The Greedy Doctor Problem
universalprior.substack.com
universalprior.substack.com
I have thought about this a lot in the context of real estate- up to recently, agents typically got a fixed commission (generally 3% each for buyer and seller in the places I’ve lived).
There is a moral hazard in that, assuming that the agent can impact my sales price favorably with effort on their part. If they’re helping me buy a house, and they get a percentage, then it’s in their interest for me to overpay. If I’m selling a house, then the margin for increasing the sale price above market value is probably not worth the effort ($300 for every $10k increase in sale price).
My solution (I have not tried it yet) is to structure a sales rep contract such that we agree on a target price for the standard commission, but increase the commission greatly for the sale price amount above the target price. Maybe 3% up to the sale price and 50% above that. That’s what I term “skin in the game”.
Assuming I have done a minimum of homework (looked for agent ratings and avoided bad or inexperienced agents; looked at comparables on Zillow to make sure the proposed sale price is market appropriate, etc.) then I should get a good outcome.
Doctors and state license boards often make it difficult to get informed consumer type information on doctors. Our insane medical system in the US almost never supports price transparency. So we have a long way to go there.
Note one other thing. I am trying to optimize for the best outcome I can assuming limited resources. If you have unlimited resources then you can hire and fire lots of doctors, overpay for performance, and generally do things that are unavailable to mere mortals. Also I am not assuming or eve trying to achieve perfect results (most beneficial result available at the lowest price point). I think that is a fool’s game.
You have to accurately put your bonus structure >= their normal compensation. That's really hard to do.
Wait. Surely for the realtor these two scenarios are identical? Every purchase is also a sale. The only difference is in what role you've cast yourself.
Outside of healthcare, for another example example, you don't need to understand engineering and tension dynamics in order to appreciate that the second floor of your home support you and your roof doesn't cave in on itself. I don't have examples on hand, but in medicine we've had cases where people do some logical variation of "the right thing for the wrong reasons." I.e., rituals that correlate with healthy outcomes because there is some not-yet-understood principle at play (i.e, you don't need to understand germ theory to benefit from cleanliness rituals).
I think this is one of those logical conundrums which falls into the trap of "in theory, in practice". The artificial constraints around the problem space result in artificial logical conundrums.
Notably: the prescription opioid epidemic is a great example of how this can go wrong in the opposite direction of healthcare providers valuing "relative" benefit vs "absolute" benefit.
My spouse is a primary care provider, and there was a period a couple decades ago where the prevailing wisdom was "if a patient says they are in pain, they are in pain, and you treat that pain. We are experts in medicine, but the patient is the expert of their own perception." This is still a complicated issue today, but there are clearly outcomes where we can make people "feel better" all the way to an early grave.
For the terminally ill, it seems absolutely appropriate to me to let the patient guide whether they wish to accelerate their death in exchange for quality of life. But "terminally ill" is often not such a black and white issue...we are all eventually mortal.
If things like heroin don't inspire a sense of horror or dread, then we just aren't speaking the same language. I, for one, don't want to live in a world where the human priority-at-large is everyone defining "living" as maximizing pleasure until their death. For the hedonists, I get why this makes a certain nihilistic sense, but I think the horror of the reality of it outweighs any momentary benefit.
In the very abstract, maybe it doesn't really matter one way or the other. In the concrete day-to-day reality of it, it's absolutely awful to see anyone struggling with any kind of addiction, out of control of their own lives...sometimes because they got on a treadmill-to-death on the _expert_ advice of someone trying to "help them out."
So a potential path will be that the incentives start to line up with measurable impact on symptoms short term (easy outcome to evaluate) instead of meaningful shift in root causes (difficult and/or slow to measure)
There is a story about doctors in ancient China who are paid by their clients only so long as the client is healthy. When they fall sick, the doctor is not paid again until the patient is healthy again. Then the goal is not a cure for a disease but the prevention of disease.
I don't know if any such thing ever really happened but it seems like it might be worth exploring even if only to illuminate the various possibilities.
It is easy to see an optimal strategy of never treating anything.
Some doctors are greedy. Some ________ are greedy. There are major factors in the waste in our health care system but by and large it isn't doctor's fees. There are so many rent seekers taking a cut.
