002 0.000 0.029 \*\*

0.000 0.001


0.000 0.000 0.031 \*\*

0.000 0.011

0.005 0.000 0.019 0.000 -0.013

0.000 0.000 0.009 0.000 -0.025

0.000 0.000 0.027 0.000 0.004


0.000

0.037 \*\*

0.000

0.021

[2]

[3]

[4]

[5]

[6]

educational

pathways

controlled

**Women**

**Men**

*Note: (a): Additionally controlled variables: 20 most common trained occupations and 5 tertiary degrees; employment classified in 25 categories according to ISCO; year of analysis (2004‒2007, 2010); language region (German, French, Italian speaking part) (see Bertschy et al. 2014, table 2: page 290 and table A1: page 303). Weighted estimation (participation probability). The estimates lead to robust variances. Source Bertschy et al. (2014: 290)*

#### Wage discrimination at career entry in Switzerland

## **4 Results**

Table 2 shows the estimation results of the pooled Ordinary Least Squares (OLS) estimation as well as the estimation for both sexes separately. The first column (estimations 1, 3, 5) shows the estimation results with controls of education/trainings completed, the second column (estimations 2, 4, 6) show the results with the additional control of professional activities carried out. According to the estimated equation, gender explains 4.7% to 5.7% of gross wage differentials. However, these estimations do not reflect that the endowment effects between the sexes not only differ, but can also have different effects on wages.

The separate regression estimates of gross wages for women and men are used for the decomposition of wage differentials into an explained part (endowment effects) and an unexplainable share of wage difference (see Table 3).

By controlling for endowments, the effective wage difference is 7.8% or 300 CHF per month. Twenty-three percent of the differences (1.8% of 7.8%) can be explained with the variables controlled and the choice of training/education. By controlling not only for the trainings completed, but also for the professional activities carried out, the explained portion drops to 7% of the wage differential. The rest of the difference remains unexplained and corresponds to the percentage, women should earn more to be paid the same as their male counterparts; commonly referred to as wage discrimination. In our model, the estimated wage discrimination is around 7.3%.

Significant explanatory factors are, in addition to gender; acquired training, (upper) secondary school type, company size, occupational categories, age, year of entry into the labor market, and region. Taking all factors into account, their influence on wage may provide 60 percent of the variation in women's wages and 57 percent of the variation in men's wages (see R-squared in Table 2, estimates [4] and [6]). Regarding the endowment effects, the trainings completed and the professional activities carried out have a crucial impact on wage level, whereas formal qualifications and competences (reading competences, final grades) do not have a significant impact. The influence of individual factors often differs between men and women (i. e. different coefficients or different significance). For example, the upper secondary school type reflects a higher wage for men (school type with advanced requirements). Regarding the size of the company, higher wages can be achieved in companies with a higher number of employees. This effect is higher for men, whereas for women it can only be manifested for companies with more than 250 employees.

It is not surprising that completed training and professional activities have a significant impact on wages. What is surprising, however, is that "productivity"-variables like formal qualifications and skills (reading skills, final grades) do not have any direct impact on wages.

#### *Do different values and different risk preferences cause gender wage differentials?*

Different values and readiness to assume a risk for the variable portion of wage are likely to only have an indirect effect on wage differentials at career entry. Men with a strong intrinsic motivation for work, who feel that "Having a job that feels like you are doing something meaningful" is "very important," earn around 5% less than those who find it "rather unimportant." In contrast, a distinct extrinsic motivation for work does not have an impact on wage levels, for women or for men. Different behavior of men and women in wage negotiations – a factor that is often pointed out and mentioned to explain inexplicable wage differentials – cannot by proofed with the data. However, there exists no significant difference among the value systems of young adults: Asked about their values and their motivation for work, such as whether a good salary or good career opportunities represent a very or less important role, there are no significant differences between sexes. However, the results should not be interpreted in a way in which those values do not play a role at all. As confirmed by empirical analysis, it cannot be excluded that value systems can be decisive factors even before choosing a career. From the beginning, with the choice of educational pathways, or later, with the choice of employer, women (or men) with distinct preferences for family can choose options from which they could expect better reconciliations of family time and work.

Additionally, gender differences and preferences for risk are unlikely to explain wage differentials at career entry. The argument, for example, that women have less preference than men for variable pay systems – although such would be beneficial, given their skills – and prefer fixed pay systems with lower salaries instead, is not reflected in different value systems. Furthermore, this argument remains highly controversial in the literature. Even if women would have weaker preferences for variable pay systems, and therefore would prefer fixed pay systems with lower salaries instead, those factors are not likely to explain wage differentials at career entry. On one hand, as shown in data from the Swiss earnings structure survey (BFS 2012), variable wage components are ten times as high in mid- and senior level positions than in lower positions. Therefore, variable wage components are expected to only be relevant at later career stages. On the other hand, according to empirical studies, differences in risk aversion can only explain a small portion of wage differentials (Le et al. 2011).

The first conclusion of our analysis is that wage discrimination is not the result of different career paths, but already exists at career entry. "Productivity" variables, like formal qualifications and skills, do not have any direct impact on wages. Also, values and risk preferences are unlikely to influence the wages at career entry.

#### *Impact of occupational segregation*

We wanted to know if strong occupational segregation impacts starting wages and unexplainable wage differentials. We performed the same analysis separately for typically female, gender-mixed, and typically male jobs. Occupations with a female representation of 70% or more were defined as typically female professions and those with a share of 70% or more men as typically male professions. <sup>1</sup> As can be concluded from the descriptive statistics in Table 1, average wages at career entry are significantly lower in female-typical occupations (3'824 CHF per month) than in those that are typically male (4'028), and slightly lower than in gender-mixed occupations (3'872). Women, on average, earn more in typically female occupations, than on average in mixed or typically male occupations, but still earn less than men in all three categories. Men even achieve significantly higher incomes than women in typically female occupations.

The wage gap in the data used can be explained, to a great extent, by the fact that woman "choose" trainings or occupations that are paid less in the labor market. In other words, for occupations with a comparatively high proportion of women, on average, lower incomes are paid. This well-known effect is also confirmed by the data (see Table 1, Average wage within segregation class: Total).

However, the results of the regression analyses, applied separately for the different segregation levels of occupations (see Table 3), show that in accordance with their training and skills, women in particular earn less in gendermixed occupations. With the variables controlled, 6.3% of 7.1% of the differences cannot be explained. This effect also persists not only the trainings, but also when the professional activities are controlled. However, in the male-

<sup>1</sup> Examples of jobs with a women share below 30% are: mechanic, electrician, gardener, computer scientist, etc. Gender-mixed professions include baker, retail employee, or teacher at secondary schools. Whereas female-typical jobs, for example, are infant teacher, clerk, medical assistant, etc. The classification is based on the Swiss Census for the year 2000. The distribution of women and men in the sample (unweighted) is as follows: male-typical occupations (number of men: 359/number of women: 79), gender-mixed occupations (182/322), typically female occupations (80/590)

typical occupations, with an 8.9% wage gap, the unexplainable part of the wage differential decreases from 6.3% to 3.1% if, in addition, professional activities are taken into account. This indicates that within male-typical jobs, women focus more on professional activities and work content that is paid a lower wage. However, decomposition results are not significant here.

In the female-typical occupations, however, men on average earn 360 CHF per month more than women. Because of their training, men should achieve higher wages ‒ about 14.5% more (decomposition A). The negative unexplained share reveals that the wage differential should be higher due to the endowment variables. With the additional control of professional activities, this discrimination effect of men in the typically female occupations is no longer significant (decomposition B) and can be explained by the fact that men take over activities and work content that is better paid within female-typical occupations.

*Table 3: Decomposition of wage differentials due to segregational level of occupations*


*Not significant results are presented in italics (Significance level 5%). Source: Bertschy et al. (2014, Table 3: 293)*

Looking at concrete examples in the data, it becomes clear that within female-typical occupations, such as social worker or primary school teacher, men fill comparatively well paid positions. Men often specialize within those same professions to activities with better salaries. For example, in the commercial sector, men take or get jobs with more responsibility or auditing activities. Additionally, women occupy niche activities in male-dominated occupations. For instance, they engage in administrative or sales activities within the metal construction industry, even though they have technical training which qualifies them for better paid activities, or they work as data processing assistants, despite having training as a computer scientist.

These gender niches within the professions have been repeatedly pointed out in the research (see, for example, Heintz et al. 1997; Krüger 2001). It is likely that these niches are even more pronounced in reality, but the database does not provide more detailed information. Focusing on niche activities – self chosen or allocated – probably provides another reason for unexplained wage differentials. This observation indicates that the discriminatory wage differentials might be slightly overestimated. It also leads to another question: Why and how are these niches chosen by young women and men or allocated (by firms, social attitudes, prejudices, etc.)? It seems economically inefficient for women to procure education and career training if they do not work in the jobs or activities they are qualified for.

Our second conclusion is that in high gender-segregated occupations, woman chose (or are allocated to) niche activities that are paid less, even if their training qualifies them for better paid activities. Men do the contrary. Unexplainable wage differentials of 7.3% at career entry might be slightly overestimated.

#### *Wages as a determinant for further gender inequalities*

From an economic point of view, and with regard to an optimal allocation of human resources, it is questionable that women do not work in activities their training enables them for, and instead escape into niches. The lower incomes of women at career entry also cause economically problematic dynamic trends. Earnings determine gender roles and participation rate in the labor market. This has been shown, for example, by Kanji (2013) and can determine further gender inequalities. If individuals in a partnership efficiently resolve the issue of household and work organization by having one partner specialize in paid work and the other focused on unpaid (house) work, and the distribution of labor and family work is affected by the hourly wages of both partners (Becker, 1995), it is still more likely, that the woman would reduce her working hours, even if she has an increasing potential to earn more than her partner due to her educational achievements. As a result of wage discrimination at career entry, women automatically start at a disadvantage. Furthermore, by establishing a family unit, division into traditional gender roles is more likely, as is an increase in wage inequality within the career (see e. g. Bispinck et al. 2008).

We saw in our analysis that young female professionals differ little from their male colleagues in their value orientation with respect to family/kids and in extrinsic or intrinsic motivation for work. No significant correlations between value orientation and real achieved wage can be determined. However, the actual division of labor in practice equals social norms (e. g. in the division of household work and employment in couple relationships) rather than the expressed value orientations. Further analyses of a sub-sample<sup>2</sup> of those respondents that live in couple relationships show that women carry out a significantly higher share of family and household work, even when working full-time, including those women who earn more than 50% of household income.

In relative comparison, those women who execute a high workload and/ or have a preference for full-time employment, carry out a lower share of the incurred household work. The greater the value orientation for family/children or if a woman already has children (true for 13% of the sample), the greater is her share of the housework. For the men, the same preferences lead to the opposite effect. The share of housework the men are carrying out, according to their own assessment, decreases when they report a stronger value orientation for family/children or if they already have children. For the women, those correlations also remain in the multivariate observation (not so for the men). There is a relationship between the division of labor within the family and the real incomes of women. A comparison of the data for career entrants in 2006 (entering the labor market before 2007) with the data in 2010 shows that women who had an above average income (compared to the other women) and/or recorded a high proportion (>50%) of the household income at the entry level, also had a more balanced division of housework with their partner four years later, i.e. on average they carry out a smaller share of housework than other women, but still more than their partner.

Analysis of wage developments in the first five years of work experience (2005–2010) shows that a traditional distribution of paid and domestic work seems to be quite economically efficient from the perspective of young households. Men's increase in salary turns out to be significantly higher than the one for women: The (standardized) gross monthly wages of men (as well as the real wages) rise by an average of 7.8%, the wages of women by only 6.4%.<sup>3</sup> Thereby, the average wage gap increases again by almost 7% over five years. This difference in wage developments between young male and female professionals can neither be explained with their formal qualifications nor with part-time employment, which is more common among women. The same dif-

<sup>2</sup> Employees in 2010 living in a relationship that, in addition to the information on their own earnings, also answered questions about household income and the distribution of housework (n = 572).

<sup>3</sup> In a 2-sample t-test the null hypothesis of identical wage development is rejected at a significance level of less than 1% (n = 256). Almost identical results in the development of real wages (N = 262, the null hypothesis is rejected at significance level of <5%)

ferences are also apparent if we only compare full-time employed women and men.4

Theories in which the partner with the higher income is less active in the household, and more active in the labor market, are supported. Even more so when we assume that this unequal division of labor will be reinforced over time due to different statistically unjustifiable wage developments for men and women and also that the majority of young professionals in couple households quickly divide the paid and house work according to the traditional pattern.

## **5 Conclusion**

Our analysis shows that wage discrimination is not the result of different career paths, but already exists at career entry. "Productivity" variables like formal qualifications and skills do not have any direct impact on wages. Also, different values (e. g. extrinsic or intrinsic work orientation) or different risk preferences are unlikely to influence the wages at career entry. We conclude that the early wage discrimination is partly due to the persistent gender segregation in the Swiss labor market and education system, which stimulate early gender-typed occupational career decisions. Findings show a wage discrimination of 7.3% at career entry. Young women have lower earnings due to lower salaries in traditional female occupations and because of wage discrimination in gender-mixed occupations as well as in typically male professions. In highly segregated jobs, women in typical male-jobs often "choose" or get allocated to specific work contents that go together with lower compensation; men do the contrary in typical female-jobs. These niche activities are economically inefficient because training qualifies women for better paid activities. Young women do not choose or are not allocated to these activities they have trained for.

### *Possible reasons for early wage gaps?*

Unexplainable wage differentials at career entry might be slightly overestimated, and also within our data set. However, the majority of the inexplicable differences in wages are likely the result of discrimination. One possibility is the so-called "statistical" gender discrimination. Companies assume that,

<sup>4</sup> The null hypothesis of identical wage development is rejected at a significance level of <5% (n = 223).

sooner or later, women tend to reduce their working hours. From the start, these companies consciously or unconsciously pay lower salaries to women, assign them lower paying jobs, or do not admit them the same development in wages as for men. In sociological and economic theory, this effect is referred to as statistical discrimination, because a specific behavior (the one of the companies) is derived from observed probabilities (women reduce their employment more often than men due to parenting). This behavior of the companies discriminates against all women, including those not complying with this assessment and those not intending to reduce their workload, with or without children.

Generous family-friendly policies, such as long maternity leaves and/ or part-time work protections, made it possible for more women to work. But they can also enforce statistical discrimination: Woman are more likely to work in low segment, lower paid jobs and less likely to be managers in countries with generous family-friendly policies, as shown by Blau and Kahn (2013) in a study comparing 22 countries. To avoid statistical discrimination, policies should be devised gender neutral. This is apparent in countries/regions including Iceland, Sweden and Quebec where parental leave policies encourage both men and women to take time off for a new child. Apart from measures against education and labor market segmentation, gender-neutral state policies could thus help to close the early gender wage gap. Secondly, solidarity among men is also likely to play a role. In sociological theory, this behavior is known as discrimination theory, which states that "the same" or "the like" is preferred. As there are many more men than women in superior positions and as such, men more often make employment or wage decisions, they may prefer applicants or employees with similar characteristics, for example. This preferential treatment can also be unconsciously done.

#### *Implications*

Implications of gender wage gap on work and family and the persistence of gender inequalities can be manifold. In order to maximize household-production, young professionals in couple households quickly divide the paid and house work according to the traditional pattern Analyses with TREE-Data show that young couples divide household labor and paid work based on hourly wages. An unequal division of labor will be reinforced over time due to the different, statistically unjustifiable wage developments for men and women. From an economic point of view, and with regard to an optimal allocation of human resources, this development is economically inefficient. The lower incomes of women at career entry also cause these economically problematic dynamic trends. The interaction of all these effects is jointly responsible for the persistence of gender inequalities.

## **References**


ter? pp. 407–442 in *From school to work: A comparative study of educational qualifications and occupational destinations,* edited by Yossi Shavit und Walter Müller. Oxford: Oxford University Press.


#### *Kathrin Bertschy*


# **Cantonal patterns of gender-specific time-inequalities in paid and unpaid work: Empirical results and political-institutional conclusions**

*Michael Nollert & Sebastian Schief\** 

## **1 Introduction**

Women continue to be disadvantaged in the working world. They have lower labor force participation rates than men and are usually paid less than men (BASS 2010; Donzé 2013). They are overrepresented among people in precarious employment and among the non-registered unemployed, and underrepresented at higher management levels in large companies (Van der Lippe et al. 2010; Miranda 2011). Similar patterns can be found in other areas: unpaid family work and housework (Schempp/Schief/Wagner 2015), care work and helping relatives and neighbors. Similar to volunteer work (Nollert/Huser 2006), women mainly perform basic tasks, whereas men often assume leadership roles with greater prestige.

There are few comparative studies that consider paid employment, unpaid family and housework, and volunteer work. Even though strong regional variations in the division of paid and unpaid work can be observed, there has been little research on the political, institutional and social structural factors responsible for these variations.

Obviously time inequality is a major issue within the working world and there is a need to analyze causes of these disparities reflecting the relevant political, cultural and economic context. Therefore, our research not only focused on gaps between men and women in paid and unpaid work (time inequalities), but also took regional differences into account. We premise that Switzerland is especially suitable for a comparative analysis because it is

<sup>\*</sup>Both authors contributed equally to the article.

culturally heterogeneous and provides with the cantons as federal subunits a comparatively large scope of decision making (Armingeon et al. 2004). The *Gender Equality Atlas* of Bühler (2001a) already indicated considerable differences between the regions of Switzerland in terms of inequality between men and women. Although our statistical analysis is limited to a single country, we premise that our conclusions can be generalized for contexts in which gender-specific time inequalities can be observed.

Regarding family models, in Western Switzerland the dual breadwinner/ external carer model is dominating. In contrast, in the German-speaking part of Switzerland a majority of families are organized around the male breadwinner/female part-time carer model (Pfau-Effinger 2005).

Finally, in the Italian-speaking canton of Ticino, a traditional family model dominates which barely supports paid work for mothers. On the other hand, if a woman has no children her probability of being full-time employed is even higher than in the rest of Switzerland (Losa/Origoni 2005). It is likely that the big cultural, political and socio-structural variance in Switzerland leads to the wide range of time inequalities.

Thus, we first grouped the cantons based on an analysis of inter-cantonal differences. In a further step, we examined the influence of individual and context features on the gender-specific inequalities. Moreover we identified historical and institutional factors that explain cantonal inequalities in the working environment.

After a sketch of theoretical considerations (section 2) and preliminary findings referring to gender gaps in paid and unpaid work, we summarize our empirical findings (Epple et al. 2014; Gasser et al. 2015; and Epple et al. 2015). In section 3.1. we present a cluster analysis of Swiss cantons based on data on time inequalities. In section 3.2., the findings of a Qualitative Comparative Analysis (QCA) on the influence of configurations of cantonal factors on gender time inequality are summarized. Section 3.3. recaps results of a multilevel analysis on parenthood and employment.

