From Brothers in Arms to Enemies? Dynamics of Infighting in Multiparty Civil Wars

Infighting in today’s complex multiparty civil wars, or what some call 'rebel fratricide', has received increased attention in the recent literature. We argue, however, that there is a tendency to overstate the frequency and significance of infighting. Particularly, our understanding of the phenomenon is hampered by two persistent gaps in the literature: First, there is a lack of a clear and coherent conceptualization of what kind of conflict between which groups meaningfully constitutes infighting. There can be a variety of social relationships between non-state groups in an armed conflict, which can be inherently more or less conflictual. Specifically, it is hard to explain why groups on the same side of a conflict should divert valuable resources from their struggle against the government to fight each other, as they face high opportunity costs (namely the neglect of the fight against the government) to do so. Especially for groups from the same social movement, mediating institutions should exist to minimize this risk. As shown by the case of Syrian Islamist opposition groups, these local modes of conflict resolution actually limited violence to a few peaks. Secondly, there is a lack of theoretical perspectives considering conflict dynamics: While infighting is an undeniable part of many multi-party civil wars, it represents only a small proportion of the overall activity of rebel groups. Moreover, it does not occur on a constant rate, but is concentrated in space and time, pointing to a need to take into account changing developments on the battlefield and in the overall conflict. Consequently, this paper addresses not only the puzzle which type of groups are poised to fight each other, but also under what conditions they are more likely to do so. Going beyond simple measures of ideology, we develop a typology of group relationships, ranging from groups standing actually on opposing sides of the conflict to groups belonging to the same movement or to different factions of the same parent organisation. While rebels might in principle share the same goals, they may diverge on how to achieve these goals or simply compete for support of a shared constituency or scarce resources. In line with our second argument, we theorize that violent conflicts between these groups are expected to break out when opportunity costs are low, which can be the case either when government pressure decreases and/or groups succeed on the battlefield (fighting for spoils) or when defeat is imminent and groups jockey for position and survival (fighting for scraps). When violence cannot be constrained by institutions and mechanisms of conflict resolution, the consequence of infighting can be the demise of the inferior groups or the split of a rebel movement. We test out assumptions both with a quantitative and qualitative approach, utilizing large-N event-data of armed conflict from the UCDP as well as a detailed case-study of the Northern theatre of the Syrian civil war based on extensive field research and interviews with members of various violent groups.