by Michael Tanner
Michael Tanner is director of health and welfare studies at the Cato Institute.
Executive Summary
As the movement for medical savings accounts (MSAs)
picks up speed in Congress, critics of consumer-based health
care reform are mounting a counterattack. An examination of
the evidence shows that their criticisms of MSAs are just
plain wrong. For example:
* Critics claim health care has become so complex that con-
sumers are no longer capable of making cost-conscious deci-
sions about their treatment. However, numerous scientific
studies show that health care consumers can and do make cost-
conscious decisions when given a financial incentive to do
so.
* Critics say consumers will forgo necessary or preventive
care to save money in their medical savings accounts, but
studies show that MSAs do not deter preventive care. Rather,
savings result from reduced use of optional services and
cost-based selection among competing providers.
* According to critics, MSAs would attract the healthy,
leaving the sick with conventional insurance. If so, that
"adverse selection" would drive up the cost of traditional
insurance. However, companies currently using MSA-style
health plans have not had significant problems with adverse
selection.
* Critics claim MSAs are regressive, providing benefits pri-
marily to the wealthy. Our current system of providing a tax
break only for employer-provided insurance is far more re-
gressive.
MSAs represent a significant step in solving the prob-
lems facing our health care system. Supporters of MSAs
should not be distracted by flawed and misplaced criticisms.
Introduction
Despite the defeat of the Clinton health care plan, the
need for significant health care reform remains. Health
care continues to cost too much. Approximately 40 million
Americans still don't have health insurance. Millions of
working men and women live in fear that if they lose their
jobs they will lose their health insurance.
One of the most popular alternatives for health care
reform likely to be considered by the 104th Congress is
medical savings accounts (MSAs). Support for this concept
cuts across party and ideological lines, with Democrats as
well as Republicans, liberals as well as conservatives,
supporting MSA proposals.
It is easy to see why this idea is so popular. MSAs
would allow individuals to save money in a tax-exempt ac-
count, in much the same way they can in independent retire-
ment accounts (IRAs) now.(1) The person could use that money
to pay routine medical expenses. Then, instead of an expen-
sive first-dollar insurance policy, he or she could purchase
a relatively inexpensive catastrophic insurance policy to
protect against major medical expenses.
It costs an employer more than $5,400 to provide health
insurance for a typical American worker today, his or her
spouse, and two children.(2) Wouldn't it be better if, in-
stead, the employer bought a catastrophic policy (with, say,
a $3,000 deductible) for approximately $2,400 and paid the
worker the $3,000 difference? The employee could then put
that money in an MSA. (See Figure 1.) Any money that wasn't
spent would roll over to the next year. Since 90 percent of
Americans spend less than $3,000 per year on health care, in
a very short time the worker would have a tidy pool of money
available to use in the future. When the balance reached a
certain level, the worker could transfer the funds to an IRA
or other retirement fund.
Most proposals for health care reform focus on govern-
ment, physicians, hospitals, and insurers. MSAs are unique
because they focus on the most important participant
in the health care system--the consumer.
MSAs would establish an incentive for consumers to act
responsibly in making their health care decisions. Consumer
behavior is a key component in controlling health care
costs. Our current system discourages cost-oriented deci-
sionmaking and encourages overconsumption and overuse of
health care services.
Figure 1
How MSAs Work
[Graph Omitted]
Source: Patient Power.
Under our current third-party insurance system, most
health care consumers do not pay for their health care.
Nearly 95 percent of hospital bills and more than 80 percent
of physician fees are paid for by private health insurance.
On the average, 76 cents of every dollar used to purchase
health care is paid by someone other than the consumer who
purchased it.(3) (See Figure 2). As a result, consumers have
little incentive to question costs and every incentive to
demand more services.
However, with MSAs, patients would be spending more of
their own money, giving them an incentive to become cost-
conscious consumers.
A second advantage of MSAs is that they would be com-
pletely portable. One of the most serious problems of our
current health care system is that insurance is so closely
linked with employment.(4) That means that if you lose your
job or change jobs, you are in danger of losing your insur-
ance. Of the estimated 37 million Americans without health
insurance at any given time, half are uninsured for four
months or fewer, and only 15 percent are uninsured for more
Figure 2
Percentage of Personal Health Expenses Paid by
Third Parties, 1965 and 1990
[Graph Omitted]
Hospital
1965 83.2%
1990 95.0%
Physician
1965 38.4%
1990 81.3%
All Services
1965 48.4%
1990 76.7%
Source: Patient Power.
than two years.(5) (See Figure 3.) Most of these temporary,
short-term spells without insurance happen to individuals
between jobs. With an MSA, individuals would have funds
available to pay for health care and health insurance during
such temporary interruptions.(6)
Moreover, expenses paid out of an MSA would entail no
insurance administrative cost. Insurance is a very ineffi-
cient way to pay for small or routine health expenses.
