[Economic Report of the President (2006)]
[Administration of George W. Bush]
[Online through the Government Printing Office, www.gpo.gov]

 
CHAPTER 4

Improving Incentives in
Health Care Spending

Health care spending in the United States has increased rapidly over
the past several decades, rising 44 percent in real per capita terms
in the past ten years alone. Some of the reasons for this marked rise
reflect higher-quality health care, such as improved technological
options for enhancing the health and quality of life of the American
people. However, other factors, such as poorly functioning markets
for health care, may have led to excessive spending and inefficient
patterns of medical care utilization. Furthermore, whether this
increased spending is of high value or not, it has put tremendous
pressures on individuals and the institutions that finance health
care spending. Family budgets are being strained as health care costs
take up an increasing share of incomes. Government health care
expenditures have also been increasing rapidly, burdening both
Federal and state budgets. If not curtailed, the increased costs
to governments will eventually lead to large tax increases, sharp
cuts in nonhealth spending, or both.
This chapter reviews the causes and consequences of health care
spending growth and discusses how spending can be more efficient and
of higher value in the context of a consumer-driven, market-based
system. The emerging consumer-driven health care movement aims to
empower consumers with improved information and ability to make
choices about their own health care, which in turn can result in
increased provider competition to better serve patients' needs at
lower costs. The key points of this chapter are:
 Growth in spending on health care has been much more rapid
than general inflation, straining consumers, employers, and
government budgets.
 Perverse tax and insurance incentives have led to
inefficient levels and composition of spending on health care. Some
increased spending has produced valuable health improvements, but in
a better-functioning health care market these improvements could be
attained at lower cost.
Promoting a stronger role for consumers is a promising
strategy for improving health care value and affordability.

The Growth in Health Care Spending

Spending in the health care sector has steadily grown from under 6
percent of GDP in 1965 to 16 percent of GDP in 2004. If current
trends continued, health care spending would be projected to reach
19 percent of GDP by 2014 and 22 percent by 2025 (Chart 4-1). Since
1965, the government share of total health spending has risen from
25 percent to over 45 percent, mainly due to increased eligibility
and generosity of Medicare and Medicaid. (Medicare is a Federal
government program that pays for health care for senior citizens and
those with certain disabilities. Medicaid, financed by both Federal
and state governments, is focused on providing health care for the
poor.) Medicare spending alone is projected to increase from 2.6
percent of GDP in 2006 to 4.3 percent by 2025. Among those without
access to Medicare or Medicaid, most expenditures are financed by
private health insurance (64 percent), provided mainly through
employers (91 percent of those with private insurance). The rising
costs of health care are reflected in premiums (employer plus
employee share) for employer-provided insurance that in 2005
averaged almost $11,000 for a family (Chart 4-2), up from $6,700 in
1999 (in 2005 inflation-adjusted dollars). Per capita health care
spending in the United States has risen from about $4,500 ten
years ago to about $6,500 today (in 2005 dollars).
The United States today spends roughly twice as much per capita
on health care as other industrialized countries, such as the other
members of the Organization for Economic Cooperation and Development
(OECD). This large difference in part reflects higher levels of per
capita income and output



in the United States, since richer countries tend to spend
proportionately more on health care, but the United States spends
a substantially larger share of GDP on health care than other
wealthy countries do. For example, the United Kingdom spends about
8 percent of its GDP on health care, compared with the United States'
16 percent. The U.S. expenditure as a percent of GDP is more than
six percentage points higher than the average in OECD countries.
Rates of spending growth, however, are much more similar across
countries. For example, from 1998 to 2003, average real health care
spending increased 4.6 percent per year in the United States as
compared to 4.5 percent in the OECD as a whole. This suggests that
many of the underlying international spending differences stem from
longer-term factors.
When looking at these statistics, it is also important to remember
that buying more health care is not necessarily equivalent to buying
more health. Health care is one of many different determinants of
health status, and for many people marginal increases in health
care consumption may be less cost-effective than marginal increases
in spending on other determinants such as a healthier lifestyle
(exercising, not smoking, eating a healthier diet). Evaluating the
relative cost-effectiveness of spending on different health
determinants can be challenging, however, in part because it is
difficult to measure the quality of health services consumed.

Where Health Spending Has Grown

There have been significant increases over time in all major
spending categories, including outpatient, acute inpatient,
long-term care, and pharmaceuticals. Both personnel costs and goods
costs have increased. Spending has grown for both privately and
publicly financed and delivered care.
One might guess that the aging of the U.S. population would
explain an important part of the increase in health care costs,
especially since about one-quarter of health care in a given year is
spent on those who die that year. Research suggests, however, that
less than 10 percent of the growth in health spending over the last
several decades can be attributed to this factor. Another
contributing factor might be America's rising prosperity, because
richer individuals and nations demand more health care, but again
this factor can only account for a relatively small portion of the
health care spending growth. Various studies have speculated about
the contribution of other factors such as rising obesity, but there
is as yet no consensus on the importance of these factors. There is
general agreement, however, that the rapid growth in development
and use of expensive new health care treatments accounts for a large
share of overall health care spending growth over time.
A useful framework for understanding increases in medical spending
breaks these spending increases into three components: (1) changes
in the quantity demanded of existing health-related goods and
services, (2) changes in the prices of those existing goods and
services, and (3) the effects of technological advances that change
the available set of health-related goods and services. The next
part of this section looks at each of these three factors.