There are other models than fee-for-service. They have tried population health models where the Provider (Dr, nurse, etc.) get $X per person per year. They are incentivized to be more preventative where they basically get to keep whatever they don't spend on patient care. However, you can lead a horse to water but can't make him drink. Ask yourself this: would you write web code where you got paid based on how many unique visitors viewed the page your wrote?
We can easily fix health care with a few simple changes but there is no real appetite to do so. One simple change is to go to referenced based pricing. You can't charge more than 1.2 the Medicare reimbursement. This change alone would reduce our spending by 25%.
Agree that this is huge. Price transparency, reference pricing. You know "the law of one price" is supposed to be a pillar of market economics, so this should be a no-brainer. Additionally, don't allow charging more than the insurance will pay with the exception of an agreed upon co-pay ahead of time. If a hospital accepts insurance, then it shouldn't be able to send you a bill later on that the insurance refuses to pay. That needs to be worked out between the hospital and the insurer. I would also like to see binding quotes before any procedure, with no surprise billing.
When asked for a choice by a medical professional (or a home improvement contractor for that matter), I will more often than not ask “if you were in my shoes, what would you do?” I don’t 100% of the time go with what they say, but I think it helps me understand their expertise and judgment better.
In the surgical case above, if the surgeon would have taken it out and given me a single recovery experience and told me about the bill when I woke up, I’m probably better off for it.
Point is, you need clarity of prices rather than the situation today where you sign a promise to pay whatever costs the hospital decides to charge you, and you may not even know what those are until weeks after the procedure. There is no meaningful way a market can operate under conditions in which blank checks are demanded in order to get anything done.
If that can't work, and the hospital insists it cannot quote you a cost at all, then go full socialized healthcare since obviously no market is possible.
The price leverage of the doctor is proportional to how much information the doctor has about how much money you have. If you can't keep it secret, you can add noise proportional to how much information the doctor likely has to cancel it out. The resulting "fair" price will be a function of the net shared information each side if the noise were removed.
Negotiations theory was a pet interest of mine, and I came to believe that in zero sum adversarial bargaining situations like that, the domainant position was the side with the most information, so the best strategy was to use noise to diminish their information, and then use whatever information you retained as arbitrage. If you think of noise in terms of signal entropy, and low entropy as signal, you can figure out whether their gambit is signal or noise by repsonding with noise and gauging the entropy of their response. This is crazy armchair stuff that nobody articultes, but variations of it gets used in bureaucracy and politics all the time, it's basically the game theory of gaslighting.
I'm not even going to read the rest of the article because the premise up front is so flawed. It assumes that doctors alone have say over pricing / compensation, and that's just not reality. It's much like engineers being blamed for things that rest entirely with management. It also makes the same flawed premise that I deal with in my profession every day - that I'm so greedy and so focused on my own pocketbook that I will screw over my clients. Are people motivated by money? Certainly. But a professional will put their client above money assuming no other perverse incentives.
After dealing with some surgeons I am not so sure about this anymore. They seem to feel very entitled to have big houses and nice cars.
1) The person doing the work often has limited say in the cost / compensation model unless they are also an owner.
2) Yes, there will always be people who prioritize money over the quality of their services. But reputation has a way of reducing their effectiveness.
This article is about game-theoretical-ideas of how to deal with the principal-agent problem. The "greedy doctor" idea is just a motivating example. You're seemingly taking offense at a knowingly fictional setup. That's akin to not wanting to learn how to solve physics problems about objects in a world where friction is assumed away, because that's not realistic. Umm, what?
Maybe this should be named "The greedy thesis advisor problem"?
Why? Because, generally speaking, the answer to this question is "it doesn't matter". A student will never get a job in academia without their thesis advisor's good word.
If a student is ready to leave and their thesis advisor isn't ready to let them go, their best option -- assuming they have at least a few publications (for CS) -- is to find an industry job and say "Sorry, my hands are tied on the start date and I need a better paying job for personal reasons. I really wish I could write a better thesis but I think there's enough work here at least to graduate."
