Escaping from hardship, searching for comfort: Climate matching in refugees’ destination choices

Do refugees settle in destinations that are ecologically similar to their origins? We assess the relevance of “climate matching” theories of migration for Venezuelan refugees in South America. Leveraging social media data, we build the first local, bilateral matrix of Venezuelan flows across the region. Validation exercises show that this matrix aligns with independent data on Venezuelan refugees at the bilateral, destination and origin levels. We measure bilateral ecological similarities in terms of temperature, precipitation, elevation, and distance to the coastline. Performing Poisson Pseudo-Maximum Likelihood gravity models of migration, we show that Venezuelan flows are more likely between ecologically similar areas. Estimates are economically significant, and model predictions explain independent measurements of Venezuelans’ settlement choices at both bilateral and destination levels.