有序选择因子结构模型的MCMC估计及应用

MCMC Estimation of An Ordered Probit Factor Structure Model and Evaluation of China's Distributional Treatment Effects

为了评估中国分布类政策效应,根据中国微观数据的变量可得性,在Heckman等构建因子结构模型的基础上,将Heckman基准模型中的连续型测度方程调整为离散型有序选择模型,建立有序选择因子结构模型,并推导出MCMC估计方法。运用该方法,结合中国样本数据,对高等教育的分布类政策效应进行实证估计。有序选择因子结构模型及其MCMC估计方法,对于经济政策的分布类效应评估具有普遍的理论适应性和实际应用价值。

This paper constructs an ordered probit factor structure model through modifying the continuous measurement equation of the baseline factor structure model, which is developed by Heckman, to a discrete ordered probit function to accommodate China＇s micro-data structure. Then, the MCMC estimation method is developed and used to evaluate the distributional treatment effects of China＇s higher education. The ordered probit factor structure model and its MCMC algorithm are both theoretically and empirically useful to evaluate the distributional treatment effects of general economic policy.