Revisiting Violence in Refugee-Hosting Areas
Why are some refugee-hosting areas targeted frequently for attack by armed groups while others experience relatively little violence? Recent subnational level research shows that the presence of refugees does not increase the likelihood of civil conflict, but locations that host refugees suffer higher rates of civilian victimization. Despite often being on the periphery of a country, refugee-hosting areas generally experience an influx of humanitarian aid and an increased presence of state authorities, making them attractive targets for armed groups. Yet we still know little about the geography of civilian abuse, particularly how geographical variables influence where and when violence occurs. In this paper, we utilize new data on refugee settlement locations and characteristics to examine how patterns of violence in refugee-hosting areas are affected by geographical and temporal considerations, including proximity to the capital, location on a border, state presence along the border, refugee settlement patterns, and timing near host country elections. We theorize that these factors influence the opportunity and willingness of sending and host country conflict actors to perpetrate such abuse. Our findings have implications for the development of effective strategies to protect refugees and other civilians and bolster host state security.