Working Paper: NBER ID: w26868
Authors: David Lagakos; Samuel Marshall; Ahmed Mushfiq Mobarak; Corey Vernot; Michael E. Waugh
Abstract: Recent studies find that observational returns to rural-urban migration are near zero in three developing countries. We revisit this result using panel tracking surveys from six countries, finding higher returns on average. We then interpret these returns in a multi-region Roy model with heterogeneity in migration costs. In the model, the observational return to migration confounds the urban premium and the individual benefits of migrants, and is not directly informative about the welfare gain from lowering migration costs. Patterns of regional heterogeneity in returns, and a comparison of experimental to observational returns, are consistent with the model’s predictions.
Keywords: Migration; Development; Poverty; Economics
JEL Codes: O11; O18; R23
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| observational returns to migration (J61) | welfare gains from reducing migration costs (F16) |
| lower migration rates (J61) | higher observational returns (I26) |
| skill composition of migrants (J61) | observational returns to migration (J61) |
| observational returns to migration (J61) | productivity of urban areas (R11) |
| experimental returns to migration (F22) | observational returns to migration (J61) |
| observational returns to migration (J61) | joint distribution of benefits and costs associated with migration (F22) |