Health sector finance and expenditures in China

PHN Technical Notes GEN 14 This paper is for distribution --within the World Bank Group only. It should not be cited in any external publication. THE WORLD BANK HEALTH SECTOR FINANCE AND EXPENDITURES IN CHINA May 1983 Population, Health and Nutrition Department * This paper is one of a series issued by the Population * * Health and Nutrition Department for the information and * * guidance of Bank staff working in these sectors. The * * views and opinions expressed in this paper do not neces- * * sarily reflect those of the Bank. * GEN 14 ABSTRACT This paper analyses the mobilization and allocation of resources in the Chinese health sector. Financing comes in about equal amounts from three main sources - private outlays (32 percent)., labor insurance (31 percent) and state budget expenditures (30 percent). The residual seven percent is financed by production brigades. The rural collective system, principally brigade and commune health facilities, delivers the largest share of health services, approximately 40 percent; the government system delivers about 32 percent; and the enterprise system, which benefits employees and dependents of state enterprises only, delivers 25 percent. Of all health resource inputs, pharmaceuticals are much the most important, accounting for 58 percent of total expenditure. Western drugs alone consume 49 percent of total health expenditures. A major feature of the Chinese health sector is the high proportion of expenditure mediated through insurance schemes. This reflects the high degree of insurance coverage, which constitutes one of the major achievements of the Chinese system. Only about 30 percent of the population are completely uninsured. A second important feature is the existence of considerable urban - rural differentials in health expenditure. Urban expenditure per capita is triple that in rural areas, principally because of the regressive pattern of public expenditures. Per capita. state subsidies for health in urban areas exceed by a factor of ten those for rural areas. This paper is Supplementary Paper Number 5 to World Bank Report No. 4664-CHA, "The Health Sector in China". Prepared by: Nicholas Prescott, Core Team, WDR 1984 Dean Jamison, PHN Contents, 1 Introduction and Summary..............,****************.......... 1 2. Sources and Methods of Finance...... ........................... 5 2.1 The State Budget.5.........********.................. 5 2.2 Labor Insurance.....uc..................................... 15 2.3 'Rural Cooperative Insurances1......... 2.4 Production Brigades....... ............... 23 2.5 Private Expenditurea2 ...................................... 24 3. Expenditure Aggregates and Strucr ............................. 32 3.1 Expenditures by Resource Input.......... 32 3.2 Expenditure Comparisons, 1957 and 1982. 34 3.3 Allocation Between Delivery Systems......................... 34 3.4 Urban-Rural Differentials..rn...... ......... ....... 38 3.5 Allocation Between Preventive and Curative Medicine......... 41 3.6 Allocation Between Traditional and Western Medicine......... 43 3.7 International Comparisons............................... 44 4. Statistical Annex.. ............................... ....... 46 Text Tables" 1. Ministry of Public Health Expenditure, 1977-81.............. 7 2. Inter-Provincial Differentials in MOPH Expenditure Per Capita, 1979-81........................................ 9 3. Intra-Provincial Differentials in MOPH Expenditure Per Capita, 1981............................................ 9 4. MOPH Expenditure: Budgetary Ratios ..............ds......... 11 5. Expenditures and Membership of the Government Insurance Scheme, 1977-1981..... 0 ....... ........... 13 6. Government Insurance Expenditure Per Primary Member in Selected Areas, 1981 ................. .......... 13 7. Labor Insurance Expenditure Per Primary Member in Selected Areas ................................... ........ 18 8. Average Charges in Selected Hospials....................... 26 9. Health Expenditures in Shanghai County, 1980....... ...... 27 Text Tables (continued) 10. Estimated Private Health Expenditures in Three Counties.,... 28 11. Estimated Distribution of Recurrent Expenditure by Source of Finance, 1981.........*..**................... 30 12. Estimated Distribution of Population by Insurance Status, 1981. . ................... ....... ................. 31 13. Estimated Recurrent Expenditure by Input, 1981.............. 33 14. Drug Expenditure Shares in Selected Facilities............. 35 15. Estimated Recurrent Expenditure by Input, 1957.............. 36 16. Estimated Distribution of Recurrent Expenditure Between Delivery Systems, 1981...e.................. 37 17. Estimated Urban-Rural Differentials in Recurrent Health Expenditure, 1981........ .................... 39 18. Estimated Urban-Rural Differentials in State Subsidies for Recurrent Health Expenditure, 1981 ...................... 40 19. Recurrent Expenditure on Preventive Services, 1981.......... 42 20. Recurrent Expenditure on Traditional Medicine, 1981......... 43 21. International Comparisons of Health Expenditure............. 45 HEALTH SECTOR FINANCE AND EXENDITURES IN CHINA 1. Introduction and Summary This paper analyses the mobilisation and allocation of resources in the Chinese health sector. Total health expenditure in 1981 is estimated at approximately Y15.0 per capita, of which Y14.3 is for recurrent expenditure and YO.6 for capital construction. This represents 3.3 percent of GDP per capita in 1981.1/ Figure 1 shows the results of dividing up recurrent expenditure in three separate ways: by sources of finance, by expenditure on resource inputs and by delivery system. The top frame of Figure 1 shows that financing comes in about equal amounts from three main sources -- private outlays (32 percent), labor insurance (31 percent) and state budget expenditures (30 percent). The residual seven percent is financed by production brigades. The bottom frame of Figure 1 shows that the structure of health services delivery is essentially tripartite. The rural collective system, principally brigade and commune health facilities, delivers the largest share of health services, approximately 40 percent; the government system delivers about 32 percent; and the enterprise system, which benefits enterprise employees and dependents exclusively, delivers 25 percent of total services.-V Private medical practice has recently been authorized, and some 1/ GDP per capita in 1981 was Y455. See IBRD Report No. 4072-CHA, Statistical Appendix Table 2.1 2/ It should be noted that enterprise employees can either receive medical services from facilities operated by the enterprise or they can be referred to county or municipal hospitals that are part of the government delivery system; in either case all expenses are covered by labor insurance. Figure 1 RECURRENT EXPENDITURE ON HEALTH, 1981: SOURCES OF FINANCE, RESOURCE INPUTS AND DELIVERY SYSTEMS SOURCES OF FINANCE A Private Expenditure 32% B Labor Insurance 31% C State Budget 30% D Brigades 7% C RESOURCE INPUTS 0 - E A Western Pharmaceuticals 49% B Salaried Personnel 20% . C Hospital Beds 13% D Traditional Pharmaceuticals 9% E Equipment 5% SSF Brigade Personnel 4% A DELIVERY SYSTEMS A Rural Collective System 40% B Government System 32% C Enterprise System 25% I D Medical Education - 3% PriVate Practice negligible S ries: T eI.s 1I. and16 Sources. Tables 11, 13 and 16. - -3- exists, but it is of negligible .gnificance. Of all resource inputs into the. health sector, shown in the middle frame of Figure 1, pharmaceuticals are by far the most important, accounting for 58 percent of total expenditure. Western drugs alone consume 49 percent of total health expenditure outlays.. A remarkable feature of the financing profile is the high proportion of total expenditure that is mediated through insurance schemes. This, reflects the extremely high degree of health insurance coverage, which constitutes one of the major achievements of the Chinese health system. As shown in Figure 2, it is estimated that only about 30 percent of the population are completely uninsured. Virtually all the uninsured are rural dwellers, and considerable urban-rural differentials in health expenditure exist in China (Figure 3). Urban expenditure is estimated at Y33 per capita, more than triple rural expenditure of Y9 per capita. More striking is the regressive pattern of public expenditures on health. State subsidies for health in urban areas exceed by a factor of over ten those for rural areas - approximately Y26 per capita compared to less than Y3 per capita. Private expenditures per capita amount to less than Y3 in urban areas but over Y5 in 1/ The cost of unpaid labor time, either voluntarily or otherwise mobilized, is not included in the discussion here because virtually no relevant data exist. Anecdotal accounts suggest, however, that substantial amounts of unpaid labor contribute to the success of many preventive efforts, particularly those associated with environmental sanitation. For example, it has been estimated that the economic cost of voluntary labor mobilised to work in the schistosomiasis control campaign in Shanghai County was approximately Y8 million between 1953 and 1978, equivalent to nearly double the financial costs incurred for drugs, surgery, molluscicides and salaried personnel. See Chen et al (1982), "Eradication of Schistoso- miasis", American Journal of Public Health 72 (9), Supplement: 50-51. Figure 2 DISTRIBUTION OF POPULATION BY INSURANCE STATUS, 1981 INSURANCE STATUS A Rural Cooperative Insurance 48% B Uninsured 29% FI C Labor Insurance 12% D Collective Industry Insurance 5% E Commune Industry Insurance 4% F Government Insurance 2% Source: Table 12. Figure 3 URBAN-RURAL DIFFERENTIALS IN RECU72ENT HEALTH EXENDITURE, 1981 Y32.46 per capita STATE SUBSIIES Y26.19 Y9.64 per capita STATE SUBSIDIZS Y2.76 PRIVATE PRIVATE EXENDITURE EXPENDITURE Y2.76 Y5.07 OTHER ___ 5.07____ Y3.51 OTHER Y1.81 UREAN RMRAL Source: Table 17. -5- rural areas. In China, as in every other developing country, the rural majority receives less health care and pays more for it than do their favored urban counterparts. The following sections analyse in detail the available data on health care costs and financing that underly the figures just presented. The most detailed information concerns sources and methods of finance, which are analysed in section 2. More brief analyses of expenditures by resource input, by delivery system and by rural or urban location are then presented in section 3. 