The distribution of benefits under China’s new rural cooperative medical system: evidence from western rural China

As one of three basic medical insurance schemes in China, the New Cooperative Medical Scheme (NCMS) is one of the most-widely covered health insurances in the world. It covers more than 800 million rural Chinese, which accounts for about 98.7% of the total rural population [1]. The scheme, which was started in 2003, aims to ensure that rural residents of China receive basic health care services, thus reducing the medical burden and achieving social fairness [2]. As a voluntary and governmentally organized scheme, it is largely financed through government subsidization and to a lesser extent through flat-rate household contributions. Unlike other social health insurance schemes with an income-related contribution (e.g., China’s Urban Employee Basic Medical Insurance), the NCMS is characterized by enabling insured citizens to enjoy the same benefit package including inpatient and outpatient by paying a flat-rate premium [3]. However, the equal policy design of the NCMS does not necessarily result in the equitable distribution of benefits.

As a means of risk-sharing, medical insurance protects the savings of enrollees from catastrophic medical costs. In addition, the income transfer in medical insurance may permit the purchase of health care that is beyond the consumer’s budget constraint and access to health care makes insurance very valuable [4]. Medical insurance achieved equitable financing and financial protection by levies insurance premium based on the enrollees’ ability to pay, while offering health services free of charge as the need arises [5]. Under the cost-sharing program (i.e., in deductibles, coinsurance, and copayments), the moral hazard effect of medical insurance can be reduced; however, the distribution of benefits may be undesirable [6]. The amount of the distributed benefits depends on both utilization and consumption of health care services, which are primarily driven by the demand for health care [7]. To a large extent, the use of services can be affected by budget constraints, i.e., enrollees with lower income have a smaller likelihood of consumption [8, 9]. Striking evidence indicates that a cost-sharing program leads to a pro-rich horizontal inequity in the distribution of health care services [10‐12]. Hence, a flat-rate premium of the NCMS combined with the potential inequity of utilization [13‐16] may result in an inequality of insurance benefit distribution.

Policy details of NCMS benefit packages play a role in balancing the distribution of benefits across an economic gradient. In terms of the NCMS policy design, different deductible and coinsurance rates were set based on the levels of medical institutions; here, a lower level of medical institutions indicates lower deductible and coinsurance rates. Table 1 shows the details of the NCMS design, using a county of western rural China as an example. Here, in 2013, deductibles ranged from 0 Yuan (1 US$ = 6.85 Yuan at current exchange rates) for township health centers (institutions at the lowest level that provide inpatient services) to 1500 Yuan for provincial hospitals, and reimbursement rates range from 90 to 100% for township health centers to 50% for provincial hospitals. These policy details are often considered to be positive because they indicate that seeking lower-level health care is likely to result in higher effective reimbursement. Different cost-sharing arrangements are partially designed to induce hesitation in enrollees before seeking health care and to forgo the use of services that are expected to result in a lower reimbursement rate. There is striking evidence that economically disadvantaged individuals, who were more sensitive to the price of health services were more likely to seek treatment in primary-level health-care facilities than economically advantaged individuals [7, 15]. Accordingly, whether and to what extent the distribution of benefits differs across the economic gradient under the NCMS in China requires close examination.

Several empirical studies have attempted to explore the distribution of benefits of medical insurances among insured populations across different economic levels. For example, Huang et al. analyzed the net benefit distribution under Taiwan’s National Health Insurance program and reported that an apparent pro-poor pattern was observed under wage-based premiums; however, in terms of flat-rate premiums, such a pro-poor pattern was less obvious [17]. This study provides examples for the net benefit distribution of co-financing models by employees, employers, and governments. Pan et al. and Tian et al. analyzed the distribution of benefits under China’s Urban Resident Basic Medical Insurance (URBMI) targeting the urban population with an exception of formal sector employees; the authors reported that the rich benefit more from this particular insurance [18, 19]. Wang et al. analyzed the equity of benefits under the rural Cooperative Medical System (CMS) in China and concluded that under the “low premium and high coinsurance benefit design” model, regardless of the individual health status, high-income groups always benefit more from the CMS than low-income individuals [20]. These findings are not fully applicable to the NCMS. The insurance financing mode of Taiwan favors the poor, while the NCMS funding is equal for each insured citizen, which means that all enrollees pay the same premium regardless of their income. Several differences exist in the target population and benefit packages, cost-sharing arrangements, and financing modes between NCMS and URBMI or CMS [3, 21, 22]; therefore, analyzing the findings of the benefits distribution under URBMI and CMS is not fully generalizable for the NCMS. Few recent studies about insurance benefits have attempted to look into the magnitude of economic-related horizontal inequity or have ignored enhancement of NCMS altogether [23, 24].

