Transformative State Capacity in Post-Collective China: The Introduction of the New Rural Cooperative Medical System in Two Counties of Western China, 2006–2008

Abstract

In 2002, the Chinese leadership announced a turnaround in national welfare policy: Local insurance at county level, called the New Rural Cooperative Medical System (NRCMS), was to cover all counties by 2010. This paper addresses the main characteristics of NRCMS as an example of ‘transformative state capacity’ in decentralised policy fields and its feature ‘responsiveness’ as a market-based means of its introduction.

Reviewing the modes of governance and comparing the introduction of local schemes based on two case studies of western China since 2006, this paper argues that the flexibility shown by local administrators in considering structural and procedural adjustments is the result not only of central directives but also of local initiatives. Forms of locally embedded responsiveness to the needs and perceptions of health care recipients are crucial in enhancing the accountability and responsiveness of local cadres. These new modes of ‘responsiveness’ or responsive regulation are important in understanding and conceptualising the transformative state capacity. Responsive settings using centrally defined local feedback loops are different from hierarchical control and the formal institutionalised representation of the interests of the local population, and are a rough but effective means of enhancing both flexibility and the efficiency of control and financing by the central state. These feedback loops, which are based on voluntary enrolment and on central state subsidies made dependent on contributions received from participants and local government, are complementary forms of governance at grassroots level.

1. Introduction

In 2002, the Chinese leadership proclaimed a turnaround in national health policy: Local voluntary medical schemes at the county level, presented as the New Rural Cooperative Medical System (NRCMS, xinxing nongcun hezuo yiliao zhidu) were to be scheduled to cover all counties by 2008.² At the end of 2007, more than 85 per cent of all counties were covered by an NRCMS plan, with a high participation rate, also 85 per cent, reported by the national media.³ Since then there have been impressive and rapid improvements in the contribution and benefit structure. This paper will address the main characteristics of transformative state capacity in the context of rural health governance⁴ since 2002. The existing literature in the social sciences focuses on extractive state capacity and modes of governance with rigid implementation lines (policy dimension) of party cadres (political dimension) in very centralised and hierarchical institutionalised forms of coercion (polity dimension).⁵ How are transformative changes governed within a Communist regime? Political campaigns at different political levels, a strict time frame with a target for coverage, and transfer payments from the central government provide evidence that this is not a bottom-up implementation. On the other hand, patients must join the schemes voluntarily and renew their contract with the local NRCMS bureau annually. The improvements, together with remarkable flexibility and responsiveness to the needs of the rural population at the local level within a loose regulatory framework, with wide decision scope for the cadres but without significant democratic participation of the farmers, challenges the existing literature on extractive state capacity and governance in China.⁶

We will argue that in traditionally highly decentralised policy fields like rural health care, state action is not only characterised by recentralisation, political campaigns or control. Instead, we find local embedded market-based mechanisms as they exist in Western capitalist societies. Responsive regulation is an institutional feature for enhancing accountability of local cadres. With increasing interdependence between state and society in multi-level modes of governance, it becomes a source of transformative state capacity.⁷

Our initial proposition, which is supported by two case studies, is that responsiveness is a distinct feature of the transformative state capacity in rural health governance. Responsiveness aims at achieving higher accountability by designing regulatory processes, which stimulate and respond to the regulatory capacities of already existing stakeholders and institutions inside and outside the political system. It entails control by enhancing the behaviour of stakeholders as a signal in local feedback loops, instead of setting up new mechanisms from above (audits) or below (elections). Guaranteeing opting in and opting out, it attempts to keep regulatory interventions to a minimum necessary to secure the designed outcomes. Responsive regulation is a market-based design aiming to intervene primarily through a more stringent integration of the perception and behaviour of existing acting groups and stakeholders, rather than a coercive integration of population groups into policy-making within supervisory boards.⁸ Its outcome can be obtained when institutions and procedures are redesigned in such a way that defined institutions or stakeholders respond appropriately to the expectations of targeted individuals.⁹

First, we will outline the policy before 2002 and analyse the existing and politically designed decision space at the local level. Using a case study in Xinjiang and additional information from a selected county in Yunnan, both provinces of western China, we will illustrate the mechanisms of this introduction and the characteristics of a step-by-step approach after 2002.¹⁰ In the final section we will summarise the characteristics of transformative state capacity.

2. Policy Changes: From CMS to NRCMS

2.1. The Central Role of NRCMS in Rural Areas

In the 1980s different forms of health care provision were possible, and even in the year 2001 CMS was a ‘not mandatory’ (bushi qiangzhi xing de) policy for local governments.¹¹

Several trials in impoverished counties in the 1990s failed.¹² Before 2002, the central state—even in cooperation projects with international donors like the World Health Organisation—did not consider it desirable to invest in the Cooperative Medical System (CMS; hezuo yiliao zhidu).¹³ After illusionist expectations that even in marginalised western regions ‘the main contributions are individual premiums, the collective will give aid and the government will give appropriate support’,¹⁴ from 2002 on these local schemes in the less well-off regions, especially in western and central China, were managed at county level and subsidised directly from the central state budget for the first time in the history of the People’s Republic, becoming the main and only form of primary health care and protection in rural areas.¹⁵ The CMS was renamed the ‘New Rural Cooperative Medical System’ (NRCMS) and defined as follows:

The New Rural Medical System provides mutual help and benefits by mainly focusing on curing serious diseases. It is organised, led and supported by the government with the voluntary participation of the farmers. The system is jointly financed by individuals, collectives and the government.¹⁶

The two main objectives are to reduce illness-induced poverty and to reimburse the cost of major illnesses.

