Data and Code for: The Long-Term Impacts of Mixing the Rich and Poor: Evidence from Conscript Dorms

To what extent is economic success determined by with whom individuals interact socially? We tackle this question by exploiting a large-scale natural experiment in the Finnish conscription to estimate the effects of peers from different family backgrounds on long-term economic and educational outcomes. Our research design is based on the alphabetization of dorms within squadrons, which is shown to induce as good as random variation in peer composition. Using data on more than 50 thousand conscripts, we find that exposure to a dormmate from a high-income family has a positive impact on earnings, with the largest effect among individuals from high-income families. For them, a one-standard-deviation increase in parental income of dormmates increases long-term earnings by around 5.7%. Effects on earnings of individuals from poorer families are small. The results support labor market networks among the rich as the key mechanism. The findings imply that social sorting reinforces wage inequality between richer and poorer families.