Turkey - Issues and options in health financing

Report No. 8042-TU Issues and Options in Health Financing in Turkey September 20, 1990 EMENA Regional Office FOR OFFICIAL USE ONLY Document of the World Bank This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  T U R K E Y HEALTH SECTOR FINANCING MEMORANDUM CURRENCY EQUIVALENTS Currency Unit = Turkish Lira [TL] Average Exchange Rate for US$1.00 1981 - TL111.22 1982 - TL162.55 1983 - TL225.46 1984 - TL366.68 1985- TL521.98 1986 - TL674.51 1987- TL857.20 1988 = TL1405.40 FISCAL YEAR January 1 - December 31 PRINCIPAL ACRONYMS USED BAG-KUR Social Insurance Agency for Merchants, Artisans and Self-Employed Professionals EEC European Economic Community GERF Government Employees Retirement Fund HMO Health Maintenance Organization ILO International Labor Organizatior. MOH Ministry of Health NHS [British] National Health Service OECD Organization for Economic Cooperation and Development SIO Social Insurance Organization  .AL, T U R K E Y HEALTH SECTOR FINANCING MEMORANDUM EXECUTIVE SUMMARY . . . . . . . . . . . . . . . . . . . . . . . . . . . . i BACKGROUND ........... .... .... .... .... ... . 1 PART ONE : RECENT TRENDS AND FINANCING MECHANISMS . . . . . . . . . . . . 4 I. Health Expenditures . . . . . . . . . . . . . . . . . . . . . . . 4 Total expenditures . . . . . . . . . . . . . . . . . . . . . 4 Private Health Expenditure . . . . . . . . . . . . . . . . . . 5 Public Health Expenditures . . . . . . . . . . . . . . . . . . 6 Major Types of Expenditures . . . . . . . . . . . . . . . . . 8 The Budget of the Ministry of Health . . . . . . . . . . . . . 8 II. The Cost of Health Care . . . . . . . . . . . . . . . . . . . . 9 Outpatient Services ............ .... .... . 9 In-patient Services . . . . . . . . . . . . . . . . . . . . . 10 III. Public Sector User Charges ..... . ....... . .. . . 11 IV. Health Insurance Schemes . . . . . . . . . . . . . . . . . . . . 13 Government Employees ............ .... .... . 13 Salaried Employees ............... ..... . . 14 Self-Employed . . . . . . . . . . . . . . . . . . . . . . . . 17 V. Population Coverage .......... ............. . 17 PART TWO : IMPROVING EFFECTIVENESS AND EQUITY . . . . . . . . . . . . . . 19 Effectiveness of Public Institutions . . . . . . . . . . . . . 19 Streamlining the User Charges System . . . . . . . . . . . . 20 Eligibility Criteria for Social Security Benefits . . . . . . 24 Revenues and Assets of Social Security Institutions . . . . . 25 PART THREE EXPLORING NEW OPTIONS . . 28 Guiding Principles ........ . ........... . . 29 Broad Features of Western Systems . . . . . . . . . . . . . . 31 A "Universal" Social Insurance Organization ? . . . . . . . . 32 A Turkish National Health Service ? . . . . . . . . . . . . . 33 A Turkish Liberal Medical System ? . . . . . . . . . . . . . . 35 What Role for the Private Sector ? . . . . . . . . . . . . . . 36 The Case for Private Insurance . . . . . . . . . . . . . . . . 37 A Catastrophic Health Insurance Option ? . . . . . . . . . . . 38 Health Maintenance Organizations ? . . . . . . . . . . . . . . 39 Financing First Levels of Care for Under-Privileged Populations . . . . . . . . . . . . . . . . . . . . . . 40 CONCLUSION .44 Tables This report wa prepared by Louis G. Vasailiou tEMlPEJ and Mehmet Tokat [tConultantl This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  EXECUTIVE SUMMARY [i] The Turkish public health system is expected to provide free preventive care to the population znd the State assumes primary responsibility for all matters related to communicable diseases and emergencies. Curative care must be paid for. Most health services are supplied under three largely autonomous systems: the Ministry of Health, the Medical Schools, and the Social Insurance Organization. In public health facilities operated by the Ministry of Health [MOH], user charges are collected directly from patients, or from a third party payor for persons covered by a health insurance scheme; however, fees are waived for the poor, who are often taken in charge by charitable foundations. University hospitals, opened to all, charge higher fees. The Social Insurance Organization [SIO] operates its own hospitals and dispensaries. The Army has a large network of facilities and covers the health needs of its active and retired members and their dependents. Some ministries, public organizations, state economic enterprises, foundations, religious groups and foreign communities still operate their own facilities. [ii] Only 59% of the Turkish population, i.e. about 31 million persons, have their health care needs covered under -a social security system consisting of four major schemes and some small private funds. This number includes: [i] government employees and their dependents [about 5.9 million persons] who have free access to MOH facilities; [ii] retired government employees and their dependents, covered by the Government Employees Retirement Fund [1.7 million persons]; [iii] members of the SIO and their dependents [13.3 million persons]; [iv] members of the Social Insurance Agency of Merchants, Artisans and Self-Employed Professionals fBAG-KUR] and their dependents [9.5 million persons]; [v] about 300,000 persons covered under various private funds, with limited benefits for dependents; and [vi] some 300,000 students in higher education institutions covered by a special scheme. [iii] Most of the remaining 22 million persons rely on the MOH for their health care needs. This is mainly the rural population but it includes lower income