2. Sources and Methods of Finance 2.1 The State Budget State budget expenditures on health comprise three types of expenditure undertaken at all levels of government, both central and local: expenditure through the Ministry of Public Health (MOPH) on health services delivery and medical education, expenditure through the Ministry of Finance on the government insurance scheme for government employees and expenditure through the Ministry of Education on medical education in the provincial medical colleges. Ministry of Public Health. State budget expenditures through the MOPH have increased steadily since the 1950s, reaching a total of Y3,881 million in 1981 (Table 1), of which 84 percent was for recurrent expenditure and 16 percent for investment in capital construction. Only about 5 percent of total recurrent expenditure and 10 percent of capital construction is 6- financed out of the central MOPH budget, 1/ the remainder being appropriated through local Bureaus of Public Health at provincial, prefectural and county levels. In nominal terms, per capita expenditure increased by three-quarters between 1977 and 1981, from Y2.22 to Y3.92. In recent years the growth in nominal MOPH expenditure has been especially rapid, equivalent to about 16 percent per year for recurrent and capital expenditure combined between 1977 and 1981, compared to only 6 percent for total state budget expenditures during the same period. The rate of growth in real terms has also been substantial, the implicit Net Material Product deflator having registered an average annual increase of only 2.4 percent between 1977 and 1981. 2/ In 1981 the major portion of Ministry of Health recurrent expenditure, 62 percent, was allocated to the subsidy of curative hospital services (Statistical Annex Table A-2). Hospitals of western medicine claimed a share of 37 percent of total expenditure, hospitals of traditional medicine only three percent and commune health centers 22 percent. Government-owned hospitals, including commune health centers, receive a subsidy equivalent to the total cost of staff salaries and maintenance costs. Commune health centers owned by collectives rather than the government, which comprise about two-thirds of the total 3/, receive only a 60 percent subsidy of salary costs. A further 17 percent of MOPH recurrent expenditure financed the 1/ Central MOPH expenditure is estimated as the residual between aggregate MOPH expenditure and total provincial MOPH expenditure (Statistical Annex Tables 1 and 7 respective3y). 2/ See IBRD Report No. 4072-CHA, Annex 2, Table 2. 3/ Li (1982), "Health care network serves the people", China Reconstructs 31 (10): 31-32. -7- network of preventive services comprising Epidemic Prevention Centers, Maternal and Child Health Centers and Drug Control Centers. Approximately five percent of recurrent expenditure was allocated to the secondary Medical Colleges at provincial level. Expenditures by the central MOPH on the 13 core medical lolleges and their affiliated hospitals, excluding capital construction, amounted to Y135 million in 1981 (Statistical Annex Table A-6), equivalent to four percent of total recurrent expenditure. A similar functional breakdown for investment in capital construction is not available. Table 1 MINISTRY OF PUBLIC HEALTH EXPENDITURE, 1977-1981 Total Expenditure Per Capita Expenditure Recurrent Capital Total Recurrent Capital Total (million yuan) (Yuan) 1977 1,809 273 2,082 1.93 0.29 2.22 1978 2,242 321 2,563 2.36 0.34 2.69 1979 2,602 421 3,023 2.70 0.44 3.13 1980 3,016 571 3,587 3.09 0.56 3.67 1981 3,274 607 3,881 3.31 0.61 3.92 Av. growth rate (%) 14.8 20.0 15.6 13.5 18.6 14.2 Source: Statistical Annex Tables A-1 and A-4. There are very substantial regional differentials in the distribution of MOPH expenditure. Detailed data are given in Statistical Annex Tables A-7 and A-8 and summarised in Table 2. For example, in 1981 recurrent expenditure per capita in Xizang Autonomous Region was seven times greater than in Anhui Province, and nearly five times greater than the provincial average.- The inter-provincial differentials are even more marked for capital expenditure -8- per capita, with a ratio of about 9 to 1 between Tienjin Municipality and Anhui Province, and over 4 to 1 between Tienjin Municipality and the provincial average. In general, per capita expenditures tend to be much higher in the higher income, more heavily urbanised provinces and in the autonomous regions. 1/ There are also significant differentials in per capita expenditure levels within provinces (Table 3). The distribution of recurrent expenditure per capita is also characterised by significant urban bias, with ratios between urban and rural,-areas of almost 3 to I in Ningxia and over 1.5 to 1 in Shandong. Although.some degree of urban bias is a natural consequence of the urban concentration of referral facilities which receive proportionately greater state budget financing, it seems clear that rural residents benefit less from these facilities than their urban counterparts. 2/ I/ Regression analysis of 1981 province-level data indicates that 80 percent of the variance in MOPH recurrent expenditure per capita is explained by variations in urban and rural incomes, urbanisation and autonomous region status. An increase (decrease) of Y100 in urban income per capita is estimated to induce an increase in recurrent expenditure of Y3.1 (equivalent to one-fifth of mean expenditure), while zhe same increase in rural incomes is associated with an increase in expenditure of only Y1.6. Autonomous Region status adds, on average, a further Y1 per capita. With a double-log specification, the elasticity of recurrent expenditure is estimated to be 2.7, almost five times greater than the rural income elasticity which is estimated at 0.6. Qualitatively similar, though quantitatively weaker, results hold for MOPH capital expenditure per capita. For detailed analysis see "Determinants and Consequences of Health -Resource Availability in China", by N. Prescott, D. Jamison and N. Birdsall, Supplementary Paper No. 4 to Health Sector Issues in China, The World Bank, 1983. 2/ Detailed data on the origin of hospital patients are scarce. However, a recent study found that 72 percent of patients in an infectious disease ward of a tertiary care hospital in Wuhan, Hubei Province, were from Wuhan City and suburban areas, although the hospital catchment population was only 10 percent urban. See Henderson and Cohen (1982), "Health Care in the People's Republic of China: A View from Inside the System", American Journal of Public Health 72 (11): 1238-1244. -9- Table 2 INTER-PROVINCIAL DIFFERENTIALS IN MOPH EXPENDITURE PER CAPITA, 1979-81 (Yuan) Recurrent Capital Coefficient Coefficient Year of variation a/ Min Mean b/ Max of variati2n a/ Min Mean b/ Max 1979 0.98 1.88 2.60 12.68 ... ... ... ... 1980 0.87 2.17 2.96 14.20 . .. ... ... 1981 0.92 2.24 3.15 15.33 1.04 0.27 0.55 2.41 Notes: a/ The coefficient of variation is the ratio of the standard .deviation to the population-weighted mean of the distribution. b/ Population-weighted mean of all expenditure within individual provinces, excluding national level expenditure by the central MOPH. Source: Statistical Annex Table A-8. Table 3 INTRA-PROVINCIAL DIFFERENTIALS IN MOPH EXPENDITURE PER CAPITA, 1981 (Yuan) Urban Municipalities Rural Counties Shandong Province 3.50 2.20 Ningxia Autonomous Region 9.20 3.14 Source: Data provided by the Bureaus of Public Health of Shandong Province and Ningxia Autonomous Region. 10 - Although provincial expenditure differentials are very wide, they appear to have narrowed slightly since the fiscal decentralisation reform introduced in 1980. 1/ Prior to 1980, the central government specified the sectoral composition of local government expenditures. The fiscal decentralisation reform delegated the authority for sectoral allocation decisions to provincial or lower levels of government, thus in principle permitting greater inter-provincial variations in expenditure. However, both in the aggregate and in detail, MOPH expenditure has increased relatively rapidly since the decentralisation. Absolute levels of total recurrent and capital expenditure by the MOPH rose by 27 percent in the post-reform years 1980 and 1981, 'While overall state budget expenditure fell by 15 percent, thus increasing the MOPH sector share from 2.4 percent in 1979 to 3.5 percent in 1981 (Table 4). Data on sectoral allocations at provincial level indicate that local trends underlying the net improvement in the sector as a whole have also been very positive in general (Statistical Annex Table A-7). Only Ningxia Autonomous Region experienced a decline in absolute levels of recurrent expenditure on health between 1979 and 1981. The lack of data on total state budget expenditures by province precludes an assessment of relative changes in health sector shares at provincial level. Government Insurance. The Government Insurance scheme, originally introduced in February 1951, provides free outpatient and inpatient health services for life (excluding the cost of hospital meals) to government employees, college teachers and students. In general, only primary members 1/ The coefficient of variation in recurrent expenditure per capita declined from 98 percent to 92 percent between 1979 and 1981 (see Table 2). Table 4 MOPH EXPENDITURE: BUDGETARY RATIOS (Percent) Recurrent Capital Total 1977 3.3 0.9 2.5 1978 3.4 0.7 2.3 1979 3.4 0.8 2.4 1980 3.8 1.4 3.0 1981 4.3 1.9 3.6 Source: Derived from Statistical Annex Table A-1, IBRD Report No. 4072-CHA, Statistical Appendix Table 5.7 and Statistical Yearbook of China: 1981, p. 405. are covered by the scheme, and dependents receive no benefit entitle- ments. 