Economic Shocks and Conflict: The (Absence of) Evidence from Commodity Prices

Abstract

The idea that economic shocks provoke conflict is perhaps the most widely-cited theories and empirical findings in the study of political instability. In developing countries, commodity prices are a large and arguably exogenous source of income levels and volatility. Hence, if economic shocks provoke instability, we should see clear evidence from export price shocks. We do not. We specify two common theories—that fighting varies with the opportunity cost of conflict, and also with the value of state rents—and specify tests that can distinguish between competing accounts. We find no evidence to support either theory, even when conditioning on common sources of conflict risk or disaggregating by commodity. We also show previous commodity price results to be highly fragile or spurious. And by these same robustness techniques, we also see that other economic shocks—like rainfall—are less robust that commonly appreciated. Null and inconsistent results are seldom pursued or published in empirical social science, a practice that we argue has adversely affected our understanding and response to civil wars. We discuss why the quantitative study of civil war may be especially prone to publication bias, and argue that this paper’s results should lessen our confidence in the most commonly asserted income-conflict theories.