PUBLICATIONS & REPORTS

Contents Introduction

This paper adopts the distribution approach to study convergence across Chinese provinces. In particular, it uses the Markov transition matrix methodology to capture the dynamics embodied in the data and to produce corresponding ergodic distributions. The results indicate that distribution of per capita income across Chinese provinces has become bi-modal over the period of 1952 to 2003. This conforms to the general impression of rising regional disparity in China and is therefore discouraging from the point of view of convergence. However, examination of the dynamics of the post-reform period separately shows that the ergodic distribution on the basis of this period’s dynamics is negatively skewed, with more provinces at the higher end of the distribution. This provides some ground for optimism regarding eventual regional convergence in China.