1/ The scheme is financed exclusively by the state budget, with no individual prepayment by its beneficiaries. Both the financing and administration of the scheme are quite independent of the Ministry of Public Health. Expenditures on the scheme are appropriated through the Ministry of Finance and administered by its Committee on Government Insurance. The level of financing within the state budget is decentralised according to the level of employment of the beneficiary. Thus, benefits for government employees at national level are financed out of the central government budget and benefits for those at provincial level are financed out of provincial budgets. It is of interest to record that these insurance benefits were introduced at a time 1/ A recent study in Shanghai County, Shanghai Municipality, reports that some family members receive 50 percent reimbursement of health service costs, but this appears to be unusual. See Ye Xi-Fu et al (1982), "Introduction to Shanghai County", American Journal of Public Health, 72 (9) Supplement: 13-18. 12 - when state cadres were paid only a minimal cash allowance, but have been retained despite the adoption of salary remuneration in 1956. Total state budget outlays on the Government Insurance scheme have increased slightly faster than MOPH recurrent expenditure, at an average annual rate of 15.3 percent during 1977-1981, reaching a.total of Y78 million in 1981 (Table 5). This was equivalent to YO.79 per capita, or equal to about 24 percent of MOPH recurrent expenditure and 20 percent of total MOPH outlays on recurrent and capital expenditure combined. Thus, the Government Insurance scheme exerts a sizeable claim on total fiscal subsidies to the Chinese health sector. There are currently about 18 million primary members of the Government Insurance scheme, or about 1.8 percent of the entire population. Annual expenditure per member averaged Y44 in 1981, representing a reasonably significant transfer to relatively high income individuals, equivalent to five percent of the average annual wage of employees of state-owned units (Y812 in 1981 2/). However, as shown in Table 6, average expenditures in some areas are well below the national average. It is plausible that expenditure levels under the Government Insurance scheme represent a sub-optimal use of health resources. Since government insured beneficiaries face a zero price for utilising health 1/ See Chen (197.6), Population and Health Policy in the People's Republic of China, Occasional Monogr0ph Series No. 9, Interdisciplinary Communications Program, Smithsonian Institution, Washington, D.C. 1/ Statistical Yearbook of China: 1981, p. 431. - 13 - Table 5 EXPENDITURES AND MEMBERSHIP OF THE GOVERNMENT INSURANCE SCHEME, 1977-1981 Expenditure per Expenditure Primary Members primary member (million yuan) (million) (yuan) 1977 423 12.37 34.20 1978 484 13.40 36.12 1979 570 15.08 37.78 1980 668 16.39 40.76 1981 780 17.69 44.09 Av. growth rate (%) 15.3 8.9 6.4 Source: Data provided by the Ministry of Public Health. Table 6 GOVERNMENT INSURANCE EXPENDITURE PER PRIMARY MEMBER IN SELECTED AREAS, 1981 Expenditure per primary member (yuan) Shandong Provincei/ 40.2 - Yexian County 38.3 - Qu Fu County 35.7 - Quingdao Municipality 37.1 Ningxia Autonomous Region2 31.0 Shanghai Municipality b/ - Shanghai County 41.2 Source: a/ Data provided by local Bureaus of Public Health. h/ Hinman & Parker (1982), "Costs of Care", American Journal of Public Health, 72 (9) Supplement: 83-88. - 14 - services (except for food 1/ and travel costs), they have an obvious incentive to utilise services at levels at which the marginal benefits are substantially lower than their marginal costs. The MOPH is aware of the inducement to inefficient utilisation but has no real control over it, since the administration of the government insurance scheme lies outside its authority. For example, in Shandong Province the MOPH attempts indirectly to influence utilisation by offering a Y60,000 incentive payment to counties in which average expenditure per beneficiary does not exceed a Y30 norm. The 'ncentive is weak, however, since the payment is a lump sum independent of the extent of the cost saving, and there are no penalties for exceeding the norm. Only 10 of the 106 counties in the province qualified for the incentive payment in 1981. Ministry of Education. Provincial Bureaus of the Ministry of Education finance the cost of medical education provided in the provincial medical colleges. These appropriations are for the medical colleges only. The state subsidy for their affiliated hospitals is appropriated through the hospital services budget of the provincial Bureaus of Public Health. No data are available on Ministry of Education allocations to the provincial medical colleges, but an indirect estimate can be constructed as follows. There are 106 provincial medical colleges (excluding the 13 core medical colleges financed by the central MOPH), with a total enrollment in 1981 of 116,596 undergraduate and 1,707 graduate students. Assuming a unit cost per student 1/ A recent study reported average food costs at approximately YO.5 per hospital day, equivalent to about 9 percent of average daily charges for drugs and services. See Henderson and Cohen (1982), "Health Care in the People's Republic of China: A View from Inside the System", American Journal of Public Health 72 (11): 1238-1244. - 15 - of Y1,250±1 per year, the total fiscal cost of recurrent expenditure can be estimated at Y148 million per year. This figure excludes the joint cost of running the affiliated teaching hospitals, which is financed by provincial Bureaus of Public Health. 2.2 Labor Insurance Workers and staff employed in state enterprises with more than 100 employees are insured by the compulsory Labor Insurance scheme, which was first introduced in February 1951. Like the Government Insurance scheme, the Labor Insurance scheme entitles primary members to free health care for life. In addition, their dependents are entitled to 50 percent reimbursement of health care costs. There are no official data on the coverage of the Labor Insurance scheme, but a rough estimate can be made. The number of primary members in 1981 can be estimated as the number of staff and workers in state- owned units, 84 million.!/, minus the number of government employees, 18 million (Table 5), that is, a total of 66 million primary members. In addition, it can be estimated that there are about 51 million dependents of primary members, based on the 1:0.77 ratio of employees to dependents reported in the 1981 Household Income and Expenditure Survey of Staff and Workers in 1/ This is approximately equal to the unit recurrent cost in all conventional universities, both central and provincial, financed by the Ministry of Education. It is slightly higher than the Y1,185 unit cost (excluding part-time students) estimated for Quingdao Municipal Medical College in 1981 and substantially lower than the Y2,268 unit cost in the core medical colleges (estimate derived from enrollment data in "Medical Education in China", by J. R. Evans, Table 1, Supplementary Paper No. 6 to Health Sector Issues in China, The World Bank, 1983, and expenditure data in Statistical Annex Table 6). 2/ Statistical Yearbook of China: 1981, p. 106. -16- Urban Areas. 1/ These estimates imply a total coverage in 1981 of 117 million persons including dependents, or about 12 percent of the total population. This estimate is consistent with a previous estimate of 10-12 percent coverage. 2/ In addition to the compulsory Labor Insurance scheme, many county collective, commune and brigade enterprises voluntarily organize insurance schemes for their employees. These schemes generally offer lower reimbursement rates than the compulsory Labor Insurance scheme, typically around 70-80 percent for primary members and partial coverage for dependents. There are no accurate data on the total numbers covered by these schemes, although a maximum estimate for primary members would be the number of staff and workers employed in collective units, about 26 million in 1981., plus the 14 million workers employed in commune enterprises in 1980. - The maximum number of dependents can be estimated at 20 million plus 30 million for the county collective and commune industry schemes respectively, based on the ratios of dependents to wage earners reported in the 1981 urban and rural household surveys. 5 1/ Statistical Yearbook of China: 1981, p. 438. This ratio is higher than the ratio of 52,000 members to 30,000 dependents (1:0.58) estimated for Shanghai County by Ye et al. (1982), "Introduction to Shanghai County", American Journal of Public Health, 72 (9) Supplement: 13-18. 2/ Hu (1981), "Issues of Health Care Financing in the People's Republic of China", Social Science and Medicine, 15C: 233-37. 3/ Statistical Yearbook of China: 1981, p. 106. 4/ IBRD Report No. 4072-CHA, Statistical Appendix Table 7.9. 5/ Statistical Yearbook of China: 1981, p. 439 and p. 441, respectively. - 17 - The financing of the labor insurance funds for health care is treated as an addition to the enterprise wage bill, with no individual prepayment by the employee. There are no precise data on the total expenditure financed by these schemes. However, the MOPH estimates that approximately Y4.4 billion was spent by the compulsory and voluntary schemes (including county collective and commune industries but excluding brigade enterprises) combined in 1981. This estimate includes the cost of health workers and small equipment for health facilities, owned by the enterprises themselves, as well as reimbursement for services provided by government health facilities. It is officially estimated thit Y11.7 billion was spent on state subsidies for free medical services for state employees between 1979 and 1981. -/ Since a total of Y2.018 billion was sperit on Government Insurance during this period (Table 5), it can be assumed that the residual Y9.682 billion was spent on the compulsory Labor Insurance scheme. The cumulative number of primary members of the Labor Insurance scheme can be estimated at 192 million between 1979 and 1981.12 This suggests that annual expenditure per Labor Insurance member averaged approximately Y50 between 1979 and 1981. This is higher than under the Government Insurance scheme, as would be expected since dependents are eligible for 50 percent coinsurance. This, in turn, implies that annual expenditure per primary member of the. voluntary 1/ Beijing Review, No. 43, 25 October 1982, p. 7. 