HEALTH CARE FINANCING: A COMPARATIVE ANALYSIS

Transcription

1 PUBLIC BUDGETING & FIN. MNGMT., 7(2), SUMMER 1995 HEALTH CARE FINANCING: A COMPARATIVE ANALYSIS Sharon M. McManus and Khi V. Thai* ABSTRACT. This article compares and analyzes health care financing in the Organization for Economic Co-operation and Development (OECD) countries along two dimensions: (a) revenue sources and overall expenditures for health care, and (b) the mechanisms for reimbursement of hospitals and physicians. Comparisons of health care financing data and health care reimbursement systems in these industrialized countries reveal several findings that should interest policy makers who are pondering health reforms. INTRODUCTION In the OECD countries, health care financing has become one of major fiscal concerns of policy makers as health care expenditures became one of several major expenditure categories of the government budget, particularly when most OECD countries have struggled to control their budget deficits. This article will focus on health care financing in the OECD countries along two dimensions: (a) revenue sources and overall expenditures for health care, and (b) the mechanisms * Sharon McManus, Ph.D., has served as a faculty member of the Department of Health Administration, Florida Atlantic University. She is currently District Program Director for Children's Mental Health Services in the Alcohol, Drug Abuse and Mental Health Program, Department of Health and Rehabilitative Services, District IX, Florida. Khi V. Thai, Ph.D., is Professor and Director, School of Public Administration, Florida Atlantic University. His teaching and research interests are in public budgeting and finance. Copyright by PrAcademics Press

2 280 MCMANUS & THAI for reimbursement of hospitals and physicians. However, before addressing the above issues, difficulties and limitations of available health data will be discussed. DIFFICULTIES IN INTERNATIONAL COMPARISONS As pointed out by Schieber and Poullier (1989), international health care comparisons are difficult for the following reasons: 1. Data are not generally comparable; 2. Systems performance cannot be easily evaluated because of the inability to measure outcomes; 3. It is difficult to measure and control for social, medical, cultural, demographic and economic differences across countries; and 4. Transferability of policies across countries is problematic. The above cited difficulties are applicable to the OECD countries although they have many characteristics in common. All OECD countries are classified by the International Bank for Reconstruction and Development as industrialized and rich with per capita income ranging from $5,930 in Portugal to $23,700 in Denmark as compared with $80 in Mozambique in 1991 and with a high life expectancy ranging from 75 to 77 years as compared with 39 years in Guinea-Bissau in 1991 (Thai & Sekwat, 1994). These countries vary considerably, however, in some aspects such as wealth, size of land, languages spoken, and size of population. Moreover, despite their current level of industrialization, all countries have their own institutional and organizational arrangements, as well as their own history. These factors have all contributed to different developments of their health care systems and different ways their health care is financed. As mentioned later in this article, health care is financed differently among the OECD countries. In addition, this study relies heavily on OECD Health Data (1995). However, although OECD has instead time and resources for health data collections, data on a number of variables are incomplete. Several important variables needed for our comparisons are not available although they are listed in the above publication. General revenue and contributions and payroll taxes, for instance, are listed as two sources of health care financing but data are not available most countries. In some other instances, data on some other variables such health tax expenditures and private health insurance are

3 COMPARATIVE HEALTH CARE FINANCING 281 available for a handful of countries and for only a few years, Finally, data on some variables such as total health care expenditures, public and private expenditures, and gross domestic products (GDP) which are relatively complete are still missing for a few countries as well as a few years. Consequently, tables presented later in this article will contain "na" indicating "not available". Those countries which have data for less than three study years will not be listed in the tables. These difficulties are described here so that the validity limitations of the data should be recognized. Compounding the above problems is the lack of comparable data on health expenditures. Indeed, there is variation among the countries examined here regarding the array of health care services purchased or delivered. Some types of services occur in every country, for example, hospital services and physician services. However, some services are purchased through the health care system in a very small number of countries, for example, mental health or substance abuse services. All countries include hospital services in their consideration of expenditures. The extent of those services and the ownership of the hospitals included do lead to differences. Some systems exclude private hospitals, but consider public hospitals, while others include both public and private facilities. Many countries consider hospital-based specialist physician services as part of hospital services, but some like the United States categorize some of these as physician services. In addition, some health care systems include long term care/nursing home services into the acute care hospital setting, for example, as occurs in France, Austria or rural hospitals in Iceland (OECD, 1992, 1994). Other countries, like New Zealand, are encouraging hospitals to become more closely integrated with ambulatory care services and community health services (Malcolm, 1995). In Japan, very small clinics with fewer than twenty beds are considered part of community health care (OECD, 1994). Thus, there is considerable variation in the functions which comprise the term hospital services, incorporating services at both the input and the output sides, as well as differentiating among the types of ownership of the organization. These differences in definition seem substantial and increased specificity in terminology, or at least in the recording of which services are so categorized, appears indicated. All countries also include physician services in expenditures, but as with hospital services, what is defined as physician services evidences variation across these health care systems. Hospital-based physician services

