Evaluating the welfare effects of a location-based program in the presence of migration: A method and some examples

Estimating the impacts of a place-based policy in a partial equilibrium setting often involves studying the effects on a few outcomes proximate to the policy in question. However, in much of the recent development literature devoted to evaluating the impact of location-specific investments, little attention is focused on measuring welfare effects when adjustments to the policy occur along a different margin: the margin of migration. This paper proposes an alternative method for estimating welfare effects of a place-based policy, taking into account migration and missing land and housing markets. We develop a simple model to show how a new infrastructure investment affects labor market outcomes and migration directly, and indirectly changes the probability of accessing other locally-specific goods. We show that if one can plausibly estimate the migration elasticity in response to the initial infrastructure investment, as well as the change in probability of accessing these other locally-specific goods in response to the new policy (i.e. the increase in the length of a "queue"), under some assumptions on the utility function and using data on prices of these local goods taken from other markets, the total impact on welfare of the initial location-based investment can be bounded.