In the second part of the paper, we formulate political-institutional conclusions that can be drawn from the presented analysis. They would pave the way for a "dual earner/dual caregiver society" (Gornick/Meyers 2008). Section 4.1 emphasizes that, first and foremost, there must be a fundamental change in working time policies and working time organization, especially for men. In Section 4.2. we argue that gender policies should be tailored to the needs of the cantons because of the huge differences, e. g. in terms of values and beliefs. Section 4.3. brings forward that the impact of welfare regimes in shaping gender-specific time inequalities is underrated and needs to be emphasized. The article ends by drawing some conclusions.

## **2 Theoretical focus and expectations**

The analytical starting point for the empirical detection of gender time inequality is the integral consideration of the three spheres of work: labor, household and family work and volunteer work. It is a major goal of gender equality policy that paid and unpaid work between women and men is symmetrically distributed (e. g. Fraser 1994; Plantenga et al. 2009). The distribution is considered as "fair" if women and men symmetrically share paid and unpaid work (cf. Gornick/Meyers 2008; Haug 2011).

It is recognized, of course, that with the expansion of the adult worker model (Daly 2011) neither the problem of unequal opportunities of access to well-paid employment of good quality nor the fair distribution of unpaid work will be solved. We should also be aware that neoliberal policies may reduce inequality by making it economically necessary for women to support their families through paid work.

Nevertheless, the integral analysis of paid and unpaid work is vital in terms of gender inequalities for three reasons: Firstly, women more than men are working part-time and are thus more dependent on sources of income outside of the labor market. Secondly, even an egalitarian division of paid work does not guarantee that the unpaid work is distributed egalitarian. Thirdly, every community depends on both care and education work within the family and voluntary engagement in non-profit organizations.

As a matter of fact, time inequality has decreased in the 20th and early 21st century in Switzerland. Thus, the proportion of couples with children under age six who lived in a traditional breadwinner model reduced from 61.5% in 1992 to 29.2% in 2012 (BFS 2013, p. 18). However, household and family work is still mainly in the responsibility of women and usually suffers from a lack of recognition and social security (BASS 2012).

The Gender Equality Atlas (Bühler 2001a) presents an index of equality that takes into account a variety of different indicators. The cities of French-speaking Switzerland and Basel-City showed the lowest levels of gender inequality, whereas in the rural areas (Uri, Schwyz) and the urban surroundings of German-speaking Switzerland high levels of inequality persisted.

Comparative research on the causes of variation in gender inequalities often emphasizes the influence of cultural factors (Bühler 2001b; Pfau-Effinger 2012). Thus, one can postulate that traditionalist, family-oriented attitudes in the population imply that men primarily earn income in labor markets whereas women do the housework and care for family members (children, grandparents) (Cyba 2002; Pfau-Effinger 2010). Those role models especially prevail in regions where Catholicism is still influential. Therefore, we expected that in traditional cantons the traditional division of labor between women and men would be more resistant to the claim for a more egalitarian work distribution pattern than in progressive cantons.

The example of Scandinavia also shows that the claim for gender equality is mainly articulated by left-wing parties. Accordingly, it is expected that gender inequality decreases with increasing strength of left-wing parties (Korpi 2000). However, feminist scholars have repeatedly criticized that the social democratic welfare state is still androcentric and corroborates the horizontal segregation in occupations (Charles 2005; Mandel/Semyonov 2006). Nevertheless, a cross-national analysis of Bambra (2004), for example, showed that female employment rates are higher in social democratic regimes. Thus, the welfare state decreases the dependence of women on the male breadwinner income (defamilization) (Bambra 2004) by inter alia promoting employment opportunities for women and by providing external childcare services (Hegewisch/Gornick 2011). The welfare state in general and the public sector in particular are assumed to help in reducing time inequality. Therefore, it is reasonable to assume an influence of the strength of left parties, the support of gender equality in the population, the strength of the welfare state and the size of the public sector on gender equality.

Finally, economic development is seen as an equalizing factor because economic actors more and more accept the view that the integration of women into paid work provides more prosperity. Similarly, it is argued that increasing urbanization and a growing tertiary sector weaken the traditional family model. In contrast, the increasing demand for care for children and especially for elderly people pushes women out of the labor market and into the family and therefore supports the traditional division of labor (Sullivan/ Gershuny 2001).

The main reason why Switzerland is especially suitable for analyzing the noted causes of time inequality is due to the fact that the social and family policy profiles of the cantons vary considerably (Armingeon et al. 2004). Moreover, cantons differ in the political-institutional design, in the party political constellations, the values and the gender culture. Cross-national studies have documented the relevance of all of these factors in explaining gender time inequality (e. g. van der Lippe et al. 2010). In addition, the inter-cantonal comparison of Stadelmann-Steffen (2007a) underlines the impact of social and family policies on the extent of female employment.

## **3 Empirical results**

Our research on cantonal patterns of gender-specific inequalities in paid and unpaid work (Epple et al. 2014; Gasser et al. 2015) presumes that in addition to discrimination in paid work, the distribution of unpaid work is a key element in the debates on gender inequality. Hence, we compared for the first time the Swiss cantons with respect to patterns of time inequality based on the gendered division of paid and of unpaid work.

#### *3.1 Time inequalities in the Swiss cantons*

In a first step we checked whether there are patterns of time inequality based on the gender division of paid and unpaid work. Using data from the Swiss Labour Force Survey several cluster analyses based on the gender gaps in time use for employment, housework and family work show six stable patterns of macro-level time inequality. Almost all cantons could be related to a gender time (in-)equality cluster. Geneva and Basel City show the most egalitarian pattern with very small gender gaps in both spheres, paid and unpaid work (modern). A second cluster shows a modernized pattern (Neuchâtel, Zurich, Vaud, Appenzell Ausserrhoden), where the gender gap in paid work is slightly bigger as in the modern pattern. The more rural cantons tend to belong to a more inegalitarian pattern of gender time inequality. Cantons with a traditional pattern (Obwalden, Nidwalden, Uri, Grisons) have big gender gaps in paid and unpaid work. Cantons with a detraditionalized pattern are Fribourg, Thurgau, St. Gallen and Aargau. These cantons are partly urban and partly semi-rural. Being historically coined as traditional, the cantons have modernized step by step lately. The familial-modern pattern was identified in the cantons of Solothurn, Schaffhausen, Glarus and Jura. It is characterized by a small gender gap in housework and family work and a high gender gap in labor. The opposite time distribution is connected to the familial-traditional pattern of the Ticino. In this particular case the gender gap in labor is slightly below the cantonal average but the gender gap in housework and family work is the highest of all cantons.

The gendered distribution of labor, housework and family work seems to reflect cantonal gender arrangements (Bühler 2001b; Pfau-Effinger 2005, 2012; Epple et al. 2014). In other words, the institutional framework at the level of cantons influences constraints and opportunities of individuals doing paid an unpaid work. Moreover, the structural influence might be different for men and women. Obviously one has to have in mind that gender time inequality is a matter of a complex interaction of several structural and personal factors (Bühler 2001a). The influence of the structure on the room for maneuver (1) is therefore built by a complex interplay of structural conditions building an institutional framework and (2) creates different constraints and opportunities for different subgroups of societies.

*Source: Gasser et al. (2015)* 

*Abbreviations: AG = Aargau, AR = Appenzell Ausserrhoden, BE = Berne, BS = Basel-Stadt, BL = Basel-Landschaft, FR = Fribourg, GE = Geneva, GL = Glarus, GR = Grisons, JU = Jura, LU = Lucerne , NE = Neuchâtel, NW = Nidwalden, OW = Obwalden, SG = St. Gallen, SH = Schaff hausen, SO = Solothurn, TG= Thurgau, TI = Ticino, VD = Vaud VD, VS = Wallis, ZG = Zug, ZH = Zurich*

#### *3.2 Causes of diff erent time inequalities*

In order to test the impact of the structural features of the cantons (see section 2), we used a fuzzy-set Qualitative Analysis (fsQCA)<sup>1</sup>. This kind of analysis is suitable for this purpose since it allows for testing configurations of conditions that lead to a certain outcome even though the number of cases is rather low (in our analysis 26 cantons). FsQCA therefore is situated between qualitative and quantitative analyses (Ragin 2008; Schneider und Wagemann

<sup>1</sup> See Table 2 of Epple et al. (2014) for detailed results. 1

2012; Leggewie 2013). In contrast to crisp-set QCA, fsQCA especially aims at handling continuous variables. Thus, fsQCA variables must not be dichotomous, e. g. 1 for "in" and 0 for "out".

The analysis helped to figure out relevant configurations of structural conditions that lead to more or less gender time inequality on cantonal level. Our research on institutions and gender time inequality in Swiss Cantons (Epple et al. 2014) compares cantons with respect to gender time inequality of paid and unpaid work, defined as the gender gap in how men and women, on average, divide their time between labor, housework and family work. We ranked the cantons on the basis of an index which measures the dissimilarity of cantonal gender time inequality (see Epple et al. 2014 for details). Ten out of 26 cantons were found to have a lower gender time inequality than Switzerland as a whole: Geneva, Basel-Stadt, Zurich, Berne, Glarus, Solothurn, Schaffhausen, Vaud, Neuchâtel, and Jura.

We addressed two research gaps: Firstly, subnational comparisons are often neglected in comparative research. Secondly, little interest is shown in how contextual factors crystallize into configurations related to macro-level gender time inequality. The crystallization can be interpreted as the cantonal structure influencing the agency of individuals within a canton.

Although there is a wide variety of contextual factors from different societal dimensions (economy, politics, culture, social structure) we had to limit the model to five indicators. We chose the indicators on the basis of preliminary analyses, e. g. traditionalism instead of Catholicism. These indicators were taken into account: traditionalism, public sector size, welfare spending, political backing of gender politics, and potential care demand (children, elderly), as a demographic indicator. Our focus was on the "big picture": How is gender time inequality related to arrangements of these five conditions? Our fsQCA suggests that strong traditionalism is a necessary condition for high gender time inequality, strengthening previous support for the role of cultural factors. The indicator for traditionalism is based on the Swiss Electoral Studies (Selects) data (Lutz 2008; FORS 2010) for the 2007 parliamentary elections. We used the mean of reported importance given by the people to four issues: preservation of tradition, strong army, law and order, and equal chances for foreigners.

However, strong traditionalism alone is not sufficient for high inequality. It is sufficient only in the right supportive context of a small public sector, a weak welfare state, or a high care demand. Addressing these supportive conditions thus has the potential to mitigate the relation between traditionalism and inequality. This leads us to another observation that strengthens the idea of cantonal gender arrangements: Several different configurations relate to gender time inequality, and all but one comprise both political-institutional and cultural factors. Two further conclusions can be drawn by the observation. Firstly, there seems to be no way around political action in order to reduce gender time inequality. Secondly, changing a complex social problem like gender time inequality requires that a whole bundle of institutions is changing. The two conditions of care demand and political backing serve to illustrate the complexity: Both are positively related to gender time inequality in one configuration, and negatively in another. Given this potential for contradictory associations, it is particularly suggestive to note that building up welfare state support, strengthening the public sector, and weakening traditional family models are all three associated with lower gender time inequality in whichever configuration they occur.

What is called *configuration* in terms of a fsQCA refers to the institutional and structural conditions that influence labor, housework and family work of people in the cantons. Hence, the term *configuration* emphasizes that it is not a single condition or a single institution but a complex interplay of contextual factors that affects gender time inequality. Apart from different structural levels influencing each other and structure influencing agency we assumed that different parts of the institutional framework of one structural level influence each other on top of that. Our results clearly support this presumption.

#### *3.3 Parenthood and employment*

At the household level the question of how to distribute paid and unpaid work is particularly crucial after transition to parenthood. Therefore, we concentrated in a further step on the impact of policies and culture on employment of parents compared to non-parents (Epple et al. 2015). To be more specific, the research analyzed employment (yes or no) as well as working time (number of hours) of parents compared to non-parents. The gendered effect of parenthood on employment is a principal driving force for gender inequality in general, as well as for the potential of a more equal division of housework and family work. The empirical findings presented here address the question whether these gendered effects are influenced by cantonal social policy and cultural contexts. In other words, the research dealt with the question of whether specific structural institutions have diverse impacts on specific groups of persons' agency. In contrast to ordinary linear regression techniques, the multilevel analysis allows us to prove or discard statistically both the influence of individual attributes (level 1) and the impact of the structural context (level 2) on individual behavior (Langer 2009).

The patterns of individual-level effects comparing the models for parents (i. e., parents of young children) with those for non-parents (i. e., men and women without a co-resident child under 15) show hardly any difference. It seems that parenthood does not change the basic pattern of individual predictors of employment and the amount of working hours.

The results confirm previous cross-national findings that social policy supports employment, potentially buffering the gendered impact of parenthood. Contrary to our expectations, however, mother's probability of employment increases with the tax load. Two explanations are plausible: Firstly, a high tax load may decrease available income and thus make a second income more desirable. Secondly, a higher tax load is associated with a more generous welfare regime, as taxes are an important way of financing welfare expenditures (Esping-Andersen 1990). Usually, a larger welfare state is accompanied with higher employment rates of mothers, as is the case in social-democratic welfare regimes.

Childcare facilities buffer the gendered impact of parenthood and raise the potential for a more balanced division of paid work and family and housework between men and women. It is related to longer working hours among mothers of young children, and lower employment rates and working hours for fathers. This confirms previous results by Stadelmann-Steffen (2007b) and Stern et al. (2013). However, the supportive role of childcare facilities for mothers only applies for those already employed. They do not increase mothers' probability of employment, but in contrast lower those of women without children. Moreover, increasing supply rates of childcare facilities lead to fewer employed men and shorter working hours of men without children. Thus, the supply rate can be seen as an indicator for the extent of the double-carermodel among families, suggesting different cantonal degrees of familialism. As the supply rate also affects those men and women without children, it furthermore can be interpreted as an indicator for the cantonal welfare arrangement and the underlying gender culture, although in some cantons the responsibility for childcare lies upon the market or communities (Epple et al. 2015).

Besides public policies, parents' employment is associated with the cultural context. Traditionalism exacerbates the re-traditionalizing impact of parenthood. In more traditional cantons mothers of small children have lower working hours while fathers are employed more often and increase working hours. The effect of traditionalism is bound especially to parents since we couldn't find an effect on the employment rate of mothers, but women without children show an increased probability of employment within that cultural context. Furthermore, male employment in all models is positively linked to a traditional context, regardless of fatherhood.

In contrast to men's employment, female employment is closely related to the language region. Although Neuchâtel, Vaud and especially Geneva are mainly Protestant and modern cantons (see 3.1.), even after controlling for traditionalism and Catholicism, in French- or Italian-speaking cantons, fewer women without children are employed. It is less surprising that employed mothers work longer hours. Obviously, the institutional conditions to work longer are better for mothers in Western Switzerland and the Ticino than in German-speaking cantons.

## **4 Political-institutional implications**

In what follows, we draw three political-institutional conclusions that can be derived from the results of the NRP60-project presented above. Following several "working time" scholars we argue that working time policy is always gender policy (Jurczyk 1998, Gross/Seifert 2010, Mückenberger 2012, Allmendinger et al. 2013): For a dual earner/dual caregiver society first and foremost a fundamental change in working time policies and working time organization would be necessary, especially for men. Secondly, since longterm values and beliefs (traditionalism) are persistent, innovative measures that strengthen gender time equality in traditionalist cantons have to be developed. Thirdly, the role of welfare regimes in shaping gender-specific time inequalities is underrated and needs to be emphasized.

#### *4.1 New working time policies and working time organization*

Our results point to the fundamental influence of working time and work organization on gender time equality. As we have seen, the organization of female and male labor is very different. Generally speaking, men have higher employment rates and work longer hours then women, especially in the case of parenthood. Gender equality policies have faced that problem and tried to trigger changes by first and foremost promoting the opportunities of women to arrange family work and employment. The main aim was to open the labor market for more women, especially mothers. The employment rates of women show that those policies of promoting female labor have been successful – at least partly. More and more women are employed, even though a significant percentage of them works part-time.

However, the fact that women are more integrated in the labor market than they used to be has only marginal influence on the way housework and family work is organized. Women are still doing the majority of housework and family work. This holds true even for couples where both partners work fulltime. Moreover, our results show that parenthood does not change the way men organize labor. On the contrary, fathers work on average longer hours than before the birth of the first child.

Policies at and for the work place therefore might not promote more gender equality in housework and family work – even if they would lead to an equal share of the work volume of labor. In our opinion, gender equality policies have to more strongly emphasize working time organization and work organization of male labor than it used to do in the past. Questions addressed should be: How can we change the cultural setting of male work as fulltime work and female work as part-time work? What incentives can be implemented for employers to reduce working hours of men without harming career opportunities?

Thus, a gender equality promoting policy needs to address both sides of the imbalance of men and women in labor, the fact that men work too long and women work too short to settle the circumstances of gender time equality within housework, family work and labor. If equality policies emphasize labor market inclusion of women only, the odds of reaching the aim of general gender time equality are rather limited. Therefore, Allmendinger et al. (2013) propose a 32 hours weekly working time as standard working time for all.

#### *4.2 Linking traditional values and beliefs to gender time equality*

Our results show that there is a strong connection between traditional values and beliefs and gender time inequality. Cantons where traditionalism prevails tend to have bigger differences in terms of housework, family work and labor between men and women. It is obvious that values and beliefs are rather stable and hard to change. Moreover, governmental attempts to change values and beliefs are hardly ever welcome in liberal democratic societies, for good reason. Finally, in order to narrow the gap between cantons with low and large inequalities, policy measures have to develop solutions tailored for rather rural cantons with traditional values and beliefs. More child care will not change gender time inequality in cantons where a strong majority of people still believe that young children should be raised at home exclusively. More jobs for women will not change anything in cantons where the majority of people believe that a woman should stay at home running the household and taking care for the kids.

However, two strategies of change are imaginable: Firstly, the minority of people who are in favor of gender time equality are to be supported. Those people work as role models within those cantons showing the advantages of more gender equality. This would be a long-term strategy whereas the question of success is rather open. It might very well be that the forerunners of gender equality are not accepted as role models since there are only few values and beliefs shared.

Secondly, the connections of traditional values and beliefs with gender time inequality should be used. While it is commonly agreed that gender equality in the sense used here is not on the agenda of traditionalism, family values, a pursuit of wealth, work ethics and justice could be links to a policy leading to more gender time equality. Moreover, gender time inequality could be interpreted economically as a waste of human capital (Allmendinger 2013).