4 282 MCMANUS & THAI are frequently included in hospital service, as for example, in Germany (OECD, 1992). While in Japan, as just noted above, clinics with fewer than twenty beds are considered as community health services. The differentiation among physician services, community health services, ambulatory care and primary care may entail the same service but provided in a different location, or the same service delivered by a different type of health care professional (nurse or physician). These inconsistencies in the definition of physician services also render comparisons about expenditures for those services across the various health care systems very difficult. Most of the OECD countries include pharmaceuticals or drugs within their health care expenditures. The majority of these countries also require some copayments or cost-sharing on the part of the consumer in purchasing pharmaceuticals. In Canada, children and senior citizens have no cost-sharing requirements for drugs (OECD, 1994). In Greece, there are some copayments required, but specified population groups are exempted from these requirements (OECD, 1994). It is unclear whether the discussions of expenditures for pharmaceuticals includes or excludes over-the-counter, nonprescription medications. Again, these variations in the definition of pharmaceuticals, specification of those included or excluded, and specification of exempted (or not) population subgroups renders the examination of those expenditures very difficult. Long term care is another component of health care system expenditures that many countries include. However, in Denmark, Germany, and Netherlands, nursing home services are not considered as part of the health care system, but rather as part of social services, distinct from health care (OECD, 1992; 1994) In Belgium, consumers pay board and lodging for nursing home care and the health care system provides the medical services component (OECD, 1992). Many of the OECD countries provide public health and/or community health services. Here again is a diversity of services within these categories without a clear delineation or distinction between them. Preventive and curative services are included together. School health services, home nursing services, health centers, health clinics, primary care, ambulance services, environmental health, diagnostic services, rehabilitative services and maternal and child health services all appear within the descriptions of these categories. Such a diverse array of activities within one or two types of expenditure impacts upon the meaning of any attempted comparison here.

5 COMPARATIVE HEALTH CARE FINANCING 283 Other types of health care services are also included in the determination of expenditures. Dental services are included for Australia, Belgium, Italy, New Zealand, Sweden and a few others, but not in Canada, where only dental services provided by oral surgeons in a hospital setting are included (OECD, 1992;1994). Austria, Netherlands, and Spain include psychiatric or mental health services, and some substance abuse services (OECD, 1992; 1994). Denmark includes physiotherapists, chiropractors and podiatrists; Switzerland includes physiotherapists and chiropractors; Norway includes physiotherapists (OECD, 1994). Both Greece and Ireland include ophthalmic services at least for some specified populations, while Netherlands includes domiciliary care and services for the developmentally disabled population (OECD, 1992;1994). And in a unique configuration, Finland includes employer-mandated occupational health services which extends to primary care services (OECD, 1994). Moreover, despite good efforts of OECD in data collections, data are not available, in some cases, for many countries, and in some other cases, for some years of a single country. However, data documented by OECD are the most comprehensive source of health data for the OECD countries. This source, as will be presented later, provides useful information for those policy makers who are considering health care reforms. HEALTH CARE FINANCING The Organisation for Economic Co-operation and Development (OECD) has also utilized a framework for comparing the health care financing systems of its member countries (OECD, 1992, 1994). This typology is based upon two dimensions: sources of financing and method of reimbursement. There are two sources of financing: voluntary or private financing, and compulsory or public financing. Along with these, there are four methods of reimbursing providers: out-of-pocket by consumers without insurance; out-of-pocket by consumers reimbursed by insurance; indirectly by third parties through contracts; and indirectly by third parties through budgets and salaries within an organization that provides both insurance and care delivery. Reimbursement in the field of public finance typically means funding from the government, however, in health care, reimbursement generally refers to the funds delivered to the providers of services regardless of the source of those funds. The consumers with insurance may also be reimbursed by insurance if the consumer reimburses the provider directly.

6 284 MCMANUS & THAI The combination of these two dimensions, sources of financing and method of reimbursing providers, yields eight models as follows (see Chart 1): Model 1, voluntary financing with out-of-pocket payment to providers by consumers, where consumers pay providers directly and do not have health insurance; Model 2, compulsory financing with out-of-pocket payment to providers by consumers where consumers are taxed and pay providers directly. [This model is hardly found in practice (OECD, 1992, p. 19)]. Model 3, voluntary financing with reimbursement of consumers, where consumers who have health insurance pay providers and are reimbursed from the insurance source; Model 4, public financing with reimbursement of consumers, where there is publicly funded insurance and consumers pay the providers and are then reimbursed by the insurance; Model 5, voluntary financing and third party payers or insurers contract with providers for direct payment, where services are usually free to the consumer and providers are reimbursed through contracts with voluntary insurance programs; Model 6, public financing and insurers contract with providers for reimbursement, where services are usually free to the consumer and providers are reimbursed through contracts with the publicly funded insurance programs; Model 7, voluntary financing with integrated insurance and provision of services within the same organization, where there are typically global budgets for hospitals and salaried physicians, services free of charge to the consumers and financing is through voluntary insurance (as in a health maintenance organization); Model 8, public financing with integrated insurance and service delivery, where there are usually global budgets for hospitals and salaried physicians, services are free of charge to the consumer and financing is through a publicly funded insurance program.

