In a comment, he writes,
As bad as it may presently be, the whole domain between the current equilibrium and a more genuinely market-like equilibrium is a vast no-man’s-land of even more dysfunctional instability. This type of circumstance is usually a big factor when bad systems are surprisingly stable.
Again, anyone that proposes a reform that relies on nudging the system towards the market-equilibrium from where we are now by taking one little step is being very naive. One has to accept the possibility that we might have to pair pro-market moves with non-Libertarian coercive regulatory actions, at least in the short-term.
These are provocative statements, with which I am inclined to agree. For a long time, I have been saying that he basic problem we face is that in the United States we undertake many medical procedures that have high costs and low benefits. There are two directions that we can turn in order to change this. The “left turn” is toward government rationing. The “right run” is toward the price mechanism, with market forces turned loose and health insurance providing only catastrophic coverage, so that consumers must confront the true marginal cost of what I call “gray area medicine,” where the benefits are low and uncertain.
Obamacare is a weak left turn, that neither solves the basic problem (high-cost, low-benefit medical procedures) nor satisfies the goals of those on the left to equalize access to health care services. My reading of Handle’s comment is that what is likely to happen if Republicans win in 2016 is a weak right turn, which will not solve the basic problem and not satisfy the goals of the right to foster a flexible, innovative health care system.
I will repeat my mantra that what individuals want is unlimited access to medical services without having to pay for them. To the extent that politicians attempt to satisfy this, they create incentives for ever-increasing spending. That describes the status quo, which Obamacare did not change.
From where we are now, it is hard to move either left or right. Households and businesses are basically satisfied, despite grumbling about higher health insurance premiums and higher deductibles. Health insurance companies are basically satisfied, despite grumbling about regulation. Health care providers are basically satisfied, despite grumbling about insurance company practices. So the politically winning approach is to tell everyone “you can keep your ___.”
This is the theory of the second best, no?
I will repeat my mantra that what individuals want is unlimited access to medical services without having to pay for them. To the extent that politicians attempt to satisfy this, they create incentives for ever-increasing spending.
Is that actually true? I do not want unlimited medical services — I only want enough to ensure my good health. I don’t like going to the doctors, and I hate hospitals. Right now my health is good, so I have no demand for health services. If I was unhealthy, I would want only enough healthcare that could in high probability restore my good health. I know there are hypochondriacs and others who grasp at any magic bean they think might save them. But I’m not sure you need to use a price mechanism to limit their access — doctors could simply say, “There is nothing we can do to make you better. Sorry.”
If you held hospitals and doctors accountable for only recommending treatments that are likely to make people more healthy, then you solve all the problems at once. You do not need to ration based on price, or ration at all, because we actually do have the resources to give all the health-benefiting care that is needed. But if doctors are not held accountable for recommending bogus treatments, then even rationing based on price won’t work, because people will be scared and conned into paying for expensive treatments that they do not need.
In other words:
1) 99% of the time, people do not know what kind of treatment they need. They have to be able to trust a doctor to give them good advice.
2) Thus, the core issue at stake in designing an incentive system for healthcare, is to give doctors the incentive to make accurate prescriptions. This could be done by requiring doctors to formally write down their treatment, and their estimated probability that the treatment will work, and then tracking the doctor’s estimates over time. If a doctor predicts continually predicts that a surgery will work out for the better 90% of the time, but it does so only 30% of the time, that doctor should be fired.
3) If you do 2), then we are already spending enough to give everyone all the treatment that they need that passes a bare minimum cost/benefit ratio. There is only so much healthcare that can reasonably be done.
I doubted whether or not unlimited demand for health services was true before I had a number of health problems. Strictly speaking, it’s not true – your demand is finite.
The problem is that health care goods and services are just proxies for attempts to buy better health, and it’s very hard to do that with the technology we have today. A lot of interventions are “well this works in 20% of cases” or “this won’t heal you but it will make you feel better some of the time” or “that drug didn’t help, we’ll maybe we can try this.” Different doctors say different things. You may go to one, and wonder what you might hear if you go to another – sometimes it’s the same story, but other times it isn’t.