Significantly more administrative costs are involved in
processing a large number of small claims than in processing
a few large claims with an equal dollar value. Indeed,
premiums generally fall as average claim size increases.(7)
MSAs would cut insurance companies out of the vast majority
of health care transactions, particularly small claims where
insurance is least efficient. That would reduce both the
overall cost of health care and the paperwork burden on
doctors.
Figure 3
How Long Do People Go without Health Insurance?
[Graph Omitted]
Uninsured for Fewer Than 6 Months 50%
Uninsured for Less Than 1 Year 72%
Uninsured for Fewer Than 2 Years 82%
Source: "Spells without Health Insurance: Distributions of
Durations and Their Link to Point-in-Time Estimates of the
Uninsured," Inquiry 27 (Fall 1990).
However, even as MSAs have gained popularity, those who
support a government takeover of the American health care
system, supported by large insurance companies who fear a
loss of premium income if MSAs become widespread, have
mounted a sustained attack on this free-market approach to
health care reform. They allege that MSAs would not reduce
health care costs, could bankrupt the health care system,
and would primarily benefit the wealthy.
Yet an examination of the most common criticisms of
MSAs shows that these critiques are either misleading or
simply wrong.
Are Consumers Stupid?
From Hillary Rodham Clinton to Alain Enthoven, critics
of MSAs have argued that health care has become too complex
for average patients to make rational decisions about their
treatment. This argument generally takes one of two some-
what contradictory tracks: a) consumers will not reduce
spending because they will mindlessly accept any treatment
proposed by their doctor, and b) consumers will forgo neces-
sary and preventive treatment to save money.
Rep. Pete Stark (D-Calif.), for example, has said that
patients either "feel they are invincible [if healthy]" or,
when sick, are "absolutely brain dead, sniveling, begging
and fantasizing ills and pains."(8) Neither is supported by
the facts.
Some of the most compelling evidence that consumers can
and do make cost-based decisions on health care comes from a
study performed by the RAND Corporation in the late 1970s.(9)
That study assigned families to four health insurance plans
with differing copayment provisions and deductibles. Some
families had no copayment or deductible, meaning that the
plan paid all their medical bills, while other families had
to pay up to 95 percent of the cost of their medical bills,
until their bills reached a deductible of $1,000 in 1973
dollars, which is the equivalent of approximately $2,850 in
today's dollars.
The RAND researchers observed how the different copay-
ment rates influenced the use of medical resources by 2,500
families for three to five years. They concluded that "the
data from the Health Insurance Experiment clearly shows that
the use of medical services responds to changes in the
amount paid out-of-pocket."(10) In particular, families with
no copayment used 53 percent more hospital services (mea-
sured in dollars) and 63 percent more visits to doctors,
drugs, and other services than did the families with the 95
percent copayment. Overall, the total use of medical re-
sources was 58 percent greater for the group with no copay-
ment, despite virtually identical health outcomes.
Even smaller copayment rates produced savings. The
study found that an individual with no copayment spent 18
percent more on health care than an individual with a 25
percent copayment. (For the results of the RAND experiment,
see Table 1.)
The RAND study essentially confirms earlier studies by
Martin Feldstein and others.(11) In addition, studies of
specific health care services such as mental health(12) and
prescription drugs(13) have shown that consumers will make
cost-conscious decisions if given an incentive to do so.
Several studies have called into question the degree to
which physicians are actually able to induce demand.(14)
University of Washington economist Michael Morrisey, for
example, points out that the inflation-adjusted median
Table 1
RAND Health Insurance Experiment: Increased Spending
over Plans with No Copayment
95% Copayment 25% Copayment
Physician visits 167% 137%
Outpatient expenses 167% 131%
Admissions 129% 122%
Inpatient expenses 130% 110%
Probability of any use 128% 110%
Probability of any inpatient use 130% 123%
Total expenses 145% 118%
Source: Manning et al. (1987)
Note: "95% Copayment" requires out-of-pocket expenditures of 95
cents on the dollar for covered services. "25% Copayment" requires
out-of-pocket expenditures of 25 cents on the dollar for covered
services.
income of physicians has declined since 1975, suggesting
that physicians are not able to generate an unlimited demand
for their services regardless of price.(15) Others suggest
that even if physicians are able to induce demand, consumer
decisions will still be influenced by their perceived costs
as distorted through third-party payments.(16)
Certainly, a person suffering a heart attack or in-
volved in an automobile accident is not going to comparison
shop for the best price. But fewer than 15 percent of
health care expenditures are emergency in nature.(17) For
nonemergency services, it is possible for consumers to shop
and compare. For example, one study found that the cost for
cataract surgery in Illinois ranged from $650 to $5,674
depending on the hospital; hernia surgery ranged from $404
to $4,329; and mammograms ranged from $35 to $178.(18)
MSA critics also warn that consumers lack the informa-
tion and expertise necessary to make such decisions. Clear-
ly, many patients will have to rely on the advice of a
physician that they know and trust. Such patient-physician
relationships have long been at the heart of health care,
but are not unique to medical goods and services. Few
Americans know all the details of automotive repair. When
their cars break down, they rely on the advice of mechanics
whom they know and trust.
