Localized Ethnic Conflict and Genocide

Abstract

Building on Fearon and Laitin, who concede in-group policing could be exploited for genocidal purposes instead of moderating interethnic hostilities, the authors seek to explain variation in the scale of ethnic conflict, using data from Rwanda and Burundi. Their computational model assumes individuals vary in their propensity to engage in violence, form independent beliefs about others, and react to public messages about current levels of ethnic aggression. In addition, the dominant ethnic group is subject to genocidal norms—defectors who fail to participate in ethnic violence face sanctions. Their results demonstrate that (1) the scale of violence varies considerably across episodes; (2) interethnic conflicts are not structurally deterministic but rather reflect endogenous interactions; (3) interethnic trust influences patterns of conflict—communities exhibiting high degrees of trust generally experience intense violence that subsides rapidly, in contrast to the persistent, moderate violence characteristic of less trusting communities; and (4) stronger genocidal norms exacerbate ethnic violence.