We consider the one-to-one matching models with transfers of Choo and Siow (2006) and Galichon and Salanié (2015). When the analyst has data on one large market only, we study identification of the systematic components of the agents’ preferences without imposing parametric restrictions on the probability distribution of the latent variables. Specifically, we provide a tractable characterisation of the region of parameter values that exhausts all the implications of the model and data (the sharp identified set), under various classes of nonparametric distributional assumptions on the unobserved terms. We discuss a way to conduct inference on the sharp identified set and conclude with Monte Carlo simulations.
One-to-One Matching; Transfers; Stability; Partial Identification; Nonparametric Identification; Linear Programming;
Cristina Gualdani, and Shruti Sinha, “Partial Identification in Nonparametric One-to-One Matching Models”, TSE Working Paper, n. 19-993, February 2019, revised June 2019.
TSE Working Paper, n. 19-993, February 2019, revised June 2019