WPS5530 Policy Research Working Paper 5530 Equity and Public Governance in Health System Reform Challenges and Opportunities for China Hana Brixi Yan Mu Beatrice Targa David Hipgrave The World Bank East Asia and Pacific Region Human Development and Poverty Reduction and Economic Management Departments January 2011 Policy Research Working Paper 5530 Abstract Achieving the objective of China's current health system prefectures and lower levels within provinces. The reform, namely equitable improvements in health increasing inequity in public expenditure at sub-national outcomes, will be difficult not least because of the levels indicates that incentives, responsibilities, and continuously growing income disparities in the country. resources at sub-national levels are not well aligned with The analysis in this paper shows that since 2000, disparity China's national priorities. To address the weaknesses in selected health outcomes has been declining across in equity and efficiency that characterize China's health provinces, largely due to earmarked central government system and health outcomes, China's health system allocations. By contrast, public expenditure on health reform may require complementary reforms to improve is increasingly regressive (positively correlated with governance for public service delivery across sectors. local income per capita) across provinces, and across This paper is a joint product of Human Development and Poverty Reduction and Economic Management Departments in the East Asia and Pacific Region. It is part of a larger effort by the World Bank to provide open access to its research and make a contribution to development policy discussions around the world. Policy Research Working Papers are also posted on the Web at http://econ.worldbank.org. The authors may be contacted at hbrixi@worldbank.org. The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent. Produced by the Research Support Team Equity and public governance in health system reform: Challenges and opportunities for China Hana Brixi1, Yan Mu 2, Beatrice Targa3, and David Hipgrave2 1 World Bank, Washington, DC, USA 2 UNICEF Office for China, Beijing, PR China 3 UNICEF Consultant, Arezzo, Italy Address for correspondence: Hana Brixi, The World Bank, 1818 H Street NW, Washington, DC 20433, USA. Tel. 1-202-458-5131, FAX: 1-202-522-3394, email: hbrixi@worldbank.org Authors' contribution Hana Brixi designed and led the analysis. Mu Yan contributed to the study design and analysis. Hana Brixi, Mu Yan and Beatrice Targa conducted the systematic review of the available literature and evidence, extracted the available data from the original statistics sources, conducted the analysis, and drafted the paper. David Hipgrave contributed to the structure and content of the final text of the manuscript. All authors provided important intellectual and material input at different stages of the work and contributed to drafting the paper. The corresponding author (Hana Brixi) had full access to all the data in the manuscript and had final responsibility for the decision to submit for publication. The authors thank Jennifer Fong for research support, John Langenbrunner for valuable comments, and participants in seminars organized by the National Development and Reform Commission, Ministry of Health and UNICEF in China for feedback. Equity and public governance in health system reform: Challenges and opportunities for China 1. Introduction In early 2009, China's Ministry of Health (MoH) announced a comprehensive set of health system reforms (HSR) designed to rectify the widely acknowledged and increasing inequity that characterized China's health services. (1) China's HSR started formally in 2006, after President Hu Jintao demanded universal access to quality basic healthcare and better health outcomes for the Chinese people, followed by an open apology for problems of affordability and accessibility to the health sector by then Vice Premier Madam Wu Yi. The State Council established a 16- ministry leading group for, and opened an unprecedented public debate on the issues and proposals for HSR, with wide participation of citizens as well as the domestic and international expert community. The reforms aim to provide "safe, effective, convenient and affordable" healthcare to all urban and rural residents through five main pillars: strengthening public health functions and services, enhancing primary care delivery, establishing universal basic health security (insurance), ensuring the safety of and access to essential medicines, and optimizing the management of public hospitals. (2) Inequity in China's health sector has been driven by the comparatively low level of government funding, averaging only 0.7-0.9% of gross domestic products (GDP) compared to about 2% in comparable middle-income countries. Out of pocket spending as a proportion of total health expenditure before the reforms was around 60%. (3) Recognizing that China's HSR required an increased financing commitment, in March 2009 Premier Wen Jiabao announced at the 11th National People's Congress an RMB 850 billion allocation (2.8% of 2008 GDP) to support HSR during 2009 - 2011 (4), and government expenditure on health increased to 1.4% of GDP in 2009. (5) This is still low compared to government expenditure on education (3.5% of GDP in 2008) (6), but brings China closer to a WHO estimate that government spending on health of 1.5-2.0% of GDP could guarantee primary health care (public health and a modest package of essential clinical care) for all in China. (7) The announcement of the priorities for and content of the reforms by national government, along with the expectation of their implementation at all administrative levels, has placed a great burden on provincial and sub-provincial health authorities. Moreover, ensuring that the concomitant injection of new fiscal resources contributes effectively to the accomplishment of HSR objectives is a challenge, since about 90 percent of public resources for health care (as well as other public services) is generally allocated at the provincial and sub-provincial levels. (8) Promoting allocative and operational efficiency in public expenditure at these levels is a particular challenge that has yet to receive appropriate attention nation-wide. In this paper, we explore whether the government's actual expenditure on health across levels support China's HSR objectives. We analyze the available national, provincial and sub-provincial statistics to explore recent trends in relation to equity in certain key health outcomes and public 2 resource allocations in China. In particular, we study to what extent public resource allocations have remained correlated with the local GDP per capita, and how and why local health outcomes have changed in this context. We also include information on governance issues pertaining to health sector funding allocations, and conclude with our perspectives on how China's HSR provides an opportunity for improving governance in public sector financing in and beyond the health sector. 2. Disparity in health resource allocation across administrative levels Burgeoning inequity despite national government initiatives Health financing in China is decentralized across five levels of government (Figure 1) and the public contribution is largely determined by local fiscal capacity. (9) Local governments, namely the lowest two out of the five tiers of government, bear the main responsibility for financing essential public services and their fiscal capacity differs widely across China even after adjusting for equalization transfers, which are formula-based grants by central government according to measures of expenditure needs and fiscal capacity cross localities. The extent of inequality in public spending on health per capita across provinces has risen since 2001 (figures 2 and 2A). This is also true in other areas of public services, such as education in the period 2000-2008 (figure 3). Another source of burgeoning inequity is that total health spending per capita within provinces is increasingly skewed toward urban areas (figure 4), where household income is higher. Within provinces and also within prefectures, evidence suggests that inequity in spending on health is also rising. Examples of rising disparity across prefectures in Inner Mongolia and Shandong and across district levels in Jinan and Zhengzhou municipalities are shown in figures 5 ­ 6. (Annex A provides descriptive statistics. Graphs in Annex figures A1 ­ A3 show similar inequity in