Abstract
This paper discusses two methodological issues in research studying the performance of political regimes. First, it compares the conceptualisation and measurement of three data sets on types of autocratic regimes developed by Geddes et al. (2012a), Hadenius et al. (2012), and Cheibub et al. (2010). The data sets are found to differ with respect to their conceptualisation of autocratic regimes (narrow vs. flexible concepts) and types of autocratic regimes (criteria for classification, number of autocratic regime types and definition of the main types), as well as their levels of measurement (simple vs. complex indicators). As a result, the choice of data sets matters for this type of research. Second, with respect to the data on performance, the paper summarises the current state of research on the issue of the availability of data for democratic and autocratic regimes, discusses possible effects on empirical findings and addresses ways to deal with the problem.
Acknowledgements
The author would like to thank the anonymous reviewers, the editors, Stephanie Bergbauer and Jan Teorell for their valuable comments and suggestions to improve the paper.
Notes
This paper has been written in parallel with the one by Wahman et al. (Citation2013).
See Brooker (Citation2009, p. 44) for these questions and the construction of typologies of autocratic regimes in general.
Some of these cases – foreign occupied, warlord, provisional – appear in .