Do Commodity Price Shocks Cause Conflict?

The natural resource curse (sometimes called “Dutch disease”) was one of my first fascinations in development economics. It represents the apparent “paradox” of a boom in natural resource wealth leading to less economic growth. There are, of course, numerous theories as to why this observation persists. One popular theory, that is repeatedly tested empirically, is that sharp and dramatic changes in the prices of these resources lead to conflict, which in turn slows economic growth.

A new review article—by Graeme Blair, Darin Christensen, and Aaron Rudkin—performs a meta-analysis on the numerous natural experiments (350 papers since 2002) that provide the opportunity to credibly estimate the effect of commodity price shocks on conflict. I found this review article helpful and I will highlight some of the important insights.

(1) Commodity price changes, on average, do not cause conflict

The authors point out that prominent studies in this literature come to seemingly contradictory findings. For example, Dube and Vargas (2013) find that when the international price of oil rises violence increased in Colombia’s oil-producing municipalities whereas Bazzi and Blattman (2014), using data from most countries in Africa, the Middle East, Latin America, and Asia from 1957 to 2007, find that price shocks have no effect on the onset of conflict.

Blair et al. perform a formal meta-analysis, which here means they standardize 102 estimates from 46 empirical studies and re-analyze results when necessary (e.g., when results are estimated with a probit regression). The authors find that when they pool studies together across all commodity types, price shocks have no effect on conflict. This null effect, however, masks important heterogeneity.

(2) Price increases of labor-intensive agricultural commodities reduce conflict

When only looking at agricultural commodities, which tend to be relatively labor-intensive (compared to capital inputs), price increases reduce conflict. This is consistent with existing theory predicting that price increases for labor-intensive commodities generate new local employment which, in turn, increases the opportunity cost of fighting (Dal Bo and Dal Bo 2011).

This effect, that increasing prices of labor-intensive commodities decrease conflict in places producing those commodities, is often referred to as the “opportunity cost” effect. This theory, which largely builds on the work of Becker (1968), highlights the idea that “rational” actors engaging in conflict require an account of the relevant alternative activities. When a price increase in some commodity also increases the profitability of an alternative activity, then this increased opportunity cost of conflict can actually discourage conflict activity.

(3) Price increases of oil, a capital-intensive commodity, increases conflict

When only looking at oil and gas, which tend to be relatively capital-intensive (compared to labor inputs), price increases increase conflict. In contrast to labor-intensive commodities which translate price increases to expanded local employment, capital-intensive commodities increase the returns to fighting by gaining control of commodity production areas (Fearon and Laitin 2003).

This effect, that increasing prices of commodities increase conflict in places producing those commodities, is often referred to as the “rapacity” effect. This is one of the most well-known theories of conflict and rests firmly on the economic-based motivations of “rational” actors. In short, price increases of a given commodity increase the “prize” to be won by controlling the production of the commodity.

(4) Price changes for “lootable” artisanal minerals increase conflict

Finally, price increases for “lootable” artisanal minerals—such as diamonds and gold—increase conflict. This “lootable” qualifier seems important here. Minerals, especially gold, are relatively easy to melt down and obscure any trace of origin. This tends to enable rebel groups to profit from the extraction and trade of these minerals at a lower cost than non-lootable minerals.

Although I think adding a discussion of the effect of rising food prices and food abundance on conflict could be worthwhile in this review, I really appreciate the effort of Blair et al. in conducting this meta-analysis. Such work is tedious and often under-rewarded. In summarizing the results from a literature as rich and important as the literature on commodity price shocks and conflict, this paper is valuable. I’d encourage anyone interested in conflict studies to read this meta-analysis in more detail than I am highlighting in this short blog post.