In my experience, most health care interventions fail partially or completely. However, on occasion they do succeed, and it isn’t always obvious ex-ante which ones are the ones which will work. Other times, they might not be necessary but you really don’t know that in advance Would I have died if I didn’t go to the hospital after getting a severe case of food poisoning? I don’t think so, but I don’t know for certain, and I certainly wouldn’t have taken my chances at the time. When you feel sick, you’re willing to try a lot of “well, there’s a 20% chance this could work” type solutions, and especially so once the deducible’s maxed out.
My scheme for incentivizing doctors via predictions really would only apply to major, pre-scheduled treatments where the patient is getting multiple opinions.
For emergency care, I think it would make the most sense for the hospital to be funded out of property taxes (since you are essentially choosing your hospital when you choose a city, same as you choose a fire department service or police service when you choose a city, you don’t choose the emergency room in the heat of the accident). The hospital should simply get enough funding to amply treat all comers. It should treat people with problems, and turn away people who are abusing the system. I don’t think you need any fancy incentives for emergency care.
Me, too, but here is the thing- as I have aged past middle age (I will be 50 next Summer), I have been given a real close-up look into old age by my parents and grandparents, who also were like the young me and you at one time. The amounts spent after age 60 are multiples of what they contributed in real resources via insurance premiums, FICA taxes, out of pocket payments, and any reasonable return on those- and my parents, who are retired, are still alive. I was reluctant to do so, too, but I now firmly believe Kling is completely correct.
If you held hospitals and doctors accountable for only recommending treatments that are likely to make people more healthy, then you solve all the problems at once.
And the AARP would have the head of any politician who proposed medicare only pay for procedures supported by strong evidence.
BTW what is generally meant by unlimited access to medical services, is any treatment with no regard to price. It some $10,000,000 treatment that increased chance of a cure by 20% existed, people believe that they should be able to get that treatment. They do want anyone to say no it costs too much so you do not get it.
Unfortunately, catastrophic coverage is the core of high cost, low benefit. There isn’t much to be saved where there isn’t much spent, and little incentive to save where most is spent.
Catastrophic coverage can be structured to provide incentive to save. They can pay 95+% of the costs so there is still incentive to save 5% of a $100,000 bill.
When you and Handle agree the likelihood that I do is basically 1.
Cool. Then let’s just mandate HSAs against all objections. Just kidding. But the point is I suspect we will not get anything we want resorting to coercion. HSAs already work despite hostility from the coercion caucus, having to carry the free-riding cost unconscious, and occupying a small market niche and having to wag the dog. And yet they still work.
You can claim that they are hard and this seems like an ironically common complaint. “HSA-eligible plan enrollees who participated in GAO’s focus groups generally reported positive experiences, but most would not recommend the plans to all consumers.” (https://en.wikipedia.org/wiki/Health_savings_account#Consumer_satisfaction). Goose, gander.
And yet they still work. So, one can’t conclude from anecdotal experience that being hard makes them not work because the studies show that they work on net. So, maybe some of what makes them hard is what makes them work, and not something that makes them not work. Maybe if more people did it, they’d be easier, so anything that artificially stands in their way could be eased. And by the way, it’s not like everyday insurance snafus, finding in-network specialists and sovereign bankruptcy are easy in the long run either. What choice do we have? And why isn’t promoting HSAs and fighting against the hostility against them a really good one?
Devin Helton: “If you held hospitals and doctors accountable for only recommending treatments that are likely to make people more healthy…”
I don’t think this is as doable as you think. Example: I swallow a small aspirin each day. Studies show aspirin reduces the odds of strokes and heart attacks. So that’s good — aspirin is cheap, strokes are expensive. But the studies also show the effect is small; if 1000 men take aspirin every day, over a five year period we can expect ONE less stroke or ONE less heart attack in that cohort. At a cost of … maybe $35,000 of aspirin, and $100,000 of medical consultations.
So let’s save money here. Aspirin works for one guy in a thousand. We just need to identify that guy and tell the other 999 they’re wasting their money, and through the magic of the free enterprise system, we’ll have improved the health care system!
But we can’t find that one guy in a thousand who was saved by aspirin, We can point to 40 or 50 guys who did have heart attacks or strokes, and say that aspirin didn’t work for them. But that leaves 950 guys who didn’t have heart attacks and strokes, and our one guy who actually did benefit from the aspirin is just one of them. He doesn’t stand out.
And if we tell this group of 1000 men to forget about aspirin, it’s a myth, it’s a silly waste of money … after five years maybe we have 51 dead guys rather than 50. So the studies say. So we do autopsies, and the pathology reports come back reading “Heart Attack, Stroke”. There’s nothing to say that an aspirin a day would have prevented any of those deaths.
