The Tom Coburn samizdat Medicare reform proposal

As reported by Ezra Klein:

“If I had the magic wand,” he told me, “I’d change how we pay for Medicare.” That’s a common enough sentiment, but the policy Coburn has in mind is a bit more radical than what’s typically offered in Washington.

“I’d change all physicians to time instead of fee-for- service,” he says. “What we’re doing with fee-for-service, and most people don’t realize this, is when you go to the doctor, they have this pressure to see X number of patients a day to meet their numbers.”

If we cut payments to doctors, Coburn says, “they’re going to cut the time they spend per patient. When a patient is in a room and you haven’t used your skills as a physician to really listen, you walk out and cover that absence of time by ordering tests. So if you say here’s all the hours we’ll pay for if you’re a Medicare doctor, and we can actually audit that time, doctors would have to demonstrate proof that they’re spending this time with patients.”

That wasn’t, I noted to Coburn, a policy that appeared in any of the bills he had sponsored, a fact he acknowledged with a laugh. “I didn’t put that in there,” he said, admitting the idea has little political support. “It’s just something I’ve thought about a long time. Nobody should be seen for less than 20 or 30 minutes if you’re doing this properly. And if I knew I was going to get paid for my time I wouldn’t be in a hurry to see the next patient.”

Here are further ideas on Medicare reform.


To quote the tennis great John Patrick McEnroe, Jr., "you CANNOT be serious".

Dammit, Ray. John is the multi-grand slam winner with a loud mouth.

Patrick is his soft-spoken younger brother with a Stanford degree.

Little known is that their full names are John Patrick and Patrick John.

Or American doctors could just earn what their counterparts in France or Germany do.

Nah - why would doctors in the U.S. want to be paid less?

And thankfully, due to the American political system's devotion to protecting certain classes of worker from international competition (yes, Dean Baker is a fine source for further study), American doctors won't be facing any pressure to reduce prices.

GPs are supposed to keep you away from specialists. Do you KNOW they aren't cost avoidant?

Are you referring to the idea of 'Hausartzt'? Or the need to get what is called an 'Überweisung' before seeing a specialist? Because that system seems to be working out pretty well here, Though the idea of having a single doctor familiar with a patient's history and needs does seem a bit old fashioned. But then, Germany has the problem of too many doctors, not too few - yes, Germany exports doctors, including to the UK (I know one - he uses Ryanair, making his commute quite cheap, as he lives maybe a half hour from the regional airport - but admittedly, RTyanair is awful).

We have a hypothesis.

I'm fine with importing doctors, but we should also lower the barriers to entry for doctors trained in the states.

Precisely this. If you want a cheaper medical industry, you should have more medical providers. Preferably many more.

It's a pretty radical idea for an economics blog, I realize, but I'm throwing it out there anyway.

Nah, I'm kidding. Just raise taxes! On the "wealthy"! Problem solved!

I've met enough dim bulbs in the medical profession that relaxing standards for entry seems to me like a move in the wrong direction.

So current policies restrict supply and we still have poor quality? Sounds like more reason for reform, not less.

I fully agree.

When I suggested that STEM fields are sometimes overly difficult to enter relative to the actual technical knowledge/skill required after college to a conservative family member he thought I was recommending that we allow bridges to be built by idiots. It's a non-starter politically - especially considering incumbent lobbying interests.

Oh well.

Doctors are a very powerful lobby, for better or for worse. Simply trying to cram down their incomes will be a waste of your time.

Yes, that mechanism is clearly at work. That's why you never, ever see a doctor working in the US who immigrated here from India, Pakistan, or the Middle East.

It's better than fee for service, true. Fee for service has all the bad incentives, this has fewer bad incentives.

Ironically, this is how psychiatrists billed insurance until very recently. Now we're switching to the same system that the rest of medicine uses.

This is still basically the model of psychoanalysts who run cash only practices. They also see a patient 4 times a week for years...

I doubt this proposal is a magic bullet. It will either encourage fraud (e.g., holding 'classes' where, with patient consent, a room full of 30 patients get to quiz their doctor and get feedback about their illnesses), or, will create bottlenecks and queuing (since hypochondriacs who run to the doctor for every ailment will have to wait in a long line to see their doctor, as he'll be busy patiently explaining things to other patients). A better solution: do as they do in all other countries outside the USA: let people who have heart attacks die (do not resuscitate) and pull the plug on terminally ill patients. Problem solved, old Eskimo exiled to iceberg style.

The "all doctors are crooks" thesis...

DNRs and living wills are pretty common in the US. Plugs are pulled very regularly.

"Fee for service has all the bad incentives, this has fewer bad incentives."

