« McCain on Iraq | Main | How Sanctions Helped Hamas »

April 12, 2007

The Progressive Case For More Cost Sharing

Here's the piece on Furman's plan I've been promising. It also reflects my thinking that there's more for liberals in HSAs than first meets the eye, but they have to be constructed progressively, smartly, and within the context of a universal structure. The American Prospect doesn't yet allow comments, so put discussion in this thread. I'll be curious to hear what you guys think.

April 12, 2007 in Consumer-Directed Health Care | Permalink


Does anyone have cost figures on the Furman plan, in particular the most progressive version (wherein the group paying for the 50% of coverage up to 7.5% of income, as well as anything above that, is the government)? If it's not substantially different from the current federal costs for Medicare/Medicaid/S-CHIP, I can see this being really attractive politically.
My only policy comment would be that, even though Ezra declared neoliberalism dead, this seems like the quintessential neoliberal plan for health care, in that it achieves a liberal end (UHC) through traditionally conservative means (HSAs). See, Ezra, neoliberals aren't so bad.

Posted by: Minipundit | Apr 12, 2007 4:34:14 PM

The cost figures are all in the paper. Furman estimates that his plan will cost 30 percent less in total. I think that's overly ambitious, but it's certainly less expensive.

Posted by: Ezra | Apr 12, 2007 4:51:55 PM

Thanks, Ezra. That definitely sounds good.

Posted by: Minipundit | Apr 12, 2007 5:04:17 PM

Between 1990 and 2003, the U.S. ranked 8th out of 18 OECD countries in average annual growth rates of total health expenditures per capita--pretty much in the middle. Countries with higher growth rates included ones with both single-payer and multi-payer health care systems. France (whose health care system U.S. proponents of major reform often cite admiringly) had a spending growth rate almost as high as that of the U.S. (3.2% for France vs. 3.6% for the U.S.) [source]

On the basis of this and other evidence I remain deeply skeptical that a single-payer system or any other kind of major reform of U.S. health care would yield a significant reduction in the growth rate of health care spending.

Posted by: Jason | Apr 12, 2007 5:59:37 PM

Ezra, Furman's numbers simply don't add up. We need to remember that insurance premiums also come out of our pockets, something that you do admit but IMO don't explore very much.

The average American pays $840/year out-of-pocket. Ok, that means a family of four paid $3360. Assuming they have employer based insurance, that family is paying around $3000/year in premiums - never mind the $8000 the employer is plopping down for them.

So in terms of money that either comes out of their bank accounts or never gets there in the first place, it adds up to $6360. If we look at that as a percentage of this average American family's average income - single earner, ~$60,000 - it's 11%. If we take their income up to $80,000 - 200% of the federal poverty threshold for a family of four - that's still 7.9% of their income.

What Furman wants is to have out-of-pocket costs - excluding premiums - equal 7.5% of our income. Well, I can't afford that. So what unnecessary care would Furman like my family to give up? My daughter's asthma medicine? It would sure be a shame to see her wheeze and cough after running for a few minutes, but I suppose it's time to stop the frivolity. Maybe my wife should have had her medically necessary C-section at home so that we could have saved on costs. Or when they heard the murmer in my son's heart I should have just told them that since 96% of those things just heal up on their own we didn't need to worry with any consults or further testing. Oh wait, I didn't fucking know that at the time!

If Americans are consuming "too much" healthcare, the problem is not with the consumers. As much as we might want to take every pill advertised on TV, doctors are the ones we pay to know about them and either prescribe them or not. That we can doctor shop until someone finally writes the script is a problem with our doctors.

Unnecessary tests and procedures? Again, this isn't the consumers' fault. If my chest starts hurting, I'm not going to the all-night angioplasty clinic, I'm going to the ER, and they're going to decide what I need. And if the ER doc says I need angioplasty, then I'm going to consent no matter what I might have read on this blog about the procedure's effectiveness, because when my chest hurts and I'm scared my world narrows a bit.

A system that takes away redundancies in levels of bureaucracy, regulates malpractice insurance companies more closely and that includes larger numbers of people in the groups of insured is what is going to help reduce costs, not some half-baked punitive system that ignores, first of all, how much money an average American family makes in a year ($97,000 my ass) and second, just who it is that really controls where the money in our system is spent.

