Under which conditions does party government matter for welfare state reforms? This research question has received ample attention in the ‘Does Politics Matter’ literature (see for an overview: Keman 2002b). Most of the time this research involved statistical (correlation) analysis on a small number of cases. These analyses include variables such as party preferences, cabinet composition, type of government, institutional arrangements and constraints. One problem with this type of small-n analysis is that the explained variance may be boosted by the low number of cases. Some recent advancements have been made by means of pooled time series analysis, but unfortunately this is accompanied by a number of problems and pitfalls. These are discussed by Kittel and Winner (2005) and Plümper et al. (2005) in their critique on the study of Garrett and Mitchell (2001) on the relationship between total government expenditure and the partisan composition of government as well as economic internationalisation. Due to these complications some are in favour of qualitative in-depth analysis of government formations and the effects on policymaking (Andeweg and Timmermans 2008). One of the reoccurring conclusions in the ‘Does Politics Matter’ literature is that mixtures of institutions matter more than certain models of democracy with a prescribed set of institutions (e.g. Lane and Ersson 2002). In other words, which combinations of actors, contexts and institutions do lead to a certain outcome?1