health spending at sub-national level.) A similar situation appears in government expenditure on education at both prefecture and sub-prefecture levels (Annex A figure A4). At household level, the availability of public resources for health also differs across population groups according to their participation under the different health security schemes available in China. Figure 7 shows that since 2002, whilst China has introduced several major new schemes to benefit previously excluded population groups and make the allocation of public resources more equitable, the size of the allocations per capita is, as yet, far from equal. Similarly, in other social sectors, including old-age pensions and education, China has been introducing programs to target the poor trying to overcome the overall bias toward better off population groups. (10) Finally, catastrophic and out-of-pocket health expenditures continue to demonstrate considerable inequity in China. During 2003-08, as National Health Service Survey reports (11) show, incidence of catastrophic health spending among low-income households remained about 10 percent in rural areas and increased from 4 to 6 percent in urban areas (figure 8). Average out-of- pocket payment (after insurance reimbursement) for a single inpatient stay in 2008 was 50-70% of annual income per capita in rural areas, about twice the levels common in urban areas (figure 9). Moreover, to some extent, the funding increases for rural health care through the restored 3 rural cooperative medical scheme (RCMS) and specifically through subsidies for hospital delivery may have been associated with cost inflation, limiting the reduction in absolute out-of- pocket payments per intervention (figure 10 showing the development in cost and out-of-pocket expenditures per hospital delivery in absolute terms ­ on left axis ­ and as a share of annual income per capita ­ on right axis -- by urban and rural typology). Some inequities are closely related to the existing inefficiencies Since the early 1980s, China's public resources have facilitated rapid improvements in availability of advanced care in specialized urban hospitals. As public resources are biased toward higher-level facilities, they disproportionately benefit higher-income households. (12) A similar bias in the allocation of public resources toward higher-level facilities and higher-income population groups is also seen in the education sector, where "key schools" at each level benefit from higher levels of public funding, and are more easily accessed by the children of the higher income households. In China's cities, advanced care is available at levels comparable with advanced economies (for example, the availability of advanced technologies per person in large Chinese cities matches the levels in most OECD countries since the mid- to late-1990s (13) and various life-cycle indicators match or exceed those of the developed world). Beijing and Shanghai, for instance, report prevalence of magnetic resonance imaging machines and other advanced medical equipment ­ funded partially by public resources ­ exceeding levels common in European cities. (14) Primary care in urban areas, however, has been unnecessarily expensive, often delivered by hospitals and specialized doctors instead of health centers and general practitioners with qualified nurses. (12) By contrast, the fraction of public resources that benefits rural (township) health centers is disproportionately small (figure 11), while poor quality of care, over-servicing and irrational use of drugs remain serious problems there, as well as in urban areas. (1) Poor rural households still face financial obstacles in accessing cheap effective care (such as management of newborn and respiratory illnesses that remain common causes of young child mortality). A government-UN review in 2006 found that during 2000-2004, preventable conditions accounted for about 70% of neonatal deaths, particularly in poor rural areas. (15) A more recent Lancet review of under-five deaths in China (16) also infers that a majority of these would be preventable with interventions that are commonly available and cheap to provide. For China's HSR, inefficiency and inequity in the distribution of financial, human and physical resources (17) (18) are an area of focus. China's HSR framework includes many appropriate elements to promote cost-effectiveness and overall operational efficiency in public resource allocation in health. Efforts to allow cost- recovery in primary care, implement the list of essential medicines, reduce the dependence on service fees in provider payment mechanisms, strengthen health centers and enhance the management of public hospitals are all included. The bias toward advanced care, however, continues to surface in the process of HSR implementation at the local level. The detailed design and implementation of most of the social protection schemes in health, including the RCMS, Medical Financial Assistance (MFA) and Urban Residents' Basic Medical Insurance (URBMI), occurs at county/district level, and for the most part excludes outpatient care. Research in 2008 estimates that only 15% of counties cover 4 outpatient as well as inpatient care in their RCMS schemes (19), generating incentives to providers to admit patients for ailments that could be treated more cheaply at home. Out-of-pocket payments as a share of medical bills remain significantly higher for outpatient care compared to inpatient care in both urban and rural areas. The most recent National Health Services Survey (NHSS) (11) shows that in 2008, 33% of patients received partial reimbursement for outpatient care compared to 85% who received partial reimbursement (in the average amount of 35% of the medical bills) for inpatient care. This is confirmed by field research. Citizens' scorecard survey shows that out-of-pocket payments as a share of medical bills remain significantly higher for outpatient care compared to inpatient care, and reach 84% of monthly per capita income among patients in the poorest quintile (compared to about 11% in the richest quintile) for an average outpatient visit, and a staggering 140% of annual per capita income in the poorest quintile (compared to about 10% in the richest quintile) for a single average hospitalization episode. (20) Earmarking of vertical programs to promote equity may also be failing in some areas To implement national priorities, the central government has been circumventing the problem of resource allocation at local level by increasingly relying on vertical programs and earmarked specific purpose transfers (that are conditional, to incentivize sub-national governments, as opposed to general-purpose transfers that are unconditional to provide general budget support). As part of China's HSR, six new public health initiatives (for hepatitis B vaccination of older children, folic acid supplementation, fluorosis prevention, cervical and breast cancer screening for rural women, cataract treatment, and rural water and sanitation initiatives) have been recently added to the hundreds of earmarked transfers already in existence in the health sector. Currently, the entire central government budget for the health sector consists of earmarked transfers (accounting for RMB 112 billion in 2009). Central government spending on health does not include any general-purpose transfers to benefit poorer localities. Although there are over one hundred earmarked transfers in the area of public health alone, their total allocation is relatively small compared to the overall inequities in public resource allocation on health between urban and rural areas, and between wealthy and poor, and eastern and western provinces. (21) In some areas of social sector spending, such as the programs to reduce maternal mortality and support 9-years of compulsory education, earmarked transfers have supported measurable improvements in outcomes. Figure 12 demonstrates a marked reduction in the ratio of maternal mortality in rural versus urban areas, from 2.6 in 1991 to 1.3 in 2009, with most of the reduction occurring since 2002 when subsidies for hospital delivery jumped up in poor rural areas. Earmarked transfers for basic compulsory education, complemented by close monitoring of student enrolments, have helped strengthen basic education across China. Similarly, earmarked transfers and intensive real-time monitoring have facilitated successful roll-out of the dibao cash transfer scheme nation-wide. The reliance on earmarked transfers for such programs, however, results in concerns about the sustainability of any observed improvements and the predictability of funding at the local level. Furthermore, the selection of programs for earmarked transfers may depend on sectoral agencies' or local governments' lobbying ability rather than to a rigorous assessment of needs. Moreover, local studies have revealed weaknesses in monitoring and, in some instances, low compliance in 5 the use of earmarked transfers. For instance, county governments may offset earmarked transfers by cutting operating budget or raising staff numbers for the transfer recipient agencies. (22) Finally, although vertical programs and earmarked transfers may help deliver results in specific areas they are not an appropriate financing strategy for effective HSR implementation. The large number of vertically-funded health interventions and the associated problems of monitoring suggest that it would be impossible to achieve the HSR's management-strengthening components and develop China's proposed universal primary health care system as a sum of vertical programs. 