Family values include the imperative of being able to enable a decent living for the family. In an ideal traditional world, it should be the man who earns the money to support the family. In lower income classes, it is necessary for both, men and women, to contribute to the household income to enable the living standard aspired. When it comes to having to decide between traditional family values and a decent living standard, people might choose the decent living standard without accepting the claim for gender equality.

Traditionalist values and beliefs include work ethics and justice. People who work hard should earn enough money. Hence, a policy of gender equality in traditionalist cantons could emphasize justice and performance of people, be it men or women. The performance of the individual thus might outperform traditional values. Once again, it is an open question which value would succeed, but policy could set a framework in order to promote gender equality. In general, gender equality policy in traditionalist cantons should emphasize shared values as mentioned above to promote gender equality instead of promoting the value of gender equality itself.

#### *4.3 More gender-sensitive welfare spending*

Our third political-institutional conclusion derives from the fact that the size and type of welfare regimes influence gender time equality. Our results give proof that higher welfare spending promotes gender time equality (see e. g. Sainsbury 1996). Of course it is always a set of conditions that affect the division of paid and unpaid work, but welfare spending persistently decreases time inequality. Nevertheless, we have to keep in mind that the way a welfare regime is organized has multiple impacts on various aspects of society. It would be reasonable to define common criteria about welfare regimes in Switzerland to safeguard a certain level of support, especially for women.

In practice, the way work-family policies in Switzerland are organized has to change. Firstly, the supply of child care has to reach a level so that it is not a burden for parents to find an affordable place for their children if they want or have to work to make ends meet. Secondly, social security has to be designed in a way that leaving the labor market for some time to care for children should not seriously harm material security of individuals or/and the family. Welfare regimes should provide flexibility with respect to unusual, segmental life courses. If so, the welfare state as lender of last resort gives way to find solutions to balance labor, family and housework individually or within a household. Thirdly, and maybe most difficult, the relationship between life and work has to be organized like a revolving door where men and women can easily get in and out and in again. As a political measure, this recommendation seems to be very difficult to plan and to put into practice. The political institutional framework can ease the transition from work to family work and vice versa, or make a reconciliation of labor and family work possible. But a major part of responsibility to reach that aim is dependent on the will of enterprises and employers to support those rather segmented life courses and labor episodes.

Creating a political institutional framework to improve the reconciliation of work and life certainly has its limits since it only can create opportunities or restraints for enterprises, households or individuals to do or not do certain things. Consequently, it is an open question to what extent and how the recipients perceive the changes and measures recommended and to what extent this leads to changes in behavior. Some of the measures will be starting points for improvements, some will change nothing at all, and some might have unintended consequences even pointing in the opposite direction of what was intended. It is therefore unavoidable to evaluate the measures and to change them in case they turn out to be ineffective, unnecessary or even counterproductive.

## **5 Conclusions**

Switzerland's paid and unpaid work is still unequally distributed between men and women. While labor is still dominated by men, women carry the majority of the burden of unpaid housework, family and care work. However, there are considerable differences between regions and cantons. This variety of time inequalities makes it difficult to generalize across all parts of the country. High inequality prevails in central Switzerland and Appenzell Innerrhoden. In contrast, time inequality in the highly urbanized cantons of Basel-City and Geneva is comparatively low.

Although a decrease in the employment gap seems to be linked to a decrease in the housework/family work gap there are deviant cases. The Italian-speaking Ticino is characterized by a strong burden of women in unpaid work but a comparatively equal distribution of labor. On the other hand, we find four cantons in which men are comparatively more engaged in housework and family work. Although gender time inequality generally decreased between 2000 and 2010 there is at the same time a trend of more regional divergence.

Our results suggest that certain conditions and combinations of conditions are connected to higher or lower time inequality. Thus, higher inequality is primarily linked to strong traditionalism combined with low social spending and/or a weak public sector and/or high care needs. Weak traditionalism, strong political support of gender equality policy measures as well as a strong public sector, including state childcare, are responsible for a lower time inequality in the highly urbanized cantons of Geneva, Basel-City as well as in Zurich and Western Switzerland. Finally, we could show that the differences in employment and working hours of mothers and fathers are at least partly moderated by the cultural and political context of the cantons.

Political-institutional measures promoting the dual earner/dual caregiver society need to consider the results described above. First and foremost, the results point to a necessary fundamental change in working time policies and working time organization, especially for men. Secondly, since long-term values and beliefs (traditionalism) are persistent, innovative measures that strengthen gender time equality in traditionalist cantons have to be developed. Thirdly, the role of welfare regimes in shaping gender-specific time inequalities is underrated and needs to be emphasized.

## **References**


Cantonal patterns of gender-specific time-inequalities in paid and unpaid work


# **Child care services – A relevant policy tool to enhance gender equality?**

*Christina Felfe, Rolf Iten & Susanne Stern*

## **1 Introduction**

Gender inequality in terms of labor market outcomes is to a large extent related to the presence of children. While in the OECD in 2000 the employment gap between men and women without children amounted to 11.8%, this gap widens to 22.9% as soon as there is one child present and to 32.3% as soon as there are two or more children around. While employment rates among men, if anything, rise in the presence of children, employment rates among women fall from 73.7% among women without children to 70.6% among women with one child and to 61.9% among women with two or more children.<sup>1</sup> In other words, the increase in the gender employment gap is mostly due to the fact that women stop working once they have children. Gender differences are even more pronounced when looking at the hours worked. In 2009, for instance, 66.2% of all women with children under the age of 15 were working, yet only 44.6% of them worked full-time, 26.1% had a workload of 50–90% (3–4 days per week), and 29.4% worked less than 50%. In contrast, men with children under the age of 15 mostly worked full-time (78.4%). Given the existing evidence on negative consequences of a reduced workload for women's career opportunities (Waldfogel, 1997; Bratti, Bono, & Vuri, 2004; Felfe, 2012), many countries consider expanding the supply of child care services (Kamette, 2011).

Increasing the compatibility of family and career is also on top of the political agenda in Switzerland. This issue has become even more pressing since the Swiss citizens accepted the "Initiative on Mass Immigration" in February 2014 – an initiative which asks for a more restricted immigration. In light of the imminent lack of qualified workers, activation of untapped labor force

<sup>1</sup> http://www.oecd.org/employment/emp/17652667.pdf access on August 27, 2015

potentials has become a political priority. Given the low employment rate among women with children, policies to stimulate employment of this group are heavily discussed. Since the organization of child care constitutes a limiting factor for mother's access to the labor market, expanding the provision of extra-familial child care services is understood as a promising candidate for an effective policy.

Despite the popularity of child care services as a policy tool to enhance maternal employment, empirical evidence whether the availability of child care services indeed stimulates the employment of mothers is scarce. The reason for this is mainly the lack of representative data on extra-familial child care services. Similarly, we do not know how fathers' employment changes due to an increase in the provision of child care services. Thus, it remains an open question whether providing child care services is an effective policy tool to enhance equality between mothers and fathers and thus ultimately between men and women.

One goal of the NFP60 was to analyze the relevance of the provision of child care services as a policy tool to stimulate mothers' labor supply and to foster gender equality.2 In the research presented here the first goal was to provide a representative database on child care services available in Switzerland. A second goal was to study the impact of the provision of child care services on mothers' and fathers' labor supply. For reasons explained below, the particular focus was on the provision of after-school care and thus on the provision of child care services available to school-age children (4–12 years old) for parental labor supply. Studying the impact of care services available for this particular age group is interesting from two points of view: first, there is basically no empirical evidence on this specific subgroup yet. Second, children of this age group are already enrolled in formal education and thus mothers are under no circumstances the sole care provider. As such, prejudices against maternal employment are likely to be less relevant in the case of women with school-age children than in the case of women with preschool children. As a result, child care services may be more likely to make a difference for mothers of school-age children than for mothers of preschool children. And indeed, the results from our empirical analysis point towards a diminishing gender gap in terms of employment: after-school care provision stimulates mothers' full-time employment and fathers' part-time employment (at the expense of fathers' full-time employment). As such, child care provision acts as an effective policy to enhance gender equality.

<sup>2</sup> The National Research Program "Gender Equality" (NFP 60) was a research program funded by the Swiss National Science Foundation with the aim to generate more knowledge about effective gender policies.

The remaining part of this chapter provides some theoretical considerations about the impact of childcare services on parental labor supply and describes the existing empirical evidence. It then continues with a description of the institutional background in Switzerland. Importantly, it provides a detailed regional picture of the provision of child care services – unique data which has been collected in the course of the NFP60 project. Since this data is the first of its kind in Switzerland and thus provides important insights about the current situation of child care services across regions in Switzerland, the focus is on both child care services available for preschool children and for school-age children. We then continue with a description of the data, methodological approaches, and finally discuss the results with respect to the impact of after-school care provision on mothers' and fathers' labor supply. Finally, the relevance of this policy for gender equality is addressed.

## **2 Theoretical considerations and existing empirical evidence**

What are the theoretical hypotheses on the role of child care services for enhancing gender equality? A simple economic model where families aim to maximize their utility helps to derive specific hypotheses. Such a model is, for instance, offered by Blundell and MaCurdy (2008). According to this model, a family's utility is determined by consumption and leisure as well as time spent with children. Consumption is only possible if the family receives an income. Income is generated by work outside the home. As such each family has to decide between time devoted to work outside the home and work inside the home, where child care is part of work inside the home. This decision is generally influenced by the relation between the opportunity costs of work inside the home – in other words, the income generated by work outside the home and hence wages – and the costs of child care. The government can influence this relation not only by regulating the price of child care or granting subsidies on child care slots, but also by levying direct or indirect taxes. The allocation of work inside and outside the home between men and women is determined by the relative wages of men and women. Yet, also non-monetary aspects such as direct utility generated by time spent with children play a role for the allocation of work inside and outside the home and thus may weaken the impact of the provision and the price of child care services on parental labor supply.

What is the empirical evidence so far supporting these theoretical considerations? There is a vast amount of literature providing empirical estimates for the consequences of child care provision for mothers' labor supply, yet only a few studies rely on a convincing empirical strategy to provide causal estimates for the impact of child care services on maternal labor supply. The focus lies here on the latter type of studies. In his seminal paper, Gelbach (2002) exploits strict rules related to the birth date of a child regarding eligibility to attend child care and thus uses quarter of birth as instruments for child care attendance in the US. He finds that the provision of public child care free of charge stimulates employment by 6–15% among married mothers and by 6–24% among single mothers. A series of subsequent studies supports these findings for other countries and other contexts: relying on expansions in child care services as an exogenous source of child care usage, the positive effect of child care for maternal labor supply has been shown, among other countries, for Argentina (Berlinski & Galiani, 2007), Germany (Bauernschuster & Schlotter, 2015), Israel (Schlosser, 2011) and Spain (Nollenberger & Rodriguez-Planas, 2011); other studies relied on the introduction of a price subsidy or a tax for public care as an exogenous source of child care usage and have confirmed the positive impact of child care services for Canada (Baker, Gruber, & Milligan, 2008; Lefebvre & Merrigan, 2008) and Germany (Gathmann & Sass, 2012).

More recent studies, however, find that maternal labor supply is on average rather inelastic to exogenous increases in child care availability. Only subgroups of mothers, such as single mothers or mothers in disadvantaged areas, react positively to an increase in public child care, see for instance Cascio (2009) and Fitzpatrick (2010) for the US, Goux and Maurin (2010) for France or Havnes and Mogstad (2011) for Norway.

Fitzpatrick (2012) discusses the reasons for the lack of consensus in the empirical findings. On the one hand, studies differ in their empirical methodologies. On the other hand, child care policies vary with respect to their institutional and socio-economic contexts. For instance, over the years, the subset of mothers who provide a labor supply in response to an expansion of public child care has potentially shrunk because of increasing maternal employment, delayed childbearing ages, and rising educational attainment. Moreover, the impact of child care supply may depend on the initial level of child care institutions as well as on the alternative child care arrangements (whether those are the extended family, nannies, friends, etc., or the mother).

As discussed in the following section, Switzerland is characterized by a rather low supply of child care and low supply of full-time employment among mothers (despite a relatively high level of overall employment among mothers). Thus, we should interpret the results presented in this chapter as the results of expanding child care in a rather conservative setting where there is low supply of formal child care that substitutes for informal care, which is mostly maternal care.

To the best of our knowledge, there is scant evidence on the impact of care institutions available for school-age children. The only exception is the study by Lundin, Mörk and Öckert (2008). The authors evaluate the effects of a price reduction of child care for 0–9 year old children in Sweden at a time when overall child care coverage was already high (80%). Their results reveal positive effects of subsidized child care on overall maternal employment, not, however, on employment of mothers with school children.

Our study contributes to this field of research by providing evidence first for Switzerland. Second, we provide evidence on the impact of child care supply on the intensive margin of mothers' employment, in other words we provide evidence on the impact of child care supply on the amount of hours worked by mothers. This margin might be particularly interesting if we talk about child care services supplementing care provided by schools. Third, the impact of child care on fathers' employment has been rather neglected in the whole debate. Given the lack of data on prices of extra-familial child care services, our empirical analysis can only speak to the role of supply of extra-familial child care services, but not to the role of the price or the quality of extra-familial child care services. Supplementary qualitative interviews, however, provide us with some insights on parental preferences regarding the organization of child care.

## **3 Institutional background**

Gendered labor division in Switzerland is even more pronounced than in many other OECD countries3 : Only in 15% of families with children (age 0–12) both parents work full-time, in 46% of these families the mother works part-time and the father works full-time, and in 25% of these families, the mother does not work while the father works full-time (see Figure 1).

To understand the formal child care system in Switzerland it is important to differentiate between early child care services on the one hand and child care services for school-age children on the other hand. Early child care services cover the age group 0 to 3 ("Frühbereich"), while after-school child care services cover age group 4 to 12 ("Schulbereich"). Depending on the canton (pre)school starts at age 4 or 5 years old.

<sup>3</sup> For an overview regarding gendered labor division in other OECD countries please refer to Chart LMF1.1.A in the following document by the OECD: http://www.oecd.org/els/ family/LMF\_1\_1\_Children\_in\_families\_by\_employment\_status\_Jul2014.pdf (retrieved on June 15, 2015)

*Figure 1: Gendered labor division in couples (age 18–62) with children age 0–12*

*Source: Own calculations based on* Strukturerhebung 2010*; N=22'284*

While early child care services such as "Krippen" or "Kindertagesstätten" run between 10 to 12 hours a day (between 7 am and 7 pm), after-school care services only cover the additional hours not covered by the half-day school system. The length of the school day and thus the duration of after-school services vary by age: while in the first years, school covers only the morning hours (8–12h), in later years school hours are gradually extended. To cover the afternoon hours, many schools provide optional child care services. While school is free of charge, parents have to pay for the optional child care services.

To promote work-family balance, the Swiss government launched a federal program in 2003.4 This program subsidizes new child care facilities as well as the expansion of existing child care facilities during the first three years after their establishment or expansion. Both public and private providers are eligible for the subsidy. By February 2010, the program had financed 25,000 new child care slots, which corresponds to an increase in child care coverage by approximately 50%, according to the Federal Social Insurance Office in Switzerland (FSIO, 2010). A recent data collection by Infras, Zurich, and the Swiss Institute of Empirical Economic Studies at the University of St. Gallen allows for unprecedented statements regarding the supply of child care slots: in 2010, the average coverage rates of extra-familial care amounted to 11% among children age 0–3 and to 8% among school-age children (age 4–12).

Child care coverage rates vary substantially at the cantonal level, with cantonal coverage rates ranging from 1% to 23% for children aged 0–3 and

<sup>4</sup> The program has been launched on February 1, 2003. It is called "Federal Law on Financial Support for Extra-Familiar Child care" ( "Bundesgesetz über Finanzhilfen für familienergänzende Kinderbetreuung") and is administered by the Ministry of Social Affairs (Bundesamt für Sozialversicherung). Article 1 of the law states the purpose of the program: "The Swiss federation provides […] child care subsidies […] so that parents can better reconcile family life with work and/or education" (own translation).

from 1% to 43% for children age 4–12 years old. Where do the differences between cantons come from? Cantons differ in their policies to support child care provision (see Table A.1 in the Appendix for an overview). For instance, 19 out of 26 cantons explicitly mention extra-familial child care as one policy to support families in their legislation; 17 cantons provide information and counseling to child care facilities that want to apply for federal subsidies; and 15 cantons contribute financially to the provision of child care.<sup>5</sup>

Coverage rates vary not only across cantons, but also within cantons (see Figure 2 and Figure 3). Heterogeneity within cantons comes from municipalities' discretion in the provision of child care, but the degree of discretion in turn depends on cantonal laws. Depending on the legal setup, either the canton alone, the municipality, or both regulate, license, and supervise the provision of child care facilities (please refer to Table A.1 in the Appendix for details).

*Figure 2: Coverage rates of early child care by cantons 2010*

*Source: Own calculations based on population survey 2010 and child care database.*

<sup>5</sup> Child care costs are generally borne by parents. Subsidies (by the canton or the municipality) are available to low-income families and are paid independently from the child care provider (in other words, public subsidies can be used to pay for a slot in a publicly or a privately organized child care institution). The availability and amount of these public subsidies, however, varies greatly across and within cantons. Unfortunately, so far no reliable data on the availability or the amount of public demand subsidies exist.

*Figure 3: Coverage rates of after-school care by municipality in 2010*

*Source: Own calculations based on population survey 2010 and child care database.*

## **4 Data**

Three types of data are necessary to answer our research question: first, data on parental labor supply; second, data on the availability of extra-familial child care services; and third, data on parental preferences regarding extra-familial child care services.

The best available micro data for parental labor supply, in particular due to its sample size, is the Swiss Structural Survey – a questionnaire filled out by a subsample of the Swiss census. In total it contains information on more than 200,000 households. Aspects covered by the Structural Survey are among others employment, mobility, education, and family. As such, we can construct information on the household composition and parental employment as well as socio-demographic information on all household members and the municipality of residence. Yet, the Swiss Structural Survey lacks both other aspects needed for the purpose of our research project: it lacks information on available extra-familial child care services and information on parental preferences.

Given the lack of national child care statistics, an important aim of this NFP60-study was to create a database on the coverage of formal child care services on a cantonal and communal level. As child care is regulated at cantonal or communal levels, child care services in Switzerland differ substantially between and within the cantons, for instance with regard to the different types of child care services, the age of the children looked after in these services or the opening hours. Therefore, a major challenge of the NFP60-project was to produce a comparable data set with the available data from cantonal or communal authorities. The data set distinguishes between early child care services for children aged 0–3 (so-called "Kindertagesstätten" or "Krippen") and child care services for school-age children aged 4–12 (so-called "modulare schulische Tagesstrukturen" and, "gebundene Tagesschulen"). Among other things it contains information about the number of available slots, opening hours, and days in operation. Family day care services (Tagesfamilien) are not included due to insufficient data at a communal level.

