A. The Need for a Guide for Macroeconomists

Whether working on a low-income country (LIC) or an advanced economy, it is important that macroeconomists are aware of issues in health economics and health policy. Health care has gained prominent recognition in development. The UN Millennium Development Goals (MDG) set 10 specific targets to be achieved by 2015, of which three explicitly pertain to health. Research studies document that millions of households in developing nations are impoverished each year by health expenditures, retarding poverty alleviation efforts, and an emerging body of research shows that investments in health can have a significant effect on economic development (Fogel, 2004; Bloom and Canning, 2000; and Bloom, Canning, and Sevilla, 2004). HIV/AIDS has undermined the development prospects of a number of African countries and threatens to weaken the growth momentum of several important Asian economies. For middle-income, transition economies, and industrial countries, the challenges posed by the health sector for macroeconomists differs, but is no less daunting. Increasingly, pressures emerging from the health sector—in part due to the aging of populations and in part due to the rapid pace of technological change in the medical sector—are affecting fiscal sustainability, inflation, and possibly even the current account of the balance of payments.

For those working in the health sector, it is also obvious that macroeconomic policies can have a measurable impact on health care. Macroeconomic guidelines relating to public expenditure targets, inflation control, tax policy, and exchange rates will have effects on the provision of health care and ultimately on the health status of the population (Glied and Remler, 2002). Since a large share of health funding comes from the government, fiscal targets will constrain how much a government can spend on health. Tax policies relating to tobacco, alcohol, and firearms will influence people’s demand for these products and ultimately their health (Jha and Chaloupka, 1999). And the exchange rate will be a factor determining the cost of vaccines and drugs. Macroeconomists concerned with the fiscal balance will also influence decisions on the respective roles to be played by government and private markets in the health sector. Such decisions will influence the pace of health cost inflation, the efficiency and quality of health care, and the degree of equity associated with its financing.

This primer represents an effort to bridge the divide, enabling particularly macroeconomists (such as those who work at the IMF) to obtain at least a basic understanding of the key health policy and health system issues that may arise which are integral to their understanding of the macroeconomy. For those macroeconomists with limited time, who only want to have the most basic grasp of health economics and health policy, the major points of this primer are briefly summarized in this introductory chapter. Section B describes the unique characteristics of health and the health sector that have implications for policy. Section C follows with a discussion of the major challenges facing the health sector. Section D characterizes some key empirical research findings with important policy implications.

Finally, Section E provides a road map to the different chapters of the primer for those who wish to go further. The primer recognizes that the key health policy issues of concern to policymakers or macroeconomists in low-income countries will not be the same as those faced by their counterparts working on advanced, middle-income, or transition economies.

B. Unique Characteristics of Health and Health Systems

Put simply, there are a few important facts that are fundamental to an understanding of the role of health and the functioning of the health system.

• Good health, broadly shared, is intrinsically valued in all societies. The concept of equity in health and equal access to health care is based on an ethical notion of fairness. Inequities are intrinsically repugnant; disproportionate illness burdens and suffering by selected groups of people offends our innate sense of justice. From this perspective, we can infer that, at the minimum, every individual should have access to basic services and medicines to relieve pain and suffering and to avoid untimely death. The most ambitious goal would be for every individual to be able to attain his or her full health potential regardless of age, gender, or socioeconomic status.

• Health care as a good differs from other essential goods and services, such as nutrition, education, and housing. The likelihood of illness or the incurring of health expenses is subject to uncertainty. Individual households face only a small probability of having a major accident or illness. But most households would have difficulty affording the cost of treatment for a major illness. As Kenneth Arrow (1963) and John Nyman (2006) show, providing health insurance increases a society’s welfare. Insurance, however, causes moral hazard and results in a loss in economic efficiency.

• Serious market failures exist in the health sector. The supply side dominates the demand side in the health services market. Professional dominance prevails due to the asymmetry of information between physicians and patients. If left unchecked, the medical profession can exercise its monopolistic power to induce demand and set high prices, leading to rapid health cost inflation and a deterioration in the quality of services. Insurance markets also experience serious market failures. Adverse selection poses a serious problem for voluntary health insurance and deters the pooling of risks between the healthy and less healthy. On the supply side, private insurance companies select the healthy and younger people to insure and leave the less healthy and poor uninsured. Thus, the good risks are not pooled with the bad, and government is left with a serious social problem and a large fiscal burden if it is left providing for the uninsured.

• The cost of serious illness can be a major factor causing poverty for households in many LICs. The cost of modern medicine (e.g., diagnostic tests, surgery, and hospitalization) is simply not affordable for most LIC households. For example, the average cost of a hospitalization typically exceeds the annual median household income. Studies have found that for several LICs, 20-30 percent of households have had to borrow or sell assets each year in order to pay for medical expenses (Russell, 1996; Sauerborn, Adams, and Hien, 1996; Liu, 1995; Gu and Tang, 1995). Households either need medical insurance or access to subsidized health services to prevent financial bankruptcy.

C. Challenges and Issues

Across the world, one can identify the following key challenges facing the health sector.

• Spending on health care strains both household and government budgets throughout the world. In many countries, health expenditure per capita has risen faster than the rise in GDP per capita. Consequently, health care costs have taken an ever-increasing share of government, employer, and household budgets and put pressure on those financing the burden of health care (through taxes or insurance contributions). Looking ahead, there are concerns in many industrial countries that an ever-increasing share of the government budget spent on health care (OECD, 2006) will crowd out resources for other important public goods and publicly provided services.

• The aging of populations is a development now confronting both advanced economies and a number of emerging market countries (notably China, Korea, Thailand, Singapore) (see Australia Productivity Commission, 2006; OECD, 2006; European Commission, 2006; and Bryant and others, 2005). Illness and health expenses increase with age. The rapidly aging population in advanced economies will further exacerbate the pressures associated with a high health expenditure inflation rate.² China in particular will, within the next few decades, face an increasing share of its population in the elderly group. Yet few countries have a coherent policy strategy to deal with the fiscal problems arising from an aging population.

• A “double disease burden” and equity issues face middle-income countries (MICs). The epidemiological transition confronts most MICs. Such countries not only have to fund and deliver services to address the communicable diseases affecting the poor but also for the chronic diseases facing middle-and upper-income groups (see Reddy and others, 2005). This will inevitably force these countries to reform their health care financing and delivery systems. Equity in health also gains greater attention politically as an MIC becomes more affluent. Typically, only the rising number of upper-income households can obtain and afford to pay for expensive sophisticated health care, while others go without. Such disparity creates social and political tensions that governments recognize the need to confront. But they may not have the knowledge or resources to do so.

• LICs challenged by HIV/AIDS are now the beneficiaries of a rising tide of donor assistance for prevention and treatment. But many countries are having difficulty absorbing this assistance and most confront a severe shortage of human resources. HIV/AIDS has weakened the economy of several LICs and strained their health care and social service capacity. Most simply lack the absorptive capacity—health care infrastructure and human resources—to prevent and treat HIV/AIDS patients. Perversely, the large inflows of donor assistance targeted to these diseases (through so-called vertical disease programs) have weakened the infrastructure and drained the human resources required for preventing and treating common diseases (such as diarrhea, and upper respiratory infections) that may kill many more people. Furthermore, multiple donors, each with their own priorities, bureaucratic requirements, and supervisory structures, have created waste and confusion with recipient nations. Lastly, an important concern is the sustainability of these vertical programs, since donors’ funds may not prove stable or longlasting. For recipient countries, these inflows have created difficult challenges in the management of the health sector.

D. General Findings and Policy Actions

Sound public policy should be based on scientific information, evidence, and solid analysis. Economic policymakers can inform themselves of several major lessons learned from research and countries’ varied experiences in organizing and financing health care. This accumulated knowledge has taught us which policies are sound and which are not. The following general findings and actions can be applied in most countries to improve health care.³

• The good health of a population significantly contributes to human capital development and economic productivity. Numerous micro-studies have found that child’s health has a large impact on his or her ability to learn and retain knowledge (Jackson, 1993; Kramer, Allen, and Gergen, 1995; and Novello and others, 1992). Adult health status affects the size of the labor force, worker absenteeism, and worker efficiency (Mushkin, 1962). Recent macro studies have found that the health of a population can significantly influence a country’s rate of economic growth (Fogel, 2004, Bloom and Canning, 2000, and World Bank, 2004).

• Health resources should be allocated to achieve three objectives: (i) an optimal level of health status distributed equitably; (ii) an adequate degree of risk protection for all; and (iii) the highest possible level of public satisfaction for the entire population. Achieving these objectives will require making difficult decisions about trade-offs, especially between equity and efficiency.

• One size does not fit all. Nations are at different stages of socioeconomic development and have different epidemiological patterns. They differ in their resource capacity, knowledge base, human resources, and institutions. What works in the United Kingdom would not likely to work in Kenya. Few general health policy guidelines can apply to all nations and an universal performance standard does not exist.

• Governments should establish institutions to finance health care and pool risk, rather than relying only on the free market. However, the way in which the market and government can work efficiently and appropriately varies by function, for example, financing versus the provision of health care. Although many countries have tried regulatory remedies to correct the market failures in the voluntary private insurance market, no country has succeeded. On the other hand, international experience shows that government-managed “free” public health services tend to be inefficient and non-responsive to patients’ needs. Market mechanisms can provide services that are more efficient and higher in quality than government-managed free services.

• As market competition is capable of addressing only the efficiency issue, the government has to be responsible for the equitable financing and distribution of essential health goods. At a minimum, the government should: (i) finance and provide public and merit goods, for example, health education, immunizations, and maternal and child health services; (ii) target and subsidize primary care and hospital services for the poor; (iii) establish national or regional health risk-pooling (e.g., insurance) mechanisms for formal sector workers and their families; (iv) establish regulations to remedy market failures and monitor market performance; and (v) educate the public to be informed consumers of health services. In carrying out these responsibilities, the government may have to increase its health budget, build its institutional capabilities, and strengthen its human resources.

• Because publicly financed health benefits in developing economies usually favor higher-income households, governments should shift their resource allocations to target their subsidies to the poor and to those in the greatest need.

• Costly technology used by advanced economies to provide sophisticated medical treatments as well as clean water and sanitation is often unaffordable for LICs. Yet the potential for developing affordable technology is not exploited by industrial country researchers or pharmaceutical firms because it is not glamorous or profitable. To increase social benefits for the poor, international organizations should promote the use of affordable technologies and support investment in international public goods, such as the development of vaccines for malaria and HIV/AIDS.

E. Road Map to the Primer

This primer aims to provide macroeconomists with the essential information they need in order to understand the economics of the health sector. They need to know which health policies may improve the equity and efficiency of health care, as well as which policies will improve the level of health status, reduce poverty, and enhance social and political stability. As a primer, it does not provide in-depth information or complete evidence for the arguments made, but offers an extensive bibliography for those who desire further information. It also highlights situations in which macroeconomists should engage health sector specialists in policy formulation exercises.

In particular, subsequent chapters address six basic questions. Chapter II examines why it is critical for macroeconomists to understand health policy issues. Chapter III discusses what basic facts (and myths) that macroeconomists should know about the health system and the financing of the health care market. Chapter IV examines both the common health policy challenges confronting all countries as well as the challenges particular to countries at different stages of development. Chapter V explores what analytical framework should be used to assess the health care system.

Chapter VI provides ideas on guideposts that can be used in considering health policy options for countries at different stages of development. This particular chapter is lengthy because health issues and their economic considerations vary significantly across countries and to be relevant, it is necessary to discuss the health issues and policy separately for each stage of development (to help the reader, Figure 1 provides a road map to steer macroeconomists toward those sections most relevant for the type of country on which they are working). Finally, Chapter VII offers concluding thoughts on what specific advice macroeconomists can provide, given their normally peripheral role in health policy, to ensure that macroeconomic policy issues are taken into account.

