Skip to content

Global: Commodities, long term development, and political instability

coffeeMost developing countries are dependent on a handful of primary commodities, and some of these commodities are more price volatile than others. What is the effect of these (mostly unpredictable) commodity shocks on economic development and politics? The answer is an ongoing side project.

For the most part, commodity booms appear to be good for growth in developing countries, while commodity busts are bad. This is not true as a rule, since commodities can be both a hope and a curse.

What seems to be especially bad is volatility. In a recent paper with Jason Hwang and Jeff Williamson, I develop a century-long time series of country commodity price indices (available for download). We find that if we look at development in the long run–from 1870 to the Second World War–countries whose commodity endowments were more volatile in prce grew much more slowly. In fact, commodity volatility seems to be among the best long-term predictors of which countries grew rich while others remained poor.

Why might commodity busts and commodity volatility reduce growth? Many reasons, but one candidate is political instability and war.

When prices go down, the people who produce them get poorer and may be more inclined to revolt. That is one story told. Another story focuses more on the center. In flush times, governments tend to invest more than they save, and often these investments are bad and difficult to reverse. When prices fall, governments bankrupt themselves and lose control of the state, empowering rebellious groups. Lower prices might also mean lower rents to extract, and so leaders might have an incentive to loot the country rather than maintain stability.

Some theories, though, predict the opposite relationship. Resources make capturing the state more valuable, and so rising prices could prompt war. These and other theories are discussed in my JEL review of the civil war literature with Ted Miguel.

In a long-running project with Sami Bazzi, I am looking to test these theories empirically with new and better commodity data–more commodities and for more country-years than ever before.

Our preliminary result: neither side seems to win. There is little evidence of a link between commodity shocks and conflict in any direction. This is true unconditionally and conditional on any number of risk factors. We can also decompose price shocks into state-centric, like oil and minerals, and peasant-centric, like agricultural crops. Even these shocks show little relationship with instability. We can reproduce others’ results to the contrary, but we show most of this evidence to be extremely fragile.

Papers