This paper examines the relevance of gender in accounting for the formation of different welfare state regimes in Sweden and Germany. Recent feminist scholarship has provided us with an array of approaches to the comparative analysis of social policies. However, it has not fully succeeded in conceptualizing gender as an analytic category. Concerning the welfare state, Joan W. Scott's a critique of feminist theorizing is substantially correct today. Gender is still often used as a substitute for »woman,« as for example in Skocpols work. Gender relations - here meaning gendered welfare state institutions - are  usually explained in terms of »non-gendered« causal relations. Thus Jane Lewis claims, referring among other countries to Sweden among other countries that gender was irrelevant to the formation of different welfare regimes.

      My hypothesis, by constrast is, that gender plays a constitutive role in the creation of the early welfare state in Sweden and Germany. The purpose of this paper is therefore twofold: first to develop a theoretical framework which moves beyond the conceptual limitations of feminist inquiry mentioned above; and second to demonstrate its empirical usefulness in a comparative case study focusing on the emergence of protective labour legislation in both countries.

      The research strategy I pursue includes three levels of analysis: 1) processes of social and economic change; 2) political forces and institutions; 3) discourses and interpretative frameworks. My aim is to explore the mutual and complex relations among these different levels, in order to avoid a slippage into a determinist epistemology, whether conceptualized in terms of discourses, institutions or social/economic structures.

      Sweden and Germany are extremely well suited to this comparison. From the perspective of social history, they share several similarities: the tradition of a strong bureaucratic state, a weak liberal bourgeoisie, and an early political mobilization of the labour movement. With regard to the subject of inquiry, they differ considerably. Germany passed numerous gendered regulations in the field of protective legislation before 1914-- such as maternal leave in 1878 (repeatedly extended), prohibition of night work, and limitation of daily working hours for women in 1891. The Swedish development was in contrast rather modest: the first protective law in 1889 did not include any gender distinctions, a maternal leave was enacted in 1900 and women's night work was prohibited in 1909. 

      This analysis shows that the policy differences between the two countries can not be adequately explained in terms of socioeconomic structures indicated, by e.g. women's employment rates. The legal differences correspond to variations in formations of national discourses. Demands for protective legislation in Germany were formulated in a gender specific way from the outsest. This was not the case in Sweden. Characteristic of the German debate was an early enmeshment of moral with scientific discourses defining women's factory work as undesirable and harmful for the female and social body, as well as an explicitly articulated masculine political interest. The interpretative  frameworks of the Swedish dabate were quite different. This is well illustrated by the strategies of the early labour movements in both countries. The close relations between the Swedish and German labour movements led Swedish social democracy to take over the Gotha programme though with one notable exeption however: the paragraph demanding special protective legislation for women was excluded.