2/ The estimate is derived by subtracting the cumulative total of Government Insurance beneficiaries (Table 5) from the cumulative total of staff and workers employed in state-owned units between 1979 and 1981, as given in the Statistical Yearbook of China: 1981, p. 107. - 18 - county collective and commune industry insurance schemes averages about Y28, / substantially lower than under the compulsory scheme. These indirect estimates are broadly consistent with fragmentary evidence available for Shanghai Municipality and Shandong Province (Table 7). Reported expenditures per beneficiary in these areas range from Y34 to Y42 unde.r the compulsory scheme, from Y29 to Y30 for county collective industries, and from as little as Y12 to as much as Y47 for beneficiaries of commune industry insurance schemes. The generally lower levels and higher variance in the expenditures of the voluntary schemes clearly reflect their voluntary character and dependence on enterprise profitability for financing. Table 7 LABOR INSURANCE EXPENDITURE PER PRIMARY MEMBER IN SELECTED AREAS (Yuan) Labor County Collective Commune Insurance Industry Industry Shanghai Municipality (1980) Shanghai County 42.12 46.61 Shandong Province (1981) Yexian County 39.47 28.80 12.40 Qu Fu County 34.34 31.31 20.00 sources: Hinman & Parker (1982), "Costs of Care", American Journal of Public Health, 72 (9) Supplement: 83-88. b/ Data provided by County Bureaus of Public Health in Shandong Province. 1/ An annual average expenditure of Y50 per primary Labor Insurance beneficiary yields an estimated total expenditure in 1981 of Y3.3 billion for the 66 million primary members. Subtracting this from the estimated total Y4.4 billion spent by the compulsory and voluntary schemes in 1981 implies a residual Y1.1 billion averaged over the 40 million primary members of the county collective and commune industry schemes. -19- 2.3 Rural Cooperative Insurance Implementation of the rural cooperative insurance system began in 1968, and represents a unique example of community health financing in developing countries. This subsection discusses the organisation and expenditures of the rural cooperative health insurance schemes. The following subsection examines related but independent sources of finance at brigade level, particularly for the services of barefoot doctors. Although individual details vary considerably throughout China, the cooperative insurance schemes generally take the form of a pre-paid medical insurance plan, organised at the level of the production brigade. Thus, they provide only a limited pooling of risk and provide no opportunity for redistributive subsidies from richer to poorer brigades. The insurance fund is typically financed jointly by annual prepayments paid by individual members of the brigade and by annual appropriations from the brigade's welfare fund. The services of barefoot doctors in the brigade health stations are financed separately on the basis of workpoint claims on collectively distributed income. In at least 14 provinces, barefoot doctors also receive a modest fiscal subsidy paid out of the provincial health budget 1/, but this is only a very recent innovation designed to prevent the collapse of the cooperative system (see Para 4.20). Beneficiaries enrolled in the cooperative insurance scheme are-generally 1/ The level of this subsidy varies widely. For example, in Ningxia Autonomous Region the government pays Y35-40 per month for doctors with a rural doctor certificate working in mountainous areas, Y30-35 per month for doctors with a rural doctor certificate working in irrigiated areas, Y15-20 per month for barefoot doctors with a barefoot doctor certificate working in mountainous areas, Y1O-15 per month for doctors with a barefoot doctor certificate working in irrigated areas, and Y5-7 per month for barefoot doctors with no certificate. - 20 entitled to free or substantially reimbursable services and drugs at the brigade health station and also at higher level referral units. The operations of the cooperative insurance scheme are well illustrated by recent data for Shanghai County, Shanghai Municipality. -Y In 1979, approximately 90 percent of the rural population in 238 brigades wf-re enrolled in cooperative insurance schemes. The annual individual prepay ,nt averaged Y1.27, over a range from zero (13 percent of brigades) to Y2.00 (28 percent of brigades). The annual appropriation per member paid out of brigade welfare funds averaged Y3.94, varying from Y1.50 (three percent of brigades) to between Y4.00 and Y17.00 (19 percent of brigades). Together, these prepayments averaged Y5.21 per member, of which 24 percent was contributed by individuals and 76 percent by the brigades. This level and pattern of cooperative insurance financing contrasts with that found in many brigades in Shandong Province in which cooperative schemes require no individual prepayment, Into the insurance fund, which instead relies exclusively on annual contributions from the brigade of only Y1.00 to Y3.00 per member. There is substantial inter-brigade variation in the benefit entitlements offered by the cooperative insurance schemes. In general, cooperative insurance outlays appear to be highly income-elastic, indicating that much of the variance in benefit entitlements is explained by variations in income levels. For example, an analysis of grouped data from 116 brigades in Qu Fu County, Shandong Province, indicates an income elasticity of 1/ Chao et al. (1982), "Financing the Cooperative Medical System", American Jour-al of Public Health, 72 (9) Supplement: 78-82. - 21 - insurance outlays of 1.84. 1/ The data for Shanghai County 2/ reveal that the majority of brigades reimbursed 100 percent of the cost of services delivered at the brigade health stations. The lowest reimbursement rate for brigade level services was 50 percent, provided in eight percent of brigades. In addition, a majority of brigades, 56 percent, reimbursed 50 percent of referral costs incurred at higher levels. Reimbursement rates for these costs ranged from 40 percent (one percent of brigades) up to 100 percent (14 percent of brigades). In addition, 32 percent of brigades imposed an absolute limit to the value of costs reimbursed for any one diagnosis. This threshold varied from Y40 to Y100, the modal value lying in the range Y25 to Y100. The coverage of the rural cooperative insurance system reached a peak of 85 percent of brigades in 1975, but has since declined substantially to only 69 percent in 1980 and 58 percent in 1981. 3/ With an average 1,140 persons per brigade in 1981, 4/ this implies a total covezage of approximately 474 million, equivalent to 48 percent of the Chinese population. As a result, nearly half of the rural population has to pay the full cost of health services, with no formal co-insurance to alleviate catastrophic risks. Regional variations in this trend are difficult to assess because of the lack of relevant data, but it is clear that recent experience has been mixed. 1/ "Rural Health Care in China: The Case of Qu Fu County" by J. van der Gaag, Supplementary Paper No. 11 to Health Sector Issues in China, The World Bank, 1983, p. 36. 2/ Chao et al. (1982), "Financing the Cooperative Medical System", American Journal of Public Health, 72 (9) Supplement: 78-80. 3/ See Health Sector Issues in China, The World Bank, 1983, Statistical Annex Table C-14. 4/ Statistical Yearbook of China: 1981, p. 134. -22- Data for Ningxia Autonomous Region 1/ indicate that in Wuzhong County the cooperative insurance schemes have almost completely collapsed, the proportion of brigades with insurance schemes having fallen from 99 percent in 1975 to only one percent in 1982. In Pingluo County, the corresponding proportions fell from 97 percent to 83 percent. In contrast, the proportion of brigades with insurance schemes, currently estimated at around 60 peient, appears to, have been relatively stable in Shandong Province. However, data for 116 brigades in Qu Fu County, in which coverage averaged about 97 percent between 1977 and 1981, suggest that stability in nominal coverage has been maintained at the cost of a sharp reduction in the value of annual prepayments, and hence effective insurance coverage, from an average Y1.82 per member in 1977 'to only Y1.00 in 1981. 2/ The degeneration of the rural cooperative system is attributed primarily to the introduction of the production responsibility system in the rural sector. The weakening of collective organisation in the countryside has impaired the mobilisation of collective resources to finance basic needs. It has also generated a substantial increase in rural incomes, which appears both to have raised the opportunity cost of barefoot doctors' services and also to have stimulated a demand for higher quality health services not perceived to be available through the cooperative system. The extreme variation between brigades in annual insurance prepayments per member precludes an accurate estimationof t e total yalu Of health expenditures mediated through the rural co6perativ& tnurance systeA. 1/ Data provided to the World Bank Health Pro ject ErparAtion Mission, February 1983. 2/ "Rural Health Care in China: The Case of Qu Fu Couttf "by JJ. Van der Gaag, Supplementary Paper No. 11 to Health Sector Issuestin China, The World Bank, 1983, p. 39. A -23- health expendlitures mediaed through the u al cooperative insurance system. A rough estima can be made by assuding that the mid-point of the Y1.00 to Y3 .00 rane pb served in Shandong is representative. This assumption is reasonåb r1p auåible, given that average rural income (as proxied by household expenditrein >handong Prov,ince s more ar less, equal to the national average* With an éstimiated total coverage of about 474 million persons, this implies an aggregate value of tural cooperative health insirance expenditures of Y9 million in 1981. 