groups in urban areas with a substantial proportion of the labor force unaccounted for, especially in family businesses and cottage industries. They are expected to pay directly for health care received through the private sector or, more likely in MOH or university hospitals. Of this total, more than 5 million [10% of the population] are probably unable to pay and would seek assistance from MOH or charitable foundations. Except in cases of emergency or very serious illness, this group is bound to suffer most from the limited coverage of the current system. From the public sector's perspective, and irrespective of their ability to pay, MOH is thus expected to cover all health care needs of some 30 million persons [taking into account active and retired civil servants and their dependents]. [liv] The Turkish health system does not operate efficiently. More than two decades of neglect, under-funding and misallocation of resources have resulted in a run-down health system and very significant disparities in health and fertility status. The on-going restructuring process of the economy has often affected the underprivileged groups who rely most on the public health system. Recognizing that social sectors, and health in particular, have lagged behind in development, the Government has stated its  ii determination to reorient its social policy, and this policy received a new impetus with the country's application to the European Community. [v] The purpose of this report is to assist In identifying corrective measures to improve the health care delivery system, strengthen its financial foundations and, in a longer term perspective, set the stage for largely self-supporting public health services. Part I analyzes recent trends and shows that health expenditures in Turkey are low by OECD, and even by middle- income country standards. In 1987, MOH budget was raised by 20 percent in real terms, and has remained at this level; however, this will not be sufficient to sustain the projected growth in total health expenditures. Part II of the report therefore suggests four major areas of intervention that could lead to substantial gains in terms of efficiency and equity: the management of public sector operations; the user charges system; the eligibility criteria for social security benefits; and the management of assets of social security institutions. Nevertheless, gains in productivity would also fall short of the actual financing needs of the sector, and Part III explores options open to the Turkish Government. [vi] In Part I, total expenditures on health are estimated at 3.2% of the country's GNP. This ratio is substantially below the ones prevailing in OECD countries [ranging from 4.2% ini Greece, to 6% in Portugal and Britain, and 11% in the US] or in other middle-income countries [4%]. On a per capita basis, total expenditures on health declined from US$40 to US$35 between 1981 and 1986, before returning to US$40 in 1987 [current prices]; this compares with a 1986 range of US$715 [Britain] to US$1926 [US] in industrialized countries. Private expenditures increased by 65% in real terms during the same period and their share in total health expenditures escalated from 38% to 46%; this brought them to US$18.48 per person; overall, private expenditures now amount to 2.3 times the budget of MOH. As regards public expenditures, MOH channels -nly a third of all public outlays; the other major sources are the STO [about a quarter], the Ministry of Defense [15- 20%], and University Hospitals [10-15%]. Scanty data make it very difficult to assess the real cost of health services; they do nevertheless confirm the relatively low cost of health services in Turkey compared with other middle- income countries. [vii] User charges are collected in university hospitals as well as in MOH facilities. Large and long established university hospitals cover an increasing proportion of their recurrent budgets from patients dues, either directly from individuals or through third party payor [80% in 1987, against 25% in 1981]. User charges in MOH are based on a general tariff, updated periodically; however, fees are rarely collected in district hospitals, and are generally waived in health centers. MOH has substantially improved the collection of user charges which increased by 62% in real terms in 1981-87 in spite of high inflation rates. Nevertheless, revenues collected through direct billings in MOH's hospitals still represent only 6% of regular budget allocations. [viii] Current demand for public as well as private health services is low and can only increase in the future. Substantial gains in productivity could be obtained through an improved utilization of the existing infrastructure, and better management and training. Moreover, to respond to the pressing and growing needs of the population, it will be necessary to upgrade the skills of health personnel, modernize and introduce additional equipment in the  iii public health sector, and improve efficiency in service delivery. This is bound to increase total costs, irrespective of any gains in productivity. Consequently, public as well as private health expenditures will gradually increase, to a level more compatible with individual incomes and economic and social aspirations. The challenge confronting the Turkish health sector will be to improve the functioning of the existing system, and to increase resource mobilization from both public and private sources, while setting- up new mechanisms that will ensure the provision of efficient