The child care data base and the Swiss Structural Survey could be merged based on the information about the municipality of residence. Based on the municipality of residence we could also add further macro data to the dataset – such as the demographic composition and socio-economic situation of municipality. Macro data are provided by the Federal Statistical Office.

To acquire some information on parental preferences regarding extra-familial child care services, we conducted a series of qualitative interviews with parents of children in preschool and school-age. Parents were selected according to a theoretical sampling scheme. As such, the aim was not to select a representative set of parents, but a set of parents with "typical" characteristics regarding our research questions. The following characteristics were considered: inactive; full- and part-time working parents; parents with formal as well as informal child care arrangements; parents with different educational backgrounds; as well as parents from urban, suburban and rural areas. Overall, we conducted interviews with 31 mothers and fathers. Interviews were conducted by telephone using a written guideline. The main focus of these interviews was on the reasons for choosing a particular child care and work arrangement (advantages and disadvantages of different arrangements, practical and organizational questions regarding child care and work) and on the specific needs and desires of parents regarding child care services (type of service, quality, price, opening hours, flexibility, etc.).

## **5 Methodology**

Is the provision of child care a relevant policy tool to enhance mothers' employment and ultimately promote gender equality? A simple comparison of municipalities with a high supply of child care and municipalities with a low supply of child care is not sufficient to answer these questions. The underlying reason is that the supply of child care is likely to be confounded with 1) efforts of the local government to improve the local living and working conditions and 2) preferences of the local population regarding work and family. For instance, the provision of child care may come in a package with other policy tools meant to stimulate employment. In addition, lobbyism constitutes one effective policy tool of citizens to exert pressure on local politicians. Thus, it is likely that parents aiming at participating in the local labor force exert pressure on local politicians to expand child care.

An instrumental variable (IV) strategy helps circumventing potential endogeneity of child care supply with respect to parents' labor market outcomes. This method relies on exogenous factors that are independent of individual or communal efforts to improve the child care system. Such exogenous factors are, for instance, differences in the cantonal laws enforcing the supply of child care. An analysis of the cantonal laws leads to basically two types of cantons: 1) cantons that enforce the supply of supplementary child care explicitly (Berne, Geneva, Solothurn, and Zurich) and ii) cantons, which abstain from doing so.<sup>6</sup> Enforcing the supply of supplementary care may only lead to an improved supply of child care independently of any further conditions favoring the supply of child care. As such, the cantonal laws exert a direct influence on the supply of child care available for school children. Unfortunately no such regulations exist regarding the care available for pre-school children. For this reason, the instrumental variable only serves to estimate the impact of offering child care available to school children on parental employment.

The assumption underlying the instrumental variable strategy is that there are no differences across cantons regarding the determinants of parents' labor market outcomes other than the child care legislations. This assumption is unlikely to hold in general, owing to two concerns: First, cantons differ in their industry structure and thus in their employment opportunities. Second, cantonal laws reflect the preferences of the local population. Thus, enforcement might occur in cantons with better employment opportunities, or in cantons with stronger preferences for policies that enable parents to work. Both of these associations most likely bias the effect of after-school care upwards. The analysis is therefore restricted to confined regions along canton-

<sup>6</sup> This classification refers to the status quo in 2010. In the following yeas further cantons enforced the supply of supplementary care in the course of reforming their school laws.

al borders, in particular to economically *integrated local labor markets* – for details please refer to Felfe, Lechner and Thiemann (2013). *Integrated local labor markets* ensure the following: i) individuals residing in a local labor market (LLM) face approximately the same job opportunities; or put differently, for any two individuals residing in the same LLM, the cost of commuting to each potential workplace must be approximately the same;<sup>7</sup> ii) individuals residing on different sides of a cantonal border exhibit the same preferences regarding the cantonal enforcement of child care supply. In other words, the described restrictions of the sample ensures that contrasts in political choices at the cantonal level do not result from contrasts in political preferences within the LLMs.<sup>8</sup>

## **6 Results**

What was the impact of the cantonal enforcement of after-school care provision on communal after-school care availability? On average, cantonal enforcement shifts the supply of after-school care slots from a level of 3 slots per 100 children to a level of 11 slots per 100 children. Inside the integrated labor markets under study, municipalities on the side of the cantonal border where no enforcement of after-school supply exists on average display a coverage rate with after-school care of 3%. Municipalities on the side of the cantonal border where there is enforcement of after-school supply on average display a coverage rate of after-school care of 11%. In other words, cantonal enforcements on average induce a shift of 8 slots per 100 children. Thus, the results presented below are to be understood as the impact of an expansion of 8 percentage points higher coverage rates (8 more slots per 1000 children). Moreover, it is important to keep in mind that the setting of the study was a setting of a non-exhaustive level of child care supply – thus demand is likely to exceed supply.

<sup>7</sup> For this to be true, we restrict the local labor market to an area of 30 minutes commuting distance.

<sup>8</sup> For this to be true we impose three conditions. First, the population inside an LLM must not comprise the majority of any of the cantonal populations. Otherwise, an LLM's population could determine cantonal laws. Second, inside any LLM, the populations on both sides of the cantonal border should have similar preferences related to work and family. Results of a recent referendum on maternity benefits at the municipality level provide suggestive evidence for this condition. Third, regions outside the LLM should drive cantonal differences in legislation. Again, the referendum on maternity benefits provides evidence on this condition.

What are the consequences of such an increase in after-school care for parents' labor force participation? Results from the instrumental variable estimations are displayed in Figure 4 and 5 below, as well as in Table A.2. in the Appendix. Overall, no statistically significant change in employment status exists, both for men and for women. Yet, a statistically significant adjustment in full-time employment for both women and men can be detected. An increase in after-school care by on average 8 slots per 100 children leads to an increase in women's full-time employment of similar magnitude: full-time employment among women in the sample rises from 4% to 12% on average. Hence, every slot created in after-school care causes one woman to increase her workload to full-time work. In contrast, full-time employment among men in our sample decreases from 96% to 87%. Hence, every slot created in after-school care causes one man to reduce his workload.

To which extent do our results provide evidence in favor of child care being an effective policy to promote gender equality? In fact, we do not find any impact on parents' probability to start working – most likely due to the fact that formal education covers already half a day and thus allows one parent to work full-time and the other work part-time. Yet, we do find an increase in mothers' full-time work due to a reduction in part-time work – an indication that additional hours of child care allow mothers extend her labor supply beyond pat-time work. The fact that we observe fathers to work only part-time might indicate a slight convergence of both genders in terms of provision of work inside and outside the home. In addition, it might indicate that parental preferences in favor of spending time with their children may weaken the positive influence of extra-familial child care services on parental labor supply.

*Figure 4: Employment effects of child Care supply for women with children (age 0‒12)*

*Figure 5: Employment effects of child care supply for men with children (age 0‒12)*

The results of the qualitative interviews gave the following additional insights: dual earner families (either full- or part-time) relied exclusively on a combination of formal and informal child care. Formal child care services were particularly crucial for families in which fathers worked full time and mothers worked more than 50%. Nevertheless, informal child care – provided, for instance, by grandparents or neighbors – represented still a very important prerequisite for the equal participation of both parents in the labor market. Reasons for complementing formal child care by informal child care were mainly the costs of formal child care services and the complex organization of everyday life, especially if a child or care person is sick, and the general lack of time. For parents using mostly formal child care services the main reason for choosing this kind of child care arrangement is that the mother wants to work outside the home or has to work outside the home for financial reasons. Parents which use mainly informal child care – especially provided by grandparents – choose this sort of arrangement mostly for social reasons, such as the win-winsituation for grandparents and grandchildren. When asked what measures would most enhance the compatibility of work and family, answers referred mostly to the quality of formal child care services as well as flexible and family friendly working conditions. All in all, the qualitative results revealed that for the decision to work outside the home, the price and the quality of formal child care services combined with the conditions on the labor market are decisive for parental decisions to join the labor force and thus to enhance gender equality.

## **7 Conclusions**

Many developed countries are currently debating policies that enhance gender equality on the labor market. One promising policy to stimulate women's and in particular mothers' labor market participation seems to be the public provision of child care services, both at the early childhood level as well as at the school level, complementarily to school services. Switzerland, for instance, launched a federal program in 2013, which provided subsidies to new or expanding care institutions. Germany is currently debating to extend its school system and to offer an increasing amount of all-day schools. Regarding maternal employment and female career opportunities, this investment might pay off: Our research provided insights that each newly created after-school care slot caused one more mother to work full-time, as opposed to not working or working part-time. At the same time, each newly created after-school care slot caused one father to reduce his workload from full-time to part-time work.

While an expansion of the child care system seems to pay off in terms of enhancing mother's labor supply and to support fathers in spending more time outside work, the general equilibrium effects are still unknown. To generate more research that allows for better policy recommendations more data on child care services – in particular longitudinal data on child care services – are necessary.

## **References**


## **Appendix**


*Table A1: Cantonal regulations regarding child care services*

*Source: Internet platform Beruf und Familie (dated 2008)*


*Table A2: Instrumental variable estimates* 

*Note: \* significant at the 1% ; \*\* significant at the 5%; \* significant at the 10% significance level. Above estimates are weighted averages of the IV estimates for each LLM (LATE). The underlying weights correspond to the number of compliers in the respective LLM. The instrument is based on the enforcement of after-school care supply in the cantonal law. The sample corresponds to 10,875 observations for women and 10,133 observations for men.*

# **Care** *and* **education? Exploring the gendered rhythms and routines of childcare work**

*Julia Nentwich, Franziska Vogt & Wiebke Tennhoff*

## **1 Introduction**

Early childhood education is a female dominated occupation (Cameron 2001). Within Europe, more than 95% of the trained childcare workers in nurseries are women, although the numbers of male childcare workers vary from 9% in Norway to less than 1% in Austria (Rohrmann et al. 2008: 4). Work in early childhood education provides little career perspectives and mobility, and is characterized by low wages as is typical for a female dominated profession (Meyer 2006; Nissen et al. 2003).

This imbalance of women and men in childcare work is nowadays also depicted as problematic. Women's numeric dominance and men's marginalization are perceived as especially problematic for boys as their needs and interests might not entirely be met (Rohrmann 2009). Men caring for small children are becoming more visible as more and more fathers assume a more active role (Nentwich et al., in prep.). As a consequence, attracting more men into the field has become an important objective for gender equality policies and initiatives (Faulstich-Wieland 2011; Buschmeyer 2013). However, such a change seems to occur only at a low pace.

Taking the gendered situation of the occupation as our point of departure, the research project on '(un)doing gender in nurseries' sets out to examine how childcare work is reinstated as gendered work. From the backdrop of the theory of 'doing gender' (West and Zimmerman 1987; Nentwich and Kelan 2014) we conceptualize gender as something that is done in everyday practices of 'doing the job' (Leidner 1991). In an organizational context this means to focus on those practices that establish and stabilize what Acker (1990) coined as the 'gendered organization', the interaction of bodies and identities, symbols and images, gendered divisions of labor and gendered interactions. In addition, those practices form the image of the 'ideal worker' – an image of the characteristics ideally displayed by the person performing the job. The 'ideal worker' is gendered too.

However, over the last decade, gender research in organization studies has further explored the relevance of gender performativity (Butler 1990). For instance, Bruni, Gherardi and Poggio (2004 and 2005) emphasized the importance of both material and discursive practices for the everyday practices of doing gender. From such a perspective, both the 'ideal worker' and gender are produced in the very same 'doing'. In conceptualizing gender as a social practice both, the 'situated practices of doing gender and institutionalizing specific forms of gender' (Gherardi and Poggio 2007: 19), are made relevant for our understanding of the persistence of gender inequality. Gender and organizations are (re-)produced by specific social practices that assemble discourses, materiality and identities.

The analysis we are presenting in this chapter tackles in particular the role of organizational rhythms and routines for this assemblage of doing gender in childcare work. Rhythms and routines in general play an important role in stabilizing social practices (Orlikowski and Yates 2002; Staudenmeyer, Tyre and Perlow 2002). They are rather powerful as they are taken for granted and not easily reflected critically. In our empirical analysis we started out analyzing how gender enters the actual work of childcare workers on an everyday level via the material and discursive practices they are engaging in. In the specific ways of how things are done, different facets of a gendered 'ideal worker' are made relevant (Acker 1990). Our analysis of 'doing gender' in this chapter hence moves beyond the idea of gender as limited to the identities of individuals or bodies, but as a powerful resource for the normative formation of the (gendered) ideal worker and, as a consequence, the ideas available on how to do the job properly.

The chapter is organized as follows: First we explore the relevance of the discourses of 'care' and 'education' and how these discourses historically construct early childhood education as a gendered occupation. Second, the ethnographic research design and methodology of our empirical investigations of the everyday rhythms and routines at four nurseries are presented. In the results section we compare these four organizational case studies. We show how the everyday practices of doing childcare work either prioritize the discourse of 'care' or the discourse of 'education', thereby constructing different facets of the 'ideal worker' as either enacting 'the good mother and housewife' or an 'early educator'. In the discussion, consequences of the two competing 'ideal workers' for doing gender as well as for gender equality and the further development of early childhood education are drawn.

## **2 Early childhood education between 'care' and 'education'**

The history of early childhood education emphasizes the relevance of the discourses of 'care' and 'education' for the gendered connotation of childcare work. Both discourses were positioned in a binary opposition. During the 19th century, kindergarten, nurseries ('Bewahranstalten' – literally institutions were children were 'kept') or preschools ('Kleinkinderschulen' – schools for small children) were introduced (Rossbach and Grell 2012: 333), often as a provision for children of working class families as both parents worked in factories. The names of these institutions illustrate the different discourses of 'education' and 'care' in early childhood provision. Institutions of early childhood education and care were seen as a lesser evil, as children should be cared for while their mother works in a factory, but as inferior to motherly care in the family. In order to counteract the idea that kindergarten would actually harm a child if he or she attends the kindergarten too early or for too many hours a day, kindergarten and nurseries were depicted as being like a traditional family home with a caring mother (Friis 2008: 23).

Within Europe, Rossbach and Grell (2012) distinguish two types of early childhood education and care according to the emphasis on either 'care' or 'education': the early education type, for example in France, and the kindergarten type emphasizing the social pedagogy tradition, for example in Germany. Within the Swiss Cantons, this contrasting focus is also reflected by the language divide. The historic development of the kindergarten seems to leave the kindergarten shifting between the poles of 'social' vs 'pedagogical' and 'private' vs 'public'. Both poles are clearly gendered, connecting the 'social' and the 'private' with 'femininity' and 'pedagogical' and 'public' with 'masculinity' (Rabe-Kleberg 2003: 31). Whereas the German kindergarten still follows the social pedagogical model, the Swiss kindergarten is now part of the education system. The Swiss kindergarten has developed from local initiatives supported by women's charitable organisations in the early twentieth century to a fully integrated part of the Cantonal education system, with defined curricula and the requirement of professional training at tertiary level.

Nurseries in Switzerland are not part of the education system but of the social welfare system and only loosely regulated. Swiss nurseries are mostly privately run and financed. Whereas teachers in kindergarten are qualified at tertiary level, the childcare workers in nurseries obtain a professional qualification in a three-year apprenticeship at secondary level, thus emphasizing a hierarchy between 'education' and 'care' as well as work with older and younger children. While kindergarten focuses on the two pre-school years, nurseries cater for children from the age of three months to six years. Nurseries are still conceptualized as a substitute for childcare provided by the family, mostly the mother. For instance, the media discourse emphasizes societal and economic benefits resulting from female employment rather than the pedagogical and developmental benefits for children (Lanfranchi and Schrottmann 2004: 10). Consequently, care is at the centre of nursery work, childcare workers do at the nursery what the caring parent would do at home, they follow homely routines of meals, sleep times, visits to the playground etc. The emphasis on care is also noticeable in the regulatory requirements. The guidelines of the Swiss Nursery Association (SKV) define requirements for structural aspects such as room size, qualification of personnel and staff-children ratio. A pedagogical concept is also requested for the permission of running a nursery, but there are no requirements stated regarding its content and the pedagogical quality (Lanfranchi and Schrottmann 2004: 19). While care aspects are defined precisely, criteria of education quality remained vague.

However, in recent years the discourse of 'education' has gained importance. Nurseries have joined the programme of 'infans' (Laewen and Andres 2011) and call nurseries 'Bildungskrippen' (educational nurseries) (Thkt 2014) in which childcare workers are trained to observe and document children's 'Bildungs- und Lerngeschichten' (stories of education and learning) (Leu et al. 2007). Nurseries are encouraged to adhere to a curricular framework (Wustmann, Seiler and Simoni 2012) and to assess and develop the quality of nursery provision, including learning activities. Large scale quality assessments found a medium to high level in the interaction and relationships between childcare workers and children in nurseries, but only medium to low level of seizing learning opportunities and providing educational impulses (Fried and Briedigkeit 2008; Pianta et al. 2005). Focussing on quality, the discourse of education enters the discourse of 'good childcare', and the demand for quality serves as an impulse for innovation and further development (Vogt 2015). Quality in early childhood education, or 'good childcare' therefore integrates 'education' and 'care'.

While both, the discourse of 'care' and the discourse of 'education', are made important for defining what counts as 'good childcare' and what qualifies as a 'good childcare worker', the discourse of 'care' is dominating. 'Care', even within 'education', is often connected with motherliness and femininity (Vogt 2002). This connection was enforced by the women's movement around 1900 to secure women's access into professional work (Evers 2012: 492). Up to the early twentieth century, teaching was considered a male occupation. The argument was put forward that women are better suited for educating and teaching young children because of their natural ability to be a mother and to provide motherly care. The discourse of 'care' was instrumental for the entry of women into a paid occupation. Depending on the economic situation and the shortage or surplus of teachers, women were called to work in schools and kindergartens or asked to leave paid work (Crotti 2005, 2011).

Work in early childhood is dominated by an understanding of providing care, whereby care is strongly connected to mothering, ensuring hygiene and providing food. This connection of 'care' and 'femininity' has immediate consequences for the position of men and women working in the field. As Vogt's (2002) analysis of interviews with kindergarten and primary teachers indicates, an understanding of caring as mothering excludes men from the work and renders the work as women's natural work. As a consequence, men's position in this field is often contested, as for instance people regularly react with surprise when coming across a man in the nursery. Men in the field have to deal with the perception of femininity of the 'ideal worker' (Nentwich et al. 2013; Cross and Bagilhole 2002).