(BTW: industry is much better than academia. The work is more interesting, the teaching/mentoring is more rewarding, and you'll make enough to retire in your 40s instead of desperately clinging to a job in the increasingly MBA-ified Higher Education Industry until your 80s. The "Academic Life" sucks pre-tenure anyways. If what you want is the so-called comfortable and care-free Full Professor days, just get financial independence in industry first and then go be a professor of practice or even ad junct somewhere for the healthcare and vacation money... which, aside from pay, are both actually low effort and care free!)
Their incentive then becomes to keep a good relationship with you so you continue to pay them. The best way to do that is to keep you healthy and happy.
I am sure if I asked my mates they would all disagree with me on this. Bad luck for the patients.
Having sued a doctor once, you might never be able to get medical treatment again.
No, I know they don't.
The idea of actual healthcare being dependent solely on optimization of greed is, of course, ludicrous.
Edit: The smarter person will charge you while you are healthy and not treat you when you need it, or underdeliver for the expectations.
(I think the actual answer to the question in the OP is, at least in part, to license more doctors. Then the cartel doesn't have as much pricing power and medical care is less of a luxury good. The rich person should really hire a lobbyist.)
https://www.beckersasc.com/asc-news/9-cms-pay-cut-for-physic...
The 80s/90s/00s were probably the golden years for them, but they got in the political spotlight, and combined with the decreasing proportion of young to old people, the government is going to push down cuts to whoever lacks political power to push back. Such as smaller businesses like independent doctor offices and pharmacies.
The government has already increased supply by letting Nurse Practitioners and Physician Assistants do stuff that doctors used to do, and I suspect a lot more price discrimination is on the way. Expect the average quality of care to fall (may not be a bad thing if overqualified people were treating pink eye), and paying more will result in being seen by someone more qualified.
If I pay doctors $400,000 a year to work 80 hours a week and then double the number of doctors and pay them $200,000 a year to work 40 hours per week, the quality of care would almost certainly improve but costs would be constant. And actually, do we pay doctors a premium for those overtime hours where they're actually probably less effective? If so, our costs would go down as quality of care improves.
The AMA does not want more competition and lower income per doctor.
Also, increasing class sizes is not without its down sides. Pharmacy schools cashed in and blew up the number of pharmacists in the last 10 years, and now the pharmacists’ employers have so much power in the negotiations due to so much supply of labor, that they can make pharmacists accept metrics which pharmacists know are excessive and unsafe since no one can possibly do the job properly in the time they are expected to check the medications and counsel people in.
A pharmacist family member says the law is to counsel patients, but the reality is anyone who did that would be fired and replaced so, in reality, people are not getting the counseling they deserve when they pick up the medications, and pharmacists are not able to properly double check the doctors.
> There are, however, two big problems with this approach. The first is that even though my life may be worth $10 million to me, most of the (huge) insurance premium to pay for this insurance would be wasted from my point of view -- there probably is not one person to whom I would want to give this much money when I die. The other problem is moral hazard -- heavy insurance may reduce my incentive to keep myself healthy, and may even create incentives for my relatives to try and hurt me. It might be a problem if I could only give my doctor a 50 percent interest in my life by taking away 50 percent of my own interest, or by giving a minus 50 percent interest to my relatives.
I get the super high-deductible catastrophic with HSA plan from my work, and then I pay $135/mo for my whole family for the DPC doctor (I could pay that $135 with my HSA and have it be on pretax money, but I'd rather invest my HSA pretax right now). The total there is less than any other insurance available to me. If I ever need anything, I just call or email my doc, most of the time we just do a (secure) video chat and I'm done and I don't need to make a trip in. If it is more severe, I can easily get a same-day appointment. He knows me, knows my history, and has an incentive to keep me healthy. The healthier I am, the more I pay him and don't see him.
I love the model!
Easy fix... "Just be wealthy!" lol... :|
I've paid for the absolute best insurance I could find this year and gone to doctors my entire life that simply don't care. I've had everything from colonoscopies to MRIs done and I haven't spent more than one day in my entire life in a hospital overnight in over 40 years.
Many would say I'm pretty healthy, but who knows?
It usually seems like doctors peddle as much fear as web MD now in order to drive me through the service bay of fees every time I have an ache or troublesome pain.
