Should I Stay or Should I Go: Microeconomic Determinants of Migration

Abstract

Moving to a more productive, economically vibrant place offers people the potential to improve their welfare. Yet economic migration remains relatively rare in low-income countries despite large spatial productivity gaps. This paper develops a model where individualized information frictions are sufficient to explain this immobility. The model shows how uncertainty over skill heterogeneity, and which location an individual is best suited for, can lead to ex post welfare losses irrespective of other migration costs. By posing migration as a dynamic choice problem under uncertainty, this paper demonstrates how multi-armed bandit policies admit an optimal solution to location choice. A laboratory experiment demonstrates how participants make optimal location choices conditional on their information set. The addition of an information subsidy illustrates how information is the limiting constraint to producing welfare gains from migration.