The previous chapter discussed rival arguments regarding the impact of party ideology on welfare state retrenchment and pointed out a number of important limitations in our understanding of this link. In particular, we know relatively little about the motivations underlying retrenchment – or what one might call the “why question.” So why do they do it? In this chapter, I argue that we need to take (party) ideology more seriously in order to approach this question. Thus, I single out the common practice of investigating the impact of partisanship based on static expert judgments and the underlying theoretical assumptions as the most pressing and underresearched problem. Regarding the operational dimension of this Independent Variable Problem, I argue that party labels based on such expert judgments are limited in terms of their concept validity as proxies for the ideological composition of government and that their usage prompts problems of empirical discrimination as regards the different theoretical arguments underlying the three competing theoretical perspectives. At the theoretical level, I argue that this measurement strategy implies a one-sided conception of government ideology as representation of (material) group interest and show that the related causal mechanism inherent to this conception stands in stark contrast to the definitions, assumptions, and findings in several adjacent strains of research. For one, the mechanisms comprising the causal chain underlying Power Resources and Partisan Theory are under pressure by the erosion of party group ties commonly referred to as dealignment. More to the point, the comprehensive work on (political) ideology, ideas, and cognitive frames is neglected.

Based on these criticisms, the remainder of the chapter then complements the hitherto dominant approach of ideology as agency for (fixed) group interests with an alternative and more literal conceptualization of ideology as belief system and cognitive frame. More precisely, I develop a cognitive framing argument as to why the ideological composition of government affects social policy choices of democratic elites in general and unemployment insurance in particular, even under conditions of high economic problem pressure. Against the theoretical background of this cognitive framing argument, I suggest a suitable measurement strategy based on the discussion of two strategies.