If it's any consolation though, my dentist, my auto maintenance shop, my lawyer, and even my local supermarket want a constant revenue stream out of me because they have a pipeline of service that they know they can get a customer hooked on if they coordinate their efforts and make their pricing variable based on being properly opportunistic.
A lot of the tactics the health care industry regularly uses to sell services (and drugs of course too) are from the nefarious book of street drug dealers... In a bad economy, the scams and opportunism are ripe. Getting a second opinion is often VERY EXPENSIVE as well.
Be careful my friends, sometimes it's better to face your fears of "WebMD prescribed" death head-on rather than to bankrupt yourself, because if you don't die, living broke or in extreme debt can be a "silent killer" and harmful mentally and physically to you too. Maybe marry a doctor... That might cut extreme health care costs... By about 5% if you're lucky... :P
Or have all doctors employed by the state on good fixed pay
I don't know if overpaying will work, because the marginal value of money goes down the more you have. Once your doctor is rich enough, he won't be overly concerned if you die and the checks stop coming in.
The real answer is the action-movie/not-that-uncommon-in-reality answer: Credibly promise to shoot the doctor if you die, or to shoot the doctor's family members.
I’ve found that my employer-provided health insurance is less than worthless for routine health care. I actually pay less by going to providers who don’t accept insurance — even though I maintain bare-minimum coverage and an HSA. I don’t think I used my health insurance at all this year — at least not enough to come close to meeting the deductible.
My primary care doctor uses this model — I pay $300/yr for basically unlimited access. I have a therapist who I pay $75 a session. I use GoodRX for prescriptions and get better prices than the post-deductible price on my insurance plan.
Basically, I pay less out of pocket by not using my insurance. I also get to choose my doctors, and they treat me as an actual patient rather than a number. Health insurance in the US is a scam.
1. The doctor's stake is pretty limited.
The old royal court doctor model is the perfect model for this. It's based on the doctor having very few (one?) patient. If the patient dies, not only do you risk your substantial income and high status, but you might even lose your life if the royal family gets suspicious enough.
It's a good solution if you're very rich. Much richer than the average doctor.
2. When you have more than one patient, at some point you probably earn more by treating less.
The less work you spend on each patient, the more patients you can take on. The more patients you can take on, the more you earn. If you simply stop treating patients - "you're perfectly healthy, don't worry" - you can basically take on an unlimited amount of patients and collect their fees until they perish. (This sounds like a caricature of the US insurance system, from my western EU perspective.)
---
What about this convoluted system:
- The doctors get paid a recurring fee that starts small and increases the longer you keep a patient around. This counters the problem that patients can just be replaced by a younger and healthier patient.
- The doctors get assigned patients randomly. This addresses the problem that doctors would choose young and healthy patients.
- The doctors can only have a certain number of patients at once. This makes each patient more valuable, and the doctor will be incentivized to actually treat patients, instead of having as many as possible.
Oh wait...
Now, the problem is that a young, ambitious, greedy doctor would be incentivized to cull their set of patients and select for the most viable ones in the long run. This actually incentivizes them to kill off patients with minor defects?
“I’m not going to kill you, but you would surprised how much suffering you can live through.”
They constantly try to escape
From the darkness outside and within
By dreaming of systems so perfect
that no one will need to be good.
But the man that is will shadow
The man that pretends to be.
--TS EliotIt's a life-time contract for the doctor, which aligns the incentives... Provided being in the highly quantized state of "alive" is all we care about, i think this is pretty bulletproof. The only place it will fall down is in the moral grey areas where euthanasia usually enters the ring, in which case a more specific criteria of alive may be necessary to prevent prolonged suffering.
However If this is applied to the multidimensional continuum of "health" it gets tricky again. Verifying you are "alive" can be done by your dog; verifying you are "healthy" would require at least as much independent expertise as the greedy doctor, and is an opinionated subject.
Interestingly though, this is kinda, roughly, sortof how public healthcare is supposed to work: everyone who is alive and in working health is capable of contributing economically and paying taxes... which fund the healthcare system, but since it's so indirect this alignment in incentives can only be appreciated at the organisational level trying to increase efficiency. e.g in theory it should encourage preventative healthcare which minimises long term cost while keeping their income stream alive, literally.
