SMU SOE Seminar (Dec 5, 2018, 11.30-1pm): Large Scale Panel Choice Models with Unobserved Heterogeneity: a Bayesian Data Augmentation Approach

This paper considers the estimation and inference of a logistic panel regression model with interactive fixed effects, where multiple individual effects are allowed and the model is capable of capturing high dimensional cross-section dependence. The proposed model also allows for heterogeneous regression coefficients. For estimating model parameters, new Bayesian and Non-Bayesian approaches are introduced. We investigate the asymptotic property of the estimated parameters. When both the cross section and time series dimensions of the panel go to infinity, we show the consistency and the asymptotic normality of the estimated regression coefficients and the estimated interactive fixed effects. A new information criterion based on the likelihood function is proposed to estimate the dimension of the interactive fixed effects. It is shown that the criterion asymptotically selects the true dimension. Monte Carlo simulation documents the satisfactory performance of the proposed method. Finally, the method is applied to study the New York City medallion drivers' efficiency performance.