Our review shows that although 'care' and 'education' are not necessarily oppositional concepts and care can also be seen as part of the professional work with young children (Vogt 2002), the discourses of 'care' and 'education' are clearly gendered along a binary opposition. While the discourse of 'care' is closely connected to femininity, the discourse of 'education' is potentially open to both masculinity and femininity. Only in its connection to the 'early years' does an educator's gender become more narrowly defined as feminine. Furthermore, the discourse of 'education' also serves as a powerful reference for men when legitimating their position (Nentwich et al. 2013; Tennhoff, Nentwich and Vogt 2015).

Assuming that gender and organizations are (re-)produced by the social practices that are assembling discourses, materiality and identities (Bruni, Gherardi and Poggio 2005), these two discourses play a major role in the everyday practice of childcare work. While the discourse of 'care' establishes the image of an ideal worker who is oriented towards the ideal of the 'good mother and housewife', the discourse of 'education' rather emphasizes the importance of 'enabling learning' and hence an ideal of an 'early educator.' Both the ideal of the 'good mother and housewife' and the ideal of the 'early educator' strive for defining what counts as 'good childcare'. And both concepts contain specific requirements with regards to doing gender. While the image of the 'good mother and housewife' cannot be separated from 'doing femininity', the 'early educator' is potentially open for both performances of femininity and of masculinity. Analyzing how they are made relevant in the everyday rhythms and routines provides us with some insights in how the construction of early childhood education as gendered is stabilized and hence inequality is perpetuated.

## **3 Methodology**

The empirical research project was conducted in Switzerland's German-speaking part. Our ethnographic research design (Gherardi 2012; Eberle and Maeder 2010; Gobo 2008; Dellwing and Prus 2012) involves multiple qualitative research methods: 20 interviews with nursery managers, 18 interviews with male and female childcare workers, photographic documentation of the spatial arrangements, and (video) observations in four nurseries during 15 days in total. Eight of the nurseries in our sample employ one or more fully-trained male childcare worker; this enabled us to conduct ten interviews with men and eight interviews with their respective female colleagues. Observations of everyday practices were carried out in four of the nurseries that employ at least one fully trained male childcare worker. Following the procedure of theoretical sampling (Glaser and Strauss 1967), the four nurseries were selected on the basis of sampling criteria such as differences in the children's economic background and the nursery's spatial arrangements. An important sampling criterion relevant for the analysis of the discourses of 'care' and 'education' was the pedagogical concept. While the nurseries *blue* and *yellow* only provided very short specifications on their websites, the concepts of nurseries *red* and *green* are more elaborated.

*Table 1: Organizational case studies and number of men and women in the teams*


Teams of two researchers spend a total of 15 days in these four nurseries, taking turns in observing and taking fieldnotes and following childcare workers with the video camera. After each day, we discussed and shared our fieldnotes and wrote short memos. From the four nurseries, a total of 50 hours of video material was collected. The material was coded using Atlas.ti. We developed a descriptive coding scheme identifying the spatial situation (doll corner, building blocks etc.), the childcare workers involved (female or male childcare worker), the activity involved (guided or unguided play, tidying up, meals, sleeping times). Coding the videos helped us to reduce the material to those scenes, where gender was made relevant. Again, we wrote memos describing and interpreting these scenes.

The photographic documentation of the spatial arrangements is analyzed elsewhere (Vogt, Nentwich and Tennhoff 2015), as well as the self-positioning of the male childcare workers in the interviews (Nentwich et al. 2013; Tennhoff et al. 2015). In this contribution, our focus of analysis lies on the organizational rhythms and routines. We draw on material from the video analysis, fieldnotes and interviews. In a contrasting reading of scenes in the four nurseries, we developed 'thick descriptions' (Geertz 1977) of organizational rhythms and routines. All names are pseudonyms.

## **4 'Care' and 'education': Preparing food in four nurseries**

A central and common task in all nurseries is preparing food for lunch or snacks. Snack- and lunchtimes are also major markers for the daily rhythms and routines. In the following section, we present a contrasting analysis of four scenes of preparing food in the four nurseries. The analysis shows that preparing food can be done differently and with contrasting consequences for the relevance of the discourses of 'care' and 'education' as well as the gendering of the ideal worker. While activities related to 'care' as well as activities related to 'education' proved to be equally important for the everyday work in all four nurseries, they differed significantly in the way these activities were organized, or *how* these aspects of 'care' and 'education' were made relevant. While in nursery *blue* and *yellow*, 'caring' activities such as preparing food, washing hands and so on dominated the daily routines, nursery *red* and *green* had more emphasis on aspects put forward by the discourse of 'education': pedagogical objectives and children's learning activities.

#### *4.1 Nursery* **blue** *and* **yellow:** *The discourse of 'care' and the dominance of household requirements*

In nursery yellow, we noted the following routines: It is close to lunchtime. The children are playing in the entrance hall; the female childcare worker is with them. Maurin, the male trainee, is laying the table. A colleague from another group is helping him to get things done in time. Maurin sorts out little symbols for the children that assign them a specific seat at the table. Everything is prepared for the children. Salad is already served on the plates, water is poured into cups. While Maurin is pouring water, the following interview takes place:

Interviewer: Is there a certain reason, why you prepare everything completely before the children are coming in?

Maurin: Hmmm just as a preparation. Because afterwards, we only have got 20, 25 minutes to eat, half an hour at the most, and it requires a lot of time, when you have to serve every child.

Interviewer: Why are there only 25 minutes to have lunch?

Maurin: Because they need to have their nap afterwards, they have to brush their teeth, and I need to have sent all the crockery to the kitchen, yes it all is, in terms of the timing, not stress, but tight, I then have to clean, then teeth brushing with the kids, and now I have not managed to get the mattresses ready for sleeping, It is mainly because of the kids, because they go for their midday nap afterwards, and it also is the case, that they fall asleep while sitting at the table, hmm.

(Video transcript, nursery *blue*)

In his description, Maurin refers to the schedule of 'to dos', which appears to be stressful, even though he emphasizes that it is 'not stress, just a bit tight with time'. Bringing the crockery to the kitchen in time as well as having to clean dominates the worker's activities and results in a mealtime routine where children are not part of laying the table and doing the dishes. Feeding the little ones is seen as a chore. What Maurin describes here is what we also observed in nursery *yellow*: household activities are the dominant element for structuring the work priorities and daily schedules.

Household requirements dominate not only the organization of meal time routines but also each day's time structure. This finding is also mirrored in childcare workers' 'articulation work' (Corbin and Strauss 1993; see Gherardi 2012: 12). For every transition from one activity or task to another, they are negotiating who is going to do what. Talking about task allocation takes quite some time out of their total talking time amongst the team. This leaves a general impression of being busy. As a result, the day's structure is divided into tiny sequences. Play and other pedagogical activities are cut short through hygiene tasks and seem squeezed in between rituals. These rituals are designed in such a way, that all children have to participate and join in the same way and at the same time. Very often we observe children waiting for others to do something, then do it themselves, then wait again. In consequence the time available for free play and for interaction among the children is rather limited.

Most of the activities are done in the group and the daily schedule defines what to do and when to do it. For instance, nappies are changed at a certain time, and not when it is necessary in a specific case. This results in a high proportion of waiting times for the children in nursery *blue*. They wait for others getting done with something, sitting on a bench in the entrance hall or in a circle at the floor. This general idea of having to do things together and not individually is even more dramatically in nursery *yellow,* as the following video memo shows. Here the dominant logic of 'everybody does the same at the same time, no matter if it is individually necessary' is often enforced, even against children's will and at the cost of escalating conflict.

Leo gets reminded several times by Michel, that he should also sit in the circle for singing. Leo is fidgety and repeatedly leaves the circle. As Leo turns away from the circle completely and starts to play with a snail, Michel gets up and puts the snail away. Leo shouts 'no'. Michel warns him that he will 'get the chair.' Leo comes back into the circle and is quite agitated, because Michel has put the chair in the vicinity. (The chair seems to be a sanction: when children are misbehaving, they need to sit on the chair).

#### (Video memo, nursery *yellow*)

Not only joint activities such as circle time are used in a disciplinary way, but also the choices of play during times of seemingly free play times. In nursery *yellow*, playing in the home corner is often not allowed although children ask for it. Instead, childcare workers suggest jigsaws, board games and art activities, which are done quietly sitting at the table. Childcare workers' suggestions for play and other activities are chosen according to time of the day, they need to fit into a general scheme of activities and should not lead to too much noise. Often, activities are offered to reinforce normative behavior, such as keeping things quiet, and not for a purpose of a more general learning.

Preventing chaos and preserving order are important objectives for the childcare workers: this is very tellingly captured by a sign put up in the entrance hall of one of the nurseries we analyzed for their spatial situation. With the general slogan *'orderliness is a necessity, chaos oh no!'* childcare workers and eventually also parents are reminded to tidy up shoes on the designated shelf, to put away dry clothes in the wardrobe and to only leave seasonal clothing in the nursery. However, preventing chaos and preserving order is not only a norm for childcare workers' behavior, but has also consequences for what counts as acceptable behavior for children. While childcare workers emphasize in the interviews that children should be able to move around, jump, and shout, such behavior is rarely accepted in the indoor-facilities of the nurseries, as in the following scene in nursery yellow:

Playing becomes increasingly wild and a boy turns a chair over several times. Michel intervenes the second time and asks, whether this is a good idea and he warns them to be quieter. As it is not getting quieter he repeats his demand and warns them, that he will finish the play if they are not quiet immediately. Few moments later the head of the nursery comes to the door of the room and says in a stern voice, that the smaller children are sleeping and that it is too loud. Michel removes the chairs 'we have to tidy up, I told you'. He asks the children to go and get board games and to sit at the table.

(Video memo, nursery *yellow*)

Rules are followed in a rigid way here. Especially in nursery *yellow*, children are often admonished to behave well and refrain from being noisy or fidgety. Although children are involved in some of the household chores, for instance bringing back their plates to the sink after finishing lunch or helping with the cleaning tasks in nursery *blue*, the tasks are mainly done by the childcare workers in an efficient way. Household activities are mostly separate from the task of enabling children's learning. In everyday rhythms and routines, these two nurseries treated 'care' and 'education' as rather separate activities and hence re-enforced 'care' and 'education' as binary oppositions.

The dominance of keeping things quiet and orderly is also mirrored by the nursery's aesthetics. Order is achieved in a systematic way for instance by using boxes and little signs in the cloakroom, extra boxes for every child at the entrance of the dormitory to deposit clothes before going for a nap, or for putting away tooth brushes in the bathroom.

While being able to retrieve the specific tools or accessories of every child is of course an important necessity when sharing organizational space with 20 or more children, it was the strong emphasis on this specific task and the aesthetics of orderliness that shows the dominance of household duties.

Furthermore, children are often asked to do things 'nicely'. With demanding that things are done 'nicely' childcare workers emphasize the order and final product but not the potentially messy learning process. As in the following example, Michel motivates children to focus on the 'nice' end product – a tall tower that stands without tumbling – and rather holds them back from what could be depicted as the learning aspect of exploring its stability, experimenting with the force of gravity or statics.

Michel constructs a high tower using building blocks with the children. The children are very excited, noticeably as they speak with loud voices. Michel is also very immersed in the task, he repeatedly tells the children to be careful to make sure that the tower is not falling down, as 'it is such a nice tower'. He does not encourage children to take part in the building process. He mentions that the children should not try to put more blocks on the tower, as the tower might fall down. This surprises me a bit, as there would be spaces to put additional blocs also at the bottom of the tower. Michel suggests to keep the tower standing there, certainly until snack time. He then mentions repeatedly, that the tower is so nice.

#### (Video memo, nursery *yellow*)

Overall, the everyday rhythms and routines are (re)producing a dominance of household tasks and hence the discourse of 'care'. Mealtimes and snack times, cleaning up after meals and activities are important activities for the childcare workers and structure their day. Time is enacted as 'chronos' here, giving priority to its linearity emphasizing effectiveness and efficiency (Orlikowski and Yates 2002). Household chores are primarily tasks that are done by the childcare workers. Even if children are involved in these activities, the scenes are handled in a way that prioritizes the efficient and effective handling of the situation rather than enabling the children to learn to handle water, plates, etc. By keeping the discourse of 'care' as separate to the discourse of 'education', household activities are not depicted as suitable for enabling children's learning. Furthermore, the relevance of learning is rather marginalized: Although many opportunities for learning are created, the priority for setting the schedule is given to household activities. It is the household and care chores that define the time frames and what needs to be done by everyone at a set time, whereas learning is 'happening' in between. Interestingly, in nurseries *red* and *green* we found a different way of organizing time and routines that we describe subsequently in the following paragraphs.

#### *4.2 Nursery* **red** *and* **green:** *Preparing food as an activity that enables learning*

Daily rhythms and routines are organized differently in nurseries *red* and *green*. Although care and catering for the children are important activities here, too, these activities are not dominating. The main objective for structuring the day in these two nurseries is the need to create as large units of time as possible for children's play and learning activities. Time is not as much understood as linear, but being build up from 'events' and their particular sense making (Orlikowski and Yates 2002). Enabling learning and hence the discourse of 'education' is setting the scene here, while household necessities are still being done. There seems to be a balance of both, household requirements and children's learning, instead of one dominating the other.

One example of how 'care' and 'education' work are organized in nursery *red* is a scene, where three children between two and three years old and the female childcare worker are baking bread. Baking bread is a meaningful household activity as this is the bread that children would eat over the next days. Every week, one of the two groups in the nursery is responsible for baking bread.

Stefanie cleans up the kitchen table and prepares all ingredients and tools for making the dough. She lets the children know that it is about time to bake bread. It is a voluntary activity for the children; they are free to join in. Two boys and one girl come to the kitchen (one girl and boy are about three years old and the other boy is two). Stefanie gets them aprons and helps them putting them on. She asks the children: What do we need for baking bread? The children are very eager to answer her questions and point out the ingredients. Now they start to prepare the dough. Children take turns in adding things into the bowl. Stefanie takes her time to give everybody a chance to participate. She points out the flour dust that is in the air after pouring flour into the bowl. The atmosphere is calm and concentrated. Stefanie starts to sing while working the dough, the children first listen, copy her movement with the dough with their hands on the table, and then join in singing. Stefanie sends one girl that has licked off the dough on her hands to wash her hands. All children are now joining in and work smaller parts of the dough. The children are 60 minutes at the table, working concentrated to bake bread.

(Fieldnotes and video memo, nursery *red*)

What is striking in this scene is that baking bread, although a necessity for the upcoming meals, is organized as a learning activity for the children. It is not the product and its efficient production that is at the centre, but getting children engaged with the activity of baking bread, learning about the ingredients, smelling the ingredients, and feeling the dough.

The nursery's schedule reflects the general principle of facilitating the maximum possible length of free and uninterrupted playtime for the children. It is the aim that children are deeply engaged in play and interact with other children in the group without being directed by childcare workers.

We are keen to enable really long and uninterrupted play sessions as far as possible and we offer less activities initiated by us, like circle time and such. That is what we do: we gather them before lunchtime, for example. But we realize that we do not want to disrupt them with things, which are our things, where we think they would benefit. They are only able to really enter into play when we leave them to it for as long as possible.

(Interview, nursery manager, nursery *red*)

This is similarly important to Reto, the male childcare worker leading the outdoor group of nursery *green*. Spending the day outdoors in the forest or the nursery's spacious garden, the most important objective for organizing the day is children's time for play. Both nurseries, *red* and *green*, organize time around a few fixed times such as breakfast and arrival, lunchtime with a brief group activity just before lunch, and a calmer part following after lunch. The snack times in the morning and the afternoon are offered as voluntary snacks: children are merely asked whether they are hungry. Children who are hungry can come and fetch something to eat, usually some fruits or little biscuits. However, it is not organized as a group event where everybody has to join. Similar to the scene of baking bread children are often actively involved in preparing snacks, such as cutting apples or bread, thereby learning how to handle a knife without getting hurt.

They can only develop their play if we leave them to it as long as possible, as, already the meals are setting structures. So breakfast, mid-morning snack, lunch, afternoon snack. These are the structures which are set, and we would like to keep them free in their choice of play, of playmate and of duration of play.

(Interview, nursery manager, nursery *red*)

While our example of baking bread in nursery *red* showed how household activities can be transformed into sites of learning, the following scene highlights the cooking activity as an important aspect of organizing space. Reto, the male childcare worker in nursery *green,* stresses in the interview the importance of preparing the spatial situation *'as activating and stimulating, space is the third teacher,'* thereby drawing on a major concept of Reggio pedagogics (Knauf 2000). The nursery's space – in his case the garden and the outdoors – have to be organized in a way that allows children to play freely and without being interrupted.

Reto: I concentrate on the environment, I cannot focus on individual children, this is intended, not because I would not be able, but I am the helping hand for the children.

Interviewer: What do you mean by preparing the environment?

Reto: The forest is already prepared in a quite good way, but to see where the limits are, where there is broken glass, but also which limits is the environment imposing on us. This is exciting, I can demonstrate, I can show what one can do, but I prefer the most, when they manage by themselves. (…) Preparing the environment also means mentally preparing, what is happening today in the forest. Sometimes there are days, weird people are walking around here, that I do not like and then I keep the children nearer, and then there are days, where it is totally relaxed.

(Video transcript, nursery *green*)

Relying on the spatial situation as stimulating as well as limiting children's activities also has consequences for the childcare worker's role. He has to be present and approachable for the children and able to observe what is going on, but not interfere with children's activities. In a video sequence, Reto points out the importance of keeping the balance between risk prevention and enabling children to play freely. While his primary focus is on leaving children to explore and play independently, he would rather go and check as soon as he has the feeling that something is going wrong.

Reto is pottering round with the kitchen utensils and then turns to Wiebke (researcher) and begins to explain: Now it suddenly got very quiet, them there in the far back, they were playing completely quietly. Now I need to see, where they are at, they were really quiet during play. And if we would even come nearby, we would disrupt them. We know, they are at the back over there, but when I get a strange feeling, and then I go quickly to have a look.... Controlling is sometimes better.

#### (Video transcript, nursery *green*)

What strikes us as interesting also from our observations is how he succeeded in organizing this balance between leaving the children on their own as well as them being safe. By making it his job to prepare lunch at the open fireplace, he manages to be part of the scene and being approachable for children, however, he is not participating in children's play. In that sense, he is making use of the necessary household activity of preparing lunch for educational means: creating a protected but nevertheless free space for children to learn. In nursery *green*, cooking became an important enabler of children's free play in a protected environment. Thus, taking care of household requirements and creating learning opportunities are not treated as in opposition, but as enabling one another.