This has been practiced by patients for a long time by asking for a second opinion. However, this no longer works, because doctors now all share your medical record, and the second doctor risks getting sued if they disagree, while they bear no risk at all for agreement with the first doctor.
So, the consensus rules have to actively break the model for collusion. I’m not sure how to do this, and actually I think this is a relatively unexamined area of game theory, because most models assume tautological correctness of consensus.
Breaking consensus requires an oracle, which is by definition, a source of truth that operates outside of the consensus model, and cannot be affected by participants. This would be something like a piece of diagnostic equipment that can’t be altered or censored. It also requires a very large incentive for correctly predicting the oracle result against the consensus opinion.
So here is a straw man mechanism: 1. Assume condition can be assessed by a diagnostic test oracle 2. See many doctors, solicit opinions on treatment 3. For each proposed treatment, the doctors bet on the post-treatment test result 4. Choose the treatment with the best most money wagered on a good result 5. Add your service payment to the pool 6. Undergo treatment, take post-treatment diagnostic test 7. Distribute all money to the correct predictors
Draw up a contract whereby you are only obligated to pay the doctor while healthy. The doctor loses money as long as your health is suboptimal and therefore it is in the greedy doctor's best interest to do everything they can to keep you healthy.
Also, what happens when people self-select such that only people who often get sick take on these contracts, and people who are normally healthy stick with pay-for-service?
Only if they cannot help you. In which case you have the same medical outcome but save your cash.
Doctor locks up money under contractual obligation. If they act dishonestly and your health declines, they lose their stake.
As long as they act honestly, they are rewarded.
The same approach can be applied to any service that requires regular intervention to avoid service interruption. If you pay someone every time the service is interrupted, then they’ll be incentivized to do a bad job so it gets interrupted more frequently (e.g. contractors tasked with ensuring traffic lights or electricity supply works all year round in all weather conditions).
Pay a flat rate for continual service. Beyond a certain reasonable/acceptable level of interruption, reduce what you pay them. Eventually if they do enough poor quality work they’ll go out of business - as opposed to the current conventional approach that incentivizes them to do poor quality work.
If this applied to certain goods as well (like electronics) we’d probably be able to get rid of planned obsolescence.
I suppose another such solution, for a patient with a given medical condition X, is to have a very large pool of doctors and see if any of them have condition X (with similar age, background, etc).
Then one could compare how treatment varies when doctors treat themselves with the same issue.
"Physician, heal thyself."
P != NP, as far as we've discovered. As a non-doctor, I can still understand medical conditions. What I don't have are the heuristics to diagnose a medical condition. The way I see it, I'm hiring a doctor (or any other skilled knowledge professional) for their ability to generate a proof certificate (in the algorithms sense) that convinces me of their declared diagnosis. Any consultant who asserts "The answer is X" without being able to explain that answer to an intelligent non-domain-expert has not done their job.
What you need is someone that tells you what do (or most likely not to do) before you get sick.
Herein lies the very big problem, if you don't get sick doctors have no job!
Or if a particular emergency department physicians keep ordering unnecessary and expensive diagnostic tests, not only may those tests not be reimbursed, but their "score" for quality of care goes down, and so will their personal reimbursement.
They charge the patients the SAME, since, to the eyes of insurance, the services supplied are basically the same. They just get to KEEP more of the difference for the hospital stakeholders with a PA than with someone with a medical degree.
But I think they are abused during their residency and 4 - 5 years of being treated like absolute shit and everyone looking at you like a spoiled idiot (but is actually poor since residency salaries basically cover only housing and food) can turn most futures doctors into jaded and disillusioned cogs for the insurance machine.
Hospitals want to get paid by the insurance companies... so will bend over backwards and make it happen to the employees detriment.
Insurance companies PWN us all. So do not get distracted and think bigger picture. Doctors actually don't know the price of anything, they just do their job, and are getting trickle down from the insurance companies... which the hospitals gut through first.
Yeah, doctors had to work to get paid that much, but its the insurance and hospital business aspect that makes the costs sky high. They have lobbied for it to be this way for decades, letting idiot doctors distract us and have become an easy scapegoat. Most doctors have no idea how they get paid, the hospital complicates it on purpose so that idiots on both the doctor and patient sides are clueless and blame anything but the accurate picture.