7 COMPARATIVE HEALTH CARE FINANCING 285 CHART 1 OECD Framework for Health Care Financing and Reimbursement Reimbursement Private/Voluntary Financing Public/Compulsory Out-of-Pocket Model 1 Model 2 Out-of-Pocket with Insurance Reimbursement Model 3 Model 4 Third Party Contract Model 5 Model 6 Third Party Budgets and Salaries Model 7 Model 8 Source: OECD (1992, 1994) No OECD country utilizes Model 2, as noted above. Moreover, most OECD health systems do not fall into a single model. The United Kingdom and Italy seem to fit Model 7 most closely. However, the Belgian system falls into Models 6 and 8 while the Canadian system fits Model 6 for physicians and Model 8 for hospitals. The French system fits Models 3, 5 and 8. The German system is dominated by Model 6, supplemented by Models 5 and 7. The system in the United States is even more complex, encompassing Model 1 for the uninsured, Model 3 for the insured and Model 6 for Medicare and Medicaid populations. Adapting the OECD framework, the remaining part of this article will examine two dimensions of health care financing in OECD countries: health care financing policy and methods of health care payment.

8 286 MCMANUS & THAI Private versus Public Funding Analysis of the levels of public and private funding of health care expenditures in the OECD countries (Table 1) reveals several important findings as follows. In a framework of the pure-public-goods/pure-private-goods continuum in the theory of public finance, some goods such as defense and education are provided purely or wholly funded by government at one end of this continuum, and some goods such as household appliances are provided purely or wholly by the private sector or market at another end the continuum. In a pure communist system, where the market system does not exist, health care is considered as a pure public good, financed by the government and provided by the government. However, in OECD countries, this continuum does not exist since in no country is health care provided and funded wholly by government or by the private sector. In almost all OECD countries, health care is close to the pure- public-good side of the continuum in terms of public/private funding. Indeed, as shown in Table 1, in the last thirty years, except in the United States, health care expenditure in all the OECD countries was mainly financed by the public sector. However, there is a great variation in the funding composition among the OECD countries. At one extreme (Tier 6, Chart 2), in the U.S. and Portugal, 44.0% and 56.2% of total health expenditures, respectively, are funded by public sources in At another extreme (Tier 1, Chart 2), in Norway and Luxembourg, 92.7% and 91.3% of total health expenditures, respectively, are publicly funded in Between these two extremes fall the remaining OECD countries as shown in Chart 2. Trends in Public/Private Funding From 1965 to 1993, three major trends in the public funding share of health care expenditures can be identified in Chart 2 as follows: - Increased share: Australia, Belgium, Canada, Finland, France, Japan, Netherlands, Norway, Spain, and Turkey; - Stable share: Denmark, Greece, Iceland, Ireland, Portugal, Sweden, Switzerland, the United Kingdom (UK), and the United States (US); and - Decreased share: Austria, Germany, Italy, and New Zealand.

9 COMPARATIVE HEALTH CARE FINANCING 287

10 288 MCMANUS & THAI

11 COMPARATIVE HEALTH CARE FINANCING 289

12 290 MCMANUS & THAI

13 COMPARATIVE HEALTH CARE FINANCING 291 Noting recent policy trends in several countries, Graig (1993) observed that the health care systems emphasizing government involvement have been moving towards more private sector involvement, while systems emphasizing the private sector have been experiencing increasing levels of government involvement. In terms of funding sources, this observation is not really accurate. Indeed, those health care systems heavily funded by the public sector have not moved more towards private sector funding involvement. CHART 2 Health Care Financing Systems Classified in Terms of Public/Private Funding % of Expenditure Countries Tier Publicly Funded % and over Luxembourg Luxembourg, Norway 2 80% % Denmark, Iceland, Belgium, Denmark, Iceland, Italy, New Zealand, Sweden, UK Norway, Sweden, UK 3 70% % Austria, Belgium Canada, Finland, France, Germany, Greece Greece, Ireland, Italy, Japan, Ireland Netherlands, New Zealand, Spain 4 60% % Finland, France, Japan, Australia, Austria, Germany, Netherlands, Switzerland Switzerland, Turkey 5 50% % Australia, Canada, Portugal Portugal, Spain, 6 Below 50% Turkey, US US Source. Derived from Table 1. Similarly, not all health care systems that are more heavily funded by the private sector have experienced increasing levels of government funding involvement. Of six countries in Tiers 5 and 6, whose the health care systems received the smallest shares of public funding in 1965 and the largest shares of private funding as compared with other OECD countries, Portugal and US remain unchanged in 1993, while four other countries (Australia,

