Wars Within Wars: Understanding Inter-rebel Fighting

Why do rebel groups frequently fight each other rather than cooperating against their common enemy – the state? This dissertation presents a theory of inter-rebel war and tests it with a combination of case studies and statistical analysis. The theory conceives of inter-rebel war as a calculated response by rebel groups to opportunities for expansion and threats generated by the civil war environment in which they operate. Insurgent organizations attack weaker coethnic groups when government forces only pose a limited threat (i.e., when they face a window of opportunity), so as to eliminate potentially threatening rivals and acquire more resources to be used against the state. Additionally, rebel groups resort to force in desperate attempts to deal with a mounting threat posed by coethnic groups or a drastic deterioration of their power relative to other groups (i.e., when they face a window of vulnerability). Rebel groups’ cost-benefit calculus about infighting is powerfully influenced by whether they are facing coethnic insurgent organizations. Coethnic rebel groups’ overlapping mobilization bases make it possible for an organization to take over the resources (in particular, recruitment pools and tax bases) of defeated rivals and consequently improve their chances in the fight against the government. Thus coethnicity amplifies both defensive and aggressive motives for inter-rebel war. This dissertation adopts a mixed-method approach, combining case studies and statistical analysis. My three main case studies are the Kurdish rebellions against Iraq (1961-1988), the Eritrean war of national liberation (1961-1991) and the insurgencies in Ethiopia’s Tigray province (1975-1991). These case studies combine secondary literature with primary sources collected during fieldwork in Iraq, Ethiopia and several European countries – including fifty-four semi-structured interviews with forty former insurgent leaders, their memoirs, and archival materials. In order to assess the generalizability of my argument across a variety of historical, geographical and political contexts, I also conducted shadow case studies of the civil wars in Lebanon (1975-89), Sri Lanka (1983-2009) and Syria (2011-), and analyzed an original panel dataset of all dyads of rebel groups pitted against the same government in multi-party civil wars in the period 1989-2011