Description
Studies of civil war have begun to explore more acts by insurgents than their use of violence against opponents and civilians, particularly on rebel governance. However, we know little about insurgents’ use of non-lethal violence during conflict despite anecdotal evidence about rebels issuing threats, blocking roads, kidnappings, or in other ways impede on everyday life in conflict zones. These activities rarely lead to fatalities or involve a “battle” with government troops so are excluded from datasets commonly used for the study of civil war. In this paper, we develop a theory about rebels’ use of non-lethal violence as determined by temporal and spatial conditions. We contend that rebels will primarily use non-lethal violence close to stronghold areas as they need to both emphasize their presence but not aggrieve the local population through killings. We also expect there to be shift in strategy during negotiations when we expect an overall shift towards non-lethal acts in all areas for similar reasons – reducing the killing signals a willingness to participate in the peace process, while non-lethal violence signals the continued relevance of the group. We systematically explore this using sub-national data from Colombia, Indonesia, and Thailand and a novel multi-level modelling strategy.