Quantity of Health Care Demanded

Do we demand higher volumes of health care today than in the past?
While we clearly consume more of some types of care (based on higher
incomes, changing medical needs, etc.), health care visits per
capita have not increased. The biggest components of health care
spending are physician and hospital services. Doctor visits per
capita dropped somewhat from 1980 through the mid-1990s, and have
increased only modestly since then. The number of hospital discharges
per capita and the average hospital length-of-stay, however, have
declined dramatically--they were 50-percent higher in 1980 than in
2000. Growth in spending within the United States does not seem
to be explained by increased visits to the doctor or hospital.
Moreover, international differences in spending cannot be
explained by differences in the quantity of physician and hospital
visits. In fact, doctor visits and hospital nights per capita in the
United States are lower than in many OECD countries. For example,
in 2000 the United States had 0.7 hospital nights per capita,
compared to 0.9 nights in the United Kingdom, 1.3 nights in
Switzerland, and 1.9 nights in Germany. Service intensity in the
United States is very different, however, with U.S. hospital
staffing levels at double the OECD median. Thus, while Americans
have fewer health care contacts, they appear to receive more
services at each contact. This difference explains in part why the
average U.S. hospital night costs three times the OECD average.

Health Care Prices

The official medical consumer price index (medical CPI), which
measures price increases for medical goods and services and is
published by the Department of Labor's Bureau of Labor Statistics,
indicates that health care prices over the last few decades have
grown more rapidly than prices of other goods and services in the
economy. From 2000 to 2004, the health care component of the CPI
grew 19 percent compared to only 10 percent for the general CPI,
indicating 9 percent real growth in health care prices. Thus of the
33 percent growth in total per capita health spending over this
period, one-quarter apparently derived from increases in the prices
of health care relative to other goods and services.
Why would health care prices rise so rapidly? One possible
explanation for these recent price increases is that supplier
consolidation has led to reduced competition among health care
providers, enabling hospitals and physician groups to leverage
market power to raise prices. For example, there were about 900
hospital consolidations during 1994-2000 (from a base of roughly
6,000 hospitals). Some of these mergers have appeared to result in
monopolistic price increases, and even some major metropolitan
areas have become dominated by just two or three hospital systems.
It is not clear how important such trends will be in the future,
however, in the face of vigorous antitrust enforcement.
Part of the apparent increase in relative prices may, however,
be the illusory result of measurement problems. Standard price
indices such as the medical CPI may overestimate price growth in
health care if they do not adequately account for improvements in
health care quality. Price indices are supposed to reflect price
changes for a given product. However, because health care quality is
constantly increasing, rising prices for a given health care visit
may reflect improved quality, rather than just higher costs for a
given level of care. For example, the coronary artery bypass graft
that the average patient receives today may result in fewer
complications and longer and higher quality of life afterward than
would have been the case for a patient receiving the procedure 10
years ago--so the higher price paid for the procedure reflects in
part the fact that the patient is receiving more ``health,'' not just
paying more for the same service.
That said, higher prices for medical services do appear to be an
important part of the explanation for why the United States spends
more on health care than other OECD countries do. For example, one
study of Australia, Denmark, France, Canada, Germany, and the United
Kingdom found that physician wages in the United States are 77
percent higher than the average across those countries. This does
not mean, however, that those countries provide a model that should
be emulated: Heavy price regulation in some countries has led to
long waiting lists for certain types of medical services. One recent
survey found that over half of patients in Canada and the United
Kingdom had to wait longer than a month for a specialist
appointment, compared to less than a quarter of patients in the
United States. Similarly, more than a third of patients had to wait
longer than four months for elective surgeries in Canada and the
United Kingdom, compared to fewer than 10 percent in the United
States.
There is a common perception that drug prices are unduly higher
in the United States than in other OECD countries, perhaps due to
aggressive price negotiation by European governments, but recent
research suggests that this may be misleading for several reasons.
First, carefully accounting for manufacturer discounts to insurers
in the United States shows price differences to be smaller than
simple retail price comparisons would suggest (U.S. prices are
discounted by about 8 percent on average). Second, U.S. consumers
use a much higher proportion of generic drugs than do consumers in
other countries (e.g., 58 percent of units in the United States
versus 28 percent in France). When comparing average prices paid
for each active ingredient (whether generic or name brand), rather
than only prices for selected name brand drugs, the international
price differences are further narrowed.
Furthermore, some experts suggest that wealthier countries such
as the United States should pay a larger share of drug development
costs than should less-wealthy countries, because of both equity
and efficiency arguments. Thus, observing lower drug prices in
developing countries than in the United States does not generate
great controversy. Many people do not recognize, however, that the
United States is also substantially richer than most other OECD
countries. For example, per capita income in the United States is
22-percent higher than in the United Kingdom. After adjusting for
differences in manufacturer discounts, use of generics, and per
capita income, average drug prices are in fact higher in many other
OECD countries.  Research has found that U.S. drug prices relative
to income are 7-percent lower in France, but 4-percent higher in
Canada, 10-percent higher in Germany, and 25-percent higher in the
United Kingdom. Thus, the United States' higher health care
spending as a share of GDP does not appear to be explained by
higher drug prices.