2.4 Production Briades The tatal value of bArefoot doctors' services, financed independently of the copratve insurance scheme, are also difficult to estimate accurately. The per capita value depends an the density of barefoot doctors per brigade and iher annual wage, both d which vary enormously even within ind"Lvidual 'Qmunes. For example, the annual incoine of 448 barefoot doctors in 12 communes\in Shanghai County, Shanghai Municipality, averaged Y614 in 1980, but rang from lýss than 'Y00 (18 percent) to over Y1000 (2 per- cent). -1, In Shandong Province,imean annual incomes are much lower but vary over an even wider range. The annual income (excluding sideline income) in 1981 of 133 barefoodGetedrs -n Qu Fu County averaged Y347 - in 1981, varying 11 In 1981 `ýral household expenditure per capita in Shandong Province was Y179 compared with Y191 for China as a whole. See Statistical Yearbook of China: 1981, p. 445. 2/ Gong and Chao (1982), "The Role of Barefoot Doctors", American Journal of.. Public IIealýh., 7 (9) upplement: 59-61. 3/ This I nea 14 one-fift,4higher than the Y297 average estimated for Yexia' County, Sh 4tong Provir ce, by the Qounty Health Bureau. - 24 - 24 from only Y108 to as much as Y1044. Assuming that the Qu Fu County data are reasonably representative, it can be estimated that the total value of collective distx buted income paid by individual brigades to the 1.4 million barefoot doctors in China was approximately Y485 million in 1981, equivalent to roughly half of the total outlays of the rural cooperative health insurance system. In addition, part-time rural health aides and midwives are often paid in workpoints. In Yexian County, Shandong Province, annual payments to health aides averaged Y30 in 1981. 2.5 Private Expenditure Consumers of health services face nominal prices at all levels of the Chinese health system, although effective prices net of reimbursements vary with insurance status. At brigade level, users nominally pay a registration fee per visit and the cost of drug -e'Descriptions. In addition, they may also pay a service fee per immunisation, although vaccines themselves are provided free by the state. Service fees are usually quite low at brigade level. For example, in Shandong Province, typical charges for outpatient registration and immunisation are YO.05 each. At higher levels of the health system, in Commune Health Centers and hospitals, patients are nominally liable for the cost of drugs and standardised fees for inpatient and outpatient services. Drugs are generally sold at a markup of 15 percent over uniform national wholesale prices for Western medicines and 30 percent for traditional 1/ "The Barefoot Doctor: Training, Role and Future" by 'M. Young, Supplementary Paper No. 10 to Health Sector Issues in China, The World Bank, 1983. - 25 - medicines. User charges for hospital services follow guidelines established in each province by the Provincial Health Bureau in consultation with the Price Bureau. In Shandong Province, outpatients pay registration fees of YO.1 per first visit and YO.05 per revisit at county level, and YO.15 and YO.10 at prefectural and provincial level hospitals. Inpatients pay a bed-day charge plus surgical fees for operations. Average charges in Shandong Province are Y2.88 per bed-day and Y20, Y10 and Y5 for major, medium and minor categories of operation, respectively. Hospital inpatients also pay for the cost of food. Hospital outpatient charges for drugs and fees are usually quite low, but inpatient charges can be relatively high compared to average incomes. Some examples are given in Table 8. Outpatient charges vary around Y2 to Y3, somewhat lower in Commune Health Centers than in major hospitals. Inpatient charges are much higher, varying from an average Y15 in Commune Health Centers in Qu Fu County, Shandong Province, up to an average Y180 for patients in a tertiary care hospital attached to Hubei Provincial Medical College. In the latter facility, total charges for three inpatients hospitalised with epidemic hemorragic fever averaged as much as Y271. Hospital charges of this magnitude are clearly very high in relation to estimated urban income per capita of Y500 and rural income per capita of Y223 1/ in 1981. 1/ Statistical Yearbook of China: 1981, pp. 438 and 441. 26 - Table 8 AVERAGE CHARGES IN SELECTED HOSPITALS (Yuan) Per outpatient visit Per inpatient admission Min Mean Max Min Mean Max Hospitals MOPH Medical Colleges (1981) a/ 1.91 2.94 3.56 101.77 132.24 156.31 Hubei Provincial M7dical College (1980) ... ... ... 77.03 180.09 270.67 Quingdao Municipal Medical College (1981) c/ ... 2.98 ... ... 119.63 Commune Health Centers Qu Fu Couny, Shandong Province (1981) - 1.22 2.11 3.08 6.86 14.83 23.90 Sources: a/ Derived from Statistical Annex Table 5. b/ Henderson and Cohen (1982), "Health Care in the People's Republic of China: A View from Inside the System", American Journal of Public Health, 72 (11): 1238-1244. c/ Health Sector Issues in Shandong Province, p. 91. d/ "Rural Health Care in China: The Case of Qu Fu County" by J. van der Gaag, Supplementary Paper No. 11 to Health Sector Issues in China, The World Bank, 1983, p. 24. Since a large portion of the Chinese population is either uninsured or faces coinsurance rates .L substantially greater than zero,.2/ it is evident that direct private outlays on health care must be considerable. However, there are no systematic data on private health expenditures in China. The 1981 Sample Survey of Household Income and Expenditure of Peasants provides no information on outlays for health. The 1981 Sample Survey of 1/ The coinsurance rate is the proportion of the price that has to be paid by the patient after deducting the proportion paid by the insurance scheme. 2/ A maximum of only 124 million persons, or 12.5 percent of the total population are fully insured primary members of insurance schemes (Table 11). - 27 - Household Income and Expenditure for Staff and Workers in Urban Areas records an average commodity expenditure on medicines and medical articles equivalent to Y2.76 - per capita per year, but does not identify non-commodity expenditures on health service fees. This represents only 0.6 percent of average total expenditures by urban households, reflecting the fact that most urban residents benefit from enrollment in the Government Insurance or Labor Insurance schemes, which entitle beneficiaries to very low average coinsurance rates. A sample survey in Shanghai County, Shanghai Municipality, indicates a much higher level of personal outlays on health care, averaging Y5.80 per person in 1980 across the different insurance systems (Table 9). Despite the low average coinsurance rate for government-insured beneficiaries, members of the Government Insurance scheme spent the highest amount per capita in direct outlays, about Y9 or nearly one-fifth as much as insured outlays. Table 9 HEALTH EXPENDITURES IN SHANGHAI COUNTY, 1980 Average Annual Expenditure (yuan per capita) Private as Insured Private % insured Government Insurance 41.15 9.10 22.1 Labor Insurance 42.12 3.10 7.4 Commune Industry Insurance 46.61 3.10 6.7 Rural Cooperative Insurance 11.-93 6.70 56.2 Average 18.29 5.80 31.7 Source: Hinman and Parker (1982), "Costs of Care", American Journal of Public Health, 72 (9) Supplement: 83-88. 1 See Statistical Yearbook of China: 1981, p. 439 and pp. 441-450. - 284 -.- -28- Beneficiaries of the Labor Insurance scheme spent the lowest amount at Y3 per capita, less than one-tenth of insured expenditure. On average, members of the rural cooperative insurance schemes spent nearly Y7 per capita. As would be expected from the substantially higher coinsurance rates in these schemes, private outlays were equivalent to around 50 percent of insured expenditure. It would be quite misleading to assume that the levels of private expenditure reported for Shanghai County are representative of China as a whole, since average income levels are much higher than the national average. However, alternative estimates in othr areas are scarce and Table 10 ESTIMATED PRIVATE HEALTH EXPENDITURES IN THREE COUNTIES (Yuan) Total Expenditure Expenditure Population Per Capita Shanghai Municipality - Jiading County (1979) a/ 70,000 490,000 0.14 Shandong Province - Yexian County (1979) a! 393,000 824,000 0.48 - Qu Fu County (1981) b/ 1,210,000 517,000 2.34 Source: a/ Chen (1980), "A Survey of the Medical Cost and Source of Income in Jiading County, Shanghai, and Yexian County, Shandong Province", presented at the WHO/UNICEF workshop on Cost and Financing of Primary Health Care, Geneva, 1-5 December 1980. b/ Estimate provided by the Qu Fu County Health Bureau. 1/ In 1981 rural household expenditure per capita in Shanghai Municipality was Y390, more than double the national average of Y190. See Statistical Yearbook of China: 1981, p. 445. - 29 - unreliable. Some examples are given in Table 10. Levels of YO.14 and YO.48 per year are reported for two counties, which appear to be implausibly low. The average Y2.34 reported by Qu Fu County, Shandong Province, appears to be much more realistic, but the basis for making this estimate is not known. A rough estimate for China as a whole can be made by deriving private expenditure as a residual, after netting out expenditure by all other sources of finance from total health expenditure. Details of this estimate are given in Tables 12 and 13. By this method it is estimated that private expenditure a'Veraged Y4.60 per capita in 1980, of which YO.48 represents individual prepayments into rural cooperative insurance schemes and Y4.12 direct private outlays. Private expenditure accounted for about 30 percent of total recurrent health expenditure in 1981. Summary. The estimated contributions of the major sources of finance for recurrent health expenditures in 1981 are summarised in Table 11. More than half of total recurrent expenditure, 53 percent, is financed through state subsidies paid through the Ministries of Public Health and Education and the compulsory Government and Labor Insurance schemes, much of which benefits urban rather- than rural areas (Table 15). Private expenditure pays for most of the remainder, 32 percent, with the residual 15 percent shared about equally between commune industry insurance and production brigade financing of brigade level health workers and rural cooperative insurance. A remarkable feature of the financing profile shown in Table 11 is the very high proportion, 43 percent, of total expenditure that is mediated through insurance schemes. Although only up to 13 percent of the population - 30 Table 11 ESTIMATED DISTRIBUTION OF RECURRENT EXPENDITURE BY SOURCE OF FINANCE, 1981 Total Per Capita Share (Million juan) (Yuan) (%) Ministry of Public Health a! 3,274 3.31 23.1 Ministry of Education b/ 148 0.15 1.0 Government Insurance a/ 780 0.79 5.5 Labor Insurance c/ - Compulsory 3,300 3.34 23.3 - Voluntary 1,100 1.11 7.7 Production Brigades * Barefoot doctors d/ 485 0.49 3.4 Health aides d/ 60 0.06 0.4 - Rural midwives d/ 18 0.02 0.1 - Cooperative insurance e/ 474 0.48 3.3 Individual - Coopcrative insurance e/ 474 0.48 3.3 - Direct outlays / 4,071 4.12 28.7 TOTAL g/ 14,184 14.34 100.0 Notes: a/ Statistical Annex Table A-1. b/ As estimated in section 2.1. c/ As estimated in section 2.2. d/ As estimated in section 2.3. e/ Assumes average annual prepayments of Y2.00 per member into the rural cooperative insurance scheme, shared equally between brigade welfare funds and individual members. f/ Estimated-as a residual. V As estimated in Table 13. - 31 - Table 12 ESTIMATED DISTRIBUTION OF POPULATION BY INSURANCE STATUS, 1981 Primary members Dependents Total Coverage (millions) Government Insurance a/ 18 ... 