The Influence of Social Insurance on Wages in China: An Empirical Study Based on Chinese Employee-Employer Matching Data

ABSTRACT

This article uses China Employee-Employer Matching Survey data (CEES) to estimate the influence of social insurance contributions on workers’ wages. The results indicate firms may transfer the increased burden of social insurance onto their workers, and the negative effect of firms’ actual contribution rate is greater for poorly educated workers, migrant workers, non-manager group (manufacturing worker or clerks) and workers in the private sector, small firms, and labor-intensive firms than for their counterparts.

KEYWORDS:

• Social insurance
• wage
• firms’ actual contribution rate
• city contribution
• Chinese firm

Acknowledgments

We are grateful to Professor Takashi Oshio (Hitotsubashi University), Professor Noriyuki Takayama (Research Institute of Policies on Pension and Aging) for their helpful comments and suggestions at the international workshop on “Pension System and Policy Reform: Comparison with China and Japan” on July 28–29th, 2018 in Beijing, China and on December 3–4th, 2018 in Tokyo, Japan. We also acknowledge the anonymous referees and editors of the Journal for their many helpful comments and suggestions, and Dr. Richard Lingwood for his helpful editing work.

Supplementary material

Supplemental data for this article can be accessed on the publisher’s website

Notes

1. Enterprises may transfer the payroll taxes burden to their workers by other means, for example, by decreasing employment, or transferring the burden to the consumer by raising the sale price of goods and services. This will be the subject of future research.

2. For detailed definitions of these three indices, please see Section 3 and Appendix B in this article. Cheng (2018) and Cheng, Du, and Park (2018) use similar indices to make empirical studies for other issues.

3. For empirical studies on Chinese labor market segmentation for the public sector and the private sector, please refer to Chen, Demurger, and Fournier (2005), Demurger et al. (2007); Luo and Li (2007), Zhang and Xue (2008), Ye, Li, and Luo (2011), Demurger, Li, and Yang (2012), and Ma (2009, 2014, 2015, 2016, 2017, 2018b, and 2018a).

4. Ma and Zhang (2018) consider the labor market segmentation by ownership types and analyze the issue, but the labor supply-side factors are not controlled.

5. The legislation for the social security system includes a housing fund but provision of the housing fund is not enforced by the government and therefore many firms have not established a housing fund. The contribution made by firms and workers to the housing fund is different for each firm and region.

6. However, the lowest wage level for the payroll taxes differs for each region. For example, in Beijing, it is 40% of the prior year’s average wage.

7. Firms may also choose to decrease employment, or to increase the sale price of goods and services. These issues should be researched in the future.

8. The economic transition pattern in Russia, and Central and Eastern European countries are distinguished by radicalism (shock therapy, big-bang approach) but by gradualism in China (Lipton and Sachs 1990; Balcerowicz 1994; Liew 1995; Sonin 2013; Shleifer and Treisman 2014).

9. The instrument variable (IV) method can be utilized to address the endogeneity problem. The city contribution rate (CR), the rate of city enforcement of social insurance (CE), and the lagged variables were evaluated as the instrument variables for the endogenous variable (the firms’ actual contribution rate: FAR) in the individual-level and firm-level wages functions but the results based on the Hausman Test indicate they are not appropriate instrument variables. As faulty instrument variables may lead to severe bias the OLS model was selected for the individual level wage function in this study. The endogeneity problem in the individual level wage function is an issue to be addressed in the future research.

10. For the CEES, the firm survey data is a panel data, while the employee data is a cross-section data caused by a change in the worker samples for each survey year; therefore, the clustered OLS model is used for individual-level wage analyses, and the clustered fixed-effect model is used for firm-level wage analyses. As shown in Section 4, the F-test, Hausman Test is used to compare the OLS, the random-effect model and the fixed-effect model. The results indicate the fixed-effect model is appropriate.

11. For the statistical description of dependent and independent variables please see Table A1.

12. In the CEES questionnaires, only the information about the total social insurance contributions is gathered. The five kinds of social insurances shown in Table 1 cannot be distinguished separately in the analyses.

13. In the firm survey questionnaire, the managers responsible for social insurance, human resources or finance are asked to give the firm social insurance contribution rates applicable to the enterprise according to local government regulations, and also the total annual social insurance contributions for the past 3 years. The labor cost refers to salaries, bonuses and social insurance contributions for employees.

14. Based on the data from the national Migrants Survey conducted by the National Bureau of Statistics (NBS) of China in 2012 the participation rates of social insurance of migrants in China are 14.3% (pension insurance), 16.9% (health care insurance), 24.0% (worker injury compensation insurance), 8.4% (unemployment insurance) and 6.1% (maternity insurance) (Cai and Wang 2013).

Additional information

Funding

JSPS KAKENHI (Grant Number JP16K03611 and 18H00863), and National Natural Science Foundation of China (71673295) supported this research.