ABSTRACT
After a decade of methodological uniformity – or something approaching it – it seems that the study of referendums is once again entering a revolutionary period. While there are certain tendencies that suggest that the reasoning voter paradigm has deserved this epithet, the study of referendums has not been characterized by the methodological uniformity of, say, neo-classical economics, let alone of theoretical physics. Yet, referendum studies were still not characterized by a single paradigm. It is difficult to say that the study of referendums has followed Karl Popper's prescribed methodology. For a start, very few of the natural sciences actually follow the models outlined by either Popper or Kuhn. During the subsequent period of "normal science", most research would be characterized by "puzzle-solving", where anomalies would be sought and explained within the framework. Until the 1990s, research about referendums was characterized by single case research.