14 292 MCMANUS & THAI Canada, Spain and Turkey) experienced increased shares of public funding. Australia and Turkey moved up one notch, from Tier 5 to Tier 4 and Canada and Spain moved up two notches from Tier 5 to Tier 3 in terms of higher levels of public funding or government involvement. Sources of Funding What are the sources of public and private funding? Are there any significant changes in those financing sources in the past three decades among the OECD countries? In OECD Health Data, there are three sources of health care finance: general government revenue, health payroll taxes, and private health insurance. In addition to the above sources, there are: - out-of-pocket payments by consumers who do not have health insurance, and - tax expenditures, that is, instead of directly financing health care costs, the government uses tax provisions to induce certain health care activities. For example, employer-purchased health insurance for employees is not considered taxable income, and hospitals are able to finance their capital expenditures through tax-exempt bonds. These tax provisions result in tax losses, namely tax expenditures. Although OECD has tried to compile health data for all OECD countries, data on funding sources are very incomplete. Information from only a limited number of countries and years is available for health payroll taxes and private health insurance (as shown in Tables 2 and 3). For information about general revenue for health expenditure, data from only one country (US) are available. Thus, it is impossible to conduct a comprehensive analysis of revenue sources for health care funding. As shown in Table 2, for those countries where data are available, Austria and Canada rely very much on health payroll taxes which cover, during the period, between 71% and 81.3% of total health costs in Austria, and about 83% in Canada. However, in Ireland and the UK where health care costs are heavily covered by government, health payroll taxes are not a major source of public revenue (about 8% or 9% of total health expenditure in Ireland and between 8% and 15% in the

15 COMPARATIVE HEALTH CARE FINANCING 293 TABLE 2 Health Payroll Taxes as % of Total Health Expenditure Country Austria Canada na na na na na na Ireland na na na na na na Switzerland na na na na na na UK US Source: Organisation for Economic Co-operation and Development (1995), OECD Health Data, Paris: OECD. UK). The United States relies much more heavily on health payroll taxes (31.1% of total health expenditure) to pay its publicly funded health programs (which cover 44% of the total health expenditure in 1993). The gap of 13.9% (44.0% %) in public funding is covered by other sources of revenues including general federal and state/local government revenues. Among the six OECD countries where data are available, there is no significant shift in the trend of this revenue source, except the United States which faced a large increase during the period of 1965 and 1970 when two major publicly funded health programs were initiated, Medicare and Medicaid. In almost all the fourteen countries for which data are available (Table 3), private health insurance has been an insignificant source of health care financing, except in the US where private insurance covers between 27.7% and 34.0% of total health care expenditure. As health care costs in Belgium and the UK are heavily funded by the public sector, private insurance has covered less than 2% of total health expenditure and 4%, respectively, during the period. Private

16 294 MCMANUS & THAI TABLE 3 Private Health Insurance as % of Total Health Expenditure Country Australia na na na na Austria Belgium na na Canada na na na na na na Finland na France na 1.4 2, Germany na na na na Italy na na na Luxembourg na na na na na na Netherlands na na na na na New Zealand na na na na Spain na na na na UK US Source: Organisation for Economic Co-operation and Development (1995), OECD Health Data, Paris: OECD. health insurance is also insignificant in Canada, France, Italy, Luxembourg and Spain. Finally, private health insurance, while small, is a very stable source of health care financing in the last three decades for these countries, except the US. HEALTH CARE SPENDING In general, there is great variation in total health expenditures and trends in health expenditure among the OECD countries although the average health expenditure in OECD as a whole has grown remarkedly in the past three decades, from 4.5% of GDP in 1965 to 8.1% in Sizes or Levels of Health Care Expenditure As shown in Table 1, the average health care expenditure of OECD countries is 8.1% of GDP in 1993, with considerable variation. The American health care system is the most expensive among the OECD

17 COMPARATIVE HEALTH CARE FINANCING 295 CHART 3 Health Expenditure Classified in Terms of % of GDP 1965 Expenditure 1993 Expenditure Over 10% Canada, US 9% - 9.9% Austria, France, Switzerland 8% - 8.9% Australia, Belgium, Finland Germany, Iceland, Italy, Netherlands Norway 7% Japan, New Zealand, Portugal, Spain Sweden, UK 6% - 6.9% Canada Denmark, Ireland, Luxembourg 5% -5.9% Australia, France, Germany, Sweden Greece, Turkey US 4% - 4.9% Austria, Denmark, Finland, Ireland Italy, Japan, Netherlands, UK 3% - 3.9% Belgium, Greece, Iceland, Norway Switzerland Below 3% Portugal, Spain, Turkey Note data are not available for Luxembourg (at 3.8% of GDP in 1970), or New Zealand (at 5.2% in 1970). They are also not available for Portugal and Turkey. However, health care expenditures in 1970 for Portugal and Turkey were very low at 2.8% of GDP and 2.5%, respectively. Thus, these two countries can be entered in this table without risk of error. Source. Derived from Table 1. TABLE 4 Tax Expenditures as a Percentage of GDP Country Canada na na na 0.2 na Finland na na Greece na na Ireland na na na New Zealand na na na na na US na na na Source: Organisation for Economic Co-operation and Development (1995), OECD Health Data, Paris : OECD.

