The causes and consequences of ethnic power-sharing

What are the causes and what are the consequences of ethnic power-sharing? Existing research generally agrees that ethnic coalitions decrease the likelihood of civil war. However, most scholars of power-sharing claim that political elites from different ethnic groups rarely form coalitions and that these coalitions are inherently unstable unless so-called power-sharing institutions provide incentives for cooperation. Since most studies of power-sharing do not measure coalitions directly, it remains unclear how accurate these claims are. This thesis complements existing research by adopting a theoretical and empirical approach that explicitly centers on elite behavior. Using global data on the government access of ethnic elites, the thesis explores what type of coalitions these actors form and how long their power-sharing pacts endure. Exploiting the variation in the ethnic composition of governments, this thesis also contributes novel insights to the study of democratization and civil war. More specifically, the empirical results indicate that ethnic elites frequently form oversized coalitions because they fear future defections by their co-ethnics and violent revolutions by members of excluded ethnic groups. Accordingly, ethnic coalitions are less stable than mono-ethnic governments. There is only limited evidence that so-called power-sharing institutions influence these dynamics. Since ethnic coalitions are usually unstable and elites face higher risks when they lose power in dictatorships, large elite coalitions are more likely to embrace democratization in order to lower their personal risks. For the same reasons, ethnic coalitions stabilize democratic regimes. Finally, this thesis finds that ethnic coalitions mainly reduce the risk of territorial civil war but neither have a positive nor a negative effect on conflicts over government power. ∗Center of Comparative and International Studies, ETH Zurich, Switzerland. Email: bormann@icr.gess.ethz.ch