Selection and Heterogeneity in the Returns to Migration

This project estimates who gains from internal migration and by how much, developing a grouped-random-coefficients GMM framework that recovers heterogeneous returns from long household panel surveys in Indonesia, Tanzania, and China. The framework extends estimation methods whose core machinery is published at Econometrica to longer panels, many migration trajectories, unbalanced histories, and weak-identification-robust inference for trajectory-specific returns and counterfactual aggregates. A Monte Carlo simulation study, run on NCI Gadi under a Macquarie Research Compute & Storage Allocation, validates the extended estimator's finite-sample performance ahead of journal submission.