Posted by: Stephen | Apr 12, 2007 6:24:40 PM

Posted by: Jason | Apr 12, 2007 2:59:37 PM

Between 1990 and 2003, the U.S. ranked 8th out of 18 OECD countries in average annual growth rates of total health expenditures per capita--pretty much in the middle. Countries with higher growth rates included ones with both single-payer and multi-payer health care systems. France (whose health care system U.S. proponents of major reform often cite admiringly) had a spending growth rate almost as high as that of the U.S. (3.2% for France vs. 3.6% for the U.S.)

And yet the French spend less on health care to gain more care. Odd that ... perhaps the level of spending per capita and the rate of health care price inflation are also factors to take into consideration.

Given that France has lower health care price inflation than the US, 3.2% increase in spending delivers more increase in health care service than a 3.6% increase in spending in the US.

OIOW, more le bange for le bucke.

Posted by: BruceMcF | Apr 12, 2007 6:25:36 PM

No question that UHI advocates must confront cost containment head on, and that even the most efficient systems will have to ration. There are a lot of good arguments against cost-sharing in the article, so I am curious why you embrace this instead of a regulatory approach to determining the limits of national health expenditures. Patients dont really drive the health care purchasing decision; the front end preventative care (which cost-sharing would reach) is more efficient and effective than expensive the late end care (which, under the proposals you cite, cost-sharing would not reach); to the extent that patients do undertake cost-conscious decision-making, they are not well informed enough to choose correctly because, hey, they didnt go to medical school; it disproportionately impacts the poor and undereducated. Of course, anyone who has fought with an insurance company knows the fear and frustration that can be brought about by a regulatory approach; my hope is that it would be more rationally constructed in a national health system that took away the profit motive for the payer to deny health care.
In any case, the nuanced approach to the cost-sharing issue -- recognizing that cost-containment is compatible with universal insurance, that sliding scale cost-sharing is possible, and that cost-sharing need not apply to every service, is a valuable addition to the dialogue. Thanks

Wow. I have to admire the courage to link to something that is at such dramatic tension with your own argument. No one will ever accuse you of censoring your sources. They might accuse you of drawing the most unbelievably tangential inferences from a field of data that is arrayed so clearly against you.

From your source
"Health spending is rising faster than incomes in most developed countries, which raises questions about how these countries will pay for future health care needs. The issue may be particularly acute in the United States, which not only spends much more per capita on health care than any other country, but which also has had one of the fastest growth rates in health spending among developed countries"

And it continues:
"It is reasonably well known that for some time the United States has spent more per capita on health care than other countries. What may be less well known is that the United States has had one of the highest growth rates in per capita health care spending since 1980 among higher income countries."

With charts!:
"Exhibit 1 shows per capita health expenditures for 2003 in U.S. dollars purchasing power parity. Health spending per capita in the United States is much higher than in other countries – at least 24% higher than in the next highest spending countries, and over 90% higher than in many other countries that we would consider global competitors. Exhibit 2 shows that per capita health expenditures in the United States also were considerably higher than in the other analyzed countries in 1990. "

Fear not, though, says Jason, our unbelievably bloated health care system is growing at roughly the same rate as the single payers. So the extent to which it gets larger every year than the single payer systems is limited. Phew. And I thought there was a problem.

Now, it is certainly true that single payer systems grow. They will grow forever unless we implement some stratgey of cost-containment. That is a function of new technology, and, in the industrialized countries that fought WWII, aging baby boomers. But if our system has larger per capita health expenditures, and it is growing at the same rate, doesn't it follow that something is growing in our system that isnt growing in the single payer systems?

Or to put it another way, the fact that single payer systems grow from X to X+1 each year doesnt mean that this system, which is already at 4X, would grow from its current base if it switched to single payer tomorrow. Isnt it possible (likely) that single payer (or, likely, most UHI plans) would restructure the system to eliminate waste and profit, shedding per capita expenditure, and then grow from that level?

If I weigh 300 pounds, and you way 200 pounds, and we are both gaining weight at the same rate eating different diets, wouldnt you recommend that I switch to your diet? Isn't single payer, in any case, liposuction?

Posted by: RW | Apr 12, 2007 6:30:29 PM

And yet the French spend less on health care to gain more care.

This vague and unsubstantiated claim, again. I think you need to describe exactly what you mean by "more care" and show how you have determined that the French get more of it than Americans do.

But in any case, I wasn't addressing that issue, but rather the claim that Furman's proposal, or some other kind of major reform, would be effective at containing the growth in health spending. The OECD data cast serious doubt on that claim.

