Edgs Working Paper

Presented as part of the " Comparative Politics Workshop " speaker series at the Equality Development and Globalization Studies (EDGS) program at Northwestern University, with generous support from the Rajawali Foundation. ABSTRACT Since the restoration of multiparty competition in the country in the early 1990s, three of Kenya's five presidential elections—1992, 1997, and 2007—have been accompanied by severe violence in the form of ethnic clashes. However, the elections of 2002 and 2013 were relatively peaceful. What explains this variation? Through sub-national comparisons and on-site fieldwork in the Rift Valley and the Coast, this paper develops a two-stage argument to account for the puzzle at hand. First, it posits that elite choices about the ethnic composition of electoral coalitions set the stage for whether or not violence will accompany any particular election. More specifically, when politicians build alliances between hostile communities, election-related conflict is unlikely to occur even in places where there has been a history of such violence. The second stage of the argument, then, identifies factors that influence politicians' decisions about the ethnic make-up of electoral alliances. With regard to this meta-puzzle, the article argues that peaceful elections in Kenya have not been a product of leaders' commitments to peace per se. Instead, they have occurred when unique circumstances drove politicians to unite rival ethnic groups in coalitions of convenience.