3. Socio-economic and health outcome disparities and decentralization in China Decentralization and China's economic development Achieving equitable improvements in health outcomes is challenging in any country. It is especially challenging in the context of China's large and growing disparities in economic development and incomes across localities and population groups. (23) Decentralization and competition among localities has played a role in the success of China's economic development over the past 30 years. (24) (25) Local governments' ability to experiment and adapt the implementation of national policies has promoted local economic performance and helped the central government gradually enhance national policies. Decentralization and competition among local governments, however, have also contributed to the growing disparities across China. Growing income disparities have arisen across localities and population groups, often within local jurisdictions. (26) (27). Along with income disparities, disparities in human development indicators, such as child mortality rates, are significant across China, and are partly driven by inequities in access to essential public services. (28) Central government initiatives to resolve disparities in social outcomes Over the past decade, the Government of China has adopted a range of policies to reverse the trend of rising disparities with good results on several fronts. Broad economic development strategies, such as Going West to support the development of 12 Western provinces (launched at the National People's Congress in 2000) have helped to contain the income disparities between the coastal and inland provinces and facilitated overall improvements in selected inland provinces. With respect to human development outcomes, earmarked transfers from the central government budget have supported programs to reduce inter-provincial and rural-urban disparity particularly in the maternal mortality ratio (MMR), and to a lesser extent also infant and under-five mortality rates (U5MR) (rural:urban disparity in the latter has also fallen by around 20% in the last decade). These earmarked transfers are of three main types: a) targeted vertical schemes, such as for nine-years of compulsory education, hospital delivery subsidies for rural women (29) now in all 2297 rural counties, improvements to the Expanded Program on Immunization, the six public health programs listed earlier and most recently the 15 yuan per capita essential public health payment ­ again implemented by county health authorities 6 according to local priority; b) social protection payments, such as the RCMS, URBMI, the MFA and social assistance dibao cash transfers for the rural poor, and c) infrastructure investments, such as construction of health facilities, improvement of drinking water sources, sanitation, and rural access roads. These earmarked transfers may have contributed to some of the reduction in disparity in selected outcomes across provinces and across the rural and urban areas nationally. The remaining inequities in health outcomes, however, have multiple roots, including social determinants, underfunding of primary care, wide-spread incentive distortions among the providers of care and financing and design weaknesses in the RCMS and MFA. (30) (31) (32) (33) (34) Comparisons in health outcomes across provinces Disparity in health indicators in relation to income per capita across China's provinces is of a similar magnitude to that observed across countries, revealing the large scope for the equalizing role of government. The MMR, U5MR and IMR, for example, while on average lower in China's provinces than in countries with a comparable GDP per capita, are as closely related to the provincial GDP per capita in China as the levels across countries are to the national GDP per capita world-wide (figures 13, 13A and 13B). Across China's provinces, disparities in maternal and child health outcomes have matched income disparities, but the correlation progressively weakened during 2000-2008 (see Annex Figure B1 for annual figures), suggesting that government initiatives to reduce inequity even before the HSR, supported by technical interventions to improve service content and quality and human resource capacity, were proving effective (figures 14, 14A and 14B). Further, although rural-urban household income disparities nation-wide grew in the last decade, rural-urban disparities in maternal and to a lesser extent infant and child mortality have declined, again predominantly over the last decade (figure 15). Comparisons within provinces Socio-economic disparities within provinces have also worsened. (35) Since 2000, all but four western provinces (Sichuan, Tibet, Xinjiang and Yunnan) registered a continuously rising ratio of urban per capita annual disposable household income to net rural household income (the ratio between per capita annual disposable income for urban households and per capita annual net income for rural households) (Annex Table B1), which in some provinces reached 4:1 compared to the national ratio of 3.3:1 in 2009. (8) With respect to health indicators, some provinces have achieved equitable improvements. Inner Mongolia (one of the few provinces for which prefecture data are readily available), for instance, has reduced infant mortality faster in poorer prefectures (figure 16). However, intra-province disparity in health outcomes remains high between and within rural and urban areas respectively, (figure 17, Annex figures B2 and B3 and Annex tables B2-B3). The infant mortality rate in the best and worst districts/counties differed 10-fold in Chengdu (36) municipality and 12-fold across prefectures in Gansu (37) province, both in 2007. 7 In conclusion, the improvements in equity in rural versus urban maternal and child mortality in China over the last decade appear linked to national initiatives in health security funding, hospital delivery subsidies and possibly infrastructure improvements and training of staff. We do not have evidence to suggest comparable improvements in mortality equity within provinces, but the decreasing equity in health funding allocations and persisting inequity in out-of-pocket payments and per capita allocations would infer the opposite. 