I received a grant from the Harry Frank Guggenheim Foundation at the early stages of what became an eight-year-long inquiry into the causes of mass violence against civilians, in particular genocide. The project’s methodological premise was that scholars (and policymakers) should seek to learn from “negative cases,” that is, situations where theory would predict genocide to occur but where it did not. The early stages of my research, which HFG funded, focused on intensive study of Côte d’Ivoire in West Africa as well as on understanding the range of variation in violence across sub-Saharan Africa since independence.
The initial research in Côte d’Ivoire prompted two developments. One was a set of new hypotheses about the determinants and dynamics of mass violence. Second was some expertise on the country. In turn, when an electoral crisis broke in 2010 in Côte d’Ivoire, I became engaged in public commentary on the country via the Huffington Post, and I published accessible academic articles in Foreign Affairs and African Affairs. In the former, Tom Bassett and I argued for the central importance of regional organizations in the management of the Ivorian violent crisis. In the latter, I demonstrated patterns of violence in the electoral crisis and distinguished the logic of electoral violence from that of exterminatory and civil war violence.
The intensity of war—specifically threat perception—shapes leaders’ willingness to use mass violence and the public’s acceptance of it.
The Ivoirian crisis, as well as my survey of violence in postindependence Africa, led to an inquiry into the dynamics of electoral violence more broadly. With Charlie Taylor, I assembled a dataset on electoral violence in Africa since the transition to multipartyism in the early 1990s. The dataset, the African Electoral Violence Dataset, uncovered a set of patterns that were unexpected given our prior assumptions. We found, for example, that about one in five African elections resulted in serious violence; that violence after war was less common than if a country had not experienced war; that income level was not correlated to electoral violence; that incumbents committed the majority of electoral violence; and that the dynamics of prevote violence differed from postvote violence. Those findings were published in an influential volume edited by Dorina Bekoe of the United States Institute for Peace. Charlie, Jon Pevehouse, and I also published an article in the Journal of Peace Research, in which we show through a battery of regression analyses that incumbent running is the strongest predictor of significant African elections.
The HFG research also launched the broader, longer project examining why genocide took place in some conflicts but not others. That project culminated in a book published by Cornell University Press in 2015, entitled Making and Unmaking Nations: War, Leadership, and Genocide in Modern Africa. Three findings stand out and all relate to the initial findings from the HFG-funded research in Côte d’Ivoire. First, the intensity of war—specifically threat perception—shapes leaders’ willingness to use mass violence and the public’s acceptance of it. Second, ideological constructs—what I label “founding narratives”—influence the ways in which political and military leaders devise strategies of violence in war. Third, other sources of restraint—such as certain kinds of economic structures—can create incentives to moderate violence. I extend the empirical analysis to Senegal, Mali, Rwanda, and Sudan, each occupying case study chapters in the book. The book has won four awards, including the 2018 Grawemeyer Award for Ideas Improving World Order.