Technological Change

Research suggests that, over time, a major source of health care
spending increases has been adoption of new, technologically
intensive health care goods and services. For example, one study
found that average spending per heart attack case in the United
States increased in real terms from $12,000 in 1984 to about $22,000
in 1998, and that about half of this spending increase could be
attributed to the adoption of more-sophisticated technologies. This
does not mean that the higher spending is not of very high value:
post-heart attack life expectancy over this same period increased
from five years to six years, with 70 percent of that increase
attributable to the adoption of better technology.
The United States appears to use some expensive technologies more
intensively than do other countries. For example, the United States
has more than 50-percent more MRI units per capita than do other
OECD countries on average. The United States' more-intensive use of
technology partly reflects its higher rate of innovation and earlier
adoption of technology. For example, angioplasty was relatively rare
outside the United States in 1990, with the U.S. utilization rate
three times higher than the next-closest country; Germany finally
reached the U.S. level by about 1998, while adoption in other
countries continued to lag.
It is worth noting that the adoption of new technologies does not
inevitably raise costs. New technologies regularly reduce costs in
many other sectors of the economy, such as the semiconductor
industry. In the U.S. health care industry, however, the combination
of technological change along with muted consumer incentives to
demand lower costs is responsible for a significant portion of
rising health care spending.

First-Dollar Insurance Inhibits Consumer
Cost-Consciousness

In most markets outside of health care, consumers decide what to
purchase by comparing the price of a good or service against the
benefit it brings them. By contrast, in the health care sector,
consumers often do not learn the prices of goods and services
consumed until bills are received weeks or months later, if ever.
Instead, physicians are expected to make health care consumption
choices for patients, despite the fact that physicians frequently
lack the incentive to match the benefits of care with its costs,
and may even lack information about the costs themselves. A major
reason for this lack of consumer incentive is the fact that many
health insurance policies provide close to "first-dollar
coverage" of health care costs. That is, people with health
insurance typically pay only a relatively small portion of the
total cost--or in some cases, literally none of the cost--of the
health care services they receive. This section reviews the
causes and consequences of first-dollar insurance coverage.

Causes of First-Dollar Insurance Coverage

Unlike most other types of insurance, health insurance in the
United States often includes first-dollar coverage of the cost of
even routine, predictable services. By contrast, most other forms
of insurance focus on protecting the insured from large and
unexpected losses. If automobile insurance had the first-dollar
coverage of even routine services that many health insurance
policies offer, it would cover the costs of oil changes and new
tires, rather than just protecting against unpredictable
catastrophes such as automobile accidents.
Health insurance policies have this unusual first-dollar coverage
feature in large part because the tax code makes it cheaper for
people to purchase health care indirectly through insurance than
directly through out-of-pocket payments (see Box 4-1). Another
factor underlying first-dollar coverage is the increased use of
managed care programs, which spread rapidly during the 1990s. Most
managed care plans are characterized by minimal cost sharing,
relying instead on gatekeepers to regulate use of resources.
Interest in managed care programs has decreased recently, because
of public backlash against the cost-containment measures used in
these programs.

---------------------------------------------------------------------
Box 4-1: Tax Preferences for Employer Health Insurance Premiums

Since the 1940s, the tax code has excluded employer payments for
health insurance premiums from the portion of workers' compensation
subject to taxation (both payroll and personal income taxes).
The total value of the tax exclusion is quite large, reducing
Federal taxes by over $200 billion in 2006 ($133 billion for the
income tax exclusion and $80 billion for the payroll tax exclusion),
which is equivalent to about 10 percent of actual Federal tax
receipts. This exclusion of health insurance premiums from taxation
was a by-product of wage-control legislation during World War II
(which established a precedent for treating employee benefits
differently from regular wages), and was not intentionally designed
to promote health insurance coverage. But this tax treatment of
employer-provided health insurance premiums has had important
consequences for insurance markets.
First, it has caused the private insurance system to become
predominantly employment-based. More than 91 percent of privately
insured individuals under age 65 receive their health insurance
through their employers. Except for the self-employed, those who
purchase insurance on the individual market (that is, not through
their employers) must do so with after-tax dollars. The
self-employed receive an "above-the-line" income tax deduction for
health insurance premiums (equivalent to the income-tax exclusion
for employer insurance), though they still owe full payroll taxes
on the income used to buy premiums. For someone in the 15-percent
income tax bracket and subject to the 15.3-percent payroll tax, a
policy with a $10,000 premium would cost roughly $7,000 if purchased
through an employer, $8,500 if the person were self-employed, and
the full $10,000 if the person were not self-employed and purchased
the policy individually. This tax treatment has created a strong
financial incentive for individuals to purchase health insurance
through their employer, even if their first choice of insurance
product is not offered by the employer. In addition, as an incentive
to buy health insurance, this tax subsidy is larger for people in
higher tax brackets (as shown in the chart), despite the fact that
a given subsidy amount would reduce uninsurance much more among
lower-income households.
Furthermore, the employer premium tax exclusion promotes
low-deductible insurance coverage with minimal out-of-pocket cost
sharing. In most cases, while insurance premiums are paid with
pretax dollars, out-of-pocket health spending must be paid for with
after-tax dollars. For example, $1,000 of health care services
covered by full insurance costs the person with employer-provided
insurance only about $700 in after-tax dollars (assuming a
15-percent income tax bracket and 15.3-percent payroll tax), whereas
those same services would cost $1,000 if paid out-of-pocket. Because
of the tax penalty for out-of-pocket spending relative to insurance
premiums, there is a strong incentive for employers to provide and
employees to select first-dollar coverage, even if they would have
preferred higher deductibles and lower premiums in the absence of
the tax provision. This has, in turn, diminished the role of
consumers as guardians against wasteful spending and unduly high
prices.