services, at an affordable cost, for the whole population. [ix] Many interventions are called for to enhance cost-effectiveness in the provision of health services. As the report focusses on financing issues, Part II addresses four priority areas. [x] First, the effectiveness of public health institutions is hampered by insufficient and obsolete equipment, and lack of vehicles, drugs and medical supplies, aggravated by shortages and high turnover of trained personnel. Evidently, this calls for sustained investments and higher levels of current spending. But the solution does not lie in a mere increase of public financing and, in the short term, the income generation potential of the user charges system is limited. Enormous benefits can be derived from improved planning and management of services, training, new job descriptions, reduction of in-patient services through expanded ambulatory care, the generalization of pre-admission tests, reallocation of excess capacity, improved management of drugs and supplies, development of nursing homes as an alternative to hospital treatment, etc. The overriding priority for the sector is to improve hospital utilization through a reduction of average lengths of stay, better referral procedures, and well targeted investments aimed at Improving the current admission policy. [xi] Second, it is essential to improve the efficiency of the user charges system. Although it would not be realistic to expect high levels of cost recovery in the foreseeable future, a strong case is made for a wider collection of user charges: equity and user charges are not incompatible, provided that the coverage of services improves and adequate safeguards are built in for the poor. Periodic and frequent adjustments are crucial at times of high inflation. Moreover, MOH should do away with the single tariff concept which calls for the same rate to be applied in all facilities for a given procedure. The charge for any given procedure should also reflect the actual cost of the service. Finally, the whole process of collection and administration of user charges in MOH facilities must be reviewed: the current system ensures sound bookkeeping practices but it is costly and cumbersome and unsuitable to health centers or even district hospitals. This simplification is possible as shown in Part III. [xii] Third, the harmonization of eligibility criteria for social security benefits is crucial in a system composed of many organizations, especially during the phase of transition towards universal coverage. This concerns the provisions regarding pensionable age for men and women, as well as the package of health benefits. As any reform in this area is likely to entail unpopular decisions, the issue can only be addressed in close collaboration with organized labor, in order to reach a national consensus. [xiii] Finally, particular attention needs to be paid to the management of the revenues and assets of social security institutions. SIO, as well as  iv BAG^KUR and the GERF operate under capitalization financing systems and, because contributions still exceed benefits, the Turkish social security system is not threatened by insolvency in the short run. However, its long- term financial position is being eroded by continuing liquidity problems and legal reserves are not maintained at the required actuarial level. Three issues are addressed in this context. First, escalating costs in recent years have resulted in deteriorating contributions-to-benefits ratios. The second issue relates to premiums collection, and the financial constraints caused by employers' delays in transferring contributions to SIO. The third issue relates to the management of the assets and the return on the investments of the social security institutions. Legal constraints, liberal lending policies, generous benefits, conservative management and rapid inflation have reduced investment returns and endangered the long term viability of the system. In spite of the drastic reforms introduced by the Government after 1986, the three institutions have not yet fully recovered. [xiv] Part III is based on the premise that Turkey is at a crossroads and that options regarding its national health system are wide open. There is a clear trend towards an extension of health care benefits to larger segments of the population. When considering the options available to the Government, three fundamental sets of factors must be kept in mind. FLrst, the Turkish social security system is an old system, deeply rooted in history and culture; this would exclude, at least for the mid-term, any radical reform or dramatic departure from the Bismarckian model, i.e. the compulsory affiliation and contributions of workers and employers, to a centrally funded and managed system. Second, the country's aspiration to join the European Community has strengthened the tendency to look towards Europe for a model, and raised the need to harmonize Turkey's system with conditions prevailing in the EEC. This would also exclude a radical shift towards a very "liberal" model, based essentially on the private sector. Third, there is not much room for exceeding current rates of overall payroll deductions, as direct household outlays already represent almost 48% of total health expenditures [57X when adding workers contributions for health coverage under social security]. [xv] In Western industrialized countries [with the exception, to a large extent, of the United States] the public sector has retained a major overall responsibility in the