Essays on the Economics of Violent Conflict

UoM administered thesis: Phd


This thesis contributes to the literature on conflict in economics with two chapters. Chapter 1 investigates the effect of the 2009 Niger-Delta Amnesty Program on the oil related conflict in Nigeria. Using a difference-in-difference estimation strategy, we compare conflict in Local Government Areas with and without oil fields in the Niger-Delta region. We find robust evidence that the amnesty policy reduced rebel and militia activities by about 27 percent. However, the reduction in conflict was short-lived. Beyond 2014, we do not find any impact of the amnesty program on conflict. Using broad trends in key social indicators, we explore how the increase in conflict in the long-run is due to the amnesty only paying existing rebel while ignoring the economic causes of violence. Further, we find evidence of a peace dividend in terms of increase in economic activities—as measured through night time luminosity data—in Niger-Delta
Local Government Areas with oil fields after the policy.

Chapter 3 is a study of the age-old herder-farmer conflict in North-Central Nigeria which estimates the impact of drought on pastoral conflict. Using an interaction of drought and the availability of pasture, we identify adverse shocks to contested land and estimate its effect on the pastoral conflict. After controlling for cattle production and the spill over effect of drought from herders’ homeland into the contested area, we find local drought effect within the agro-pastoral region. Adverse weather shocks to productive land substantially decrease pastoral conflict in the contested area. We attribute this effect to lower agricultural productivity decreasing farmers’ incentive to engage herders in violence. Importantly, contrary to existing research, we tend to find a positive relationship between precipitation and conflict—a rapacity effect—in the
labour-intensive agricultural sector.


Original languageEnglish
Awarding Institution
Award date2021