« Political Science Blog Bleg | Main | Economists In The Blogosphere »

June 25, 2007

Counterintuitive Health Chart of the Day

The other night, I had a long argument with some health care types* that hinged, partly, on whether individuals have seen sharp shifts towards out-of-pocket spending in the past few years.  Whether, in other words, the Great Risk Shift was heaping out-of-pocket costs onto individuals in some new way.  My recollection of the data was that it wasn't, that out-of-pocket spending was actually going down, and that ever more money was going into insulation and premiums.  This was not a popular viewpoint.  So I was glad to come across this CBO graph detailing the story I was trying, and failing, to tell:

Out Of Pocket Spending

That's a sharp drop, particularly considering how much health costs have risen over the same period.  What you're seeing, to be sure, isn't less total spending on the part of consumers, but more going into premiums, and being taken out of wages.  And this has some negative effects.  As the CBO says, "Consumers facing lower out-of pocket costs tend to demand more health care services than consumers facing higher out-of pocket costs. At the same time, rising healthcare costs (as a share of income) have probably led individuals to seek more extensive insurance in order to keep the variability of their out-of-pocket expenses from increasing." 

So it's a bit of a perverse cycle.  Costs go up, unsettling consumers, who buy more protection, which encourages somewhat more use of health care, which then brings costs up...rinse and repeat.  This critique is often used by libertarians to argue that we shouldn't have any insulation at all.  That's foolish.  But I could see a world in which first-dollar cost sharing is increased in a
targeted way to both bring down costs and increase quality.  Indeed, I've even written about such a world...

*Yes, I'm exciting.

June 25, 2007 in Charts, Health Care | Permalink

Comments

Wait ... this chart measures percentage of health expenses paid out of pocket. But if you're trying to measure the consumer pinch, shouldn't you measure out-of-pocket expenses as a percent of discretionary income?

... Okay, found the right table. Exhibit 5 in this Kaiser study lists historical health spending as a share of GDP. Taking the midpoint of each bit, we see that out-of-pocket health expenses were 2.6% of GDP in '75, 2.7% in '85, just under 2.3% in '95, and around 2.6% in 2005.

Okay I confess I have no idea what to think of that bit of information. Except to say that no one would be complaining about health care costs if the income gains since the mid-'70s had been more broadly distributed.

Posted by: Nicholas Beaudrot | Jun 25, 2007 11:21:35 AM

But the issue here isn't5 absolute spending, which has gone up across the board, but the proportions borne by different payers. In other words, I'm interested in whether businesses have actually forced a shift towards more vulnerability for individuals -- and I'm just not seeing the evidence for it. Yet.

Posted by: Ezra | Jun 25, 2007 11:29:35 AM

Ezra,

You are right on point here-- addressing the growth in costs is the key issue for health care reform, at the end of the day (not that others aren't important such as UHC, etc. but there are reasonable proposals on the table for those).

However, its also the primary reason why a single-payer system is the wrong approach. As I've said in other threads, my concern is the lack of competition on methods of cost-control that would exist in a single-payer system. There are a lot of different ideas on how best to control costs (your thoughts on targeted first-dollar cost sharing is one of many)-- the reality is no one knows which is most effective, fair and/or equitable. We need a system that allows for experimentation of cost control methods, consumer choice (i.e. some may be willing to pay more upfront for less restrictions and vice-versa) and different options on levels of covered procedures (e.g. on the spectrum of standard-of-care versus experimental). Preserving competition over these elements is fundamental to the long-term success of our health care system in preventing costs from reaching the current astronomical and unsustainable projections, and this competition would not exist in a single-payer system.

Posted by: wisewon | Jun 25, 2007 11:31:18 AM

It is also relevant, though, that out-of-pocket health care costs as a share of family budgets has been rising significantly. Here's a useful Commonwealth Fund summary on that subject:

http://www.cmwf.org/publications/publications_show.htm?doc_id=347500

Posted by: anrig | Jun 25, 2007 11:41:21 AM

dude, ezra, chicks really dig the health care wonkery. some chicks.