planning, organization, supervision and, in many instances, the delivery of health services. Most health expenses are covered by social security institutions, but payments are still generally made on a case by case basis or per hospital day. Many countries are also gradually moving towards systems where financial annual allocations to health care providers are decided ex ante, on a capitation basis, as in the United Kingdom model which has inspired the 1980 reform in Italy. Financing institutions assume a more and more important role, by imposing their rules and standards on health care providers. In the OECD countries, some 80X of health expenditures are financed by public outlays; in Europe, the proportion ranges from 71% in France and 78% in West Germany, to 90% in the United Kingdom and the Scandinavian countries; in the United States, it is only 411. Public financing mechanisms are based either on general taxation or on compulsory insurance, with wide variations between the two basic models. Most European countries rely on compulsory health insurance schemes funded by compulsory employers and workers payroll deductions; the Government regulates the system and often provides general or specific subsidies; this is basically the Bismarckian model which has also prevailed in most of Latin  V America and was adopted in Turkey. Western systems initially provided virtually full coverage but, havirg been affected by very rapidly escalating costs, have adonted, or contemplate, measures such as: deductibles [especially for out-patients services and prescription drugs] and co- payments: lump sum payments to service providers, based on diagnosis related groups; sub-contracting of ancillary hospital services; etc. [xvi] In Turkey, various models could influence the on-going evolution. The first alternative is evidently a Universal Social Insurance Organlzation, using SIO as the cornerstone of a national system extending health coverage to the whole population. In 1987, the last year for which actual data are available, assuming SIO had been able to provide health coverage to 29.6 million more persons [i.e. the rural and informal sectors], under the same rates, it would have required raising US$1151.7 million from these new beneficiaries. It could be argued that, as two-thirds of health contributions in the modern sector are paid by employers, the State should have provided, through increased taxation, a direct subsidy of about US$768 million; this compares with a total MOH budget for 1987 of some US$421.7 million, with only half of it earmarked for curative care. But these numbers are misleading: on the one hand, the cost of pub'Lic health services, presently very low, is bound to increase; on the other hand, the sudden access to social security coverage by large numbers of presently underserved individuals would trigger a dramatic increase in demand for health services. So far, SIO has managed to fund health benefits with relatively modest premiums; its foundation would be jeopardized if it were to provide any type of universal health coverage at present levels of contributions. [xvii] The British model has a strong appeal in Turkey, as in many other countries, and a Turkish National Health Service, financed at 90X by taxation as in the United Kingdom, would be perceived as a "free" system, to provide good quality coverage to the whole population. But this model requires a wide taxation base, and an efficient collection mechanism ensuring high yield on direct taxes, which still does not exist in the country. As workers of the modern sector paying regular income taxes represent at best one third of the civilian labor force, a national health scheme inspired by the British model would have to be financed through a regressive system based on indirect taxation. Assuming that, in 1987, Turkey operated a national health system similar to the British one; assuming further that its total health expenditures were 51 of its GNP [instead of 3.2X, but compared to UK's 61], total health expenditures would have been US$3.4 billion [instead of US$2.1 billion]. Because, as in the UK, private health expenditures would have fallen to about 101 of the total [against 461], public expenditures would have jumped to US$3.0 billion [against US$1.2 billion]. The financing gap, i.e. the additional resources to be raised through taxation to finance this Turkish NHS, would have been close to US$1.9 billion. This would have represented, for 1986, 611 of personal income taxes levied by the Treasury [US$3.1 billion], and more than all corporate income taxes together [US$1.3 billion]. [xviii] At the other end of the spectrum could be advocated a Turkish Llberal Medlcal System tailored on the United States model. The impressive technical and medical results of the liberal model are achieved at very high economic and social costs, and the system is being more and more criticized by beneficiaries, financing institutions, and the medical establishment. In Turkey, it would exclude a substantial proportion of the population unable,  vi or unwilling, to purchase private health insurance. More importantly, this liberal model goes against social and cultural factors that would make its adoption extremely difficult. This does not mean that Turkey, as many other countrics, cannot learn from the reforms and structural changes introduced in the United States or other countries to cushion the excesses of the liberal approach to health care delivery, e.g. Health Maintenance Organizations, pre-determined lump sum payments