Militant Splinter Groups and the Use of Violence

Existing research portrays militant splinter groups as more violent than their parent organizations due to factors like more extreme preferences or capacity-building needs. Though widely held, the assumption that splinters are particularly violent has not been systematically tested. In this paper, we develop and test an alternative explanation for splinter behavior. We argue splinter groups often appear less violent than their parents due to an underlying selection effect. Splinters break away where there are large organizational barriers to internally address a faction's grievances. These barriers tend to exist in well-organized parents that are also capable of high levels of violence. Splinter groups lack this established organizational infrastructure, resulting in lower levels of relative violence. We test this logic with an original dataset on parent and splinter groups and a pair of comparative case studies. We find that splinters are less violent than parent organizations, challenging conventional wisdom.