Growth decomposition bias when accounting for heterogeneous regimes: Evidence from China

When exploring the logic why some economies grow faster than others, previous studies commonly assume that all economies follow a universal growth path. This paper explores the heterogeneity of growth regimes across economies and then investigates the decomposition bias of growth sources in traditional methods. Using a panel data of China's provinces, the empirical results show that a finite mixture model with three classes is best to describe the data, revealing that there are multiple growth regimes across provinces. Also, some provinces switch regimes over time while the others remain stable. Further, neglecting heterogeneous regimes overestimates the importance of factor endowment and underestimates the importance of sector productivity, while it does not greatly influence the importance of of factor market efficiency. In particular, the decomposition bias embodies in physical capital and energy input rather than labor. Our findings indicate that the existing literature may underestimate the contribution of sector productivity. Thus, it is critical to account for heterogeneous regimes when exploring the sources of economic growth.