Secondary Logo

Journal Logo

Controlling U.S. Health Spending: Opportunities for Academic Health Centers

Anderson, Gerard F. PhD

Health Policy

This article begins by examining the factors that explain the level and rate of increase in health care spending. Expenditures per capita for health care in the United States are more than double the expenditures per capita in Canada, the United Kingdom, France, and most other industrialized countries. The main reason for the higher expenditures is not that Americans have access to or receive more health care, but that the prices Americans pay for medical services are two to three times higher than the prices in other industrialized countries. The author examines three actions that leaders of academic health centers (AHCs) could take that could reduce the burden of these higher costs on the American public. First, leaders of AHCs could compare the costs in their hospitals to the costs in comparable hospitals in other countries to find out why hospitals in the United States are so much more expensive. Second, they could examine how much they charge the uninsured for hospital services at AHCs—generally two to ten times more than they charge people with insurance. Third, including more people with multiple chronic conditions in clinical trials could make the findings of the clinical trials applicable to a larger patient population and thereby reduce the substantial geographic variation of health care that exists in the United States.

Dr. Anderson is professor of health policy and management, Johns Hopkins Medical Institutions.

Correspondence should be addressed to Dr. Anderson, Johns Hopkins Medical Institutions, 624 North Broadway, Baltimore, MD 21205; telephone: (410) 955-3241; fax: (410) 955-2301; e-mail: (

Fortunately for the mathematically challenged, the only equation in this article is E = P × Q. Medical care expenditures (E) are a product of the prices (P) and the quantity (Q) of medical goods and services. To understand the level and the rate of increase in medical care spending and to understand actions that leaders of academic medical centers (AMCs) can take to control medical care spending in the United States, it is necessary to understand how prices and quantity determine medical expenditures.

This article begins by examining two questions about health care prices: (1) are prices higher in the United States than in other industrialized countries; and (2) are health care prices in the United States increasing faster than most other goods and services? It will then examine three questions with respect to quantity: (1) do we use more health care services in the United States than those in other industrialized countries use; (2) is the use of health care services increasing faster than the population is increasing; and (3) do different regions of the United States use the same level of care? Finally, I will examine several actions that leaders of AMCs can take to reduce the burden of the high cost of health care on the American public.

Back to Top | Article Outline

The Price of Health Care in the United States

Prices are significantly higher in the United States

Compared to 29 other industrialized countries, Americans often pay two or three times as much for medical services as do their counterparts in the other industrialized countries. This is not true for most other goods and services, such as food, gasoline, or computers, but Americans pay significantly higher prices for hospitals, physicians, and drugs—the three largest expenditure categories in the United States.

According to 2003 data from the Organization for Economic Cooperation and Development (OECD), the average price of a day in a U.S. hospital in 2003 was $2,086.1 This compares to $804 in Canada, $1,041 in France, and a median of $512 in the 30 largest industrialized countries. Although it is true that the United States has, on average, a shorter length of stay than that of many other countries, the difference in length of stay does not explain most of the difference in expenditures per day.2 It is difficult to compare the complexity and severity of illness among these 30 countries, but there is little reason to believe that the United States has hospital patients with an average complexity of illness that is more than twice as great as patients in Canada, the United Kingdom, or Germany.3

Physician income is three times higher in the United States compared to the average industrialized country, and physicians receive substantially higher payments for similar services, such as office visits, than do physicians in other industrialized nations.4 These comparisons are made after all practice expenditures, including the high cost of malpractice insurance in the United States, are taken into account. For economists, perhaps more significant is the finding that the ratio of average physician income to average employee compensation is 5.5:1 in the United States compared to 1.5:1 in the United Kingdom and Sweden.5

Perhaps an easier price comparison involves pharmaceuticals because it is possible to compare the same drug at the same dose with the same method of administration. Two recent studies have compared pharmaceutical prices between the United States and other countries. One study in 2003 compared a market basket of the 30 drugs with the highest prices in the United States and found that they cost 52% less in Canada.6 A study that included a wider array of pharmaceuticals and used data from 1999 found that prices were 33% lower in Canada.7 In Canada, pharmaceutical prices are negotiated by a single entity—the government—whereas in the United States pharmaceutical prices are negotiated by multiple entities. This gives each entity less market influence and less incentive to keep prices low.