Of course, nurseries *red* and *green* are relying on time structures to organize the rhythm of their days too. For instance, in nursery *red* the daily schedule is displayed on a complex chart in the group's kitchen. The day is organized in small sequences. The chart maps out childcare workers' activities such as cooking, changing nappies, brushing teeth, but also preparing a group activity or holding a meeting. Stefanie, the group lead childcare worker explains that this schedule is important to organize the work and to let trainees and apprentices know what they should do. Furthermore, childcare workers take it in turn to be responsible for organizing the duties and activities for one day or for half a day. Like that, tasks are allocated clearly and childcare workers do not need time for organizing (nursery *red*, interview with Stefanie).

Despite this rigid-looking plan, the everyday schedule doesn't give the impression of being rigid or split in tiny micro-activities. Furthermore, we never observed the micro-articulation work that was inevitable in nursery *blue*. In the everyday practice the schedule is handled rather flexibly and can be adjusted spontaneously when needed. The planning and collegial but structured work allocation provides a structure for the smooth running of the day. There is less need to negotiate between workers. This enables childcare workers to focus on the pedagogical aspects of the work.

Furthermore, as the following example from nursery *green* shows, even lunchtime as an important corner stone in the everyday schedule can be changed and done differently if necessary.

I come in the morning and I have my routines, and so, that is me, that is how I am, I improvise and am spontaneous. So for example once I noticed a lot of chestnuts on the ground on my way to the nursery. So I went collecting chestnuts with the children. Of course it took some time until the children were all ready and dressed, and then while walking, they needed to look at everything, okay, but we went. I had planned something entirely different for the afternoon, and the food was also bought, but then, it took ages, you would walk it in five minutes, but we almost took two hours to get there, to pick up chestnuts, to look at every single chestnut, to compare them, to sort them, and all what is part of that, and then it was lunch time and I just knew I need to now order pizza, (laughs) well that is now maybe not very good, and not very creative, but it was the quickest to do and I also had the staff to cover it. And so we sat on a rock there and ate pizzas and later in the afternoon, we made our way back.

(Interview, childcare worker Reto, nursery *green*)

This general flexibility is also mirrored in interaction with children. Childcare workers would stop an activity that is no longer interesting and reorganize their plans according to children's needs. As a matter of fact, children are also allowed to get lost in an activity and to decide autonomously what to do. As a consequence, we observed long sequences of free play in nursery *red* and *green* without childcare worker's interventions. Careful scripts for transitions are crucial here. For instance, one of the childcare workers in nursery *red* would leave the lunch table earlier, go to the bathroom and start helping children brushing their teeth, washing their hands and changing nappies. Children who have finished their meals can get up, tidy their dishes away and leave for the bathroom to get ready for play or a nap without having to wait for others. The same can be observed with the transition for going to the garden. One of the childcare workers would check if the gates are locked and stay outside, receiving the children ready to play in the garden, while the second one would help children putting on their shoes, hats and sun cream. This way of organizing transitions allows the children to get ready in their own pace and not having to wait for others, and already start with their activities. Childcare workers on the other hand are rarely in need of disciplining children for not participating or not waiting patiently.

Children in nursery *red* and *green* are free to choose in which room they engage in what kind of activity. Children's play is most often self-organized. Childcare workers would decide on the basis of a child's playing behavior and interests what to do. They place a great emphasis on enabling long sequences of uninterrupted activity. This is also true for guided activities. In one scene in nursery *red*, Stefanie reads books to the children. She does that in such a performative and interesting way that children are mesmerized and highly attentive. Participation is created by capturing children's attention and interest and not by disciplinary means, also allowing for individual freedom. For instance, it is totally acceptable for children to leave the book-reading scene for a moment and come back later. While tidying up and keeping things in order is also an important issue in these nurseries, the main focus is on children's competences to install this order by themselves. Tidying up the toys they have been playing with is used as a learning event.

Also the issues surrounding order are enacted in a different way. In nursery *red* a system of little colored dots is used to inform children in the creativity room, which material is freely available and for which material they would have to ask a childcare worker. A similar rule applies in nursery *green*, here children have to ask the childcare worker when they are interested in material that is kept in the construction trailer in the nursery's garden. All other things are freely available. The order that is installed in these two nurseries relies very much on self-organization, children's autonomy and getting children to implement order on their own initiative. Again, tidying up or doing things in a specific order ‒ which is considered an important objective in all the Montessori-based activities in nursery *red* ‒ is important for children's learning, and not because things might look nicer.

In contrast to the dominance of the discourse of 'care' in nursery *blue* and *yellow*, nurseries *red* and *green* are prioritizing 'education' and 'learning' as the major organizing principles. Although care, hygiene and healthy diets are important here as well, they are not dominating the everyday rhythms and routines. The daily schedule is organized to provide children as much space for play and other autonomous activities as possible. Rituals and routines are critically reflected for their support of children's learning and often reduced in order to meet the objectives of serving children's and childcare worker's needs.

## **5 Discussion**

The discourse of 'care' and 'education' have shown to be relevant not only for the gendered construction of the occupation in a historical perspective, but also for the making of the 'ideal worker' in the everyday practices of childcare work. With our ethnographic perspective we especially focused on the everyday rhythms and routines in four nurseries in German speaking Switzerland. We found profound differences between our organizational case studies. In nurseries *blue* and *yellow* the dominant organizing principle builds upon the discourse of 'care'. Organizational rhythms and routines are prioritizing the discourse of 'care' and thereby marginalizing the discourse of 'education.' This was different in nurseries *red* and *green*. Here the demands of 'household' and 'learning' where no longer seen as in a binary opposition, on the contrary, household activities were used as learning opportunities for children. It can be concluded, that although both discourses, 'care' and 'education', are stressed as highly relevant for childcare work, they are prioritized differently in the everyday practices across organizations.

Our observations in nurseries *blue* and *yellow* revealed that the two aspects, 'care' and 'education', were perceived as opposites here. Childcare workers seemed to be stressed, as they didn't have enough time to both lay the table *and* interact with the children in a meaningful way. Managing chaos and noise was a major challenge in these nurseries and children were often admonished and disciplined. On the contrary, nursery *red* and *green* seemed not to have problems around noise, order and behavior. Their approach of using household activities as an opportunity for learning enabled them to provide both 'care' *and* 'education' in early childhood education.

These two major principles in organizing early childhood education in nurseries lead to different concepts of the 'ideal worker'. While in nursery *blue* and *yellow* childcare workers feel obliged to secure order, to prevent chaos, and to produce products adhering to a certain aesthetic, they enact the ideal of 'the housewife'. In contrast, nursery *red* and *green* encourage the workers to create opportunities for learning, to be attentive, to observe children's development and to turn everyday situations into learning events, therefore fostering the ideal of an 'early educator'. With its strong connection to femininity, the 'ideal housewife' reinforces the femininity of the occupation, while the 'early educator' remains potentially open to both women and men. As our in-depth analysis of the interviews with male childcare workers shows in greater detail (Nentwich et al. 2013; Tennhoff et al. 2015), both ideal workers are frequently referred to. However, while the ideal of the 'early educator' was frequently cited by the interviewed men in order to produce legitimacy, the 'housewife' ideal either remained as a taken for granted or it had to be examined in a more critical stance enabling the speaker to distance himself from this ideal worker without being perceived as unprofessional. While the discourse of the ideal worker as an 'early educator' would allow both, doing femininity and doing masculinity, the discourse of the ideal worker as 'housewife' comes with implicit images and connotations about the ideal subjects engaging in that work being female.

A particular strength of our theoretical framework is that it enables differentiation between *men and women doing something* while at work in the nursery on the one hand, and the *organizational practices* at work on the other. Both are relevant for explaining doing gender in an organizational setting. However, our analysis clearly shows that male/female ratio of childcare workers did neither explain the major differences in everyday rhythms and routines nor the two facets of the ideal worker that were enacted in a nursery. Individual men or women are not doing things differently within their organizational logic. Rather it is the organizational rhythms and routines that seem to be relevant beyond the individual worker. They are responsible for (re) producing the two distinctive ideal workers of the 'housewife' and the 'early educator' resulting in different consequences for doing gender.

## **6 Conclusion**

Focusing on the organizational rhythms and routines that enact gendered discourses of 'care' and 'education' sheds some light on the practices that stabilize traditional gender concepts and therefore counteract the aim of gender equality. Although the curriculum for Swiss nurseries equally emphasize both 'care' and 'education' as important aspects (Wustmann Seiler and Simoni 2012) and the quality criteria also include both aspects (Qualikita 2014), 'care' still seems to dominate 'education' in two of our cases whereas the other two nurseries seem to be more balanced in that respect. As the traditional connection between 'care' and 'femininity' is still made relevant in the 'ideal worker' as well as in the organizational rhythms and routines, change is easily resisted here. As organizational rhythms and routines play an important role in stabilizing social practices (Orlikowski and Yates 2002; Staudenmeyer, Tyre and Perlow 2002), we conclude in more general terms that it is crucial to focus on these tacit practices that we are often unaware of when implementation of new pedagogical concepts is at stake. Furthermore, the relevance of the feminine 'ideal worker' as well as the discourse of 'care' for the dominance of household activities in nurseries' rhythms and routines proves to be a vital starting point for explaining the stability of gender inequality in this field. Moving beyond early childhood education and the nursery, our findings point to the relevance of the material and discursive practices that bring the gendered subject into being (Nentwich 2014). It is exactly these everyday rhythms and routines that seem to counter-act gender equality policies in the long run.

## **References**


# **5 Postscriptum**

# **Resisting change: A critical analysis of media responses to research on gender equality**

*Andrea Maihofer*

## **1 Introduction**

What kind of media response could be expected when, in a country like Switzerland, a national research program on gender equality<sup>1</sup> presents its final report to the public; a research program which spanned several years and was approved and promoted by the national research institution (the Swiss National Science Foundation) as well as the Federal state? Its topics cover a broad spectrum: wage inequality; sexual harassment at work; the consequences of tax and social policy; childcare; the continued segregation of paid work; and questions regarding unpaid care work. For sure, not all findings came as a surprise. And some important topics are absent. Amongst them, the situation for men, as well as discrimination against marginalized gender and sexual modes of existence. Yet, overall, the projects delivered substantial knowledge and a series of important results. This includes findings on a range of persistent inequalities and discriminations on the basis of gender as well as important indications why they persist and the mechanisms whereby they persist. In addition, the projects formulate concrete recommendations for the necessary social, political and legal measures that must be implemented to attain the desired goal of gender equality.

Considerable interest in the research could have been expected in view of the abundance of detailed insights into the persistence of old and the emergence of new forms of inequality as well as the findings about current devel-

<sup>1</sup> Note on the translation: The German term *Gleichstellung*, as used in the title of the research program *Gleichstellung der Geschlechter* and elsewhere, denotes an action which is carried out by an unmentioned agent and which brings about equality. As such it suggests a process towards equality, which is absent from the common English terminology *gender equality* that is employed in the official national research program reports as well as in the present article.

opments in gender relations in Switzerland; Especially since it is becoming increasingly evident that – although much has been undertaken in the last years to achieve gender equality, and although much has also been achieved – the fundamental problem has not at all been solved (see also Derungs et al. 2014).

Yet, as a first cursory glance at the media response shows, public interest in the findings of this research program could at best be described as reserved. Most articles report rather plain, neutral information. Alongside them, there are, however, a few articles that could be allocated to the spectrum of socalled 'anti-genderism' (see Maihofer/Schutzbach 2015). This second group of articles decidedly criticize the program's results. In addition, they express a clearly recognizable wish to question the scientific standards of its findings. Mostly, discrediting reference is made to "genderism" (Huonder 2013: 5) or "gender-ideology" (Petition 2014).

How should one explain this discrepancy between, on the one hand, what presently counts as an obvious demand for gender equality, and, on the other hand, the rather reserved interest in the findings about ongoing deficits? Does this confirm the assumption of a merely "rhetorical modernization" (Wetterer 2003)? Or does it point towards something different? How, in this context, should we interpret the partly quite vehement attacks on the research results? What is being fought about? And what does this mean for gender research?

In what follows, I would like to address these questions by means of an analysis of example press articles. My aim therewith is also to work out the frame of reference within which findings from gender research are formulated. With which cognitive and affective resistances are they confronted? When are findings seen as scientifically or politically relevant? Or how are they made irrelevant? What mechanisms of trivialization but also scandalization do such findings encounter? And last but not least – why this reservation?

## **2 Media response – Gender equality still remains a contested project**

#### *2.1 Between slight impatience and reservation*

The most print media articles about the findings of the National Research Program 60 mainly offer overviews of the results. At times some projects are also treated in greater detail and once in a while reference is made to events related to the topic. Generally the findings are summarized neutrally, closely oriented along the synthesis report and without any additional commentary or even critical assessment of its significance for gender relations – as if the authors were worried that they might say something wrong; or more precisely, that they might say too much (whatever that might be). And, although most articles are accompanied by soft approval of the findings, indignation about the persistence of inequality can only seldom be sensed. Why?

Consider the headlines. Alongside the rather plain assertion that "Women and men are not equal in Switzerland" (*Luzerner Nachrichten* 27.5.2014; *Bieler Tageblatt* 27.5.2014) or "Equality only partially realized" (*Freiburger Nachrichten* 2.7.2014; *Aargauer Zeitung* 28.5.2014), there are also titles in which a certain impatience about the slowness of the development can be detected, when it is emphasized for example that "Women and Men still not equal" (*Die Botschaft* 28.5.2014) or it says: "No equality yet" (*Appenzeller Volksfreund* 28.5.2014, *Solothurner Zeitung* 27.5.2014). According to these headlines, Switzerland has already been on route for a long time, and has made good progress, but still has a long way to go. As Kubli states, "We have, admittedly, made progress but we are clearly still on the way" (*Oberbaselbieter Zeitung* 19.6.2014). Some authors express their impatience more explicitly: "Men and women as before not yet equal" (*Appenzeller Volksfreund* 28.5.2014) and emphasize that women continue to "wait" for equality (*Freiburger Nachrichten* 2.7.2014).

What is striking about these formulations – despite general approval of the research results – is the clear effort to characterize the present situation *positively*: the relations are not *unequal*, but *not yet* or *only partly equal*. This recurring "not yet" or "partly" furthermore invokes the image of consistent progress. Even if it is (too) slow, Switzerland is on the right track regarding equality: it is moving forward, step by step. Explicitly *negative* formulations in which reference is made to "gender inequality" are rare (*folio* 13.8.2014; *Aargauer Zeitung* 28.05.2014). There is visible avoidance to explicitly call current gender relations a situation of continuing inequality or even illegal.

Correspondingly, reference is only seldom made to the anchoring of equality in the Federal Constitution or, as the *Freiburger Nachrichten* puts it, to the "constitutionally prescribed equality of man [!] and woman" (2.7.2014; see also *Appenzeller Volksfreund* 28.5.2014; *folio* 13.8.2014). There seems to be little knowledge of the constitutional anchoring of equality. Likewise, there is as little knowledge of the fact that, accordingly, gender discrimination is explicitly declared illegal. The media response seems to imply that somehow these prescriptions are in the constitution and belong to the normative standard of Switzerland, but one is not quite sure about this. After all, this right often rather comes as something like a concession; insisting on it is not seldom perceived as feminist exaggeration, maybe even something (slightly) ridiculous. In any case, the impression persists that this right to equality should be carefully handled, that it should not be insisted upon too loudly and not too vehemently, and then only with much acknowledgement of everything that has already been achieved – and of course, with gratitude.

Accordingly, the articles furthermore do not contain any explicit reference to the social mandate to bring about equality. It also does not seem obvious that one could call on this constitutional mandate and its detailed elaboration in the Federal Law of 1995. There is also no certainty whether it is really sensible or opportune to do so. Hence the necessity of (additional) measures is addressed: "Nothing happens voluntarily," writes the *Die Wochenzeitung* regarding wage equality (19.6.2014; regarding care work, see *folio* 13.8.2014 or generally, *FritzFränzi* 25.8.2014). However, it is at the same time assured that measures to "realize equality" cannot be ordained "simply from above" but must be "carried by all members of society" (*folio* 13.8.2014; *FritzFränzi* 25.8.2014). Social responsibility is thereby again relativized or more precisely individualized in that all members of society are addressed as equally involved in the realisation of gender equality.

But what exactly does it mean if – as is evident in the media response – talk about "more equality" seems more opportune and speaking about ongoing inequality or even injustice is avoided? Is the present situation not perceived as illegal? Or does this express uncertainty whether there is really a socially shared consensus? I think both. This would, however, point to the fact that the often cited thesis of rhetorical modernization (Wetterer 2003) falls too short. Because, different from what the thesis of rhetorical modernization assumes, it is to date clearly not (yet) obvious that gender equality is a shared norm in Switzerland. Surprising as it may be, there is no shared consensus: neither regarding what exactly the content of gender equality is, nor whether there is in fact a right to equality, nor whether there is a state and social responsibility to see to it that such equality is actually brought about. Even supporters of equality have different views on this.

The main problem, hence, is not primarily the (still) insufficient realization of equality. Or as Sylvie Durrer declares in relation to the findings of the NFP 60: "Whilst equality continues to count as recognized value, it is long not applied in practice. There is, admittedly, an awareness, but concretely little is happening" (*Soziale Sicherheit* 8.9.2014). On the contrary, gender equality and the prohibition of discrimination on grounds of gender and sexuality remain socially contested. They are by no means a "recognized value" that should merely be realized. Instead, the thesis that gender equality is widely acknowledged, conceals a deep disagreement and thereby contributes, even if unintentionally, to the persistence of gender and sexual inequalities and discrimination. But how can this hesitation, this existing 'normative' disagreement, be concretely named and explained in its different facets and consequences? Why this caution to be more offensive in tackling the social debate on equality as general normative standard? A debate that, in my view, is increasingly urgent as the attacks on equality presently have become more vehement – possibly more so the more equality is realized and gender relations change.

Against this background the reservation in the media responses can be interpreted as expression of the only too justified concern that the stigma of feminism virtually inevitably sticks to criticism of gender relations, however diluted such criticism may be. The effect is that such criticism becomes discredited as "killjoy", as Sarah Ahmed pointedly remarks (2010: 65). Such branding and disparagement are, however, not merely familiar attempts to devalue criticism and rob it of its political relevance. These disparaging comments also establish a fatal interplay of affects which are prompted by (feminist) critique and which are at the same time revived by the stigmatization of (feminist) critique. Thus, in many contexts, only a minute critical allusion to existing gender relations (such as sexism, wage inequality, or the "glass ceiling") suffices to trigger a complex mixture of bad feelings, such as irritation, annoyance, feelings of guilt, up to contempt and scorn. Such critical comments sensitively disrupt the dominant need for a common world view and a homogenous we-feeling. They apparently question the assumed agreement on existing conditions and the approval of a shared consensus on values – within seconds one becomes a spoilsport and outsider. For many, such remarks are not at all harmless; they are rather – virtually instinctively – perceived as (personal) attack. They are, in every case, perceived as a disturbing diminishment of the efficacy of "symbolic violence" (Bourdieu 2005: 63) and a dwindling of the recognition of "masculine domination" (ibid.). Critical comments cut through the effects of the dominant politics of truth, namely the "acceptability" (Foucault 1992: 33) of existing gender relations, and must therefore be sanctioned. Correspondingly, according to Ahmad, the history of feminism is "a history of making trouble" (2010: 60) and "saturated with unhappiness" (ibid.: 65).