Posted by: Jason | Apr 12, 2007 6:34:36 PM

Let's see how this would work today. Let's assume a family of 3 spends money over the year for 6 to 12 providers (noting that one action can spur multiple billings, often from different organizations/providers: get an x-ray and pay for the x-ray with facilities, and also pay the MD to interpret the x-ray, and then pay a different MD to determine what to do about the diagnosis, and then further bills to check on progress.)

Each of these is paid by VISA card or check, in advance (without insurance, no one gets care). No cash, no treatment.

Now who or what determines they have reached thier 7.5% limit? Who is keeping the books? And how do the 'exempt' services get paid by a third party? And who is this third party?


In order to do that, you need an integrated system capable of setting system-wide priorities. As Furman, in fact, says, in a surprising admission for an economist associated with the centrist Hamilton Project, "The simplest and cleanest way to implement income-related cost sharing would be as part of a far-reaching fundamental health reform.

So, this whole Furman HSA thing is just a wheel on the carriage of some integrated system. Which we don't have and the HSA companies are bitterly against, along with the drug companies, part of the AMA, and the traditional free-for-service and HMO-type insurance companies. Will they be won over because of the cost-sharing features into supporting a single integrated system.

I'm afraid that my view is short-sighted about this Furman/Klein cost sharing because I don't see how this advances us much down the court toward either universal coverage or a single-payer system.

If the argument is that with all these 'promised' savings we can afford a single payer system that covers everything, and therefore all parties to the issue will immediately support it, then I demur, because there is no evidence that is the case.

If the argument is that with all these 'promised' savings the government or insurance companies will offer a HSA-type system that will motivate millions of Americans to give up their current employer-based health insurance and the uninsured to rush to their insurance agent for a HSA, I call bullshit.

Cost savings of this type are an administrative detail, not a plan for national reform of our health access system. It bears comparison to a drug formulary or a list of covered or not-covered diseases and procedures.

Maybe I just don't understand, but I think I do. It lacks a structure (a health coverage plan) to be imbedded into. It doesn't begin to resolve how we overcome objections to ANY national health plan, of whatever structure.

And finally, exactly how are these savings achieved? Should I haggle and shop for an othopedist when my leg is broken and I'm shock? What 'buying pool' is negotiating for me?

Sorry if this comment sounds snarky or stupid, but I just don't see how this addresses the overarching health care need.

Posted by: JimPortlandOR | Apr 12, 2007 6:37:04 PM

The growth rates for American care in the 90s were anomalistic because of the savings offered by managed care. Those, however, turned out to be a one-time, and very temproary, thing. As for the fact that the French pay less per unit of care, there's no debate on the point. Tis just that way.

Posted by: Ezra | Apr 12, 2007 6:41:44 PM


It's true that the U.S. had one of the highest growth rates in health spending between 1980 and 2003 (second only to Luxembourg), but over the more recent period of 1990 to 2003, its growth rate is firmly in the middle, as I said. The spending growth rate was higher in seven of the eighteen countries shown in the chart: Britain, Norway, Luxembourg, Ireland, Belgium, Austria and Australia, and almost as high (within a few tenths of a percentage point) in another four countries: Netherlands, Japan, Iceland and France.

Given this data, and the wide variety of types of health care system used by these nations (from single-payer systems like Britain's to complex multi-payer systems like France's), it is not plausible that major reform of the U.S. health care system will lead to a significant reduction in the rate of spending growth, especially if the reform will provide significantly greater health care services to the 40 million Americans who are currently uninsured with producing any serious reduction in services to those who have insurance.

Posted by: Jason | Apr 12, 2007 6:50:29 PM

Sorry, I just cant get enough of the data in Jason's "supporting" document

"Exhibit 3 shows the average annual growth rates for per capita health spending for the analyzed countries, adjusted for inflation in each country.8 The U.S. average annual growth rate (4.4% from 1980 to 2003) was the second highest among the countries analyzed. The combination of a relatively high level of per capita health spending in 1980, and a relatively high level of growth in that spending between 1980 and 2003, resulted in the very high level of health spending per capita that we see now in the U.S. relative to other countries."

Even in the 90's, we were growing faster than the average OECD country

"Annual increases in per capita health spending slowed in the U.S. between 1990 and 2003, but even that growth rate (3.6%) equaled or exceeded the rates in more than half of the analyzed countries over the period."

