带固定效应面板分位回归方法的构建与比较  

Construction and Comparison of Quantile Regression Method With Fixed Effects Panel

作　　者：陶丽 邰凌楠 田茂再[1,2,3,4] Tao Li;Tai Lingnan;Tian Maozai(Center for Applied Statistics,Renmin University of China,Beijing 100872,China;School of Statistics,Renmin University of China,Beijing 100872,China;School of Statistics,Lanzhou University of Finance and Economics,Lanzhou 730020,China;School of Statistics and Information,Xinjiang University of Finance and Economics,Urumchi 830012,China)

机构地区：[1]中国人民大学应用统计科学研究中心,北京100872 [2]中国人民大学统计学院,北京100872 [3]兰州财经大学统计学院,兰州730020 [4]新疆财经大学统计与信息学院,乌鲁木齐830012

出　　处：《统计与决策》2020年第17期9-13,共5页Statistics & Decision

基　　金：国家自然科学基金资助项目(11861042)。

摘　　要：文章将现有的固定效应面板分位回归估计方法按估计类型分为惩罚估计、两阶段估计、工具变量估计和最小距离估计,分析各类方法的估计原理及其优缺点。提出了一种基于Hausman-Taylor工具变量的面板分位回归自适应惩罚方法。案例分析主要从供需两方面研究了房价的影响因素,结果发现,房地产开发住宅投资额和上一期房价对于当期房价的影响显著;几种估计方法的系数估计值稍有差别,但走势大致相同。This paper divides the existing methods of quantile regression estimation of fixed effects panel into penalty estimation,two-stage estimation,instrumental variable estimation and minimum distance estimation,and then analyzes the advantages and disadvantages of each reviewed method.Finally,the paper presents an adaptive penalty quantile regression method for dynamic panel data by using Hausman-Taylor instrumental variables.The part of case analysis mainly studies the influence factors of housing price from the two aspects of supply and demand,and the results show that investment in real estate and the previous housing price have a marked effect on current price,and that the coefficient estimates vary slightly in several ways,but the trend is roughly the same.

关 键 词：固定效应 面板数据 分位回归 HT工具变量 

分 类 号：F224.0[经济管理—国民经济]