The importance of hate radio pervades commentary on the Rwandan genocide, and Rwanda has become a paradigmatic case of media sparking extreme violence. However, there exists little social scientific analysis of radio's impact on the onset of genocide and the mobilization of genocide participants. Through an analysis of exposure, timing, and content as well as interviews with perpetrators, the article refutes the conventional wisdom that broadcasts from the notorious radio station RTLM were a primary determinant of genocide. Instead, the article finds evidence of conditional media e fects, which take on significance only when situated in a broader context of violence.
Why do some multiparty elections lead to political violence while others do not? Despite extensive literatures on democratization, civil war, and violence against civilians in civil war, the topic of electoral violence has received less attention. We develop a set of theoretical propositions to explain this variation, testing them on an original dataset on African elections from 1990 to 2008. We find that elections in which an incumbent presidential candidate is running for re-election are significantly more likely to experience electoral violence, both prior to the election and after voting has taken place. We argue that clientelism is behind this pattern, and that clients often resort to electoral violence to protect a reliable incumbent patron. On the other hand, when an incumbent candidate is not running for office, we argue that clients are less willing to assume the risks of participating in electoral violence because candidates in the election have not established a record of delivering patronage at the executive level. We further find some evidence that pre-existing social conflicts increase the risk of pre-election violence. We suggest that this finding is due to the tendency of political elites to mobilize voters around pre-existing political and economic grievances to promote their candidacies, in turn heightening tensions and divisions. We also examine, but find little support for, a number of other possible determinants of electoral violence, such as regime type, income level, international observers, ongoing civil war, pathway to power, and first elections after civil war. The article contributes not only to a small but growing literature on electoral violence but also to a burgeoning literature on political behavior in African elections.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.