Order Code RS22437
May 5, 2006
CRS Report for Congress
Received through the CRS Web
Health Savings Accounts:
Some Current Policy Issues
Specialist in Social Legislation
Domestic Social Policy Division
President Bush is urging Congress to expand the availability and attractiveness of
health savings accounts (HSAs). Among other things, his FY2007 budget proposes (1)
new tax deductions and refundable income tax credits for purchasing individual
insurance policies that allow one to have an HSA, (2) increases in allowable HSA
contributions, (3) a refundable income tax credit for contributions not made by
employers, and (4) permission for employers to make larger contributions on behalf of
employees and family members who are critically ill. The Administration argues that
the proposed changes, which would have a 10-year revenue cost of $156 billion, would
slow rising health care costs by enabling additional consumers to play a more direct role
in their health care decisions.
In considering the proposals, Congress might review three important issues in light
of recent developments and the likelihood, given the initial market response, that HSAs
are here to stay. One is the tax treatment of contributions, which now favors higher
income individuals and families. A second is the purpose of health care savings, for
which it appears there are multiple and sometimes competing objectives. Finally, there
is the question whether HSA policies should favor or be neutral towards employmentbased and individual market insurance. This report may be updated to reflect further
analysis and debate.
Health Savings Accounts (HSAs) are one way that people can pay for their
unreimbursed medical expenses (deductibles, copayments, and services not covered by
insurance) on a tax-advantaged basis. The basic rules for 2006 are straight-forward,
though some aspects are complex:1
Individuals can establish and fund HSAs when they have qualifying high
deductible health insurance and no other coverage, with several
exceptions. The annual deductible for self-only coverage must be at least
For more information, see CRS Report RL33257, Health Savings Accounts: Overview of Rules
for 2006, by Bob Lyke.
Congressional Research Service ˜ The Library of Congress
$1,050, with an out-of-pocket ceiling for covered benefits not to exceed
$5,250; for family coverage, the deductible must be at least $2,100, with
an out-of-pocket ceiling not over $10,500. Deductibles need not apply
to preventive care.
HSA contributions may be made by account owners, their employers, or
others up to an annual limit of the lesser of $2,700 or 100% of the
insurance deductible for self-only coverage, or usually the lesser of
$5,450 or 100% of the deductible for family coverage. Individuals who
are at least 55 years of age but not yet enrolled in Medicare may
contribute an additional $700.
HSA tax advantages can be significant for some people: contributions
are deductible if made by individuals or excluded from their income if
made by or through employers, withdrawals are not taxed if used for
medical expenses, and account earnings are tax-exempt. Unused
balances may accumulate without limit.
HSAs are part of what some call consumer-driven health care. Although definitions
vary, in general this burgeoning movement involves a high deductible insurance plan and
a tax-advantaged savings account, ideally coupled with readily-available health care
information and a choice of providers whose charges and performance measures can be
known in advance.2 Proponents argue that these arrangements encourage more prudent
use of health care because people who pay expenses with their own money tend to consult
doctors less often and to seek out less expensive options. For example, they might use
home remedies for colds and be more willing to try generic drugs. Frugal but informed
use of health care is claimed to be the most practical way to hold down steadily rising
costs. Proponents also argue that HSAs enable people to pay for health care providers and
services of their own choosing, circumventing untoward restrictions of managed care. In
addition, it is said that HSAs can help people build up resources for future health care
needs. In light of growing demands on public budgets, HSAs may be a good way for
families to pay for costs not covered by Medicare as well as for long-term care expenses.
However, HSAs are controversial. The principal objection is the requirement that
one have high deductible insurance when contributions are made to the account. Critics
argue that high deductibles are more attractive to people who are young and healthy. As
they drop out of plans with lower deductibles, the premiums for those who remain steadily
rise, forcing people who are older and the chronically ill to pay ever higher rates. There
is concern as well that some people who need care will not have money in their accounts
to pay medical expenses below the deductible. They or their employer may not have
made contributions, or they may have spent the funds previously. In addition, critics
question why tax savings associated with the accounts should flow disproportionately to
higher income people.
