Robust Two-Stage Estimation in General Spatial Dynamic Panel Data Models  

作　　者：DING Hao JIN Baisuo WU Yuehua 

机构地区：[1]Department of Mathematics and Statistics,York University,Toronto,Ontario,M3J 1P3,Canada [2]Department of Statistics and Finance,School of Management,University of Science and Technology of China,Hefei,230026,China

出　　处：《Journal of Systems Science & Complexity》2023年第6期2580-2604,共25页系统科学与复杂性学报（英文版）

基　　金：supported by the Natural Sciences and Engineering Research Council of Canada under Grant No.RGPIN-2017-05720;the National Natural Science Foundation under Grant Nos.12201601,71873128,11571337,71631006,and 71921001;the Anhui Provincial Natural Science Foundation under Grant No.2208085QA06。

摘　　要：This paper proposes a robust two-stage estimation procedure for a general spatial dynamic panel data model in light of the two-stage estimation procedure in Jin,et al.(2020).The authors replace the least squares estimation in the first stage of Jin,et al.(2020)by M-estimation.The authors also provide the justification for not making any change in its second stage when the number of time periods is large enough.The proposed methodology is robust and efficient,and it can be easily implemented.In addition,the authors study the limiting behavior of the parameter estimators,which are shown to be consistent and asymptotic normally distributed under some conditions.Extensive simulation studies are carried out to assess the proposed procedure and a COVID-19 data example is conducted for illustration.

关 键 词：Asymptotic normality CONSISTENCY M-ESTIMATION model selection OUTLIERS 

分 类 号：O212.1[理学—概率论与数理统计]