Civil-Military Relations and Civilian Victimization in Civil War

Why do some states victimize noncombatants during civil war? Scholars have identified regime type, international norms, and battlefield conditions as important factors explaining variation in outcomes. Here I argue that understanding variation in civilian victimization requires the identification of the institutional interests of those in control of the state. Civilian victimization is likely when the military controls pre-war planning and execution because of the institutional goal of winning wars quickly and efficiently by attacking every major source of enemy power. This often includes noncombatants. Most civilian leaders’ institutional goals, however, are centered around governance. Thus, these leaders prefer restraint from victimization because they often believe such barbarity will result in future difficulties for governance. I test my argument alongside others through a binary regression analysis of 103 conflict dyads between 1989 and 2010 and find that the variable for militarily dominated governments maintains significance across model specifications and features the largest effect size of any variable.