Greed, grievance, or graduates? Why do men rebel?

Education is widely known for its positive effect on conventional politics and nonviolent protest as well as its suppressive effect on violence. However, recent studies have connected education to violence through its effect on relative deprivation and frustration–aggression mechanisms. We address these divergent findings by presenting a theory of the conditional effect that education has on violence. To do so, we build on literature addressing education’s positive association with political participation, such as voting and protests, and question how this relationship translates to contexts in which conventional and nonviolent channels are unlikely to be effective – specifically, in cases of politically excluded ethnic groups. We argue that education increases ethnic group members’ resources and desire to address grievances. Yet, because the ethnic group is politically excluded, opportunities for conventional politics and nonviolent protest are limited. Educated ethnic group members are limited in political options, and are thus more likely to address their grievances through the support of rebel groups, increasing the probability of violence. Violence then occurs in localities where members of a politically excluded ethnic group are located and where those members have higher levels of education. Using geo-spatial data and statistical analysis, we demonstrate that education has divergent effects on violence in areas populated by politically excluded versus politically included ethnic groups in Africa and Central America. Areas with highly educated politically excluded ethnic group members are the most likely to experience violent events.