Coding competitive authoritarianism

Abstract

Hybrid regimes have posed an empiric, methodological and conceptual challenge to academics since their emergence in the early 1990s. One of the most ambitious studies of the nature and behavior of hybrids in the past decade is Steven Levitsky and Lucan Way’s 2010 book, Competitive Authoritarianism: Hybrid Regimes After the Cold War (Cambridge University Press).

In this “era of unprecedented abundance of cross-national political data” (Mudde and Schedler, Pol Res Quart 63(2):410–416, 2010, p. 410), the Levitsky and Way framework both organizes data into a uniquely operable model and contributes to it by generating new information on regime type. Yet those who look to build and expand upon it must be aware of the “structural problems of information about data supply and data quality” (Mudde and Schedler, Pol Res Quart 63(2):410–416, 2010, p. 410) inherent in current political quantitative data and models. The competitive authoritarian (CA) model is no exception to these problems. In order to reproduce it and to use the regime model and theory for further research, it is necessary to address a number of conceptual, methodological and empirical deficiencies present in the work.

This article assesses Levitsky and Way’s CA regime classification model by independently reproducing it in the 14 Sub-Saharan African countries in the original analysis as well as in a limited number of additional cases in Sub-Saharan Africa. The paper opens with a presentation of the core concepts of competitive authoritarianism, an argument for the purpose of the present study, and an overview of the CA model. The replication exercise is described and the findings are reviewed in detail. The paper closes with a systematic analysis of the work, drawing on Munck and Verkuilen’s framework (Comp Pol Stud 35(1):5–34, 2002) for assessing quality of democracy data. I describe the most significant empirical indeterminacies that arise from the methodological indeterminacies identified in the replication exercise. I conclude that the inability to replicate the case selection undermines the usefulness of the proposed model to systematically identify CA. I consider how the broader conclusions of the original work are impacted by these findings, specifically how the persistence of competitive authoritarianism in this region is impacted by variations in case selection. The article offers recommendations on how to address shortcomings in the model in order to strengthen it and adapt it to the study of other non-CA hybrid regimes.