18 296 MCMANUS & THAI countries, at 14.1% of GDP while the Turkish system is the least expensive, at 2.7% of GDP. It is interesting to note that in 1993, in the US which had the lowest commitment to public funding in 1993, health care expenditures are the highest of all OECD countries. Similarly in 1965, Canada had very low commitment to public funding (51.7% of total health expenditure) but health care expenditures were the highest in the OECD countries. In contrast, in Luxembourg and Norway which have the highest commitment to public funding (Chart 2), health care expenditures are below average (6.9% of GDP and 8.2%, respectively). If tax expenditures for health care were included, the total health care expenditures of those countries using tax provisions to induce some health care activities would be much larger. Again, OECD (1995) provided very limited data on tax expenditures on health care, available for only six countries and, worse, very few studied years (Table 4). Although the data provided by six countries are not complete, it seems that tax expenditures were used more extensively in the United States and Greece than in Canada, Finland, Ireland and New Zealand. Health Care Expenditures and Satisfaction There is no agreement on a good indicator to measure the quality of health care systems of the OECD countries in order to determine whether there is a correlation between health care costs and quality. There is no comparative analysis of health care quality of all OEDC countries. One indicator of quality is the level of satisfaction expressed by consumers. Five years ago, a comparative study of ten OECD countries were conducted by Blendon, Leitman, Morrison and Donelan (1990). According to this study, except in Canada, where a little over a half of its citizens were satisfied with the health care system, the majority of citizens were not pleased with their health care systems (Table 5). In France and Germany, about 41% of the population believed that their health care system worked pretty well. In Australia, Sweden and Japan, about one third of the population was pleased with their health systems. Meanwhile, Italians and Americans were not satisfied with their health care systems. Moreover, there is no association between the expenditures and the quality of the health care system. In the United States and Canada, where health care was the most expensive among the ten surveyed countries (ranked number 1 and 2, respectively), the Canadian health care system is the best in terms of quality, as measured by satisfaction, in the ten countries and the

19 COMPARATIVE HEALTH CARE FINANCING 297 American counterpart was the worst. The British health care system which cost the least among the ten countries is ranked number 8 in term of quality/satisfaction (only 27% of population felt that it worked pretty well); while the Dutch health care system was ranked the second best and cost the third lowest among the ten studied countries. (1) Finally, there is no association between the quality of a health care system with the size of public funding or private funding. In the United States, where the private sector comprised the largest portion of health care expenditure (57.5%) as compared with other surveyed countries, the American health care system was the least satisfactory. TABLE 5 Health Care Costs and Quality/Satisfaction of Ten OECD Countries Country Health Care Quality Health Expenditure Rank % of Rank % of Public Private (Quality/ Population (Most GDP Funding* Funding* Satisfaction) Satisfied Expensive) Canada Netherlands Germany France Australia Sweden Japan United Kingdom Italy United States * Funding as a percentage of total health expenditure Source. Derived from Blendon, R. Leitman, R., Morrison, I., and Donelan, K. (1990), "Satisfaction with Health Systems in Ten Nations," Health Affairs, 9, p.188 (for health care satisfaction); and from Organisation for Economic Co-operation and Development (1995), OECD Health Data, Paris: Author (for health expenditure).

20 298 MCMANUS & THAI Trends of Health Care Expenditure In general, health care expenditures increased rapidly in the OECD countries, from 3.1% of GDP in 1965 to 8.1% in 1993, as shown in Table 1. Health care spending increased by 6.% of GDP during , from 4.5% GDP to 5.1%. It exploded during the period of , from 5.1% of GDP to 6.4% (or an increase of 1.3% of GDP). Then, the growth of health spending declined during the decade of , from 6.4% of GDP in 1975 to 6.9% in 1980, and particularly during the first half of the 1980s, from 6.9% of GDP in 1980 to 7.1% in 1985 (or an increase of merely.2% of GDP). Health care spending started to increase at a faster rate during the second half of the 1980s, from 7.1% of GDP in 1985 to 7.6% in The health care expenditures exploded again, from 7.6% of GDP in 1990 to 7.9% in 1991, an increase of.3% of GDP in a single year. In summation, during the past 30 years, overall health care expenditures in the OECD, as a group of industrialized countries, grew faster than the economic growth. Several common factors (OECD, 1987; Abel-Smith, 1984; Pfaff, 1990) account for the continuing increase in health care expenditures in the OECD countries: - an expansion in the coverage of health insurance under public and private programs; - demographic change, particularly the increasing population of elderly in the population; - general economic inflation; - increases in utilization of new technologies involving both higher capital expenditures and operational costs; and - intensity of services as a result of the expansion of doctors per 1000 population. In the OECD countries, while there are increased health care expenditures and common factors causing the continuing increase of health care spending, marked differences exist among individual countries. The limited space of this paper does allow further analysis of these differences. Briefly, the rising health care expenditures have been the one universal driver for health care reform. Industrialized countries all faced pressures to reform the finance and delivery side of their health care systems.

