A0857
Title: Regularized estimation in dynamic panel with a multifactor error structure
Authors: Shou-Yung Yin - National Taipei University (Taiwan) [presenting]
Abstract: The limiting behavior (consistency) of the regularized estimation in the dynamic panel with a multifactor structure is derived. Compared with the OLStype common correlated effects estimator (CCE), the regularized CCE approach does not need to assume that the sample size should be larger than the number of the variables of the approximated model. Therefore, consistency can be established under regularized CCE approach. The simulation results confirm that regularized CCE approach can outperform the OLS CCE approach and is robust even when p is relatively small. We also apply the proposed approach to investigate the determinants of economic growth in 157 countries covered from 1970 to 2019. The results show that after considering the regularization, the patterns of estimated coefficients are different in these countries compared to the results of OLS type estimation.
