THE DIFFUSION OF ATROCITIES: A SPATIAL ANALYSIS ON THE ROLE OF REFUGEES

A range of theories have attempted to explain the existing variation in the level of civilian victimization across countries. To date, most of these theories have been focused on the influence of the strategic environment in which these atrocities take place or they have emphasized the organizational difference between the involved actors. Less attention is, however, devoted to the possible spillover effect of these atrocities. This study fills this niche by analyzing the role of refugee flows on the diffusion of atrocities. We do so through statistical analyses of refugee from neighboring countries and the occurrence of atrocities in Africa during the period of 1995-2010, controlling for other possible explanation of atrocities. Our study is the first to systematically examine the effect of refugees on the likelihood of atrocities in refugee-recipient states. We do this by employing a spatial lag model with a temporal component with two different spatial weighting matrices. The preliminary results of the analyses suggest that refugees indeed influence the amount of atrocities and that atrocities are spatially determined. Furthermore, civilian killings is primarily caused by strategic factors such as the number of atrocities and rebel groups in neighboring state and the number of rebel groups and battle deaths in the host country.