Moreover, the market is already generating an increas-
ing number of resources to provide consumers with informa-
tion on health care prices, quality, and availability.
Automated medical information lines offer prerecorded facts
and figures, while live lines such as Ask a Nurse, Doctors
by Phone, and Pharmacy Question? offer person-to-person
contacts with health care professionals.(19) As more and more
consumers take control of their health care decisions, such
information resources can be expected to proliferate.
Academic studies of consumer ability to make cost-
conscious health care decisions are confirmed by real-world
experience. Several companies have established MSA-style
insurance plans and have realized significant savings as a
result of changed consumer behavior by their workers. Among
these are Golden Rule Insurance Company, Dominion Resources,
Forbes Inc., Quaker Oats, and Indresco Corporation.(20)
If it is shown that consumers do change behavior ac-
cording to financial incentives, MSA critics then move on to
part (b) of their argument: consumers will forgo necessary
or preventive care to save money. As a result, they will
end up sicker and cost the system more money in the future.
However, once again the facts dispute the critics'
contention. Analysts studying the RAND Health Insurance
Experiment concluded, "We reject the hypothesis that less
favorable coverage of outpatient services increases total
expenditure . . . by deterring preventive care."(21)
The RAND experiment found that the reduced expenditures
are not caused by individuals forgoing truly necessary
health care. Health outcomes were virtually identical.(22)
Rather, the savings resulted from a reduced use of optional
services and cost-based selection between competing provid-
ers.
There is even evidence that MSAs will increase the
likelihood of seeking preventive care, particularly among
low-wage earners. Under conventional insurance, individuals
receive no reimbursement until they have met the deductible.
That places all the spending disincentive on the first
expenditures of the year, expenditures that are most likely
to involve preventive care. For low-wage earners, who may
lack the resources to pay for these expenses out of pocket,
that disincentive creates a strong likelihood that they will
forgo preventive care.
MSAs, in contrast, flatten the spending curve, spread-
ing the spending disincentive over the entire $3,000 rather
than focusing it on the first expenditure.
Indeed, MSAs would actually provide low-wage earners
with a pool of money that they could use to pay for preven-
tive care. A survey of Golden Rule employees with MSAs
found that 20 percent used their MSAs for medical services
that they would not have purchased with traditional plans.
Yet, overall health spending declined.(23)
MSAs and Health Care Costs
Critics of MSAs also claim that even if consumer behav-
ior changed it would have little impact on overall health
care costs because the majority of costs occur at a level
above the $3,000 deductible envisioned by MSA supporters.
As Alain Enthoven puts it, "The $3,000 deductible does
nothing to motivate reduction of expenditures on high-cost
treatments.... Once someone is told she is pregnant, or has
cancer, or must be admitted to a hospital, she might as well
write off the $3,000 and say 'bring on more technology.'"(24)
It is clearly true that the vast majority of health
care expenditures are made by only a tiny fraction of the
population. The evidence suggests that the spending curve
steepens sharply after $3,000. In a typical insurance pool
about 4 percent of the people spend approximately 50 percent
of the health care dollars. (See Figure 4.)
Even so, a substantial portion of health care expendi-
tures, between one-third and one-half of all health care
spending, is on bills below $3,000.(25) If spending were
reduced on just this portion of health care expenditures,
overall costs would be significantly affected.
Second, the incentive structure created by MSAs could
prevent some expenditures from ever reaching the $3,000
level. For example, if an individual seeking to preserve
the $3,000 in an MSA avoids an unnecessary $6,000 operation,
she or he reduces spending above the $3,000 level as well as
below.(26)
Perhaps most important, this argument ignores the
impact that reduced prices will have throughout the health
care system. If the price of an x-ray is reduced, it will
Figure 4
Distribution of Medical Expenses among 50 People
[Graph Omitted]
Source: Patient Power.