---------------------------------------------------------------------

Consequences of First-Dollar Insurance Coverage

The original purpose of health insurance, like other forms of
insurance, was to protect individuals from catastrophic and
unexpected costs by spreading risk across a larger population.
However, as discussed, health insurance in the United States has now
also become a vehicle for financing relatively low-cost, routine
expenditures. This use of insurance as "prepaid medical care" has
three important consequences: (1) It encourages consumers to
overuse certain types of health care. (2) It gives little incentive
for consumers to search for the lowest-price providers. (3) It
distorts incentives for technological change. Rather than focusing
research incentives on cost-effective technology, it induces
adoption of technologies for which costs exceed incremental
benefits, while undermining the development of cost-saving
technologies.  We discuss each of these points.
First, heavily insured individuals, being insulated from most
health care costs, have the incentive to overconsume certain types
of care, a phenomenon referred to as moral hazard. An allergy drug
may have great value for patient A who has serious symptoms, but
little value for patient B who has only mild symptoms. If the two
patients faced the market price of $100/month, then A might decide
the drug is worth the cost but B might forgo it, given its
negligible benefit for him. With first-dollar insurance coverage,
however, B might instead choose to continue taking the drug as long
as the expected benefits to him were greater than zero. In this
case, B's decision would inefficiently drive up health care
spending at a loss to society, since the benefit of the drug would
be less than the real cost.
Some would argue that such scenarios are rare because physicians
should not prescribe the drug for person B if it would be wasteful
or of little practical use in improving his health. But in fact
physicians may not have enough information to fully evaluate the
benefit to patients, and often have little incentive to limit
inappropriate care to highly insured patients. Providing extra
services increases their incomes and protects them from the charge
that they did not take every action with conceivable benefit to
the patient.  Box 4-2 discusses the role of medical malpractice
liability in increasing medical expenditures.
In order to quantify the moral hazard effects of first-dollar
insurance coverage, the RAND Health Insurance Experiment
randomized individuals into health insurance plans with different
co-insurance levels. (Co-insurance refers to the percentage of
health insurance spending above the deductible an individual must
contribute.) A higher co-insurance level gives both the patient and
the doctor greater incentive to avoid the use of drugs or procedures
that are costly and have low expected benefit. The study found that
changing the structure of health insurance does affect the behavior
of patients and their

---------------------------------------------------------------------
Box 4-2: Medical Liability Costs

Substantial costs in the U.S. health care system are associated
with the medical liability system. This affects health care
spending in several ways. First, the cost of malpractice damage
awards, the legal costs of malpractice lawsuits, and the costs of
underwriting malpractice insurance policies are passed on to
providers through malpractice insurance premiums and then to
patients through out-of-pocket payments and insurance premiums.
Second, defensive medicine-ordering tests and procedures solely
to guard against potential malpractice claims-may have an even
bigger effect on health care spending than the direct costs
associated with malpractice suits.
The President has called on Congress to pass liability reforms to
make the system fairer and more predictable while reducing wasteful
costs. The trend toward greater consumer decision making in health
care may have complementary effects in reducing liability costs
associated with defensive medicine. Consumers with first-dollar
insurance coverage have little incentive to decline many of the
tests and procedures suggested by physicians, even if they and
their physicians understand that there may be very little health
benefit from the increased spending. But as consumers pay for a
greater portion of noncatastrophic care, they may decide to forgo
expensive and unnecessary tests and procedures suggested by
physicians primarily to avoid lawsuits rather than to improve
patients' health.
---------------------------------------------------------------------

