Electoral Violence in New Democracies: Institutionalizing Peaceful Elections

Abstract

In the last few decades, thousands have died in election-related violence. In the past ten years in Africa alone, Nigeria, Ethiopia, Algeria, Madagascar, Central African Republic, Côte d'Ivoire, Uganda, Chad, Angola, Togo, and Kenya have experienced severe instances of electoral violence. However, a number of other transitioning states, for example Ghana and Benin, held peaceful elections in the same period. The main question I ask in my dissertation is why some new democracies experience electoral violence while others do not. I argue that elections are credible commitment problems in which candidates running for election face incentives to use fraud and or violence to win the election. The incentives to subvert the electoral process increase when previously marginalized ethnic groups select competitive candidates to compete in the election. However, independent electoral management bodies (EMBs) can resolve the commitment problem by convincing candidates and their supporters that a loss at the polls today does not mean a loss of power forever. In addition, an independent EMB limits the opportunity for election rigging, making it harder to subvert the electoral process. In the following chapters, I use both qualitative and quantitative methods to test the relationships outlined above. To test the effects of independent EMBs, I collected and coded data on more than 200 elections in sub-Saharan Africa between 1990-2010. To test the causal process I conduct multiple comparative case studies of elections in Kenya and Ghana between 1992 and 2013. I find that ethno-political exclusion and perceptions of competitive elections significantly increase the risk of electoral violence, while de facto independent EMBs mitigate this risk by reducing electoral fraud and extending the time horizon of the candidates.