Abstract
Many existing datasets used to categorize political regimes present problems for researchers, either because they were not specifically designed for that purpose, or because they employ dichotomous coding schemes that do not adequately capture the mix of democratic and authoritarian practices that coexist in many political systems. This article discusses these problems and introduces a dataset designed to overcome them - an expanded and updated version of the Political Regime Change dataset, which was designed specifically for the task of classifying regime types and transitions. Although the Political Regime Change dataset is comparable to existing datasets that measure institutional and political characteristics, its use of a broader set of indicators and categories by which to classify regimes creates important differences. The importance of this broader classification scheme becomes increasingly apparent in many 'third wave' transitions, where semi-democracies have been a common outcome of regime change.