Note: Assumes a $250 deductible and a 20 percent copayment on the
next $5,000 of expenses. Period of coverage is one year.
be reduced for people who spend more than $3,000 in a year
as well as for those who don't. There is overwhelming
evidence that prices will be reduced in response to cost-
sensitive purchasing by consumers. For example, a study by
Joseph Newhouse and Charles Phelps found that a 10 percent-
age point increase in out-of-pocket expenditures resulted in
a 2 percent reduction in the price of physician services.(27)
Likewise, several studies have found that increased third-
party payments have led to price increases, making it likely
that decreased third-party payments would lead to price
decreases.(28)
That criticism also ignores the tremendous potential
for administrative savings with MSAs. Administrative costs
amount to between 19.3 and 24.1 percent of total American
health spending.(29) Considerably more claims are submitted
for expenses below $3,000 than above $3,000. Thus adminis-
trative expenses are disproportionately concentrated in low
dollar claims. By eliminating much of the paperwork and
other administrative costs associated with third-party
payment of these claims, overall health care spending could
be reduced significantly. Some estimates indicate that MSAs
could save as much as $33 billion per year in reduced admin-
istrative costs.(30)
Will MSAs Bankrupt the System?
One of the most widely circulated recent criticisms of
MSAs has been leveled by John Burry, CEO of Blue Cross and
Blue Shield of Ohio.(31) Burry claims that MSAs "would create
a large financial shortfall that would bankrupt our health
care system."(32) Because Burry's criticism of MSAs appears
more scientific than others, it is worth examining his
claims in detail.
Burry bases his argument on his analysis of the claims
experience of 38,729 families currently insured by Blue
Cross and Blue Shield of Ohio. According to Burry, total
claims by this group were $159 million, or an average of
$4,113 per family. Approximately 10 percent of the families
were responsible for 55 percent of this total, or approxi-
mately $88 million, with the remaining 90 percent of fami-
lies incurring claims of approximately $71 million. For the
10 percent incurring the highest claims, the average charges
were $22,747, compared with an average of $2,045 for the
remaining 90 percent.(33)
Burry then assumes that a catastrophic insurance policy
purchased in conjunction with an MSA would cost $1,200. His
assumption is based on the Clinton administration's claims
that the cost of health care for the average American family
is $4,200 (very close to the $4,113 in charges for Ohio).
Burry suggests that if the family subtracts $3,000 for
deposit in an MSA, there remains $1,200 to purchase a cata-
strophic policy. That would result in a total premium
payment of $46.5 million.(34)
Next, Burry estimates that 68 percent of the families
would spend less than the $3,000 in their MSAs. These
families would average spending $961 from their MSAs, a
total of $25.3 million. The remaining 32 percent of fami-
lies would spend all $3,000 in their MSA, totaling $37.2
million.(35) Thus, the total spending from MSA funds would be
approximately $62.5 million. Adding the $62.5 million in
MSA spending to the $46.5 million in catastrophic insurance
premiums yields total health care spending of $109 million.
But, since claims totaled $159.3 million, this results in a
$50.3 million deficit.(36) Projecting this deficit over the
entire U.S. health care system leads to a potential short-
fall of $83.6 billion.(37)
However, Burry's analysis is far from accurate. For
example, Burry confuses "billed charges" with insurance
premiums. Under Burry's example, the families had average
charges of $4,113. However, assuming an industry standard
loss ratio of 75 percent, the families would pay an average
premium of $5,484.(38) That means that after putting $3,000
in an MSA, the family would have $2,484 left over, more than
enough to pay for a catastrophic policy. This would yield
total premiums of $96.2 million. Adding this revised premi-
um total to the $62.5 million in MSA expenditures equals
total expenditures of $158.7 million. If total charges in
the system were $159.3 million, the results are a statisti-
cally meaningless deficit of only $600,000.
Second, Burry assumes that the family currently has no
out-of-pocket exposure. But that is highly unlikely.
Nearly all policies sold by Blue Cross and Blue Shield of
Ohio require both copayments and deductibles. If the fami-
lies in question averaged only $500 in out-of-pocket expens-
es, there would be an additional $19.4 million in the sys-
tem, providing an actual surplus of nearly $19 million.
Most important, all of this assumes no change in behav-
ior among health care consumers and therefore no reduction
in health care expenditures. But the entire concept of
MSAs is that it would change consumer behavior.
Similar flaws can be found in a discussion of MSA
distributions by the Medical Savings Account Working Group
of the American Academy of Actuaries. The actuaries con-
clude that "the savings to employers of replacing low-
deductible plans with very high-deductible plans would be
substantially less than the change in the deductible that
the workers would have to pay."(39) Their logic is very simi-
lar to Burry's. Because "10 percent of covered individuals
account for between 70 and 80 percent of all health insur-
ance claims" subsidies by nonconsuming insured individuals
are required to hold down premium costs for the catastrophic
policies.(40)
However the actuaries also repeat Burry's errors. "For
simplicity," the actuaries made three very flawed assump-
tions:
1. Current plans "pay all claims (i.e., that the de-
ductible amount and coinsurance are all zero)."
