Econometric analysis of the social and economic costs of civil war

Beyond the direct effects of gunfire, civil conflicts may also affect populations not directly in harm's way. In Colombia, for example, civil violence is largely a rural phenomenon, so the direct costs of conflict are suffered by residents of the countryside. At the same time, attacks by insurgents have led millions of rural dwellers to flee the countryside for the relative safety of the country's urban areas. The purpose of this dissertation is to examine to what extent these migrations affect the poor living in large metropolitan areas. My estimates suggest that forced migration contributes to deteriorating labor market conditions for low skilled workers in cities, have a crowding effect on educational attainment of poor children and are positively related to an increase in non violent crime.;The case of Colombia offers a different type of instrument with which to study the effects of migration-related labor supply shocks. Detailed data on the location and timing of civil violence and violence-related migration let us establish several important facts. First, large migration flows in Colombia are tied directly to massacres of civilians in rural areas. This suggests that the timing of migrants' migration decisions is driven by the timing of violence rather than by the timing of favorable conditions in destination labor markets. Second, the timing of violence in rural areas is not related to conditions in nearby urban labor markets. Third, workers seeking safety from rural violence generally relocate nearby, most often in large cities. These facts motivate our instrument for migration-related supply shocks. For each urban labor market in Colombia, the instrument is equal to the number of deaths due to civil violence in the previous year, weighted by the inverse of the distance between the urban labor market and the site of the violence.;The first chapter studies the effects of these migrations on urban labor markets. My results show that a 10 percent supply shock increases the likelihood of informal-sector employment by 12.5 percent, has no effect on native employment; and while reducing wages in the informal sector by 2.1 percent, it has no effect on wages in the formal sector. Given the widespread problem of civilian displacement during civil wars in the developing world, and the robust relationship between poverty and civil wars, our results have broad implications for welfare and economic development.;The second chapter focuses on the effects of these migrations on educational attainment, concentrating on the effects of these migrations on schooling decisions of non-migrant children, rather than the migrants themselves. My results indicate that there are adverse effects of these migrations on educational attainment of older children and young adults in urban areas. This result underscores one previously undocumented way that civil war can have long-term effects on economic growth.;The third chapter explains the links between fierce competition for jobs in the legal economy and its contribution towards the expansion of urban crime. My results suggest that at least some of what appears as criminal violence may be more directly linked to political violence related to the civil war.