第三类Tobit模型的半参数估计方法

本文给出第三类Tobit模型的一种新的半参数估计方法。在独立性假设下，利用主方程和选择方程中可观察受限因变量的条件生存函数所满足的关系式，构造第三类Tobit模型参数的一步联立估计量。在已知选择方程中参数一致性估计量的条件下，这种方法也可用于构造主方程模型参数 的两步估计量。本文证明了所提出的一步联立估计量和两步估计量的一致性和渐近正态性。实验模拟表明，我们提出的估计量在有限样本下具有良好表现，且一步联立估计量的有限样本表现优于或接近于Chen（1997）的估计量。

Semiparametric Estimation for the Type-3 Tobit Model

This paper proposes an alternative semiparametric estimation method for the Type-3 Tobit model. Under the independence restriction, we apply a relationship satisfied by the conditional survival function of the limited dependent variables in the main and selection equations and construct a one-step simultaneous estimator for the Type-3 Tobit model. Moreover, given a consistent estimator for the parameter vector in the selection equation, the proposed method can be used to construct a two-step estimator for the parameter vector in the main equation. We establish consistency and asymptotic normality of the one-step simultaneous estimator and the two-step estimator. Finally, Monte Carlo simulations are constructed to show that our estimators perform well in finite samples and that the one-step simultaneous estimator outperforms or is close to Chen (1997)’s estimators.