Common factors and common breaks in panels : an empirical investigation

Abstract

Unobserved common factors and common breaks are two important features in empirical studies using large panels. Recently, Baltagi, Feng and Kao (2016, 2019) extended Pesaran's (2006) common correlated effects (CCE) approach by allowing for common breaks and endogenous regressors in large heterogeneous panels. In this note, we empirically investigate these estimators and compare them with CCE and Bai's (2009) interactive fixed effects (IFE) estimators in the context of China's provincial infrastructure investment.