2.2. Objective One: Reducing Illness-induced Poverty

While it is true to say that before 2002 the goals of the Ministry of Health (MoH) to reform and develop health care were promulgated together with the Ministry of Finance (MoF), the Ministry of Agriculture (MoA) and National Commissions and sometimes even the Communist Party, issued as joint declarations, we argue that the outlined policy tools were indeed self-conflicting and indicated a horizontal fragmentation of central government at that time. The problem of the ‘medical poverty trap’ (the ‘return into poverty due to [the financial costs of] illness’, yin bing fan pin) was already being addressed by political leaders in 1997.¹⁷ However, the introduction of the CMS (as a new form of cost for farmers) conflicted with the 1990s goal of decreasing the farmers’ burden (jianqing nongmin fudan).¹⁸ Consequently, the MoF did not actively support the introduction of CMS, but underlined the illegitimacy of collecting CMS premiums.¹⁹ There was no consensus on whether the MoH or the Ministry of Civil Affairs (MoCA) should administer the management of the schemes.²⁰ CMS local pilot studies in the 1990s did not result in any further central enforcement or regulation, thanks to the deadlock at central level.

This situation changed after 2002 and before the outbreak of SARS in China, when several ministries agreed that the MoH was to take the lead by introducing the system. The dissent between ministries was overcome. A consensus was reached that the premiums of NRCMS ‘should not be seen as an increase of the farmers’ financial burden’ (buneng shiwei zengjia nongmin fudan),²¹ and that NRCMS should be regarded as in line with the anti-poverty efforts of the MoF.²² From 2002 onwards, it became one of the key measures against poverty in rural areas.

2.3. Objective Two: Reimbursing the Cost of Serious Illnesses

While before 2002 the scope of the health services was left open to adaptation to prevailing local conditions and financial resources, after 2002 NRCMS has consequently focused on risk-sharing, especially for the main diseases, which demanded hospitalisation and thereby often caused impoverishment.²³ Three alternative models were made available to the county level during the introductory phase of NRCMS, which commenced in 2004. Changes between these alternatives are at the discretion of the counties concerned: (I) separate accounts for in- and outpatient treatment; (2) one single account, but with different reimbursement ratios for in- and outpatient treatment; (3) coverage only for inpatient treatment.²⁴

3. Procedural Characteristics

3.1. Forced Introduction in Selected Pilot Counties

Neither the era of the Cultural Revolution nor the period from de-collectivisation until 2002 can be defined as experimentation policy-making intended by the state.²⁵ During the time of the Cultural Revolution, CMS could be organised at the level of the communes or the brigades, or as a joint scheme of selected brigades. After the disastrous experiments of the Great Leap Forward, national policy was left open to adaptation to local conditions. Model communes demonstrated that even in poor and remote settings a sustainable primary health care provision was possible in highly decentralised patterns. The communicated ‘success’ of these experimentation points with campaigns of revolutionary ‘best practice’ has indicated that further intervention, unification and subsidies are not required.²⁶

This cellular pattern of rural health care was perpetuated in the 1990s. Affirming the status quo of decentralised finance and organisation and non-action at the central level, it was already obvious in the 1980s that these schemes, operating on a market base, were not sustainable in poor counties.²⁷ Pilot counties (shidian xian) in selected areas mainly attracted funding from international organisations or Chinese scientific research agencies in a situation where these experiences were not unified at central level, whereas the national policy was left unchanged for 20 years.

From the year 2002 onward, the central state has finally declared that without getting involved in the grassroots constellation, change is impossible.²⁸ The goals of NRCMS are defined, and there is a national policy for local schemes at county level. The aim is not to encourage local innovation in selected and separated experimental points giving policy input in central-local interaction, as in the case of the policy of financial regulation,²⁹ but to introduce schemes in a clear time frame. This approach is characterised by the creation of pilot studies, and central transfer payments to the county go along with unification of the organisation model within the county government.

Pilot studies are an integrated part of the introduction of NRCMS. In 2003, provinces were to have approved at least two to three pilot counties (shidian xian).³⁰ Those provinces that failed the evaluation were to be restricted in setting up new pilot counties. Criteria for selecting pilot counties are:

1. willingness of local political leaders;

2. participation rate of the farmers;

3. guarantee of local subsidies;

4. high management skills of health care administrators.³¹

There are several advantages to this approach:

1. Since the county administration has to apply and is selected from among other counties,³² it is likely to be more motivated. State subsidies can be used more efficiently because local commitment is a prerequisite for introducing the health scheme. The effectiveness of state subsidies is thereby enhanced.