The perceivable reservation in the media articles is, not in the last instance, also the outcome of internal censorship. It is based on the caution or even the fear that too clear a positioning may result in accusations of exaggerated political correctness or stigmatization as a 'women's libber' or feminist. That these fears are not unfounded can be seen from a closer examination of the present frame of resonance within which the articles on the results of the NFP 60 are situated.

#### *2.2 Contestations – between "fear and scorn"*

Partly truly vehement attacks on Swiss equality politics and gender relations come from the churches, such as the Christian fundamentalists, as well as the conservative right. Central in the responses by the churches, is the allegation that "gender ideology" (Petition 2014, n. p.) seeks nothing less than the "deconstruction of the natural gender order and the breaking up of the heterosexual norm" (ibid). Such deconstruction is connected, not only to the threat of homosexualization, but to the pluralisation of the genders as such (see also Zukunft CH 2015, n. p.). Correspondingly Bischof Huonder warns that "genderism" considers "every sexual practice (lesbian, gay, bisexual, transsexual) as of equal value with heterosexuality" (2013: 7).

From this perspective – besides the danger of the pluralisation of gender and sexual modes of existence (Maihofer 1995) – questioning the naturalness of the traditional gender order furthermore implies the dissolution of the 'natural' connection between femininity and motherhood. With that goes the collapse of the bourgeois family with its traditional gendered division of work. This means, according to Huonder, nothing less than a fundamental "[a]ttack on marriage and family as the structures that carry our society" (2013: 5). This is because, in eroding the bourgeois family as hegemonic life-form, the family does not only lose its central function as guarantor of the reproduction of (purportedly natural) gender differences. Rather, the family loses its function as guarantor of the reproduction of the existing gender order as such. Accordingly, the petition "No gender in curriculum 21"<sup>2</sup> , demands the removal of gender and equality, as well as all topics connected to it, from the school curriculum: what counts is securing the reproduction of the traditional bourgeois gender order.

What dominates in the right-conservative mainstream media, in contrast, is the scenario of the levelling of difference [*Gleichmacherei*]. An example for this is the article by Markus Somm "Man and woman are equal, more equal, most equal" (Basler Zeitung 2014). In this article two studies from the NFP 60 are commented upon: one records the persistent gender segregation in training and career trajectories amongst young people;<sup>3</sup> the other gives evidence of the reproduction of stereotypical gender differences in preschool centres.4 According to Somm, both studies belong to "the microcosm of purported 'gender injustices'" (ibid.: 3). He classifies these studies thus, because, when compared to earlier "serious discrimination such as the fact that a woman could not sign

<sup>2</sup> The curriculum reform concerns the whole of Switzerland; the petition was launched by Christian organisations as well as members of the Young Swiss People's Party (SVP) and the Federal Democratic Union.

<sup>3</sup> NFP 60 (2014): Geschlechterungleichheiten in Ausbildungs- und Berufsverläufen.

<sup>4</sup> NFP 60 (2014): Puppenstuben, Bauecken und Waldtage: Gender in Kinderkrippen.

a contract without the approval of her husband", gender equality in Switzerland has been achieved and is generally acknowledged as an "obvious matter" (ibd.: 4). Schär too finds such studies superfluous, after all, as far as "gender justice" is concerned, Switzerland has for a long time occupied "top positions" (Weltwoche 2014: 2). In comparison to these achievements, the remaining differences that the research has shown are not of any further importance.

If, "long after legal gender equality has been written into our constitution," there are nevertheless "still very few female electricians and carpenters" (Basler Zeitung 2014: 2), this only goes to prove for Somm that these must be natural and unchangeable gender differences. Accordingly, he is of the opinion that there is also "something totalitarian in how these scholars want to shape humans afresh: a Frankenstein without gender. It should not and cannot be that a girl prefers to play with dolls more than a boy does" (ibid.: 5). In opposition to this 'levelling from above', he insists on the idea of a liberal society in which "gender equality" is an obvious "liberal concern" (ibid.). However, for Somm gender equality cannot mean "that in the end everyone must be the same. We like to differ, and that also concerns man and woman" (ibid.). It is, anyhow, "not the duty of the state to level out gender differences" – especially since "this would hardly be possible" (ibid.).

Considered more closely, Somm's plea for difference does not, however, aim towards the recognition of difference and variety. As in the argument on pluralisation above, the scenario of levelling is used here as a strategy to ward off changes that are taking place. In addition, reference to the purportedly undeniable natural differences between men and women is also a way of upholding and legitimizing the inequalities and discriminations that are associated with these distinctions.

In this argument a 'new' mechanism of domination becomes evident, one that exists alongside naturalization, which has already often been shown to be a central mechanism of domination that contributes considerably to the persistence of traditional gender differences (see, as representative for many others, Goffman 1977). This new myth is that gender equality has been achieved and that there is no further need for social measures. As the NFP 60 shows, it is exactly this combination of 'new' and 'old' myth that is a central reason why – notwithstanding all the changes that have occurred – gender inequalities in Switzerland still persist (see in general also McRobbie 2009).

There is, however, something else at stake too: namely the important distinction between formal and material equality. Somm and others understand equality merely in terms of *formal equality.* For them formal equality covers the full extent of social and state responsibility. Further demands for material or factual equality count as exaggerated, even totalitarian, as they are in no sense the duty of the state and society. Because, in this view, equality is achieved as soon as the legal discrimination of women is abolished – for example the exclusion from higher education or the prohibition against entering into contracts. The rest is the responsibility and the autonomous choice of individuals. According to this (neo)liberal view, whether women use their opportunities, is solely a question of their individual will and the force of their assertiveness.

What this formalist opinion leaves out though, is that the Federal Constitution is not merely concerned with the levelling of "differences" (Somm 2014: 5) or simply with "legal equality" (ibid.: 2). The constitution has a much more fundamental objective, namely the removal of *inequalities* and *discriminations* on the basis of gender. In brief, it is about creating "legal and factual equality" amongst the genders (Art. 8, § 2 and 3). According to this article, the state and society explicitly have the duty to create the material conditions under which women can also realistically realize their potential. When Somm and others refer to basic individual responsibility and exonerate the state and society from this obligation, it contradicts the constitutional responsibility enshrined in the constitution and elaborated in Federal Law, namely that the realization of factual equality also requires the creation of the necessary social conditions.

Analogous to this example from Somm, articles by Michael Bahnert "Gender justice skewed" [*Gleichberechtigung in der Schieflage*] (Basler Zeitung 2014) or Claudia Wirz "Please no imposed happiness!" [*Bitte keine Zwangsbeglückung!*] (Neue Zürcher Zeitung 2015) illustrate how fundamental the disagreement still is – or is especially now – about the understanding of constitutionally enshrined gender equality.

Bahnert considers scandalous the change in the law pertaining to breastfeeding as decided by the Bundesrat, notwithstanding the fact that a similar regulation is common in other countries. According to this new regulation, mothers in paid employment can now count as work-time the time needed to breastfeed or pump off milk. This – for Bahnert – is a clear indication that equality has become "skewed" (Basler Zeitung 2014). To him it constitutes blatant discrimination against "men, who are unable to pump off milk and breastfeed" [*abpump- und stillunfähigen Männer*]" (ibid.) Thus he develops – ironically – a 15-point program for legally ensured compensation. The program includes the right to "vouchers for after-work beers," "lifelong access to Playboy.tv and the right, for the duration of breastfeeding or pumping, to visit the site on a daily basis," the right that "home-side football games count as holidays," as well as "high-heels-duty for female apprentices and interns" (ibid.), and more of the same. As is well known, irony is often a means 'to, anyway, be allowed to say' what is politically incorrect. In addition, irony puts those who take seriously what has been said, in danger of counting as humorless "killjoys" (Ahmed 2010).

When one does, however, take the article seriously, it becomes clear that Bahnert does not share the understanding of equality that underlies the change in the law. He does not acknowledge the injustice of disadvantages that mothers experience in their jobs; nor does he consider it justified that positive discrimination is used to create the conditions for factual equality in the working lives of mothers. Equality is understood purely as legal equality. There is, accordingly, neither any awareness of injustice in the form of structural discrimination, nor any idea about the social responsibility to overcome such injustice.

According to Wirz too, it is only "legal equality" that is at stake, and this only as an "urliberal concern" (*Neue Zürcher Zeitung* 2015), which has already been long achieved. In her view, all additional measures, such as the creation of equal wages or quotas for women in leadership positions are inappropriate and lead to a dangerous "levelling of every gender difference in social cohabitation" (ibid.). All of this aims at nothing but a threatening "modification of society in the sense of gender justice"; at a "state feminism" that "imposes happiness" from above (ibid.)

In this, as in many other articles and announcements<sup>5</sup> , there is not only a deep disagreement in Switzerland about the understanding of gender equality enshrined in the constitution. Seen more closely, there is also presently a basic social quarrel about the direction in which the gender order is developing. However, not everyone sees this dispute in the same manner, and people act quite differently in this context.

The myth that gender equality as "urliberal concern" (ibid.) has long been achieved, plays an important role in these quarrels. On the assumption that gender equality already exists, it is easy to discredit and rob of its political relevance, research that is conducted on persistent inequalities and discrimination on the basis of gender. It is equally easy to discredit measures aimed at creating factual equality and to rob them of their significance. Depicting as superfluous the preoccupation with trivial and unchanging (because natural) 'remnant inequality', is a powerful myth with which to undermine such research and measures. Noticeable about the argumentation strategy is, however, how this academic research and these social measures are *at the same time* also stylised as highly threatening interference from above, as totalitarian violations and re-education programs from the side of the state and its "state-feminist" collaborators (ibid.). This calls up and reproduces a pattern that Joan Scott had already identified in her historical analyses: a paradoxical

<sup>5</sup> On this, see the extensive analysis of 'antigenderism' in Switzerland in Maihofer/Schutzbach (forthcoming 2015).

affective mixture of fear and scorn. Through criticizing the social and gender order, feminism, according to Scott, "opened questions about the system's original design and about the need for rethinking it. That was (and is) the power and the danger of feminism, the reason it provokes both fear and scorn" (1998: 12).

However, in these articles it is gender research rather than traditional feminism that now is the target of criticism. Gender research now counts as "exaggerated feminism" (Lusser 2014). Labelling it as so-called "genderism" (Huonder 2013: 5) it is stylised, like feminism in earlier times, as at one and the same time laughable and extremely threatening. For this reason, there is also a mounting trend towards talk of 'antigenderism' (On the current shift from antifeminism to 'antigenderism', see Maihofer/Schutzbach (2015)).

#### *2.3 Discrediting gender research*

In many of the articles, it is not only individual positions or results of gender research that are criticized; instead, the attacks are directed more at gender research as such. The aim of these discrediting attacks is specifically to deny gender research its scientificity and to disavow it as pure ideology (for a more extended argument, see Maihofer/Schutzbach (2015)). Gender research's alleged unscientificity is mainly tied to its more fundamental rejection of the naturalness of gender difference. From the perspective of these critics, feminism still argued (and continues to argue) within the framework of the heterosexual dual gender order and has been advocating equality within a clear gender dichotomy. In contrast, gender research (at least of deconstructive provenance) does not only question the natural givenness of traditional gender differences; it also questions the gender binary as such. Correspondingly, gender research denies the existence of any "innate differences between the genders, that is differences traced back to biochemistry and evolution" (Schär 2014: 2). Connected with that, (deconstructivist) gender research also denies the naturalness of heterosexuality. Furthermore, in this view, the knowledge and findings of gender research are neither based on knowledge from the natural sciences, nor are they "proven according to the methods of the natural sciences" (Walker 2014, op. cit.; similarly Lusser (2015, op.cit.). In other words, for the critics of gender research, traditional empiricist (natural) science is the benchmark of scientificity. That the humanities and social sciences, and with them gender research, have established an own understanding of science with its own criteria of scientificity, is eclipsed, or devalued as scientifically deficient. Thereby the criticism does not only strike gender research alone; yet, because of its specific themes, such criticism does hit gender research in particular. Equally disregarded by this criticism, is the fact that gender research engages extensively with knowledge from the natural sciences and that, in the meantime, there is highly sophisticated gender research within the natural sciences. But why at all this urgent wish to discredit gender research in its scientificity? What danger is connected with gender research?

From the perspective of the critics, the repudiation of the naturalness of the heterosexual gender binary does not only go hand in hand with the levelling of different gendered and sexual modes of being and with that the dwindling of heterosexuality as hegemonic norm; the repudiation of the naturalness of the heterosexual gender binary also entails a pluralisation of genders and sexualities. With the "disruption to the stable binaries of sexual difference" (McRobbie 2009: 63), however, vital cornerstones of the dominant gender order, such as the bourgeois family as germ cell of society, are fundamentally questioned.<sup>6</sup> Furthermore, the dominant gender order and its accompanying inequalities and discriminations are robbed to an even larger extent of their legitimation in the supposed natural gender difference.

An additional reason why its critics perceive gender research as unsettling, is that it again and again (as also in the framework of the NFP 60), gives empirically founded evidence for the continued existence of inequalities and discrimination on the basis of gender and sexuality, and that it also reveals the reasons for their persistence. Moreover, this research offers a number of concrete recommendations for measures to overcome these inequalities and this discrimination. That means: in a scientifically grounded manner and insistently, gender research draws attention to what should rather remain unsaid: merely formal equality does not suffice. What is needed is much more the creation of factual gender equality; and for that a range of social, legal and political measures still remains indispensable.

All of these findings produced by gender research seem to be considered so threatening because gender studies has established itself – however precariously – in the academic system. This may also have contributed to the construction of a partly really monstrous conception of gender studies as an enemy. A reason for the increased hostile focus on gender research is that with it, gender as a theme has advanced to an autonomous science. Gender research has become part of an admittedly increasingly disputed, but still socially powerful institution (see also Hark 2014). The point of the attacks is to ward off a knowledge project that, within the framework of one of the most significant institutions, provides evidence of the persistence of inequality and discrimination. This project thereby makes visible, in a scientifically founded

<sup>6</sup> This does not, however, as is mostly assumed, mean the erosion of the significance of family as such, only of its hitherto hegemonic form. As can be presently observed, it is possible that the pluralisation of familial arrangements is accompanied by an intensification of the significance of family (for greater detail, see Maihofer 2014).

manner, the continuing necessity of social and state action to overcome existing gender relations.

Not least, these attacks are also an attempt to repudiate a science that points to the historicity of gender relations and the fact that these relations are socially constructed; a science which insists on the critique of existing gender relations as well as their changeability. In contrast, many of these articles draw an inference – in a positivist empirical sense of science – from empirical facticity to how things *should* also be. As Horkheimer and Adorno (2009) have already criticized, this evidentness of the factual, implies an authoritarian attitude, an attitude that is neither willing to engage in critical self-reflection nor signals the will to engage in a broader social debate on the further development of the social and gender order. This authoritarian attitude claims for itself, on the contrary, an unerring ability to know what is the case and what should be the case, how one thinks, feels and lives and how one is allowed to think, feel and live. In other words, in the final count, what is at stake in this dispute is also – in the Foucauldian sense – a specific form of "politics of truth" (1978: 51). It is a battle to define what is true and what is right and how one is allowed to live and how not. And in this case in particular, it is about securing the bourgeois gender order as natural, true social order. What simultaneously becomes evident is that this is not only a scientific disagreement. It is also a political dispute about the further development of the social and gender orders in Switzerland – even though not everyone perceives this dispute in the same manner.

## **3 Conclusion**

Why is the movement towards equality in Switzerland so slow? This question is repeatedly raised. How can this persistence in change be explained, this reluctance or resistance? The aim of my article could not have been to answer conclusively this broad question. Rather, by means of an analysis of media reporting on NFP 60, I attempted to work out central elements of the frame of resonance within which research results on the state of equality are formulated. My interest is especially the question: What cognitive and affective resistance confronts this research? And which mechanisms of trivialization but also of scandalization do these results encounter? In addition, I wanted to show with which lines of argument (feminist) criticism of existing gender relations must reckon till today. Furthermore, I sought to explicate how, presently, the frame of resonance for gender research is shifting (e. g. from anti-feminism to so-called anti-genderism), and also how gender research contributes to changes in this frame of reference. My hope was, in this way to add some additional suggestions to the many made by the NFP 60. And indeed, I think, some fundamental aspects of the resistance to change have become visible. Although they are not really new, these aspects do, in my view, play too little a roll in discussions on equality. To conclude this essay, I want to bring these findings to the point.

**First:** Contrary to its own self-understanding, in Switzerland there is still no consensus – neither about what the contents of equality is, nor whether there is in fact a right to equality, nor whether there is a state and social responsibility to realize equality. Even amongst its advocates, there are clearly different views. Seen exactly, there is still not even really a general collective awareness that something illegal is being done. To be precise, injustice both in terms of Switzerland's own constitution and its explicit prohibition of discrimination as well as the Federal Law of 1995, and injustice in the sense of international UN declarations. Correspondingly, the existing relations are widely, either not at all even perceived as inequality and discrimination, or as hardly mentionable 'remnant inequalities', and in no need of further action. Parallel to this, however, the awareness that one has a right to equality also appears rather vague, as does the certainty what is meant by such a right. An important reproduction mechanism for all of this is the naturalization of gender and gender differences as well as the myth that equality has long been achieved.

**Second:** Against this background, the problem of the view that there is a supposed longstanding general consensus regarding gender equality, and that all that is lacking is the application, becomes clear. This assumption reveals itself as a further significant mechanism in the persistence of inequality in gender relations. The view that there is such a consensus conceals – consciously or unconsciously – the deep-running disagreement about the understanding of equality, its content, the right to it, and its application. It furthermore conceals the fact that there are also quite intensive social disputes. What is at stake is thus not, as assumed, merely the gap between the widely recognized norm of equality and the practical application of the norm. The insufficient realization of this norm is much more also due to the fact that it is not yet at all generally shared and recognized. Moreover, the view that there is consensus obstructs the insight that norms such as freedom and equality are social-cultural phenomena and therefore the determination of their content, their social significance, as well as their realization are never fixated once and for all. On the contrary, because they are tied to different interests, associations and goals, they are necessarily constantly socially contested and their spectrum of validity depends essentially on the given social power relations. In view of this, and as the media response confirms, it can be suspected that there has never been a general social consensus regarding the question of equality in Switzerland, even amongst the majority who supports equality. At any rate, there is currently no (longer) a general consensus. Even more, there are strong social and political forces with decidedly different ideas. Not only do these forces not share a further overcoming of traditional gender relations, they also hinder, and even want to reverse much of it.

