Living with conflict: The effect of community organizations, economic assets, and mass media consumption on migration during armed conflict

This dissertation is an examination of migration during armed conflict and the individual and community characteristics that shape this relationship. Although evidence consistently shows that conflict affects migration on an aggregate level, there is little theoretical or empirical work at the micro-level that addresses the individual- and community-level determinants that make people willing and able to migrate during conflict.;To address this gap, I develop multi-dimensional theoretical models that analyze migration decisions at the individual level and the role of community organizations, employment and economic status, and consumption of mass media in systematically altering the way individuals react to armed conflict. Using the recent Maoist insurrection in Nepal as a case study, I empirically test these theoretical models with prospective survey data and detailed records of specific violent events.;I find that specific violent events have different effects on migration, with gun battles increasing and bomb blasts decreasing the likelihood of migration. Within this context, community organizations can provide economic and social support that mitigates the influence of conflict on individuals' lives. I find evidence that organizations such as markets, employers, farmers' cooperatives, and religious institutions dampened the effect of violent events on migration. Economic indicators also moderated the conflict-migration relationship. Location specific characteristics that an individual could lose upon migration, such as employment and land ownership, decreased migration after any violent event. Conversely livestock, which are a more liquid asset, increased migration after violent events. Finally, results show that use of the mass media positively affected migration during the conflict, likely through the provision of information and influencing exaggerated perceptions of threat.;Overall, this dissertation contributes to the academic literature through the construction and empirical testing of theoretical models of individual migration decisions during armed conflict. I demonstrate that individuals react to violence differently, depending upon their individual and community circumstances which affect their experience and perceptions of violence and the utility and ability to migrate away. I also demonstrate that detailed measurement of the specific events that constitute armed conflict is necessary to effectively study subsequent behaviors.