21 COMPARATIVE HEALTH CARE FINANCING 299 MECHANISMS FOR HEALTH CARE REIMBURSEMENT Now this article turns to the second dimension of health care financing, the mechanisms for provider reimbursement. Various mechanisms for reimbursement are influenced and created by the context within which the flow of funds or fiscal transactions take place. None of the OECD countries is 100 percent publicly-funded. Since hospitals and physicians are the primary health care service providers, they will be discussed separately here. The examples illustrate the dominant reimbursement mechanism present in that country. Hospitals Beginning with hospital services, there are several channels through which these institutions are reimbursed by the health care system. Most systems have both public and private hospitals, although the proportion of each varies. Japan forbids for-profit private hospitals (Yajima and Takayanagi, forthcoming). The first and most centralized reimbursement mechanism flows from the national government to a regional health authority, responsible for purchasing hospital services (and other services) for a geographically defined population. The authority is given a set budget, which may have the population size, age and other variables factored in to the calculated amount, and services for that population are to be delivered within that budget. The United Kingdom, New Zealand, Italy and Spain are examples of this approach (Maynard and Bloor, forthcoming; Malcolm, 1995; Taroni, Guerra and D Ambrosio, forthcoming; OECD, 1994). A second approach, the block grant, involves the national government dispersing a set amount of funds to state or provincial governments, usually based on size of population, and delegating responsibility for health care services to that level of government. Canada and Australia are examples of this approach. In Canada, the provincial government then negotiates with the provincial hospital association for reimbursement levels. In addition, the provincial governments can supplement the federal block grant through other sources of provincial revenues such as taxes (Manga, forthcoming; OECD, 1994). Australia renegotiates the federal block grants to the state governments every five years (Harris and Harris, forthcoming). In the United States, the Medicaid program follows a similar approach, with state governments setting reimbursement rates for that program. Although there are several mechanisms typically utilized by states, prospective payment for Medicaid and other

22 300 MCMANUS & THAI payers, prospective payment only for Medicaid admissions, selective contracting and retrospective reimbursement (Thorpe, 1992), state government carries out this function. In several smaller countries, the county/municipality, rather than the state/province sets the reimbursement. Denmark, Sweden and Finland function in this manner (OECD, 1994). A third approach evidences less direct government involvement and increasing private sector or quasi-governmental health insurance activity but the focus remains at the national level. Here, the federal government determines a national fee schedule and all third party payers reimburse a standard per diem rate for hospital services. Japan exemplifies this approach in which the set fees are negotiated between the federal government and a council of health care providers, payers and consumers (Yajima and Takayanagi, forthcoming). Greece also functions in a similar manner with the federal government setting allowable fees (charges) and the third party payers (insurance, sickness funds) then reimburse hospital services on this per diem rate (Niakas & Petsetakis, 1995). In France, public and some private nonprofit hospitals receive global budgets but private hospitals are reimbursed through rates negotiated with the statutory insurers (OECD, 1992). In the United States, the Medicare program for those over the age of 65 years, follows a similar approach, utilizing a set reimbursement amount for each client based upon diagnosis related groupings. A fourth approach provides for an increasing level of private sector activity and somewhat less government involvement, exemplified by Germany, the Netherlands, Belgium and Switzerland (OECD, 1992; 1994). Here, hospitals negotiate reimbursement fees with sickness funds and private insurers based upon prospective global budgeting for the hospital. In Germany, the sickness funds are largely independent of each other and both federal and state governments. A fifth approach evidences a higher level of private sector, free-market based activity, exemplified by the United States and to some degree, also by Turkey and Austria (OECD, 1994). In the United States, private insurance typically reimburses on a percentage of the usual and customary charges by the hospital for that procedure. While private insurers are increasingly shifting to a fixed reimbursement form through a managed care mechanism, setting reimbursement in advance for the enrolled population, there remains much more fluidity in what amounts hospitals can charge and in what third party payers pay than in other health care systems. The government-funded