2. "There are no administrative costs or profits asso-
ciated with the (traditional) plan."
3. There is "no change in behavior" for people with a
medical savings account.(41)
Every one of these assumptions is incorrect. As noted
above, nearly all current insurance policies require copay-
ments and deductibles. Heavy administrative costs are
associated with traditional insurance plans, as much as 33.5
cents of every premium dollar, according to some esti-
mates.(42) And, as stated above, the purpose of MSAs is to
change behavior. Indeed, elsewhere in their report the
actuaries themselves propose that MSAs will make Americans
"more conscious of their health care and more thoughtful
about the casual utilization of care" providing "a substan-
tial role in reducing health care inflation."(43)
The actuaries do not agree with Burry that MSAs would
bankrupt the health care system, but merely contend that the
cost of catastrophic insurance would increase, leaving a gap
between the savings an employee would have to deposit in the
MSA and the deductible. However, even if this did occur,
the vast majority of individuals with low medical expenses
would be unaffected. Those few who did face out-of-pocket
expenses would be likely to have less exposure than under
traditional deductibles and copayments. Finally, because
there would be additional contributions to the MSAs each
year, any shortfall is likely to be a one-time occurrence.
Are MSAs Regressive?
According to critics, MSAs would primarily benefit "the
rich, because they get the biggest benefit from tax-free
investments."(44) They argue that MSAs, like the IRAs they
resemble, disproportionately benefit taxpayers in the high-
est tax brackets, because all tax deductions are worth more
to those individuals.(45)
However, it is today's system of tax breaks only for
employer-provided insurance that really favors the wealthy.
In this country we give American workers and their families
very generous tax relief on their medical expenses, but only
on two conditions. First, they must obtain their medical
care through health insurance. And second, they must obtain
their health insurance through their employers.
As a result, if one works for a Fortune 500 corporation
that provides an all-inclusive first-dollar insurance plan,
the worker receives it tax free. But a small business
owner, waitress, or truck driver has to pay for health care
out of pocket or purchase insurance, and receives no tax
break at all. MSAs would help to level this playing field,
giving these workers a chance to save on a tax-free basis
for the health care.
A study by the Congressional Budget Office noted how
regressive the current tax structure for health care is.
The CBO points out:
The tax exclusion provides a subsidy for employ-
ment-based health insurance premiums that increas-
es with the size of premiums, the share of the
premiums paid by the employers, and the marginal
tax rate. These factors all increase with in-
come.... Moreover, families with higher incomes
are much more likely to have employment-based
health insurance than families with lower in-
comes.(46)
A study conducted by Lewin-VHI in 1994 for the Heritage
Foundation found that households with incomes greater than
$50,000 per year received $35 billion in tax relief, while
households with incomes under $20,000 received only $2.7
billion in tax relief.(47) Urban Institute economist Eugene
Steurle concludes that the current tax treatment of health
care provides a family in the top fifth of income earners
almost six times as much benefit as a family in the lowest
quintile.(48) (See Figure 5.)
All tax deductions are by nature regressive. MSAs are
certainly no more so than the current system. Indeed, the
fact that contributions are limited to $3,000 makes them
slightly less regressive than the current open-ended tax
exclusion. Congress can take steps to replace the current
tax exclusion with a universal health care tax credit. Such
a credit could even be made refundable. However, such
actions are independent of support for MSAs.
A corollary of the argument over regressiveness is the
assertion by some critics that the poor would lack the money
to contribute to MSAs. The Center for Budget and Policy
Priorities notes that pre-1986 IRAs were used primarily by
those with higher incomes. In 1986, for example, 66 percent
of individuals with incomes in the top 4 percent of taxpay-
ers took IRA deductions, while only 4 percent of workers
Figure 5
Estimated Average Value per Recipient Household of
Federal Tax Exclusions for Employer Health Insurance,
Fiscal Year 1992
[Graph Omitted]
Quintile of Household Income
Lowest $270
Second $525
Third $690
Fourth $1,025
Highest $1,500
Source: C. Eugene Steurle, "The Search for Adaptable Health Policy
through Finance-Based Reform," in American Health Policy: Critical
Issues for Reform, ed. Robert Helms (Washington: American
Enterprise Institute, 1993).
with incomes below $15,000 did so. They contend that MSAs
would work the same way.(49)
However, contributing to an IRA required an individual
to divert income that could have been used for another
purpose. It was, in effect, an additional expenditure. For
the vast majority of Americans who receive their insurance
through their employer, MSAs merely represent a different
way of receiving their insurance benefits. No additional
expenditure would be required.
Those individuals who do not receive employer-provided
insurance would have to make an additional expenditure.