4. The influence of public sector governance weaknesses on HSR implementation The weaknesses in equity and efficiency in public expenditure on health and other social services may be partly explained by weak incentive structures in the provincial and sub-provincial governments. Efficiency and particularly equity in public resource allocation have not been influential performance indicators for governments and service providers at sub-national levels. (38) Local government officials are not accountable for equity in local health outcomes and for equity and efficiency in public resource allocation. In this respect, health differs from education, where the achievement of universal 9-year compulsory education is subject to strict monitoring and performance evaluation at the local levels. Introducing equity in health outcomes and public resource allocation at sub-national level has also been a challenge also in the development of the monitoring and evaluation framework for HSR implementation. As a result, provincial and sub-provincial governments may not have the incentives or capacity to comply with China's HSR objectives. (39) (20) Surveys (39) (20) and insights offered by government officials interviewed for this analysis suggest that provincial, prefecture and municipal governments may partly withhold public resources originally targeted to counties, townships and villages in need. In particular, interviews have confirmed that beyond the earmarked transfers and selected nation-wide priorities, provincial and lower levels of governments favor spending "close to home"; that is, mainly in the provincial capital city and at the prefecture and municipal level. Transfers ­ even if targeting specific counties ­ are transmitted through provincial governments/autonomous regions/municipalities and prefectures/districts (as illustrated in figure 1) and would be disproportionately spent on urban development (mainly urban infrastructure), leaving a shortage of resources at the county level and below to spend on essential social services. (39) It is worth noting that weaknesses in sub-national public sector governance also complicate the enforcement of health-related laws, regulations and standards. For instance, although China passed a strict national food safety law and introduced a series of food safety standards in 2009, implementation of the law is poorly regulated, and food safety problems (such as the resurgence of melamine-contaminated dairy products in 2010) have persisted. Chinese media have also reported enforcement weaknesses in the areas of environmental, road, patient and drug and vaccine safety among others. Possible conflicts of interest at local level (such as local economic growth versus public health safety) are not systematically monitored and addressed. Sub-national governments are yet to become truly accountable for local performance in the areas of regulatory and law enforcement, policy implementation and the financing and delivery of services, such as health care. Recently, the case of Shenmu County in Shaanxi province, where 8 local government has been strongly committed to a high-profile HSR pilot, has illustrated the difficulty of building an adequate surveillance and enforcement capacity at the local level. In spite of the effort so far, the National Audit Office reported misappropriation of health resources, including fake invoices. (40) Ensuring appropriate implementation of HSR may require improved monitoring and management of government performance across the different sub- national levels. International experience has shown that factors outside the health system, namely the governance and macroeconomic and social policy environment, may act as constraints on health system strengthening. (41; 42) The health system strengthening agenda across countries has largely taken these constraints as given and exogenous. China, however, approached HSR with such a high level of government commitment that recognizing the broader institutional constraints of HSR might in fact help motivate improvements in public sector governance. 5. Public sector governance considerations Our analysis indicates that public resource allocation and the underlying incentives at the sub- national levels may need to be better aligned with China's national priorities in order to facilitate HSR implementation. The public sector governance challenge for HSR may be how to ensure that sub-national governments have their responsibilities clearly defined in line with the national policies, standards, laws and regulations; how these responsibilities are implemented and how this is independently, reliably and regularly monitored, and that they have and allocate adequate resources so as to fulfill said responsibilities. International experience suggests that this may require strengthening accountability relationships across government levels and agencies, and between government agencies, providers of care and citizens. (43) Stronger accountability relationships would, in turn, allow for an increasing reliance on general-purpose equalization grants ­ in recognition that incentives of sub-national governments match national policy priorities ­ as opposed to earmarked transfers. Given China's size and decentralization in financing and delivery of public services, it may be crucial to strengthen the role and accountability of provincial governments. Provincial governments may have to become explicitly responsible for equity and efficiency in public resource allocation, for national policy implementation, enforcement of laws, standards and regulations, and for adequate health system performance within the entire province. The central government could specify viable fiscal targets for expenditure on primary health care across provinces and define the outputs and outcomes that each province should achieve in an equitable and cost-effective manner in the context of HSR. Making provincial governments explicitly responsible for results in HSR implementation at all levels may strengthen their commitment to improving public resource allocation, compliance and performance monitoring across levels in each province. Furthermore, HSR implementation may be facilitated by centralizing key financing responsibilities and schemes at the provincial level. A single agency at the provincial level, for instance, could manage all social protection schemes in health (including the RCMS, Urban Employee BMI, URBMI and MFA schemes). Pooling resources for each of these schemes at the 9 provincial level would help address intra-provincial inequity. In addition, provincial-level agencies could develop capacity to establish a viable contracting and performance evaluation arrangement with the providers of care. Provinces could be incentivized to explore alternative service purchasing and payment mechanisms, based on schemes already piloted abroad, to improve efficiency of service provision. Moreover, provincial governments could boost their capacity to monitor the use of public resources by replacing the existing cascading system of transfers (which moves resources through several levels of government before they are actually spent) with direct payments from the provincial treasury system (via a treasury single account that is already operational in many provinces). Importantly, the central government will have to effectively monitor and evaluate the use of public resources, policy implementation and overall service delivery performance across provinces, holding the provincial governments to account. As an innovative measure, citizen score card surveys could become a useful tool to gather citizens' feedback regarding their experience with public services (including their ability to utilize primary care and other services, the required fees and out-of-pocket payments, their ability to access relevant information and resolve complaints, and their satisfaction with services and with the performance of service providers, insurance schemes, local government agencies and others). Such a direct mechanism for obtaining citizens' feedback would allow the central government to better assess policy implementation performance at the local level, particularly with respect to equity and quality in service delivery. (20) The independent assessment could effectively feed into a comprehensive performance management system and help strengthen accountability at the provincial and sub-provincial levels (across government agencies and providers) for the delivery of health care and other public services and their outcomes. The strong monitoring, evaluation and performance management system, internalizing and addressing citizens' feedback, with respect to service delivery (outputs) and its outcomes as well as public resource allocation at the local level, will boost incentives (and hence allow for a greater autonomy) at the provincial and sub-provincial levels. In fact, overall improvements in public sector governance reforms are likely to generate equitable improvements in the health of Chinese citizens beyond HSR. This is because public governance reforms would enhance essential public service delivery in line with the national policies across sectors improving the social determinants of health, such as access to safe water, sanitation, basic health education, housing, rural access roads, social assistance, and others. (44) Acknowledgements We thank John Langenbrunner (World Bank) and Sarah Barber (WHO) for comments, and Jennifer Fong for assistance in data collection and analysis. 10 References: 1. Blumenthal, D and Hsiao, W. Privatization and Its Discontents The Evolving Chinese Health Care System. The New England Journal of Medicine. 353 2005, pp. 11651170. 