based on Diagnosis Related Groups, Preferred Provider Organization contracts between large corporations and organized health care providers, "global budgets", etc. They all include elements of particular interest during this transition phase in Turkey. [xix] As regards the Private Sector, physicians already provide a substantial share of primary care, even for rural populations; but for inpatient services, the country relies essentially on the public sector, as private hospitals represent less than 5% of total beds and the larger ones are non-profit institutions. For the private sector really to take off in Turkey would require: [i] either that S10 gradually abandons its function of health care provider to become a financing organization, or [ii] that health insurance develops more rapidly, as an alternative or as a complement to social security institutionis. But, as the trend towards increasing membership in SIO is unlikely to be reversed, the development potential of private health Insurance is limited in the near future. Civil servants and the military are the only organized groups not yet served by SIO; if the provision of. free medical care for these personnel in MOH or Army facilities were abandoned in favor of a social security type of coverage based on payroll deductions, it is very doubtful that the Government or beneficiaries themselves would opt for private insurance instead of a participation to SIO. The private insurance industry could then play a role as a voluntary complement to SIO, to cover deductibles as in some countries. The prospects for private health insurance as an alternative to public cove-age under SIO appear even more remote. [xx] The introduction of catastrophic health insurance would also be confronted with major problems. To fulfill its objective, such a scheme would presumably be compulsory. This would correspond to a direct tax levied on under-privileged groups excluded from SIO; yet, among t1'ese 20 million or more persons, those who cannot pay the [low] user charges of the Ministry of Health, are already protected by the de facto safety net, being exempted from user charges or supported by charitable foundations. For cultural reasons, they are also likely to resist any :ype of health insurance. Members of SIO and BAG-KUR simply do not need this catastrophic insurance as most of their health expenditures are already covered. Catastrophic health insurance would become a necessity only if MOH [and University hospitals] were to raise user charges dramatically, and collect them more rigorously, beyond the support capacity of the foundations. [xxi] Could health maintenance organizations [HMO] constitute a complement or, in a longer term perspective, an alternative to the existing system ? Their development would be hampered by major legal and institutional obstacles: participation in the social security system is compulsory for all enterprises with paid workers; there are not yet institutions capable of and ready to enter into contractual agreements for the large scale provision of a comprehensive package of health services; HMOs require large numbers of participants, to be economically viable; finally, they can only be set-up by organized groups. Nevertheless, HMOs should be encouraged, as they would  vii stimulate modern management techniques, promote cost-effective measures in health service delivery and generate constructive competition among health care providers. BAG-KUR self-employed members, who presently subscribe to SIO health coverage on a voluntary basis, represent a potential market for HMOs. On an experimental basis, the law regarding compulsory participation in the SIO could be amended. The experience could be carried out in the metropolitan areas where industries as well as the private health care providers are already concentrated. Fiscal advantages could be considered to encourage physicians to set up group practices, or for changing the status of the largest hospitals run by for-profit organizations, medical schools or by foundations. The potential benefits would largely outweigh the risks or financial cost to the Treasury of the experience. [xxii] To illustrate the choices facing the Government, the report then speculates on how the system could evolve over the next few years. As the taxation base in Turkey is still too narrow to permit the early introduction of a National Health System, tailored on the British modc.l, the distinction between [i] modern sector workers and their dependents, and [ii] the rest of the population living in rural or periurban areas, is likely to persist well into the next century. The number of wage earners in the modern sector is bound to increase and, since participation in the Social Insurance Organization is compulsory for all paid employees, this will lead to a parallel increase in the number of SIO beneficiaries; conceivably, BAG-KUR members, civil servants and military personnel could also be integrated into SIO. Coverage of SIO can then be expected to grow significantly before it is replaced by ar,. national scheme, and SIO will probab'ly be the cornerstone of such a scheme, whatever its characteristics. The government will therefore need to prevent the development of a heavy bureaucracy and the escalating administrative costs of mature systems as in France or Latin America; it should also stand ready to revise existing laws and regulations regarding SIO's mandate and the management of its assets. For the foreseeable future, private insurance is not a -riable alternative to