2. Resources and audits can be concentrated on the most needy or the most suitable cases first.

3. New pilot studies in other counties can learn from the experience in other pilots.

3.2. Time Frame for Implementation

The shift from a non-compulsory, decentralised policy to a centrally guided policy with mandatory implementation can be clearly seen in the establishment of a time frame for national coverage. According to the Chinese government in 2002, all counties were to run a scheme in the year 2010.³³ A detailed schedule for the years 2006 to 2010 was not published until 2006: in 2006, 40 per cent of all counties were to have set up pilots, rising to 60 per cent in 2007; by 2008, every county was basically (jibenshang) to have set up an NRCMS scheme, with full coverage obtained by 2010.³⁴ Central government accelerated the introduction in 2008 and is currently trying to reach full national coverage this year, and the year 2010 is no longer mentioned.³⁵ Even though NRCMS had to be introduced by all county administrations by 2008 and the coverage ratio was defined by the central state, there are restrictions to a top-down approach: individual participation remains voluntary, albeit with a high participation rate desired as a necessary precondition for economic and political success.

3.3. Management Only at County Level

Until 2002, CMS could be operated at all levels from village to county. These schemes were not subsidised from the central coffers. After 2002, all funds were to be managed at county level, with the county being the lowest level of the state health administration. Models and reimbursement quotas can differ, but central government suggests that the differences between neighbouring schemes should not be too huge. Unification with health schemes at city or even provincial level has not yet been foreseen. Since the county is responsible for financial shortfalls (shou dingzhi),³⁶ there is no liability on the part of central government and the county is therefore forced to strike a balance between contributions and expenditure.

3.4. The Voluntary Participation of the Farmers

Because of regional differences with respect to economic, social and cultural conditions, there are inherent limits to the enforcement of policies such as NRCMS in a top-down approach. International public health advisors advocate a compulsory system.³⁷ But they have ignored the different opinions held by the ministries concerned. In the 1980s it was already apparent that forced enrolment had increased mistrust in the managing cadres, both during the Cultural Revolution and in later periods. Therefore, enrolment should be voluntary.³⁸ The negative historical legacy determined the path dependency of health insurance development: obviously, policy-makers were conscious of these constraints and stressed the necessity of confirming voluntary participation. The CCP leadership therefore affirmed, as early as 1997, that farmers could enrol or withdraw from the schemes and pay their premiums voluntarily on an annual basis (‘nongmin ziyuan jiaona de hezuo yiliao fei’;³⁹ ‘nongmin ziyuan canjia’).⁴⁰ By making enrolment voluntary, the decision of affordability was shifted from the state on to the farmers. Voluntary enrolment or withdrawal also signals their degree of satisfaction with and trust in the scheme.

The voluntary nature of the scheme presents both advantages and disadvantages. The advantages of this approach include the following:

1. Participants can decide whether the system meets their specific needs. Their willingness to join and pay for NRCMS is vital for sustaining the system.

2. Administrators must therefore convince participants of the effectiveness and efficiency of the system and build motivation through high quality service.

3. High turnover rates signal deficiencies in the system to the administrators, and enhance flexibility/changes in the service package/reimbursement.

4. Voluntary participation enhances the sense of ‘ownership’ and generally induces participants to make a more judicious use of the system.

On the other hand, the disadvantages encompass the following points:

1. Adverse selection occurs, whereby it is mainly households at risk that enrol (with cases of serious or chronic illnesses, and old people).

2. There is a high cost of mobilisation campaigns for the annual enrolment in autumn.

3. The problem of anticipating participation makes actuarial budgeting difficult and might require rapid changes in contribution levels and benefit policies.

One key pillar of CMS/NRCMS is voluntary enrolment, including the right to opt out of the system annually. Unlike elections, it does not institutionalise the interest of groups in the governmental bodies, but guarantees ways out of/into the system.

3.5. Earmarked Central Transfers as Matching Funds

In comparison to policies implemented before 2002, the most important innovation is the funding of CMS. While before 2003 the financial contribution to health insurance funds was left to the discretion of the collectives, to local government or to market forces, now central government is ready to contribute to the funding of NRCMS. After the promulgation of this policy change in 2002, in 2005 the central budget began to subsidise NRCMS pilot schemes in all non-urban areas in western and central China.⁴¹

In 2006, central government increased its contribution from 10 to 20 yuan per enrolled participant for all western and central regions with a rural population of more than 70 per cent and for the pilot counties in Liaoning, Jiangsu, Zhejiang, Fujian, Shandong and Guangdong.⁴²

The state transfers have earmarked matching funds. They are allocated in accordance with the actual number of individual participants and are earmarked for inpatient treatment. The receipt of subsidies is not subject to bargaining between higher and lower levels, but depends on enrolment into the schemes by the farmers.⁴³ We argue that this approach is intended to enhance the responsiveness of patients, local state agencies and health care providers. The premiums paid by participants and the financial support by local government determine the focus and scope of the financial transfer from the centre in two ways.

First, these earmarked subsidies are related to the actual enrolment of the rural population and the payment of a premium of not less than 10 yuan by each participant. The behaviour and perception of patients now determine the financial support by the upper levels of the political system. Central government contributions in a given year were originally based on enrolment statistics for the previous year and on expected enrolments for the next year, but will now be adjusted to current figures of the current year.⁴⁴ The role of the rural population is thus strengthened.

