Essays in Development Economics: On Civil Conflict in Nigeria

Abstract

The three essays in this dissertation explore the cause and consequences of civil conflict in Nigeria. The first chapter examines the effect of international oil prices on civil conflict in Nigeria. The analysis uses time variation in global oil prices and cross-sectional variation based on the initial distribution of oil production across Nigerian districts. According to our estimates, an increase in oil price increases the risk of civil conflicts in districts that produce oil by 50 percent. Using data on intergovernmental transfers, public attitude surveys, labor outcomes and pollution, we test for popular theoretical mechanisms of the resource curse and show that positive oil price shocks affect conflict through rising competition for resource rents and through a reduction in social capital. We do not find evidence in favor of mechanisms related to changes in the opportunity cost of engaging in conflict or grievances about pollution and its implications.The second chapter in this dissertation explores the effect of terrorism on child health. We exploit geographical variation in terror attacks of Boko Haram in Nigeria to show that prenatal exposure to fatalities from terror attacks lead to 0.23 percentage point difference in birth weight between exposed and unexposed cohorts. These effects are stronger for mothers with less education. In addition, we find evidence that parents compensate for exposure to terror during pregnancy by increasing investment in postnatal health care. The final paper in the dissertation examines the impact of violent conflict on household risk coping strategies by estimating the two-fold effect of conflict victimization and income shock on household consumption in Nigeria. Using panel data, we find that shocks reduce food consumption by 17 percent for victimized households but have little effect on nonvictimized households. We test for the mechanisms of consumption smoothing and show that victimized households receive more remittances and invest more in informal support groups. Our results are robust to using a sharp increase in local violence as an event study