The theme of our research is attitudes towards the welfare state. Our guiding question is: Under what conditions do welfare state attitudes become consequential for political behavior? With political behavior we mean behavior during popular votes and elections. We will study this question by re-analyzing surveys on behavior in popular votes about social policy issues and on behavior in general elections for the period 1984-2004 in Switzerland. In two smaller sub-projects we conduct an international comparison of social policy issues in general elections in OECD countries, and a experimental study on the direction of the relation between beliefs and behavior. Popular votes specifically address one policy issue, and we can assume that this issue is in most cases the decisive factor in deciding the popular vote. In contrast, general elections do not specifically address only one issue. In electoral decisions, stances towards social policy issues are influenced by the saliency of this issue in the electoral campaign, the saliency of the issue for the individual voter and his/her attitudes towards the issue. Popular votes on social policy issues allow for the identification of the relative impact of factual, normative and utilitarian beliefs on the specific policy decisions. Comparing the impact of these explanatory variables on specific decisions to their impact on general elections allows for the identification of the conditions under which welfare state attitudes become decisive for the electoral decision. These findings, from in-depth studies of Swiss popular votes and elections, will be corroborated by an international comparison of welfare state attitudes, welfare state reforms and electoral outcomes in a comparative analysis. In addition, in an experimental study we will analyze the direction of the causal arrow between beliefs concerning the welfare state and behavior in decisions about social policy issues. The underlying problematique is the following: support for the welfare state is extremely broad among the Western public. It has been argued that any attempt by government to retrench the welfare state is an exercise in ‘blame avoidance’; otherwise, governments will be punished at the next election. However, this argument is not entirely convincing. There have been governments that cut back welfare state expenditures without losing votes. And there are voters who favor the consolidation or even expansion of the welfare state who still choose to vote for political parties that aim to reduce welfare state expenditure. Our research will contribute to the understanding of the politics of welfare state reform in Switzerland between 1984-2004. In addition it will deal with questions such as this one: Under what conditions will welfare state retrenchment lead to electoral losses for governing parties in democratic polities?