There are sharp disagreements about likely reductions in health care use. While
people who switch to high deductible insurance generally use less care, the magnitude of
Other types of tax-advantaged savings accounts include health reimbursement accounts, flexible
spending accounts, and medical savings accounts. For similarities and differences, see CRS
Report RS21573, Tax-Advantaged Accounts for Health Care Expenses: Side-by-Side
Comparison, by Bob Lyke and Chris L. Peterson.
the reduction may be less than some imagine. Because most aggregate health care
spending is beyond usual plan deductibles, even a 10% reduction in spending below the
deductibles would probably reduce total spending by less than 3%.3 Still, a small
percentage reduction could represent significant savings for a country that now spends
more than $2 trillion on health care.
Some people might view their HSAs as insurance instead of savings, particularly as
their account balances grow. When HSAs are seen as readily-available (as well as taxfree) funds for paying cost-sharing requirements, they could lead to increases in health
care spending, just the opposite of the usual prediction. But assuming that people with
HSAs generally consult doctors less often, would this be good or bad? Lessons from the
oft-cited Rand Health Insurance Experiment 30 years ago still seem pertinent: costsharing deterred patients from seeking highly effective care as much as care that is rarely
effective. Patients became more frugal, but not necessarily better, consumers of medical
care. For most people, a reduction in health care use had minimal or no effects on health
status, but people who were both poor and sick were on average worse off. Today some
consumer driven health care arrangements provide ready access to information and
advice, whether this mitigates adverse consequences from cost-sharing, let alone leads to
sustainable improvements in health is yet unknown.4
Assessing the arguments about HSAs is difficult because there are not yet many
accounts. A recent report by America’s Health Insurance Plans (AHIP) estimates that
there were 3.2 million people who were covered by qualifying high deductible insurance
plans in January 2006; the number includes both policyholders and their family members.
The number of HSAs that have been established with these plans is not known, but it
likely would be smaller. The number of covered lives was triple the number estimated
in March 2005, which was double the number estimated in September 2004. So whatever
the number of accounts, it likely has been growing rapidly as well.5
Little is publicly known about the HSAs that do exist, including the incomes of
people who establish them, how much they and their employers contribute, and what
amounts are being used to pay health care expenses. Some of this information will
become available from the Internal Revenue Service (which is receiving information
returns from account trustees and supplementary tax forms from account owners), though
at first only for calendar year 2004 when HSA plans were just starting. It will be several
See Alan C. Monheit, “Persistence in Health Expenditures in the Short Run: Prevalence and
Consequences,” Medical Care, vol. 47 #1 (2003), pp. III-53 - III -64. Table 1 shows that in 1997
70.4% of aggregate expenditures were beyond a threshold of $4,155 (in 1997 dollars) and 84.3%
were beyond a threshold of $1,786 (also in 1997 dollars).
See CRS Report RL32467, Health Savings Accounts, by Bob Lyke, Chris Peterson, and Neela
Ranade, pp. 19-20; and Joseph P. Newhouse, Consumer-Directed Health Plans and the RAND
Health Insurance Experiment,” Health Affairs, vol. 23, no. 6, 2004, pp. 108-109.
For further information, see CRS Report RS22417, Data on Enrollment, Premiums, and CostSharing in HSA-Qualified Health Plans, by Chris L. Peterson. The AHIP study can be accessed
at [http://www.ahip.org/content/default.aspx?docid=15302]. The 3.2 million covered lives
represent a little more than 1% of the noninstitutionalized population under age 65.
years before there is adequate knowledge about how HSAs change over time, such as how
many account owners maintain eligibility to make contributions, how long it takes to
build balances beyond the insurance deductible, and so on.6
Much of the current debate over HSAs is speculative. While some is based on
actuarial experience with patient cost-sharing and economic analyses of employers and
households, more than a little, both for and against, appears to be founded on
unsubstantiated assumptions and philosophic differences. Even the best analyses involve
assumptions about consumer rationality, insurance markets, and the response of health
care providers that may not be borne out.
