摘要
本文为一类具有异质性非参数时间趋势的面板数据模型提出了一种简单估计方法。基于局部多项式回归的思想,首先去除数据中的时间趋势成分,然后由最小二乘法来估计公共系数,同时得到时间趋势函数的非参数估计。在一些正则条件下,研究了这些估计量的渐近性质,即在时间维度T和横截面维度n同时趋向无穷时,建立了各个估计量的渐近相合性和渐近正态性。最后通过蒙特卡洛模拟,考查了这种估计方法的有限样本性质。
In this paper, we propose a simple estimation method for heterogeneous nonparametric trending panel data models. We firstly use the local polynomial regression to detrend the data, and then estimate the common coefficients by the pooled ordinary least squares method based on the detrended data. Given the estimators of common coefficients, for each individual we can obtain the estimator for nonparametric trend functions. The paper discusses the asymptotical properties of these estimators. Under some regular conditions, we establish the consistency and the asymptotic normality of the estimators for both common coefficients and time trend functions when T and n pass to infinity simultaneously. A small set of Monte Carlo simulation is conducted to investigate the finite sample performance of our es- timators.
出处
《数量经济技术经济研究》
CSSCI
北大核心
2012年第3期116-127,共12页
Journal of Quantitative & Technological Economics
基金
浙江农林大学科研发展基金(2351000996)资助
关键词
面板数据
异质性
非参数时间趋势
局部多项式回归
Panel Data
Heterogeneous
Nonparametric Time Trends
Local Polynomial Regression