The NCMS has undergone rapid enhancement and changes during recent years [25]. To equip the population with both affordable and equitable basic health care, the Chinese government unveiled a new Health Care Reform in 2009 [26]. Under this plan, the Chinese government enhanced its investment in basic medical insurance schemes. According to macroeconomic data, the expenditure of China’s government on the subsidy of medical insurance schemes reached 442.9 billion Yuan (roughly US$ 64.7 billion) in 2013, which amounts to 46% of the total government health expenditure [1]. This figure was close to three times the amount of 2008 [1]. The NCMS has improved with the growth of financing and coverage, which led to an enhancement of the reimbursement ratio and financial protection. Table 1 shows the changes of NCMS financing and benefit packages from 2008 to 2013. It became clear that, compared to 2008, the NCMS has greatly improved the financing and benefits package, i.e. higher reimbursement rates, higher reimbursement ceiling, expansion of the scope of reimbursement, simplification of the reimbursement method, and lower deductibles in township health centers. In the data used for this study, the ratio of actual out-of-pocket payment was approximately 50% of the total medical expenditures for NCMS inpatients in 2013, while it was 70% in 2008.

So far, empirical studies evaluating the equality in NCMS benefits under the background of the increasing benefit design and government subsidies of the scheme are limited. Theories associated with social health insurance indicate that the change of insurance generosity may affect the individual demand for health care by changing the out-of-pocket payment at the time of consumption, and furthermore affecting the distributional consequences [6, 7, 27]. Within the health area, the renowned RAND health insurance experiment of the 1970s investigated how copayment affects the use of health care services. Analyses of the RAND experiment reported that participants with lower income were more sensitive to cost-sharing than those with higher income; furthermore, participants with higher deductibles and coinsurance rates reduced their use and expenditures of health care services [8]. In addition, previous studies reported that simplification of the reimbursement procedures and expansion of the scope of reimbursement may improve health care utilization among the poor [13, 28]. Therefore, the distribution of benefits across economic gradients may be improved in response to the improved generosity of the NCMS.

The main purpose of this study was to investigate whether and to what extent the distribution of benefits differs across economic groups in 2008 and 2013; furthermore, whether and to what extent the magnitude of inequality in benefits changed under the background of increasing benefit design of NCMS was also investigated for the time-frame from 2008 to 2013. The equity of the basic medical insurance system is a central indicator to evaluate the New Health Care Reform [29]. Our results provide evidence for the further reform of the basic medical insurance in China and provide meaningful policy implications toward universal coverage of health insurance among other countries that face similar challenges.

Although the goal of the NCMS is to provide equal financial protection for all enrollees, the distribution of benefits between the rich and the poor is not necessarily equitable. In this study, using two variables as proxies to measure benefits, i.e., the probability of receiving reimbursements and the absolute amount of received reimbursements, we assessed the distribution of NCMS benefits. This was conducted under the background of the increasing benefit design of the NCMS, using two-wave cross-sectional household health surveys conducted in 2008 and 2013. We found that the benefits of the NCMS are concentrated among higher economic groups; however, horizontal inequities in the NCMS favorably benefiting the rich decreased over the investigated period, while the level of insurance benefits increased.

Specifically, the regression results firstly show that, compared to the poor, the rich benefit more from the NCMS. These results are consistent with recent studies, showing that insurance benefits were more directed to the rich under the Urban Residents’ Basic Medical Insurance in China [18, 19]. The analysis of inequality further indicates clear pro-rich horizontal inequity in benefits distribution under the NCMS in both 2008 and 2013. Moreover, the inequality-reducing effect of NCMS reimbursements relative to the income benchmark was observed in 2013. After controlling for differences in health needs across economic groups, the obtained values of CIs slightly changed. The relatively small difference between unadjusted and adjusted CIs may be due to the relative equilibrium distribution of health need variables across an economic gradient. Comparing the degree of inequity between 2008 and 2013 showed that need-adjusted CIs for both probability and amount of receiving reimbursements in 2013 were significantly lower than in 2008. This indicates that the pro-rich inequality of the NCMS decreased. The trends of inequality in benefits distribution observed in this study are consistent with the trends reported in previous studies on inequality in access to health care under the NCMS [13], suggesting that the inequity in inpatient utilization decreased along with the improved generosity of the NCMS.

