The Effect of Non-Employment-Based Health Insurance Program on Firm’s Offering of Health Insurance: Evidence from the Social Health Insurance System in China

Achieving universal health insurance coverage has been a top health policy priority in many countries and has been particularly challenging in developing countries due to the lack of resources. Many countries in the developing world have established a segmented health system, where different population groups defined based on income or employment status have access to different health insurance programs. In countries like China, Mexico, and Indonesia, formally employed workers obtained employment-based health insurance, which has a high premium and more generous benefits; individuals who are not formally employed obtain government-subsidized non-employment-based health insurance, which has a low premium and less generous benefits. Despite that these segmented systems serve the purpose of achieving universal or nearly universal coverage with limited resources, the efficiency of these systems merits further research. Importantly, the subsidized non-employment-based health insurance may reduce firm’s offering of employment-based health insurance. This crowd-out effect would lead to efficiency loss of the health care system and misuse of public funds.

In this study, we examine this crowd-out effect using firm-level data in China. Specifically, we examine the effect of the Urban Residents Basic Medical Insurance (URBMI), a non-employment-based insurance program, on firm’s decision to offer the Urban Employee Basic Medical Insurance (UEBMI), an employment-based insurance program. We take advantage of a novel firm-level dataset: the National Tax Survey Database (NTSD) in 2007-2011. The data come from administrative tax record information maintained by the State Administration of Taxation. The data contain rich information about firm characteristics and a firm’s contributions to employment-based health insurance. To establish a causal relationship, we exploit city-by-year variations in the enactment of URBMI. The program was piloted by the central government in 2007 and was gradually rolled out in all cities over several years. Using a difference-in-differences model and a propensity score matching difference-in-differences model, we find robust evidence that the establishment of URBMI reduced a firm’s offering of UEBMI by 0.93-1.30 percentage points (a 2.3%-3.3% decrease relative to a base offering rate of 40.3%). Subsample analyses suggest that the effect is pronounced among domestic private firms, new firms, and firms that are individual-owned. Moreover, the decrease in UEBMI offering is not accompanied by an increase in wage, implying an overall decrease in total compensation to workers.

These results suggest that the segmented health insurance system in China is potentially associated with a significant welfare loss due to the firm-driven crowd-out effect. The enactment of non-employment-based insurance program led to a significant decrease in employment-based insurance offering by firms. These findings also highlight the importance to evaluate firm behavior when examining health insurance crowd-out effects.