Bayesian Analysis on Dynamic Multivariate Discrete Choice Models and Its Applications

点击翻译按钮获取中文摘要

In this research project, I consider the Bayesian method for estimating the multivariate ordered probit model and the dynamic panel ordered probit model. Further, I propose a method for estimating the coefficients of ordinal endogenous explanatory variables from the posterior results obtained for multivariate ordered probit model. I apply the methods to the estimation of subjective well-being equations by using the micro-level survey data. Since one of the merits of Bayesian analysis is that the value of latent variable for probit models can be obtained directly from the posterior results, I propose a new inequality measure for subjective well-being by using the posterior results.