Urban Interactions

This paper studies social-tie formation when individuals care about the geographical location of other individuals. In our model, the intensity of social interactions can be chosen at the same time as friends. We characterize the equilibrium in terms of both social interactions and social capital (the value of social interactions offered by each agent) for a general distribution of individuals in the urban geographical space. We show that greater geographical dispersion decreases the incentives to socially interact. We also show that the equilibrium frequency of interactions is lower than the efficient one. Using a unique geo-coded dataset of friendship networks among adolescents in the United States, we estimate the model and validate that agents interact less than the social first best optimum. Our policy analysis suggests that, given the same cost, subsidizing social interactions yields a higher total welfare than subsidizing transportation costs.