View Full Size

A Road Map for Reading the Primer

Citation: IMF Working Papers 2007, 013; 10.5089/9781451865776.001.A001

• Download Figure
• Download figure as PowerPoint slide

A. Basic Facts about the Health Sector

Certain basic facts about the health sector and health care are not widely known. This section explains what macroeconomists should know on this subject as they participate in the formulation of health policies.

• Trade-off between equity and efficiency. Equity becomes a paramount consideration when everyone needs basic health care to maintain life and relieve pain and suffering. Without insurance, health expenses can also be a primary cause of falling into poverty. Therefore, basic health care and risk-protection have to be equitably distributed and price should not be used as the primary rationing tool. This implies a significant trade-off between equity and efficiency. A society’s social values and political philosophy influence how a country approaches this tradeoff.

• Resource allocation within the health sector also involves complex trade-offs. The trade-off goes beyond equity and efficiency. Health systems are developed to achieve multiple outcomes. Contrary to simplistic belief, the allocation of resources in the sector cannot be based solely on cost-effectiveness, where effectiveness is measured only as health gains. The uncertainty of potentially large financial burdens arising from serious illness creates a legitimate demand for insurance against expensive hospital services. In public and private financing of health care, resource allocation has to aim at three common objectives: improving the population’s health, protecting people from financial catastrophe, and meeting the public’s expectations as to the availability of health care services. These multiple goals inevitably imply difficult trade-offs.

• Markets alone cannot produce efficient outcomes in the health care sector. The health care sector consists of more than a dozen markets, most of which suffer serious failure due to an asymmetry of information, imperfect agency relationships, barriers to entry, and moral hazard. Evidence shows that the supply side (technology and induced demand) drives health cost inflation more than consumer demand. Specific market failure examples such as adverse selection (Rothschild and Stiglitz, 1976; Cutler, 1996) and risk-selection⁸ (Luft, 1986; Holahan, 1997), seriously impair the efficient operation of insurance markets. Moral hazard from insurance also produces inefficiency. In the service-provision market, physicians have the market power to practice price discrimination (Kessel, 1958) and induce demand (Yip, 1998). In factor markets, patent law protections offer monopolistic profits for pharmaceutical products. Licensing laws restrict free entry and competition. These market failures cause inefficiency, high health expenditure inflation, and inequity. International experience has taught us that some of these market failures can be corrected, but others are beyond our current capability to address. In summary, a blanket policy to allow the health sector to operate fully on a free market basis will not yield efficient outcomes. Market competition can be used only selectively for the benefit of a country’s health and welfare.

• The distribution of health expenditure is highly skewed. In rich and poor countries alike, approximately 20-25 percent of a country’s total health expenditure in any year is spent on one percent of the population, and approximately 50 percent of expenditure is spent on five percent of the population. For 20-25 percent of the population, there is no spending for health care in a given year. Table 1 gives the contingency table of US health expenditure (i.e., percent of health expenditure by percent of the population), which is fairly typical of most countries. But our ability to predict which individuals are at highest risk in any given year is very limited. Thus, this skewed distribution has at least three important economic implications: (i) health insurance is necessary, as shown in the seminal paper by Kenneth Arrow (1963); (ii) adverse selection inhibits the establishment of a stable health insurance market, as shown by substantial empirical evidence (Rothschild and Stiglitz, 1976; Cutler and Zeckhauser, 1998); and (iii) risk-selection by insurers leaves the burden of covering the high-risk population to the government (Davis, 1975) and reduces the quality of medical services (Newhouse, 1992).

• The supply side has a greater impact on health care efficiency, quality, and spending than does the demand side. Moral hazard on the demand side is well known. What is less understood and measured is the fact that monopolistic power on the supply side has a greater impact on efficiency, quality, and spending. A US government-funded ($100 million) study by the Rand Corporation in the early 1970s measured the price elasticity of demand for medical services in the US and found the elasticity of health expenditure to be about 0.2 (Newhouse, 1977). Other studies have found an elasticity range of 0.1 to 0.4 for LICs. It is, however, important to note that the elasticity could be greater than 1.0 for poor families in LICs (van der Gaag, 1991; and Gertler and van der Gaag, 1990).

Table 1.

United States: Contingency Table of Health Care Expenditure, 2002

(For all age groups)

Source: Yu and Ezzati-Rice, 2005.

Table 1.

United States: Contingency Table of Health Care Expenditure, 2002

(For all age groups)

Percent of Total Population	Percent of Total Health Care Expenditure
1	22
5	49
10	64
50	97

Source: Yu and Ezzati-Rice, 2005.

View Table

Table 1.

United States: Contingency Table of Health Care Expenditure, 2002

(For all age groups)

Percent of Total Population	Percent of Total Health Care Expenditure
1	22
5	49
10	64
50	97

Source: Yu and Ezzati-Rice, 2005.

Table 2.

United States: Health Expenditure for Five Most Costly Conditions, 2002

Source: Olin and Rhoades, 2005.

Table 2.

United States: Health Expenditure for Five Most Costly Conditions, 2002

	Condition	Billions of Dollars	Percent of Total Health Expenditure
1	Heart disease	70.0	8.3
2	Trauma	55.8	6.9
3	Cancer	48.4	6.0
4	Mental disorder	47.5	5.9
5	Pulmonary conditions	45.3	5.6
	Total	267.0	32.7

Source: Olin and Rhoades, 2005.

View Table

Table 2.

United States: Health Expenditure for Five Most Costly Conditions, 2002

	Condition	Billions of Dollars	Percent of Total Health Expenditure
1	Heart disease	70.0	8.3
2	Trauma	55.8	6.9
3	Cancer	48.4	6.0
4	Mental disorder	47.5	5.9
5	Pulmonary conditions	45.3	5.6
	Total	267.0	32.7

Source: Olin and Rhoades, 2005.

On the supply side, the asymmetry of information between the physician and the patient as well as the urgency of some medical conditions give physicians the power to induce demand and set prices. The elasticity of induced demand (viz., the elasticity of prescribed demand to price) has not been rigorously measured because of data limitations. Nonetheless, the US government accepted the value of 0.5 for policy-planning purposes. For LICs, the stylized facts show that this elasticity is much higher than in the US, because professional ethics and quality regulations are less developed. It is well known that physicians in LICs can drastically change the type of drugs prescribed, tests ordered, and length of hospital stay ordered in response to price changes that affect their income. The vast number of case studies indicate that the elasticity of inducement would be near 1.0 in these countries. In summary, the elasticity of induced demand is comparatively larger than the price elasticity of demand.

In advanced economies, where there is an adequate stock of physicians, an increase in the aggregate supply of physicians raises total health expenditure. This is because disease etiologies provide almost endless opportunities for specialization. An increase in the number of physicians usually leads to growth in subspecialties, which leads to a greater number of referrals. Consequently, the quantity of services and health spending rises. This has happened, for example, in the US, further fragmenting health care with questionable benefit for many patients. In summary, to promote efficiency and manage expenditure inflation, it is more important to control the supply side than to regulate demand.

• Per capita health spending increases with age. Figure 2 shows that, in the US, those between ages 65 and 74 typically spend three times more on health care than those aged 18-64. This ratio increases to four times for the group aged 75-84. Other advanced economies show similar relationships.⁹ By the year 2010, most advanced economies will begin to experience an accelerated rate of increase in the elderly dependency ratio, with the ratio tripling for many by 2030. MICs and LICs are experiencing similar demographic transitions, though starting a decade or two later (Heller, 1998). As a country’s population ages, pay-as-you-go (PAYG) methods of public financing for health care will place an increasing tax burden on the working age population. This burden will have an impact on both the labor market and the national saving rate.

View Full Size

US: Average Health Care Expenditure by Age Group, 1999

(In dollars)

Citation: IMF Working Papers 2007, 013; 10.5089/9781451865776.001.A001

Source: U.S. government, Center for Medicare and Medicaid Services 2005

• Download Figure
• Download figure as PowerPoint slide

B. Myths about Health Care Systems

As discussed, health policy deliberations are hampered not only by ignorance about the basic facts, but also by the prevalence of several widely-held myths. Some of these are based on standard neoclassical economic theory that is itself unsupported by empirical evidence. Nonetheless, these myths have strongly influenced policy debates around the world, particularly in the US. Five such myths are debunked below, in the hope that policymaking can be based on solid evidence:

• Myth #1: National or social insurance restricts patients’ choice of providers; private insurance gives much greater choice.

This myth is widely held in the US and has distorted the public’s understanding of social insurance. In fact, Canadians, covered under social insurance, have total freedom to choose providers under their national health insurance; and Germans have a much greater freedom of choice under their social insurance than do Americans under private insurance and managed-care plans.

• Myth #2: Health expenditure inflation cannot be managed because it is driven by the demand for services produced by new, cost-increasing technology.

Health expenditure inflation rates have indeed increased since the late 1960s. Economists have hypothesized two reasons for this trend. First, three decades ago, most US economists believed that expenditure inflation was caused by the moral hazard arising from increasing insurance coverage. This hypothesis turned out to be unfounded when insurance coverage in the US remained level or declined in the 1980s and health expenditure inflation continued unabated. Second, many argue that high health expenditure inflation can be attributed to the rapid advance in, and increasing availability of, expensive medical technologies. Economists argue that new technology is endogenously determined by health insurance (Weisbrod, 1991). This hypothesis is not valid for all types of health insurance, though it might be valid for the US, which relies largely on private insurance. In OECD countries with extensive insurance coverage, however, wide use of new medical technology still coexists with much lower rates of expenditure inflation.

Evidence shows that all major advanced economies except the US have been able to manage their health expenditure inflation since the mid-1970s. Prior to the mid-1960s, advanced economies spent 4.0-5.5 percent of GDP for health care and experienced similar health expenditure inflation. In the late 1960s, all advanced economies experienced rising rates. Countries have tried different methods to control and manage this problem. By the mid-1970s, all major advanced economies except the US were able to keep the total share of national health expenditure in GDP under control. As shown in Figure 3, the rise in expenditure share for the major advanced economies became less steep, while the slope of that for the US remained basically unchanged. These other advanced economies have, however, been adopting new medical technology just as rapidly as the US. Thus, international experience demonstrates that the main cause of health expenditure inflation is not new medical technologies. The cause is something more fundamental.

View Full Size

Selected Countries: Total Health Care Expenditure, 1960-2003

(Percent of GDP)

Citation: IMF Working Papers 2007, 013; 10.5089/9781451865776.001.A001

Source: OECD 2005

• Download Figure
• Download figure as PowerPoint slide

View Full Size

United States: Concentration of Health Care Expenditure Over Time

Citation: IMF Working Papers 2007, 013; 10.5089/9781451865776.001.A001

Sources: Berk and Monheit, 2001, and Yu and Ezzati-Rice, 2005

• Download Figure
• Download figure as PowerPoint slide

Other advanced economies have been able to manage health expenditure inflation pressures by establishing an effective budget constraint over the entire health sector. A “closed” budget encourages greater efficiency, including the curtailing of less cost-effective medical practices. Health expenditure inflation rates differ because most major advanced economies have been able to break the endogenous link between medical technology and health insurance by deciding exogenously on the level of overall health insurance expenditure. The US health financing system provides an “open” budget to the health sector. This gives providers the ability to shift costs from one plan to another. In other words, the US does not have a hard budget constraint for its health sector.

Americans argue that other countries have been able to manage their inflation rates by under-investing in health care, which has resulted in long waits for certain tests and surgical procedures. However, patients in most advanced economies—the United Kingdom (UK) and Canada are the two exceptions—have not had to wait long for non-emergency tests and surgeries. Recently, the UK has sought to eliminate long waiting lists by increasing its health spending from 6.8 percent of GDP to about eight percent of GDP (see Aaron, 2006). Canada has waiting lines since the late 1980s due to the drastic cuts made in health budgets during the economic downturn, but that situation is improving after the economy revived and the federal government increased the health budget.

• Myth #3: Government-financed health services inevitably become underfunded, resulting in shortages, inferior quality of services, and long waits.

