The comparative analysis of welfare states has been greatly advanced by rights-based measurements of social provisions. Social insurance replacement rates have figured prominently here. Apparently, there is considerable confusion about the validity of replacement rates and their comparability across different datasets. The purpose of this study is to outline a refined institutional perspective in the comparative analysis of welfare states focusing on the character of social citizenship rights. We show that social insurance replacement rates from different datasets differ in their underlying theoretical framework for policy analysis and therefore capture different aspects of how welfare states secure the livelihood of citizens in periods of work incapacity. Analysing validity solely on the basis of replacement rate point estimates is therefore misleading. We show that the close focus on social citizenship rights and programmatic design in the Social Citizenship Indicator Programme (SCIP) carries great potential for causal welfare state analysis.