A study comparing health expenditures in the United States with those of 29 other industrialized countries reached the conclusion, “It’s the Prices, Stupid.”8 The study found that the higher prices charged in the United States were the primary reason why the United States spent $5,635 per capita on health care in 2003, whereas other industrialized countries spent a median of only $2,280 per capita. A series of papers have looked at other explanations—use of services, the aging of the population, the threat of malpractice litigation, defensive medicine, the lack of waiting lists, the number of uninsured, and administrative costs—and have found that these factors do not explain most of the variation in health care spending between the United States and other industrialized countries.2,8–11

Back to Top | Article Outline

Increases in U.S. rate of health care spending are similar to those of other countries

Most U.S. policymakers, however, have not focused on the price differentials between the United States and other industrialized countries; instead, they have focused on the annual rate of increase in health care spending. This is somewhat surprising because the overall rate of increase in real spending in the United States (after adjusting for inflation) is comparable to the rate of increase in spending in other industrialized countries over the period from 1960 to 2003.2

Table 1 shows the rate of increase in health care spending from 1960 to 2003 in five-year increments. Perhaps more important, however, is the accompanying breakdown of the factors responsible for the expenditure increases, also presented in five-year intervals. Table 1 shows that economy-wide price increases were responsible for almost two fifths of the increase. In addition, health care prices are increasing faster than overall inflation and these increases in excess of economy-wide price increases were responsible for nearly another fifth. Together these price increases were responsible for almost three fifths of all health spending increases in the 1960 to 2003 period. Changing demographics (the aging of the population) were only responsible for a little over 10%. The greater intensity of services or increased quantity of services was responsible for the remainder (32.2%). These increases are mostly accounted for by technological improvements and additional services.12 To reduce U.S. health care spending, it is clear that policymakers must address the source of the problem—health care prices.

Table 1

Table 1

Back to Top | Article Outline

Policymakers should confront price levels

As noted earlier, policymakers focus more on price increases than price levels, in spite of the fact that the United States is an outlier in the levels of health care spending, not the rate of increase in health care spending. This is because it is easier for policymakers to control the rate of increase in spending than to evaluate the appropriate level of spending. For example, because of payment systems such as the Prospective Payment System (PPS) and the Resource-Based Relative Value Scale (RBRVS), federal policymakers are able to control the rate of increase in funding they give to hospitals, physicians, nursing homes, and other providers who participate in the Medicare program.

Most recently, the debate among policymakers has focused on the appropriate rate of increase in physician spending. The Congress enacted a formula along with the passage of the RBRVS that tied the rate of increase in the physician fee schedule to the volume of services provided by physicians. In setting up the formula for establishing the RBRVS, Congress recognized that physicians could increase their income by ordering more services. As a result, Congress included a provision in the payment formula called the sustainable growth rate (SGR), which reduced the fee schedule if utilization of physician services increased faster than anticipated.

Every year the Medicare actuaries estimate the rate of increase in the physician services that are anticipated in the Medicare program based on demographics and other factors. For many years, actual expenditures for physician services have exceeded the estimate, and for several years the Medicare payments to physicians were reduced because of the formula. As physicians began to complain about the formula, Congress confronted a substantial budget problem. If, in response to physicians’ complaints, Congress decides to eliminate the volume adjustment and increase payments to physicians by the Medicare Economic Index (a general price index), it will mean an additional $154.5 billion in Medicare spending over a ten-year period.12 Congress made the adjustment to compensate physicians according to the Medicare Economic Index for 2006, but it will become increasingly costly, and therefore difficult, in subsequent years to ignore the SGR.

To some extent, the payment system is a “zero sum game”—if one provider category gets more, then the other providers will get less. Under the Medicare Modernization Act of 2003, the Medicare program is expressly prohibited from setting pharmaceutical prices. This could change, however, if the hospitals, physicians, and other health care providers in the Medicare system receive lower annual increases than pharmaceutical companies. These other providers will want the pharmaceutical industry to operate under the same rules as they do, and they could appeal to policymakers to regulate pharmaceutical industry compensation accordingly, potentially by lowering prices of pharmaceuticals. Leaders of academic health centers (AHCs) also need to understand the implications of payment being a zero sum game, because if physicians get more, then hospitals and/or nursing homes will probably get less.

The same debates occur in the Medicaid program and among private insurers and managed care plans. Hospitals, physicians, and other providers are paid on some type of fee schedule by most insurance plans, and the fee schedule needs to be updated annually. Unfortunately, as is the case among congressional policymakers, the rate of increase in prices, not the absolute level of prices, is the focus of most of the negotiations between the insurers and health care providers.

Back to Top | Article Outline

The Quantity of Health Care Services in the United States

Fewer services are available to and used by Americans

Surprisingly, Americans have access to and utilize fewer medical services than people in most other industrialized countries. For example, the United States has fewer physicians, physician visits, hospital beds, and hospital bed days per capita than 15 other industrialized countries.* Perhaps even more surprising is that many other countries have more computed tomography scanners and magnetic resonance imaging units per capita—two technologies synonymous with highly technological medicine.