The regression results indicate that health need factors were independently associated with the probability and amount of receiving reimbursements. Females were more likely to receive reimbursements partly due to hospitalized delivery. The lower amount of reimbursements among female enrollees may be associated with their lower medical consumption caused by the lower social and economic position of females in rural China. In addition, the results indicated that marital status was significantly associated with both the probability and the amount of received reimbursements, which was consistent with previous observations on health services utilization [40]. Our study did not find a significant correlation between the presence of commercial medical insurance and NCMS benefits. In China, very few individuals of the rural population purchased commercial medical insurance and no adverse selection was apparent [41].

This distribution of benefits is the most direct result of the NCMS. In terms of the “equal treatment for equal needs” concept, NCMS benefits should be decided by the health care needs alone, and not by any other factors such as the enrollees’ economic level; consequently, individuals with the same health care needs should receive the same benefits. However, the obtained findings indicate that the NCMS favors the rich. The underlying reasons for this inequity are mainly the inequity of the enrollees’ health care utilization and the policy design of the NCMS in the financing and benefit package.

The launch of a social medical insurance is considered a crucial step to reduce inequality regarding the access of health care. However, evidence of developing countries indicates that voluntary public insurance schemes, particularly those with high premiums and coinsurance rates, may have little impact for the improvement of the health services access for the poor [42, 43]. Theoretical models in the literature predicted that price elasticity of the demand for health services is likely to be higher for individuals at the lower end of the economy [44]. Therefore, the cost-sharing program with a higher coinsurance is unlikely to reduce the discrepancies of health care utilization between the rich and the poor or promote their equity. As suggested by Nyman’s theoretical research on the value of health insurance, health insurance can help low-income enrollees to surpass budgetary constraints to access health care; however, the degree of access depends heavily on the insurance payment system [4]. Numerous empirical studies reported a significant positive correlation between the access to health services and the socioeconomic status of enrollees. Studies in developed countries showed that high-income individuals are more than twice as likely to use medical services compared to those with low income [45]. Reports from developing countries also suggested that the rich access more health services [46]. It seems that NCMS contributed less to reducing income-related inequality in health service utilization and the relative increase in utilization of hospitalization was higher among rich enrollees [14, 15, 47].

The NCMS has been designed to collect flat-rate premiums and to provide an equal package of benefits for all enrollees, regardless of their ability to pay. Contributions toward the financing of social insurance may redistribute income [34]. However, with regard to flat-rate premiums, the economically disadvantaged have to pay the exact same premium as all others; therefore, health insurance contributions comprise a decreasing share of income with increasing ability to pay. Although previous publications of theories and empirical evidence suggested that income-based (a progressive premium method) social insurance schemes tended to have a pro-poor pattern in the distribution of net benefits [5‐7, 17], it is difficult to adopt a wage-based or income-based insurance program for rural residents without well-defined income or economic conditions. The NCMS covers such a large number of enrollees that it is also difficult to identify poor or vulnerable groups; therefore, a uniform standard of benefits package for all has been adopted.

Although the policy design of the system can achieve only limited effectiveness in achieving fairness, the NCMS is trying to improve and adjust. Compared to 2008, the NCMS has greatly improved the benefits package after the new medical reform. Correspondingly, in our results, the pro-rich inequality in benefits have decreased for 2013. The improvement of the NCMS is characterized by higher reimbursement rates, higher ceiling, expanded scope of reimbursement, simplified reimbursement method (for example, immediate reimbursement, i.e., patients can be compensated immediately after paying medical expenses), and lower deductibles in the primary medical institution. When the level of insurance package is high, medical expenditure is less dependent on the personal economic conditions [8]. In this case, for disadvantaged enrollees, enhanced access to health services increases the opportunities to receive benefits from the NCMS. Recent findings regarding access to health care showed that hospital admissions increased from 6.8% in 2008 to 8.4% in 2011 in rural areas in China, and inpatient reimbursement rates likewise increased from 32.9% in 2008 to 43.7% in 2011 [22].