But wait, you say, how do we know that the growth reductions of the HMO era arent permanent. I dont know, but Jason's source sure doesnt think so
"After a brief respite in the mid-1990s, significant annual increases in health care spending over the past few years have refocused U.S. policymakers on the impacts that rising health care costs have on businesses and individuals and on federal and state budgets. "

No matter -- we all know Jason's position about comparing health systems. Sure, we may be spending more, but we cant really say we arent getting more, right? I mean, his source backs him up at least on that key point, the center of his argument, right?
Not so much
"Despite this relatively high level of spending, the U.S. does not appear to provide substantially greater health resources to its citizens,10 or achieve substantially better health benchmarks, compared to other developed countries.11 "

This was fun. It was really really great that you linked to this document. You are a truth seeker through and through.

Tell you what, Jason, you've given me a good idea. Tomorrow, everybody has to cite sources they disagree with. It'll be like, you know, one of those touchy feely group therapy exercises.

Posted by: RW | Apr 12, 2007 6:59:07 PM

It's hard to see how the period 1990-2003 qualifies as an anomaly. The U.S. has long spent a higher proportion of its GDP on health care than comparable nations. Look at the spending figures in Exhibit 5. Even in 1970, almost 40 years ago, the U.S. spent around half as much again of its GDP on health as many comparable nations. That table also shows that health care spending as a fraction of GDP has grown every decade without exception in all the surveyed countries since at least 1970. They're all headed in the same direction as the U.S. I think there's a plausible explanation for this. As a country becomes richer, the relative value of health care services increases and thus the proportion of its GDP the country spends on health also increases. It wouldn't surprise me if several decades from now health spending consumes 40 or 50% of our GDP.

And the growth of managed care demonstrates that the market is able to produce its own mechanisms for containing costs.

Posted by: Jason | Apr 12, 2007 7:10:36 PM

Wouldnt surprise me either. All systems grow. All systems need cost-containment.
The difference is, as a historical average, the US consistently grows faster. And its bigger.
Why is the US so much bigger than the single payers?

Posted by: RW | Apr 12, 2007 7:13:14 PM

I think Medicare, which already spends more than $400 billion per year, has the critical mass and market power to get favorable prices from providers. Indeed, it already dictates what it will pay hospitals and doctors; it doesn't negotiate. Instead of the current $992 deductible for each hospitalization, $131 deductible plus 20% co-pay with no out-of-pocket maximum for Part B services plus a $265 deductible, 25% co-pay up to the donut hole and then a 5% co-pay after that for drugs under Part D, perhaps Medicare would like to redesign its system along the progressive deductible model that Mr. Furman outlined. If it proves successful in reducing costs, then maybe it could compete with private insurers on a level playing field to win the business of the under 65 population.

If we want to reduce utilization, I would look at reforming our approach to end of life care for low hanging fruit. Why should we provide any surgical intervention at all for people with advanced Alzheimer's or severe dementia? Should 90+ year old people get bypass surgery, knee and hip replacement, kidney dialysis, chemotherapy? Why don't we have a politically independent Board modeled after the Federal Reserve that can examine the scientific evidence and decide whether or not we should pay for the newest ultra expensive biotech cancer drug based on its cost-effectiveness? If a doctor certifies that someone has less than six months to live, perhaps the default protocol should be hospice and comfort care unless the family clearly indicates otherwise. Even then, a doctor should have to certify that providing such care would not be futile. This is where the dollars are. Let's go after them. If we do, there will be plenty of money to waive co-pays for statin drugs and hypertensives for everyone who needs them.

Posted by: BC | Apr 12, 2007 7:26:40 PM


Even in the 90's, we were growing faster than the average OECD country

Wrong. For 1990-2001, the OECD average real per capita growth rate in health expenditures was 3.4%. For the U.S., it was 3.2%. [source]

No matter -- we all know Jason's position about comparing health systems. Sure, we may be spending more, but we cant really say we arent getting more, right?

Unless you can produce some kind of serious evidence that we aren't "getting more" (whatever that vague phrase is supposed to mean, exactly), that's right. You can't say it. Unsubstantiated assertion is not a substitute for evidence. And please don't tell me that you know we aren't "getting more" because we have a lower average life expectancy at birth or a higher infant mortality rate. For reasons I have explained at length, those statistics do not tell us anything meaningful about "how much" we're "getting" from our health care system.