doctors. Specifically, individuals with
first-dollar coverage had 45-percent higher health expenditures
than individuals who were randomly assigned insurance plans with
95-percent co-insurance up to a catastrophic out-of-pocket maximum
level (the out-of-pocket maximum was about $3,500 in today's
dollars). Importantly, the extra care received in the first-dollar
coverage plans produced no discernible extra health benefits in the
studied sample as a whole. There were, however, some health
benefits for select subpopulations of low-income and chronically
ill individuals, suggesting that care should be taken not to expose
lower-income families to excessively high cost sharing relative to
their income, and that certain preventive measures such as
chronic-disease management are important to exempt from cost
sharing. For most services consumed by the majority of the
population, however, the RAND study showed that higher cost sharing
can be a powerful tool to induce consumers to take responsibility
for focusing their health care spending on only those products and
services with the highest value.
A second consequence of first-dollar insurance coverage is that
consumers are less sensitive to the prices of health care consumed,
an outcome that dulls the competitive forces that keep prices down
in most other markets. Many insurers attempt to reduce the range of
choices available to enrollees through mechanisms such as selective
contracting and preferred provider networks, but such practices are
even more effective when the consumer is also price-sensitive.
Imagine two hospitals that provide the same service, but hospital A
charges $1,000 and is located in an older facility while hospital B
charges $2,000 but is located in an updated facility with a wide
array of amenities and equipment on site. Given these choices, a
consumer facing the actual price may prefer hospital A, but in a
world of first-dollar coverage, most people would choose hospital
B, even if the extra amenities of hospital B provided only modest
benefit. As a result of this structure of incentives, health care
providers may compete for patients by providing greater convenience
or amenities with little incentive to control costs. This lack of
price sensitivity on the part of the consumers of health care is
one of the major forces underlying the rapid growth of health care
costs.
A third consequence of first-dollar insurance coverage is
distorted incentives for technological development. One type of
distortion is that new technologies may be developed and marketed
even when they are of low incremental cost-effectiveness relative to
other available options. For example, if a new drug is even slightly
more effective than an existing drug, a person with first-dollar
insurance coverage may demand the new drug even if it is priced well
above existing satisfactory and effective alternatives. When
consumers have dulled price incentives pharmaceutical companies will
invest in bringing a new drug to market even if it provides little
new value. In a world in which most consumers had high-deductible
insurance and were sensitive to the full cost of drugs, the
pharmaceutical company might choose not to spend the large amount
of resources necessary to complete clinical trials and bring the
drug to market if they knew its incremental improvement over
existing drugs would be small.
Likewise, dulled price sensitivity on the part of consumers
reduces the incentive to develop cost-reducing technologies. In many
other sectors of the economy, such as computer memory chips,
technological progress results in cheaper and more cost-effective
products each year as producers look for more-efficient
manufacturing processes and product innovations to keep them ahead
of their competitors. In health care, this type of technological
innovation is much rarer, since few consumers have the incentive to
adopt a cheaper product, particularly if it has even slightly
lower effectiveness. If more health care consumers were to become
price sensitive, the health care sector would have the incentive
to pursue more such cost-reducing technologies that could, over the
long term, help reduce the rate at which health care spending is
growing.
Some observers have expressed concern that changes to the current
system might be harmful if they result in reduced innovation, but
these observers have often failed to distinguish cost-effective
from cost-ineffective innovations. Life expectancy at birth has
increased from 70 to almost 78 years since 1962. In addition to
living longer, we are also enjoying more years in better health and
with fewer disabilities. While some of these health improvements
have been due to lifestyle changes, some can clearly be traced to
medical technologies, such as those that have reduced infant
mortality, improved survival rates after heart attacks, improved
treatment of depression and other mental illnesses, and improved
the management of chronic illnesses. Research suggests that on
average our spending on new medical technology has indeed been
cost-beneficial. This indicates that, as a society, we would not
want to return to the health spending levels of 1960, for example,
if doing so also meant returning to the types of medical care
available in 1960. But economic efficiency depends on each
("marginal") individual new technology being cost-beneficial, not
just the average of all technologies. The fact that on average our
investment in medical technology has paid off does not preclude
the possibility that our system contains significant inefficiencies,
and that some of the new technology may have contributed little
compared to the amounts spent on it. If consumers were given the
information they need about the actual costs and benefits of
various treatments, as well as the incentives to compare those
costs and benefits, it might be possible to eliminate some of that
wasteful spending.

Consequences of Inefficient
Health Care Spending

Rising health care spending is a burden to employers, consumers,
and taxpayers. Employers who offer insurance complain that rising
premiums strain their labor relations and threaten their balance
sheets. Rising premiums make health insurance less affordable,
contributing to the ranks of the uninsured. Those who are insured
face rising out-of-pocket costs and lower cash wage growth. And
taxpayers must finance the rapidly increasing costs of publicly
provided health care for seniors, the disabled, and the poor.

Private Spending

As consumers spend more of their budgets on health care, they must
spend less on other goods and services. Since 1980, for example, the
share of consumer spending that has gone to medical care has
increased from 10 percent to 17 percent, while the shares of
spending on items such as food and clothing have decreased. Of
the $7.5 trillion increase in personal income since 1980, $1.5
trillion has been devoted to health care. Similarly, of the $2.19
real increase in hourly compensation over the past five years, $0.54
(25 percent) has gone toward higher health insurance premium costs.
Thus, take-home pay has grown more slowly than total compensation
(including health insurance and other benefits) (Chart 4-3).
The costs of health care would be of less concern if most health
care spending reflected optimal decisions by consumers weighing the
costs and benefits of the services they buy.  For example, the fact
that consumer spending on DVDs increased 31 percent in 2004 alone
has not alarmed anyone nor led to calls for government intervention.
But spending on private health care is different, because health
care is considered a "merit" good deserving of government support
for those that cannot afford it, because of the government's
extensive role in the health care market, and because of the
forces that interfere with the efficient allocation of resources.
Employers have also been affected by increasing health care
costs. In particular, firms that have promised generous health
benefits to retirees have borne increasingly heavy costs. The
economic consequences of this may include the need for restructuring
of some of these firms, loss of expected benefits for some retirees,
and potential costs to taxpayers if some of these retirees increase
their reliance on public health insurance. Rising costs for current
employees have also affected employer behaviors. Some employers
have tried to reduce their insurance costs by hiring more part-time
workers (who are generally