SIO; nevertheless, it could play a growing role, to complement SIO benefits, by covering co-payments. [xxiii] Soon, SIO will have to decide whether it will continue to be a provider of health services, or whether it should limit its role to being a financing institution. The experience of other European countries shows that the latter is more likely and that SIO will eventually divest itself of its facilities [either in favor of MOH or by selling them to the private sector". If SIO became a purely financing institution, this would have an immediate impact on the development of the private sector as the number of family practitioners and small in-patient facilities would grow very rapidly; the private sector would be further strengthened by the purchase of SIO hospitals. The notcome would be a system comparable to the French one where, for curative care, social security beneficiaries have the free choice [within limits] of private or public practitioner and health facility, the public sector being limited to teaching hospitals and large specialized institutic-ts. [xxiv] While workers protection calls for compulsory health insurance, this does not imply compulsory participation to any given scheme, or to SIO in particular. If SIO were to relinquish its responsibilities in service delivery, and if workers had a choice, it is unlikely that many would decide to leave SIO; it is even more doubtful that they would opt for private health  III insurance. Rather, such a move could sustain the development of Health Maintenance Organizations (HMOs] in the major cities. Individuals [or groups of individuals] could enter into direct contractual agreements with HMOs; workers and employers premiums would be paid directly to HMOs, which would compete among themselves for membership; SIO would retain, at best, a regulatory function. But such a drastic departure from the European model is unlikely. A more plausible development would see contractual agreements between SIO and HMOs for the provision of health services to their members, at predetermined rates. In the short run, University Hospitals would provide the best foundations for HMOs, following the pattern of some large American universities. [xxv] The 22 million persons deriving their income from the primary and informal sectors, who cannot join SIO, BAG-KUR or any regular paying scheme, will only gradually diminish. The State must therefore continue to assume its responsibilities in this area and, for many years to come, MOH will remain a major provider of curative care. Because of the obvious resource constraints, this population will have to increase its direct contributions to its health coverage. In recent years, MOH has been quite successful in improving cost recovery, in spite of high inflation rates. It will have to continue its efforts in this direction: user charges must not only be adjusted to more realistic rates, they must also be levied more systematically at the lower levels of the system. While there is logic in not collecting user charges in health centers that provide only minimal services, this tolerance would be difficult to justify, once investments have raised service quality. A more systematic but pragmatic implementation of the user charges system, with the direct involvement of the communities concerned, is therefore recommended in the report. The scheme should not be seen as a government device to escape its obligations in matters of public health, or even to reduce expenditures. It is designed primarily for areas already underserved, even by the private sector; it would create only limited competition for qualified private practitioners whose services will always be needed, especially in mid-size and large cities. [xxvi] It would be futile to try to identify the financing scheme most suited to Turkish conditions; such a scheme is likely to be a combination of different approaches that will emerge from a national consensus taking into consideration historical and cultural factors as well as political, social, and financial constraints. What is attempted here is a presentation of alternative models that could influence the final choice. Fundamental questions still need to be answered, clearly defined objectives must be identified and major policy choices must be made before embarking on any reform. In particular, what should be the respective roles of the public and private health sectors; what type of services should the public sector provide, to whom, and at what price; should social security institutions be at the same time providers of health services and financing bodies; etc. [xxvii] The issues debated here extend well beyond the traditional areas of concern of the Ministry of Health, and call for the active involvement of the major health care providers of the public and private sectors, as well as representatives from workers and employers organizations and financial institutions. They also call for a type of expertise in short supply in Turkey. It would be a very serious mistake to embark in isolation on any reform, without the benefit of the experience of other countries which have been through similar exercises. It is also essential for Turkish decision  makers to gain a first hand knowledge of the functioning of the various social securlty systems likely to influence the current evolution in their country. Once a preliminary consensus emerges on a possible reform, it will be essential to open the debate to outside participants, to submit any proposal to the scrutiny of international experts, particularly