Why do some civil wars generate more refugee flows than others?

Abstract

Although civil wars are widespread, they produce highly unequal refugee flows. Because displacement reshapes politics and economies in origin, host, and broader regions, understanding what drives volume is essential for effective management. Existing research on conflict-induced forced migration highlights state violence, local insecurity, and social networks in countries of asylum; however, it tends to underplay rebels—the primary belligerents whose behavior structures civilians’ exit choices. This thesis addresses that mid-level gap by centering rebel–public relations. I theorize that three coercive practices—forced recruitment, child recruitment, and forced funding—generate negative rebel reputations that simultaneously signal personal danger, illegitimacy, and organizational desperation. These signals operate as push factors that increase the propensity to flee. To evaluate the argument, I assemble a new dataset by merging sources on refugee movements, rebel reputation, and intrastate conflicts, yielding 206 country–rebel group–year observations from 1980 to 2011. The statistical results do not support the specific hypotheses. Nonetheless, limitations in data availability, coverage, and operationalization—especially sparse reputation measures and coarse temporal aggregation—caution against strong inferences. Rather than disconfirming the theory, the null findings indicate where measurement and design must improve. Substantively, the project clarifies mechanisms linking rebel conduct to displacement and motivates future data collection on coercion, finergrained temporal designs, and models that incorporate interactions with state behavior and local protection networks.