We reevaluate the hypothesis and empirical result that ethnic civil wars lead to higher skilled emigration (Bang and Mitra, 2013). We develop a simple conceptual framework that predicts contrasting results depending upon if the economy is assumed to be agglomerating in skilled labor or non-agglomerating with network effects. In the latter case, non-ethnic wars may lead to higher skilled emigration. A regression model that accounts for the time-varying definition of migration and includes important explanatory variables shows that non-ethnic wars as opposed to ethnic wars may lead to more skilled emigration.