Selected Research

How does a transitory income shock affect household migration decisions in low-income countries? We show that a drop in income precipitated by an agricultural strike in Mali reduced household migration rates by approximately 32% over a six-year period. The short duration of the strike allows us to cleanly identify the long-run impact of a sudden change in household income on migration choices. Our results demonstrate how cash on hand is a binding constraint to labor migration for a poor population, which suggests that economic development policies intended to reduce migration may end up increasing migration rates.

Uncertainty over the value of migration can create inefficiencies when people remain in areas that offer relatively little economic welfare. I construct and test a new model of migration choice that admits immobility as a rational response to noisy signals over the value of moving to new locations. I show that inefficient location choice can be caused by imperfect information about location match value. Migration as a repeated-choice dynamic problem is computationally intractable, but can be solved as two separable static problems: where and when to go. I show that multi-armed bandit algorithms can derive efficient solutions to these problems under uncertainty. I create a migration video game to capture the dynamics of the model, and test it in an incentivized lab experiment in the United States and Ethiopia. My experiment also includes a randomized treatment that provides extra information over location-match values. The experimental results show that players’ sequential decisions in the lab matched the optimal bound achievable by bandit algorithms. Providing extra information improved optimization by 20%, showing significant participant response to the quantity and quality of information received. My findings show surprisingly optimal performance in solving a dynamic choice problem under uncertainty, suggesting that inefficient migration trajectories are the result of limited and noisy signals rather than behavioral idiosyncrasies. Policies that can provide individualized information over the value of location-matches may provide successful in incentivizing migration from low-welfare areas.


Understanding Migration Choice

Understanding how people make migration choices under uncertainty


The Yale Research Initiative on Innovation and Scale (Y-RISE) advances research on the effects of policy interventions when delivered at scale. While evaluation techniques for pilot-scale programs are well developed, complexities arise when we contemplate scaling up interventions to create policy change.


courses and lectures

I will be teaching a seminar on the politics and economics of migration at Yale for Fall 2019. The course will be cross-listed with the South Asian Studies department.

I also teach the development economics undergraduate lecture ECON 325 at Yale.