When it comes to social policy development, Switzerland has always been a laggard in international comparison. Whereas most other OECD countries have seen rapid growth of the welfare state during the post-war decades of welfare state expansion, social policy was still very limited in Switzerland up until the 1970s. In international comparisons, the Swiss welfare state was therefore often qualified as 'liberal' or 'residual' (e.g. Esping-Andersen 1990). From the late 1970s onwards, however, social policy in Switzerland has gradually expanded, bringing the country closer to the ideal type of a continental European welfare state (Armingeon 2001). This catch-up effect first affected the traditional social policy schemes such as unemployment insurance (which became compulsory only in 1984), occupational pensions (in 1985), as well as health insurance (in 1994). Family policy benefits, by contrast, remained on a very modest level and care services were virtually non-existent until the late 1990s. It was only at the turn of the millennium that a series of successful reforms gave a boost to Swiss family policy. The most important reforms include the implementation of a federal program to subsidize childcare infrastructure (in 2003), the introduction of compulsory maternity insurance (in 2005), as well as a nationwide harmonization and--in some of the cantons--increase of family allowances (in 2006). Observers seeking to explain why the development of Swiss social policy lags behind the pack (Armingeon 2001: 152) generally point to the specific institutional and political features that characterize Switzerland. For instance, Bonoli (2007: 767) nicely shows how the development of legislation in the field of social policy was "institutionally delayed" by direct democracy, federalism and the interaction between the two. Direct democracy was a major brake to public sector expansion, as voters tend to oppose new taxes. The introduction of a maternity insurance, for instance, was rejected no less than four times at the polls in direct democratic referenda between 1984 and 1999. Due to federalism, the central state is weak, and therefore unable to propel the development of social policy--unless a constitutional change (subject to referendum) explicitly assigns to it competences and resources that are required. Family policy is until to-day mainly organized at the cantonal and the local levels. Until the harmonization in 2006, the cantonal differences, e.g. in the level of child allowances, were as large as between the most generous and the most liberal European countries (Hausermann 2007).
On the political dimension, the power balance between political parties in Switzerland is unfavorable to welfare state development. Both in Nordic and in Continental countries (Esping-Andersen 1990; van Kersbergen 1994), Social Democrats and Christian Democrats were the major driving forces of welfare state development. In Switzerland, both of them have always been weaker than in the neighbouring countries, and they traditionally faced a dominant market-liberal party, which opposed welfare state expansion in general and family policy expansion in particular. Under this strong liberal influence, the Swiss Christian Democrats also played a somewhat more cautious role than Christian Democrats in other Bismarckian countries, such as Germany or the Netherlands, where they were usually among the main drivers of (transfer-oriented) family policy (van Kersbergen 1994). An additional political impediment--especially in the field of family policy--was the weakness of the feminist movement during much of the 20th century, due among other factors to the belated introduction of female suffrage in 1971 (Martin 2002).
The growth of the Swiss welfare state since the 1970s, however, and in particular the introduction of new family policy schemes since the late 1990s, raise questions with regard to the factors that may explain these developments. Indeed, while the reference to the institutional obstacles and the political weakness of the left in Switzerland can plausibly explain the delay of welfare state development in Switzerland with respect to other countries (see also Olivia Thoenen's contribution to this issue), they seem less helpful in explaining its recent expansion, especially in the field of family policy, given that both the institutional framework as well as political power relations have not shifted dramatically. Institutionalist theories of the welfare state generally suffer from the weakness that they are very capable of explaining inertia and resilience, but have difficulties explaining policy change (see van Kersbergen 2002). And when it comes to power-and interest-based explanations, they seem more helpful to explain recent welfare retrenchment (Allan and Scruggs 2004) rather than the expansion we see in the realm of family policy.
For all these reasons, family policy development in Switzerland (as well as in Germany and other European countries, see the contributions to this issue) calls for explanations that go beyond the traditional approaches. This is all the more true given that developments in Swiss family policy since the mid-1990s denote not an incremental, but a major change (Hausermann 2006; Kubler 2007), both in terms of the underlying conceptions of social risks to be addressed, as well as in terms of the values and models of society embodied in the policy goals that are pursued.
The objective of this article is therefore twofold. On the theoretical level, we aim to provide an analytical framework that is able to complement the traditional factors, institutions and interests, to achieve more analytical leverage. More precisely, we will argue that a stronger emphasis on policy frames, ideational factors and political coalitions as drivers of reforms is needed to understand the politics of family policy reform in Switzerland. It is only by analyzing the dynamics of coalition-formation on the background of the ideational frames that underlie these dynamics and the institutional incentives that shape them that we can account for recent policy development. On the empirical level, we provide a systematic account of recent changes of Swiss family policy at the national level, and gauge the prospects of future developments in this field.
In order to do this, the article is divided in three sections. In a first, theoretical section, we discuss various idea-based approaches to (family) policy change and present the analytical framework that is used for the subsequent empirical analysis. The second section assesses recent developments in three major issues of family policy in the light of this theoretical perspective: parental leave schemes, childcare services, as well as family and child allowances. On this basis, we identify essentially two different patterns of winning coalitions that form around measures of family policy that allow specific combinations of different family policy frames. We conclude on the prospects of future family policy reform in Switzerland.
2 Explaining family policy change
2.1 Ideational approaches and the concept of frame
Family policy has always posed particular challenges to traditional approaches in welfare state analysis. While the three welfare regimes (Esping-Andersen 1990) can largely be recognized also in family policy benefits and services (see e.g. Lewis 1993; Ferrarini 2006), the pace and direction of development has oftentimes been out of step with major other social policy programs. On the one hand, the evolution of family policy in OECD countries differed significantly from the development of other components of the welfare state (Gauthier 1999). Family policy expanded less rapidly than other policy sectors during the "trente glorieuses", the three decades following the end of WW2. State support for families was also less affected by cuts in public spending since the 1970s; it was even expanded in many European countries, in spite of an often difficult budgetary environment (Henderson and White 2004). In addition, the traditional explanatory factors identified for other fields of social policy seem to have little explanatory power in the recent development of family policy. At best, the evidence is mixed. While Bonoli (2006) shows that the generosity of family policy still correlates with the traditional variables explaining welfare generosity (mainly left power), Henderson and White (2004) find no evidence for this link, nor for an association of recent family policy development with institutional factors (such as the existence of veto points). In a similar vein, Gauthier (1999: 963) concludes that family policies appear to be driven by different forces than other components of the welfare state. This does not come as an utter surprise, since family policy has always been much more than an instrument of distributive class conflict. In particular, family policy has always been particularly value-laden, expressing ideas and ideologies on the organization of family and gender roles in the society (Lewis 1992; van Kersbergen 1994). There-fore, it is particularly important to take ideational factors into account when analyzing family policy.
Several authors have recently come to the conclusion that the development of family policy in Switzerland since the turn of the millennium denotes major policy change that must be analyzed in the context of ideational transformations. Seeking to explain the strengthening of gender equality policy in Switzerland since the 1990s, Martin (2002) argues that the presence of feminist ideas in political and discursive arenas played an important role, as did alliances that women's movements constructed with parties and organized interests. Cognitive framing, combined with coalition building, she argues, was important for the development of policy measures in this realm. A similar line of argument is followed by Ballestri and Bonoli (2003)...