Posted by: Jason | Apr 12, 2007 7:27:50 PM

Why is the US so much bigger than the single payers?

There are probably many reasons but one of the most important is probably that it is primarily private and market-driven, and thus more responsive to demand. The total amount of spending is determined mainly by the demand from consumers, not by government bureaucrats deciding how much of their fixed budget to spend on health. Markets are good at matching supply and demand. Governments are very bad at it. That's why the health care systems of countries like Canada and Britain are characterized by shortages, rationing and waiting lists.

Posted by: Jason | Apr 12, 2007 7:38:09 PM

I hate to say it, but I think Jason is right in this respect - changing the payment methodology in healthcare isn't going to solve the bigger (or biggest, I'd say) problems in healthcare costs. I don't, necessarily, think that's about the rate of growth in costs, but I think it's hard to see how single payer, or some other solutions, can dramatically reduce costs. The rate of increase might be reduced, but that's not going to be enough; even slowing the rate of increase in healthcare costs is still continuing an upward spiral that's ultimately unsustainable. Managed care has not, really, shown that costs can be contained; it has shown that consumers (and providers) balk at efforts to tightly control patient access to care, which really became the way to control costs. If managed care's effects on controlling rate of growth really had taken hold, you wouldn't see a return to the rates of growth seen prior to their acension. Until you start to examine why healthcare costs are so high, and what can be done to reduce or eliminate some of these costs, you can't really touch the rate of growth. The healthcare systems have really seen to that.

That said, I don't think "status quo" is sustainable or desirable, and something's got to give (and, Jason, I remain curious just what you are arguing for, not what you argue against). And I think it's why people who are really committed to changing health care in this country need to move beyond the Universal Coverage discussion and thread in other crucial reforms - about changing the way care is delivered, refocusing on prevention over catastrophic and emergency care, and addressing high cost aspects like long term elder and rehabilitative care. Insurance isn't a solution - it's a payment method. And payment methods are a discussion already at some remove from the issues at the core of getting more Americans better health care. That last, I'm sure, will get a reminder from Jason that American health care is marvelous... and it is, for some. I continue to think that we can improve care to more people, though it involves trade-offs, and those trade-offs involve educating the public. Telling the public that insurance (or a "Medicare for all" type solution) will solve all of what's wrong is not giving a full picture, and indeed, I think it's confusing and misleading.

I say all of this, by the way, merely in reaction to comments already posted. I am interested to read the proposal, but have not yet had the chance.

Posted by: weboy | Apr 12, 2007 7:42:25 PM

"Wrong. For 1990-2001, the OECD average real per capita growth rate in health expenditures was 3.4%. For the U.S., it was 3.2%."

I stand corrected.

Note, then, that when you add the beginning of the this decade, the growth rate increased again. And even these charts show how obese our system is absolute terms. US tops the chart in average per capita spending at $4887. Second place? Those heartless green eye shade Swedes at $3167. I'd say the burden is on advocates of the present system to show what exactly is is that we are getting for our $1,720 a year.

I'm sorry I havent been able to find the perfect piece of evidence that shows that we get less of services that are good for us than do other countries. Ezra's produced evidence that we receive less in hospital times. You said Hospital stays might not reflect value. He's produced evidence that we receive less in doctor face time. You said that might not reflect value. Someone else produced evidence that we receive less favorable health outcomes. That wasnt good enough either because we couldnt prove .

You have produced a reference to "Megan McArdle" that you havent linked to, and a series of documents whose authors think that, ahem, "Despite this relatively high level of spending, the U.S. does not appear to provide substantially greater health resources to its citizens,10 or achieve substantially better health benchmarks, compared to other developed countries".

If you think the evidence produced during this discussion supports your position it can only be because you are willing to assume that any conceivable gap in the data breaks in your favor.

Great if you are applying a reasonable doubt standard in a court room. But insofar as we know that we are spending out the freakin gills for health care, I'd say you (advocates against reform)are the one with the burden here.

I appreciate the dialogue, though. You've made good arguments for what you have.

Posted by: RW | Apr 12, 2007 7:43:53 PM

Should read
"Someone else produced evidence that we receive less favorable health outcomes. That wasnt good enough either because we couldnt prove that the difference in outcomes was linked to health care."