18 1.8 Labor Insurance b/ 66 51 117 11.8 Collective Industry Insurance c/ 26 20 46 4.7 Commune Industry Insurance c/ 14 30 44 4.4 Rural Cooperative Insurance d/ ... ... 474 47.9 Other e/ ... ... 290 29.3 TOTAL f/ 989 100.0 Notes: a/ As estimated in section 2.1. b/ As estimated in section 2.2. c/ As estimated in section 2.2. / As estimated in section 2.3. e/ Estimated as a residual. T/ - Estimated mid-year population in 1981. benefits from full insurance coverage 1/ (Table 12), an additional 60 percent receive partial coverage either as dependents of primary members of the Government or the Labor Insurance schemes (both compulsory and voluntary) or as members of rural cooperative insurance schemes. Only about 30 percent of the population are completely uninsured. This degree of health insurance coverage is unparalleled in any other developing country. 1/ Beneficiaries of full insurance coverage are defined as primary members of the Government Insurance, Labor Insurance, collective industry insurance and commune industry insurance schemes. -32- 3. Expenditure Aggregates and Structure 3.1 Expenditures by Resource Input Approximate estimates of the total recurrent costs of the major health sector inputs in 1981 are given in Table 13. Although only approximations, the aggregation of these input costs provides the only available method of estimating total health expenditures in China. The basis of most of the estimates is explained elsewhere in this chapter. Comment here is confined to the two major inputs, salaried health personnel and medicines. The average salary of health personnel is assumed to be equal to the average annual wage of staff and workers in state-owned units in the science, culture, education and public health sector in 1981, Y759 per year. This is probably a reasonably accurate approximation, although in Shandong Province the average salary of state health personnel was slightly lower at Y704 per year in 1981. Detailed data on the average salary paid to different categories of health manpower are not available (salary ranges for Shandong Province are given in Statistical Annex Table A-9). Data on drug production refer to the gross value of output produced by the China National Pharmaceutical Corporation (Western medicines) and the China National Herbal Medicine Corporation (traditional medicines). Although some Western and traditional medicines are produced locally outside of the state enterprise sector, for example by hospitals, commune health centers, and brigade enterprises, there is no direct evidence of its magnitude. Since the 5:1, ratio between the value of Western and traditional medicines shown in Table 13 is broadly consistent with the composition of drug sales in Chinese hospitals, it can be inferred that the value of local production is relatively small. Thus, no additional value is imputed to local production in these estimates. - 33 - Table 13 ESTIMATED RECURRENT EXPENDITURE BY INPUT, 1981 Unit Cost Per Number Cost = Total Cost Capita Share (millions) (Yuan) (million Yuan) (Yuan) (%) Barefoot doctors a/ 1.397 347 485 0.49 3.4 Health aides a/ 2.007 30 60 0.06 0.4 Rural midwives a/ 0.585 30 18 0.02 0.1 Salaried personnel 3.796 b/ 759 c/ 2,881 2.91 20.3 Hospital beds 2.017 b/ 912dc/ 1,840 1.86 13.0 Western medicine e/ ... ... 6,900 f/ 6.98 48.7 Traditional medicine e/ ... ... 1,300 &/ 1.32 9.2 Medical equipment e/ ... ... 700 0.71 4.9 TOTAL 14,184 14.34 100.0 Notes: a/ As estimated in section 2.3. b/ Statistical Annex Table A-5. c/ Assumed equal to the average annual wage of staff and workers in state-owned units in the science, culture, education and public health sector. See Statistical Yearbook of China: 1981, p. 436. d/ Assumes average capital cost per hospital bed of Y10,000 annualised over 20 years at a cost of capital of 7.5 percent per year. e/ Data on the output value of the China National Pharmaceutical Corporation, the China National Herbal Corporation and the China National Medical Equipment Production Corporation provided by the State Pharmaceutical Administration. f/ Assumes a 15 percent retail markup over wholesale prices and that gross output is valued at wholesale prices. / Assumes a 30 percent retail markup over wholesale prices and that gross output is valued at wholesale prices. The value of local production additional to that of state factories operated by the State Pharmaceutical Administration is excluded. -34- The most striking feature of the input structure revealed in Table 13 is the very high proportion of total costs contributed by drug consumption. In aggregate, 58 percent of total health"expenditures are attributed to drug costs, with Western medicines alone accounting for almost 50 percent of the total. A similar pattern is evident in the cost structure of individual facilities (Table 14). Although comparisons with other countries are complicated by relative price differentials, 1 the high drug consumption share of total expenditure does suggests an inefficient use of drugs which is consistent with impressions by physicians of widespread over-prescription of drugs in the Chinese health system. 3.2 Expenditure Comparisons, 1957 and 1981 Table 15 presents a rough estimate of total recurrent health . expenditure in 1957, constructed similarly to the 1981 estimate shown in Table 13. Total recurrent expenditure in 1957 is estimated at Y2.86 per capita at 1981 prices, indicating an average annual growth rate of 6.7 percent in per capita expenditure between 1957 and 1981. This rate of growth was relatively rapid compared t&j;rerall per capita GDP growth, yielding a gradual increase in the health sector scheme of GDP. 3.3 Allocation Between Delivery Systems Rough estimates of the distribution of recurrent expenditures on health between different service delivery systems are shown in Table 16. The 1/ For example, wages paid to Chinese health personnel are relaEively low by international standards. Thus, labor probably receives a lower weight and other inputs, especially drugs, a correspondingly higher weight in aggregate expenditure than in other countries. -35- Table 14 DRUG EXPENDITURE SHARES IN SELECTED FACILITIES % of recurrent expenditure Shanghai County, Shanghai 1Iuicipality / b/ County Central Hospital (1980) 40 Qui-Yi Commune Health Center (1980) 44 Hong-Quia Commune Health Center (1980) 53 Cooperative Insurance Funds (1979) 50 Yexian County, Shandong Provincec/ Xu Jia Commune Health Center (1981) 50 Xiao Han Brigade Health Station (1981) 68 Qu Fu County, Shandong Provinced/ Commune Health Centers (1981) 55 Sources: a/ Hinman & Parker (1982), "Costs of Care", American Journal of Public Health, 72 (9) Supplement: 83-88. b/ Chao et al. (1982), "Financing the Cooperative Medical System", American Journal of Public Health, 72 (9) Supplement: 78-80. c/ WHO/UNICEF/IBRD/PRC Interregional Seminar on Primary Health Care, Yexian County, Shandong Province, 13-26 June 1982. d/ "Rural Health Care in China: The Case of Qu Fu County", by J. van der Gaag, Supplementary Paper No. 11 to Health Sector Issues in China, The World Bank, 1983. Q -36 Table 15 ESTIMATED RECURRENT EXPENDITURE ,Y I¶T 1957 (at 1981 prices) Unit Cos yer Number Cost Total Cost4 Capa Share (millions) (Yuan) (Mil on Yua ) (Yuan) (%) Salaried Personnel 1.254 759 952 1.47 51K3 Hospital Beds 0.295 912 269 0.42 1 3 Pharmateuticals ' --- --- 636 0.98 34.3 TOTAL 1,857 2.86 00.0 Notes: a/ Unit costs valued at 1981 prices, as given in Tabl, 13 b/ Stzkistical Yearbook of China: 1981, p., 477. c/ Statistical Yearbook of China: 1981, p. 476. d/ Includes medical equipment. The 1957 total for Tmedical equipment and western and traditional medicines equals one-fourteenth the 1981 total estimated in Table 13. I;I Table 16 ESTIMATED DISTRIBUTION OF RECURRENT EXPENDITURE BETWEEN DELIVERY SYSTEMS, 1981 Expeåditure Share C ~(million yuan)() Medical Education - KOPR Coie Colleges 64 - HOPH Secondary Medical, School& 170 - Provincial Medical Collegéý 148 Sub-total 382 2.7 Government Health System - ~7ersonyel -b. 811 - Beds 779 - Druga and Equipment d 2934 Sub-total 4528 31.9 n-rpwis,,alth System e .. 