Esping-Andersen (1990; 1999) was interested in identifying the similarities between countries within the three categories of welfare states that he distinguished: the by now familiar types of ‘Scandinavian’ or ‘social democratic’, the ‘Anglo-Saxon’ or ‘liberal’, and the ‘continental’ or ‘conservative’ welfare states. Scharpf and Schmidt (2000a; 2000b) went further and took a closer look at differences within these categories. They studied how countries reacted to common challenges of internationalization, and found quite a few variations between countries – even within the same welfare regime. They argue that policy responses have been influenced by three sets of factors: (a) the nature and intensity of policy challenges; (b) governance institutions defining actor constellations and their modes of interaction; and (c) perceptions and preferences of the actors involved (Scharpf and Schmidt, 2000a: 1–20). This implies that the affiliation to a welfare state regime cannot alone explain or predict the modes of adaptation and reform countries choose when confronted with problems of internationalization. Does this make EspingAndersen’s typology less useful or even obsolete? Not quite. The type of welfare state, that is the manner in which it provides income to its citizens – by general taxes, by insurance, or work and residual poverty protection – and the way in which it provides services (broadly public, by the family or by subsidizing private schemes) is still a major explanatory variable of reforms. It was typical for all conservative welfare states, for example, first to try to shift the increasing unemployment problem away from the labour market, either by sending people into early retirement, into invalidity, into motherhood or into longer training, whereas it was typical for liberal welfare states to create flexible employment patterns and lower wages. Scandinavian countries had developed active labour market policies very early and, therefore, faced labour market problems later and less dramatically than most conservative welfare states. Though Esping-Andersen’s typology of welfare state families is still useful, it seems worthwhile to analyse them in more depth in order to understand the reforms and performance of welfare states when faced with challenges of internationalization. The present paper deals with EspingAndersen’s ‘continental’, ‘conservative’ and strongly ‘corporatist’ family of welfare states, made up of Austria, Belgium, France, Germany, Italy and the Netherlands. These welfare states are rooted in the ideas of Bismarck and von Taffe (Esping-Andersen, 1990: 24). They are dominated by a system of social insurance which implies that social entitlements are derived from employment rather than citizenship (contrary to the social-democratic model) or proven needs (as in the liberal model), and which is committed to a preservation of status differentials. Social protection tends to be differentiated by occupational classes. Benefits mirror status and earnings rather than redistributive ambitions (Esping-Andersen, 1990: