Estimating Separable Matching Models
summary
Proposes two practical estimators for separable transferable-utility matching: a minimum distance estimator that matches observed cell shares to the model using the generalized entropy from Galichon–Salanié, and a Choo and Siow reformulation as a generalized linear model with two‑way fixed effects that reduces to weighted least squares. Methods deliver closed‑form mappings, a chi‑square specification test, fast Python implementations (sub‑second on examples), and extensions to imperfect transferability and many‑to‑one or multipartite settings.
AI-written summary · about