Three worlds' typology: Moving beyond normal science?

There is no doubt that Esping-Andersen's three worlds' typology has been extremely valuable. However, the literature inspired by it shows signs of Kuhnian normal science, which is impairing empirical and theoretical progress. We explain normal science, demonstrate that it characterizes recent empirical regime studies and ask why this has come about. We show that the welfare regime literature has a tendency to confuse the terms typology' and ideal-types'. This has prevented the emergence of anomalies that are needed for progress. We argue that normal science is fostered by the combination of researchers' tendency to prefer certainty (e.g. to address solvable research problems) and the environmental pressures they face (especially the publish or perish' culture and the need to frame research problems in terms of variation rather than similarity). In the discussion section, we suggest several routes by which the welfare state literature can move beyond normal science.