However, at least they would receive a tax break for that
expenditure, which is more than they receive under the
current unfair system. The question of whether and to what
degree there should be subsidies to assist these individuals
to buy insurance is independent of MSAs.
Finally, it should be noted that if one is concerned
about the impact of health care reform on the poor, no one
would benefit more than the poor from the lower health care
prices that MSAs are designed to bring about.
The Adverse Selection Problem
The final criticism leveled by critics of MSAs is that
such a plan would appeal primarily to the young and healthy,
leading to adverse selection that will drive up the cost of
traditional first-dollar coverage.
In analyzing MSAs in 1994, the Congressional Budget
Office warned that "the availability of the catastrophic-
plus-MSA option would exacerbate the problem of adverse
selection."(50) According to the CBO, as healthy people in-
creasingly choose MSAs combined with catastrophic insurance,
the pool of people purchasing traditional low-deductible
insurance will become steadily sicker. To compensate,
insurers will have to increase the cost of such insurance.
Eventually, low-deductible insurance will become so expen-
sive as to "threaten the existence of standard health insur-
ance."(51)
The problem of adverse selection becomes even worse if
individuals can move freely from one insurance policy to
another. As the Center for Budget and Policy Priorities
warns, "There would be strong incentives to accumulate tax-
advantaged savings [in medical savings accounts] at times
when few health care expenses are anticipated.... When medi-
cal expenses became more probable, the individual could
simply switch to comprehensive coverage and keep the MSA
accumulation as savings."(52)
The experience of companies currently using MSA-style
health plans has not shown significant problems with adverse
selection. At companies such as Golden Rule, employees with
chronic illnesses have chosen MSAs nearly as frequently as
have healthy workers. The reason is that the copayments and
deductibles under traditional insurance leave those workers
with chronic conditions facing significant expenses every
year. Under such a traditional policy, a worker with a $250
deductible and a 20 percent copayment up to $3,000 and a
chronic condition costing more than $15,000 per year can
anticipate paying $3,250 out of pocket every year. With an
MSA, the worker would have little or no out-of-pocket ex-
pense.
In addition, the likely alternative to MSAs, managed
care with its limits on the choice of physician and treat-
ment and restrictions on access to specialists, can be
unpopular with the chronically ill. Indeed, a study from
the National Center for Policy Analysis demonstrates that it
is the chronically ill who are most likely to be short-
changed under managed care.(53) Therefore, MSAs may well be a
popular alternative for this group.
However, if adverse selection did occur and increase
the cost of low-deductible policies, that would not neces-
sarily be a bad thing. It is traditional, low-deductible
insurance that is driving up the cost of health care. If
such policies become unsustainable and most Americans move
to an MSA plus catastrophic coverage, the result will be
lower health care costs. The market will merely have pro-
vided an incentive for a socially beneficial change in
behavior.
The issue of individuals using MSAs when healthy and
shifting to traditional insurance when they become sick
would exist primarily in an environment where insurance was
not properly risk rated. If insurers are permitted to base
premiums on an individual's health status and to refuse
coverage to individuals with preexisting conditions, it
would not be possible to "game" the system in this manner.
Therefore, this critique becomes an argument against commu-
nity rating and "guaranteed issue," not against MSAs.
Conclusion
MSAs are not a "magic bullet" that will solve all our
health care problems. However, they will have a significant
impact on reducing health care costs, while expanding access
to care and preserving consumer choice and the quality of
our health care system.
Ultimately, only three entities can control health care
costs: government, through rationing; insurance companies,
through managed care (another form of rationing); or indi-
vidual consumers. MSAs provide the incentive for individual
consumers to make cost-conscious decisions. Recent criti-
cisms of MSAs are not accurate.
1. Consumers are capable of making cost-conscious
decisions about health care purchases.
2. In making cost-conscious decisions, consumers
do not forgo necessary or preventive care.
3. MSAs will reduce costs throughout the health
care system, not just on spending below $3,000.
4. MSAs will not bankrupt the health care system.
5. MSAs are no more regressive than the current
health care system.
6. The adverse selection problem has been over-
stated.
Medical savings accounts may not be perfect, but they
represent a significant step in solving the problems facing
our health care system. Supporters of MSAs should not be
distracted by flawed and misplaced criticisms.
Notes
(1) For a full discussion of MSAs see John C. Goodman and
Gerald Musgrave, Patient Power: Solving America's Health
Care Crisis (Washington: Cato Institute, 1992).
(2) KPMG Peat Marwick Survey of Employer-Sponsored Health
Benefits, 1994.
(3) Goodman and Musgrave, p. 77.
