Semiparametric qualitative response model estimation with unknown heterosecedasticity or instrumental variables

摘要

This paper provides estimatiors of discrete choice models, including binary, ordered and multinomial response (choice) models. The estimators closely resemble ordinary and two-stage least sqares. The distribution of the model's latent variable error is unknown and may be related to the regressors, e.g., the model could have errors that are hetero-scedastic or correlated with regressors. The estimator does not require numerical searches, even for multinomial choice. For ordered and binary choice models the estimator is even for multinomial choice. For ordered and binary choice models the estimator is root N consistent and asymptotically normal. A consistent estimator of the conditional error distribution is also provided.