Posted by: RW | Apr 12, 2007 7:47:14 PM

I think we can all agree that 网站优化 makes some interesting points. ;)

Posted by: weboy | Apr 12, 2007 7:56:48 PM

I think 网站优化 has clearly shown us the way on health care policy.

Posted by: RW | Apr 12, 2007 7:58:06 PM

Honestly? If we weren't a society that has a strong ideological bias toward consumerism, we would find this paper embarrassingly stupid. It is the proverbial silk purse from a sow’s ear.

First of all, the authors don’t really seem to have a solid underpinning for where their savings are going to come from. The 70/20 rule has a rather huge impact here and is rather casually (and unconvincingly) dismissed.

For instance:

Much of the expenditure of this high-spending 20 percent may still be subject to cost sharing. The typical high-deductible health plan includes cost sharing up to about $7,000 of total health-care expenditures for individuals and $14,000 for families. In total, 86 percent of nonelderly households, representing 41 percent of total expenditures, fell under these limits and thus were subjected to cost sharing throughout the year.

And later on they say:

Total healthcare spending falls by 13 to 32 percent in the scenarios. Even though the bulk of health care is purchased by a relatively small number of people with very high spending, enough spending takes place in the range affected by cost sharing to have a substantial effect. Under the HSA plan or the 7.5 percent income limit, about 80 percent of families would end the year with out-of-pocket expenses below the maximum, and thus would still be facing cost sharing at the margin.

Frankly, saying that 80% of families would have costs under the maximum says very little about how much SPENDING is under the maximum. It’s a dodge, and it’s a cheap dodge and I thought the people at Brookings were better than that.

Assuming that all families could afford premiums (a point which I'll get to later), most of these “high-spenders” would be spending LESS out of pocket under the authors’ plan than they currently are in HSAs. The median household income in the US is $46,000 (I’m cheating a bit here, because I couldn’t find the non-elderly household median, sorry). The authors’ plan exempts anyone earning less than $30,000 from any co-payments. That’s a LOT of households, and there’s gotta be a significant number of “high spenders” in that demographic. In addition, there’s the lowered co-pays in the next demographic (up to roughly $40,000), and there’s the 7.5% out-of-pocket maximum for the rest. Overall, this seems like a recipe for less cost-sharing by “high-spenders” than under current HSA proposals. Frankly, it looks like Brookings was finessing this point, and the authors don’t have a solid estimate as to what percentage of healthcare costs would even be affected by their plan. That’s a rather important data point to be missing, considering they’re making a very strong prediction about how much overall healthcare costs would decrease.

Second, who are they going to sell these cost-sharing plans to? It's certainly not anybody in that 20% making up 70% of costs. If we’re selling HSAs, the risk-adjusted premiums for this population, to protect them from their significant costs, would be spectacular. So, they're probably only selling these plans to the 80% of people making up 30% of costs, and they're exempting smart preventive care from out of pocket costs...um, since most of what this healthy population is getting IS preventive care, we're essentially cracking down on the excessive cost of unnecessary preventive care. Those massive savings are not gonna be enough to make up for all the cross-subsidies we lost when these healthy folks jumped out of the big pool and bought a plan with underwritten premiums that a sick person could never afford.

Third, the writers frequently assert low-end cost-sharing will generate cost pressures in areas where low-end cost-sharing has not been proved to actually create cost pressures (and where, frankly, one would not expect it to). For example, on page 22, the authors conclude: "Relatively low cost sharing reduces demand for technologies that save money, yet guarantees virtually unlimited demand for expensive new technologies." While much evidence in this paper supports the statement that relatively low cost containment pressure reduces demand for technologies that save money, there is nothing to support the more specific conclusion that cost-sharing--especially of the sort they propose--would exert a true cost containment pressure on providers in the context of buying capital intensive medical technologies. It is simply asserted. (And, in fact, the RAND study they cite suggests the opposite, noting that once consumers come in contact with the medical system, they seem to do little to contain costs by exerting a consumer pressure. Consumers in the study only saved money by being so sensitive to costs that they didn't go to the doctor or hospital.)

Fourth (and this one is just snarky), the writers make a long digression into economic theories of risk aversion. While interesting, I really hope this orthodoxy didn’t dominate their thinking. Because, um, current economic theory rather sucks when it comes to understanding the behavior of consumers in vivo. For instance, a recent paper demonstrated that current economic theories of risk-aversion can’t explain contestants’ choices on the TV show “deal or no deal.” If economic theory can’t predict a simple gamble like that, it’s a bit terrifying to rely on it to predict consumer behavior in the complicated situation of “your money or your life.”