23 COMPARATIVE HEALTH CARE FINANCING 301 insurance mechanisms reimburse hospital services through set fees determined either federally (Medicare) or at the state level (Medicaid). These insurance programs cover 34 million people (Moon, 1994) and 30 million people (Rowland, 1994) respectively. The large majority of the population, some 77%, is covered by private health insurance offered through some 800 private commercial companies and 73 Blue Cross/Blue Shield plans (Whitted, 1993). However, it is important to explore two other factors which influence these mechanics of hospital reimbursement: precisely what services are included in hospital services and charges, and, does hospital reimbursement distinguish between capital expenditures and operating expenditures, and if so, how does this occur? As described in a prior section, hospital services is not a consistent group of services across all countries under consideration here. Some countries combine acute care services and long-term care into hospital services. For example, both Japan and Italy include long-term care services in hospital services (Yajima and Takayanagi, forthcoming; Graig, 1993; Taroni et al., forthcoming). In New Zealand, hospitals are moving towards service management, an approach which combines hospital and community-based (outpatient, social, rehabilitative services) care (Malcolm, 1995). In Germany, all hospital, physician and laboratory services are included in the hospital per diem (Graig, 1993), while in the United States, most of these services are billed separately. Thus, the definition of hospital services and what is included in the per diem charge or the global budget is variable from system to system. To collapse all of these definitions for the purpose of comparing hospital expenditures or costs renders the comparison meaningless. The distinction between hospital operating expenditures and capital outlays for service expansion or purchase of technology is another part of the lack of comparability in the definition of hospital services and their costs although it is beyond the scope of this analysis to elaborate further here. Recent trends in reimbursement for hospital services demonstrate that the health care systems with more centrally-controlled expenditures are building in some characteristics of free-market competition, while the least centrally-controlled system is moving towards greater fixed and prospective reimbursement. For example, both the United Kingdom and Italy allocated fixed and captitated budgets to regional health authorities which are responsible for the purchase of hospital services to a geographically specified

24 302 MCMANUS & THAI population (Maynard and Bloor, forthcoming; Taroni et al., forthcoming). Both countries have recently instituted mechanisms to promote competition between provider hospitals for the reimbursement within regions in hopes that this will encourage greater efficiency. And in the United States, the majority of the private health insurance plans/third party payers are moving into managed care: a fixed and capitated reimbursement mechanism allowing (forcing) competition between providers in contracting with the payers to provide services to the plan s enrollees or subscribers. In sum, each hospital receives a specified, set reimbursement amount from which it is to deliver the necessary services whether the mechanism is: * global budgeting, a fixed amount set in advance from the government (federal or state) or governmental agency (regional health authority) given to the hospital from which services are to be delivered for a specified population; or * managed care, a fixed amount set in advance from the private insurer given to the hospital under contract from which services must be delivered for a specified population; or * negotiated fees from sickness funds, a fixed reimbursement rate for the enrolled or covered population. Only two countries appear to allow an escape hatch for potential failure to remain within the prospective budget or estimated cost projections. In Greece, if a non-profit hospital was unable to cover its costs from the per diem charges set by the Ministry of Health, it becomes eligible for direct government funding. According to Niakas and Petsetakis (this symposium) such hospitals are now under direct government control. In addition, some large private hospitals are not dependent upon sickness funds and set heir own rate to attract the self-pay or privately insured consumer (Niakas and Petsetakis, 1995). And in the United States, if a hospital (whether public, forprofit private or not-for-profit private) is unable to cover the cost of services from the fixed reimbursement sources (Medicare, Medicaid, managed care plans), it shifts those costs to the non-fixed sources of reimbursement, standard indemnity health insurance plans and direct-pay clients, by charging these groups of consumers higher prices. Physicians

25 COMPARATIVE HEALTH CARE FINANCING 303 There are also several channels through which physicians are reimbursed by the health care system, again varying in the degree of centralization and government involvement. The emphasis here will be upon primary care physicians or general practitioners, those physicians who tend to be the consumers first contact or point-of-entry to the health care system. The most centralized and direct approach involves the federal or national government employing or contracting directly on a full-time basis with the physician. The clearest example of this mechanism occurs in the United Kingdom. Here, general practitioners contract with the National Health Service and the physicians salaries are comprised of three components: (a) a base salary which covers practice costs, (b) a capitation rate based upon the number of clients who have signed on with the physician, and (c) fees received from services to clients such as tests and immunizations (Graig, 1993; Potter and Porter, 1989; OECD, 1992). Specialized physicians can be employed (a) full-time in the NHS which allows up to an additional ten percent of their NHS salary to be earned in private work, or (b) part-time in the NHS, whereby these specialist physicians can also deliver unlimited private sector services (Graig, 1993). Such a centralized and direct physician reimbursement approach also occurs, although on a much smaller scale, in the United States with federal government salaried physicians in the Veterans Administration, the military and the Public Health Service. In Italy, a similar approach has been taken. The regional Local Health Offices (USLs) have contracts with the federal government for the purchase of health care services for a specified geographic region. From a national list of ambulatory services, the USLs set the maximum fees to be paid to physicians. General practitioners are self-employed but contract with the government to provide regular and comprehensive care to clients who have registered with them. Primary care physicians act as gatekeepers for referrals to specialists. The USLs purchase, on a fee-for-service basis, services from the general practitioners. Thus, with mandatory client registration with physicians, the physicians annual per capita fee is based upon the number of clients on his/her list (Taroni, et al., forthcoming; OECD, 1994). A very similar mechanism is utilized in New Zealand where four Regional Health Authorities with capped and capitated budgets purchase services from primary care practitioners (Malcolm, 1995; OECD, 1994). Other countries with a similar approach include Finland, Iceland, Ireland, Norway, Portugal and Spain (OECD, 1992; 1994).