This is cheaper, and produces better outcomes than the alternative.
That the Doctor is greedy is in the premise, but why would anyone settle for a greedy doctor at all?
Doctors did due diligence on their clients and advised them preventive. They did not want sickly households.
Also the incentive is on keeping the whole house in good shape.
I learned the opposite once: never let a capitalist near your health care program. They make money when you are sick, and stop making money when you are healthy.
Wouldn't the same rule apply for IT Sysadmins?
Is there even a theoretical possibility that this could be done other than after the fact?
I’m starting to think most readers on HN don’t really read the articles and go straight into the comment section..
Edit: Part of the problem is that the analogy is such an amateur caricature as to be offensive. I am not a doctor.
What makes it more poignant is that I typically go straight to the comments myself, assuming that intelligent commentary on an article will convey what the article was about faster than the article itself. Which would probably be a good assumption - if I was the only member of the commentariate to make it.
[1]http://harmful.cat-v.org/people/basic-laws-of-human-stupidit...
To use a specific example there was once a claim about a board seat for Tesla going to someone that people tend to dislike. The comments were largely in agreement that this would happen and it would happen basically because it was outraging and seemed evil which agreed with the commenters preconceptions. Anyone who posted contrary to this - which I did, quoting a primary source which disagreed with the claim, got downvoted. Ultimately I was right and the board seat didn't go to the 'evil' person.
This isn't intended to be a critique of me getting downvoted. Instead I'm trying to point out that if you read all the comments the consensus in the comments doesn't approximate the correct answer. So what do you have to do? You have to go primary sources in order to be able to get an unbiased estimator. Maybe the comments link to one, which is nice, but notice: you can't rely on the comments alone. Which means it was the people who read the primary sources, not the people who read the comments, which get the best outcome individually.
Unfortunately, the people who are doing this aren't becoming popular for doing this when it helps them. Quite the opposite. When this is effective, it is effective entirely because it is not in line with the popular opinion.
It's a particularly low-value form of comment, and also very common, I think mostly because we try to secure our identity in opposition to others. Since this mechanism appears to work the same way in everyone, such comments are repetitive and extremely tedious.
https://abcnews.go.com/Health/whistle-blower-helped-expose-m...
What a nightmare for the people involved.
Such massive generic topics are like black holes that suck in unsuspecting threads that fly too close. Solution: consciously steer clear.
https://hn.algolia.com/?dateRange=all&page=0&prefix=true&que...
Maybe this is the kind of market that can never be free, because you can never step away from a deal, as the alternative is often death. As such perhaps in these matters we should just have some faith in humanity. You know, it works for a lot of open source software. Capitalism is not a silver bullet, people have a consciousness for a reason, use it, have faith in it. You will see that that is the exact thing that will make it flourish.
Btw, all "solutions" would piss me off as a doctor that just wants to help. The vast majority of people deserve and want to be trusted. We work better that way. What a cold piece.
Add on top that by being independent, doctors will have their own company and be subjected to a much lower tax rate than a "normal" person earning 400k$/year.
Less overwork results in better care, and more jobs. I'm all for it.
Yes, in theory you could "just add more doctors". But there is clearly already less doctors than there needs to be. Policy that encourages doctors to work less exacerbates this issue.
You can't just say "add more doctors" as if that helps anything. Shortages exist for a reason, and analyzing and fixing the system which creates the shortage is the only way to resolve the shortage. You can't just tell poor people to "make more money" and solve poverty in the blink of an eye.
Why wouldn't it?
There are more than enough people who would want to become a doctor, and would be capable. Especially if the work/life balance would be better. At least in my part of the world, it seems that the system is rigged to keep the number of docters artificially low.
I explained in my comment: There is a shortage of doctors. Unless you know 50,000 people with medical degrees searching for a job, you are going to have to fix the situation that created the shortage rather than just wishing more doctors would appear.
This is what dasudasu was doing. They identified a cause of the shortage (doctors are encouraged to work 60% of a work week) and your response was "so what, more doctors will just appear". More doctors have not been appearing.
We detached this subthread from https://news.ycombinator.com/item?id=29270645.