A closer look at oil, diamonds, and civil war

Abstract

Studies of natural resource wealth and civil war have been hampered by measurement error, endogeneity, lack of robustness, and uncertainty about causal mechanisms. This paper develops new measures and new tests to address these problems. It has four main findings. First, the likelihood of civil war in countries that produce oil, gas, and diamonds rose sharply from the early 1970s to the late 1990s; so did the number of rebel groups that sold contraband to raise money. Second, exogenous measures of oil, gas, and diamond wealth are robustly correlated with the onset of civil war. Still, these correlations are based on a small number of cases, and the substantive effects of resource wealth are sensitive to certain assumptions. Third, petroleum and diamond production lead to civil wars through at least three different mechanisms. Finally, the only resource variable robustly linked to conflict duration is a measure of "contraband," which includes gemstones, timber, and narcotics.