2. CCCPC. The Central Committee of CPC and the State Council's Joint Guidelines for Deepening the Medical and Health Sector Reform. Beijing : CCCPC, 2009. Circular No. 60. 3. Organization, World Health. World Health Report. Geneva, Switzerland : WHO, 2008. 4. China Government Net. The Central People's Government of China. [Online] 03 14, 2009. [Cited: August 10, 2010.] http://www.gov.cn/english/official/200903/14/content_1259415.htm. 5. China Health Economics Institute. China National Health Accounts Report Abstract. Beijing : Ministry of Health, 2009. 6. Ministry of Education, National Bureau of Statistics, Ministry of Finance. The 2008 Statistical Notice on National Education Expenditure. Beijing : s.n., 2009. 7. World Health Organization. Health in China's Harmonious Society: Building Health System to Benefit All. s.l. : WHO, 2007. 8. National Bureau of Statistics. China Statistical Yearbooks. Beijing : China Statistics Press, Various Years. 9. Feltenstein, A and Iwata, S. Decentralization and macroeconomic performance in China: regional autonomy has its costs. Journal of Development Economics. April 2005, Vol. 76, 2. 10. Xiao and Xiao, Y. Equity of Pension System in China,. Shanghai Economic Research. 2008, Vol. 8. 11. Centre for Health Statistics and Information, Ministry of Health,. An Analysis Report of National Health Services Survey in China in 2008. Beijing : Press for China Union Medical University, Dec 2009. 12. Meng, Q. Equity, Efficiency and Sustainability of Health Financing in China. Health Economics Study,. 2007, Vol. 4. 13. OECD. OECD Health Data. Paris : Organization of Economic Cooperation and Development, 2010. 14. Chen, B. Allocation Planning for Big Medical Equipment in China. China Medical Equipment. 2007, Vol. June. 15. Ministry of Health, World Health Organization, UNICEF and UNFPA. Joint Review of Maternal and Child Survival Strategy,. Beijing : In Press, 2006. 16. Rudan, I*, et al. Causes of deaths in children younger than 5 years in China. The Lancet. on behalf of WHO/UNICEF's Child Health Epidemiology Reference Group (CHERG), 2010. 17. Anand, S, et al. China's human resources for health: quantity, quality, and distribution. The Lancet. 2008;, Vol. 372, 1774 1781. 18. United Nations Development Programme, China Office. Human Development Report China 2007/2008 . [Online] http://unpan1.un.org/intradoc/groups/public/documents/undpadm/unpan042714.pdf. 19. Hu, SL. The Implementation and Evaluation of the Rural Cooperative Medical Insurance Scheme. China Health Economics. February 2008. 20. Brixi, H. China: Urban Services and Governance. . Washington, DC : World Bank , 2009. Policy Research Working Paper No. 5030. . 21. Chen, CH and Li, SP. Discussion on the Ways of Chinese Central Finance Health Transfer Payment. China Health Economics. 2010, Vol. 1. 22. Liu, M et al. The National Audit Report on Central Budget Implementation and Uses of Other Financial Resources revealed relatively high levels of misappropriation in the audited accounts. 2009. 23. World Bank Group. From poor areas to poor people: China's evolving poverty reduction agenda. Washington DC : World Bank, 2009. 24. Qin, Y, Jin, HH and Weingast, B. Regional Decentralization and Fiscal Incentives: Federalism, Chinese Style. s.l. : University of California at Berkeley, 2001. 25. Lin, J YF and Liu, ZQ. Fiscal Decentralization and Economic Growth in China. Economic Development and Cultural Change,. October, 2000, Vol. 49(1). 26. Xing, L, et al. Intra Rural Income Disparity in West China. China Economic Quarterly. 2008, Vol. 8, No. 1. 11 27. Zheng, M, Fu, Q and Wang, XH. Comparative Study on Structural Changes in Income Disparities in Urban Households in Chongqing Municipality, Shanghai Municipality and Sichuan Province. Journal of Reform and Strategy. 2008, Vol. 5. 28. Tang, S, et al. Tackling the challenges to health equity in China,. Lancet. 2008, Vol. Oct 20. 29. Feng XL, Shi G, Wang Y, Xu L, Luo H, Shen J, et al. . An impact evaluation of the safe motherhood program in China. Health Economics . 2010, Vol. 19(S1). 30. World Bank. Reforming China's Rural Health System. . Washington, DC : World Bank, 2009. 31. Ying, YZ. Benefit for Women and Children in Rural Cooperative Medical Scheme. Beijing : UNICEF Report, 2009. 32. Ma, J and Zhao, M. Improve the Design of Medical Financial Assistance Programme and Build a Harmonious Society ­ Implementation of MFA in Shanghai. China Health Economics. 2009, Vol. Nov. 33. Yu Baorong, Meng Qingyue, Tang Shenlan, Lennart Bogg, . Health Service Utilization by Rural Residents in Shandong and Ningxia Provinces . China Health Economics. 2008, Vol. May. 34. Singer Babiarz, Kimberly et al. New evidence on the impact of China's New Rural cooperative Medical Scheme and its implications for rural primary healthcare:multivariate diferenceindifference analysis. BMJ, 2010. 341:c5617. 35. UNDP and China Institute for Reform and Development. China Human Development Report. Beijing : s.n., 2008. 36. Chengdu Development and Reform Commission, Chengdu Bureau of Statistics. Chengdu Social Development Report 2008. Chengdu : s.n. 37. Gansu Provincial Bureau of Statistics. Gansu Social Progress Report. Lanzhou : s.n., 2008. 38. Guo, Y. et al. Tracking China's health reform. Lancet. March 27, 2010 , Vol. 375 . 39. Subprovincial Intergovernmental Fiscal Transfers, presentation delivered at 2006 Annual China Fiscal Reform Forum. Liu, MD. s.l. : UNDP Conference Volume., 2007. 40. National Audit Office. National Audit Report on Central Budget Implementation and Uses of Other Financial Resources. Beijing : National Audit Office, 2010. 41. Hanson et al. Expanding access to priority health interventions: a framework for understanding the constraints to scalingup. Journal of International Devleopment. 2003, Vol. 15, 114. 42. Travis, P. et al. Overcoming healthsystems constraints to achieve the Millennium Development Goals. Lancet. 2004, Vol. 364, 9437. 43. World Bank . Making Services Work for Poor People. Washington DC : World Development Report, 2004. 44. World Health Organization. Commission on Social Determinants of Health. 2008. http://www.who.int/social_determinants/thecommission/finalreport/en/index.html. 12 Figure 1: China's fivetier government Nation Municipality Autonomous Province directly under Region State Council Prefecture / Autonomous Prefecture / Prefecture level Municipality Rural County / Urban District Town / Urban Township Community Source: Authors 13 Figure 2: Provincial government operating expenditure on health in relation to local GDP, 20012006 450 Provincial government 400 operating expenditure on health per capita (RMB) 350 300 250 200 150 2006 100 50 2001 0 1500 15000 Provincial per capita GDP (RMB, log scale) Note: A comparable figure for 2000 is not available. A new government budget classification methodology was introduced in 2007. Operating expenditure on health was a pre2007 term, primarily including government subsidies to providers and excluding some health security programmes. From 2007, the definition of health spending was expanded. Source: China Statistical Yearbooks Figure 2A: Provincial government expenditure on health in relation to local GDP, 20072008 900 Provincial 800 government spending on 700 health per capita (RMB) 600 500 2008 400 300 200 100 2007 0 5000 50000 Provincial per capita GDP (RMB, log scale) Source: China Statistical Yearbooks 14 Figure 3: Provincial government expenditure on education in relation to local GDP, 20002008 2000 Provincial government 1800 spending on education per capita (RMB) 1600 1400 1200 2008 1000 800 600 400 200 2000 0 1500 15000 Provincial per capita GDP (RMB, log scale) Source: China Statistical Yearbooks 15 Figure 4: Rural and urban total health expenditure per capita within provinces, 2000 and 2006 150 rural health expenditure (2005 PPP US$) 100 50 2000 2006 urban health expenditure (2005 PPP US$) 0 0 50 100 150 200 250 300 350 Source: Gapminder (www.gapminder.org). 