基于面板数据的分位数回归及实证研究

Research on Quantile Regression in Panel Data Modeling and Its Application

在面板数据模型和分位数回归原理分析的基础上,对分位数回归方法在面板数据模型中的应用作了深入的研究.利用1998-2007年间31个省市的地区生产总值与专利申请量的面板数据分析了自主创新与经济增长的关系,比较了最小二乘法和分位数回归在面板数据模型中的估计结果,结果显示分位数回归方法在面板数据模型估计时具有一定的优势.

This article principally introduces the concepts and estimation equation of quantile regression in panel data modeling and its application. It analyze the relationship between endogenetic innovation and economic growth with actual data, and contrast the estimation efficiency of two kinds of methods by theoretical and empirical analyses： LS, PLS and QR, PQR, and find that PQR or QR has certain relative advantages over LS or PLS on estimation of panel data models.