Posted by: belle waring | Jun 25, 2007 11:56:51 AM

I think you're using the wrong perspective on this graph, and that the graph may be hiding significant details. There's a huge drop from 1975 to 1995, but since 1995 the numbers have stayed fairly close (as has the percent-of-gdp number). So most people don't see the drop in percentage of costs, they only see the absolute real increases -- and, as Beaudrot points out, against a background of flat or falling real wages.

We also don't know what the detailed picture is, e.g. what groups are getting hit more or less hard, and how that affects the perception of risk. (Anecdotally speaking, we wouldn't have noticed our insuror's change in prescription co-pay policy from $10 per prescription to 50% of cost if it hadn't been for those lifesaving medicines that happened to cost $750 a vial.) You can have the same aggregate amount going out, but people will rightly perceive their risk -- in the common sense -- as increasing if the variance rises.

Posted by: paul | Jun 25, 2007 12:03:01 PM

With health costs growing 5 times as fast as per capita income, 80 years out we will be spending 300% of (today's) per capita income on medicine. Luckily, we can expect 400% expansion of per capita income by then.

All to the good: old people may be sprouting propellers while the young leap tall building with a single bound. In any case we will need something to employee people at -- by then, we should have the same percentage of people in manufacturing as we do today in farming.

Posted by: Denis Drew | Jun 25, 2007 12:18:18 PM

i'm a little surprised you didn't go where you've gone before - the preferential tax treatment of premiums vs OOP spending push more and more health spending into the former away from the latter.

(OOP + employee share of premiums) as a share of total health spending is actually pretty flat since the early 1980s.

the one study i've seen that tries to take a cut at OOP spending by income group shows that distribution mattered a lot in the 1990s: high income groups saw (absolute) OOP declines of something like 40% while low and middle-income groups saw declines of well under 10%.

Posted by: josh bivens | Jun 25, 2007 12:24:20 PM

"I'm interested in whether businesses have actually forced a shift towards more vulnerability for individuals"

Shouldn't the absolute inflation-adjusted amount be taken into account also? If that's the actual risk the individual is bearing, it shouldn't be expressed strictly as a percentage of the overall whole. If total costs double, and out of pocket costs go up fifty percent, risk has increased, no?

Part and parcel of the "Great Risk Shift" argument, of course, would also be that fewer people can rely on their employers for their medical coverage, that might also bear investigating.

Posted by: Anthony Damiani | Jun 25, 2007 12:38:51 PM

What about the people who simply stop purchasing services at all due to health insurance being unaffordable or unavailable? Would that not also drive down the overall out of pocket costs?

Posted by: fiat lux | Jun 25, 2007 1:19:13 PM

For single payer advocates:

I saw a reference recently in Health Affairs to a study sponsored by the AARP a couple of years ago which found that roughly 50% of healthcare costs in the U.S. are attributable to the 65 and older population. This includes payments by Medicare, Medicaid (mainly long term care costs), Medicare Supplemental insurers, long term care insurers, beneficiary Medicare premiums, and out-of-pocket payments. This sounds about right as the elderly account for roughly 15%-16% of the population and consume approximately three times more healthcare per capita than the rest of the population. These figures imply that current healthcare spending by and on behalf of the elderly works out to between $20,000 and $25,000 per person. Since Medicare is already, in effect, a single payer system (complete with dictated prices and supposedly low administrative costs) for this population, the question is: how does this level of per capita spending on our 65 and older population compare to the per capita spending on the elderly (65 and over) in other countries? If it turns out that our per capita spending on the elderly is 50%-100% higher than other countries' per capita spending on their elderly populations, it implies that Medicare for All is NOT the answer for the U.S. While there are undoubtedly administrative inefficiencies in our current fragmented system, the more important cost drivers probably relate to such factors as differences among countries in the approach to end of life care, level of provider compensation, fear of litigation driven defensive medicine, lack of interoperable electronic medical records, higher prices in the U.S. for brand name drugs, and inadequate to non-existent price and quality transparency.