16 Figures 5 and 5A: Prefecturelevel government expenditure on health versus local GDP in Inner Mongolia and Shandong, 2000 and 2007 Inner Mongolia Government spending on health per 500 400 2007 capita (RMB) 300 200 100 2000 0 1500 15000 Prefecture GDP per capita (RMB log scale) Shandong Goverm,emt Health Spending p.c (yuan) 350 300 250 200 2007 150 100 50 2000 0 1000 10000 100000 Prefecture GDP per capita (RMB log scale) 17 Figure 6 and 6A: County and districtlevel government expenditure on health versus local GDP in Jinan and Zhengzhou Municipalities, 2000 and 2006 Jinan County/district health spending per capita 90 80 70 60 2006 (RMB) 50 40 30 20 2000 10 0 3000 30000 County/district GDP per capita (RMB, log scale) Zhengzhou County/district health spending per capita 160 140 120 2006 100 80 (RMB) 60 40 20 2000 0 1000 10000 County/district GDP per capita (RMB, log scale) 18 Figure 7: Government financing per participant across health security schemes introduced during 1950 2007 3000 Per capita government Employees for Public Administrative Units and Organs (10 mil) 2629 2500 funding (RMB) 2000 1500 Urban Medical Financial Assistance Family Members of Employees of Public 1000 (4.44 mil) 797 Revolutionary Martyrs Sector Services Units and and ExServicemen Organs (39 mil) 533 (4.68 mil) 510 500 Rural Medical Financial Assistance RCMS (7.6 mil) 396 (815 mil) 0 78 Urban Residents' Basic Medical Public Health Security Insurance (199 for Civil Servants 500 (Gongfeiyiliiao) mil) 40 year of 1950s 1990s 2002 2003 2007 launch Note: Bubble size is equivalent to the number of participants. Number of participants is shown in parentheses. Government spending per participant is shown in red. Government funding figures are annual per person, except for the rural and urban medical financial assistance, reported per case. Source: National Health Account Report 2009 and China Health Statistical Digest 2010 19 Figure 8: Catastrophic health expenditure among lowincome households, 2003 and 2008 12 Incidence ratio of catastrophic spending 10 8 6 % 4 2003 2008 2 0 Source: China National Health Services Survey (NHSS) 2003 and 2008. Note: According to China's NHSS, outofpocket health spending is catastrophic when exceeding 40% of annual household (HH) consumption. Low income HHs are those with per capita annual income below 50% of the mean annual HH income in the jurisdiction. 20 Figure 9: Average cost and outofpocket expenditure on inpatient care after insurance reimbursement, relative annual income per capita, by urbanrural typology, 2008 12000 70 10000 60 % of annual income p.c. 50 8000 40 cost 6000 30 4000 20 2000 10 0 0 average total cost average OOP share Source: NHSS 2008 and 2003. Note: The China MoH urbanrural socioeconomic typology system is described in the 2006 joint governmentUN review 21 Figure 10: Outofpocket expenditure for hospital delivery, 20032008 117 2000 120 108 105 2008 OOP as % of 2003 OOP 2008 Hospital 89 100 delivery average out Hospital delivery oop 1600 ofpocket payment (after NRCMS 80 reimbursement) 1200 2003 Hospital delivery outof 60 pocket payment (without NRCMS 800 coverage) 40 Hospital delivery outofpocket 400 payment in 2008 20 relative to 2003 (%) 0 0 Large Medium Small Type I Type II Type III Type IV cities Cities Cities Urban Rural Source: NHSS 2008 22 Figure 11: The distribution of government subsidy across providers, 2008 2200 2000 1800 Government subsidy 1600 Medical service fees 1400 Drug revenues 1200 Other revenues 1000 800 600 400 200 0 City hospitals County hospitals Township centers Community centers Source: 2008 National Health Financial Report, MoH 23 Figure 12: China's overall, rural and urban maternal mortality ratios, 1991 ­ 2009 Deaths per 100,000 live births 120 112.5 100 88.9 80 Rural Total 60 Urban 45.9 40 34 31.9 20 26.6 0 Source: Ministry of Health, China Health Statistical Yearbook, 2009 (1991 ­ 2008 data) Ministry of Health, China Health Statistical Digest, 2010 (1990, 2009 data) 24 Figures 13, 13A and 13B: Maternal, child and infant mortality and GDP per capita: Relationship by province in China and by country internationally 10000 Maternalmortality (per 100,000 live births, log scale) 1000 Countries, 2006 100 China's Provinces, 2006 10 1 300 3000 30000 GDP per capita (US$, log scale) Underfive mortality (per 1,000 of live births, log scale) 350 Countries, 35 2006 China's provinces, 2006 3.5 400 4000 40000 GDP per capita (US$, log scale) Infant mortality (per 1,000 live births, log scale) 100 Countries, 2006 China's provinces, 10 2006 1 300 3000 30000 GDP per capita (US$, log scale) Source: Gapminder (www.gapminder.org) 25 Figures 14, 14A and 14B: Maternal, infant and child mortality and GDP per capita across provinces, 2000 and 2008 500 Provincial maternal Tibet 450 mortality (per 400 100,000 350 live births) 300 250 200 150 2000 100 2008 50 0 1500 15000 Provincial GDP per capita (RMB, log scale) 60 Provincial infant mortality Xinjiang (per 1,000 live 50 births) 40 2000 30 20 10 2008 0 1500 15000 Provincial GDP per capita (RMB, log scale) 70 Provincial under five mortality (per Xinjiang 1,000 live births) 60 2000 50 40 30 20 2008 10 0 2000 20000 Provincial GDP per capita (RMB, log scale) Source: DevInfo (UNICEF), a database system endorsed by the United Nations Development Group for monitoring human development. The mortality data in China used here is gathered through China's official maternal and child mortality surveillance system which until 2008 covered 116 surveillance sites covering 73million people in 31 provinces at the time of the last national census. 26 Figure 15: Falling rural ­ urban ratios in health outcomes and rising urban ­ rural ratio in income per capita, 19912009 3.5 3 IMR rural/urban ratio 2.5 U5MR rural/urban ratio MMR rural/urban rate 2 urban/rural income ratio 1.5 1 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 Source: China Health Statistical Digest 2010 27 Figure 16: Inner Mongolia infant mortality rate (IMR) by prefecture versus GDP, 2000 and 2007 60 Prefecture infant mortality (per 1,000 of live births) 50 40 2000 30 20 10 2007 0 1500 15000 GDP per capita (RMB, log scale) 28 Figure 17: IMR by prefecture within three of China's provinces, 2007 35 30 Prefecture infant mortality (per 1,000 of live births) Xinjiang 25 20 Inner Mongolia 15 10 5 Hebei 0 3000 30000 GDP per capita (RMB, log scale) 29 ANNEX A PUBLIC EXPENDITURE ON HEALTH Figure A1: Public Expenditure on Health per Capita and GDP per Capita Across Prefectures Within Provinces East Jiangsu (province) 2008 Tianjin (province) 2008 (GDP 2007) 500 500 Prefecture health spending per Prefecture health spending 400 400 Nanjing (capital) per capita (RMB) capita (RMB) 300 300 200 200 100 100 0 0 10000 100000 5000 50000 Prefecture GDP per capita (RMB, log scale) Prefecture GDP per capita (RMB, log scale) 2007 Public health spending per capita Public health spending per capita Highest/ Standard Co efficient Highest/ Standard Co efficient M ean Highest Lo west lo west deviatio n o f variance M ean Highest Lo west lo west deviatio n o f variance 185.88 386.76 114.47 3.38 90.61 0.49 234.98 466.45 124.63 3.74 97.16 0.41 Liaoning (province) 2008 Hebei (province) 2007 300 Shenyang (capital) 400 Prefecture health spending Prefecture health spending per Shijiazhuang 250 (capital) per capita (RMB) 300 200 capita (RMB) 150 200 100 100 50 0 10000 100000 8000 Prefecture GDP per capita (RMB, log scale) Prefecture GDP per capita (RMB, log scale) Public health spending per capita Public health spending per capita Highest/ Standard Co efficient Highest/ Standard Co efficient M ean Highest Lo west lo west deviatio n o f variance M ean Highest Lo west lo west deviatio n o f variance 163.14 278.57 95.21 2.93 53.15 0.33 159.53 380.85 22.30 17.08 110.78 0.69 Shandong (province) 2008 Shandong (province) 20002007 250 350 Prefecture health spending Prefecture health spending per Jinan (capital) 300 200 per capita (RMB) 250 150 capita (RMB) 200 Jinan (capital) 150 2007 100 Jinan (capital) 100 50 50 2000 0 8000 80000 1000 10000 100000 Prefecture GDP per capita (RMB, log scale) Prefecture GDP per capita (RMB, log scale) Public health spending per capita Public health spending per capita (2007) Highest/ Standard Co efficient Highest/ Standard Co efficient M ean Highest Lo west lo west deviatio n o f variance M ean Highest Lo west lo west deviatio n o f variance 145.84 208.09 97.78 2.13 31.58 0.22 100.27 313.70 22.10 14.19 69.32 0.69 Public health spending per capita (2000) Highest/ Standard Co efficient M ean Highest Lo west lo west deviatio n o f variance 27.04 120.76 2.32 51.97 27.18 1.01 30 Centre Anhui (province) 2008 Jilin (province) 2008 200 300 Prefecture health spending Prefecture health spending 250 Changchun 150 per capita (RMB) (capital) per capita (RMB) 200 100 150 Hefei (capital) 100 50 50 0 0 5000 50000 10000 Prefecture GDP per capita (RMB, log scale) Prefecture GDP per capita (RMB, log scale) Public health spending per capita Public health spending per capita Highest/ Standard Co efficient Highest/ Standard Co efficient M ean Highest Lo west lo west deviatio n o f variance M ean Highest Lo west lo west deviatio n o f variance 41.70 160.42 5.86 