Localized Ethnic Conflict and Genocide: Accounting for Differences in Rwanda and Burundi

This paper seeks to explain the variation in the scale of violence across episodes of ethnic conflict, using data from Rwanda and Burundi. To do so, we explore the "dark side" of in-group policing -- when it is exploited for mass killing, instead of being used as a mechanism to reduce ethnic violence. Our efforts build upon Fearon & Laitin (1996), who concede that this mechanism could backfire if an ethnic group announces its intent to attack a rival ethnic group, rather than to pursue cooperation. We develop a computational model that departs from Fearor & Laitin by assuming individuals vary in their propensity to engage in violence, form independent beliefs about ethnic rivals, and respond to catalysts, namely messages about the scale of ethnic voilence that is occurring across the country. In addition, members of the politically dominant ethnic group face sanctions for non-participation. Given these assumptions, our model yields significant variance in the scale of violence across episodes. Our analysis has important implications for the containment of ethnic conflict. We demonstrate that (1) the interaction between nominal ethnic rivals is rarely deterministic, and as a result, the preference for tracking structural factors may be somewhat misguided; (2) that changes in aggregate levels of trust influence the patterns of violence -- communities exhibiting high-levels of inter-ethnic trust are more likely to experience intense episodes of ethnic violence that subside rapidly, in contrast to moderate violence that is sustained over a long period of time; (3) metanorms that sanction non-participants within an ethnic group lead to more extensive episodes of violence.