Monday, June 15, 2015
June 15, 2015
Health care cost-sharing prompts consumers to make big cuts in medical
By Ben Handel
Is that surgery really worth it? Do I really value that cancer
screening? Is that extra imaging service necessary?
These are the kinds of questions consumers ask themselves when their
insurance plans require higher cost-sharing for medical services. This
is a new reality in the US health care system as large employers
offering coverage have moved aggressively toward less generous,
high-deductible insurance offerings.
This shift was accelerated by the "Cadillac Tax" provision contained in
the Affordable Care Act (ACA), which, starting in 2018, places an excise
tax on employers offering insurance plans that cover very high levels of
medical spending. Further, many of the consumers enrolling in the public
state exchanges created under the ACA have enrolled in lower- coverage
financial plans that cover an average of 60% (bronze) or 70% (silver) of
medical expenditures, similar to typical high-deductible coverage.
Though these policies are in part motivated by the government's need to
reduce its share of total health care spending, they are also driven by
the expectation that they will lead consumers to use higher-value,
lower-cost medical services.
In my recent research with Zarek Brot-Goldberg, Amitabh Chandra, and Jon
Kolstad, we dug into the mechanisms for how and why consumers reduce
medical spending when faced with higher cost-sharing.
To do this, we studied the medical claims and medical spending of more
than 150,000 employees and dependents from one large firm that moved
everyone from an insurance plan that provided completely free health
care to a high-deductible plan covering 78% of medical spending on average.
During the switch, the in-network providers that consumers could access
and the services covered remained the same. As a result, this switch
presented an excellent opportunity to assess in detail how consumers
respond to markedly increased cost-sharing.
Primarily, we wanted to know whether employees would reduce their
medical spending as a result of the change and, if so, by how much.
Further, we hoped to learn where specifically they'd cut back. Would
they spend less on nonessential services or reduce spending across the
board? Would they try to find cheaper sources of health care? Do some
employees cut more than others? Do employees correctly perceive the true
marginal price of care in a complex insurance contract?
Health care spending plunges
We first established that increased cost-sharing does reduce medical
spending at the firm. Age- and inflation-adjusted medical spending
dropped by 19% – from a base of approximately US$750 million in spending
– when employees switched to high-deductible coverage.
Strikingly, many of the spending reductions come from the sickest
employees. The sickest 25% (based on prior diagnoses each year) reduced
spending by one-quarter after shifting coverage. This is especially
notable, and somewhat surprising, since these employees earn relatively
high incomes and their maximum out-of-pocket payment in high-deductible
coverage for the calendar year was approximately $6,500 for a family.
How did consumers reduce spending? A detailed data analysis reveals that
medical prices did not go down. Further, we show that consumers did not
price-shop after the switch – that is, they did not move toward cheaper
providers, for example, when they were going to undergo a specific
It turns out that all spending reductions were directly linked to
quantity reductions: consumers just consumed less medical care.
Cuts across the board
Importantly, it didn't seem like consumers were particularly choosy
about what kind of health care they cut: consumers appeared to reduce
consumption across a range of medical services, from low to high value.
For example, quantity reductions led to a 22% drop in spending on
imaging services (such as MRIs or CT Scans), some of which are likely
unnecessary. However, consumers also reduced how many preventive health
services they used – which policymakers typically believe are
underutilized – by 16%.
The cuts were across the board. Spending on mental health care fell 8%,
inpatient and outpatient hospital services declined 14% and 17%,
respectively, drug purchases dropped 20%, and emergency room services
plunged 27%. Of the top 30 medical procedures (by revenue) that we
investigated, we found that consumers reduced spending for 23 of them.
Simply put, consumers did not look for cheaper services but consumed
less medical care, and did so across almost the entire range of medical
Insurance price misperceptions
One possible reason for why these sick, relatively high-income consumers
reduced potentially valuable medical spending is that they perceived the
marginal price of their medical care to be higher than it actually is.
Take the example of a consumer who knows he/she is quite sick entering
the year and expects to spend a lot on health care. That consumer should
not worry about the deductible and cost-sharing when making medical
decisions early in the year because he/she knows that by the following
January, all medical care used after passing the plan's out-of-pocket
limit will be free. Thus, the true marginal price of health care for
this predictably sick consumer is close to $0, no matter how high the
deductible is, so care consumed early in the year is essentially free as
But we found that these consumers substantially reduced spending early
in the year when under the deductible, but once they passed it spent
more. In other words, many consumers whose true marginal price for care
throughout the year is essentially zero because of their impending high
spending don't treat incremental care as free when under the deductible.
Instead, they respond as if the price of care under the deductible is
the relevant price, despite the fact that they will spend that money
during the year regardless.
This suggests that they misperceive their own health risks, misperceive
how much medical care costs or don't understand how the high-deductible
insurance contract actually works. Similar consumer price misperceptions
are also documented in Medicare Part D, electricity markets and
What this means for reform efforts
Giving consumers direct incentives to think about their health care
spending is a cornerstone of health reform in the US and plays a large
role in several national health systems around the world, such as in France.