Second, and by the same logic of responsiveness, the financial commitment of county and provincial governments is a precondition for the co-payment from central government: their respective contribution to the insurance funds per enrolled participant was not to be less than 10 yuan. Beginning in 2007, the contribution of the county and provincial governments was to be increased by 5 yuan per annum in 2006 and 2007.⁴⁵

This tendency of increasing central subsidies and the local matching funds is thereby accelerated. It is a synchronised introduction: subsidies are increased on condition that all other payers (local government and participants) increase their funding accordingly. The total contribution was to be increased to 100 yuan according to a quota of 4:4:2 in 2008. The subsidies from central government are set to increase to 40 yuan, when the contribution of local government is not less than 40 yuan and the households pay 20 yuan per member.⁴⁶ There is evidence that there are exemptions for poor (and politically sensitive) regions like the southern part of Xinjiang.⁴⁷

Since central government only transfers monies when the farmers have actually enrolled and the local governments have paid their contribution, the role of the participants is again strengthened. Administrators are asked to perform well by ‘responding’ to the needs and perceptions of the participants as measured by the participation ratio. Consequently, central subsidies did not induce a stronger administrative involvement in the daily administration: The reimbursement ratio should not differ widely within one province, but levels are not specified.⁴⁸ The monitoring of the schemes is delegated to the provincial level and below, within generic guidelines. Further, controls and evaluations (jiancha pinggu) do not include any bonus payments for the administrative or medical staff.⁴⁹

The main characteristic dynamic at the central level is the augmentation of central subsidies in 2005 and 2007. Under this annual ‘regime of responsiveness’, the critical question is how to guarantee an efficient usage of the increasing central subsidies in the implementation process. The main focus of the process is the increase not of innovative policy-making at central level but of local adaptability and efficiency of central resources. The central state is trying to find its way back into forms of financing and administering rural health care.

Coverage is the effect of top-down implementation, but participation is the result of responsive regulation. The system is flexible because the time frames are defined for a single year. At the end of the year, every peasant has to enrol again. This time frame and the strategic higher role of the recipients by voluntary enrolment enhance the flexibility and adaptation to the perceptions and needs of the users of NRCMS. The voluntary decision of the rural population is the precondition for responsive regulation, because it institutionalises the perception of the farmers as critical stakeholders.

4. The Introduction of NRCMS in Two Counties of Western China

The following section discusses similarities in the introduction of NRCMS in two multi-ethnic and poor counties in northwest Xinjiang and southwest Yunnan provinces.

Xinyuan County in Xinjiang Uyghur Autonomous Region consists of 10 townships (xiang) and one town (zhen), with a total population of 300,000 in 2006 and 305,800 in 2007 (an increase of 2 per cent). Contrasting with this increasing overall trend is a decreasing rural population that fell from 187,400 in 2006 to 174,900 in 2007 (a drop of 6.7 per cent).⁵⁰ Of the total population, 44 per cent are ethnic Kazaks, 39 per cent Han and 9 per cent Uyghurs. The majority of the Kazaks live in semi-nomadic conditions during the period of the summer pasture in the mountains, turningsome 1,090,000 cattle and sheep out to grass.⁵¹

Consisting of 16 township/villages and with a population of 505,000, Xundian County in Yunnan Province is much more densely populated. It is dominated by Han Chinese, with minorities reaching a quota of 21 per cent.

4.1. Introductory Phase: Legitimation as State Pilot

In 2005 the Health Bureau of the Xinjiang UAR approved the request of Xinyuan County to be selected as one of the first state pilot counties (guojiaji shidian xian) for testing NRCMS. The central government was not involved in this decision.⁵²

The pilot began in January 2006 with 161,164 paid-up participants from among 187,400 registered rural inhabitants, representing a coverage ratio of 86 per cent. By 2007, participation had only marginally increased to 162,723 (I per cent in absolute numbers). Nevertheless, due to a drop in the rural population to 174,971, the coverage ratio had increased to 93 per cent (a rise of 7 percentage points).⁵³ Participation between townships was rather uneven. In one of the townships visited, the reported participation was only 77 per cent.⁵⁴ Our customer satisfaction survey, carried out among Kazak herders in August 2006 (i.e. the first year of NRCMS introduction), showed that only 60 per cent found it worthwhile, 30 per cent were not satisfied, and 10 per cent had no opinion. The three main reasons for dissatisfaction were the inadequate NRCMS coverage for outpatient treatment, the low reimbursement rates for hospital costs, and procedural difficulties.⁵⁵

In the much poorer region of Xundian County, NRCMS began in 2007, probably because of its poorer financial situation compared to the counties like Xinyuan County. Figures are therefore less readily available.

4.2. Main Administrative Responsibility at County Level

The implementation of NRCMS has been delegated to the county level. The county administration dominates the management and provides subsidies in cash and kind.

(a) According to the local regulations, the Xinyuan County CMS Administrative Committee is responsible,⁵⁶ but it has yet to play a significant role. All positions in the health bureau, including those responsible for the CMS operations, are established and financed by the same administrative level (fenji guanli) in Xinyuan County, which also covers all operational costs through its offices for personnel allocation (renshiju) and for finances (caizhengju), in form of a flat rate of 0.5 yuan per rural inhabitant.⁵⁷ This is a departure from the situation before 2002, when 5 per cent of the CMS budget could be used for paying administrative overheads.⁵⁸ All contributions to NRCMS are now exclusively dedicated to the reimbursement of health service costs.

