Risk attitudes and migration decisions

This paper analyzes the relationship between individuals' attitudes towards risk and their decision to migrate. We consider migration in the United States across metropolitan statistical areas (MSAs) between 1997 and 2015, based on data from the Panel Study of Income Dynamics (PSID). Using random-effects specifications, we find that being relatively more willing to take risks is positively and significantly related to cross-MSA migration. For cross-state migration and migration across larger distances, risk attitudes are slightly more important. A recursive bivariate probit model of simultaneous equations addresses the potential endogenous nature of attitudes towards risk and allows ruling out reverse causality.