Multiple Measurements, Elusive Agreement, and Unstable Outcomes in the Study of Regime Change

Abstract

This comprehensive analysis of regime change indicators reveals that problems of conceptualization and measurement are major reasons why current research fails to draw compelling conclusions that foster cumulative knowledge. The paper first argues that even though the literature discusses the conceptualization of regime types at length, there is little attention to defining regime change. Furthermore, quantitative studies of regime change largely elide conceptual and measurement challenges. Second, although indicators of regime type are highly correlated, agreement between indicators of regime change is extremely low. Third, focal points such as elections and coups drive agreement among these indicators, suggesting that such measures often reflect notable events instead of regime change per se. Finally, a robustness check of nine articles on regime change published in top journals demonstrates that findings are often not robust to alternative indicators, implying that indicator choice influences the results of quantitative studies.