(4) The link between employment and insurance is largely a
historic accident. During World War II, American businesses
simultaneously faced a labor shortage and wage-price con-
trols. As a result they began to offer health insurance
benefits as a way to lure workers. After the war, the
practice was sufficiently widespread that it became en-
sconced in the tax code, making employer-provided health
insurance a tax-free benefit, while individually purchased
insurance received no tax break. Today, nearly 85 percent
of people with health insurance receive it through their
employers. For a detailed look at how this policy devel-
oped, see Goodman and Musgrave, pp. 137-63.
(5) Katherine Swartz and Timothy McBride, "Spells without
Health Insurance: Distributions of Durations and Their Link
to Point-in-Time Estimates of the Uninsured," Inquiry 27
(Fall 1990).
(6) Under current law, individuals who leave an employer who
employs more than 25 workers and provides health insurance
are entitled to pay the premiums and extend their coverage
for up to 18 months. However, most newly unemployed persons
are not able to pay the premiums and allow their insurance
to lapse. An MSA would provide a pool of funds that could
be used to pay the premium during this period of unemploy-
ment.
(7) See, for example, Alan Sorkin, Health Economics: An
Introduction (New York: Lexington Books, 1992), pp. 168,
182.
(8) Remarks by Congressman Fortney "Pete" Stark to a confer-
ence on "Prospects for Health Care Reform Under Clinton,"
Washington, D.C., January 14, 1993.
(9) Joseph Newhouse et al., "Some Interim Results from a
Controlled Trial of Cost Sharing in Health Insurance," New
England Journal of Medicine (December 17, 1981): 95-112.
See also Willard G. Manning et al., "Health Insurance and
the Demand for Medical Care: Evidence from a Randomized
Experiment," American Economic Review (June 1987): 251-73.
(10) Manning et al., p. 258.
(11) See, for example, Martin Feldstein, "Econometric Studies
of Health Economics," in Frontiers of Quantitative Econom-
ics, ed. D. Kendrick and M. Intrilligator (Amsterdam: North
Holland Press, 1974); Richard Eichhorn and Lu Ann Aday, The
Utilization of Health Services: Indices and Correlates: A
Research Bibliography, NTIS No. PB-211 720 (1972); Avedis
Donabedian, Benefits in Medical Care Programs (Cambridge:
Harvard University Press, 1974).
(12) Richard Frank, "Pricing and Location of Physician Ser-
vices in Mental Health," Economic Inquiry (1985): 115-33;
John Wallen, Paul Roddy, and Michael Fahs, "Cost Sharing,
Mental Health Benefits, and Physical Complaints in Retired
Miners and Their Families" (Washington: American Public
Health Association, 1982).
(13) Alan Liebowitz, Willard Manning, and Joseph Newhouse,
"The Demand for Prescription Drugs as a Function of Cost-
Sharing," Social Science and Medicine (1985): 1063-69.
(14) See, for example, David Kenkel, "Consumer Health Infor-
mation and the Demand for Medical Care," Review of Economics
and Statistics (1990): 587-95; Roger Feldman and Frank
Sloan, "Competition Among Physicians Revisited," Journal of
Health Politics, Policy, and Law (1988): 239-62.
(15) Michael Morrisey, Price Sensitivity in Health Care:
Implications for Health Care Policy (Washington: NFIB Foun-
dation, 1992).
(16) David Dranove, "The Five W's of Utilization Review," in
American Health Policy: Critical Issues for Reform, ed.
Robert Helms (Washington: American Enterprise Institute,
1993), pp. 239-55.
(17) Rita Ricardo-Campbell, The Economics and Politics of
Health Care (Chapel Hill: University of North Carolina
Press, 1991).
(18) Joseph Bast, Richard Rue, and Stuart Wesbury, Why We
Spend So Much on Health Care (Chicago: Heartland Institute,
1992).
(19) "Hold the Phone," Harvard Health Letter, November 25,
1992.
(20) For a complete discussion of private-sector experience
with MSAs, see Peter Ferrara, "More Than a Theory: Medical
Savings Accounts at Work," Cato Institute Policy Analysis
no. 220, March 14, 1995.
(21) Manning et al., p. 262.
(22) There were three exceptions: hypertension, myopia, and
dental care. However, researchers suggest that "programs
targeted specifically at these problems would be much more
cost-effective in achieving these gains." Ibid.
(23) "Answering the Critics of Medical Savings Accounts: Part
II," National Center for Policy Analysis Brief Analysis, no.
133, September 16, 1994.
(24) Alain Enthoven, letter to the editor, Wall Street Jour-
nal, July 22, 1994.
(25) "Answering the Critics of Medical Savings Accounts: Part
I," National Center for Policy Analysis Brief Analysis
no. 132, September 16, 1994.
(26) "Unnecessary" in this context does not mean the opera-
tion has no value, but that the value is insufficient to
justify the expense.