8i6 -Beeds 365 Dtugs and Equipment '2290 Sub-tot4l 3521 24.8 Collective Health System - Perso l 1371 - Beds 696 Drugs and Equipment 3625 Sub-total 5692 40.1 Private Realth Syste= - Peraonnel 14 Drugs and Equipment f. 47 Sub-total 61 0.4 T 14184 100.0 Notts. al Expenditures on the MOPH core colleges and secondary medical schools are fr= Statistical Annex Tables 6 and 2 respectively. Expenditures on the provincial medical colleges. re as estimated In section 2.2. b/ Derived as a residual between total expenditure as estimated in Table 13 and all other expenditure categories. c/ umber of hospital beds derived from <atistical Annex Table 5 and valued as explained in Table 13, note d. d/ Derived as a residual between total expenditures on drugs and egpbent as"st-iated in Table 13 and the sum of estimated expenditurt4f on drugs and equipment in the enterprise, collective and private health systema: e/ Estimated as the number of salaried health personnel giyen in Statistical Annex Tabl& 5 multiplied by the average annual wage of salaried health personnel assumed in Table 13. fl Derived by multiplying expenditures on personnel by the aggregate ratio of expenditure on drugs and equipment to expenditure on salaried health workars. ?luä Särefoot doctors estimated in Table 13., For the collectivdéhealth system the personnel expenditure. base includer personnel in commune health centers plus barefoot doctors only. Includes expenditure on salaried health workers in cammune health centers and expenditures on barefoot doctors, rural health aides 4nd midwives,' an estimated in Table 13. Expenditure on salaried bHelth workeys i commune health centers is estima,ted by äoltiplying thei úmber of personnel given in Statistical Annex Table 8 by the average annual wage assumed in Table 13. hl Statistical Annex Table 5. As eatiated in Table 13. * . - 4 1 Y - 38 - collective health system comprising brigade and commune level services for the rural population is the major delivery mechanism, accounting for about 40 percent of total expenditure. The government system, which serves all segments of the population, is the second largest with a 32 percent share of total 4expenditure. Most surprising is the importance of the enterprise health system which benefits exclusively the 21 percent of the population which are primary or dependent beneficiaries of the compulsory and voluntary Labor Insurance schemes, and which consumes about 25 percent of total health expenditure. This of course understates the benefits accruing to labor insurance beneficiaries, since they also utilise government health facilities. Approximately two-thirds of total labor insurance expenditure finances the enterprise health system and one-third the utilisation by its beneficiaries of the government system. Although private health services do exist they are relatively insignificant, accounting for less than one percent of total expenditure. 3.4 Urban-Rural Differentials There are very substantial urban-rural differentials in health expenditure in China, which are significantly exacerbated by public expenditure patterns. Approximate estimates of the urban-rural distributions of total expenditure and its state subsidy component are given in Tables 17 and 18. These suggest that per capita expenditure is more than three times as -39- Table 17 ESTIMATED URBAN-RURAL DIFFERENTIALS IN RECURRENT HEALTH EXPENDITURE, 1981 Total Per Capita h/ Expenditure Expenditure Share (million yuan) (yuan) URBAN State Subsidies a/ 5,337 26.19 81 County Collective Industry Insurance b/ 715 3.51 11. Privata Expenditure c7 562 2.76 .9 Sub-total 6,614 32.46 100 RURAL State Subsidies a/ 2,164 2.76 29 Commune Industry Insurance d/ 385 0.49 5 Production brigades e/ 1,037 1.32 14 Private Expenditure f/ 3,982 5.07 53 Sub-total f/ _7,569 9.64 100 TOTAL Aj 14,184 14.34 Notes: a/ As estimated in Table 18. b/ Estimated assuming average insurance outlays of Y27.50 per primary member (see section 2.2) and 26 million primary members (see section 2.1). c/ Estimate based on the 1981 Sample Survey of Household Income and Expenditure of Staff and Workers in Urban Areas (see section 2.5). d/ Estimated assuming average insurance outlays of Y27.50 per primary member (see section 2.2) and 14 million primary members (see section 2.1). e/ As estimated in Table 12. f/ Derived as a residual. / As estimated in Table 13. h/ Urban and rural population estimated from the 1982 Census distri- bution (20.6 percent urban) and th estimated 1981 mid-year population (Statistical Annex Table A-1). -40 Table 18 ESMATED URBAN-RURAL DIFFERENTIALS IN STATE SUBSIDIES FOR RECURRENT HEALTH EXPENDITURE, 1981 (million yuan) Urban Rural Total Ministry of Public Healtha/ Hospitals b 895 298 1,193 Hospitals of Traditional Medicine / 77 25 103 Commune Health Centers 721 721 Anti-Epidemic Activities c/ 97 377 475 Drug Control Institutes N 7 28 35 Maternal and Child Health c 14 54 68 Middle Level Training Scho ls c 34 134 169 Scientific Research N 23 90 114 Cooperative Health Services 23 23 Other N 75 292 368 Ministry of Education N 30 117 148 Government Insurancel/ 780 ... 780 Labor Insurance - 3,300 ... 3,300 TOTAL 5,337 2,164 7,501 Per Capita N Expenditure (Yuan) 26.19 2.76 7.58 Notes: a/ Statistical Annex Table E-2. b/ Allocated 75 percent urban and 25 percent rural. c/ Allocated 20.6 percent urban and 79.4 percent rural (the 1982 Census distribution). d/ As estimated in section 2.1. e/ As estimated in Table 5. f/ Refers to compulsory Labor Insurance only. As estimated in section 2.2. -41- high in urban than in rural areas, approximately Y33 compared with Y10. 1/ The substantially regressive urban bias of public expenditure is chiefly respons.ible for this wide disparity. State subsidies per capita are nearly ten times greater for urban residents than for rural residents. At Y26 per capita in urban areas, the state subsidy for health expenditure is equivalent to five percent of urban income per capita. In stark contrast, the state subsidy to rural residents averages less than Y3 per capita, or only one percent of average rural income per capita. 2/ In urban areas, state subsidies finance more than three-quarters of total health expenditure and private outlays contribute less than one-tenth of the total. In rural areas, private expenditure pays for more than half of total outlays and state subsidies contribute less than one-third. Thus in China, as in every other developing country, the rural majority receives less health care and pays more for it than their advantaged urban-counterparts. 3.5 Allocation Between Preventive and Curative Services An estimate of the volume of resources allocated to preventive interventions is given in Table 19. The major component, 86 percent, 1/ This estimate for rural expenditure is very close to a recent estimate of per capita health expenditure in a rural brigade in Yexian County, Shandong Province. See WHO/UNICEF/IBRD/PRC Interregional Seminar on Primary Health Care, Yexian County, Shandong Province, 13-26 June 1982. 2/ Urban and rural incomes per capita averaged Y500 and Y223 respectively in 1981. See Statistical Yearbook of China: 1981, pp. 438 and 441. -42 comprises MOPH expenditure on the publicly financed preventive network of Epidemic-Prevention, Drug Control and MCH Centers. The rest is accounted for by an imputed allocation of 20 percent of barefoot doctors' time to preventive work. Although the latter seems to be low in light of the stated policy emphasis on preventive work in rural areas, it is consistent with results obtained in recent studies of the allocation of time by barefoot doctors. Overall preventive interventions consume only about 5 percent of total recurrent health expenditures, equivalent to YO.69 per capita. Table 19 RECURRENT EXPENDITURE ON PREVENTIVE SERVICES, 1981 (million Yuan) Epidemic-Prevention Centers a/ 475 Drdg Control Centers a/ 36 MCH Centers a! 68 Barefoot Doctors b/ 97 TOTAL 676 Notes: 'a a Statistial Annex Table A-2a b/ Imputed at 20 percent of total value of barefoot doctors' services, as estimated in Table 13. 1/ See "The ,3arefoot Doctor: Training, Role and Future", by M. Young, Supplementary Paper No. 10 to Health Sector Issues in China, The World Bank, 1983, and Gong & Chao (1982), "The Role of Barefoot Doctors", American Journal of Public Health, 72(9) Supplement: 59-61. Young's tentative findings concerning barefoot doctors in Qu Fu County, Shandong ?rovince, are that substantially less than 20 percent of their time is currently spent in preventive work. -43- 3.6 Allocation Between Traditional and Western Medicine Expenditures on Chinese traditional medicine represent a relatively small, although still significant, proportion of total health expenditures. Estimates given in Table 20 suggest that traditional medicine accounts for about 14 percent of total recurrent health expenditure in 1981, or about Y2 per capita. This may be a slight underestiLate, since the valuation of hospital beds relates to beds in hospitals of traditional medicine only, excluding beds used for traditional medicine patients in non-specialized facilities. Table 20 RECURRENT EXPENDITURE ON TRADITIONAL MEDICINE, 1981 Number Unit Cost a/ Total Cost (millions) (Yuan) (million Yuan) Salaried Personnel 2.133 b/ 759 673 Hospital Beds 0.886 c/ 912 . 