Regional Divergence of per capita GDP in China: 1991–99

Abstract

Using China's regional data from 1991 to 1999, this paper investigates inter-regional income inequalities. GMM estimation has been used to explore a dynamic panel data model based on the Solow growth model. We find that regions conditionally converge to their own steady states at an annual rate of 8%, indicating around 8 years for a region to halve the deviation from its balanced growth path. However, the panel data exploration of convergence could not explain the catching up phenomenon. To see whether poorer regions can grow faster than richer ones, sigma and absolute beta divergence have been employed. We find that when the regional income gap enlarges during the 1990s, the initially poor regions do not catch up with the initially rich regions.

Keywords:

• Growth
• convergence
• divergence
• income inequality

Acknowledgments

Acknowledgements

The author is very grateful to Dr Maurice Kugler and Mr Raymond O’Brien for valuable comments.

Notes

1.	People's Daily Overseas Editions, October 1995. p. 4.
2.	There are various versions of TFP estimates in the literature. All of them are different from each other because of differences in the assumptions, data sets, estimation methods etc. Hence, we only include labour growth rates in the regression to avoid unnecessary problems, and labour growth rate is approximated by population growth rate.