Comparisons differ. Advocates often focus on average cases, which make HSAs
look attractive because most people are healthy in any one year; opponents draw more
attention to the smaller number of people who need continual care or suddenly have a
serious accident or illness. The former perspective gives more weight to what are likely
to be typical economic gains, the latter to atypical economic losses. Similarly, compared
with no insurance at all, high deductible insurance usually reduces short-term financial
risk; compared with most comprehensive coverage, it usually increases it.
Some Current Issues
As described in the “Summary,” President Bush is proposing a number of steps to
expand the availability and attractiveness of HSAs. The Administration argues that the
proposed changes, which would have a 10-year revenue cost of $156 billion, would slow
rising health care costs by enabling additional consumers to play a more direct role in
their health care decisions. Regardless of the extent to which this would occur, Congress
might use the proposals as an opportunity to review several features of HSAs in light of
the likelihood, given the initial market response, that they are here to stay. In an era of
steadily rising health care costs, declining coverage from employers, and growing
demands on public budgets, it would seem prudent to periodically examine how HSAs,
like other financing arrangements, can meet the country’s health care objectives.
Tax Treatment of Contributions. HSA contributions are tax deductible if made
by individuals and excluded from their income if made by employers. Because tax
savings from deductions and exclusions are based on marginal tax rates, they are often
larger for higher income individuals and families. For example, single filers with taxable
incomes of $100,000 would save $280 for making a $1,000 contribution, while those with
taxable incomes of $20,000 would save $150. Basing tax savings on marginal rates might
be justified for catastrophic health care expenses (and by extension for catastrophic health
insurance) because, conceptually in a progressive tax system, deductions for economic
losses ought to be subject to the same rates that apply to equivalent economic gains.
Some inferences might be cautiously drawn from information released by insurers and benefit
consultants. For example, a sample of several thousand HSA-eligible insurance plans sold in the
first half of 2005 through [http://www.eHealthInsurance.com] (an online marketing site) shows
that about 6% of enrollees had annual incomes under $15,000; 22% under $35,000; 42% under
$50,000; 63% under $75,000; and 78% under $100,000. Bearing in mind the sample might not
be representative, the distributions suggest that HSAs are likely to be held by individuals with
a wide range of incomes though more so by people with higher income. Source: eHealth, Inc.
Catastrophic health care expenses are almost always seen as losses in a way that large,
discretionary expenditures are not. But it is more difficult to justify tax savings based on
marginal rates for health care expenses below catastrophic levels, the very expenses that
HSAs are supposed to cover. Indeed, it would seem that if the government were to give
everyone a certain sum to pay these expenses, the first thought, prior to considering
special needs or other subsidies, would be that everyone should get the same amount.7
One of the President’s proposals would allow a small refundable income tax credit
to offset employment taxes on HSA contributions not made by an employer. The
objective is to provide equitable treatment for those who make after-tax contributions
from income on which they earlier paid employment taxes. But if equity is a goal, it
would seem more important to replace the current exclusion (for employer contributions)
and deduction (for individual contributions) with a refundable tax credit, such as 20% of
the contributed amount, that is available to everyone. Although not providing complete
equity, this change would generally seem fairer for this type of welfare transfer. It might
also provide more incentive for lower income individuals to make contributions.8
Savings Objectives. The President would raise contribution limits up to the outof-pocket ceilings for covered benefits, $5,250 for self-only coverage and $10,500 for
family coverage. The higher limits would permit contributions almost twice as large as
current law and theoretically enable some people, even in the first year with an HSA, to
eliminate nearly all risk they could not pay the insurance deductible.9 After several years,
only negligible exposure for deductibles and copayments would remain. Attractive as this
might sound to some, as a practical matter few people have the resources and inclination
to save these amounts, even with the tax deduction.10 The higher limits would likely
make a difference primarily to better-off households, especially if they could shift existing
savings (or divert future savings) from other taxable and nontaxable accounts.
Of course, there are arguments to be made for people having savings they can use
for health care, particularly as they age. If anything, Americans may be underestimating
the resources they will need simply to continue the benefits and services they now have.