In Xundian County, CMS administration is part of the county administration: The head of the Administrative Committee is the head of Xundian County. The director of the Xundian County Health Bureau is simultaneously his deputy and director of the CMS bureau.

In addition, 17 of 20 posts within the Administrative Committee are held by different departments of the county administration, while only two are held by the township head and one representative of the farmers.⁵⁹

The county administration in Xinyuan County recruits the NRCMS management bureau. In 2007, the CMS office in Xinyuan County employed 10 staff with a payroll of 120,000 yuan, all paid by the county. The director of the CMS office, a Han Chinese, was a former union representative in the County Hospital of Chinese Medicine. Six staff members were responsible for claims processing and control.⁶⁰ Medical costs incurred at village and township level are processed by the CMS offices in the health centres of the townships. The county NRCMS office only processes claims for medical expenses incurred at the county level and above. It also processes the monthly accounts of all other CMS offices in the county.⁶¹ The operational NRCMS programme and budget is decided by the county government, which evaluates the performance of the CMS bureau head, who in turn oversees the operations of the CMS offices in the health centres of the townships and towns.

(b) The county government of Xinyuan guarantees the financial viability of the insurance scheme and is expected to make up for any shortfall. To reduce this liability, a risk fund has been established by deducting 2 yuan per annum for each participant from the combined state contributions until 10 per cent of total annual contributions has been reached. Xundian County sets aside 5 per cent of annual contributions until 10 per cent of total annual contributions has been reached.⁶²

These high similarities in administration signify the clear significance of the central implementation guidelines at the local level. The next section will show the modifications of material health politics.

4.3. Financial Contributions to NRCMS

In these counties, levels of funding and forms of internal distribution are different.

Xinyuan County has fixed the NRCMS annual contribution due from each participant at 25 yuan, based on the average income of the rural population. This is close to the ceiling of 30 yuan established by the government of the province. This amount is supplemented by 10 yuan from the county government, 10 yuan from Xinjiang UAR, and 20 yuan from central government, for a total of 65 yuan. The public share (central state, province and county combined) thus represents 61.5 per cent. New enrolments and premium collections are due every year between August and November. Thereafter, no enrolments are possible.⁶³

Xinyuan County administration has chosen model no. I suggested by the central government (see 2.3) of separate accounts for in- and outpatient treatment. The participants’ contributions go into a family account for outpatient treatment. The combined state contributions (less than 2 yuan per participant for the risk fund—see 4.2b) go into a pool account, out of which expenses for hospitalisation and ‘major illnesses’ (not further qualified) are reimbursed. They may not be used for the reimbursement of outpatient treatment. In applying the guidelines of the central state agencies, NRCMS in Xinyuan County is therefore heavily biased in favour of inpatient treatment. The contributions are pooled to provide coverage for all household members who are registered in the county.

When NRCMS was introduced in Xundian County a year later, it had set an annual contribution of each participant at only 10 yuan, but the contribution of the province is double that of Xinyuan County (20 yuan). The reason could be that the county does not or cannot contribute anything, but 20 yuan as a contribution of local government is a precondition of the flow of 20 yuan from central government. The earmarking of central government is a very strict condition, matched by co-payments at local levels. Xundian County has chosen model I, with separate accounts for in- and outpatient treatment. However, the contributions from participants are pooled with the state contributions, and the combined amounts (after deducting 5 per cent for the risk fund) are split 70:30 between an inpatient and an outpatient fund, and 25 per cent of the medical costs of all outpatient treatment are reimbursed, up to an annual maximum of 100 yuan.⁶⁴

For both counties we see remarkable flexibility through a constant funding situation, with a further increase in subsidies from central government to 40 yuan in 2009, requiring an equivalent funding from local government.

4.4. The Role of Local Cadres

No targets appear to have been set by central government with regard to the percentage of eligible participants to be covered by NRCMS. There is, however, evidence that such targets for mobilisation are set locally,⁶⁵ and there are forms of targets in both counties.

For NRCMS in Xinyuan County this level has been set at 80 per cent.⁶⁶ Responsibility to achieve and maintain at least this level does not lie with the administrators in the CMS bureau. It is also not part of their performance targets, and they receive no premiums if the 80 per cent mark has been reached or surpassed.⁶⁷ At the same time they are evaluated on the effective management of the scheme, and positive evaluation is only possible with high participation rates. The township people’s governments (xiang renmin zhengfu) and villagers’ committees (cunmin weiyuanhui) are responsible for the mobilisation of the population. The party secretary who was responsible for health work in one township explained, that he could ‘mobilise a team of 50 community workers and 80 village cadres’ to induce farmers and herders to join the CMS and to ensure that ‘the enrolment percentage established by the CCP was attained’.⁶⁸

Xundian County set explicit targets: ‘Through extensive propaganda it should be ensured that more than 95 per cent of all farmers benefit from the New Rural Cooperative Medical System.’⁶⁹ Reaching at least 80 per cent or 95 per cent of coverage, however, appears to be only a ‘soft’ performance goal for the village and township cadres without any integration into the cadre responsibility system. The idea behind this seems to be that the voluntary character of NRCMS is so important that it should not be jeopardised by any means.