But, $35000 to save a life is a nice cheap example. That’s the cost of three colonoscopies, let’s say. How many colonoscopies does it take to prevent one death by colon cancer? Three? A hundred? A thousand? Can you predict who is the one person in three or a hundred or a thousand that will actually benefit from a colonoscopy? Identify that one person in advance, so we don’t waste money on the people who aren’t going to get cancer?
I gather you think this is all very easy. I disagree.
Mike shupp-
The fact that there are borderline cases where the cost-benefit is not obvious, and is hard to measure, does not mean that rationing on price is better than rationing based on doctor’s advice.
For most cases in healthcare, we know there is a benefit. Broken bones need to be set. Torn ACL’s need to be stitched up. Heart attack victims need emergency care. Children with nasty cases of strep throat need antibiotics. Etc. If we did not know there are obvious benefits in most cases, then there would be no point in healthcare at all.
I needed surgery for a sports injury. My doctor proposed a semi-risky procedure that had a 90% chance of success. But I had know way of knowing whether based on track record his predictions were accurate or not. If it was really closer to 50%, then maybe I don’t want the surgery. I was basically placing blind trust in him. I’m not claiming my system would be perfect, but I think it would be a lot better than blind trust. It was obvious whether the surgery succeeded or not (it did). Why not have third party auditing so that a doctor’s track record was known, and so his opinion and recommendations could be trusted?
There are also cases where we don’t know whether the benefit is significant. Those will always be hard calls, in any system. For standard, preventative medicine, like aspirin, the insurance company would probably just have blanket policies over what it would cover or not. Because in those cases, there would be no real way to judge a particular doctor based on his particular advice.
I’d also note, that for both healthcare of dubious value, the cost in terms of iatrogenics is greater than the cost in money. All drugs, all treatment, all surgeries, are by default net harmful. So even if something is free in monetary terms, a consumer only has an incentive to want that treatment if they are quite sure that the treatment is helpful.
Apologies in advance for the length. If I had a blog I’d post this there…
“I will repeat my mantra that what individuals want is unlimited access to medical services without having to pay for them.”
I also want unlimited internet access, steak, and high-end cars without having to pay for them. The problem in healthcare isn’t the lack of a pricing mechanism. It’s how to set up a pricing mechanism given the multiple interlocking complications:
“Reality” Structural
(1) Poor/non-existent outcome metrics. There are diseases for which there are relevant data (e.g., cancer survival rates…though interestingly data from some registries can’t be shared). There are far more where there’s no available data. Which ortho surgeons have the highest proportion of ACL repair patients who return to full performance? Which neurologists have epilepsy patients with the best control of their disease? Value = outcome/cost and, in general, there’s no good measure of outcome.
(2) Unpredictable cost of care. The cost to provide a given procedure can be highly variable and driven by multiple factors. Some of those factors are mostly controllable (central line infection). Some are predictable (a hemophilia patient that needs Factor VIII is going to be really expensive). Some are semi-predicable or semi-controllable (BMI is a contributing factor to cost-increasing complications). Some are neither (patient has an allergic reaction to a medication). Controllable and entirely unpredictable/uncontrollable can be managed with a large enough risk pool. But do you charge high-BMI patients more for their known and unknown costs? What about hemophiliacs? Especially since a bad outcome isn’t a % increase in cost it’s a multiple. If you say ‘let the market decide’ then some providers will get selective about patients (back-door differential pricing).
(3) Which leads into the issue of ‘fairness’ in pricing. There are known and suspected drivers for both disease and relative cost/complication rates. Under what circumstances will we allow individuals to be charged more for the same service? Controllable (like BMI) is OK but genetic is not? How much control? How strong a relationship between the factor and the cost driver? Pricing mechanisms may actually work on the margin, but there would probably need to be some public augment for patients with obviously uncontrollable/genetic conditions. But given how special needs in education has evolved, that will tend to undermine a support system over time.
(4) There is more than one healthcare market. Cancer treatment. Ortho trauma. Diabetes. Asthma. Depression. Schizophrenia. Infectious disease. Some of these ‘markets’ use the same resources (operating rooms, CT scanners). But they differ more than they are similar. Some are discrete episodes, some are chronic conditions. Some are caused or strongly influenced by the ‘customer’ and some are an ‘act of God’ (deserving vs. undeserving poor?). Some have clear, measurable outcomes, some have entirely subjective outcomes. Expecting one ‘system’ to work for all of these ‘markets’ is folly.
