Comparing measures of democracy: statistical properties, convergence, and interchangeability

Abstract

By comparing and analysing four cross-national measures of democracy, this article provides novel information regarding the statistical properties, convergence, and interchangeability of some of the most frequently used measures of democracy. The author points out limitations related to the statistical properties of these measures and finds that even if measures of democracy are highly convergent, their interchangeability is weak. This means that the choice of the measure of democracy has considerable consequences for the conclusions of a given study. Especially so in studies covering the last few decades, because the author finds that in general the interchangeability of democracy measures has decreased since the 1980s. In choosing one measure over another, scholars should be aware of the limitations identified in this article. To overcome problems related to weak interchangeability, if a single measure cannot be credibly chosen on theoretical grounds, the author recommends users of the measures to validate their findings with multiple measures of democracy.