**Third:** In this sense the media analysis reveals a deep-running disagreement about the future social and gender orders. What is at stake is not only what family, gender, sexuality, masculinity and femininity are or what these should be, which ways of thinking, understandings of scientificity, truth and norms are right or false; but also at stake generally is the right to define which lives can be lived and which not. Thereby the forces are particularly intensive that wish to fend off or to reverse the presently emergent really fundamental changes in social and gender relations (see McRobbie 2009; Maihofer 2014). This 'countermovement', that in itself is not uniform, and amongst which socalled 'antigenderism' is one of the most explicit variants, must also be seen as a reaction to the change that is taking place. It strives to stem this change, or better, to steer this change in a specific direction. In short: what is presently happening is more than a "gender conflict" (Lenz 2013); particularly since the disputes are increasingly not between 'men' and 'women' (with different social and sexual modes of existence), but also amongst 'men' and amongst 'women'. In the growing knowledge that the changes in gender relations represent and demand deeply influential social (i.e. economic, cultural, legal and political) changes, what is actually taking place is a cultural struggle about the future development of the social and gender orders. In this dispute many increasingly formulate the clear goal, namely to protect the maintenance of the existing social and gender orders with their inequalities. Some even advocate a retraditionalisation – however modified to fit a neoliberal fashion. Against these offensive positionings, attempts at the emancipatory further development of the social and gender orders are often rather reserved in their criticism and the alternative visions they propose. They are careful, as if held back by inner censorship, and fearfully aware not to come into the vicinity of stigmatization as killjoy, feminist or 'women's libber', and thereby have their position discredited. In other words: in Switzerland, the criticism of existing gender relations in the meantime has strongly moved into the defensive, not lastly because of the reactions to its success. But also – and this too seems to me to become clear – because the realization of factual equality itself has come to a stall. It has become increasingly clear, not only what a fundamental and complex social change the realization of factual equality requires, but also that it needs new attempts and new types of measures. Furthermore, this takes place at a time in which the disputes are growing more vehement. It is hardly surprising then if the application of equality moves ahead only hesitantly.

**Fourth:** What, however, does this now mean for gender research? When Judith Butler emphasizes, that according to her, "feminist theory has no other work than in responding to the places where feminism is under challenge" (2009: 286), she strikes a central point, even if formulated somewhat apodictically. Especially when she continues: "And by responding to those challenges, I do not mean a defensive shoring up of terms and commitments, a reminding of ourselves of what we already know, but something quite different, something like a submission to the demand for rearticulation, a demand that emerges from crisis" (2009: 286). These attacks and difficulties are to be used, but not defensively, as that would only reinforce the already defensive position. Likewise, repeated self-representation as powerless or even as 'victim,' should be avoided. Instead, there should be an awareness that the attacks and the attempt to discredit gender research are based on the threat that gender research has in the meantime come to pose to many. Feminism and now gender studies too are significant agents in social development. In brief: They should take more seriously the social significance and the efficacy of their research.

Furthermore, the opportunity that lies in these attacks, as Butler rightly emphasizes, exists in yielding to "the demand for rearticulation". This requires, in my view, working out more carefully the social dissensus at stake in the present disputes. What exactly is being vehemently attacked and what is being insisted upon? Which forms of life feel threatened and which ones should be precluded? Specifying this could allow a better understanding of what the objects of these disputes are, as well as what the meaning is of the attempts to discredit gender research as science. What connexus of power, knowledge and truth politics is this about?

In other words, what needs to be explained in greater detail – and for that the results of the NFP 60 are a good starting point – is how, despite all change, the persistence of gender inequality can be explained and how this persistence should be dealt with. Here it would be decisive to increasingly take a multidimensional perspective on the constitutive social interrelations amongst different forms of inequality, discrimination, and disciplining. On the other hand, these results are also to be further explored in terms of social theory and diagnoses of the times, more than has been the case till now. In this sense the current theoretical and empirical challenges in fact constitute an opportunity for gender research to reflect (self)critically. Such reflection might open the way to new formulations of central perspectives, concepts and insights in gender research; this might also lead to its scientific and social repositioning.

## **References**


Butler, Judith (2004): Undoing gender. New York/London: Routledge.


#### *Sources*


# **About the authors**

Andreas Balthasar, Dr. rer. pol., is adjunct professor in the Department of Political Science since 2010 and in the Department of Health Science and Health Policy at the University of Lucerne, Switzerland, since 2013. His research is focused on social and health policy and evidence informed policy. Recent publication: Balthasar, A./Müller, F. (2014): Die Verbreitung von evidenzbasiertem und gleichstellungssensitivem Wissen in den Entscheidungsprozessen der kantonalen Steuer- und Sozialtransferpolitik: eine quantitative Analyse. In: Swiss Political Science Review 20, 1, pp. 70–95.

Kathrin Bertschy, lic.rer.pol, is an economic researcher and since 2011 member of Swiss Parliament. During the National Research Program 'Gender Equality' (NRP 60), she worked as consultant and project leader at the economic research and policy consultancy ECOPLAN in Bern and lecturer at the University in Basel, Department of Sociology. Her work focusses on labor market, education and social policy issues. Recent publication: Bertschy K./Walker P./Baeriswyl A./Marti M. (2014): Lohndiskriminierung beim Berufseinstieg. Eine quantitative Analyse für die Schweiz. In: Schweizerische Zeitschrift für Soziologie, 40, 2, pp. 279–305.

Christina Felfe, PhD, is assistant professor at the University of St. Gallen, Switzerland, as well as a consultant for Empirical Economic Research at the Swiss Institute for Empirical Economic Research. She obtained her PhD in 2008 from the Universitat Pompeu Fabra, Barcelona. The focus of her work lies on Education Economics and Labor Economics, with a special focus on gender equality and human capital formation.

Gesine Fuchs, Dr. phil, works and teaches as an independent researcher and consultant in Basel, Switzerland. Her main fields of expertise are law & politics, political participation and representation as well as gender equality policies. Recent publications: Fuchs, G. (2015): Substantielle Repräsentation im Schweizer Parlament: Zum Agenda Setting beruflicher Gleichstellungspolitik 1996–2011. In: Femina Politica 24, 1, pp. 73–83. Fuchs, G./Bothfeld, S./Leitner, A./Rouault, S. (forthcoming): Gleichstellungspolitik öffentlicher Arbeitgeber. Analysen aus der Schweiz, Deutschland und Österreich. Opladen: Barbara Budrich.

Daniela Gloor, Dr. phil., is sociologist and cofounder of the institute "Social Insight" for social research, evaluation, and consultation in Schinznach-Dorf, Switzerland. An important research focus of the institute is on violence, gender and equality. Recent publication: Gloor, D./Meier, H. (2014): "Nach dieser Zeitspanne fragt man sich wirklich, ob das jetzt ein Witz ist". Erfahrungen gewaltbetroffener Frauen mit Interventionen des Rechtssystems. In: Juridikum, Zeitschrift für Kritik, Recht und Gesellschaft, 3, pp. 327–337.

Gisela Hauss, Dr. phil., is professor at the University of Applied Sciences Northwestern Switzerland. Hauss' research and teaching interests include unemployment, welfare, social inequality, gender, history and theory of Social Work. Recent publications: Hauss, G. (2014): Generating Productive Citizens or Supporting the Weak? Ambivalences and Contradictions in Working with Young Welfare Recipients. In: European Journal of Social Work, Special Issue, 17, 5, pp. 656–671. Hauss, G. (2014): "She's a migrant, she's got children, and she's a single mother". Welfare programs as sites for the (re)commodification of mothers. In Harrikari, T. et al. (Ed.) (2014): Social change and cocial work. Farnham/Burlington: Ashgate, pp. 121–139.

Rolf Iten, Dr. oec. publ., is economist and managing partner of INFRAS, an independent consulting group providing policy analysis and implementation services in Switzerland. He works in the areas of energy, social and general economic policy with a special focus on socio-economic analyses of political measures in the fields of energy and the environment, health and gender equality as well as research for sustainable asset management. Together with Michael Lechner of the University of St. Gallen he was responsible for the NRP60 project "Child care services and gender equality". Recent publication: INFRAS and University of St. Gallen (2013): Familienergänzende Kinderbetreuung und Gleichstellung.

Andrea Leitner, Dr. rer. soc. oec, is senior researcher at the Institute for Advanced Studies in Vienna, Austria. She is specialized on policy analyses and gender indicators focusing the interface between evaluation, gender politics and gender studies in the field of education, higher education and research as well as labor market. Further results of the NRP 60 will be published in the book: Fuchs, G./Bothfeld, S./Leitner, A./Rouault, S. (forthcoming): Gleichstellungspolitik öffentlicher Arbeitgeber. Analysen aus der Schweiz, Deutschland und Österreich. Opladen: Barbara Budrich.

Andrea Maihofer, Dr. phil., is philosopher, sociologist and gender researcher; she is professor of Gender Studies at the University of Basel and Director of the Center for Gender Studies. Her research focusses on the analysis of change and persistence in gender relations in the context of current processes of social transformation. She works on forms of family life, socialization, training and career trajectories, transformation of masculinities, sexualization and sexuality, norms and normalization. Furthermore, her research concerns social and gender theory as well as moral, legal and constitutional theory.

Hanna Meier, Dr. phil., is sociologist and cofounder of "Social Insight", the institute for social research, evaluation, and consultation in Schinznach-Dorf, Switzerland. An important research focus of the institute is on violence, gender and equality. Recent publication: Gloor, D./Meier, H. (2014): "Nach dieser Zeitspanne fragt man sich wirklich, ob das jetzt ein Witz ist". Erfahrungen gewaltbetroffener Frauen mit Interventionen des Rechtssystems. In: Juridikum, Zeitschrift für Kritik, Recht und Gesellschaft, 3, pp. 327–337.

Franziska Müller, lic. rer. soc., studied sociology and public law at the University Bern, Switzerland. She is head of the migration and equal opportunity department of the private consultant office INTERFACE, Switzerland. Her work focusses on evaluations and consulting related to the topics of equal opportunity for women and men, the promotion of integration, and measures in the area of asylum. Recent publication: Balthasar, A./Müller, F. (2014): Evidenzbasierte und gendersensitive kantonale Steuer- und Sozialtransferpolitik: Erfahrungen und Folgerungen. In: LeGes – Gesetzgebung und Evaluation, 25, 2, pp. 215–230.

Eva Nadai, Dr. phil., is sociologist and professor at the School of Social Work of the University of Applied Sciences Northwestern Switzerland. Her research fields include gender, social policy, organization, and ethnography. Recent publications: Nadai, E. (2015): Un/Abhängigkeit, Un/Sicherheit, Emanzipation − Geschlechterverhältnisse im Post-Wohlfahrtsstaat. In: Nadai, E./Nollert, M. (2015): Geschlechterverhältnisse im Post-Wohlfahrtsstaat. Weinheim/ Basel: Juventa, pp. 7–25. And: Nadai, E./Canonica, A. (2014): Gleichstellung am Rand des Arbeitsmarkts? Sozialinvestitionen und Verwirklichungschancen aus einer Genderperspektive. In: Schweizerische Zeitschrift für Soziologie, 40, 2, pp. 349–364.

Julia Nentwich, Dr. rer. soc., is associate professor in psychology at the University of St. Gallen, Switzerland. Together with Franziska Vogt she was responsible for the NRP 60 project "(Un)doing gender in the nursery". Her research interests cover gender and diversity, change agency and resistance, and discourse analysis. Recent publication: Nentwich, J.C./Poppen, W./Schälin, S./ Vogt, F. (2013): "The same and the other: Male childcare workers managing identity dissonance". In: International Review of Sociology, 23, 2, pp. 325–344.

Michael Nollert, Dr. phil., is professor at the Division of Sociology, Social Policy and Social Work, Department of Social Sciences, University of Fribourg, Switzerland. His research focusses on social inequality, labor markets, unpaid work, and economic sociology. Recent publications: Epple, R./Gasser, M./Kersten, S./Nollert, M./Schief, S. (2014): Institutions and gender time inequality: A fuzzy-set QCA of Swiss cantons. In: Swiss Journal of Sociology, 40, 2, pp. 259–278. And: Gasser, M./Kersten, S./Nollert, M./Schief, S. (2015): Geschlechtsspezifische Ungleichheiten in der Arbeitswelt: Kantonale Muster der Zeitungleichheit. In: Schweizerische Zeitschrift für Soziologie 41, 1, pp. 9–31.

Sophie Rouault, M. A., is research fellow at the Bremen University of Applied Sciences, Germany. Her PhD project deals with the governance regime for equal employment in France and Germany, and particularly with the instruments involved in this policy field. Recent publication: Fuchs, G./Bothfeld, S./ Leitner, A./Rouault, S (forthcoming): Gleichstellungspolitik öffentlicher Arbeitgeber. Analysen aus der Schweiz, Deutschland und Österreich. Opladen: Barbara Budrich.

Sebastian Schief, Dr. phil, is senior lecturer at the Department of Social Sciences, University of Fribourg, Switzerland. His areas of research include international comparative research on precariousness, welfare regimes, social policy, gender equality, economic sociology. Recent publications: Gasser M./Kersten, S./Nollert, M./Schief, S. (2015): Geschlechtsspezifische Ungleichheiten in der Arbeitswelt: Kantonale Muster der Zeitungleichheit. In: Schweizerische Zeitschrift für Soziologie, 41, 1, pp. 9–31. And: Schempp D./Schief, S./Wagner, A. (2015): Determinants of detraditionalization of the division of house and family work in Swiss couple households. In: Schweizerische Zeitschrift für Soziologie, 41, 1, pp. 33–57.

Susanne Stern, Dipl. Social Geographer, Exec. Master Gender Management, is an associate partner of INFRAS, an independent consulting group providing policy analysis and implementation services. She primarily works in the areas of family, educational, integration and gender policy, where she specializes in socioeconomic analyses and evaluations. Together with Christina Felfe she was leader of the NRP60 project "Child care services and gender equality". Recent publication: Stern, S./Felfe, Ch./Schwab, St. (2014): "Was bringt die familienergänzende Kinderbetreuung für die Karrierechancen von Müttern?" In: Die Volkswirtschaft. Plattform für Wirtschaftspolitik, 6, pp. 19–21.

Wiebke Tennhoff, M. Ed., is a doctoral student at the University of St. Gallen. She is currently co-leading a counseling project that aims at facilitating men's entry to the field of early childhood education day care centers. From 2011– 2013 she was involved in the NRP 60 research project "(Un)doing gender in the nursery". Her research interests are early childhood education, doing masculinities and discourse analysis. Recent publication: Tennhoff, W./Nentwich, J./ Vogt, F. (2015): Doing gender and professionalism: Exploring the intersectionalities of gender and professionalization in early childhood education. In: European Early Childhood Education Research Journal, 23, 3, pp. 340–350.

Franziska Vogt, PhD, is professor and head of the Institute for Research in Teaching and Learning at the University of Teacher Education in St. Gallen, Switzerland. Her research interests cover early childhood education, professionalization. Recent publications: Vogt, F./Nentwich, J. C./Tennhoff, W. (2015): Puppenecken und Bauecken: Gender und Raumordnung. In Müller, C./Amberg, L./Dütsch, T./Vogt, F./Wannack, E. (2015): Perspektiven und Potenziale der Schuleingangsstufe . Münster: Waxmann, pp. 113–124. And: Vogt, F./Nentwich, J. C./Tennhoff, W. (2015): Doing und Undoing Gender in Kinderkrippen: Eine Videostudie zu den Interaktionen von Kinderbetreuenden mit Kindern. In: Schweizerische Zeitschrift für Bildungsforschung, 37, 2, pp. 227–247.

Thomas Widmer, Dr. phil., is professor of political science, in particular evaluation, and the head of the research unit "Policy-Analysis & Evaluation" at the Department of Political Science, University of Zurich, Switzerland. Widmer's research and teaching focus on evaluation, public policy, Swiss politics and methodology. Within the National Research Programme "Gender Equality" (NRP60), he was the leader of the project "Development and steering of gender equality policy at work".

Christine Zollinger, lic phil., is a political scientist affiliated at the Department of Political Science, University of Zurich. Her research interests are in Swiss politics, policy-analysis and comparative social policy, with a particular focus on the intersection between gender and welfare state-research. Currently, she works on a project about the development of childcare policies in Swiss municipalities. Recent publication: Widmer, T./Zollinger, Ch. (2014): Vereinbarkeitspolitik als Gefahr für die Gleichstellung und den sozialen Zusammenhalt? In: CHSS Soziale Sicherheit, 4, pp. 219–223.

# **About the editors**

Karin Gottschall, Dr. phil., is professor of sociology and head of the section 'Dynamics of Inequality in Welfare Societies' at the Research Center on Inequality and Social Policy SOCIUM at University of Bremen, Germany. She was a member of the steering committee of the National Research Program 'Gender Equality' (NRP 60) of the Swiss National Science Foundation from 2009–2015. Her publications and research focus on sociological theory and gender, female employment and labor market segregation, welfare state and social services and the state as employer. Recent publication: Criblez, L./Gottschall, K./ Périsset, D. (eds.) (2015): Gleichstellung der Geschlechter im Bildungsbereich. Special Issue: Schweizerische Zeitschrift für Bildungswissenschaften, 37, 2.

Brigitte Liebig, Dr. phil., is sociologist and psychologist; she is professor of organizational science at the University of Applied Sciences and the Arts Northwestern Switzerland, Department of Applied Psychology, and lecturer at the University of Basel. She was president of the National Research Program 'Gender Equality' (NRP 60) of the Swiss National Science Foundation from 2008–2015. Her work focusses on gender and diversity issues, work and organization, professionalism and entrepreneurial universities. Recent publication: Liebig, B./Levy, R. (eds) (2015): Critical transitions: The production of gender (in)equality in the life course. Special Issue: Equality, Diversity and Inclusion: An International Journal, 32, 2.

Birgit Sauer, Dr. phil., is professor of political science at the Department of Political Science, University of Vienna, Austria. She was director of the graduate school "Gender, violence and agency in the era of Globalisation" (GIK) and is speaker of the research network "Gender and Agency" at University of Vienna. Her research fields include gender and governance, gender and affects, rightwing populism and racism. Recent publication: Gresch, N./Sauer, B. (2014): Topographies of gender democracy in Austria. In: Galligan, Y. (2014): States of democracy. Gender and politics in the European Union. London/New York: Routledge, pp. 33–49.