Using firm-level data for five countries in Latin America, we find a negative and statistically significant link between social conflict in rural areas and ownership of mines. We apply an instrumental variables approach and find that this link may be causal. The instrument employed is altitude of the mine location—which we claim is uncorrelated with the dependent variable, firm ownership—but is correlated with social conflict. This variable serves as an ideal instrument, as it complies with the exclusion restriction. Our results hold to a formal test of changes in specification.

ACCESS PAPER