The effect of oil and diamonds on democracy: is there really a resource curse?

This paper examines the effect of natural resources on the level of democracy in a set of countries. The main model is a fixed effects regression model, where the focus is on within-country variation over time. The effect of different resources is investigated, namely the effect of oil, diamonds and agriculture. Furthermore, a distinction is made between two broad types of resources, diffuse and point resources, to explore whether the effect on democracy is similar or different. Criticism in existing literature on the presence of the resource curse is taken into account. Production data on natural resources is used and not the common variable ‘resource exports-over-GDP’, the latter being flawed. A possible endogeneity problem is taken into account, as well as the persistence of democracy over time. I find evidence for a resource curse of oil on democracy. It is present in different model specifications, such as models with either fixed effects or a lagged dependent variable. There only seems to be very weak evidence for a negative effect of point resources on democracy, compared to the effect of diffuse resources. It is argued that this might be due to the geographic concentration of these types of resources, which enables governments and elite groups to capture resource rents.