But since there is no consensus on the objectives of accumulating health care savings, it
The example in the text compares savings at two statutory rates, 28% and 15%. Some taxpayers
with lower taxable incomes would have a 10% rate (and thus even smaller savings), while some
with higher incomes would have 33% or 35% rates (and thus larger savings). Taxpayers with no
taxable income, or whose tax liability is now offset by tax credits, would receive no savings.
(Taxable income is income after exclusions and deductions have been taken into account.) In
most states, there would be additional savings from state income taxes.
A level percentage credit might not seem sufficiently equitable since higher income taxpayers,
having greater ability to make contributions, would still on average receive larger transfers. At
the same time, a level percentage credit would represent a larger percentage of the income of
lower income taxpayers, which by itself could give them added incentive to contribute.
All risk might not be eliminated because accounts can be used for expenses not covered by
insurance. Even if the money is not saved, larger contributions would allow more health care
spending to occur on a tax-advantaged basis, seemingly the opposite of the ultimate goal.
Average contributions to other tax-advantaged savings vehicles are often far below the
maximums allowed. See CRS Report RL33116, Retirement Plan Participation and
Contributions: Trends from 1998 to 2003, by Patrick Purcell.
is difficult to decide whether particular policy changes are appropriate. If the principal
objective is to ensure that families have enough money to pay their deductibles, perhaps
for several years, maybe there should be limits on the cumulative savings allowed. Even
adding other tax-free uses of HSAs (such as paying health insurance premiums when
unemployed), it would seem that at some point accumulations would be enough. Having
no limits whatsoever allows accounts to be used as tax shelters. Alternatively, if the
objective is to encourage families to save for needs beyond health care, perhaps the
contribution limits should be coordinated with those for other tax-advantaged accounts.
Some observers have proposed that families would be better served by a single account
that could be used for health, education, home purchases, and other priorities. A single
account would simplify the tax code, but the loss of dedicated health care accounts might
make it more difficult to determine appropriate insurance cost-sharing requirements.
The President is also proposing that employers be allowed to make larger HSA
contributions on behalf of employees and family members who are chronically ill.
(Employers are not required to contribute to employees’ HSAs, but if they do, their
contributions generally must be the same dollar amount or the same percentage of the
insurance deductible.) Although it is yet uncertain how this idea would be implemented,
the proposal suggests that there may be ways to “insure” account owners against
unexpected costs that otherwise would make them reluctant to have high deductible
insurance. This may be especially important for families without much other savings.
Alternatively, some employers would like to make larger HSA contributions if their
employees adopt healthy lifestyles, such as losing weight, stopping smoking, and
exercising more. Although these two ideas may seem contradictory (in the first,
additional money is available for illness, while in the other it is available for being
healthy), together they indicate that HSAs might be fine-tuned to increase the odds of
achieving desirable outcomes. Presently there is much to be said for letting insurers and
employers experiment with guarantees and incentives to learn what is cost-effective. At
some point, though, Congress might review how accounts are being used to see who is
benefitting from larger contributions and whether the payments are consistent with its
Market Neutrality. Several of the President’s proposals attempt to provide equal
tax treatment for HSA plans sold in the individual market and those offered by employers.
On equity grounds alone, there is much to be said for this. Arguably, tax benefits for
health care should not depend on whether one happens to be employed, let alone works
for certain employers. Nonetheless, some argue that employment-based coverage is
preferable to individual market insurance since it provides better risk pooling. They
would advocate that preferential treatment for employment-based plans be retained.
However, the President’s proposals would not extend equal tax treatment to
individual market insurance that is not HSA-qualified (i.e., does not meet the minimum
deductible and maximum out-of-pocket requirements). This would not be consistent with
current tax policy, which does not differentiate tax benefits by type of coverage. One
argument for the President’s position is that it is important to start moving the population
toward a form of coverage that would reduce health care spending. Others, however,
would say that HSA plans are too new to predict outcomes with any certainty and it would
be better not to give tax incentives that favor one form of coverage.