Bayesian Methods for Economic Panel Data

在本研究中，我考虑了使用马尔可夫链蒙特卡罗（MCMC）对经济面板数据的贝叶斯估计方法。首先，我提出了一种贝叶斯方法来估计Deaton和Muellbauer提出的几乎理想的需求系统，使用包括零支出的面板数据。在模型中明确地处理了零支出问题，并估计了满足累加条件的近乎理想的需求系统。此外，我使用MCMC估计了不可观测参数，包括商品消费、总消费和等效规模，并使用它们的后验分布来计算不平等度量、总消费弹性和价格弹性。其次，提出了基于Hausman和Taylor模型的静态伪面板模型的贝叶斯方法。进一步将静态伪面板模型扩展到动态伪面板模型，并提出贝叶斯估计方法。值得注意的是，与非贝叶斯方法相比，贝叶斯方法不需要额外的变化数据（即工具变量）。我为这个项目写了三篇文章。其中一篇论文已经提交给了一家国际期刊。我让其他人准备好服从。

In this research, I considered the Bayesian estimation methods for economic panel data using Markov chain Monte Carlo (MCMC). First, I proposed a Bayesian method for estimating the almost ideal demand system proposed by Deaton and Muellbauer using the panel data that includes zero expenditures. I dealt explicitly with the problem of zero expenditures in the model and estimate the almost ideal demand system that satisfy the adding-up condition. Furthermore, using MCMC, I estimated unobservable parameters including consumption of commodities, total consumption and equivalence scale and used their posterior distributions to calculate inequality measures, total consumption elasticities, and price elasticities.Next, I proposed the Bayesian method for the static pseudo panel model which is based on Hausman and Taylor model. Further, I extended the static pseudo panel model to the dynamic pseudo panel model and proposed the Bayesian estimation method. It is noteworthy that in the Bayesian approach no additional variation data (i.e., instrumental variables) are required in contrast with non-Bayesian approach.I wrote three articles on this project. One of them has already submitted to an international journal. I make the others ready to submit.