4.5. The Voluntary Nature of NRCMS, and Fluctuation of Membership

Requests to enrol may not be refused for reasons such as age or health condition in any county—this policy is applied in both Xinjiang⁷⁰ and Yunnan.⁷¹

One inbuilt stabilising factor to discourage adverse selection resulting from frequent opting in and out of the scheme is the requirement that persons belonging to the same household must join NRCMS together. All persons with household registration (hukou) or who have actually resided in Xinyuan County for at least five years, and who are not covered by another public health insurance scheme, are eligible to join. The NRCMS accounts are kept on a household basis.⁷²

The two schemes show evidence that the management is similar because of a top-down unification. Management depends mainly on the health ‘line’ (tiao). The responsible agent is the county health bureau and its established NRCMS bureau. Contributions flow into the risk fund of both counties, until 10 per cent of the total annual contributions is reached. Besides the focus on serious diseases, both counties are in line with state regulations that the enrolment of the population is voluntary, by household. Administration costs are not deducted from contributions and are paid by the county government. See Table 1.

Table 1.

Synopsis of Funding and Reimbursement in Xinyuan and Xundian County, 2006–9

Item	Xinyuan County
	January– May 2006	June– December 2006	January– April 2007	May 2007	January 2008– December 2008 (2009)	2007	2008 (2009)
1. Contributions
1.1. Participant	¥25	same	same	same	¥25	¥10	(¥20)
1.2. Village	—	same	same	same	—	—	—
1.3. County	¥10	same	same	same	¥20	—	(—)
1.4. Province	¥10	same	same	same	(¥40)	¥20	(¥40)
1.5. Central state	¥20	same	same	same	(¥40)	¥20	(¥40)
1.6. Total	¥65	same	same	same	¥105	¥50	(¥100)
2. Reimbursement
2.1. Franchise  (each hospitalisation)
2.1.1. Township	¥100	¥100	¥100	¥80	¥80	(Very poor township—¥80) ¥100	same
2.1.2. County	¥300	¥300	¥300	¥200	¥200	¥200	same
2.1.3. City	n/a	n/a	n/a	n/a	n/a	¥500	same
2.1.4. Province	¥600	¥500	¥500	¥500	¥500	¥600	same
2.2. R-Rates (hospital)
2.2.1. Township	50%	60%	60%	70%	70%	60%	70%
2.2.2. County	40%	50%	50%	55%	55%	50%	60%
2.2.3. City	n/a	n/a	n/a	n/a	n/a	40%	50%
2.2.4. Province	30%	40%	40%	40%	(45%)	35%	50%
2.3. R-Rates (outpatient)	100%	100%	100%	100%		25%	25%
2.4. Ceilings (hospital  p.a.—all levels)	¥1,2000	¥1,2000	¥1,2000	¥1,2000	(¥1,5000)	¥5,000	¥15,000
2.5. Ceilings (outpatient)	Available balance in family account, based on individual contributions as shown below:					No separate family account
	¥15	¥15	¥25	¥25	¥25	¥100 p.a.	¥200 p.a.

Sources: Xinyuan xian, Xieyishu; Xunzhengfa 2007-4

Both counties have similar funding arrangements; changes in contributions are much less flexible than those for reimbursements. A further increase of the central funding to 40 yuan will also increase the contributions of the individuals to 20 yuan.⁷³ There are two tendencies in both counties:

1. While reimbursement ceilings are higher at the higher level institutions, the applicable franchise increases and the reimbursement rates decrease the higher the administrative level of the hospital. This should encourage patients to seek service from the nearest possible health care institution, which is also the most inexpensive health service provider. It also discourages patients from insisting that their treating physicians refer them to higher level institutions for fear that the services of the township health centres are not adequate.

2. In both schemes the state contributions are rather high for 2007: in Xinyuan County 69 per cent and in Xundian County more than 80 per cent are subsidies from state agencies.

Overall, and exemplified here for Xinyuan County, the benefits provided by NRCMS must still be considered as very modest, bearing in mind that the average annual income per capita of the population selected for the case study (Kazak pastoral households) was 2,867 yuan in 2006 (an increase of 32.3 per cent from 2,167 yuan in 2003), and average out-of-pocket expenses for each case of hospitalisation were 3,826 yuan.⁷⁴ This huge difference is explained by the fact that many surgical interventions are not covered by the scheme, and many farmers were unfamiliar with these exceptions.⁷⁵ More importantly, even a further increase in the current 61.5 per cent of state co-financing of the scheme will not bring a dramatic change until rural subsistence households are generating considerably more cash income. This is compounded by the absence of state contributions to outpatient services, placing an additional burden on the shoulders of the rural population, who are likely to fall into the poverty trap whether having to pay for in- or outpatient treatment. At least in Xinyuan County, medical financial assistance has not yet been implemented.

4.6. NRCMS Improvements between 2006 and 2009

From the inception of NRCMS to the beginning of 2009, reimbursement levels in Xinyuan County have increased thrice and in Xundian twice. The deductible amount (franchise) in the case of hospitalisation has decreased and the reimbursement ceiling increased, while contribution rates remained unchanged until 2007 (see Table 1). In response to the dissatisfaction expressed by participants, a number of benefit improvements were made while keeping contribution levels constant:

1. It should be noted that since 2007 migrant workers have been covered by the schemes in Xinyuan County. This acknowledges the increasing mobility of the rural workforce and provides some relief to migrant workers who, through often precarious contractual situations, remain uninsured at their place of work.⁷⁶ This feature was implemented in Xundian at a very early stage.

