Work package 4

China – The impact of social health insurance delivered through the New Cooperative Medical Scheme in rural areas


In China, there is extensive evidence of cost preventing people from using health services when they need them, and yet at the same time a large fraction of the population incurring health expenses that can be considered catastrophic. This is especially a problem in rural areas, which account for the bulk of China’s population, and where insurance all but collapsed following the dismantling of the old commune-based health system in the 1980s.

The intervention

The government’s response has been the introduction of a rural health insurance program known as the New Cooperative Medical Scheme (NCMS) which is being rolled out to the entire population between 2003 and 2008. The primary goals of NCMS are to prevent farmers from being impoverished by medical expenses and to improve their access to health care. Conspicuously absent from the stated goals are improvement of health and reduction of the large waste in health care delivery.

The Chinese government has allocated significant new resources to the NCMS, targeted towards the poor western and central regions. The government subsidized each farmer in western and central provinces 40 RMB (1 RMB = US$ 0.125), shared equally between the central and local governments, if the farmer pays an annual premium of 10 RMB or more to enrol in the NCMS. For the richer provinces in the East, the total subsidy is 20 RMB. The NCMS is targeted to reach 100% coverage by 2010.

The NCMS incorporates two important policy guidelines: voluntary enrolment and coverage of catastrophic illnesses. Apart from this, design of the program is left to the local governments. Consequently, there exist a variety of models, in terms of benefit package design, financing arrangements between the Central and local governments, provider payment methods and other broader purchasing relationships between the insurers and providers. A recent survey of 354 counties found that most existing NCMS models cover only inpatient services, some cover inpatient plus several selected outpatient services for major acute illnesses, and a few cover both inpatient and outpatient services. In the western and central regions, the most commonly found model combines a medical savings account and high-deductible catastrophic insurance for inpatient services. More recently, some benefit package has been extended to include specific health conditions, such as chronic conditions, maternal and child care, etc. The different models also involve varying levels of coinsurance rates, payment caps and deductibles.

On financing, there have been recent moves towards increased funding from the local government and reduction in coinsurance rates. These are, however, more likely to be found in counties with better and improving economic conditions. Other interventions include using medical assistance funds from the Department of Civil Affairs to support contributions of premiums and paying for deductibles and coinsurance for low income residents.

On the supply-side, while the dominant form of provider payment remains fee-for-service, a number of counties are experimenting with new ways of paying providers with goals to improve efficiency and quality such as village doctors, township health centers and county hospitals.

Very recently, some counties are experimenting with tying payment to treatment protocol for certain chronic conditions, e.g. hypertension and diabetes. 


The aim will not be to evaluate the basic NCMS, which has already been done, but rather to evaluate interesting designs that are intended to overcome the shortcomings of the common version of the scheme mentioned above. Two issues will be explored:

  1. How the design of the benefit package influences the program’s impact on utilization and OOP payments. The project will evaluate the standard scheme against designs that provide coverage against the cost of outpatient and inpatient care that kicks in immediately (first dollar coverage).
  2. How the impact of NCMS depends on the way in which providers are paid. The project will evaluate the standard NCMS against designs that pay providers in a way intended to discourage supplier-induced demand and provide incentives to improve efficiency.