讲座：Estimating Panel Data Models with Common Factors 发布时间：2023-06-02
题 目：Estimating Panel Data Models with Common Factors
嘉 宾：冯 曲, Associate Professor, Nanyang Technological University
Estimating panel data models with common factors or interactive fixed effects has been extensively investigated in the recent panel data econometrics literature. To model the potential correlation between regressors and error factors and loadings, we generalize Mundlak-type projection to the form of interactive fixed effects. After the projection, the errors are serially correlated, cross-sectionally dependent, and heteroskedastic. Thus, we consider robust testing techniques and wild bootstrap procedures for inference. In addition, we also consider the case of endogenous regressors by combining the Mundlak-type projection with the control function approach. Finally, Monte Carlo simulations are conducted to verify the properties of the proposed estimators and compare them with existing estimators in the literature in finite samples.
冯曲 FENG Qu is currently the head of Economics Division, and Associate professor at School of Social Sciences, Nanyang Technological University, Singapore （南洋理工大学）. His research interests include econometrics, Chinese economy and financial markets. His research papers appear in top economics journals, including Journal of Econometrics, Journal of Development Economics, Journal of Applied Econometrics, Journal of Economic Behavior and Organization, etc. He is currently also an associate editor of Singapore Economic Review, a journal of the Economic Society of Singapore.