The simplified reimbursement process may play an important role for the reduction of inequality in NCMS benefits. Specifically, the method of immediate reimbursement was used more often than before under the NCMS. For the disadvantaged economic population, later reimbursement, i.e., paying the full hospitalization cost first and then receiving reimbursements later, may cause current and future income difficulties for their family. The poor with need for treatment are therefore more likely to refuse treatment when expecting high health care costs. Immediate reimbursement, especially for the poor, can partially alleviate these current economic constraints and may therefore increase healthcare utilization [28].

Another shift in the NCMS policy was that the level of insurance benefits for lower-level medical facilities was greatly improved. For township health centers, deductibles were even removed, and reimbursement rates were raised to 90% or 100%. Numerous findings show that the poor were more likely to seek treatment in primary-level health-care facilities [48]. This design has lowered the threshold and economic burden for seeking treatment for enrollees with low economic background, thus benefitting the poor who need healthcare services from the medical insurance.

Furthermore, over time, the NCMS system has been progressively extended, increasing policy awareness confidence and gradually improving the status quo bias (i.e., the original habit of medical service utilization still adhered to in the case of insurance) among poor enrollees [49]. Thus, more health care needs may be released. In addition, other reforms within the new medical reform, such as the national essential medicines program, the establishment of a primary healthcare system, and a defined package of basic public health services for the population, may also play a facilitating role for the alleviation of the current inequality in NCMS benefits [50].

Finding the appropriate trade-off between policy feasibility and equity has become a challenge for the formulation of social health insurance funding and benefit packages in developing countries. However, this inequality can be gradually reduced through adjustments of the medical insurance scheme. For example, one promising adjustment could be the introduction of more favorable reimbursement methods, e.g., where the enrollees do not have to pay the full cost for each treatment in advance but only have to pay for the remaining portion after reimbursement. Currently, this approach is only implemented by China’s Urban Employee’s Basic Medical Insurance. This may further eliminate economic obstacles, especially for the poor. Another possible improvement might be to set distinct standards in premiums and benefit packages for vulnerable groups who are easily identified, such as the elderly and minimum living guarantee households.

To assure equitable access to health care and benefits distribution, insurance coverage alone is insufficient. The implementation of further relevant health policies can also be beneficial for reducing inequality in health insurance benefits. For instance, when the capacity and scope of primary health care facilities are improved, the disadvantaged may obtain added access to better medical services from primary health care facilities, which have lower deductibles and coinsurance rates.

This study has several limitations that require careful consideration. First, only four observable variables (age, sex, chronic conditions, and self-rated health) were used to adjust for health care needs; however, there may be other unobservable health statuses. Additionally, if the socioeconomic gradients disparities of the four health need variables are not homogenous, the socioeconomic disparities in health care needs may be weakened [51]. Second, we have been limited to repeat cross-sectional analyses. Longitudinal analysis of enrollees would allow increased scope for changes in economic-related inequity of benefits. Third, the quality of certain health-care services offered to identical patients may have varied between population groups; however, this study did not control for confounding variables related to the quality of health services as it was not feasible with the given large-scale household survey data. Further studies on insurance benefits could take the possible differences in the quality of health services received by enrollees into account across economic groups. Fourth, children below the age of 15 were excluded from this analysis due to their lack of available information regarding the required socio-economic characteristics and health status. Furthermore, outpatient reimbursement was not included as part of the NCMS benefits in this analysis, since the reimbursement information of outpatients is not available in our data neither for 2008 nor for 2013. These limitations could possibly lead to an underestimation of the benefits from NCMS for all enrollees and potentially lead to a bias of the results in terms of the distribution of benefits. However, this bias could be limited because the outpatient pooling fund of NCMS began in 2009 in the sample area and accounted for only a small part of the total pooling fund of the NCMS (less than 20%) in 2013. Despite these limitations, the results of this study provide the most realistic reflection of the benefits distribution and its inequities under the NCMS of western rural China.

This study forms a first step towards evaluating the magnitude of inequities in the NCMS benefits in western rural China during different periods (before and after the New Health Care Reform). The NCMS, which is characterized as a flat-rate premium and equitable benefit package, led to pro-rich horizontal inequity in the benefits distribution in both 2008 and 2013. A reduced pro-rich inequity in the NCMS benefits was observed toward 2013; however, it may be difficult to achieve the equity goals without a more comprehensive policy design for the NCMS that specifically and effectively targets disadvantaged enrollees. For developing countries, which attempt to extend social health insurance coverage using government subsidies, our results provide evidence for further reforms and bears meaningful policy implications.