Concept-measure inconsistency in contemporary studies of democracy

If measurement validity is to be safeguarded, concept-measure consistency needs to be high. More particularly, insofar as a concept is multi-dimensional, the theoretical structure of the concept must be reflected in the aggregation procedure used to combine the sub-components. In this article we show that many of the present attempts to create indices or typologies of democracy and non-democracy are characterized by severe concept-measure inconsistencies. The concept itself is normally treated in terms of the Aristotelian notions of necessity and sufficiency, which means that the proper aggregation procedure is either the minimum rule or multiplication, depending on whether the sub-components stand in an interactive relationship or not. But all too often simple addition is used instead. We replicate a number of such analyses and show that the use of proper aggregation rules makes a salient difference empirically. Also, we show that the identified analyses are not lone occurrences but indicate a more general problem in the literature.