As shown in Table 1, the rate of increase in the quantity of services is responsible for less than one third of the increase in health care spending. In spite of this, most of the policy initiatives in the United States over the past 20 years have focused on such issues as controlling the growth in length of hospital stays and eliminating unnecessary visits to health care providers. The international data suggest that these initiatives have been successful in controlling the overall growth in the quantity of health care services in the United States. But health care expenditures in the United States still far surpass those of other industrialized countries, further supporting the fact that in order to decrease U.S. health care expenditures to be more in line with international levels, the high price of U.S. health care services must be reduced. International comparisons also show that, in an effort to further reduce the quantity of health care services consumed, the United States has more aggressively pursued a hospital downsizing strategy. Probably the best example of this strategy in the United States has been the promotion of ambulatory surgical centers. As an alternative to inpatient hospitalization, ambulatory surgical centers not only process and treat patients more quickly, using a smaller quantity of resources than a regular hospital stay, but the widespread use of these centers also means that hospitals end up downsizing their numbers of available beds and, ultimately, the quantity of care provided.

Back to Top | Article Outline

Regional variation in utilization in the United States

Although efforts to reduce the overall quantity of health care consumption have been successful, policies have been much less effective in diminishing the substantial variation in health care utilization rates and expenditures across the nation. Jack Wennberg started writing about small area variation in health care utilization in 1973; he and his colleagues have continued writing on this same topic for more than 30 years.13 These studies have shown persistent regional variations in utilization rates despite efforts to control spending in high use areas, and even though other studies show similar outcomes of usage rates in high- and low-use areas.14

Back to Top | Article Outline

How Can Leaders of AHCs Control Health Care Spending?

Analysis of trends and levels of health care spending suggests that two factors deserve the primary consideration of the leaders of U.S. AHCs—the substantially higher prices for medical services and the substantial regional variation in health care utilization and expenditures.

Back to Top | Article Outline

Examining high prices of health care services

Leaders of AHCs can (and probably should) compare the prices of health care services at their institutions to those prices in other industrialized nations. At the same time, they can conduct studies to better understand why U.S. health care prices are so much higher than those of other industrialized nations. First, they could examine why hospitals in other countries such as Canada, the United Kingdom, or Australia can provide hospital care at substantially lower costs per day than the hospitals in the United States. Second, AHC leaders could constrain the growth in their hospital charges. Currently, hospital charges are increasing much more rapidly than hospital costs. Unlike patients whose insurance covers the majority of hospital charges, self-pay patients (mostly the uninsured) are expected to pay full charges when they enter hospitals. Based on my analysis of Medicare Cost Reports data from 1996–2003, the increases in charges for hospital services at AHCs averaged 13.2% in that period. Over the same time period, costs increased only 7.7%. In 2003, the average charge-to-cost ratio at AHCs was 2.8:1. In other words, self-pay patients (the uninsured) were being asked to pay almost three times the cost of the services they received. At one AHC, the ratio was >10.0:1.

Leaders of AHCs could also investigate why they are paying much higher prices for pharmaceuticals than AHCs in other countries. The impact on patients of the higher pharmaceutical prices directly affects patient care because high pharmaceutical prices are a major reason why patients do not take the drugs they are prescribed. A recent study showed that if the United States paid the same prices as other countries for pharmaceuticals, the prescription drug benefit offered to Medicare beneficiaries could be significantly more generous.6 It would be possible to eliminate completely the “doughnut hole,” or gap, in Medicare prescription drug coverage.

When they argue for higher payments to physicians, leaders of AHCs also need to understand that physicians’ income in the United States is substantially higher than that in other industrialized countries. It is important to recognize that the years of training in most industrialized countries are similar to the number of years required in the U.S. model of medical education, and that physicians in the United States do not see more patients on average than do physicians in other countries. And although the typical medical student graduating in the United States will have a substantially higher debt than graduates in other countries, his or her higher earning potential means that this debt can theoretically be paid off in only a few years.

Back to Top | Article Outline

Exploring regional variation of utilization and expenditures

Probably the greatest area where AHCs can affect health care spending is by reducing the amount of regional variation in health care utilization. Regional variation occurs primarily when there is uncertainty about the appropriate way to treat a particular patient.15 Leaders in academic medicine can help reduce the uncertainty by working together on practice guidelines and other approaches to standardize the practice of medicine.

One reason for the considerable geographic variation in medical practice is that most clinical studies exclude the types of patients that most physicians treat. As a result, there is uncertainty among physicians about how to best treat a patient and that uncertainty leads to practice variations. Clinical trials are generally conducted on patients with few coexisting conditions.16 However, most patients have multiple coexisting conditions. It is not surprising, therefore, that when practicing physicians are treating a patient and they pick up a clinical paper in search of treatment options, they wonder if the clinical trial in question applies to their patients because of the trial’s exclusion criterion.17 This uncertainty can explain part of the considerable geographic variation that occurs in medical practice. If physicians are unsure of a particular clinical trial’s applicability to their own patients, some may follow the treatment that is recommended by the study while others conclude that it does not apply to their particular patient. By reducing the number of exclusions in most clinical trials, it could be possible to reduce geographic variation because more physicians would see that the results of the clinical trials apply to their specific patients, and this would reduce the uncertainty that many physicians experience when evaluating a clinical trial’s relevance.

