A1341
Title: Testing for structural change in heterogeneous panels using common correlated effects estimators
Authors: Peiyun Jiang - Tokyo Metropolitan University (Japan) [presenting]
Abstract: A new test is proposed to detect structural breaks in heterogeneous panel data models with potentially strong cross-sectional dependence. An unknown number of common factors captures the error structure, and correlations between unobserved factors and explanatory variables are allowed. The common correlated effects (CCE) method is applied to eliminate the unknown factors such that it does not require estimating the number of latent factors. The asymptotic analyses indicate that the detecting statistic has the same asymptotic distribution regardless of cross-sectional dependence, as N and T go to infinity. Monte Carlo simulations show good performance of the test in the presence of strong or weak cross-sectional dependence. The method is applied to investigate the relationship between the income of a country and its emissions of chemicals such as carbon dioxide and confirm the environmental Kuznets curve before and after the breakpoint.