26 304 MCMANUS & THAI A second approach decentralizes government authority away from the federal level to the state or provincial level. This mechanism is exemplified by the Canadian approach where the provincial governments, operating within the parameters of the federal block grant and the provincial global budget for health services, negotiate physicians fees with the provincial physicians association (e.g. Manga, forthcoming; OECD, 1994). From these negotiated provincial fees, physicians are reimbursed on a fee-for-service basis, with, for the most, few controls on volume of services. There are provincial variations on volume limits or reimbursement limits, especially notable in this regard is the province of Quebec (Evans, et al., 1989; OECD, 1994). State determination of physician reimbursement also occurs in the United States within each Medicaid program, the insurance program for the low-income population. State governments individually set the reimbursement rates they will pay for the federally specified services included in Medicaid. Related to this approach is the decentralization of these functions to the county or municipality (local) level. Examples of this approach are Denmark and Sweden (OECD, 1994). A third approach to physician reimbursement involves the federal government setting a national fee schedule for physician services. This can be seen in Australia, Japan and Greece. The Australian national health insurance, Medicare, sets fees for both general practitioner and specialist services and reimburses 85 percent of the schedule fee (Harris and Harris, forthcoming; OECD, 1994). There is no limit on the volume of services that can be provided. In Japan, all insurers within the mandatory, universal health insurance system reimburse on the basis of a national, fixed and itemized fee schedule. Physicians are then reimbursed on a fee-for-service basis from the appropriate insurer. There are no limits on volume of services. In addition, hospital and physician services are not separated and physicians not only prescribe but also dispense drugs (Yajima and Takayanagi, forthcoming; OECD, 1994). It is useful to note that in the United States, a national, set reimbursement mechanism for physicians is utilized by the Medicare program, an insurance program for those the age of 65 years. There is a Medicare reimbursement rate for physician services with fixed, set fees determined through a Resource-Based Relative Value Scale (RBRVS). The calculation of the reimbursement to a physician involves physician work, practice expense, malpractice insurance, and a single national value conversion factor. The final reimbursement amount is a geographically

27 COMPARATIVE HEALTH CARE FINANCING 305 weighted sum of the three components times the conversion factor (Koch, 1993; Hsaio, Braun, Yntema & Becker, 1988; Hsaio, Dunn, Becker, DiNicola & Ketcham, 1988; Grimaldi, 1991). A fourth approach involves less government activity in physician reimbursement, whether national, state/provincial or local level. Luxembourg, Austria, Belgium, Switzerland, France, Germany and the Netherlands exemplify this avenue with funding sources negotiating with physicians regarding fees. Government is involved either as a party in the negotiations or governmental approval of the negotiated agreement is required. In Germany, the sickness funds negotiate with associations of physicians for reimbursement rates (Graig, 1993, Henke, 1990; OECD, 1992). Physicians (general practitioners) must accept sickness fund reimbursement as full payment (Graig, 1993). Physicians send vouchers for services delivered to their clients to regional physician associations for reimbursement. These associations then reimburse the physicians, based upon the negotiated amounts from the sickness funds and they also monitor the volume of services delivered by each physician. Physicians providing services in hospitals and those who serve outpatient and ambulatory populations are two distinct groups and provide services in either the hospital or outpatient setting only (Stone, 1980; OECD, 1994). While private insurance exists, it usually is purchased by the affluent as a supplement to sickness funds. This insurance reimburses physicians at much higher rates than the sickness funds (Graig, 1993). A similar approach to the German system has been developed in the Netherlands, however, the Dutch federal government is a more active participant in negotiations with physicians about reimbursement levels. The government and physicians negotiate a norm income, or an income range which then becomes one of three factors utilized in calculating physician payment levels with sickness funds (Graig, 1993; Kirkman-Liff, 1989). The other two factors are the norm patient-list size (a standard negotiated practice size) and the norm practice costs. These three factors determine the sickness fund s monthly capitation rate to the physician (Graig, 1993; Kirkman-Liff, 1989). Private insurance is also present in this health care system and physicians who contract with private insurers are reimbursed at a fixed price for specified services. Most general practitioners treat clients covered by both types of insurance and are reimbursed on a capitation basis for those covered by sickness funds and on a fee-for-service basis for those