2. In both cases the conditions for reimbursement of outpatient treatment have improved. In Xinyuan County, the disposition of the family account for outpatient services has been increased. Originally, of the premium of 25 yuan only 15 yuan was credited to the family account (jiating zhanghu) and could be used for outpatient treatment, the difference being paid into the social risk fund for major illnesses (dabing tongchou jijin). Now, the total participant contribution of 25 yuan has become available for outpatient treatment, the cost of which is deducted from the family account directly. Each family member may draw on the total amount available in the family account, and unspent balances are carried forward to the following year.⁷⁷ The family account strengthens intra-family risk-sharing, because private premiums do not flow into the social risk fund. Nevertheless, the family account for outpatient treatment is quickly exhausted when looking at the price list for medication or services in health care facilities of Xinyuan County.⁷⁸

In both cases, the overwhelming numbers of farmers are seeking outpatient treatment. In Xundian County, outpatient treatment is reimbursed from both individual and state contributions. This improves the perception of participants that the state participates in the cost of all treatment, although reimbursement ceilings for outpatient treatment are low.

In 2007, Xinyuan County discussed a similar change to a joint in- and outpatient account.⁷⁹ In Xundian County, from 2009 the ceiling for outpatient treatment is set to increase dramatically from 5,000 to 15,000 yuan.⁸⁰

One major system pitfall remains, which needs further consideration. The initial aim of introducing NRCMS was to prevent farmers from falling into the poverty trap as a consequence of illness. Both schemes show that the financial support from the government and the conditions of reimbursement are still insufficient to provide substantial relief. Additional funding is needed to establish higher reimbursement levels and bring about the desired outcome, i.e. the improvement of the economic and social status of the rural poor. Official reports mention that poor households do not join NRCMS because they cannot afford the premiums, and no local subsidies are available.⁸¹ In accordance with Art. 18.I of the NRCMS implementation guidelines for Xinyuan County, the contributions for the abjectly poor village households are advanced from the cooperative medical finance assistance (hezno yiliao buzhu). This fund still awaits implementation.

5. Conclusion

In the past, rural areas have seen excessive and extreme forms of governance. Political campaigns for NRCMS in the Maoist era either were not sustainable or led to a total retreat of state action. After privatisation in the post-collectivist society, local insurance schemes were set up in industrialised or wealthy rural areas, mainly in east China.⁸² Former breakdowns, failures of top-down delegation and insufficient sustainability of international projects in the 1990s in a highly decentralised setting are a ‘negative’ heritage for rural health care governance in the new century.⁸³

One of the most astonishing features of NRCMS is the wide acceptance of the local scheme and the flexibility in benefit improvements. Local administrators have responded to the needs of enrolled participants. This can be seen from changes introduced or considered in several areas since the inception of the scheme in 2006 in Xinyuan County, including:

1. improvements in reimbursement levels of outpatient treatment;

2. freedom of choice of health services, including facilities outside the county household registration.

We argue that the case of NRCMS, although and because its introduction has high political priority, is a vivid example of the astonishing decision space⁸⁴ accorded to local actors. Therefore, the implementation of NRCMS is neither a strict top-down implementation nor a form of experimentation policy prevailing in economic policy fields leading to policy innovation.⁸⁵

This form of governance combines political pressure and top-down implementation with an institutionalised high decision space for local governments. NRCMS is of high political priority and is introduced according to a central schedule. Local cadres cannot resist implementation. But instead of direct control by the central state, we observed a form of centrally installed mechanism that is characterised by ‘responsiveness’. Responsiveness describes a form of governance integrating local feedback loops mainly between the local administrators and the rural population. It is not the evaluation of central government that is crucial here, but the perception and behaviour of the local population leading to allowing or stopping payment of central subsidies to the county administration. It enhances forms of local feedback processes between different levels of state agency, health providers and patients:

1. Voluntary and annual enrolment enhances flexibility in the insurance conditions.

2. Central transfers depend on the enrolment of the rural population.

3. The county administration is responsible for any shortfalls. Furthermore, the county has in principle agreed to cover any financial shortfalls of NRCMS, and additional funds are being made available from the central health budget to ensure the long-term sustainability of the scheme.⁸⁶ These annual local feedback loops between local administrators and patients should ensure correct and efficient state subsidies.

These hybrid forms of governance need further differentiation from other modes of governance prevailing in the discussion of rural governance.

1. While the introduction of NRCMS is a high priority policy and coverage is defined by the central state, the enrolment rate (participation) has not been made a target in the cadre responsibility system or bonus payments of the administrators of the scheme, at least in these two counties. This introduction therefore differs from other policy fields where the importance of the cadre responsibility system is emphasised. Traditional forms of rural governance, such as campaigns and enforcement by compliance, would not have been as effective, because participation in NRCMS is voluntary and must be renewed on a yearly basis.