Fifth, I don’t find it credible to simply assume that insurance companies would pass on the savings from decreased costs (due to more cost-sharing) to consumers in the form of lower premiums. While there has certainly been premium competition in the past, market consolidation and an increasing number of for-profit companies mean that today’s insurance market is very different from even 10 years ago. I mean, the Blues are nearly all for-profits! Much more of the market has a primary responsibility to meet or exceed quarterly profit expectations to keep its shareholders happy. This mitigates strongly against insurers getting into a price war on premiums, especially as the overall market size is stagnating due to the increasing unaffordability of healthcare. They ain't gonna party like it's 1999. Furthermore, the strong underwriting in HSAs strongly mitigate against premium savings being passed on (and the more of the market they get, the worse this gets). Who the hell wants to make prices 5% lower to get 3% more market share, when that market share is made up of the people who were too sick to afford your premiums when they were 5% more expensive? Insurance companies try to make their adverse selection to get healthier people, not sicker ones.

Finally, and to me, most importantly, the paper ignores the fact that variations in care appear to be multi-factorial. While I love the RAND study, it was a study not of healthcare costs, but of a healthcare population’s response to out of pocket costs. So, RAND is a lot more illuminating about the way consumer incentives work in the healthy end of the market (where a majority of the consumers are) than it does in the sick end of the market (where the costs are). So, yeah, RAND shows us why Uncle Phil won't pay out of pocket for the statins, but it doesn't explain at all why doctors under the same Medicare incentives treat hip fractures more aggressively in Florida than they do in Minnesota. That kind of variation, which is quite financially significant, and exists across a wide number of conditions, currently exists WITHOUT any consumer input whatsoever. The authors make all kinds of hay about the significance of out-of-pocket costs, while totally ignoring this kind of variation, which is far more cost-consequential than whether someone makes two doctors' visits a year or three.

This leads me to what I think is the most important sentence in this piece, utterly buried on page 17: "It is also true, however, that health-care systems in other industrial countries do not rely primarily on demand-side constraints, but instead use global budget caps and rationing in ways that have no analogy in the United States."

I simply fail to understand why anyone would put a lot of stock in cost-sharing as a particularly effective cost-containment policy for the kind of costs that are making the system unaffordable.

Fundamentally, all these consumer-based approaches seem to me to be nothing than grand experiments in how to save money by denying unnecessary care to basically healthy people. I don’t have a problem with this, per se, except to note that the authors themselves admit that we’re not really all that sure what effective & ineffective care is. And, um, I’m not sure insurance companies will really spend so much money trying to find out what they should be paying for, instead of somebody else.

Not to mention the fact that these authors are not only ignoring the bigger problems in our healthcare system, and the areas where we could save money as well as the downside of cost-sharing approaches. Cost-sharing schemes do tend to lead to a “blame the victim” mentality. You start paying for pap smears, and men start saying things like “well, I shouldn’t pay for this! I don’t have a cervix!” And so on down the slippery slope. I mean, a lot of things in healthcare lead to “I don’t want to pay for smokers to get cancer treatment,” but it seems like any blog post about HSAs and cost sharing tend to go there in the first three comments, and it’s not just the trolls who are making the argument.

And, along without discussing the philosophical downsides of cost-sharing, the authors are rather flagrantly ignoring premiums. While I understand the desire to limit an issue, the fact is that cost-sharing is strongly associated with HSAs, which are all cherry-picking schemes that remove the cross-subsidy from the healthy to the sick out of our healthcare system. They justify it by saying there will be huge savings. And, I'm sorry, the savings generated by spending less on unnecessary care for healthy people simply isn't enough to make healthcare the least bit more affordable for people who actually need it (not that the authors of this study point it out, with their crazy estimates of 30% savings.) Not to mention, healthy folks are pretty damn entitled, and they want their own savings from the fortune of their good health to go into their own pockets. Cherry picking is, frankly, a much bigger problem in our system than the kind of minor dysfunction that low-end consumer cost sharing can fix, and it’s kind of weird to set aside cherry-picking while extolling low end cost-sharing.