While I've seen plenty of comparisons among countries related to overall healthcare spending as a percentage of GDP and per capita spending for the total population, I have never seen any comparisons that focus on the 65 and older population where a single payer system has been in place in the U.S. since 1965!.

Posted by: BC | Jun 25, 2007 1:20:52 PM

Hmm, is it just me that finds this graph as evidence against single payer? More insulation equals higher and higher costs.

Posted by: Jason | Jun 25, 2007 2:19:47 PM

Of course the whole argument hinges on the slender thread that moral hazard is a significant factor in health care spending. With the exception of the rare cases of Munchausen syndrome, I think more people tend to avoid doctors and hospitals if they can. The far greater hazard is that people tend to let simple conditions go untreated until the complications are too serious to ignore and require serious and expensive medical intervention. Is there any good research out there that shows people with top of the line insurance policies become medical wastrels?

Posted by: justawriter | Jun 25, 2007 2:29:11 PM

Writer,

Your focus was pre-diagnosis, I'd agree that few people choose to go the doctor just because they have coverage. Its the post-diagnosis consumption of health care where moral hazard comes into play, i.e.

-- do you take the generic old drug or the new branded drug? (There's only a $5 difference in co-pay even though the real difference may be $100-200/month with no real clinical difference)

-- the patient with back pain-- do you get the MRI now to make sure nothing serious is going on or wait a month, according to standard-of-care?

-- you've been recovering in the hospital for a week ready to be discharged in the late afternoon, your ride can't pick you up at the hospital until tomorrow morning, do you stay another day? (This one happens more often that you can imagine)

happy to provide more examples, if interested... and yes-- the studies are there as well (for example, its the flipside of the HSA studies, which suggest patients forgo treatment when they are exposed to a greater proportion of the cost)

Posted by: wisewon | Jun 25, 2007 3:00:30 PM

Of course the whole argument hinges on the slender thread that moral hazard is a significant factor in health care spending. With the exception of the rare cases of Munchausen syndrome, I think more people tend to avoid doctors and hospitals if they can. The far greater hazard is that people tend to let simple conditions go untreated until the complications are too serious to ignore and require serious and expensive medical intervention. Is there any good research out there that shows people with top of the line insurance policies become medical wastrels?

wisewon makes some good points. I'd like to add that spending a lot on health care isn't necessarily a bad thing. Who wouldn't want to spend money on something that can extend your life or improve your quality-of-life, especially in a country as rich as ours? There is a growing gray area between health problems and health improvements. If I break my leg, clearly that is a health problem that needs to be fixed. But what about, say, knee replacement surgery.

Insurance companies offering blanket coverage either much charge higher premiums to cover these newer procedures, or not cover them. Likewise, a single payer system will face the same problem: charge higher taxes, or ration. The solution is less insulation, not more. Sure, you can have that knee replacement, but you are going to have to pay for part of it. An individual consumer can then decide what is in their financial best interest.

Posted by: Jason | Jun 25, 2007 4:38:54 PM

Uhhh ... another thing to look at is the distribution of medical expenses. I think the evidence that medical bankruptcy is on the rise is more than just anecdotal. And that's the Hacker-esque risk shift. Instead of everyone paying 13% of their expenses out of pocket, a whole lot of people are paying 5% ... and a growing number are paying 60%

Posted by: Nicholas Beaudrot | Jun 25, 2007 4:48:50 PM

I'm not sure why this is so hard to get. If you want to keep prices under control for healthcare, you basically have two options:
1. Minimize insurance so that people pay for almost all of their care out of pocket, with the consequence that a minority (maybe 5% of the young and 25% of the old) either go without care or lose all their money because they get very sick and can't pay their bills.
2. enact government price controls that are tied to well-marked budgets that the voting public understands are tied to taxes. Just like with any tax, people aren't eager to enlarge it, and as a result healthcare budgets (and thus fees) are constrained. This works EVERYWHERE. It would work in America. Are Americans more tax-happy than Europeans or Japanese?