I don't agree. Under an approach like this, there would be no incentive for innovation or high productivity -- physicians would receive the same payment for a day's work regardless of how many or how few patients they saw in those hours. See one patient for 40 minutes or two for 20 minutes each, and the fee is the same, but all the other costs (scheduling, billing, etc) are twice as high in the 20-minute-per-patient model. So appointments would get longer, physician productivity would decline, and costs would rise.

Cleveland Clinic and Mayo Clinic are two places where doctors are salaried, not pay-per-service, and they get very good patient results at lower cost.

Lots of people think the more things the doctor does to you, the healthier you are. Which isn't true.

The physicians may be salaried, but the Clinics do charge fee for service, and at their annual reviews, the doctors are certainly made well aware of their productivity.

The "all doctors are lazy and only in it for the money" thesis...

So, are you saying that incentives have no effect on doctors?

Actually, most human beings are lazy, and generally speaking they're in their day-jobs for the money.

Doctors, on the other hand, are uncommonly non-lazy (pre-med curriculum, med school exams, 70-hour workweek in residency.) A primary care physician is not terribly money-sensitive, given that he/she could have had a substantially better salary as a specialist.

This sounds like a debate between one person who thinks taxi drivers should be paid per passenger and another person who thinks taxi drivers should be paid per mile or per minute.

If you pay the taxi driver a large base fare (which is by the passenger), then he will try to get you to your destination as quickly as possible, and find the next fare ASAP. But he will also refuse to take passengers who need to travel longer distances.

If you pay the taxi driver a small base fare and pay him by the meter instead (which is by speed, i.e. time and/or distance), then he will try to take as few fares as possible, and take the most circuitous route possible for each.

The difference between a taxi driver and a doctor is that time with a taxi driver is bad, and time with a doctor is good. Other than that the analogy seems to hold.

So why would you want to pay only by time? Doctors will see fewer patients (or deliver fewer services) if they are paid for their time rather than by the service. Each patient will get more attention, each service will be rendered more carefully, but unless the doctor's waiting room is always full, he has less incentive to see many patients.

Likewise, why would you want to pay only by service? Doctors will see as many patients as possible (or administer as many services as possible) if they are paid by the service rather than by their time. That'll get lots of patients served, but won't it get them very careful service.

Taxi company managers set both a significant-bu- limited base fare plus a meter-based fare for a reason -- they're trying to maximize the efficiency of their drivers, given the drivers' incentives. Who will do the same in the healthcare market?

You also have to account for the fact that if someone is a local and hires a taxi driver, he'll know if the taxi driver is taking the best route or not and can hold the taxi driver accountable. Is the same true for patients and doctors.

In any case, I find all of this "top-down" engineering of incentives obnoxious.

To your first point, the analogy to doctors would be if patients know whether the doctor is taking longer to keep him/her healthy and thereby keep him accountable. But a big difference is that we're talking about Medicare -- it's paid for by a third party. So passengers will keep a taxi driver more accountable to the extent they can, but patients will not. They will hold the doctor accountable for wasting their own time, perhaps, but not for wasting the government's money (two different costs).

The your second, yeah, I mean I find it obnoxious too, but it's important to keep two things in mind. One, we're talking about Medicare, which wants to guarantee a certain level of care for seniors. If you accept that that's a worthwhile policy goal, the question then becomes how to pay for it, which necessarily involves a lot of top-down engineering of incentives. If you don't accept that as a policy goal, then there isn't any need for top-down engineering of incentives, but will the market take care of seniors to the extent that you personally (or "we" as a society) would like? That's a tough question.

...oh and two: it's also worth thinking about private-sector top-down incentive-engineering. Look at the taxi driver for example -- his incentives are engineered not by the consumer but by his employer, who is adjusting the base fare vs. the fare-per-mile in an attempt to maximize profits, which (hopefully) maximizes utility for the passengers (I think it does). That is a private, profit-driven top-down engineering of incentives, but it is top-down engineering of incentives nonetheless.

The "all taxi drivers are crooks and looking for anyway to take all your money, just like doctors" thesis...

Lift caps on medical school seats?

And/Or put in a pathway to doctorizenship for nurses.

Isn't there already widespread 'pushing down' in medicine. I know this awesome NP, prolly better than most GPs.

Actually, the age-old, evolved, system of payment called "capitation" seems to work best. Doctor is paid by how many patients have the right to see him. Incentive is for doctor to keep you well at low cost. Major, not the only, remaining problem is end-game, so doctor must have right to transfer you to hospital where doctors could be salaried.

Also, in US, produce more doctors.

In other words, the concierge practice also known as accountable care organization, previously known as health maintenance organization, where a group of doctors provide all your general health needs with reinsurance to cover things beyond his means.