Back to Top | Article Outline

Why Leaders of AHCs Should Care about Higher Prices

But why should leaders of AHCs care that prices for medical services are much higher in the United States than in other industrialized countries? There are two answers to this question. The first involves international competitiveness. The recent financial problems at General Motors are an example of the problems faced by American industry as it tries to compete internationally. General Motors spends $1,525 per car on health insurance benefits for its workers and retirees. One (but certainly not the only) reason why General Motors is having difficulty competing with auto manufacturers in other countries is that foreign automakers do not have such high health care costs. No one purchases a car at a higher price because the workers have more expensive health insurance. High health care prices can have a crippling effect on businesses—and, by extension, on international competitiveness—and leaders of AHCs should do what they can to alleviate these high costs.

Second, it is important for the leaders of AHCs to recognize that Americans cannot continue to spend an ever-increasing share of the gross domestic product (GDP) on health care. The percent of the GDP spent on health care has increased from 7% in 1970, to 9% in 1980, 12% in 1990, 14% in 2000, and is projected to be 20% in 2015. Economists do not agree on the maximum percent of the GDP the United States can afford to spend on health care, but one study showed that 100% of the productivity increases in the entire economy will be spent on health care by 2039 if current productivity trends continue.18,19

The data show that U.S. health care expenditures far surpass those of other industrialized nations and comprise a growing portion of overall U.S. spending each year. To take part in reducing the growing burden of health care costs on corporations, insurance providers, and self-pay patients, leaders of AHCs must scrutinize current practices when it comes to health care pricing and be willing to make significant changes. When it comes to reducing U.S. health care expenditures, AHC leaders have the responsibility to make this possible by addressing the high price of health care services in the United States.

Back to Top | Article Outline


1 Organization for Economic Cooperation and Development (OECD). Health Data 2005 ( Accessed 17 May 2006.
2 Anderson GF, Hussey PS, Frogner B, Waters H. Health spending in the United States and the rest of the industrialized world. Health Affairs. 2005;24:903–14.
3 Redeimeier DA, Fuchs VR. Hospital expenditures in the United States and Canada. N Engl J Med. 1993;328:805–7.
4 Bodenheimer T. High and rising health care costs. Part 3: the role of health care providers. Ann Intern Med. 2005;142:996–1002.
5 Bodenheimer T. High and rising health care costs. Part 1: seeking an explanation, Ann Intern Med. 2005;142:847–54.
6 Anderson GF, Shea DG, Hussey PS, Keyhani S, Zephyrin L. Doughnut Holes and Price Controls. Health Affairs ( Accessed 17 May 2006.
7 Danzon PM, Furukawa MF. Prices and Availability of Pharmaceuticals: Evidence from Nine Countries. Health Affairs. ( Accessed 17 May 2006.
8 Anderson GF, Reinhardt UE, Hussey P, Petrosyan V. It’s the prices, stupid: why the United States is so different from other countries. Health Affairs. 2003;22:89–105.
9 Anderson GF, Hussey PS. Comparing health system performance in OECD countries. Health Affairs. 2001;20:219–32.
10 Reinhardt U, Hussey P, Anderson GF. Cross-national comparisons of health systems using OECD data. Health Affairs. 2002;21:169–81.
11 Reinhardt UE, Hussey PS, Anderson GF. U.S. health spending in an international context. Health Affairs. 2004;23:10–25.
12 Congressional Budget Office ( Accessed 17 May 2006.
13 Wennberg JE, Cooper MM. The Dartmouth Atlas of Health Care. Chicago: American Hospital Publishing, 1999.
14 Wennberg DE, Wennberg JE. Perspective: Addressing variations: Is there hope for the future? ( Accessed 17 May 2006.
15 Fisher ES. Medical care—is more always better? N Engl J Med. 2003;349:1665–67.
16 Anderson GF. Medicare and chronic conditions. N Engl J Med. 2005;353:305–9.
17 Boyd CM, Darer J, Boult C, et al. Clinical practice guidelines and quality of care for older patients with multiple comorbid diseases. JAMA. 2005;294:716–24.
18 Chernew ME, Hirth RA, Cutler DM. Increased spending on health care: how much can the United States afford? Health Affairs. 2003;22:15–25.
19 Pauly MV. Should we be worried about high real medical spending growth in the United States? ( Accessed 5 June 2006.

*Canada, Australia, Austria, Belgium, the Czech Republic, Denmark, France, Germany, Greece, Iceland, Italy, the Netherlands, Norway, Portugal, and Switzerland.2
Cited Here...

© 2006 Association of American Medical Colleges