53 Traditional Medicine --- --- 1,300 -a/ TOTAL 2,026 Notes: a/ As estimated in Table 13. b/ Assumes that the 31:100 ratio of traditional to western doctors (Health Sector Issues in China, Statistical Annex Table C-3, The World Bank, 1983) characterizes the distribution of all salaried personnel. c/ Health Sector Issues in China, Statistical Annex Table C-9, The World Bank, 1983. -44- 3.7 International Comparisons Available data suggest that per capita health expenditure in China is relatively high for a country at its low per capita income level. Comparisons with 19 LDCs with 1975 per capita GDP lower than US$2000 are presented in Table 21. For these LDCs, the relationship between per capita health expenditure and per capita GDP is almost perfectly log-linear, with an estimated income elasticity of health expenditure of approximately 1.2. At constant 1975 prices, per capita health expenditure in China is estimated at US$7.1, which is 25 percent greater than the US$5.7 per capita expenditure level predicted from the LDC sample. -45- Table 21: INTERNATIONAL COMPARISONS OF HEALTH EXPENDITURE Health Expenditure GDP Health Expenditure LDC Sample (1975) a/ per capita per capita share of GDP (US$) (US$) (%) Malawi 2.1 138 1.5 Kenya 8.9 241 3.7 India 3.4 146 2.3 Pakistan 7.2 189 3.8 Sri Lanka 4.1 185 2.2 Zambia 16.6 494 3.3 Thailand 14.8 350 4.2 Philippines 8.7 376 2.3 Korea 16.3 583 2.8 Dlalaysi 21.1 781 2.7 Colombia 23.2 564 4.1 Jamaica 43.9 1406 3.1 Syria 16.5 718 2.3 Brazil 44.2 1149 3.8 Romania 48.0 1742 2.8 Mexico 64.3 1465 4.4 Yugoslavia 72.6 1663 4.4 Iran 48.4 1587 3.1 Uruguay 66.3, 1308 5.1 Mean 27.93 794 3.3 China (1981) Actual 7.1 214 3.3 Predieted 5.7 c/ Notes: a/ Derived from Kravis et al (1982), World Produet and Income: International,Comparisons of Real Gross Product, Baltimore: Johns Hopkins University Press, Summary Multilateral Tables 6-1 and 6-3. b/ World Bank estimates of 1981 values at 1975 constant dollar prices. c/ Prediction based on the following OLS equation estimated for the LDC sample: LN HLTHEXP = -4.525 + 1.166 LNGDP, R2 0.936. T-statistics a!re given in parentheses. -46- Statistical Annex Table A-I State Budget Expenditure on Health, 1957-81 Table A-2 Allocation of Recurrent State Budget Expenditure on Health, 1981 Table A-3 Health Sector Budgetary Ratios, 1957-80 Table A-4 Per Capita, ritate Expenditure on Health, 1957-81 Table A-5 Costs and Revenues of Medical College kffiliated Hospitals, 1980-81 Table A-6 MOPH Subsidies to Core Medical Colleges, 19O-82 -8 Table A-7 Provincial Distribution of State Health Expenditures, 1979-81 Table A-8 Provincial Distribution of State Health Expenditures Per Capita, 1979-81 Table A-9 Salaries of County-Level Health Personnel, Shandong Province, 1982 - 47 Table A-l: STATE BUDGET EXPENDITURE ON HEALTH, 1957-1981 (.Billion Yuan) Recurrent Capital Total Central Local Total Healtha/ Government Total Government Government Insurance 1957 0.482 N.A. N.A. 0.069 N.A. N.A. 0.551 1965 0.929 N.A. N.A. 0.087 N.A. N.A. 1.016 1970 1.047 N.A. N.A. N.A. N.A. N.A. N.A. 1975 1.992 N.A. N.A. 0.178 N1.A. N.A. 2.170 1976 2.095 N.A. N.A. 0.186 N.A. N.A. 2.281 1977 2.232 1.809 0.423 0.273 0.158 0.115 2.505 1978 2.726 2.242 0.484 0.321 0.195 0.126 3.047 1979 3.172 2.602 0.570 0.421 0.236 0.185 3.593 1980 3.684 3.016 0.660 0.571 0.263 0308 4.255 1981 4.054 3.274 0.780 0.607 0.280 0.327 4.661 / ncludes recurrent expenditure on health by the Central Ministry of Public Health and local Bureaus of Health. Sources: 1957-1976: China: Socialist Economic Development, World Bank Report No. 4072-CHA, June, 1981, Statistical Appendix," Table 5.7. 1977-1981: Data provided to the World Bank October 1982 Rural Health and Medical Education Mission by the Ministry of Public Health. 0 -48- Table A-2: ALLOCATION OF RECURRENT STATE BUDGET EXPENDITURE ON HEALTE, 1981 Yuan Percent (Millions) of Total Hospitals 1,193.99 36.47 Hospitals of Traditional 103.59 316 Chinese Medicine Subsidy to Commune 721.44 22.04 - Health Centers Anti-Epidemic Activities 475.33 14.52 Drug Control Institutes 35.86 1.10 Maternal and Child Health 68.30 2.09 Middle Level Training 169.68 5.18 Schools Scientific Health 114.09 3.49 Cooperative Health 23.55 0.72 Servicesal Other 368.09 11.24 Total 3,273.92 100.00 a/ This provides a subsidy for the training of barefoot doctors but excludes salary subsidies. Source: Data provided to the October 1982 Rural Health and Medical Education World Bank misslon by the Ministry of Health.' ____________________________________________ *. - ~ *) - 49 - Table A-3: EALTH SECTOR BUDGETARY R cIOS, 1957-1980 Percent of Total State Budget Expenditure 1/ Year Recurrent Capital Total 1957 2.7 0.6 1.8 1965 3.0 0.6 2.2 1970 3.0 NA. N.A. 1975 4.0 0.5 2.6 1976 4.2 0.6 2.8 1977 4.1 0.9 3.0 1978 4.1 0.7 2.7 1979 4.2 0.8 2.8 1980 4.8 1.6 3.7 1/ Demominators are recurrent, capital and total state budget expenditures respectively. Sources: Table E-1 and China; Socialist Economic Development. World Bank Report No. 4072-CHA, June 1981 Statistical Appendix, Table 5.7. -`7- 50 -0 Table A-4: PER CAPITA STATE EXPENDITURE CN HEALTH 1957-1981 (Yuan) Year Recurrent Capital Total 1957 0.76 0.11 0.87 1965 1.30 t 0.12 0.42 1970 1.29 ..A. N.A. 1975 2.19 0.20 2.39 1976 2.26 0.20 2.46 1977 2.38 0.29 2.67 1978 2.87 0.34 3.21 1979 3.29 0-44- 3.73 1980 3.77 0.59 4.36 1981 4.10 0.61 4.71 Source: Data provided to the October, 1982, World Bank Rural Health and Medical Education mission by the Ministry of Public He4lth. l Table Å.ý6- EXPENDITURES OM MEDICAL COLLECES BY TRE MINISTRY OV PUBLIC HEALTH, 1980-1982 (In thousands of Yuan) Medical Educatlon AffIllated Hospital Capital Contribution 19,80 198A 1982 1980 1981 1982 1980 1981 1982 DeijIng Traditional Medical 2480 3000 3490 1,620 2470 3830 3230 2830 2610 College Cuangzhou Traditional Medical 2440 2620 3430 225Q 2580 2660 1525 8750 1580 Collage Beljing.ýedlcal College 5670 6150 7600 7600 13530 16960 7880 8611.0 MO Shanghal. Pirat Medical College 5570 5800 7600 8950 10950 11200 4060 2500 3110 Zbangsban. Medical College 1330 5800 79,10 5900 7370 7800 24w 31.50 2060 Shandong bledical College 3000 3930 4650 2170 3140 3290 2060 1840 1400 Hitnan Medlcal College 3320 3710, 4300 5200 6440 - 6840 1760 1495 1510 ÄA wuhaft kledteal College 4170 4670 5160 4620 5350 5670 2420 2230 -2050(ý Sicliuan Medical College 4990 5580 6730 4850 6220 6520 2240 3010 2500 Xlan bledlcal, College 3410 3310 4460 4600 5730 6050 2470 2010. 1990> (,Sikaanxl Province) Slieiiyajig Me 3860 4160 5010 6160 7420 7840 3210 3540 3180 Cittna Betliktne Medlcal College 4070 4660 5370 6050 7110 7540 392Ö 2940 2170'1 Capital Medical College 1160 1600 240 --- HA Nå of'Cltina Täl,tal' 49,740 55,410 64,400 5,9870 7,9280 _ý8,8200 37,225 35,435 32DO10 Sdurce: Data provided to the October, 1982, World Bank Rural Health and Medical Education mission by the Miftlotry of PubItc Health. -53 Table A-7: PROVINCIAL DISTRIBUTION OF STATE HEALTH EXPENDITURES, 1979-81 (in millious of yuan) Recurrent Government Insurance Capital Construction llrovinces 1979 1980 1981 1979 1980 1981 1979 1980 1981 NATSONAL TOTA. 2,506.17 2,889.30 3,118.22 570.35 667.85 785.33 NA NÅ 546.29 North Region Beijing 60.10 64.64 76.11 22.79 21.52 32.26 NÅ NÅ 16.21 Tienjin 46.57 51.94, 66.40 11.25 13.46 15.66 NÅ NÅ 18.21 febei 104.37 117.67 124.05 29.00 34.78 40.17 NÅ NÅ 25.28 Shanxi 65.51 73.19 74.12 16.94 19.28 22.11 NÅ NÅ 11.88 Nei Monggol 72.64 75.81 81.60 17.04 18.09 22.34 NA NÅ 8.11 Northeast Region .iaoning 121.33 144.24 150.05 24.79 27.70 32.84 NÅ NA 45.37 Jlin 80.18 94.33 107.82 13.62 15.64 18.55 NÅ NA 8.89 Heilongjiang 100.56 120.92 128.08 20.07 22.95 26.47 NÅ NÅ, 18.49 East Region Shanghai 71.01 87.54 103.30 17.99 19.90 21.59 NÅ NÅ 11.20 Jiangsu 128.29 147.87 162.71 24.06 28.86 32.39 NA NÅ 28.03 ZheJiang 88.69 109.05 122.19 18.78 22.14 ý24.62 NA NÅ 14.68 Anhui 97.64 110.70 110.00 19.07 22.03 24.14 NÅ NÅ 13.04 Fujian 66.92 76.30 91.67 14.84 16.61 19.78 NÅ NÅ 17.81 Jiangxi 77.89 86.29 92.59 14.15 18.00 22.46 NÅ NÅ 12.57 Shandong 174.54 201.75 210.61 30.40 36.60 4i.61 NA NÅ 26.93 Central South Region Henan 140.25 160.03 173.15 48.18- 55.80 62.53 NÅ NÅ 24.98 Rubei 127.43 157.76 163.03 33.53 40.53 53.66 NÅ NA NÅ Runan 111.94 134.15 139.20 23.15 28.91 32.58 NÅ NÅ 22.39 Guangdong . 125.87 150.19 169.68 28.90 35.75 46.74 NÅ NÅ 48.81 Cuangxi 77.20 81.98 89.46 17.95 21.35 26.14 Nk NÅ 15.69 Southwest Region Siehuan 182.75 212.63 224.96 47.22 53.28 60.72 NÅ NÅ 35.92 Guizhou - 61.27 63.11 70.55 13.16 14.37 17.32 NA NÅ 8.80 Tunnan 70.69 77.90 86.79 16.74 19.19 21.79 NÅ NÅ 18.74 Xizang 23.07 26.12 28.67 1.03 8.90 9.40 NÅ NA 2.93 Northwest Region Shaanxi 76.22 89.88 97.58 19.34 22.88 28.01 NÅ NÅ 16.46 Ca3u 52.91 56.27 56.81 11.04 12.66 13.95 NA NÅ 11.05 Qinghai 25.34 26.67 27.75 3.90 4.89 6.17 VA NÅ' 6.92 Ningxia 20.43 20.32 20.61 2.93 3.36 4.17 NÅ NÅ 6.32 Xnjiang 53.90 64.05 68.08 7.89' 10.28 12.62 NÅ NA 11.24 Source: The provincial expenditure data were provided to the Rural Health and Medcal Zducation mission by the Ministry of Public Heal:h. Population data used i from Table A-6. 54 Table A-8: PROVINCIAL DISTRIBUTION OF STATE HEALTH EXPENDITURES PER CAPITA, 1979-81 (in yuan) Recurrenc Covernaenc Insurance CapitaI ConstructLon Provinces 1979 1980 1981 1979 1980 1981 1979 1980 1981 NATIONAL TOTAL 2.60 2.96 3.15 0.59 0.68 0.79 NA Å Å 0.55 North Region selijing 6.98 7.34 8.44 2.65 2.44 3.58 NÅ NA 1.80 Tienjin 6.47 7.04 8.77 1.56 1.82 2.07 MA NK 2.4t Hebei 2.06 2.29 2.38 0.57 0.68 0.77 NÅ NÅ 0.48 Shanxi 2.69 2.97 2.97 0.70 0.78 0.89 NÅ NÅ 0.48 Nei Monggol ...a/ 4.06 4.30 ..8/ 0.97 1.18 NÅ NA 0.43 Northeast Region Liaoning .0../ 4.16 4.26 ...a1 0.80 0.93 NÅ MA 1.29 Jilin .../ 4.29 4.84 ... / 0.71 0.83 NA NA 0040 -eilongjiang ... / 3.79 3.97 ...i 0.72 0.82 NÅ NÅ 0.57 East Region Shanghal 6.33 7.66 8.87' 1.60 1.74 1.85 NA NÅ 0.96 -Jiangsu 2.19 2.50 2.72* 0.41 0.49- 0.54 NÅ NA 0.47 Zhejiang 2.35 2.86 3.17 0.50 0.58 0.64 NÅ NA 0,38 Anhui 2.04 2.29 2.24 0.40 0.46 0.49 NA NA 0.27 Fujian 2.72 3.05 3.60 0.60 0.66 0.78 NÅ NÅ 0.70 Jiangxi 2.43 . 2.66 2.82 0.44 0.55 0.68 NÅ NA 0.38 Shandong 2.43 2.77 2.86 0.42 0.50 0.58 NÅ NÅ 0.37 Cencral South Region Henan 1.96 2.21 2.36 0.67 0.77 0.85 NÅ NÅ 0.34 Rubei . 2.77 3.38 3.45 0.73 0.87 1.14 NÅ NÅ NÅ Hunan 2.16 2.55 2.61 0.45 0.55 0.62 NÅ NÅ 0.42 Cuangdong 2.23 2.62 2.91 0.51 0.62 0.80 NÅ NÅ 0.84 Guangxi 2.25 2v34 2.50 0.52 0.61 0.73 NÅ NÅ 0.44 Southwest Region Sichuan 1.88 2.17 2.27 0.49 0.54 0.61 NÅ NA 0.36 Cuizhou 2.26 2.29 2.52, 0.49- 0.52 0.62 NA NÅ 0.31 Tunnan 2.27 2.47 2.71 0.54 0.61 0.68 NÅ NA 0.49 Xizang 12.68 14.20 15. 0.57. 4.87 5.03 NA NÅ 1.57 Northwest Region shaanxi 2.73 3.18 3.42 0.69 0.81 0.98 NÅ NA 0.58 Canau 2.81 2.95 2.94 0.59 0.66 0.72 NÅ NÅ 0.57 QInghai 6.87 7.09 7.25 1.06 1.30 1.61 NÅ NÅ 1.81 Ningxfa 5.69 5.51 5.44 0.82 0.91 1.10 NA NA 1.67 Kinjiang 4.33 5.06 5.29 0.63 0.81 0.98 NÅ NÅ 0.87 1 Boundary changes in nid-1979 preclude escimation of per capita totala for 1979. Source: These data vere derived from Stastical Annex Tables A-6 and g-7. - 55- Table A-9: SALARIES OF COUNTY-LEVEL HEALTH PERSONNEL, SHANDONG PROVINCE, 1982 Category Grade range Salary range (Yuan per month) Public Health Directoi 7 - 1 147 - 316 Senior doctor 14 - 6 59 - 168 Assistant doctor 17 - 6 41 - 168 Technician 19 - 15 31 - 54 Clinical Director 7 - 1 147 - 316 Visiting doctor 11 - 6 86 - 168 Seior doctor 14 - 9 59 - 115 Assistant doctor 17 - 11 41 - 86 Pharmacy Diree,tor• 7 - 1 147 - 316 Pharmacist 14 - 6 59 - 168 Assistant pharmacist 17 - 11 41 - 86 Technician 20 - 15 29 - 54 Auxiliary Fead nurse or midwife 14 - 6 59 - 168 Nurse or midwife 17 - 11 47 - 86 Nurse aides and MCH personnel 20 - 15 29 - 54 Source: These data were obtained from the Bureau of Public Health, Shandong Province.