Bad idea. Look what this has done for the auto industry. Back a century ago you bought cars a la carte so you paid for only the car parts you wanted and oversaw all the design and construction because consumers are smarter than car manufacturers. Today cars are so much more expensive because the automakers add all sorts of unneeded features to cars against your will just to take more of your money because they bought government to get a government takeover of the entire auto industry way back in 1920. Look at how terrible socialized autos are...

Reminds me of the (probably apocryphal) stories of Communist Russia. Politburo decides to reward Glass-factory managers based on tonnage and they produce these huge blocks of thick plate glass. Politburo wises up and decides to now reward on the basis of square area of glass produced and the factories retool to produce these super thin sheets, again hardly of any use.

Incentives are a tricky business.

So if the standard checkup takes half an hour, then prices will have to rise to clear the market, yes? Checkups become much less affordable for a lot of people. That is, a lot less affordable for Medicare recipients.

Also what Rahul said.

If a half hour four times a year to monitor medication, with time spent educating and motivating, prevents multiple trips to the ER with hours of waiting time, doesn't that reduce the demand for doctors?

Somehow restricting by price access to doctors has not made people healthy, so the ER gets lots of demand for preventable medical problems.

Any economist would automatically think of the substitution effect because it is so integral to price and market theory.

"If a half hour four times a year to monitor medication, with time spent educating and motivating, prevents multiple trips to the ER with hours of waiting time, doesn’t that reduce the demand for doctors?"

How does 300 million people x .5 hours x 4/year = 600 million doctor hours/year sound to you? We'd need 7.5 million doctors doing nothing but checkups.

doctors only work 80 hours/year?

I think mulp was referring to Medicare recipients, since that is who Saturos referenced, and Medicare recipients require more doctor time than younger adults.

But if the average for all Americans was 2 half-hour visits per year, and a good portion of those visits could be handled by NPs, that doesn't sound impossible if America could figure out a way for that not to cost an arm and a leg.

There has recently been an intriguing proposal offered here in Switzerland. The basic problem is we have too many specialists, and data that shows that each extra specialist costs the system an extra CHF500,000 per year.

The proposal is to require specialists to have a permit in order to practice, and to limit the number of permits available in each specialty. Then auction off the available permits for a 5-year period, with the permits going to the doctors who will accept the lowest reimbursement.

Lot's of details to bedevil the implementers. How to set the number of permits by specialty and region (its been suggested it could be a board of "experts" in each Canton). How to keep tabs on quality. Etc. Austin Frakt has been talking about competitive bidding for insurance companies, but doing this at the provider level makes a lot of sense.

Restricting entry probably won't work, just as restricting hospital construction under certificate of need statutes didn't work.

What you need to do is focus on the payment mechanism--something like accountable care organizations, hmos or staff doctors with pay for performance payment mechanisms.

Docs and hospitals like restricted entry.

I think doctors at the Mayo Clinic are all salaried.

This is the only way to limit he amount of useless stuff doctors do. Providing uncapped payments to doctors based on how many bubbles they fill in is a recipe for a lot of filled in bubbles. Whether those bubbles are patients seen or hours spent with patients, it doesn't really matter, because neither clearly leads to better health outcomes and each is as easy as the other to game.

It as though we have created a mutiple choice test whose correct answers are unknown to us. Some people think we should grade the test based on how many questions the test taker has answered, regardless of which answser choices she has given. Others think we should grade the test based on how completely the test taker has filled in each of the bubbles on the answer sheet, regardless of how many bubbles have been filled in. A third group thinks we need to do some combination of these two. Until we know what the right answers to the questions actually are, we cannot really start talking about the proper way to grade the test.

I was going to say something similar.

Just throwing one proposal out there: if you choose to operate within the Medicare system, you are paid a base salary with a bonus structure based on patient feedback and/or some other metrics that are as ungameable as possible.

Maybe number of patients seen and time spent per patient (patients would have to sign off at the conclusion of each visit).

And as with any base salary, in return, you are expected to provide at least a certain level of service (x number of patients per year, minimum).

So, you want a death panel to decide how doctors get paid for the services they provide based on a package of services with measured outcomes instead of a la carte?

That is what the IPAB is chartered to do.

Doesn't that only shift the reward from the doctors to Mayo?

It is also called practicing law.

That way, medical practices can become as dysfunctional as law firms and doctors who want a life get out of hourly work as soon as they can.

+1 this. Billing for services and billing for time both have diadvantages and can create bad incentives, but I would give the advantage to billing for services over billing for time.

If he had a magic wand he would probably want to change how chefs cook in kitchens, how CEOs organize their todo lists, etc... And why again are polititions mucking around in what should be private enterprise again? Whatever happened to that idea of smaller government...

Coburn is a doctor.