Besides the public good elements of the health sector, most countries use general revenue to finance health services for reasons of equity and risk-protection. Most countries also accept the principle that health care is a basic need—that it is inappropriate to use price to ration health care according to people’s ability and willingness to pay. Because governments of LICs and MICs are financially unable to meet all demands, their public health services are usually underfunded. Consequently, these services are rationed by other means, such as long waits for treatment. However, this is not the case for advanced economies. Most have been able to balance supply with demand. The notable exception is the UK, which relies on the political process at the central government level to set its national budget for health. UK health care has to compete with other national priorities, such as education, social security, and national defense. The public, acting through the political market, seems to be willing to accept wait-rationed treatment.

• Myth #4: All preventive care is cost-effective.

Often policymakers make the erroneous assumption that the cost of preventing a disease is always less than the cost of treatment. This is not so. Besides the cost of supplies, effective prevention programs often require the identification of the population at risk and that a high proportion of that population complies with the preventive regimen. These efforts can be very costly. For example, studies show that it is more cost-effective to treat TB cases rather than to prevent it (Borgdorff et al, 2002). Influenca vaccination is not cost-effective for healthy working adults (Bridges et al., 2000). In addition, some preventive measures produce serious side effects that result in permanent disability. Preventive policy should thus be judiciously based on cost-effectiveness analysis, not on impressions.

• All private sector providers are more efficient than public sector providers.

Similarly, often macroeconomists assume erroneously that private sector production of health insurance and health services is more efficient than public sector production. This is not so. Studies consistently have found that the cost of private insurance is higher than public insurance because of the difference in marketing cost. In the provision of health services, studies in the US have found that private sector production could be more efficient if there were sufficient competition. On the other hand, Singapore has found that the private sector hospitals charge much higher prices for hospital services that are of the “same” quality as those provided by public hospitals.

C. Basic Health Economics: Health Markets and Market Failures

As noted above, empirical studies of the past three decades have confirmed the presence of serious market failures in the health sector. Most of these failures are caused by factors pervasive to the various markets of the health sector, viz., an asymmetry in information, imperfect agency relationships, and moral hazard.

• In the financing market, the asymmetry of information between the consumer and insurer about the former’s health condition results in significant adverse selection. The fact that health risks are concentrated in a small portion of the population results in serious risk-selection efforts by insurers. Although insurance is needed to cover the uncertainty of future illness, insurance creates moral hazard.

• In the health service provision markets, one would not expect competition to work when patients suffer from urgent or life-threatening medical problems. In such situations, patients are unlikely to “shop around” given the absence of sufficient information to discriminate among sources of supply. Moreover, asymmetry of information gives health practitioners strong monopolistic power to set prices and induce demand. Regarding the supply of health practitioners, to assure patients’ safety, the government and the medical profession have erected high barriers of entry as to who can practice medicine.

• In the pharmaceutical and medical-device markets, patent laws give monopolies to new drugs and new medical technologies in order to encourage research and development. Although these barriers to entry and monopolies have been established for sound social and economic reasons, they have nonetheless impaired the competitiveness and efficient operation of markets.

Recognizing these serious market failures (or the absence of the prerequisite conditions for a workable competitive market), many countries have turned to the government to finance and provide health services. The last 50 years have shown both the benefits and limitations of state action, especially in the promotion of health. Governments have helped to deliver substantial improvements in education, health, and economic security. Without this government role, sustainable development, both economic and social, would have been impossible (see World Bank, 1997:2). However, not all experiences with state action in the health sector have been encouraging. Government decisions are, of course, mostly based on bargaining among political alliances. Thus, the relative powers of different interest groups can greatly affect decisions on resource allocation and on those who benefit from and pay for public programs. Often, a disproportionate amount of public health services goes to the affluent and the urban middle class rather than to the poor. Moreover, curative hospital services are favored over cost-effective primary and preventive services.

Because most governments operate by “command and control”—using bureaucratic rules to manage operations—and public facilities usually operate as a monopoly, even the best-intentions of bureaucrats can atrophy from a lack of information and insulation from patients. Health practitioners’ interests often begin to supersede those of patients and the general welfare. Politics can then dominate public health services, turning them into major centers for patronage employment, particularly when labor unions become bases of political support. In many LICs and MICs, because of the absence of appropriate checks and balances, corruption, fraud, kickbacks and under-the-table payments to physicians and nurses have become widespread. Hence, public health services often become unresponsive to the needs of patients, and the efficiency of public health services deteriorates.

How a country finances and provides health care depends on the objectives that a country pursues. Because there are serious market and government failures, the policy question is not one of either market or government, but rather the relative degree of each and how to harmonize them. The respective roles of the government and of the market are ultimately decided by societal objectives set for the health system, with these objectives shaped by the values embraced by the society. For instance, markets normally cannot produce equitable health or health care because of income, gender, and racial inequalities in a society. If a country gives priority to equity, then the government has to take a primary and strong role in financing health care. In sum, a country’s social values and priorities determine the extent to which a country relies on the market as opposed to the government. Nonetheless, it will enhance rational policymaking if objectives are clearly set forth and both market and government failures are well understood.

When there are serious failures of both market and government, sound public policy becomes much more difficult to develop. Policymakers may then choose a strategy that will produce a better, but not the best, outcome—in essence, a second-best solution.

A. Major Issues Confronting All Nations

While specific issues confront nations at different stages of development, several issues confront all nations throughout the world. We first present the universal issues, then the ones for each stage.

Most nations, rich or poor, face the problem of per capita health care costs rising faster than per capita GDP. Consequently, health care outlays are absorbing an increasing share of government, employer, and household incomes. This constant fiscal pressure forces nations to confront two basic questions: how to finance this rising burden and how to contain the pressures for health expenditure growth.

Several common factors on the demand and supply sides have caused the rapid rise in health care costs globally. On the demand side, the HIV/AIDS pandemic and people’s heightened expectations and demands are the principal causes. Influenced by worldwide distribution of medical news, sensational headlines on medical breakthroughs, and aggressive drug advertisements by pharmaceutical companies, people are bombarded with new information about diseases and their treatments. We have become much more conscientious about our health and illnesses and expectations are raised about medicine’s ability to treat and cure. Some new expensive technologies may yield only small marginal benefits, but news of their discovery distorts consumer perceptions. On the supply side, medicine and the medical profession itself have become more commercialized. Profit has also become a strong motivating force in biomedical research. Driven by the profit motive and protected by patent laws, new medical science and technology produce mostly cost-increasing new treatments. Concomitantly, hospitals and physicians who adopt new sophisticated medical technologies gain in both reputation and profit.

Another driver increasing costs involves the socialization of medical practices throughout the world. The US has become the dominant force in medical education and research. Its emphasis on specialty medicine and expensive high-technology treatments has profoundly influenced global medical practices, and this is reinforced by regular education and continuing education programs, role models, and telemedicine. In effect, foreign physicians have been socialized to adopt the “American practice standard,” regardless of whether their country has the resources to pay for it.

The balancing of the demand and supply for health services and drugs is another critical issue confronting all nations. Few nations have the resources to satisfy people’s needs and wants in health care. Consequently, health care has to be rationed. Instinctively, economists would argue for a rationing by price. However, reasonable health care is a necessity of life, so that equity concerns must also be a key factor in policy decisions. Also, illnesses are uncertain and income is inequitably distributed in most countries. Thus, price rationing favors the affluent, while lack of insurance impoverishes most families that have large medical expenses. For those reasons, most nations provide nearly free basic health services or cover their citizens under social health insurance programs. As a result, price is not used as the primary rationing tool. Queues, waiting time, travel distance, lesser quality of service and outright unavailability, frequently serve as effective rationing tools. But when carried too far, rationing can generate public dissatisfaction and political whiplash. This was the experience in the UK and Canada and, more recently, in the US with respect to managed care. Bedside counseling by physicians to induce patients to demand less services has been used successfully in many advanced nations, but the internet and the popularization of medical information have weakened this constraining influence. Most recently, the followers of free market ideology have continued their push to ration by price, labeling it as “consumer-driven” health plans. This approach is not likely to work because the asymmetry of information places providers in a dominant market position and individual patients’ choice is not likely to exert much competitive force. Rapid health cost inflation is unlikely to be contained under this strategy. How to ration scarce health resources will thus remain a critical issue.

Finally, a health system is a dynamic organism. All nations confront the medium-term challenge of adjusting their health systems to meet changing epidemiological and demographic conditions. Disease patterns do change with socioeconomic development, demographic change (e.g., with aging, the prevalence of chronic diseases such as Alzheimer’s will increase), and, in the future, possibly with climate change. New infectious diseases can emerge, as illustrated by the recent emergence of SARS and Avian flu. The prevention and health service delivery systems will thus have to change accordingly. New more effective technologies and drugs, typically more costly, present a continuing challenge to governments and insurance companies forced to decide whether to sanction (and finance) their use.

Are these insoluble problems? To deal with health cost inflation, authorities may consider: (i) placing an effective global budget for the health service sector so providers have to lower costs and compete for patients. Limiting price increases encourages the providers to adopt new cost-reducing technologies and new management tools, and reform their organizational structure for efficiency gains; (ii) mandating that households save a portion of their wages annually in order to prefund the fight costs of health care that will be incurred later in life; (iii) offering incentives with public funding to develop cost-reducing technologies; (iv) altering the payment system to promote cost-reducing technology; (v) evaluating new technologies in terms of their cost-effectiveness (do they produce better clinical outcomes? how do new and established technologies and drugs compare in terms of cost and relative effectiveness?) and (vi) establishing an effective long-term budget for the prevention of HIV/AIDS, allocating a specific portion of budget for this goal.

To control the pressures of new technologies, medical education programs of advanced economies should seek to take account of the differences in the economic status of countries by: (i) developing medical standards varying by nations’ economic status and introduce these different standards into the medical curriculum; and (ii) efforts should be made to identify role models who practice appropriate medicine for different stages of socioeconomic development and publicize them.

Balancing Supply and Demand: authorities may seek to influence the expectations and demand for health care by: (i) promoting truth in advertising for drugs and medical devices; (ii) providing government funding for social marketing programs to balance the role of private advertising; (iii) establishing standard protocols of treatment that take into account resource constraints; (iv) capping medical insurance benefits to exclude coverage for “heroic” and experimental medical services; (v) establishing mandatory medical saving accounts (MSA) for long-term care and catastrophic medical expenses after a given age (such MSAs could be used to purchase insurance in order to pool the risks); and (vi) establishing reverse mortgage and low-interest loan programs for catastrophic medical expenses.

B. Challenges Facing Advanced Economies

Advanced industrial countries also face several issues of their own. First, most high-income countries confront the prospect of an aging population. People are living longer while fertility rates are declining. The elderly dependency ratio is projected to increase rapidly in the next two decades. Under any PAYG system for health insurance, population aging will have profound implications for the fiscal burden placed on future generations of workers, labor market conditions, and saving and consumption patterns. Table 3 shows how the shares of the elderly and support ratios will evolve through 2050 for selected advanced and middle income economies.

Table 3.

Selected Countries: Projected Shares of Elderly and the Support Ratio 2000-2050

Source: United Nations, Department of Economic and Social Affairs, Population Division (2006) Note: The support ratio is defined as the ratio of the population aged 15-64 to the population aged 65 and over.

Table 3.