When push comes to shove, those who oppose universal healthcare are forced to abandon their pseudo-economic arguments and resort to an ideological argument. Some version of: it just isn't right to force some people to pay for other people's healthcare.

Posted by: jd | Jun 26, 2007 12:06:14 AM

By the way, the greatest irony of that table showing declining out-of-pocket expenses is that the decline corresponds exactly to the growth of HMOs. The big, bad HMOs were also the principle agent in the last 30 years for reducing out of pocket expenditures. People replaced their Blue Cross "major medical" plans that had high deductibles with HMO plans that had no deductibles.

When the backlash happened starting around 1997-98, HMOs lost the ability to manage care and providers regained leverage to engage in price increases and unnecessary care, which was the main reason for the boom in health care costs starting in 1999.

Rather than fight it, HMOs decided to ride the wave and charge a little more on top. They got killed for controlling costs for a few years and their reputation has never recovered. In fact, they are now blamed for high costs when they are the only market force that has made a real and effective effort to reduce the cost increases, even if they only did it for a few years on a national level. How's that for standing history on its head?

Posted by: jd | Jun 26, 2007 12:19:11 AM

25 years ago when I was newly minted engineer, the only health insurance available was an 80/20 plan where the insurer paid 80% of the cost and the insured paid 20% until reaching a deductible, then the insurer took over at 100%. With the HMO and PPO plans that are more common now, the insured pays some token amount per visit. Under this new plan for someone with routine medical needs, the yearly out of pocket is far less than their Starbuck's bill.

In my experience there is quite a bit of discretionary medical care. Extra screening is very popular among my HMO/PPO patients. They frequently request extra labs, x-rays, MRIs, plastic surgery, full body "screening CTs", gastric bypass, and even extra Pap smears! My cash paying patients, eh, not so much.

Posted by: J Bean | Jun 26, 2007 12:50:43 AM

We need a system that allows for experimentation of cost control methods, consumer choice (i.e. some may be willing to pay more upfront for less restrictions and vice-versa) and different options on levels of covered procedures (e.g. on the spectrum of standard-of-care versus experimental).

It's entirely possible to do that within a single-payer-plus system along the lines of the French model, albeit one with highly devolved administration. You create structures for innovation on a B2B level, or on a B2C level with supplementary insurance.

As I said in my previous reply, there are going to be experiments that don't work, and you either commit to the equivalent of a bail-out or you don't. The NC mental healthcare system had an experiment with local-management-entity provision, and it's gone badly wrong. The result has been that fewer severely mentally ill people in certain regions -- people who are a risk to themselves and others, and aren't really the private insurers' dream customer -- have access to care.

Do you make patient-consumers foot the bill for cost-control experiments that go tits-up? Or, contrariwise, how do you reduce the impulse among primary providers to send people for MRIs because those machines aren't going to pay their $2m bills for themselves?

Posted by: pseudonymous in nc | Jun 26, 2007 12:55:28 AM

2. enact government price controls that are tied to well-marked budgets that the voting public understands are tied to taxes. Just like with any tax, people aren't eager to enlarge it, and as a result healthcare budgets (and thus fees) are constrained. This works EVERYWHERE. It would work in America. Are Americans more tax-happy than Europeans or Japanese?

Horrible idea. Tragedy of the commons: it is in an individual's best interest to consume as much as possible; therefore everyone starts consuming as much as possible.

The real way to control costs is to stop insulating consumers from the cost of health care through a third payer system.