27.36 44.08 1.06 188.53 240.86 137.94 1.75 35.13 0.19 200 Hunan (province) 2008 Changsha 3000 Jiangxi (province) 2008 Prefecture health spending per Prefecture health spending (capital) 2500 150 per capita (RMB) 2000 capita (RMB) 100 1500 1000 Nanchang 50 (capital) 500 0 0 5000 50000 5000 50000 Prefecture GDP per capita (RMB, log scale) Prefecture GDP per capita (RMB, log scale) Public health spending per capita Public health spending per capita Highest/ Standard Co efficient Highest/ Standard Co efficient M ean Highest Lo west lo west deviatio n o f variance M ean Highest Lo west lo west deviatio n o f variance 119.44 169.35 91.22 1.86 22.97 0.19 1053.88 2854.04 247.98 11.51 795.96 0.76 250 Henan (province) 2008 Prefecture health spending per 200 Zhengahou (capital) 150 capita (RMB) 100 50 0 9000 Prefecture GDP per capita (RMB, log scale) Public health spending per capita Highest/ Standard Co efficient M ean Highest Lo west lo west deviatio n o f variance 143.39 234.19 112.44 2.08 28.13 0.20 31 West Chongqing (province) 2008 Guangxi (province) 2008 250 Prefecture health spending 400 Prefecture health spending 200 per capita (RMB) per capita (RMB) 300 150 Nanning (capital) 200 100 50 100 0 0 3000 30000 7500 Prefecture GDP per capita (RMB, log scale) Prefecture GDP per capita (RMB, log scale) Public health spending per capita Public health spending per capita Highest/ Standard Co efficient Highest/ Standard Co efficient M ean Highest Lo west lo west deviatio n o f variance M ean Highest Lo west lo west deviatio n o f variance 162.40 236.65 75.95 3.12 32.17 0.20 179.24 347.26 4.80 72.34 88.62 0.49 Inner Mongolia (province) 2008 Inner Mongolia (province) 20002007 700 500 Prefecture health spending Prefecture health spending 600 400 per capita (RMB) 500 per capita (RMB) 400 300 2007 300 200 200 Houhot (capital) 100 Houhot (capital) 100 2000 0 Houhot (capital) 10000 100000 1500 15000 Prefecture GDP per capita (RMB, log scale) Prefecturde GDP per capita (RMB, log scale) Public health spending per capita Public health spending per capita (2007) Highest/ Standard Co efficient Highest/ Standard Co efficient M ean Highest Lo west lo west deviatio n o f variance M ean Highest Lo west lo west deviatio n o f variance 266.90 601.37 145.14 4.14 129.45 0.49 221.43 453.20 87.69 5.17 122.77 0.55 Public health spending per capita (2000) Highest/ Standard Co efficient M ean Highest Lo west lo west deviatio n o f variance 56.26 122.63 15.73 7.79 38.00 0.68 600 Sichuan (province) 2008* Sichuan (province) 2008** 300 Prefecture health spending Prefecture health spending 500 250 per capita (RMB) per capita (RMB) 400 200 Chengdu (capital) 300 150 200 100 100 Chengdu (capital) 50 5000 50000 Prefecture GDP per capita (RMB, log scale) 5000 Prefecture GDP per capita (RMB, log scale) Public health spending per capita Public health spending per capita Highest/ Standard Co efficient Highest/ Standard Co efficient M ean Highest Lo west lo west deviatio n o f variance M ean Highest Lo west lo west deviatio n o f variance 189.07 189.07 189.07 1.00 103.64 0.55 158.42 254.72 103.21 2.47 39.01 0.25 * The two points within the circled area contain prefectures, Aba ** This graph excludes the two prefectures, Aba (population 894,077)and Ganzi and Ganzi . (population 975,207). Each prefecture attributes to 1% of the total population of 81,331,778. c 1200 Xinjiang (province) 2007 1000 Prefecture health spending 800 per capita (RMB) 600 400 200 Urumqi 0 (capital) 2000 20000 Prefecture GDP per capita (RMB, log scale) Public health spending per capita Highest/ Standard Co efficient M ean Highest Lo west lo west deviatio n o f variance 234.98 466.45 124.63 3.74 97.16 0.41 Source: China Statistical Yearbooks and Provincial Statistical Yearbooks 32 Figure A2: The Relationship Between Public Expenditure on Health per Capita and GDP per Capita Across Prefectures Within Provinces (Trend lines for each selected province across prefectures), for the most recent available year (2007 or 2008) Xinjiang 500 Prefecture health spendign per capita (RMB) 450 400 Inner Mongolia 350 Tianjin Hebei Jiangsu 300 250 Sichuan Jilin Guangxi 200 Liaoning Chongqing Shandong 150 Anhui 100 Sichuan* Henan 50 0 5000 50000 Prefecture GDP per capita (RMB, log scale) * This trend line excludes the two prefectures, Aba (population 894,077) and Ganzi (population 975,207). Each prefecture attributes to 1% of the total population of 81,331,778. Source: China Statistical Yearbooks and Provincial Statistical Yearbooks 33 Figure A3: Public Expenditure on Health per Capita and GDP per Capita Across Subprefectures in Selected Provinces Changsha (prefecture) 2006 Jiangsu (prefecture) 2008 100 600 500 spending per capita (RMB) spending per capita (RMB) 80 Subprefecture health Subprefecture health 400 60 300 40 200 20 100 0 10000 100000 5000 50000 Subprefecture GDP per capita (RMB, log scale) Subprefecture GDP per capita (RMB, log scale) Public health spending per capita Public health spending per capita Highest/ Standard Co efficient Highest/ Standard Co efficient M ean Highest Lo west lo west deviatio n o f variance M ean Highest Lo west lo west deviatio n o f variance 66.40 89.55 40.99 2.18 16.80 0.25 149.34 478.95 70.48 6.80 80.44 0.54 70 Xi'an (prefecture) 2006 Jilin (prefecture) 2008 1000 spending per capita (RMB) spending per capita (RMB) Subprefecture health Subprefecture health 60 800 50 40 600 30 400 20 200 10 0 3000 30000 8000 Subprefecture GDP per capita (RMB, log scale) Subprefecture GDP per capita (RMB, log scale) Public health spending per capita Public health spending per capita Highest/ Standard Co efficient Highest/ Standard Co efficient M ean Highest Lo west lo west deviatio n o f variance M ean Highest Lo west lo west deviatio n o f variance 45.08 66.67 12.36 5.40 19.24 0.43 547.53 852.74 291.16 2.93 157.27 0.29 Jinan (prefecture) 20002006 Zhengzhou (prefecture) 20002006 100 150 spending per capita (RMB) Subprefecture health spending Subprefecture health 80 2006 100 per capita (RMB) 60 2006 40 50 20 2000 2000 0 0 3000 30000 1000 10000 Subprefecture GDP per capita (RMB, log scale) Subprefecture GDP per capita (RMB, log scale) Public health spending per capita (2006) Public health spending per capita (2006) Highest/ Standard Co efficient Highest/ Standard Co efficient M ean Highest Lo west lo west deviatio n o f variance M ean Highest Lo west lo west deviatio n o f variance 61.84 76.87 30.78 2.50 13.62 0.22 45.08 66.67 12.36 5.40 19.24 0.43 Public health spending per capita (2000) Public health spending per capita (2000) Highest/ Standard Co efficient Highest/ Standard Co efficient M ean Highest Lo west lo west deviatio n o f variance M ean Highest Lo west lo west deviatio n o f variance 23.05 33.39 10.82 3.09 6.34 0.28 14.46 30.70 5.29 5.80 8.28 0.57 Source: China Statistical Yearbooks and Provincial Statistical Yearbooks 34 Figure A 4: Public Expenditure on Education per Capita and GDP per Capita Across Prefectures and Subprefectures in Selected Provinces Jiangsu (province) 20012008 Jiangsu (prefecture) 20012008 1500 2000 Prefecture education spending spending per capita (RMB) Subprefecture education 2008 1500 per capita (RMB) 1000 1000 2008 500 500 2001 2001 0 0 5000 50000 2000 20000 200000 Prefecture GDP per capita (RMB, log scale) Subprefecture GDP per capita (RMB, log scale) Education spending per capita (2008) Education spending per capita (2008) Highest/ Standard Co efficient Highest/ Standard Co efficient M ean Highest Lo west lo west deviatio n o f variance M ean Highest Lo west lo west deviatio n o f variance 661.53 1375.04 407.92 3.37 311.09 0.47 548.26 1674.40 272.78 6.14 304.07 0.55 Education spending per capita (2001) Education spending per capita (2001) Highest/ Standard Co efficient Highest/ Standard Co efficient M ean Highest Lo west lo west deviatio n o f variance M ean Highest Lo west lo west deviatio n o f variance 166.63 250.88 100.81 2.49 56.42 0.34 137.69 252.84 73.66 3.43 45.72 0.33 Jilin (province) 20002008 1000 Jilin (prefecture) 20002008 800 Prefecture education spending Subprefecture education spending per capita (RMB) 800 600 2008 per capita (RMB) 600 2008 400 400 200 200 2000 2000 0 0 3000 30000 3000 30000 Prefecture GDP per capita (RMB, log scale) Subprefecture GDP per capita (RMB, log scale) Education spending per capita (2008) Education spending per capita (2008) Highest/ Standard Co efficient Highest/ Standard Co efficient M ean Highest Lo west lo west deviatio n o f variance M ean Highest Lo west lo west deviatio n o f variance 577.03 712.60 430.16 1.66 90.21 0.16 