随机效应变系数空间自回归面板模型的估计

Estimation of Varying-Coefficient Spatial Auto-Regression Panel Model with Random Effects

具有良好可读性和稳健性的变系数模型在各学科领域应用广泛。本文构建了一种新的随机效应变系数空间自回归面板模型,运用截面极大似然估计方法,导出了模型的估计量,证明其具备一致性和渐近正态性,蒙特卡洛模拟研究显示估计量的小样本表现效果良好。

Varying coefficient models with superior readability and robustness have been widely used in a variety of research fields. In this paper, it proposes the new varying coefficient spatial autoregrcssive panel model with random effects. By the profile maximum likelihood estimators are constructed, it gets the consistency and asymptotical normality of these estimators. The Monte Carlo simulation shows that the estimators have good performances under finite samples.