Having beat up on this paper, I think a lot of the things in it are right, and it makes me optimistic to see them moving to the center of the healthcare debate. I’d love to see more time and energy devoted to restructuring our system in a way that incentivizes primary care and preventive care. (I’ve got a pet notion that we’d save money if we actually paid chronically ill patients to take their statins, but I suspect I’m in the minority on this.) I think we should be spending a LOT more money empirically figuring out what preventive and chronic care treatments don’t work, and how to get them implemented. I can see where cost-sharing sort of pushes us this way, though I’m not completely convinced it’ll yield any more in the way of concrete results than capitation did. I’m glad to see that centrists have realized that it’s not just morally wrong but stupid, to make chronically ill poor people pay for preventive care. I’m glad to see the diminishing marginal returns and risk aversion treated differently in poor and rich populations. (OTOH, it’s not clear to me why we’re capping co-pays for rich people). I also like the clear statement that the market will not do these things on its own, even if they are better.

But all in all, this is just a massive oversell of consumer based care, even as it promises it’s not a “consumer based care is a magic bullet” argument. I think it’s great that a centrist think-tank is pointing out the way to do consumer-based healthcare better, especially given the fact that conservatives keep pushing bad consumer based-based healthcare as a cureall. But, not only are the centrists overselling consumer based healthcare but…so what? Seriously, does this idea amount to a hill of beans on its merits?

If a centrist think tank were seriously proposing a national single-payer healthcare system, whose key feature was that everyone would be enrolled in a premium-free HSA with the same features as the one they're suggesting, I would be so excited that I would wet my pants. Because, see, the savings we’d get by creating a monopsony buyer in the federal government (or even 50 state governments!) would be huge. It would be a platform to force all kinds of efficiencies onto the system, and we’d also get the added nickel savings of discouraging unnecessary preventive care for healthy people. Woo hoo. Hell, the savings we’d get from bypassing the insurance market alone would be staggering, and they would probably offset the expansion costs.

But this? There’s no real savings here, there’s no change here, there’s no discussion of downsides, there’s overpromising. It’s basically just a pean to how much money we’d save if only we put “the people” more in charge. I’m happy to make this kind of silly “power to the people will fix everything” argument if consumer-based initiatives are tied to a broader reform plan that actually DOES something. But this reform plan is just ways to put a little good consumer-based healthcare into our existing system. That’s like putting a pinch of salt on a burnt turkey. It ain’t gonna save Thanksgiving. And it just might give ya high blood pressure.

Posted by: anonymous | Apr 12, 2007 10:10:30 PM

Oh dear... well, now that I've read Furman's piece, I think I'd just make my last point more strongly - I think the proposal is way too focused on a payment structure, and not nearly enough on other fundamental changes and reforms that go hand in hand with figuring out how we pay for it. It seems unrealistic to me to dream of some accounting system that could track a "7.5% of income" cap and not see the kind of mischief such a program would introduce, full of unintended consequences (question 1 - define income. How many pages of IRS regs is that alone?). Moreover, I question Furman's emphasis on "families" which raises a number of questions in my mind about whether what he proposes really makes sense for singles; I get the sense he's cutting a plan to fit the "average family of four" example (which, statistically, is no longer quite the norm anyway, but never mind) while ignoring what it would mean for a young person just out of college, or a person on a fixed income by themselves, or a number of other specific examples.

I think Ezra's got some good background stuff in his piece which lay out things we've been discussing here for a while in a nice, concise way (pardon me for hearing reiteration of objections I've raised at various points along the way). But what I think starts out with a promising, challenging opening (the idea that "Universal Health Care" sounds very nice, but isn't fully formed, and an acknowledgement that cost containment matters as much or more than reinventing how we pay for care) kind of peters out into accepting Furman's notion that cost containment will... just happen, somehow. And that's no answer.

And I think Jim's post above is dead-on about why this proposal's big failure is that it lacks a framework, and raises more questions than it answers.

Which goes back to Jason's (or at least my take on his) point - changing the way we pay doesn't mean cost containment. Really containing costs means looking at things that are, in many ways, not about insurers, or Medicare, or other payers. They're about how we deliver care, how health care professionals work (or don't work) together, and how many things in the systems we have are really at cross purposes and create perverse incentives. Issuing insurance policies tomorow to everyone won't change that. Now what?

Posted by: weboy | Apr 12, 2007 10:17:10 PM

I'm confused. The Jason posting in this thread is not Jason Furman the author of the paper correct?

Posted by: DRR | Apr 13, 2007 3:09:55 AM

The comments to this entry are closed.