to avoid the mistakes made in other countries. * * *  B A C X G R O U N D 1. The Turkish public health system is expected to provide free preventive care for the whole population, and the State assumes primary responsibility for all matters related to communicable diseases and emergencies. Curative services must be paid for. Most health services are supplied under three largely autonomous systems: the Ministry of Health, the Medical Schools, and the Social Insurance Organization. In public health faciiities operated by the Miristry of Health (MOH], and except for persons covered by a health insurance scheme, patients are charged directly. However, fees may be waived for patients who present an indigent certificate delivered by the Muniicipality; the poor are also often taken in charge by one nf the numerous charitable foundations. University hospitals, which cover a large part of the demand for curative care, also charge fees, substantially higher than the ones collected in MO0 facilities. The Social Insurance Organization [SIO] operates its own hospitals. The Army has a large network of facilities and covers the health needs of its active members, retirees and their dependents but very little is known about its operations. Other ministries (Education, Youth ard Sports for example], public organizations (Post and Telecommunicatio.-s, Railways], and state economic enterprises still operate their owr. hospitals. In addition to the very active private for- profit sector, many foundations and hospitals for religious groups and foreign communities still exist in Turkey. 2. Health care needs of about 31 million persons, or 59X of the population, are covered under a social security system consisting of four major schemes and some small private funds: a. Government employees (about 1.5 million persons] and their dependents (4.4 mil- SOCIAL SECURITY COVERAGE lion] receive free medical care in IMOH facilities; mo- re precisely, the budget of each Ministry or public or- s,o 133 ganization includes a spe- "'ERF 17 cial allocation (Article Ova SOrafits 59 180] to cover medical ex- BGKus 9 s penses incurred by its em- Stuents 03 ployees; S 0, Others O 0 b. retired Government emplo- yees and their dependents Not Covered 22 [1.7 million] also receive free medical care, these services being billed by Million persons MOH to the Government Employees Retirement Fund [GERF];  2 c. the Social Insurance Organization (SIO] operates its own network of hospitals and dispensaries for some 13.3 million workers, retirees and their dependents who receive quite comprehensive pre-paid curative care; SIO also purchases services for its beneficiaries from MOH or University hospitals, in the absence of a facility or on referral cases; d. the Social Insurance Agency of Merchants, Artisans and Self- Employed Professionals (BAG-KUR] provides, since the end of 1988, health benefits to 9.5 million persons, including dependents; e. private funds cover the health care needs of some 300,000 persons, with limited benefits for dependents; finally, f. a special scheme provides basic health coverage to some 300,000 students in higher education institutions. 3. Most of the remaining 22 million persons fin addition to active and retired government employees and their dependents] rely on MOH, and to some extent on university hospitals, for their health care needs. This is mainly the rural population. The number also includes lower income groups in urban are&s with a substantial proportion of the labor force unaccounted for, especially in family businesses and cottage industries which still employ large numbers of children and apprentices. 4. The Turkish health system does not operate efficiently. More than two decades of neglect, under-funding and misallocation of resources have resulted in a run down health system and very significant disparities in health and fertility status. The on-going restructuring process of the economy has often affected the underprivileged groups who rely most on the public health system. Recognizing that social sectors, and health in particular, have lagged behind in development, the Government has stated Its determination to reorient its social policy. This received a new impetus with the country's application to join the European Community. 5. The purpose of this note is to assist in identifying corrective measures to improve the effiziency of the health care delivery system, strengthen its financial foundations and, in a longer term perspective, set the stage for largely self-supporting public health services. Part I analyzes recent trends in expenditures and financing, during the 1981-1987 period, when the country's GNP increased by about 30X in real terms [15X per capltal. The analysis shows that health expenditures in Turkey are low by OECD, and even by middle-income country, standards. It also provides evidence of structural changes in recent years, where the decline in public expenditures was partly offset by increasing households outlays. Since 1987, a growth in public health budget allocations has taken place but, even if this were to continue, it would not be sufficient to sustain the projected growth in total health expenditures. Part II of the note therefore suggests four major areas of intervention that could lead to substantial gains in terms of effectiveness and equity. This section also draws from the findings of a previous review' and from data and observations gathered during the preparation of the first health project submitted for World Bank financing. I