Selected Countries: Projected Shares of Elderly and the Support Ratio 2000-2050

	Percentage of Total Population Over Age 60		Percentage 80 Years or Older		Potential Support Ratio 3
Country or area	2006	2050	2006	2050	2006	2050
Asia
China	11	31	10	23	9	3
Japan	27	42	19	37	3	1
Republic of Korea	14	41	10	31	7	2
India	8	21	10	16	12	5
Indonesia	8	24	7	14	12	4
Singapore	13	38	12	37	8	2
Thailand	11	28	8	21	10	3
Europe
Turkey	8	23	7	15	12	4
Denmark	21	28	20	30	4	3
Ireland	15	32	17	24	6	2
Norway	20	30	24	32	4	2
Sweden	24	31	22	31	4	2
United Kingdom	21	29	21	30	4	3
Greece	23	37	16	26	4	2
Italy	26	41	21	37	3	1
Portugal	23	36	17	27	4	2
Austria	23	37	20	35	4	2
Belgium	23	33	21	32	4	2
France	21	33	23	33	4	2
Germany	25	35	18	35	3	2
Netherlands	20	31	19	32	5	2
Switzerland	22	34	21	37	4	2
North America and Pacific
Canada	18	32	20	31	5	2
United States	17	26	21	28	5	3
Australia	18	30	20	29	5	3
New Zealand	17	30	20	31	5	3

Source: United Nations, Department of Economic and Social Affairs, Population Division (2006) Note: The support ratio is defined as the ratio of the population aged 15-64 to the population aged 65 and over.

View Table

Table 3.

Selected Countries: Projected Shares of Elderly and the Support Ratio 2000-2050

	Percentage of Total Population Over Age 60		Percentage 80 Years or Older		Potential Support Ratio 3
Country or area	2006	2050	2006	2050	2006	2050
Asia
China	11	31	10	23	9	3
Japan	27	42	19	37	3	1
Republic of Korea	14	41	10	31	7	2
India	8	21	10	16	12	5
Indonesia	8	24	7	14	12	4
Singapore	13	38	12	37	8	2
Thailand	11	28	8	21	10	3
Europe
Turkey	8	23	7	15	12	4
Denmark	21	28	20	30	4	3
Ireland	15	32	17	24	6	2
Norway	20	30	24	32	4	2
Sweden	24	31	22	31	4	2
United Kingdom	21	29	21	30	4	3
Greece	23	37	16	26	4	2
Italy	26	41	21	37	3	1
Portugal	23	36	17	27	4	2
Austria	23	37	20	35	4	2
Belgium	23	33	21	32	4	2
France	21	33	23	33	4	2
Germany	25	35	18	35	3	2
Netherlands	20	31	19	32	5	2
Switzerland	22	34	21	37	4	2
North America and Pacific
Canada	18	32	20	31	5	2
United States	17	26	21	28	5	3
Australia	18	30	20	29	5	3
New Zealand	17	30	20	31	5	3

Source: United Nations, Department of Economic and Social Affairs, Population Division (2006) Note: The support ratio is defined as the ratio of the population aged 15-64 to the population aged 65 and over.

Population aging, combined with a changing epidemiological pattern, will pose two other issues: pressures for a change in the health care delivery system and the likelihood of a more rapid rise in health care costs. Chronic diseases have increased dramatically and these particularly afflict the elderly. The elderly also demand different kind of health services, especially those who suffer from functional disability, dementia, and Alzheimer’s disease. The organization of health care delivery will inevitably have to be reformed. Housing, social services, and nursing home care will need to be integrated with medical services. Also, as noted above, the growing share of the elderly may lead to a more rapid increase in health expenditure, as health spending rises with age.

A new pattern of medical expenditure has also emerged from changes in the prevalence of chronic illnesses and their treatments. Medical expenditure is becoming less concentrated on a small number of acutely ill patients. Previously, researchers found that the last episode of illness consumed a huge share of an individual’s lifetime health expenditure, as costly medical technologies were used to prolong lives (though for only a short time). As a result, aging may have had only a modest effect on the rapid rise in health care costs. Recent trends now suggest that the concentration of medical expenditure on a minuscule group of the population has been reduced, perhaps because new, though costly, medical technologies have been developed that prolong lives rather than rescue patients from immediate death.

Moreover, affluence has changed diets and lifestyle. As a result, obesity has emerged as a serious health problem, and may prove to be a factor leading to widespread and serious long term chronic medical problems such as diabetes, high blood pressure, and cardiovascular illness (see Olshansky et al, 2005; Olshansky, 2005; Daviglus, 2005; and Lakdawalla et al, 2005).

C. Challenges Facing Middle-Income Countries

Most MICs are faced with a “double disease” burden. As a nation’s economy develops and matures, it has more knowledge and resources to address problems of clean water, sanitation, malnutrition, communicable diseases, and basic health care. As infectious diseases and infant and maternal mortality decrease in a nation, middle and higher income urban households suffer more from chronic illnesses while lower-income households and the rural population continue to suffer primarily from infectious diseases. Moreover, as a nation’s economy grows larger, policymakers and the public become more aware of the disparity in the availability of health care and the reality that the burden of unforeseen health expenditure has been a critical factor leading to impoverishment.

MICs thus increasingly recognize that they have to confront this double disease burden. They must transform their preventive and curative care delivery systems to deal with chronic diseases, while at the same time maintaining vigorous efforts to address the burden of infectious diseases. The double disease burden has serious implications for resource allocation, health costs, medical education, and the organization of health care delivery.

Also characterizing developments in MICs is the increasing demand by formal sector workers and upper middle-class citizens for health insurance. There is growing pressure to rationalize and reform the fragmented health financing schemes that normally characterize LICs. There is demand for the elaboration of a more universal system whereby all citizens can have access to basic health care and equitable risk protection. To meet these demands, MICs will have to increase their investments in health care. Tax revenue can be a source of financing. However, most nations will probably look to social health insurance as a potentially more desirable and politically viable approach. Either approach will have macroeconomic implications.

Another critical issue confronting MICs is that public health facilities are increasingly seen as inadequate as a consequence of the emergence of for-profit private sector providers. Public health service providers tend to be slow in responding to socioeconomic change and in raising the quality of health services to meet the demands of a rapidly emerging middle class. Typically, such facilities are under-funded, with resources allocated more on the basis of planning formulae rather than demand factors. Management operates under out-dated bureaucratic rules, and health workers are often organized in strong public service unions (e.g., in Latin America.) Consequently, a quality-of-service gap develops between supply and demand.

The for-profit private providers move in to deliver services that meet the demands of the middle class. Besides high returns to investor-owned private facilities, physicians in private practice can earn much higher incomes from fee-for-service payment paid by upper income and middle class patients. As a result, many public sector-employed physicians set up a dual practice, operating private clinics during off-duty hours. This can lead to a further decline in the quality of the public health service as physicians’ absenteeism rises to a high level and many highly qualified senior physicians leave for full-time private practice. Health care becomes a markedly two-tiered system, with the public health services losing the political support of the upper and middle classes. The inequity in the health system becomes apparent, creating social and political unrest.

Aging is also likely to pose a problem for some MICs in Asia, particularly China. The demographic transition—longer life expectancy and lower fertility rate—would eventually create an aging problem for any nation in the long-term. When a nation goes through an accelerated demographic transition, the time horizon of the aging process becomes considerably shortened. By adopting a one-child policy, China created an aging problem which should intensify by 2030. China and other accelerated-aging Asian middle-income countries now have to begin planning for how to finance the future high costs of long-term care and transform their health care delivery systems to meet the needs of the elderly (see Heller, 2006).

D. Challenges Facing Low-Income Countries

LICs are not progressing very well in the health arena. Prospects are poor for meeting the MDG goals. To meet the MDG of reducing under-five mortality would require mortality rates to fall on average by 4.3 percent per annum, in contrast to the recent experience of annual reductions of 2.3 percent. The maternal mortality rate in LICs is falling by only 2.4 percent a year compared with the 5.4 percent annual reduction that would be required to achieve the MDG target.¹⁰ In sub-Saharan Africa, no country appears on track to attain the MDG targets for under-five mortality or maternal mortality. Low-income countries face several common issues. Among them are: HIV/AIDS, inadequate fiscal space, and limited human resources for health.

Many LICs have been particularly affected by HIV/AIDS, with profound macroeconomic effects on labor supply, human capital, foreign investments and economic growth. Currently, African nations suffer the most from HIV/AIDS (UNAIDS (2004, 2006), Haacker (2004)). A human tragedy, HIV/AIDS pandemic has created havoc in the health system of many nations, particularly in Africa. Financial and human resources have been drawn away to treat and prevent this disease and the emphasis on addressing the illness has in some cases severely diminished countries’ capacity to provide basic health care services to its population. A summary of studies described the macroeconomic impact of HIV/AIDS is provided in Table 4.

Table 4.

Summary of Studies of the Macroeconomic Impact of HIV/AIDS in Africa

Sources: UN Economic and Social Affairs Department, Population Division (2004). NOTES: References to effect on GDP growth or per capita GDP growth rates refer to average annual growth rates for the period mentioned, expressed as percentage-point differences from a “No AIDS” scenario.

^aFor ‘extreme” assumption about future AIDS prevalence.

Table 4.

Summary of Studies of the Macroeconomic Impact of HIV/AIDS in Africa

Study	Countries and Period of Economic Data	Period of Most Recently Used HIV/AIDS Data	Results (comparison with non-HIV/AIDS scenario)
			Growth of GDP	Growth of GDP per capita
Dixon, McDonald and Roberts (2001)	41 countries (1960-98)	Late 1990s	GDP growth rates reduced by 2-4% a year; large variation across countries, in line with prevalence of HIV
World Bank (2001b)	Swaziland	Early 1990s	Average annual growth rate of GDP during 1991-2015 will be 1.3% lower	Average annual growth rate of GDP per capita during 1991-2015 will be 0.2% higher
World Bank (2001a)	Namibia	Early 1990s	Average annual growth rate of GDP in 1991-2015 will be 0.8% lower	Average annual growth rate of GDP per capita during 1999-2015 will be 0.1% higher
World Bank (2000)	Lesotho	Early 1990s	Average annual growth rate of GDP during 1999-2015 will be 1.4% lower	Average annual growth rate of GDP per capita during 1999-2015 will be 0.3% lower
Bonnel (2000)	About 50 countries (1990-1997)	Mid 1990s		Rate of growth of GDP per capita in Africa reduced by 0.7% per year in the 1990s (1.2% for a country with HIV prevalence of 20%)
Quattek and Fourie (2000)	South Africa	Mid 1990s	Average rate of GDP growth over next 15 years will be 0.3-0.4% lower per year
Arndt and Lewis (2000)	South Africa	–	Annual growth rate of GDP is lowered by about 0.5% in the late 1990s, rising to 2.5-2.6% during 2008-2010	GDP per capita will be 8% lower in 2010 than in the absence of AIDS; implies that AIDS lowers average annual growth rate of GDP per capita by 0.7% during 1997-2010
Greener, Jefferis and Sifambe (2001)	Botswana	Late 1990s	During 1996-2021, annual growth rate of GDP reduced by 1.1-2.1%, 1.5% in the scenario considered most likely	Little effect: annual per capita GDP growth rate between 0.6% lower and 0.4% higher due to AIDS; 0.1% lower in the scenario considered most likely
BIDPA (2000a)	Botswana	Late 1990s	Average rate of growth of GDP in 2000-2010 reduced by 1.5% per year
Bloom and Mahal (1995)	51 countries (1980-1992)	Early 1990s	Statistically insignificant effect on income growth
Cuddington and Hancock (1994)	Malawi	Early 1990s	Average rate of growth of GDP in 1985-2010 reduced by up to 1.5% per year	Average growth of per capita GDP reduced by up toa0.3% per year
Cuddington (1993a, 1993b)	United Republic of Tanzania	Early 1990s	Average annual rate of growth of GDP in 1985-2010 reduced by up to 1.1%	Average annual growth reduced by up to 0.5%
Kambou, Devarajan and Over (1992)	Cameroon	–	GDP growth rate over 1986-1991 reduced by 1.9% per year
Over (1992)	30 sub-Saharan countries	Early 1990s	Average annual growth rate of GDP during 1990-2025 reduced by 0.9% on average (up to 1.5% in 10 worst affected countries)	Average annual growth rate of GDP per capita reduced by 0.15% per year (up to 0.6% in 10 worst affected countries)

Sources: UN Economic and Social Affairs Department, Population Division (2004). NOTES: References to effect on GDP growth or per capita GDP growth rates refer to average annual growth rates for the period mentioned, expressed as percentage-point differences from a “No AIDS” scenario.