An important prerequisite for these reforms to be successful is that
consumers, who may or may not be making medical decisions in conjunction
with physicians, understand the costs and benefits of different health
care services. Our evidence suggests that consumers don't seem to be
responding to increased cost-sharing with nuanced expertise and instead
reduce consumption across the range of medical services, some valuable
and some likely wasteful.
Additionally, they reduce care heavily when sick and under the
deductible, even when their true marginal price of care is very low.
Thus, while increased consumer cost-sharing can be an effective
instrument for reducing health care spending, it may be a blunt
instrument for encouraging higher value medical spending, especially
relative to supply-side interventions that target physician incentives
or interventions that reduce the use of high-cost low-value medical
As health reform pushes forward, policymakers will need to recognize the
limits of consumer cost-sharing policies and focus more on how to
appropriately incentivize providers to deliver high-value, low-cost care.
"What Does a Deductible Do? The Impact of Cost-Sharing on Health Care
Prices, Quantities, and Spending Dynamics" by Zarek Brot-Goldberg,
Amitabh Chandra, Benjamin Handel and Jonathan Kolstad (43 slides):
Comment by Don McCanne
A cursory glance at this article suggests that it is simply one more
study that confirms that high deductible plans decrease health care
spending, and since you already know that you might be tempted to pass
on reading today's message. But don't skip this one if you wish to
better understand just what impact high deductibles really do have.
What is unique about this study is that it evaluated the patterns of the
change in health care spending when a large firm switched about 85
percent of its employees from a PPO plan that provided first dollar
coverage (no deductibles, no coinsurance, and $0 out-of-pocket maximum)
to a HDHP (high deductible health plan with $3,750 deductible, 10%
coinsurance, and $6,250 out-of-pocket maximum). The health care
providers were the same both before and after the change was made. This
is about as pure of a study as you could devise on this topic - the same
employees, the same health care providers, but with a change to a high
deductible with coinsurance and a new patient responsibility for up to
$6,250 in cost-sharing.
As expected, spending abruptly declined - by about 19%. So was this a
result of better price shopping, as the advocates of these
consumer-directed HDHPs tout? No. Medical prices did not go down after
the switch was made. These health care consumers did not shop prices.
What went down was the quantity of health care provided. In fact, the
sickest employees reduced their use of health care services even more -
by about 25%. The reductions in utilization were across the board -
inpatient services, outpatient services, emergency room services, mental
health care, drug purchases, imaging, and preventive health services.
Most of these are beneficial services.
Another interesting finding is that those individuals with significant
disorders who knew that they would reach their maximum out-of-pocket
spending nevertheless reduced their utilization of health care services
while they were still under the deductible. They did not need to reduce
their use of these services since after the out-of-pocket maximum is
reached, their marginal cost of additional health care is essentially
zero. Their net costs are the same regardless of their utilization. It
is likely that these sick individuals were needlessly forgoing
beneficial health care services.
The author states, "consumers appeared to reduce consumption across a
range of medical services, from low to high value." Clearly policies
that reduce the consumption of high value care are undesirable, and, for
this reason alone, deductibles and coinsurance should be eliminated. But
what about low value care? What is low value care? Is that the MRI that,
in retrospect, turned out to be normal? Wasn't there some benefit in
excluding potential pathology? Attempting to ferret out low value care
can be detrimental if it consequentially results in the blunt
elimination of high value care as well.
Besides, how much spending reduction would we really see with the
reduction in beneficial health care services that results from
deductibles? Remember that the 20 percent of individuals with greater
health care needs consume 80 percent of our health care services. Most
of this spending is well above the maximum out-of-pocket costs and thus
cost-sharing has very little impact on this spending. The deductibles
might influence utilization for the other 80 percent of us, but that
would reduce spending by only a fraction of the 20 percent of health
care that we use. Anyway, is the amount of health care used by us
low-utilizers really an egregiously excessive amount of care? We usually
have a legitimate reason for going to the doctor.
In this article, Ben Handel states, "while increased consumer
cost-sharing can be an effective instrument for reducing health care
spending, it may be a blunt instrument for encouraging higher value
medical spending, especially relative to supply-side interventions that
target physician incentives or interventions that reduce the use of
high-cost low-value medical technologies."
Instead of using detrimental demand-side patient cost-sharing
instruments to reduce spending, just think of what could be accomplished
on the supply-side using a well designed single-payer monopsony for
financing health care: global budgeting of institutions such as
hospitals, dramatic reduction of administrative waste, negotiation of
rates for services and products, bulk purchasing of pharmaceuticals,
avoiding excess capacity through planning and separate budgeting of
capital improvements, and establishing a global budget for the entire
health care delivery system.
There is no need to assess financial penalties (deductibles and
coinsurance) merely for accessing beneficial health care services. With
a single payer system, patients simply obtain the health care that they
need, when they need it. That's the way it should be.
at 2:36 PM