Seems like a recipe for long chit-chats with the doc about life, your last vacation, how the kids are doing, etc.

As long as government is involved, incentives will be completely screwed up. There is no way around that.

More fundamentally, there is no magic set of incentives that will wipe away the principal-agent problem. All you can do is try to find some workable set of incentives that works well enough, and that is also politically acceptable given the players and their respective power. My impression is that the first problem, while never perfectly solveable, is not all that awful to solve. But the second problem is very hard indeed.

One major problem is that both private insurance and medicare/medicaid involve huge and distant bureaucracies trying to manage the incentives of people very far away from them, and both the doctors and patients have different incentives than the bureaucracy, and what we want ("we" as in the voters of the US, or the company that employs the insurance company, or whatever) is still different. It would be nice if this weren't necessary, but since normal people generally can't afford the kind of costs that come up for serious medical problems, there seems to be no good alternative. To a first approximation, there aren't any 30 year olds who have enough saved up to deal with, say, a breast cancer diagnosis and all the treatment involved, so either some outside group pays with all the principal-agent problems that involves, or 30 year old women who aren't rich and get breast cancer just die.

Coburn should talk to clients of big law firms to see how well the billable hour works.

As a client of big (and small) law firms, the billable hour works pretty well in areas when my company is a repeat player and information asymmetry is low. As a labor and employment attorney, I get excellent value from our regular L&E practitioners in my company's primary locations; not so much from attorneys in our smaller locations or areas where we don't have in-house specialists.

As a former law firm associate, the billable hour is terrible.

"As a former law firm associate, the billable hour is terrible."

On one matter last month I had 21 entries for less than four hours of work.

Well, Tom Coburn seems to be signalling to Obama and Sibelius that he wants to be the head of the IPAB which is chartered to change how Medicare pays doctors and hospitals.

I hope Coburn is being vetted for the IPAB and ideally appointed to give it the bipartisan endorsement it needs.

This is actually 100% correct.

It doesn't solve everything, but 30 minutes with a patient and you can reduce a lot of uncessary testing. And procedure. Not to mention errors.

This is silly. It is an idea that he must of thought of in the shower without having any thought as to implementation.

It could lead to more testing since what are they going to do for that extra time. If the additional procedures and test are free to doctor and patient while not provide the additional service. As far as errors, I went in for hernia surgery, and everyone asked me if I was on any medication. I brought in the bottle and told everyone the same answer, and they still go it wrong. Why are they writing down the answer ten places instead of one place that I can double check?

Lastly, I get a prescription that by law cannot get refills. I have to see my doctor every month. He check my blood pressure and fills the prescription. I usually have to wait a long time because some of the older patients take a long time, but some people go quickly. Can you imagine the queue if he couldn't customize the level of service?

The government makes me go to the doctor, and now I will have to stay there longer. I know that incentives work, but it is a professional relationship, and if I have control as a consumer, I will be able to determine the level of service works for me.

We tried converting to time rather than fee for service (ffs). Failed miserably. The amount of work accomplished dropped dramatically. I think a big part of the reason why it works at Mayo, is that it is an engrained cultural value. I am not sure what they do is exportable. We are trying it again, but with some productivity incentives. Working much better.

If you want to have more physicians in the hope of bringing down costs, you should be aware of supplier induced demand. Docs, especially specialists, are able to create a lot of their own demand. At some point, an increased number of MDs should lower their salaries, but until you reach the point, you are likely to see costs soar.

The use of extenders and mid levels is already increasing. It is a way to decrease costs in some situations. It does have problems. While I know several who are as good or better than many docs, I also know many who cannot function independently. Yet, they all have the same education and experience. How will we differentiate? The time and effort that will go into supervising, training and evaluating these people will be quite expensive. The way we have been doing it is having them work for us an extended period of time, usually at least 3-5 years, before we separate out whom can work more independently. Many have declined to work more independently when offered. We pay them more (you wont save quite as much as you might think), but the hours are an issue. Docs are not hourly workers. Many mid-levels build their lives around the idea of having regular hours. If a catastrophe happens and a doc ends up working until midnight, that is just part of the job. Not so much for most mid-levels.


Proposals by doctors to fix the system invariably involve securring higher or more secure payments for doctors.

Coburn is no different. Nor, is Paul Rand's proposal for medicare deductables which give docs undiscounted payments for first dollar coverage.

HMOs, ACOs, competitive bidding for capitated customers--that's the model. But, docs and hospitals don't like it. They like fee for service. Wonder why.

The obvious response to this would be to further delegation of patient care to lower cost nurses or physician's assistants. Just like the dentist -- you talk to the nurse for an hour, then the doctor shows up for five minutes before you leave.

Comments for this post are closed