“Policy” structural
(5) Government interventions impacting pricing signals. Andrew_M_Garland had a post on the same thread noted in the OP where he talked about the distortions that Medicare (and commercial insurers) introduce. Those distortions not only complicate pricing, they are a strong disincentive to publishing prices (real prices as opposed to ‘charges’)
(6) Actual government reimbursement is structurally below the cost of care. One example: marketing is not an allowed cost in the Medicare Cost Report. But any provider is probably going to be spending some money on marketing/ advertisement. So revenue is structurally below fully loaded cost. Which means some amount of cost-shifting is required. Which means some amount of differential reimbursement is required. It’s not a ton of money. But in an industry with average margins <5%, a little money is a lot of money. So you can't charge a single price for any provider seeing Medicare patients. Medicaid patients are typically reimbursed lower than Medicare, by the way.
(7) Regulation. Lots of licensing (limits supply of providers). Lots of regulatory constraints on practice (clinical, who can own what, how payments can be made & shared, when providing care is mandatory, etc etc). There are obvious arguments in favor of maintaining quality (esp given #1 above) but they work against a free and fluid market in search of a more efficient care delivery mechanism.
I think there might be a way to make incremental steps by explicitly carving off parts of the "overall market" (#4) and managing them differently. For example:
* emergency services are a municipal service much like fire and police
* little or no social coverage for 'diseases of old age' (most cancer, alzheimers). You save for them as part of the expected cost of aging.
* catastrophic life-insurance-like coverage for prime-of-life diseases like early onset alzheimers
* Pre-tax HSAs for anything 'optional' or grey (ACL repair? you don't NEED an ACL…)
* Public (fixed?) payments for a defined schedule of conditions, mostly early onset (type I diabetes, childhood cancer) or genetic.
Some might work, some not. But each sub-market carved out of the larger HC market would both create a functioning healthcare sub-market and simplify the remaining hairball.
All that said, I think Handle's fundamental point holds. It may be possible to make a little "right" step or two without much coercion. But at best they would be confidence-building steps before more serious changes in the regulatory landscape. And a little step or two without a coherent system in mind as the endpoint is a random walk.
Interestingly, it also appears difficult to make large "left" steps. Little steps are no big deal…CMS does that already. But Obamacare required a full-court press by all three branches of gov't.
I agree that the problem with US health care is path dependency. Changes involve trade offs, which for many tilt heavily in the cost direction.
If we were starting from scratch, it would be relatively easy to have a cheap safety net program and let the chips fall where they may in the market for everyone else, but for a sizable number of people who are currently in or expecting to fall into the expensive safety net, it is a downgrade.
“I will repeat my mantra that what individuals want is unlimited access to medical services without having to pay for them.”
I actually got that in 1997: I was the first person in the world with my kind of non-Hodgkins lymphoma to be treated with the breakthrough monoclonal antibody rituximab, and I didn’t have to pay for it because it was a clinical trial.
It was great.
I sort of have to laugh when the right talks about ‘government rationing of health care’ when the private system we have does far, far, far more rationing than any single payer or direct-provision system does in other countries.
You have an entire industry set up to deny care to people. In contrast, wait times tend to go up in the UK when Tories push through weaker NHS budgets.
But the private system is enormously lucrative for providers and insurers, so hey, let’s do that!
And they actually believe that denying others access will lower prices for themselves, rather than lowering the scale of treatment and increasing prices or reducing investment in the first place.
I would love to read what Arnold thinks would happen if it was made much easier to become an MD, PA, RN, LPN, to open a hospital etc. That is supply side reforms.
BTW here is my preferred plan on the demand side here:
The state would provide insurance to all Americans but the annual deductible would be equal to the family’s trailing year adjusted income minus the poverty line income (say $25,000 for a family of 4) + $300. So a family of 4 with a trailing year adjusted income of $30,000 would have a deductible of $5,300. A family of 4 with a trailing year adjusted income of $80,000 would have a deductible of $55,300. Middle class and rich people could fill the gap with private supplemental insurance but this should be full taxed. This would encourage the middle class and rich, who are generally capable people, to demand prices from medical providers and might force down costs.