Parameter Heterogeneity,Economic Convergence and Regional Economic Development in China

Zhou Yean,Zhang Quan
2008-01-01
Abstract:In this paper, the authors use quantile regression estimator to study China's urban economic convergence. The existing literature usually adopts conditional mean regression, a method that could hardly model regional heterogeneity as well as different types of convergence. Using cross-sectional data on city level, covering the period from 1988 to 2005, we examine the stylized facts of urban economic convergence in China. OLS estimator and quantile regression estimator extended by Koenker and Hallock (2001) are compared. Evidence strongly supports partial parameter heterogeneity, which is at odds with the OLS results. The results from the quantile regression do not confirm the neoclassical prediction of conditional convergence. We find that convergence is not a common phenomenon across the conditional growth distribution. The regions with growth distribution which in lower quantiles are characterized with conditional convergence, but regions with growth distribution which in higher quantiles are otherwise. These findings are helpful for policy makers to balance China's regional development.
What problem does this paper attempt to address?