2. The document analysis and field study in Xinyuan County confirm that central government provides financial subsidies but without enforcing a recentralisation of the system beyond the administration of central funding. Increased financial subsidies do not correspond to higher forms of direct intrusion of the central state bureaucracy.⁸⁷ The introduction of NRCMS is promulgated in the form of recommendations.

3. Until now, central decisions have only regulated the financial management and usage of the central subsidies at local level.⁸⁸

4. Central government has not defined any rights of the farmers, which are effective tools for rightful resistance in rural China, where extractive state capacity is high.⁸⁹ It is an analytically important aspect of these case studies in Xinyuan County and Xundian County that the impact of voluntary participation and the need for consensus within the administration play a more important role than the institutional framework of the Administration Committees of the New CMS in Rural and Pastoral Areas (xinxing nongmuqu hezuo yiliao guanli weiyuanhui) or Surveillance Committees (jiandu weiyuanhui) and their sub-organisations at township and village level, where participants can in principle voice their rights and concerns through their representatives in the village CMS committees,⁹⁰ but where the actual role of these representatives has so far been insignificant because these institutions do not work properly.⁹¹ In Xundian County, the farmer representative is only one of 20 CMS Administrative Committee members. Instead of a malfunctioning and lack of trust in the surveillance committees, we observe a system of checks and balances that works not within governmental bodies and organisations, but within multi-level linkages of different agents responding to the needs of the participants. Where institutionalisation of the interests (‘voice’) of the population is weak and audits are ineffective or corrupted, responsive settings are more minimal and strengthen the voluntary and annual enrolment (‘exit’) as an important signal in regulative local feedback loops replacing ‘central-local interaction’.⁹²

5. Instead of massive forms of coercion in traditional campaigns or rigid implementation policy fields like birth control, the central state limits the coercion on the local cadres and provides slowly increasing subsidies paid according to the responsive institutional setting. Thanks to a one-year time limit on insurance contracts, administrators have been forced to show flexibility and adjust the benefit structure, including reimbursement levels, according to the actual situation.

6. The provincial administration is controlling and auditing the usage of the central subsidies. Local governments, especially the county health bureau and its NRCMS management bureau, are expected to adapt the funding and reimbursement quotas of the local NRCMS to local conditions in order to achieve sustainability of the schemes and higher efficiency of central funding, increasing from 10 yuan to 40 yuan per enrolled capita in the western regions between 2005 and 2009.

7. It is a policy experimentation not for further innovation but for consolidation of central subsidies. Defined goals should be optimised according to the local conditions without, or with only marginal, consequences for central policy-making. Reimbursement percentages should be adapted to local needs and perceptions and guarantee efficient usage of central funding. There has been only marginal change or concretisation of central policy between 2002 and 2007, but there is a tendency to increase financial subsidies from central government. Instead of innovation, higher efficiency of these funds in reimbursement procedures and ratios is needed. The state subsidies have a varying impact on the different health providers. Because of the NRCMS focus on serious illnesses, state subsidies can only be used in hospitals belonging to the state health care system. They are not available for consultations, treatment and drugs prescribed on an outpatient basis. In fact, these subsidies to the farmers are a form of user-oriented subsidy for state health providers.

A more empirical policy-focused analysis of neglected areas like health care shows a diversity of rural governance. This discussion of modes in rural health governance does not aim to negate existing concepts of analysis of Chinese politics, but tries to show the focus on a very limited number of policy fields. We argue that forms of intervention or participation depend highly on the structure of the analysed policy field and forms of state capacity.

Successful transformative state capacity creates or enhances the interdependence of different actor groups. In a highly decentralised policy field like rural health care, the health administration can play only a limited role. In a situation of weak or inefficient administrative regulation, local agents will be more able to respond in the intended ways if central regulators rely on measures to embed their goals into the motivations and commitment of the stakeholders or patients as a source for transformation state capacity. This strengthens the goals of the central state institutions without necessarily involving higher direct forms of political control or administrative intrusion. Therefore, strong transformative state capacity is not characterised by weak societies, but relies on embedded strong local agents, societal agents or local feedback loops. Responsiveness, as outlined above, to introducing NRCMS has generated new interdependencies between state agencies and the rural population. As in policy patterns with a withdrawal of the central state from the traditional down-the-line (tiao) legislation and administration, particularly on issues not requiring nationwide uniformity and cohesion resulting in new policy innovation in the financial sector, local interaction is critical.⁹³

This study suggests that enhanced local accountability and responsiveness relies neither on informal and intact small communities⁹⁴ nor on infrequent very objective audits from above. It operates within a non-hierarchical institutional decision-making structure (‘polity’), guided only by centrally established framework regulations and finding new interdependencies between society and state. Where public-private partnerships or non-governmental organisations can rely on established regulations and polity to work and protect the rights of citizens by integrating their voice and resources, in responsive settings their perception and behaviour can be integrated by responsive regulation as a resource for transformative state capacity.

Responsive and experimental governance is a resource for guided transformative state capacity. Annual enrolment generates high costs of mobilisation and administration. Nevertheless, while economically speaking the approach used for the introduction of NRCMS is not optimal, it is particularly attractive to policy fields where the state is weak and searching for effective ways to distribute additional funds in a more effective way to less controlled local administration.