ineligible for insurance benefits), asking employees to contribute
more to premiums, reducing the generosity of the plans they offer,
or discontinuing health insurance benefits altogether.
In the long run, however, it is not the employers but rather the
workers who bear the burden of rising health insurance costs.
Economists have shown that even though employers may make the bulk
of the payments to cover the health insurance premiums of workers,
these payments are treated just like wages or any other component
of workers' total compensation. This total compensation depends on
worker productivity and labor-market supply and demand.  Rising
insurance premiums may thus change the mix of workers'
compensation by increasing health benefits and decreasing wages,
but if they do not affect workers' productivity they will not lead
firms in competitive markets to raise total compensation.
Institutional factors such as minimum-wage laws and sluggish wage
adjustment may mean that health insurance premiums affect employer
profits in the short run, but in the long run most or all of
increases in health insurance costs are shifted to employees in the
form of wages that are lower than they otherwise would have been.

Public Spending

When per capita spending on health care rises rapidly, the
pressures on government programs become particularly intense. First,
if the standard of care received by enrollees in government programs
is not to differ too radically from that of the general public, the
costs of helping those already enrolled in the programs will rise
as well. Second, rising insurance premiums may cause some people to
drop private insurance and to rely instead on public insurance such
as Medicaid or on safety-net providers (e.g., uncompensated hospital
care) subsidized by taxpayers. Not only does rising uninsurance lead
to higher government costs, but uninsured people often consume
health care resources inefficiently--for example, by failing to
obtain preventive care, delaying necessary care, or overusing
emergency rooms relative to less-costly clinic settings.
The largest government programs that finance health care for
those not otherwise insured are Medicare and Medicaid. These
programs are becoming increasingly expensive to taxpayers. For
example, according to projections, if current trends were to
continue unchecked, Medicare costs would increase from the current
share of 2.6 percent of GDP to 6.9 percent by 2050. Medicaid,
jointly financed by the Federal and state governments, is also
becoming an increasingly large share of budgets, with just the
Federal portion of spending projected to increase from 1.5 percent
of GDP today to 2.5 percent by 2050. The costs of these public
programs are unsustainable under any reasonable projections.
Closing the currently projected 75-year deficit in just the Hospital
Insurance (HI) portion of Medicare would require tax increases of
107 percent or benefits reductions of 48 percent. Ultimately, the
benefits paid by these programs must be significantly pared back,
the taxes dedicated to their support must be increased, or major
reforms must be enacted that slow the rate of growth in health care
spending.


Strengthening the Role of Health Consumers
Through Public Policy

This chapter has discussed the central role of first-dollar
insurance coverage in dulling the incentives for consumers to shop
carefully for cost-effective health care. By giving consumers both
the incentives and the information needed to become better shoppers
for health care, public policy can help control the growth in
health care costs and improve the efficiency of the use of health
care resources.
The President has proposed a wide-range of measures to help make
health care more efficient and accessible, such as improving
community health centers, reforming medical liability laws,
creating Association Health Plans for small businesses, allowing
insurance to be more portable and purchased more easily across state
lines, and many other reforms. This section will focus specifically
on proposals that help improve incentives for consumers.
An important policy advance has aimed to reduce the bias toward
first-dollar insurance coverage by allowing more out-of-pocket
health care expenditures to be paid with pretax dollars through the
innovative mechanism of Health Savings Accounts (HSAs).
Complementary initiatives to improve information available to
consumers for making appropriate health care choices can help
facilitate the movement toward HSA-based consumer-directed health
care.
The potential benefits of reforms that slow spending growth could
be great. Consider a scenario in which new policies successfully
reduce future national health spending by one percentage point per
year, through a combination of short-run quantity decreases,
medium-term price decreases, and long-run increases in cost-reducing
technological change. If spending were to grow by 6 percent
per year, instead of by 7 percent per year as currently projected,
by 2025 the expected health share of GDP would be reduced from 22
percent to 18 percent, a substantial difference.

Health Savings Accounts (HSAs)

