Migration and regional inequality: changing characteristics of China’s economic inequality

Abstract

Scholars studying economic inequality in China have maintained that regional inequality and economic divergence across provinces have steadily increased over the past 30 years. New studies have shown that this trend is a statistical aberration; calculations show that instead of quickly and sharply rising, regional inequality has actually decreased, and most recently, remained stable. Our study suggests that China’s unique migratory regime is crucial to understanding these findings. We conduct a counterfactual simulation to demonstrate how migration and remittances have mitigated income inequality across provinces in order to show that without these processes, we would have seen more of a rise in interprovincial income inequality. We conclude by arguing that inequality in China is still increasing, but it is changing and becoming less place-based. As regional inequality decreases, there are signs that point to the increasing importance of interpersonal inequality.

Keywords:

• China and regional economic inequality
• migration
• remittances
• economic convergence
• hukou system and inequality

Notes

1. Sometimes, the term “regional inequality” is used interchangeably with geographically based inequality. However, this causes confusion given that geographical inequality can emerge along any subnational administrative unit such as regions, provinces, or an existing rural–urban divide. While these units may overlap, they are not always interchangeable, so we attempt to be more specific in our usage of these terms. Specifically, provinces are often aggregated into either regions or coastal and non-coastal zones. However, rural–urban inequality, while geographically based, traverses rather than overlaps the boundaries of regions and provinces.

2. Chan (2007) points out that there are a number of misunderstandings and misconceptions about definitions and statistics used by scholars researching China’s cities and related economic and social phenomena. This includes the multiplicity of urban boundaries, the two statistical systems that exist in China, changing urban definitions, and the hukou system, all of which have all lead to confusion and produced misleading assertions and findings. However, in this article, we focus on the hukou system and how it shapes population statistics that are used to calculate GDP.

3. Li and Gibson (2013, 22) surveyed studies published over the last 15 years that used interprovincial GDP per capita and found that only 4 of 14 had adjusted for the distortions that emerge from using the registered rather than resident population in the denominator.

4. It is very likely that there is a similar issue with intra-provincial migration as well, which would affect calculations of inequality within provinces, as captured by calculations of rural–urban inequality. We correct for this issue across both intra- and interprovincial migration.

5. There are 31 provincial units in the Chinese Mainland. The units are normally divided into eastern, central, and western regions according to the level of economic development. The eastern region includes 11 provincial units: Beijing, Tianjin, Liaoning, Hebei, Shandong, Shanghai, Jiangsu, Zhejiang, Fujian, Guangdong, and Hainan. The central region includes eight provincial units: Heilongjiang, Jilin, Shanxi, Henan, Anhui, Hubei, Hunan, and Jiangxi. And finally, the western region includes 12 provincial units: Sichuan, Chongqing, Guangxi, Yunnan, Guizhou, Tibet, Inner Mongolia, Shaanxi, Ningxia, Gansu, Qinghai, and Xinjiang.

6. These figures are estimates based on census data and sample surveys covering the period 1987–2005.

7. Arthur Lewis initially proposed the dual sector model and the law of migration in 1954 in his article “Economic Development with Unlimited Supplies of Labor.”

8. This claim is contested and evidence is somewhat contradictory. For further details, see Chan (2010), Yuan et al. (2015), and Knight, Deng, and Li (2011).

9. Li and Gibson (2013) point out that the Chinese Government eventually switched from using the registered population for the denominator in per capita economic statistics to using the resident population. According to these authors, this switch creates a discontinuity and reverse trend which happens in 2005. This does not impact our data since it ends in 2005. Also we exclude the 2010 Population Census because it counted migrants twice: at their registered residence and at their permanent residence. Although some statistical manipulations have been developed to deal with this double count, migrants are still overestimated, and we have yet to find an acceptable solution that would make the data compatible (Xi 2011).

10. In statistical yearbooks from each province in China, the per capita GDP in many provinces was calculated using the registered population (hukou population) instead of the resident population, which created distortions in inequality statistics. To avoid this problem, we collected the data on resident population and GDP in each province and then recalculated the per capita GDP.