"Ideas are the very stuff of politics. People fight about ideas, fight for them, and fight against them. Political conflict is never simply over material conditions and choices, but over what is legitimate. The passion in politics comes from conflicting senses of fairness, justice, rightness, and goodness. ... Political fights are conducted with money, with rules, with votes, and with favors, to be sure, but they are conducted above all with words and ideas" (Stone 2002: 34).
Despite the central importance of ideas in democratic politics, and the significant role of ideas and political discourse in classical sociological writings, they have played only a minor role within the discipline of political science in general, and especially in (comparative) welfare state research until the 1990s (for overviews see Campbell 2002 and Beland 2005). In regards to welfare state reform analyses institutionalist and structural approaches dominate the literature. According to institutionalist approaches significant social policy reforms in Germany are said to be highly unlikely due to various veto players, the large welfare state clientele, and the specific party competition between two welfare state parties (cf. Pierson 2001). Structural analyses measure the extent of reforms in relation to the effective solution of identified 'problems' and often conclude that reforms have not gone far enough or were ill-designed (cf. Streeck/Trampusch 2005). Despite their merits, these approaches are insufficient to answer the question, why change occurs in the first place and what meaning it has.
In this paper I will not theorize about the opportunities or obstacles and the adequacy or inadequacy of reforms, but demonstrate in how far the normative and ideational foundations of social policy in Germany have changed significantly since the 'golden' era of welfare state capitalism. For this purpose I will first outline the relevance of social constructivism to social policy analysis; second, I will characterize the ideal normative foundation of the conservative welfare state in the 'golden' post-World War II era, before scrutinizing the new ideational framework increasingly guiding the reforms since the 1980s. This section mainly builds on a content analysis of party programs and parliamentary debates. Finally, I argue that the ideational basis for the German welfare state is no longer Conservatism, but Liberal Communitarianism.
2. Social Constructivism and Social Policy Analysis
According to the public policy literature ideas are primarily of relevance in the process of agenda setting. They are said to determine the problem definition and policy options (Kingdon 1995). Problems are not a natural given or "mirrors of objective conditions", as many welfare state analysts (implicitly) argue, but are "projections of collective sentiments" (Hilgartner/Bosk 1988: 53; cf. Blumer 1971); or in the words of Majone (1989: 23 f.), "[o]bjective conditions are seldom so compelling and so unambiguous that they set the policy agenda or dictate the appropriate conceptualization." Therefore, although 'objective' challenges may contribute to the instability of an institutional equilibrium, they are not directly causal for policy change. This approach to policy analysis is rooted in the sociology of knowledge, initially developed by Karl Mannheim (1964), whereby 'reality' is socially constructed. The philosopher Ian Hacking (1999) demonstrated that the perception of what constitutes 'reality' depends on conceptualizations of 'facts' and of the processes used to measure them. If we talk about 'facts' or 'problems' challenging the welfare state and if we want to know whether they have any influence on the development of future policies, we must first determine whether these 'facts' or 'problems' are 'real' in the political world. Hence, this approach further builds on the so-called Thomas Theorem: "If men define situations as real, they are real in their consequences" (Thomas 1951: 51). …