Do Commodity Price Shocks Cause Armed Conflict?
A Meta-Analysis of Natural Experiments: A Discussion with Darin Christensen and Graeme Blair
This interview with Aaron Rudkin and CEGA affiliates, Darin Christensen and Graeme Blair is part two of a two-part interview series with CEGA affiliates who presented at the 2020 annual meeting of the Empirical Studies of Conflict (ESOC). Every year, ESOC brings together scholars of political violence from across the country to examine the economics of conflict, including crime and policing, refugees, civil war, and foreign intervention.
Below, Christensen and Blair discuss their research about whether large changes in the price of commodities that countries export affects political instability in those countries, as well as their career trajectory and professional motivations. These interviews are excerpts from a larger piece that was originally prepared and published by the UC Institute on Global Conflict and Cooperation.
Q. What is the problem to solve?
A. Many countries — often low-income countries — depend heavily on primary commodity exports, things like palm oil or iron ore. As global demand for these commodities fluctuates, the ensuing price swings can have large economic and political consequences. Our interest is in whether and under what conditions these global price swings generate instability and conflict.
Q. Describe your research.
A. We conduct a formal meta-analysis of 46 high-quality empirical papers in economics and political science that address the question of whether commodity prices shocks impact armed conflict. These papers collectively study more than 200 countries.
Q. What have you learned?
A. When you pool together all different types of commodities, there’s no effect of prices shocks on armed conflict. But when you break out the different types of commodities, we find that price increases for a capital-intensive commodity, such as oil, increases armed conflict. The opposite is true of agricultural commodities, which are labor-intensive. We also find that prices increase for artisanal minerals — think gold and diamonds produced in small-scale mines — increase the likelihood of conflict. This confirms past claims that such minerals are highly lootable and, thus, a lucrative target for attacks when prices are high.
Q. What led you to become political scientists?
A. I like how social scientists think — how we combine theory and data to better understand why social problems arise and, thus, what might be done about them. [Christensen]
I got interested in the politics of natural resources talking to Albert Brownell, a Liberian environmental lawyer and activist, who works with communities that have shut down timber production and use that leverage to bargain with the government and the companies for fair operating terms and compensation. I found the idea that ordinary people can affect social change through disrupting an important source of government revenue really powerful. I spent my graduate work studying how communities living near oil infrastructure in Nigeria organized around the same logic: interrupting oil production through fighting and protests to demand a fair share of oil profits. [Blair]
Q. Why do you care about this stuff?
A. We work with a number of NGOs and governments in resource-rich countries, where these are pressing questions. It’s motivating to think that our research findings could inform the problems they target and the policies they put in place. [Christensen]
Policymakers are increasingly looking to implement evidence-based policies but are faced with an array of mixed if not contradictory evidence. When two sets of prominent scholars publish papers that come to the opposite conclusions, as in this space, it’s hard even for experts to figure out how to make policy decisions. Meta-analysis is one tool to help cut through these disputes and guide decisions about when and where, for example, commodity shocks are likely to be a problem for economic and social policy. It’s a tool that’s very common in medical research, but hasn’t been adopted quickly in the social sciences. If we want policymakers to take academic research seriously, we need to provide tools that guide decision-making. [Blair]