HSAs are tax-favored accounts to which individuals can contribute
funds they can then use to pay current and future out-of-pocket
medical expenses. These accounts were signed into law by the
President in 2003 and went into effect in 2004. HSAs represent a
major improvement over previous tax-preferred medical spending
accounts such as Flexible Spending Arrangements (which must be
exhausted each year, a factor that limits their use) and Health
Reimbursement Accounts (which are owned by employers, not
consumers). In contrast, HSAs are owned by individual consumers
regardless of employer, and unused account balances can be retained
and grow from year-to-year without penalty. HSAs are designed to be
used in conjunction with high-deductible health plans, defined as
plans having minimum deductibles (currently $1,050 for individuals
and $2,100 for families) with annual out-of-pocket limits
(currently no more than $5,250 or $10,500 for individuals and
families, respectively). Deductibles and out-of-pocket limits are
indexed to adjust over time with inflation. Certain types of
preventive care may be provided with first-dollar coverage if
deemed appropriate by the insurer.
HSA enrollees with qualifying insurance plans may contribute
annually up to the lesser of the plan deductible or $2,700
(individuals)/ $5,450 (family). These contributions are excluded
from income taxes both at the time of deposit and at the time of
"qualifying" withdrawal; the funds may be used to pay for
out-of-pocket medical expenditures, rolled over indefinitely, or
withdrawn after age 65 (in which case they are taxed as ordinary
income if not used for health expenditures).
A key benefit of HSAs is that they lower the previous tax bias
toward low-deductible or first-dollar health insurance relative to
higher-deductible policies with higher out-of-pocket spending. To
illustrate this point, consider a sample health insurance purchaser
facing the choice of a low-, medium-, or high-deductible plan.
Table 4-1 illustrates how this person's premiums depend on the
plans' deductibles, according to actuarial estimates for a
representative person. The premium for a $250 (low) deductible
policy with a $2,000 out-of-pocket limit would be $4,000, but
that premium could be lowered by $1,600 (or 40 percent) by moving
to a catastrophic policy with a $2,500 (high) deductible and an
out-of-pocket limit of $5,000. Suppose that this person had no
health expenditures in the first year of coverage, but a $15,000
catastrophic event in the second year. How is her total two-year
spending on health care under these plans affected by the tax code?
 If there are no tax preferences: If she buys the
traditional (low deductible) plan, her spending is $4,000 in
premiums in each year plus $2,000 out-of-pocket in year two,
totaling $10,000. If she buys the catastrophic (high deductible)
plan, her spending is $2,400 in premiums in each year plus $5,000
out-of-pocket in year two, totaling $9,800. Thus, she would be
slightly better off financially under the catastrophic plan in
the absence of tax preferences.
 If insurance premiums (but not out-of-pocket spending) are
tax-preferred: Under the traditional plan, if she is in the
30-percent marginal tax



bracket, she receives a $2,400 tax subsidy (over two years), but
under the catastrophic plan she only receives a $1,440 tax subsidy.
Thus, the tax subsidy makes her prefer the traditional plan where
she might otherwise have preferred the catastrophic plan.
 If tax-preferred HSAs are available: If she contributes
the maximum $2,500 to the HSA in both years, she would receive a
new $1,500 tax subsidy by using the HSA to pay her out-of-pocket
expenses in year two with tax-free dollars. This mitigates the
previous tax-induced bias against catastrophic plans, again making
her better off financially under the catastrophic policy.
This illustration of course simplifies many dimensions of the
comparison between policies. For example, it ignores the fact
that catastrophic events are rare, so that most people would be
able to accumulate many more years of premium and HSA savings,
further increasing the attractiveness of the HSA-qualified plans.
In addition, the example ignores the moral hazard effect of reduced
health care utilization in the catastrophic plan, as the patient
now has increased incentive to shop carefully for health care.
Not all individuals will benefit equally from moving to a
high-deductible policy. First, some poorly informed consumers may
forgo recommended care, such as preventive services--care that they
might have received under a traditional low-deductible policy.
The HSA provision that allows plans to waive the deductible for
preventive care is designed to mitigate this possibility. Second,
some chronically ill individuals with persistently high spending
may be relatively worse off, to the extent that high-deductible
policies lead to less cross-subsidization from healthier people in
their risk pool. This could be mitigated while preserving the
beneficial effects of cost sharing, for example, through improved
insurance benefits for the chronically ill, differential premium
cross-subsidies in employer insurance, or targeted high-risk-pool
subsidies in the individual market. Third, credit-constrained
enrollees and those in lower tax brackets will benefit less from
provisions allowing tax-free HSA contributions and accumulation.
This is also true of the tax exclusion for employer health
insurance premiums. These concerns must be balanced against the
potential benefits of greater price sensitivity by health care
consumers: As more consumers shift into high-deductible plans,
there is greater potential for slowing price growth and long-run
increases in cost-reducing technology, which could benefit even
consumers in traditional insurance plans.
Since the inception of HSAs in 2004, the number of people enrolled
in high-deductible HSA-qualified plans has increased rapidly. The
new tax benefits that further lower health costs for high-deductible
plans have made them attractive not only to the uninsured and
small businesses, but to large firms as well. Although HSAs are new
enough that comprehensive data are difficult to obtain, as of
January 2006, at least 3 million people were covered by
HSA-qualified plans sold by insurance company members of the
industry group America's Health Insurance Plans (AHIP). Of the
people covered by AHIP-related plans, about half purchased their
plans in the individual market and 14 percent through small
businesses.
Additional tax-code changes could make high-deductible
HSA-qualified plans even more attractive and affordable, further
strengthening incentives for more consumers to be well-informed,
cost-conscious health care decision makers. The President's 2007
budget aims to expand HSAs through proposals that include:
 Raising the HSA contribution limits up to the plan
out-of-pocket maximum. Current law allows contributions only up to
the deductible level, which is often less than half of the
out-of-pocket maximum. This change would further limit the
tax-induced bias against out-of-pocket spending for medical care.
It would also increase the attractiveness of HSA-qualified plans,
in particular for the chronically ill who have a higher probability
of out-of-pocket spending above their deductible.
 Further reducing disparities in tax treatment of HSA
contributions versus insurance premiums. Currently, individual
contributions to HSAs are excluded from income taxes but not payroll
taxes (employer contributions are excluded from both). The President
proposes to provide a new income tax credit equal to the payroll
taxes paid on the HSA contribution amounts. This will further
remove distortions that have encouraged first-dollar insurance
coverage. When combined with the first new proposal discussed above,
Americans with HSAs would be able to pay all of their out-of-pocket
expenses with pretax earnings.
 Equalizing tax preferences for purchasing HSA-qualified
insurance in the employer and individual markets. The President
proposes to exclude from income taxes the value of HSA-qualified
insurance premiums if purchased on the individual market. In
addition, taxpayers purchasing these policies on the individual
market would receive a new income tax credit equal to the payroll
taxes paid on the premium amounts. Thus, all taxpayers would receive
the same tax treatment of HSA-qualified insurance premiums, even if
working for one of the 40 percent of employers that do not offer
health benefits.
 Helping the chronically ill. In addition to allowing all
out-of-pocket expenses to be paid tax-free through an HSA, the
President also proposes allowing employers to make larger HSA
contributions for their chronically ill employees so that employers
can make HSA-qualified plans equally attractive to all employees
regardless of health status. Finally, the President proposes $500
million in annual grants to states to test innovative solutions to
subsidize insurance for the chronically ill, in order to enhance
the functioning of markets for individual insurance. For example,
states could use the funds for risk-adjusted premium subsidy
programs, or for creative enhancements of state high-risk pools
such as funding HSA accounts for enrollees.
 Enhancing affordability via a tax credit for low-income
people purchasing HSA-qualified insurance in the individual market.
The credit would be worth up to $1,000 for one adult, $2,000 for
two adults, or $3,000 for families (not exceeding 90 percent of the
premium). It would phase out at incomes of $30,000 for individuals
and $60,000 for families. The credit would be advanceable, paid
directly by the government at the time of insurance purchase.