^aFor ‘extreme” assumption about future AIDS prevalence.

View Table

Table 4.

Summary of Studies of the Macroeconomic Impact of HIV/AIDS in Africa

Study	Countries and Period of Economic Data	Period of Most Recently Used HIV/AIDS Data	Results (comparison with non-HIV/AIDS scenario)
			Growth of GDP	Growth of GDP per capita
Dixon, McDonald and Roberts (2001)	41 countries (1960-98)	Late 1990s	GDP growth rates reduced by 2-4% a year; large variation across countries, in line with prevalence of HIV
World Bank (2001b)	Swaziland	Early 1990s	Average annual growth rate of GDP during 1991-2015 will be 1.3% lower	Average annual growth rate of GDP per capita during 1991-2015 will be 0.2% higher
World Bank (2001a)	Namibia	Early 1990s	Average annual growth rate of GDP in 1991-2015 will be 0.8% lower	Average annual growth rate of GDP per capita during 1999-2015 will be 0.1% higher
World Bank (2000)	Lesotho	Early 1990s	Average annual growth rate of GDP during 1999-2015 will be 1.4% lower	Average annual growth rate of GDP per capita during 1999-2015 will be 0.3% lower
Bonnel (2000)	About 50 countries (1990-1997)	Mid 1990s		Rate of growth of GDP per capita in Africa reduced by 0.7% per year in the 1990s (1.2% for a country with HIV prevalence of 20%)
Quattek and Fourie (2000)	South Africa	Mid 1990s	Average rate of GDP growth over next 15 years will be 0.3-0.4% lower per year
Arndt and Lewis (2000)	South Africa	–	Annual growth rate of GDP is lowered by about 0.5% in the late 1990s, rising to 2.5-2.6% during 2008-2010	GDP per capita will be 8% lower in 2010 than in the absence of AIDS; implies that AIDS lowers average annual growth rate of GDP per capita by 0.7% during 1997-2010
Greener, Jefferis and Sifambe (2001)	Botswana	Late 1990s	During 1996-2021, annual growth rate of GDP reduced by 1.1-2.1%, 1.5% in the scenario considered most likely	Little effect: annual per capita GDP growth rate between 0.6% lower and 0.4% higher due to AIDS; 0.1% lower in the scenario considered most likely
BIDPA (2000a)	Botswana	Late 1990s	Average rate of growth of GDP in 2000-2010 reduced by 1.5% per year
Bloom and Mahal (1995)	51 countries (1980-1992)	Early 1990s	Statistically insignificant effect on income growth
Cuddington and Hancock (1994)	Malawi	Early 1990s	Average rate of growth of GDP in 1985-2010 reduced by up to 1.5% per year	Average growth of per capita GDP reduced by up toa0.3% per year
Cuddington (1993a, 1993b)	United Republic of Tanzania	Early 1990s	Average annual rate of growth of GDP in 1985-2010 reduced by up to 1.1%	Average annual growth reduced by up to 0.5%
Kambou, Devarajan and Over (1992)	Cameroon	–	GDP growth rate over 1986-1991 reduced by 1.9% per year
Over (1992)	30 sub-Saharan countries	Early 1990s	Average annual growth rate of GDP during 1990-2025 reduced by 0.9% on average (up to 1.5% in 10 worst affected countries)	Average annual growth rate of GDP per capita reduced by 0.15% per year (up to 0.6% in 10 worst affected countries)

Sources: UN Economic and Social Affairs Department, Population Division (2004). NOTES: References to effect on GDP growth or per capita GDP growth rates refer to average annual growth rates for the period mentioned, expressed as percentage-point differences from a “No AIDS” scenario.

^aFor ‘extreme” assumption about future AIDS prevalence.

Table 5.

Progress on MDGs in Fragile States Compared to Other Poor Countries, 2000

Source: DFID (2005) “..” indicates that data are not available. Note 1: DFID’s definition of fragile status covers those “countries where the government cannot or will not deliver core functions to the majority of its people, including the poor.”

Table 5.

Progress on MDGs in Fragile States Compared to Other Poor Countries, 2000

		Low Income Fragile States	Other low and Middle income states
Population		871m	4361m
MDG1: Number living on less than $1 a day		343m	821 m
MDG2: Primary education enrollment		33%	15%
MDG3 Primary education: female: male enrollment ratio		0.84	0.92
MDG4: Child mortality rate per 1,000 (2002)		138	56
MDG5: Maternal mortality rate per 100,000		734	270
MDG6: # of people living with HIV/AIDS		17.1	21.4
	Malaria death rate per 100,000	90	7
MDG7: Proportion of Population without access to safe water		38%	18%
MDG8: Telephone and cellphone subscriptions per 100 people		4.5	18.8

Source: DFID (2005) “..” indicates that data are not available. Note 1: DFID’s definition of fragile status covers those “countries where the government cannot or will not deliver core functions to the majority of its people, including the poor.”

View Table

Table 5.

Progress on MDGs in Fragile States Compared to Other Poor Countries, 2000

		Low Income Fragile States	Other low and Middle income states
Population		871m	4361m
MDG1: Number living on less than $1 a day		343m	821 m
MDG2: Primary education enrollment		33%	15%
MDG3 Primary education: female: male enrollment ratio		0.84	0.92
MDG4: Child mortality rate per 1,000 (2002)		138	56
MDG5: Maternal mortality rate per 100,000		734	270
MDG6: # of people living with HIV/AIDS		17.1	21.4
	Malaria death rate per 100,000	90	7
MDG7: Proportion of Population without access to safe water		38%	18%
MDG8: Telephone and cellphone subscriptions per 100 people		4.5	18.8

Source: DFID (2005) “..” indicates that data are not available. Note 1: DFID’s definition of fragile status covers those “countries where the government cannot or will not deliver core functions to the majority of its people, including the poor.”

By virtue of the fact of low per capita incomes, LIC governments also lack the financial resources to provide a minimal package of health care services. The CMH report sought to establish a reasonable estimate for the minimum amount of spending required to provide basic preventive and curative services. Compared to its estimate of around $30 per capita, one can find few LICs that annually spend even half this amount on health on a per capita basis, and most governments annually spend only $6-10 per capita. While there may be some room for additional government spending on health—by rationalizing unproductive expenditure, by commercializing some government services, by raising the government revenue share to at least 15-18 percent of GDP, and by some additional borrowing, the amount of effective “fiscal space” for health created by such measures is limited because of competition with other pressing priorities in other sectors (Heller (2006), Foster (2005)).

The stark reality is that to minimally provide basic health services, most LICs will need external assistance for some time (preferably through long-term sectoral budget support in the form of grants). While recent years have seen a dramatic increase in funding for health by the donor community, much of these resources have been targeted to vertical disease programs—through the efforts of bilateral donors to finance the Global Fund to Fight HIV/AIDS, Malaria, and Tuberculosis as well as from recent US government initiatives with respect to HIV/AIDS and malaria initiatives. As noted above, these vertical initiatives have, in anything, actually reduced the available fiscal space for spending on other preventive and basic curative services.

The recent surge in funding for HIV/AIDS has also highlighted the severe limitations of LICs in terms of available human resources for health. The recent report of the Joint Learning Initiative (2004) suggested that sub-Saharan Africa is short approximately 1 million health workers (largely doctors, nurses, and midwives, but also including laboratory workers, pharmacists, and trained community health workers) to provide basic health services. LICs are bedeviled by limits in their capacity to train doctors and nurses as well as by poor conditions of services—low salaries, poor opportunities for medical training, severe and risky working conditions in health clinics (in part because of the AIDS epidemic), and difficult living conditions for doctors and nurses posted with their families to rural health clinics. They also find that many of the doctors that they have trained have emigrated to either industrial countries or to externally-funded clinics in other LICs (to work at higher than public service salaries in AIDS clinics).

Besides these common issues, there are many fragile states among the LICs. The World Bank identifies 34 such countries as ‘Low Income Countries Under Stress’ (LICUS), representing 418 million people. The populations of these countries suffer from poor health and the failed health systems of these countries typically result in households, when struck by serious illness, being forced into impoverishment as they deplete their meager resources in seeking health care. The causes for their health problems and policy remedies defy generalization, but include:

• countries in armed conflict or in the early stages of a peace process;

• countries with a history of political instability and military coups;

• new or emerging states;

• countries under international sanctions;

• countries with closed political systems.

Not only are LICUS countries failing to meet MDG targets but, as conveyed in Table 6, they are about 25-35 percent worse off than low-income countries in terms of per capita income, per capita total expenditure on health, infant and maternal mortality rates, and the incidence of TB.

Table 6.

The Objectives of a Health System

Table 6.

The Objectives of a Health System

	Objectives (Outcomes)
Dimensions of Outcomes	Health Status	Risk Protection	Consumer Satisfaction
Average level
Degree of equity in the distribution

View Table

Table 6.

The Objectives of a Health System

	Objectives (Outcomes)
Dimensions of Outcomes	Health Status	Risk Protection	Consumer Satisfaction
Average level
Degree of equity in the distribution

A. Ends: Objectives

The objectives of a national health system can be ascertained from a country’s legislative history and policy documents. There is a convergence of views around the world about the goals, viz., to improve people’s health, prevent health-related impoverishment, and gain public support and satisfaction. The equitable distributions of these benefits have to be considered along with their average levels. At the same time, since resources are scarce and nations are constrained to optimize these goals subject to a budget constraint, affordability is often stated as an important health system goal. These objectives can be decomposed into their equity and efficiency dimensions, in effect going beyond the usual economic concerns that exclusively focus on efficiency (Okun, 1974). Table 7 provides a summary of these multiple objectives and also illustrates the possible trade-offs between them.

Table 7.

Health Care Financing and Service Provision, by Stage of Economic Development

(Percentage shares relate to proportion of population in each category of coverage)

^*2004 GNI, using the World Bank’s Atlas method, which smoothes exchange rate fluctuations by using a three-year moving average, price adjusted conversion factor.

Table 7.

Health Care Financing and Service Provision, by Stage of Economic Development

(Percentage shares relate to proportion of population in each category of coverage)

	Stage I	Stage II	Stage III	Stage IV
Country classification	Low income	Lower middle income	Upper middle income	High income
per capita GNI*	below $826	$826 -$3,255	$3,256 -$10,065	Above $10,065
Financing sources
General revenue and donor aid	50-60%	40-50%	20-40%	Countries with a high share of general revenue: National Health System (United Kingdom, New.Zealand.), Medisave and catastrophic insurance in Singapore
Social insurance	Only for civil servants and formal sector employees	10-20%	30-60%	Direct provision: National Health Insurance (Canada & Australia) Indirect provision: Bismarkian (Germany, Japan)
Private insurance	Little	5-10%	15-40%	Managed care plus Medicare (United States), or supplementary in many countries
Self pay	35-45%	20-40%	15-25%	15-25%

^*2004 GNI, using the World Bank’s Atlas method, which smoothes exchange rate fluctuations by using a three-year moving average, price adjusted conversion factor.

View Table

Table 7.

Health Care Financing and Service Provision, by Stage of Economic Development

(Percentage shares relate to proportion of population in each category of coverage)

	Stage I	Stage II	Stage III	Stage IV
Country classification	Low income	Lower middle income	Upper middle income	High income
per capita GNI*	below $826	$826 -$3,255	$3,256 -$10,065	Above $10,065
Financing sources
General revenue and donor aid	50-60%	40-50%	20-40%	Countries with a high share of general revenue: National Health System (United Kingdom, New.Zealand.), Medisave and catastrophic insurance in Singapore
Social insurance	Only for civil servants and formal sector employees	10-20%	30-60%	Direct provision: National Health Insurance (Canada & Australia) Indirect provision: Bismarkian (Germany, Japan)
Private insurance	Little	5-10%	15-40%	Managed care plus Medicare (United States), or supplementary in many countries
Self pay	35-45%	20-40%	15-25%	15-25%

^*2004 GNI, using the World Bank’s Atlas method, which smoothes exchange rate fluctuations by using a three-year moving average, price adjusted conversion factor.