Posted by: Jason | Jun 26, 2007 1:24:50 AM

jd,

Your second post was dead-on, I'm completely share the same view on HMOs in the 90's. Your first post, however, missed the mark. You've laid out the two extreme examples, but there is a gray area. Ezra gave one example in his post-- targeted first-dollar cost sharing-- but there are others as well, i.e. expose consumers to sufficient cost-sharing that there is some dollar impact for consuming health-care, but done in a way that there is still sufficient insurance protection.

Posted by: wisewon | Jun 26, 2007 6:39:40 AM

NC,

Yes and no. The french model as you suggest, could work well for offering consumer choice and levels of coverage. However, it wouldn't work so great for cost-control, unless the basic coverage provided by government was extremely minimal. In other words-- I would assume cholesterol lowering medications would fall under the government provided basic plan-- the question is-- how do you cost share with patients to ensure patients have some "skin in the game" to choose the very suitable generic drug versus the latest and greatest? But also look at the flip side (people don't talk about this much) how do you incent patients to take any cholesterol lowering medication at all? For both cost-sharing and incenting patients-- there are a number of ideas here that will be lost by having basic insurance providing by government. Multiply the cholesterol medication example by a few hundred examples, and a fair amount of cost control innovation (and incentives on preventive medicine) is lost.

On your MRI link-- The reform is clear-- doctors should not be able to self-refer to their own MRI centers (so-called Stark laws), but they have found loopholes to continue the practice. In a nutshell, however, these are the stories that single-payer advocates like to avoid, because they demonstrate other weaknesses of that approach so clearly. While its clear to all involved what the solution would be (as just described) lobbying gets in the way of what's best for patients.

Posted by: wisewon | Jun 26, 2007 6:52:02 AM

As I said in my previous reply, there are going to be experiments that don't work, and you either commit to the equivalent of a bail-out or you don't.

On other thought-- I"m not sure what you have in mind for experiments that don't work, but I don't think they require a bail-out. I think two things can happen:

1. Cost controls are too draconian/complicated/consumer un-friendly
2. They are too loose, for whatever reason, and cost rise high than they should be

If its either 1 or 2-- people simply switch plans the following year. Neither needs a bailout. 1 is a plan that is simply painful for its users, people have those types of plans today and just switch. 2 will lead to out-of-control spending-- which in real terms means a 10-20% increase in premiums from one year to the next. Again, it would lead to some pain for users, but this type of thing happens today as well-- its painful for a year and patients move on.

Posted by: wisewon | Jun 26, 2007 7:07:13 AM

However, it wouldn't work so great for cost-control, unless the basic coverage provided by government was extremely minimal.

I think we might be talking at cross-purposes about 'cost-control' here. I look at a healthcare system that's remarkable for its un-joined-upness, if you pardon the ugly term. As Ezra has pointed out (in the linked article), end users are just not very good at discriminating between savings that are a long-term good in terms of both health and cost. In that regard, I think that cost controls are generally better implemented through smart contracting within the system itself: provider-insurer, and provider-provider.

The generic/brand issue has to be widened to look at DTC advertising and pharma marketing to doctors, but that's the kind of thing that would be at least partially addressed by economies of scale in collective bargaining. Admittedly, Big Pharma is going to choke on this, and there's the issue that tomorrow's generic is today's patented drug, but it's time for them to start selling their drugs on the merits, not the prime-time ad campaign.

On bailouts: I focused on mental healthcare because it's the area I know best, but it's also the area most resistant to market modelling. 'Painful for a year' is frankly not good enough if that means someone with severe mental illness is left untreated or unhospitalised.

On MRIs: here's one situation where I say 'rationing? no big deal.' It's fairly clear that imaging services have invested in plant as a nice little earner, with the intention of getting referrals by hook or by crook, and that GPs have developed a habit of sending patients off to the magic machine without a second thought.

Posted by: pseudonymous in nc | Jun 26, 2007 9:07:33 AM

The comments to this entry are closed.