摘要
传统的面板数据分析主要是基于条件均值方法进行研究.文章讨论了只含有固定效应的面板数据模型,从理论上提出了两阶段K步差分的分位回归方法,并证明了其大样本性质.同时进行了蒙特卡洛模拟,结果显示该方法在误差非正态时比均值回归方法更有效.最后对我国城镇居民医疗保健消费进行了实证分析,得到了一些有趣的发现,这可以供决策者进行参考.
The study of traditional panel data is mainly based on the conditional mean regression methods.The paper discusses the panel data model with fixed effects and theoretically gives a method called K-step differences quantile regression estimator based on two-stages as a proof of its large sample properties.This method is shown to be more effective than the mean regression by Monte Carlo simulations when the error distribution is non-normal.Finally,the paper draws some conclusions through empirical analysis of the health care expenditure of urban residents,which can be taken as a reference information for decision-making.
出处
《系统科学与数学》
CSCD
北大核心
2015年第9期1037-1048,共12页
Journal of Systems Science and Mathematical Sciences
基金
国家自然科学基金(11271368)
北京市社会科学基金重大项目(15ZDA17)
教育部高等学校博士学科点专项科研基金(20130004110007)
国家社会科学基金重点项目(13AZD064)
中国人民大学科学研究基金(中央高校基本科研业务费专项资金)资助项目成果(15XNL008)
兰州财经大学“飞天学者特聘计划”资助课题
关键词
面板数据
分位回归
K步差分
固定效应
医疗保健消费
Panel data
quantile regression
k-step differences
fixed effects
health care expenditure