Achieving multiple goals under a budget constraint requires difficult trade-offs. Each nation has to wrestle with two types of trade-offs: inter-sectoral and intra-sectoral. In deciding how much to spend in total on health (i.e., the budget constraint for the whole sector), governments also have to consider the relative benefits produced by spending on nonhealth sectors—education, housing, environmental protection, research, and defense both in term of improved health outcomes as well as other societal objectives.

Intrasectoral tradeoffs must also be considered in choosing how to achieve different goals within the health system. For example, at the margin, there is a tradeoff between improving the average level of a country’s health status and achieving an equitable distribution of health status. Often the major policy debate may overlook key tradeoffs. For example, the common argument given by economists—that increasing direct OOP by patients will improve economic efficiency—ignores the possible consequence of a greater inequality in health status for the poor. But rarely are these tradeoffs explicit or obvious to citizens of a country. The implicit boundaries to trading off among different objectives exist in deeply rooted historical processes as well as in fundamental social values, which limit the range of available reform options. The health systems of European countries, for example, are deeply rooted in egalitarian traditions, and policy proposals violating this basic foundation of equity have little overall appeal regardless of how much they would enhance efficiency (Saltman and Figueras, 1997). On the other hand, the health care system of the US is rooted in libertarian traditions. Compulsory health insurance to cover all Americans remains elusive after more than 60 years of public debate (Marmor and Barr, 1992).

The most frequent debate is about the trade-off between efficiency and equity within the health system. For examples, at the margin, every nation has to trade off between excluding coinsurance to advance equal access to health care and imposing coinsurance to promote efficiency; or between investing in the most advanced expensive cancer treatments to improve the average level of health and building hospitals in remote areas to improve equity in health.

B. Means: Structural Elements

In the architecture of a health system, there are five major causal structural elements (the means), as identified by previous research, to achieve these goals: financing, organization, payment, regulation, and persuasion (Hsiao, 1999, Roberts et al. 2003.) We provide a brief summary of each:

Summary of Ends and Means

Chart 2 illustrates the relationship between the structural elements (means) and outcomes (ends.) We can use this analytical framework to analyze and explain how the change in outcomes arising from the introduction of a NHI system are likely to be influenced by the characteristics of its structure in terms of the five means described above.

View Full Size

Means, Intermediate, and Final Ends of a Health System

Citation: IMF Working Papers 2007, 013; 10.5089/9781451865776.001.A001

• Download Figure
• Download figure as PowerPoint slide

C. Control Knobs for Policy

A. Characterizing Different Stages of Health System Development

This chapter presents an analytical framework to help macroeconomists understand health systems and assess health policy for countries at different stages of development. To state the obvious, the same health system structure cannot be applied to all countries. Systems differ enormously among countries, due to variations in socioeconomic development. What works in the UK, say, may not work in Kenya. On the other hand, must every country be treated differently? Or can they be grouped into somewhat homogenous categories, and general conclusions drawn for each?

Using cluster analysis, we can identify three “distinct” groups of countries. For ease of understanding, we distinguish the groups of countries by per capita GDP. Of course, the boundaries between these groups are artificial, because the distribution of countries by income is continuous. Thus, although we term each grouping a “stage of health development,” this term should not be misinterpreted as referring to discrete stages. The health system also evolves because as income and disease patterns change, so does the health system gradually change. Table 8 presents examples of countries in each of the three stages of health development and summarizes the financing and service provision in each stage. A more detailed description for each stage follows the table.

Table 8.

Economic Development Stage I. Three-Tier Health Care System

Table 8.

Economic Development Stage I. Three-Tier Health Care System

	Source of Financing	Services Provided
Affluent households	Self-pay	Disease prevention and drugs Primary care and secondary services fromprivate sector
	Government	Tertiary services public health
Middle-income households, civil servants and formal sector employees Others	Social insurance	Disease prevention and drugsprimary care and secondary services from theprivate sector
	Self-pay	Disease prevention and drugsprimary care and secondary services from theprivate sector
	Government	Tertiary services public health
	Self-pay	For minor illnesses: self-care, self-medication, and indigenous and private practitioners
Poor and low-income households	Government	Primary care, public health and disease prevention. For serious illnesses: specialist services and in-patient hospital services

View Table

Table 8.

Economic Development Stage I. Three-Tier Health Care System

	Source of Financing	Services Provided
Affluent households	Self-pay	Disease prevention and drugs Primary care and secondary services fromprivate sector
	Government	Tertiary services public health
Middle-income households, civil servants and formal sector employees Others	Social insurance	Disease prevention and drugsprimary care and secondary services from theprivate sector
	Self-pay	Disease prevention and drugsprimary care and secondary services from theprivate sector
	Government	Tertiary services public health
	Self-pay	For minor illnesses: self-care, self-medication, and indigenous and private practitioners
Poor and low-income households	Government	Primary care, public health and disease prevention. For serious illnesses: specialist services and in-patient hospital services

For each stage of development, it is important to understand the history behind the variations in level of health status and risk-protection. Most non-Western countries’ governments paid little attention to health care until Western medicine was introduced—largely by missionaries and colonial governments. After World War II, colonial powers (the UK, France, Germany, the US, and the USSR) introduced their systems of health care financing and organization into their colonies and spheres of influence. However, these systems were used only in the application of Western medicine (e.g., in the training and certification of health practitioners, the regulation of drugs, and the provision of government-financed services). Patients continued to use indigenous medicine in combination with Western medicine.

B. Stages I and II: Low-Income and Lower-Middle-Income Countries

Low income countries (per capita GDP below $826). The 59 countries in this group include inter alia, Bangladesh, Haiti, India, Kenya, Mali, Nigeria, Peru, Senegal, Tanzania, Vietnam, and Yemen.

Lower-middle income countries ($826-$3,255 per capita GDP). The 54 countries in this group include, inter alia, China, Ecuador, Egypt, Indonesia, Morocco, Peru, Philippines, and Tunisia.

In the early stage of health system development, before state action, people rely on their own knowledge of hygienic and preventive practices. When they become ill, they largely rely on self-care, including self-medication. Patients also seek services from indigenous health care practitioners and pay out-of-pocket. As Western medicine enters a country, the government plays an ever-increasing role in health care. Stage I countries share many of the following characteristics with, of course, several glaring exceptions.

1. General description

Disease patterns

• For LICs, communicable and infectious diseases are prevalent and are the major cause of death. A large portion of the population has no access to clean water and latrines. A high percentage of infants and children suffers from malnutrition, and many pregnant mothers are anemic. Other characteristics include a high rate of infant mortality (ranging from 70 to 130 per 1,000 live births) and a low life expectancy (ranging from 45 to 65 years).

• For LICs, the urban population begins an epidemiological transition; limited public resources have to deal simultaneously with communicable and chronic diseases.

Health spending, the use of resources, and health outcomes

• In LICs and lower MICs, the government finances a large portion of national health expenditure, but a significant share—often close to half—is paid directly by patients. However, government statistics on national health expenditure usually understate the amount spent, because they omit or underestimate private spending. Figure 5 shows the difference between actual national health expenditure and official/government statistics.

• In general, Stage I countries allocate only a small portion of their government budgets (less than $2 per capita, on 1997 PPP basis) to public health and prevention. Two-thirds or more of the public health budget is spent on hospital services. The principal medical center, which is usually located in the capital city, receives half the total hospital budget. Although most of the population lives in rural areas, most of the public health budget is spent on services used by the urban population.

• Health outcomes vary greatly among these countries. For example, although India and Sri Lanka are neighbors, their infant mortality rates are respectively 73 and 17 per 1,000 live births, respectively. The difference does not reflect the level of lie in health spending since Sri Lanka spends less of its GDP on health than India. China also spends less of its GDP on health than India, yet its rate of 42 per 1,000 live births is substantially better than that of India.

View Full Size

Selected Countries: Comparison of Actual National Health Expenditure and Government Statistics on Health Expenditure

(As percent of GDP)

Citation: IMF Working Papers 2007, 013; 10.5089/9781451865776.001.A001

Source: Berman and others, 1999.

• Download Figure
• Download figure as PowerPoint slide

Government capacity and performance

• In all Stage I countries, central government revenue (as a percent of GDP) is relatively low—typically below 15 percent. These governments, therefore, have very limited resources to fund health care (see Figure 6).

• Health issues are usually given a low priority by Stage I countries, resulting in a lack of national strategy for health development. For example, governments usually divorce their policies with respect to medical education and the pharmaceutical industry from their policies for the health sector. Health spending by ministries of education are often as large as those of ministries of health, because the former funds and manages the medical centers that provide clinical training to medical graduates. This spending, however, may have little impact on improving the level of a population’s health status.

• Public health services are inefficient: excess capacity is often found at lower-level facilities. Using quantitative and econometric models, studies of several LICs have found the inefficiency to be between one-third and one-half of spending (Osei et al, 2005 and Zere et al., 2006).

View Full Size

Government Revenue as a Fraction of GDP, by Average per Capita Income Level

Citation: IMF Working Papers 2007, 013; 10.5089/9781451865776.001.A001

Source: IMF 2002

• Download Figure
• Download figure as PowerPoint slide

Supply and use of services

• Although public health and disease prevention programs sponsored and funded by donors may be especially effective, other public health programs are often underfunded and poorly managed.

• For minor illnesses, most of the population relies on self-care and self-medication. In rural areas, most of the governments in these countries establish and fund health stations, staffing them with modestly trained health practitioners or recent medical school graduates. However, in fact, these primary care services exist only on paper. Those that do exist offer a poor quality of service, inadequate drug supplies, and inconvenient clinical hours. Many health stations are staffed with two health practitioners and see only three or four patients on an average day. Patients often prefer to go to an indigenous medical practitioner and pay out-of-pocket. In urban areas, patients generally go to private practitioners. Surveys commonly find that urban patients obtain more than half their outpatient services from the private sector.

• For serious illnesses, the affluent seek hospital services from the private sector, when available. Most, however, use the free (or heavily subsidized) government hospital services. These serve as de facto health insurance for most people. These services, however, are rationed by waiting time and by poor service quality. Frequently, the experienced hospital medical staff hold two jobs—in a hospital and in a private fee-for-service practice. In many of these countries, physicians earn the vast majority of their income from private practice, spending only a modest amount of time at the public hospital, the rest at their private practice. In public hospitals, physicians self-refer the more affluent patients to their private clinics.

2. A three-tier system in terms of access

This section describes the common characteristics of health care for countries in the first stage of development in order to give policymakers a general understanding of health care systems. However, as with any generalization, there are exceptions and variations.

Stage I governments accept responsibility for providing all necessary health care for their citizens. But do these governments allocate sufficient funds for health programs? And how effectively are these services delivered?

Most of these countries attach insufficient priority in the allocation of budgetary resources to public health, disease prevention, and maternal and child health services. When international donors support these programs, most governments establish vertical programs to deliver the specified services. Common donor-funded preventive or primary services include programs for extended disease prevention and immunization, maternal and child health, family planning, and treatment for malaria, TB, and HIV/AIDS. Each vertical program creates its own bureaucracy, clinics, and supply system. These programs often overlap, competing for the limited number of trained health practitioners and equipment. The sustainability of these programs is therefore in jeopardy once the donors withdraw their funding.

On paper, many of these governments also take responsibility for organizing, managing, and delivering primary, secondary, and tertiary curative services to all citizens. However, few governments can fund and deliver these services adequately. Consequently, the financing and provision of health services is effectively segmented across three tiers of the population, depending largely on a patient’s ability to pay for services.

The first tier consists of affluent households who pay directly and demand services from better-qualified, private sector physicians.¹¹ These affluent households also demand secondary in-patient hospital services from private hospitals (both for-profit and not-for-profit). Since private physicians and hospitals charge high fees for their services—these services are therefore rationed by price. However, affluent households demand sophisticated and expensive tertiary services from public teaching hospitals, which charge very low fees. These services are delivered by public hospitals partly because of the lack of private capital to invest in this level of facility and partly because the few qualified medical specialists are on the faculty of the medical schools. For other services, affluent households demand care from private sector providers rather than obtain the almost-free services from public providers because in the private sector there is greater availability of drugs and supplies, shorter waiting times, more choice of physicians and better personal attention and other amenities.

