Conflict between returning and non-migrant populations is a pervasive yet frequently overlooked issue in post-conflict societies. While scholars have demonstrated how out-migration can exacerbate civil war, less is understood about what happens when the same populations return. This dissertation interrogates how legacies of forced migration influence conflict dynamics in countries-of-origin. I argue that return migration creates new social divisions in local communities based on where individuals lived during the war – in-country or abroad. These new cleavages become sources of conflict when institutions – like land codes, citizenship regimes, or language laws – provide differential outcomes to individuals based on their migration history. Using ethnographic evidence gathered in Burundi and Tanzania between 2014 and 2016, I demonstrate how refugee return to Burundi after the country’s 1993-2003 civil war created new identity divisions between so-called rapatriés and résidents. Local institutions governing land disputes hardened competition between these groups, leading to widespread, violent, local conflict.

Consequently, when Burundi faced a national-level political crisis in 2015, prior experiences of return shaped both the character and timing of renewed refugee flight. By illuminating the role of reverse population movements in shaping future conflict, this study demonstrates why breaking the cycle of return and repeat migration is essential to conflict prevention.