Authoritarian breakdown: Empirical test of a game theoretic argument

Abstract

This paper continues a project on the links between different authoritarian institutional structures and the likelihood and mode of transition to democracy. In earlier work, I have argued that different types of authoritarian government have different effects on the incentives confronting regime supporters when faced with challenges to the status quo. To summarize that argument, military regimes tend to split when challenged, personalist regimes to circle the wagons, and single-party regimes to coopt their challengers. This argument implies that military regimes are more likely to negotiate their own withdrawal and to democratize; personalist regimes rarely leave office voluntarily and more often end in popular uprising, revolution, invasion, or assassination; and single-party regimes tend to survive longer than the other two forms of authoritarianism, even in the face of severe economic crisis. I test these hypotheses using a new data set that includes information on nearly all the authoritarian regimes that have existed since 1946. To summarize the findings, military regimes, on average, survive less long than other types. They are more likely to negotiate their extrications and to be followed by competitive political systems. They are less likely to end in coups, popular uprising, armed insurgency, revolution, invasion, or assassination. Personalist regimes, in contrast, are more likely than other types to end in violence and upheaval. Their ends are also more likely to be precipitated by the death of the dictator or foreign pressure, and they are more likely to be followed by some new form of authoritarianism. Single-party regimes last the longest, but when uncontrollable popular opposition signals that the end is near, like the military, they negotiate the transition.