The second tier consists of middle-income households, which are covered by the government’s essentially limited pay-as-you-go social insurance system. This insurance system typically covers only civil servants and workers employed by large enterprises. The social insurance plan selects and contracts with public and private facilities to provide services. These facilities charge the plan on a fee-for-service basis. The fee schedule may be negotiated between the parties or set by the insurance plan. Tertiary services are contracted from medical centers. Services are rationed by waiting time and poor service quality (e.g., unfriendly health practitioners). Some large enterprises have their own hospitals and clinics, which are built to assure cost control and the availability of high-quality services.

The third tier consists of poor and low-income households (most of the population), which rely on public facilities for health services. Most poor households reside in rural areas or in urban slums. Although public health services are nearly free, the waits are often long, physicians are not on duty, clinic hours are inconvenient, facilities are dilapidated and crowded, drugs and other supplies are unavailable, and providers unfriendly. Therefore, when the problem is not life-threatening, such households often resort to self-care, self-medication, or indigenous practitioners. As a result, household expenditure surveys consistently find that poor and low-income households spend a significant portion of their income on drugs and indigenous medicines. Table 9 summarizes this three-tier system.

Table 9.

Selected Low-Income Countries: Health Expenditure and Results by Model

Source: World Bank 2006 Note: L, M, and H denote low, medium, and high, respectively.

Table 9.

Selected Low-Income Countries: Health Expenditure and Results by Model

				Percent of GDP Spent on Health		Health Status Indicators (2004)
		Per Capita GNI (2004 WB Atlas method)	Per Capita GDP (2004 PPP $US)	Total (2003)	Government (2003)	Life expectancy	Infant mortality rate (per 1,000)	Under-age- five mortality rate (per 1,000)	Risk Protection
Government-oriented model
	Bangladesh	440	1,875	3.4	1.1	63.5	56.4	77.0	M
	India	620	3,115	4.8	1.2	63.5	61.6	85.2	M
	Kenya	460	1,063	4.3	1.7	48.4	78.5	119.5	L
	Nigeria	390	1,113	5.0	1.3	43.7	101.4	196.6	L
	Sri Lanka	1,010	4,173	3.5	1.6	74.4	12.0	14.1	H
Market-oriented model
	Indonesia	1,140	3,583	3.1	1.1	67.4	29.6	38.4	L
	Philippines	1,170	4,558	3.2	1.4	70.8	26.0	34.4	M
Central-planning model
	China	1,290	5,495	5.6	2.0	71.4	26.0	31.0	L
	Vietnam	550	2,704	5.4	1.5	70.3	17.4	23.2	M

Source: World Bank 2006 Note: L, M, and H denote low, medium, and high, respectively.

View Table

Table 9.

Selected Low-Income Countries: Health Expenditure and Results by Model

				Percent of GDP Spent on Health		Health Status Indicators (2004)
		Per Capita GNI (2004 WB Atlas method)	Per Capita GDP (2004 PPP $US)	Total (2003)	Government (2003)	Life expectancy	Infant mortality rate (per 1,000)	Under-age- five mortality rate (per 1,000)	Risk Protection
Government-oriented model
	Bangladesh	440	1,875	3.4	1.1	63.5	56.4	77.0	M
	India	620	3,115	4.8	1.2	63.5	61.6	85.2	M
	Kenya	460	1,063	4.3	1.7	48.4	78.5	119.5	L
	Nigeria	390	1,113	5.0	1.3	43.7	101.4	196.6	L
	Sri Lanka	1,010	4,173	3.5	1.6	74.4	12.0	14.1	H
Market-oriented model
	Indonesia	1,140	3,583	3.1	1.1	67.4	29.6	38.4	L
	Philippines	1,170	4,558	3.2	1.4	70.8	26.0	34.4	M
Central-planning model
	China	1,290	5,495	5.6	2.0	71.4	26.0	31.0	L
	Vietnam	550	2,704	5.4	1.5	70.3	17.4	23.2	M

Source: World Bank 2006 Note: L, M, and H denote low, medium, and high, respectively.

5. Comparison of performance

Table 10 shows the outcomes by model for selected LICs. In terms of cost-effectiveness, the central-planning model clearly outperforms the other two models. The countries that adopted the central-planning model spent less of their GDP on health care but achieved better health outcomes and risk-protection for their citizens. Table 11 shows how well the three models performed in attaining health objectives.

Table 10.

Low-Income Countries: Comparison of Health Care System Performance

F, O, I, and R denote financing, macro-organization, incentives, and regulations, respectively. L, M, and H denote low, medium, and high, respectively.

Table 10.

Low-Income Countries: Comparison of Health Care System Performance

F, O, I, and R denote financing, macro-organization, incentives, and regulations, respectively. L, M, and H denote low, medium, and high, respectively.

View Table

Table 10.

Low-Income Countries: Comparison of Health Care System Performance

View Expanded

F, O, I, and R denote financing, macro-organization, incentives, and regulations, respectively. L, M, and H denote low, medium, and high, respectively.

Table 11.

Input and Output Indicators of a Best-Performing Low-Income Country: Sri Lanka

Source: World Bank Word Development Indicators 2006.

Table 11.

Input and Output Indicators of a Best-Performing Low-Income Country: Sri Lanka

INPUT INDICATORS		OUTPUT INDICATORS
		Aggregate			Equity in			Efficiency in
		Level of Health Status		Risk- Protection	Net Benefits	Access to Health Care	Risk- Protection	Health Care	Risk- Protection
		Infant Mortality Rate, (2004, per 1,000 live births)	Life Expectancy (2004)
Health expenditure per capita, (2003, $US, exchange rate)	31
Health expenditure, total (2003, as percent of GDP)	3.5
		12	74.4	H	H	H	M/H	H	H
Hospital beds (1999, per 1,000 people)	2.2
Physicians (2004 per 1,000 people)	0.5

Source: World Bank Word Development Indicators 2006.

View Table

Table 11.

Input and Output Indicators of a Best-Performing Low-Income Country: Sri Lanka

INPUT INDICATORS		OUTPUT INDICATORS
		Aggregate			Equity in			Efficiency in
		Level of Health Status		Risk- Protection	Net Benefits	Access to Health Care	Risk- Protection	Health Care	Risk- Protection
		Infant Mortality Rate, (2004, per 1,000 live births)	Life Expectancy (2004)
Health expenditure per capita, (2003, $US, exchange rate)	31
Health expenditure, total (2003, as percent of GDP)	3.5
		12	74.4	H	H	H	M/H	H	H
Hospital beds (1999, per 1,000 people)	2.2
Physicians (2004 per 1,000 people)	0.5

Source: World Bank Word Development Indicators 2006.

6. Guideposts for macroeconomists

In general, macroeconomists should rely on the guidance of the World Bank and WHO in appraising the performance of a country’s health care system and should support the policy advice offered by these institutions. A particularly useful source to draw upon is the work of the Disease Control Priorities Project, which has drawn on the expertise of a team of researchers drawn from the World Bank, academia, and the National Institutes of Health. Their recent report (Jamison et al, 2006) examines the key intrasectoral policy choices in the health sector of LICs, the choice of interventions to be delivered to the target population based on cost effectiveness criteria, and the relevance of policy choices in other sectors.¹² Some broad background, however, is useful for macroeconomists to be aware of:

• In assessing a health care system’s performance, macroeconomists can compare a country’s input and output indicators with those of a best-performing country and a median-performing country. The input indicators could be health care spending, available facilities, and human resources. The output indicators could be the level of health status, risk-protection, and the equity and efficiency of both. Table 12 gives a benchmark country for Sri Lanka—which we characterize as a “best” performing country and Table 13 gives a benchmark for Belize, which is more of a “median” performing country among these two stages of development.

• Broad recommendations: A country should take the following important policy measures in order not to repeat other countries’ mistakes:

  • Formulate a rational and coherent overall health sector policy, so that the allocation of resources can be rationalized and health-related programs in various ministries coordinated;

  • establish a cabinet-level committee, chaired by the deputy prime minister, to review and monitor health policy implementation;

  • formulate a list of essential drugs, so bulk purchases can be made through an international bidding process, and so that an organized, nationwide distribution system can be established;

  • Avoid a fee-for-service payment system for social and private insurance. This type of payment system increases health care cost inflationary pressures;

  • be cautious about accepting donor assistance for tertiary hospitals. Because the operations and maintenance costs at this level can absorb a significant portion of a country’s health budget, a careful assessment of the costs and benefits of accepting such assistance should be made; and

  • avoid situations whereby a ministry of health and/or social insurance plan funds and manages its own facilities. Ministries can easily become “captured” by ministry-employed health practitioners, resulting in their interests dominating that of patients.

Box 1.

Sri Lanka: A Country with Exceptional Performance

Among the Stage I countries, Sri Lanka stands out for its health achievements under the government-oriented health service model. In 1996, other countries in this category had infant mortality rates of 61-114 per 1,000 live births, while Sri Lanka had a rate of 17 per 1,000 and a life expectancy of 73 years, and spent only $2 per capita (on a 1996 PPP basis) for health care. These enviable accomplishments seem to be the result of two key factors. First, Sri Lanka has created a professional culture for its health care practitioners: they place professional commitments first, self-interest second. Despite low pay and poor working conditions, they are dedicated to their public health duties. Although most experienced public sector physicians have private practices on the outside and earn most of their income from these, they still work their full shift for the public facilities and perform their duties faithfully. They do so voluntarily, without tight management. Second, Sri Lanka has established a well-organized system at the village level, staffed by midwives who provide effective health education and basic primary care for mothers and children.

• Some elementary policy strategies—certain policy strategies will improve health care system performance, including the following:

  • introduce or increase excise taxes on tobacco;

  • avoid across-the-board cuts in health programs. If cuts are necessary, they should be made in tertiary and university hospitals, which are less cost-effective and do not offer risk-protection to low- and middle-income households; and

  • develop a coherent health financing plan that includes: developing social insurance to fund tertiary and hospital services for the urban population; shifting current government funds in order to subsidize the urban poor and rural populations; developing community-level financing for rural residents; targeting hospital subsidies to ward services; charging user fees for A-class services, based on their full cost plus a profit margin; and using the profits of such a system to cross-subsidize ward services.

Table 12.

Input and Output Indicators of a Median-Performing Low-Income Country: Belize

Source: World Bank, 1999a.

Table 12.

Input and Output Indicators of a Median-Performing Low-Income Country: Belize

INPUT INDICATORS		OUTPUT INDICATORS
		Aggregate			Equity in			Efficiency in
		Level of Health Status		Risk- Protection	Net Benefits	Access to Health Care	Risk- Protection	Health Care	Risk- Protection
		Infant Mortality Rate (Per 1,000 live births)	Life Expectancy
Health expenditure per capita, (1997 PPP basis)	158.0
Health expenditure, total (as percent of GDP)	4.7
		44.6	73.2	M	M	L	M	M	L
Hospital beds (per 1,000 people)	2.7
Physicians (per 1,000 people)	0.6

Source: World Bank, 1999a.

View Table

Table 12.

Input and Output Indicators of a Median-Performing Low-Income Country: Belize

INPUT INDICATORS		OUTPUT INDICATORS
		Aggregate			Equity in			Efficiency in
		Level of Health Status		Risk- Protection	Net Benefits	Access to Health Care	Risk- Protection	Health Care	Risk- Protection
		Infant Mortality Rate (Per 1,000 live births)	Life Expectancy
Health expenditure per capita, (1997 PPP basis)	158.0
Health expenditure, total (as percent of GDP)	4.7
		44.6	73.2	M	M	L	M	M	L
Hospital beds (per 1,000 people)	2.7
Physicians (per 1,000 people)	0.6

Source: World Bank, 1999a.