547.53 852.74 291.16 2.93 157.27 0.29 Education spending per capita (2000) Education spending per capita (2000) Highest/ Standard Co efficient Highest/ Standard Co efficient M ean Highest Lo west lo west deviatio n o f variance M ean Highest Lo west lo west deviatio n o f variance 88.88 113.56 71.68 1.58 13.45 0.15 81.85 176.23 51.79 3.40 22.77 0.28 35 Sichuan* (province) 20002008 Sichuan** (province) 20002008 1200 700 Prefecture education spending Prefecture education spending 1000 600 800 500 per capita (RMB) per capita (RMB) 2008 600 400 300 400 2008 200 200 2000 2000 100 0 0 2000 20000 2000 20000 Prefecture GDP per capita (RMB, log scale) Prefecture GDP per capita (RMB, log scale) Education spending per capita (2008) Education spending per capita (2008) Highest/ Standard Co efficient Highest/ Standard Co efficient M ean Highest Lo west lo west deviatio n o f variance M ean Highest Lo west lo west deviatio n o f variance 441.18 441.18 261.91 1.68 197.68 0.45 387.74 387.74 261.91 1.48 96.91 0.25 Education spending per capita (2000) Education spending per capita (2000) Highest/ Standard Co efficient Highest/ Standard Co efficient M ean Highest Lo west lo west deviatio n o f variance M ean Highest Lo west lo west deviatio n o f variance 74.43 186.79 40.06 4.66 35.59 0.48 65.97 125.74 40.06 3.14 22.37 0.34 * The two points within the circled area contain prefectures, Aba ** This graph excludes the two prefectures, Aba (population 894,077)and Ganzi and Ganzi . (population 975,207). Each prefecture attributes to 1% of the total population of 81,331,778. c Liaoning (province) 20002008 1000 spending per capita (RMB) 800 Prefecture education 2008 600 400 200 2000 0 2000 20000 Prefecture GDP per capita (RMB, log scale) Education spending per capita (2008) Highest/ Standard Co efficient M ean Highest Lo west lo west deviatio n o f variance 541.27 867.12 352.55 2.46 186.83 0.35 Education spending per capita (2000) Highest/ Standard Co efficient M ean Highest Lo west lo west deviatio n o f variance 117.39 202.75 67.49 3.00 42.48 0.36 Source: Provincial Statistical Yearbooks 36 ANNEX B: DISPARITIES IN INCOME LEVELS AND HEALTH OUTCOMES Figure B1: The Relationship Between Infant Mortality Rate, Underfive Mortality Rate and Maternal Mortality Rate, and GDP per Capita, 20002008 (Trendlines for each given year across provinces) GDP per capitaIMR Relationship 20002008 60 50 2000 Provincial IMR (per 1000 of live births) 40 2001 2002 2003 30 2004 2005 2006 20 2007 2008 10 0 1000 10000 100000 Provincial GDP per capita (RMB, log scale) GDP per capitaMMR Relationship 20002008 140 Provincial MMR (per 100,000 of live births) 120 100 80 60 40 20 0 2000 2001 2002 2003 2004 2005 2007 3000 30000 2006 2008 Provincial GDP per capita (RMB, log scale) GDP per capitaU5M Relationship 20002008 40 Provincial U5M (per 1000 of live births 35 30 25 20 15 10 5 2008 2000 2001 2002 2004 2006 2007 0 2003 2005 3000 30000 Provincial GDP per capita (RMB, log scale) Source: DevInfo 37 Table B1: Urban/Rural Ratio in Annual Disposable Income Per Capita, 2000 and 2008 Province 2000 2008 Direction of variation Anhui 2.74 3.09 Beijing 2.25 2.32 Chongqing 3.32 3.48 Fujian 2.30 2.90 Gansu 3.44 4.03 Guangdong 2.67 3.08 Guangxi 3.13 3.83 Guizhou 3.73 4.20 Hainan 2.46 2.87 Hebei 2.28 2.80 Heilongjiang 2.29 2.39 Henan 2.40 2.97 Hubei 2.44 2.82 Hunan 2.83 3.06 Inner Mongolia 2.52 3.10 Jiangsu 1.89 2.54 Jiangxi 2.39 2.74 Jilin 2.38 2.60 Liaoning 2.27 2.58 Ningxia 2.85 3.51 Qinghai 3.47 3.80 Shaanxi 3.55 4.10 Shandong 2.44 2.89 Shanghai 2.09 2.33 Shanxi 2.48 2.89 Sichuan 3.10 3.07 Tianjin 2.25 2.46 Tibet 5.58 3.93 Xinjiang 3.49 3.26 Yunnan 4.28 4.27 Zhejiang 2.18 2.45 Source: DevInfo 38 Figure B2: Rural/Urban Ratio in Underfive Mortality, 20002007 Anhui Beijing 6 6 Chongqing Underfive mortality rural/urban ratio (all provinces) Fujian WEST Underfive mortality rural/urban ratio Gansu Xinjiang Guangdong 5 Guangxi 5 Yunnan Guizhou Hainan Sichuan Hebei 4 Heilongjiang Qinghai Henan 4 Hubei Ningxia Hunan 3 Inner Mongolia Inner Mongolia Jiangsu 3 Guizhou Jiangxi 2 Jilin Gansu Liaoning Ningxia Chongqing Qinghai 2 1 Shaanxi Tibet Shandong Shanghai Shaanxi Shanxi 0 Sichuan 1 Guangxi 2000 2001 2002 2003 2004 2005 2006 2007 Tianjin Tibet National Average Xinjiang Yunnan 0 Zhejiang National average 2000 2001 2002 2003 2004 2005 2006 2007 4.5 7 EAST Underfive mortality rural/urban ratio CENTRAL Underfive mortality rural/urban ratio 4 6 3.5 Liaoning Shanxi 5 Hebei 3 Jilin Tianjin Heilongjiang Beijing 2.5 4 Henan Shanghai 2 Anhui Shandong Jiangxi 3 Zhejiang 1.5 Hunan Fujian Hubei Guangdong 1 National average 2 Hainan Jiangsu 0.5 National average 1 0 2000 2001 2002 2003 2004 2005 2006 2007 0 2000 2001 2002 2003 2004 2005 2006 2007 Source: DevInfo Table B2: Rural/Urban Ratio in Underfive Mortality Rate, 20002007 (Descriptive statistics) Underfive mortality Standard Coefficient of Mean Highest Lowest (rural/urban ratio) deviation variation 2000 2.06 0.93 0.45 4.63 0.66 2001 1.92 0.82 0.43 3.86 0.76 2002 2.01 0.95 0.47 5.30 1.08 2003 1.83 0.97 0.53 4.61 0.60 2004 1.80 0.66 0.37 3.52 0.76 2005 1.85 0.94 0.51 4.63 0.96 2006 1.95 1.05 0.54 5.93 0.82 2007 1.75 0.63 0.36 3.73 0.98 Source: DevInfo 39 Table B3: Underfive and Infant Mortality Rates, and Maternal Mortality Ratio, Rural&Urban Areas, 20002007 Standard Coefficient of Highest/ Mean Highest Lowest deviation variation Lowest U5M rural 2000 36.97 17.71 0.48 74.30 (Xinjiang) 5.20 (Beijing) 14.29 2001 32.19 14.62 0.45 63.10 (Sichuan) 7.60 (Beijing) 8.30 2002 32.89 15.50 0.47 62.10 (Xinjiang) 7.70 (Beijing) 8.06 2003 28.71 13.32 0.46 60.00 (Sichuan) 7.20 (Beijing) 8.33 2004 24.82 11.87 0.48 52.90 (Sichuan) 5.80 (Beijing) 9.12 2005 23.04 11.27 0.49 46.60 (Sichuan) 6.10 (Beijing) 7.64 2006 20.06 9.60 0.48 40.10 (Tibet) 5.40 (Tianjin) 7.43 2007 17.96 8.20 0.46 35.45 (Tibet) 5.89 (Beijing) 6.02 U5M urban 2000 18.98 9.60 0.51 42.60 (Sichuan) 6.50 (Shanghai) 6.55 2001 18.08 9.29 0.51 42.60 (Sichuan) 7.80 (Beijing) 5.46 2002 17.53 8.97 0.51 41.70 (Sichuan) 6.60 (Beijing) 6.32 2003 18.06 10.97 0.61 45.10 (Heilongjiang) 5.70 (Beijing) 7.91 2004 14.47 7.82 0.54 39.00 (Heilongjiang) 4.60 (Shanghai) 8.48 2005 13.59 7.51 0.55 39.20 (Sichuan) 4.40 (Shanghai) 8.91 2006 10.87 4.32 0.40 21.59 (Tibet) 4.77 (Shanghai) 4.53 2007 10.73 5.65 0.53 34.36 (Tibet) 3.88 (Shanghai) 8.86 IMR rural 2000 30.32 13.56 0.45 62.40 (Xinjiang) 4.00 (Beijing) 15.60 2001 25.22 10.88 0.43 46.60 (Sichuan) 5.60 (Beijing) 8.32 2002 26.35 11.60 0.44 52.40 (Xinjiang) 5.50 (Beijing) 9.53 2003 23.65 10.40 0.44 43.70 (Sichuan) 5.80 (Beijing) 7.53 2004 20.15 8.87 0.44 35.90 (Guizhou) 4.80 (Beijing) 7.48 2005 18.60 8.09 0.43 31.00 (Guizhou) 4.80 (Beijing) 6.46 2006 16.47 7.66 0.47 30.90 (Shaanxi) 4.50 (Tianjin) 6.87 2007 14.78 6.73 0.46 27.86 (Xinjiang) 4.72 (Beijing) 5.90 IMR urban 2000 15.63 7.59 0.49 34.90 (Xinjiang) 4.80 (Shanghai) 7.27 2001 14.59 7.27 0.50 32.00 (Sichuan) 5.60 (Shanghai) 5.71 2002 14.39 7.16 0.50 31.20 (Sichuan) 5.60 (Beijing) 5.57 2003 14.64 7.96 0.54 34.00 (Sichuan) 5.90 (Beijing) 5.76 2004 11.80 6.13 0.52 30.80 (Heilongjiang) 3.70 (Shanghai) 8.32 2005 10.86 5.11 0.47 24.40 (Sichuan) 3.70 (Shanghai) 6.59 2006 8.71 3.44 0.39 16.30 (Gansu) 2.20 (Hainan) 7.41 2007 8.79 4.54 0.52 26.65 (Tibet) 2.89 (Shanghai) 9.22 MMR rural 2000 86.31 99.65 1.15 531.80 (Tibet) 13.40 (Beijing) 39.69 2001 72.88 71.42 0.98 349.40 (Tibet) 8.40 (Beijing) 41.60 2002 26.35 11.60 0.44 433.7 (Xinjiang) 12.90 (Tianjin) 4.06 2003 74.32 75.33 1.01 404.60 (Tibet) 12.60 (Tianjin) 32.11 2004 64.04 58.72 0.92 321.60 (Tibet) 8.90 (Tianjin) 36.13 2005 60.52 55.58 0.92 307.50 (Tibet) 15.20 (Zhejiang) 20.23 2006 51.87 45.91 0.89 249.50 (Tibet) 5.40 (Tianjin) 46.20 2007 45.43 45.95 1.01 260.03 (Tibet) 2.60 (Tianjin) 100.01 MMR urban 2000 38.86 20.98 0.54 82.90 (Xinjiang) 7.10 (Beijing) 11.68 2001 37.70 19.47 0.52 89.10 (Xinjiang) 13.80 (Beijing) 6.46 2002 33.12 20.21 0.61 84.00 (Xinjiang) 7.30 (Tianjin) 11.51 2003 43.20 48.14 1.11 259.70 (Tibet) 6.90 (Beijing) 37.64 2004 34.33 21.41 0.62 120.30 (Tibet) 5.80 (Shanxi) 20.74 2005 32.77 24.77 0.76 120.10 (Tibet) 1.50 (Shanghai) 80.07 2006 32.66 36.11 1.11 208.90 (Tibet) 7.10 (Beijing) 29.42 2007 32.82 24.34 0.74 140.20 (Tibet) 6.93 (Shanghai) 20.23 Source: DevInfo 40 Figure B3: Coefficient of Variance for Underfive Mortality, Infant Mortality and Maternal Mortality, Rural and Urban, 20002007 1.20 Coefficient of variance 1.10 1.00 0.90 U5M rural 0.80 U5M urban IMR rural 0.70 IMR urban 0.60 MMR rural MMR urban 0.50 0.40 0.30 2000 2001 2002 2003 2004 2005 2006 2007 Source: DevInfo 41