Censored Regression Estimation Under Unobserved Heterogeneity: A Stochastic Parameter Approach

Abstract

This article presents a methodology for addressing the problem of unobserved heterogeneity in the context of regression models based on censored samples. The effectiveness, feasibility, and usefulness of the proposed approach is illustrated by means of an empirical application as well as a simulation experiment. The article demonstrates that censored regressions that control for the presence of unobserved heterogeneity perform substantially better in comparison to their counterparts in which the problem of unobserved heterogeneity is ignored.

KEY WORDS:

• Foreign direct investment
• Maximum likelihood
• Nonnested hypothesis testing
• Semiparametric estimation
• Simulation experiment
• Tobit model