Table 13.

Upper-Middle-Income Countries: Health Care Expenditure and Results by Model

Source: World Bank 2006

^*WB classifies these as lower-middle income

Note: L, M, and H denote low, medium, and high, respectively.

Table 13.

Upper-Middle-Income Countries: Health Care Expenditure and Results by Model

				% of GDP spent on health		Health status indicators (2004)
		Per Capita GNI (2004 WB Atlas method)	Per Capita GDP (2004 PPP $US)	Total (2003)	Government (2003)	Life expectancy	Infant Mortality Rate (Per 1,000)	Under-Age-5 Mortality Rate (Per 1,000)	Risk Protection
Public health service model
	Malaysia	4650	9760	3.8	2.2	73.5	10.2	12.4	H
	Turkey	3750	7710	7.6	5.4	69.9	28.3	32.0	M
Social insurance model
	Colombia*	2000	7121	7.6	6.4	72.6	17.5	20.5	M
	Argentina	3720	12723	8.9	4.3	74.6	16.2	18.2	M
	Costa Rica	4670	9805	7.3	5.8	78.7	11.3	12.6	H
	Thailand*	2540	8179	3.3	2.0	70.5	18.2	21.2	H

Source: World Bank 2006

^*WB classifies these as lower-middle income

Note: L, M, and H denote low, medium, and high, respectively.

View Table

Table 13.

Upper-Middle-Income Countries: Health Care Expenditure and Results by Model

				% of GDP spent on health		Health status indicators (2004)
		Per Capita GNI (2004 WB Atlas method)	Per Capita GDP (2004 PPP $US)	Total (2003)	Government (2003)	Life expectancy	Infant Mortality Rate (Per 1,000)	Under-Age-5 Mortality Rate (Per 1,000)	Risk Protection
Public health service model
	Malaysia	4650	9760	3.8	2.2	73.5	10.2	12.4	H
	Turkey	3750	7710	7.6	5.4	69.9	28.3	32.0	M
Social insurance model
	Colombia*	2000	7121	7.6	6.4	72.6	17.5	20.5	M
	Argentina	3720	12723	8.9	4.3	74.6	16.2	18.2	M
	Costa Rica	4670	9805	7.3	5.8	78.7	11.3	12.6	H
	Thailand*	2540	8179	3.3	2.0	70.5	18.2	21.2	H

Source: World Bank 2006

^*WB classifies these as lower-middle income

Note: L, M, and H denote low, medium, and high, respectively.

C. Stage III: Upper-Middle-Income Countries (per capita GDP Incomes of $5,001-12,000, on 1997 PPP basis)

3. Comparison of performance

Table 14 compares the performance under the three models. The public health service model (e.g., Malaysia) seems to outperform the other two. The high spending and low outcome under the social insurance model may be particular to Latin America because social insurance has often proven to be very politicized in that region. Consequently, social-insurance-managed health facilities appear very inefficient and at times a source of corruption.¹³

Table 14.

Upper-Middle-Income Countries: Comparison of Performance by Model

F, O, I, R denote financing, macro-organization, incentives, and regulations, respectively. L, M, and H denote low, medium, and high, respectively

Table 14.

Upper-Middle-Income Countries: Comparison of Performance by Model

F, O, I, R denote financing, macro-organization, incentives, and regulations, respectively. L, M, and H denote low, medium, and high, respectively

View Table

Table 14.

Upper-Middle-Income Countries: Comparison of Performance by Model

View Expanded

F, O, I, R denote financing, macro-organization, incentives, and regulations, respectively. L, M, and H denote low, medium, and high, respectively

Table 15 compares the performance of the three models and analyzes how they affect the level and distribution of health status and risk-protection, as well as the cost-effectiveness of their systems. As for equal access to health care, the upper middle-income countries have the economic capacity to provide equal access to only “reasonable” health services. The affluent and upper-middle-income households do pay out-of pocket for services that provide better personal quality, such as convenience, amenities, and physician/hospital choice.

Table 15.

Input and Output Indicators of Two Best-Performing Upper-Middle-Income Countries

Source: World Bank World Development Indicators 2006a.

Table 15.

Input and Output Indicators of Two Best-Performing Upper-Middle-Income Countries

INPUT INDICATORS		OUTPUT INDICATORS
		Aggregate			Equity in			Efficiency in
		Level of Health Status
		Infant Mortality Rate, (2004, per 1,000 live births)	Life Expectancy (2004)	Risk- Protection	Net Benefits	Access to Health Care	Risk- Protection	Health Care	Risk- Protection
Part A: Malaysia	163
Health expenditure per capita, (2003, PPP basis)
Health expenditure, total (2003, as percent of GDP)	3.8
		10.2	73.5	H	M	H	M/H	M/H	M/H
Hospital beds (2001, per 1,000 people)	1.9
Physicians (2000, per 1,000 people)	0.7
Part B: Costa Rica
Health expenditure per capita, (2003, PPP basis)	305
Health expenditure, total (2003, as percent of GDP)	7.3
		11.3	78.7	H	H	M	H	M	H
Hospital beds (2003, per 1,000 people)	1.4
Physicians (2000, per 1,000 people)	1.3

Source: World Bank World Development Indicators 2006a.

View Table

Table 15.

Input and Output Indicators of Two Best-Performing Upper-Middle-Income Countries

INPUT INDICATORS		OUTPUT INDICATORS
		Aggregate			Equity in			Efficiency in
		Level of Health Status
		Infant Mortality Rate, (2004, per 1,000 live births)	Life Expectancy (2004)	Risk- Protection	Net Benefits	Access to Health Care	Risk- Protection	Health Care	Risk- Protection
Part A: Malaysia	163
Health expenditure per capita, (2003, PPP basis)
Health expenditure, total (2003, as percent of GDP)	3.8
		10.2	73.5	H	M	H	M/H	M/H	M/H
Hospital beds (2001, per 1,000 people)	1.9
Physicians (2000, per 1,000 people)	0.7
Part B: Costa Rica
Health expenditure per capita, (2003, PPP basis)	305
Health expenditure, total (2003, as percent of GDP)	7.3
		11.3	78.7	H	H	M	H	M	H
Hospital beds (2003, per 1,000 people)	1.4
Physicians (2000, per 1,000 people)	1.3

Source: World Bank World Development Indicators 2006a.

4. Guideposts for macroeconomists

• Useful data—in assessing health system performance, three types of data should be sought: the National Health Account (NHA), compiled according to international standards; the latest statistics on the infant mortality rate, the under-age-5 mortality rate, and life expectancy, and per capita public health spending for the poor. The NHA should reveal what share of the public health budget is allocated to public health, disease prevention, and maternal and child health. This share should at least be 10 percent.

• In assessing performance—macroeconomists should make a quick assessment of the health care system’s performance by comparing a country’s input and output indicators with those of a best-performing and a median-performing country. The input indicators could be health care spending, available facilities, and human resources. The output indicators could be level of health status, risk-protection, and the equity and efficiency of both. Tables 16, 17, and 18 provide benchmarks for Stage III countries (covering two best performing countries, Malaysia and Costa Rica, and one median-performing country, Colombia.)

• In reviewing policy programs, it is important to determine whether public resources are misallocated. A poor level of health status, especially a high infant mortality rate, indicates that the country either underfunds public health and preventive programs or is ineffective in the delivery of these services. These problems are likely to occur for poor and low-income communities, particularly in the rural areas and urban slums.

• If poor health status is due to inefficiencies in the public health and social-insurance-operated facilities, poor management (or corruption) can be remedied by separating the financing function of the social insurance fund from the health-service-provision function. Using the principles of competition to improve efficiency and quality of services, patients should be given a choice as to where to seek services— from the public or private sector—and the payment for the services should follow the patient. In other words, the public or social-insurance-operated facilities should not automatically receive a budget—they should have to compete for patients.

• Targeting public funds—For the countries operating under either the social insurance or the free market-oriented models, most of the public budget should be allocated to subsidize poor and low-income households.

Table 16.

Input and Output Indicators of a Median-Performing Upper-Middle-Income Country: Colombia

Source: World Bank, 1999a. Note: L, M, and H denote low, medium, and high, respectively.

Table 16.

Input and Output Indicators of a Median-Performing Upper-Middle-Income Country: Colombia

INPUT INDICATORS		OUTPUT INDICATORS
		Aggregate			Equity in			Efficiency in
		Level of Health Status
		Infant Mortality Rate 1997 (Per 1,000 live births)	Life Expectancy 1997	Risk- Protection	Net Benefits	Access to Health Care	Risk- Protection	Health Care	Risk- Protection
Health expenditure per capita, (1997 PPP basis)	359.6
Health expenditure, total (as percent of GDP)	6.9
		24	70	M	M	M	M	M	M
Hospital beds (per 1,000 people)	1.3
Physicians (per 1,000 people)	1.0

Source: World Bank, 1999a. Note: L, M, and H denote low, medium, and high, respectively.

View Table

Table 16.

Input and Output Indicators of a Median-Performing Upper-Middle-Income Country: Colombia

INPUT INDICATORS		OUTPUT INDICATORS
		Aggregate			Equity in			Efficiency in
		Level of Health Status
		Infant Mortality Rate 1997 (Per 1,000 live births)	Life Expectancy 1997	Risk- Protection	Net Benefits	Access to Health Care	Risk- Protection	Health Care	Risk- Protection
Health expenditure per capita, (1997 PPP basis)	359.6
Health expenditure, total (as percent of GDP)	6.9
		24	70	M	M	M	M	M	M
Hospital beds (per 1,000 people)	1.3
Physicians (per 1,000 people)	1.0

Source: World Bank, 1999a. Note: L, M, and H denote low, medium, and high, respectively.

Table 17.

High-Income Countries: Health Expenditure and Results by Model

Source: World Bank 2006.

Table 17.

High-Income Countries: Health Expenditure and Results by Model

	Per Capita GNI (2004 WB Atlas method)	Per Capita GDP (2004 PPP $US)	% of GDP spent on health		Health status indicators (2004)
			Total (2003)	Government (2003)	Life expectancy	Infant Mortality Rate (Per 1,000)	Under-Age-5 Mortality Rate (Per 1,000)	Risk Protection
National health service model e.g., UK	33,940	30,843	8.0	6.9	78.5	5.3	5.8
National health insurance model e.g., Canada	28,390	31,129	9.9	6.9	79.8	5.2	5.7
Social insurance model e.g., Germany	30,120	28,147	11.1	8.7	78.5	4.2	4.7
Voluntary health insurance model e.g., US	41,400	39,618	15.2	6.8	77.4	6.7	7.6
Medisave with catastrophic insurance model e.g., Singapore	24,220	27,273	4.5	1.6	79.3	2.6	3.3

Source: World Bank 2006.

View Table

Table 17.

High-Income Countries: Health Expenditure and Results by Model

	Per Capita GNI (2004 WB Atlas method)	Per Capita GDP (2004 PPP $US)	% of GDP spent on health		Health status indicators (2004)
			Total (2003)	Government (2003)	Life expectancy	Infant Mortality Rate (Per 1,000)	Under-Age-5 Mortality Rate (Per 1,000)	Risk Protection
National health service model e.g., UK	33,940	30,843	8.0	6.9	78.5	5.3	5.8
National health insurance model e.g., Canada	28,390	31,129	9.9	6.9	79.8	5.2	5.7
Social insurance model e.g., Germany	30,120	28,147	11.1	8.7	78.5	4.2	4.7
Voluntary health insurance model e.g., US	41,400	39,618	15.2	6.8	77.4	6.7	7.6
Medisave with catastrophic insurance model e.g., Singapore	24,220	27,273	4.5	1.6	79.3	2.6	3.3

Source: World Bank 2006.

Table 18.

Comparison of Performance: Advanced Economies

Table 18.

Comparison of Performance: Advanced Economies

View Table

Table 18.

Comparison of Performance: Advanced Economies

View Expanded

D. Stage IV: Advanced Economies ¹⁴