Informed Consumers Are Better Consumers

It is important to provide incentives for consumers to choose
health care providers and services sensibly, but providing those
incentives does not guarantee that consumers will in fact be able
to make good choices. Consumers must also have access to the
information they need to make good health care decisions.
Key information includes:
 Provider prices. Few medical providers today advertise
their prices in a way that allows for comparison shopping.
Several insurers have taken an important step by beginning to make
available schedules of physician fees to their enrollees. Hospital
fees raise more-difficult issues, since prices negotiated between
hospitals and insurers are frequently subject to confidentiality
agreements, despite the fact that consumers eventually observe the
prices on bills presented to them after the fact. Of even greater
use to consumers would be information on "package prices" for
complete treatments of medical bundles or episodes. For example, a
knee replacement without unusual complications might have ten major
components of care, each of which is now billed separately. A
package price for the entire treatment would provide an estimated
cost for the entire operation, hospitalization, and follow-up
treatment. This information could be combined with revised billing
procedures, which would allow patients to identify more easily the
costs associated with the treatment they had received. The
President strongly supports efforts to increase price transparency
in the health care market. He has called for hospitals, physician
groups, insurers, employers, and other health groups to cooperate
in speeding the transition toward a market in which Americans can
easily obtain user-friendly and comparable information on prices
when shopping for health care.
 Data on provider quality and value. Price information by
itself is not sufficient for good decision making in the absence
of comparative quality data. There is growing interest in providing
accurate and usable measures of the quality of care offered by
individual health care providers such as hospitals and physician
groups. Great progress has been made by researchers in improving
the methodology for developing reliable measures, and insurers are
now helping to improve the effective dissemination of such data.
Measures that combine price and quality data into indicators of
overall value are not yet as well developed, but would be another
useful decision-making tool.
Better information would also be of use to providers of medical
services, who would then be better able to help their patients make
sound, cost-effective decisions. Examples include:
 Practice guidelines. One key barrier to more-efficient
health care spending is the lack of a research base on the
appropriate treatment in many medical situations. There is a clear
role for government in this area. For example, the Agency for
Health Research and Quality (AHRQ) is sponsoring comparative
effectiveness research studies relating to medical practice, as
authorized under the 2003 Medicare Modernization Act. Such research
can produce high returns in terms of improved health care
efficiency. Further work to translate such guidelines into
educational materials for health care consumers would also greatly
enhance the ability of consumers to make wise health care choices.
 Cost-effectiveness studies. If the usage of expensive but
low-value technologies is to be reduced by the actions of
better-informed consumers in consultation with their doctors,
then more information is needed about the cost-effectiveness of
various technologies and procedures, and about how
cost-effectiveness depends on particular factors such as the
patient's age and specific condition. Private insurers sponsor
some such studies, but the private sector will tend to underinvest
in this type of "public good" research. Government support for
research in this area, such as the research being conducted by
agencies such as AHRQ, has a strong economic justification.


Conclusion

As the United States grows richer and older and as new life-saving
technologies develop, Americans are likely to continue to spend a
rising share of their growing incomes on health. Indeed, our health
care spending overall has returned good value, with Americans
living longer and healthier lives. We could achieve this improved
health at lower cost, however, by promoting a greater role for
consumer decision making in health. Health Savings Accounts provide
one tool for doing so, by leveling the playing field for people who
prefer to save money by moving toward higher-deductible health
insurance policies. As health researchers, the insurance industry,
and government work to develop better consumer decision-making
tools, more consumers will be able to benefit from moving to such
plans. In the long run, the payoff to allocating health care
resources toward higher-value and more cost-effective care would
be great.