Decentralised growth: have central fiscal transfers changed the landscape of the Chinese social security regime?

ABSTRACT

China’s social security expenditure has increased more than twentyfold since 2000. Most studies attribute the increase to local governments, which are responsible for most social expenditure, but some research emphasizes the role of central fiscal transfers. Using provincial-level panel data spanning 2000 to 2017, we find that special transfers significantly stimulate both social assistance expenditure and social insurance expenditure and that general transfers significantly encourage social insurance expenditure. Disparities among social assistance and insurance expenditure across provinces remain largely unchanged. It reveals that central state’s fiscal capacity plays a major role in the expansion of a decentralized social security regime.

KEYWORDS:

• Central fiscal transfers
• decentralization
• social security expenditure
• social insurance
• social assistance

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1. A series of social assistance programmes were established, including minimum living standard schemes for urban residents (2000) and rural residents (2007), employment services for migrant workers (2006) and for college graduates (2009) and medical assistance for the poor (2003). The social insurance system was expanded from covering employees of state-owned enterprises to employees of private and foreign enterprises and was then extended to both rural residents in the form of health insurance for rural residents (2003) and urban residents (2007), and pension insurance for rural residents (2009) and urban residents (2011; Hu, 2016; Ngok & Huang, 2014).

2. The employee social insurance system is made up of five types of insurance: pension insurance, medical insurance, unemployment insurance, maternity insurance and industrial injury insurance. The resident social insurance system covers urban residents who are unemployed and rural residents. It includes urban and rural resident pension insurance and medical social insurance.

3. State-owned enterprises, banks, railways and insurance companies have turned over all enterprise income taxes to the central government since 1994, and local enterprises have submitted 75% of their enterprise income taxes to the central government since 2002.

4. Tax refunds in year t = tax refunds in year _t-1×(1 + tax growth rate in year t)×0.3.

5. General fiscal transfers (yibanxing zhuanyi zhifu 一般性转移支付) were initiated in 1995 under the name of transitional transfers (guoduqi zhuanyi zhifu 过渡期转移支付). They were then renamed as financial transfers (cailixing zhuanyi zhifu 财力性转移支付) in 2002. The current name was finally confirmed in 2009.

6. Social assistance expenditure per capita is the provincial-level social assistance expenditure divided by population. It has been termed ‘social security and employment expenditure’ since 2007. To avoid confusion with the overall social security system, we refer to it as ‘social assistance’.

7. We follow the literature in calculating social insurance generosity using social insurance expenditure divided by the population insured (Huang, 2015). Thus, social insurance expenditure per capita is calculated as the sum of the expenditure of each social insurance programme divided by the population insured. The social insurance programmes include pension insurance for rural and urban residents, pension insurance for urban employees, health insurance for rural and urban residents, health insurance for urban employees, unemployment insurance, maternity insurance and employment injury insurance.

8. In this article, local revenue is defined as locally-sourced revenue which does not include fiscal transfers.

9. In China, 60 is the age threshold for receiving a pension.

10. Labour strength is often measured by the proportion of union members. However, in China a union is more of a quasi-official organization than a basis for workers to organize for their own interests. Thus, we use labour dispute cases (laodong zhengyi 劳动争议), which describe labour–capital conflicts over social insurance, salary, contracts and other labour-related issues, to measure the strength of labour. In fact, about one third of labour dispute cases are over social insurance.

11. Standard errors are clustered at provincial level.

12. Dicky–Fuller tests are used to determine whether a unit root exists.

13. We run models with lagged dependent variables to correct for serial correlation in the error terms (Beck et al., 1998). We also estimate random effects models. See Appendix A1 and A2 for more details.

14. In 2000 and 2017, the population with pension insurance in Guangdong was 9.98 million and 47.18 million, respectively, both ranking first among all provinces. Pension expenditure in 2000 and 2017 was 11.05 billion yuan and 189.8 billion yuan, respectively.

15. Data for Figures 5, 7, 8, and 9 are calculated using fiscal transfers data at the provincial level. The data source is specified in the ‘data sources’ section.

Additional information

Funding

The authors would like to thank the National Social Sciences Fund of China for providing a grant to support this research project (Grant Number: 19BSH042).

Notes on contributors

Xue Li

Xue Li is an associate professor in the Department of Sociology at Fudan University. Her research interests include comparative political economy, social inequality, and global analysis. She has published articles on economic and institutional globalization, social inequality and social welfare policies in China, including American Sociological Review and Chinese Journal of Sociology.

Changquan Jiao

Changquan Jiao is an assistant professor in the School of Sociology at Beijing Normal University. His current research focuses on social welfare, fiscal institutions and the central-local relationship of China. His work has published in Social Sciences in China, Chinese Journal of Sociology, Sociological Studies, and other peer-reviewed journals.