The Shadow of Coups and Multiparty Elections in Authoritarian Regimes.

Electoral authoritarianism has become prevalent during the late twentieth century. Why do some authoritarian leaders adopt multiparty elections, running the risk of losing power? The conventional explanation emphasizes the role of internal and international pressures in explaining the emergence of multiparty elections in autocracies. Yet, many autocrats introduce multiparty elections absent these pressures for political liberalization. To answer this question, the theory presented herein focuses on a conflict between a dictator and his ruling coalition. Opening the political arena to electoral contestation diminishes the payoff from seizing power for his ruling coalition. At the same time, this allows his ruling coalition to pose a more peaceful challenge to him, defecting to the opposition. When coup threat is low, however, the dictator does not need to hold elections since they unnecessarily enable his ruling coalition to threaten him by the threat of exit. Accordingly, only when faced with high coup risk, does the dictator find multiparty elections attractive. I evaluate the empirical implications of my model based on a cross-national sample of autocratic non-electoral spells from 1960 to 2006. This empirical analysis is, to my knowledge, the first systematic large-N cross-national examination of why autocrats introduce national multiparty elections. My analysis, first of all, partially corroborates the conventional wisdom. I find little evidence for the bottom-up theories of political liberalization, which emphasize the shadow of revolutions rather than coups, but I do find that international factors are systemically associated with the introduction of multiparty elections. Next, consistent with my theory, the statistical results imply that greater coup risk makes dictators more likely to adopt multiparty elections, particularly executive elections that contest the office of the incumbent. This result demonstrates that there is an additional path to the establishment of competitive elections, so far little examined.PhDPolitical ScienceUniversity of Michigan, Horace H. Rackham School of Graduate Studieshttp://deepblue.lib.umich.edu/bitstream/2027.42/95972/1/namkyu_1.pd