Understanding the Resource Impact Using Matching

We investigate the resource impact on economic growth using matching. Using a nonparametric minimum-distance matching method, we match the countries according to their observable characteristics, and estimate the relative growth rates of each matched pair. This way we are able to analyze the impact of the resources on relative growth rates, rather than on absolute growth rates as it has been done in the literature. Assuming correlation between observables and unobservables, the matching based on observables may control for unobservables as well. If this assumption is satisfied, matching allows us to control for more variables and to single out the direct effect of the resource abundance variable. We use different measures of resource abundance to check the robustness of such a relationship. The empirical results suggest that there is a strong negative relationship between relative exhaustible resource abundance and relative economic growth. For nonexhaustible resources, the results are mixed, with often a positive impact on relative growth. We discuss the contrary evidence in Sala-i-Martin et al. (2004) and highlight the differences in methodology and estimation that potentially may create differences in the results