Please use this identifier to cite or link to this item: http://hdl.handle.net/1783.1/2567

Simulation estimation of polychotomous-choice sample selection models

Authors Lee, Lung-Fei
Issue Date 2001
Source Proceedings of 13th International Symposium in Economics Theory and Econometrics. In Nonlinear statistical modeling. New York : Cambridge University Press , 2001, P. 71-118
Summary This paper considers simulation estimation of sample selection models. Simulation estimation techniques are useful when sample selection criteria are complicated with polychotomous choice alternatives and correlated disturbances. Simulated likelihood and two-stage estimation methods are considered. For two-stage estimation, possible simulation estimation of outcome regression equations based on Gibbs sampler and variance-reduction techniques are considered. Various generalizations of the GHK simulator for likelihood simulation are possible. Monte Carlo results are provided for finite sample comparison on performance of various simulation methods. Issues on sensitivity of parameter estimates for distributional assumptions are investigated. Semiparametric estimation methods with simulation are also studied in our Monte Carlo experiments.
Subjects
Rights © Cambridge University Press 2001. This paper was (will be) published in Nonlinear statistical modeling and is reprinted with permission.
Language English
Format Conference paper
Access
Files in this item:
File Description Size Format
sim_sel.pdf pre-published version 3342911 B Adobe PDF