(27) Joseph Newhouse and Charles Phelps, "New Estimates of
Price and Income Elasticities," in The Role of Health
Insurance in Health Sector Services, ed. Robert Rosset (New
York, National Bureau of Economic Research, 1976).
(28) See, for example, Frank Sloan, "Effects of Health Insur-
ance on Physician Fees," Journal of Human Resources (1982):
331-57.
(29) Steffie Woolhandler and David Himmelstein, "The Deterio-
rating Administrative Efficiency of the U.S. Health Care
System," New England Journal of Medicine 324, no. 18 (May 2,
1991): 1253-58.
(30) Stan Liebowitz, "Why Health Care Costs Too Much," Cato
Institute Policy Analysis no. 211, June 23, 1994, pp. 20-21.
(31) John Burry Jr., Medical Savings Accounts: Bad Medicine
for the U.S. Healthcare System (Columbus: Blue Cross and
Blue Shield of Ohio, 1994). Burry makes essentially the
same argument in a second monograph, A Windfall for the
Healthy: How Medical Savings Accounts Will Hurt Americans
and Hurt Business (Columbus: Blue Cross and Blue Shield of
Ohio, 1994).
(32) Burry, Medical Savings Accounts, p. 4.
(33) Ibid., p. 7.
(34) Ibid., pp. 7-10.
(35) This assumption is contrary to the usual distribution of
medical expenses. Generally, only 10 percent of Americans
spend more than $3,000 per year on health care. See, for
example, Blue Cross and Blue Shield Association, Reforming
the Small Group Health Insurance Market, March 1991, p. 6.
An Ohio think tank, the Buckeye Center for Public Policy
Solutions, which examined Burry's numbers, concluded that
Burry's inflation of the number of Americans spending more
than $3,000 per year overstates the drain on the health care
system by at least $3.5 million (Bradley Smith and Samuel
Staley, "Medical Savings Accounts and 'Real World' Health
Care Economics in Ohio," Buckeye Center Policy Brief, June
29, 1994). However, for the sake of discussion, this paper
accepts Burry's projection.
(36) Burry, Medical Savings Accounts, pp. 9-10.
(37) Ibid., p. 12. To obtain this total, Burry divides his
$50.2 billion shortfall for Ohio by the 38,729 in his study
for an average per family shortfall of $12,967. He then
multiplies that by 6.45 million American high-use families,
representing 10 percent of the 64.5 million American fami-
lies.
(38) Burry subsequently claimed that Blue Cross and Blue
Shield of Ohio has overhead costs of only 10 percent (letter
to the editor, Cleveland Plain Dealer, September 17, 1994).
However, other industry professionals dispute this claim.
(39) Letter from the American Academy of Actuaries, Medical
Savings Account Working Group, to the United States Senate,
August 29, 1994.
(40) Ibid.
(41) Ibid.
(42) Richard Koenig, "Insurers' Overhead Dwarfs Medicare,"
Wall Street Journal, November 15, 1990.
(43) Letter from the American Academy of Actuaries, p. 4.
(44) "Don't Be Seduced by Medisave," New York Times, August
16, 1994, p. A26.
(45) Iris Lav, "Medical Savings Accounts Impede Universal
Coverage," Center on Budget and Policy Priorities, August
17, 1994.
(46) Congressional Budget Office, "The Tax Treatment of
Employment-Based Health Insurance," March 1994.
(47) See Stuart Butler, "A Policy Maker's Guide to the Health
Care Crisis: Part II: The Heritage Consumer Choice Health
Plan," Heritage Talking Points, March 5, 1992.
(48) C. Eugene Steurle, "The Search for Adaptable Health
Policy Through Finance-Based Reform," in American Health
Policy: Critical Issues for Reform, ed. Robert Helms (Wash-
ington:American Enterprise Institute, 1993), pp. 334-61.
(49) Lav.
(50) Congressional Budget Office, "An Analysis of Congressman
Michel's Health Proposal," August 29, 1994, pp. 5-7.
(51) Ibid.
(52) Lav, p. 4.
(53) John Goodman and Gerald Musgrave, "A Primer on Managed
Competition," NCPA Policy Report no. 183, April 19, 1994.
Published by the Cato Institute, Policy Analysis is a
regular series evaluating government policies and offer-
ing proposals for reform. Nothing in Policy Analysis
should be construed as necessarily reflecting the views
of the Cato Institute or as an attempt to aid or hinder the
passage of any bill before Congress.
Contact the Cato Institute for reprint permission.
Printed copies of Policy Analysis are $4.00 each
($2.00 each for five or more). To order, or for a complete
listing of available studies, write to: Cato Institute,
1000 Massachusetts Avenue NW, Washington, DC, 20001.
(202)842-0200 FAX (202)842-3490 E-mail cato@cato.org
World Wide Web http://www.cato.org/home.html.