Structural Changes in Heterogeneous Panels with Endogenous Regressors
This paper extends Pesaran (2006) and Baltagi et al. (2016) by allowing for endogenous regressors and unknown common structural changes in large heterogeneous panels. Thus, an empirically appealing panel data model is provided to accommodate important features of heterogeneity, cross-sectional dependence, endogeneity and structural breaks that revail in applied studies. This model can be estimated by combining Pesaran’s common correlated e?acts (CCE) approach with the least squares method proposed by Bai (1997a, 2010). We show that in this model CCE approach is still valid to deal with cross-sectional dependence due to error factors even in the presence of endogenous regressors and structural changes in slopes and error factor loadings. Monte Carlo experiments are conducted to examine the proposed estimators in this paper.