The New Yorker‘s Atul Gawande on “What a Texas town can teach us about health care“: McAllen is in Hidalgo County, which has the lowest household income in the country … McAllen has another distinction, too: it is one of the most expensive health-care markets in the country. Only Miami—which has much higher labor and living costs—spends more per person on health care. In 2006, Medicare spent fifteen thousand dollars per enrollee here, almost twice the national average. The income per capita is twelve thousand dollars. …
Conceivably to:Treatments that aren't subject to medicare/medicaid price limitations.Treating people who wouldn't have gotten treated before. (Say, cases w/ only small benefits.)Treatments that didn't exist before.
Robin – your blog isn't the venue to determine the political palatableness of a reasonable sideways policy-tug.Josh – yes, yes, your criticism and your proposals are all well and good, but they're as a improbable as Robin's innocuous suggestion. And just because something is illegal, doesn't mean it doesn't frequently occur, more so for illegals acts that are hard to identify and lightly punished.
Note that I said “compensation for individual physician services diminishes every year…”
And yet the price of medicine is going up. Is it new technology? an aging population? drugs? Maybe we're not spending enough on services. I happen to think primary care is especially neglected. But something has to give, since we can't spend 120% of GDP on health care.
Note that I said "compensation for individual physician services diminishes every year...". I'm referring to the government-mediated agencies, Medicare and Medicaid, and their price codes for specific services. For example, I was reimbursed $2600 in 1979 for a certain surgical procedure that I performed, in 1979 dollars. When I retired in 2005, I was being reimbursed $700 for the same procedure, this time in devalued 2005 dollars. The difference in purchasing power is about 8-fold. Note that I could not collect any more than this Medicare allowance to make up the difference, even if the patient volunteered. As Casey Stengel often said: "You can look it up."
Compensation for individual physician services diminishes every year and with each “reform”, perhaps a factor encouraging unnecessary interventions . . . healthcare costs are out of control . . .
I keep hearing stuff like this. Where is all that money going?
Employers can and do consider employee health when hiring, and deciding what wage to offer. They also can and do encourage unhealthy employees to leave.
Are you sure? I'm almost positive both of those actions are illegal. Anyway, a better response than the one I had earlier is that employer-based insurance isn't ideal either.
Since governments are almost all geographically defined, I see no other way to decentralize via government than geographically.
You could pool people together randomly and then have them vote, via proxy if necessary, on a menu of policy choices. Or you could vote on a metric for well-being, propose a few different plans, and then set up a betting market to decide which plans were likely to lead to the highest well-being. The highest-ranking plans would be then administered for several of the random pools. There are lots of other ways to do it.
Discussions of healthcare reform infrequently look at the question of who will provide the care. Robin Hanson has, in several earlier posts, pointed out that allied-health entities (physician assistants, nurses, midwives, etc.) deliver equal or superior care to that of physicians. These studies involve mostly uncomplicated health issues. For more serious problems, one prefers an especially intelligent, especially well-trained, especially skilled attendant. What will attract such individuals to the field of medicine? One powerful intangible attractant is respect for the occupation, which is rapidly disappearing, and for disciples of Hanson, has vanished. Money comes to mind, yet most participants in this blog, and indeed the general public, feel that doctors are overpaid. Compensation for individual physician services diminishes every year and with each "reform", perhaps a factor encouraging unnecessary interventions. In my college class, those who completed accounting and began practice as a CPA at age 22, working 40-hour weeks, out-earned their classmates who opted to be surgeons until the surgeon reached about age 55, working 60-80-hour weeks, including nights, weekends, and holidays. What would encourage someone with the requisite potential to see the choice of medicine as having the greatest utility? I readily understand that healthcare costs are out of control. At the same time, experienced doctors are dropping out at unprecedented rates, and about half of all US medical school applicants are accepted, down from about 1-in-20 in my day. I'm out of the game as a provider, but I dread to think that as a consumer, my surgery will be performed by the low bidder.
El Paso now spends half of what McAllen spends without running afoul of federal rules on what is acceptable coverage. So clearly there is a lot of scope for local feedback to reduce spending even with current levels of spending discretion.
Josh, yes of course there is a tradeoff between discretion and central rules. Employers can and do consider employee health when hiring, and deciding what wage to offer. They also can and do encourage unhealthy employees to leave. Since governments are almost all geographically defined, I see no other way to decentralize via government than geographically. There is the fully private option, but this was intended as a political compromise, and that private option does not seem an acceptable compromise to most. Yes of course this proposal isn't fully fleshed out. If someone were interested in doing so, I'd be happy to help.
Josh, wide discretion only conflicts fatally with an expansive concept of universal coverage, one that knows exactly who should be treated when and how.
That's no discretion! You can go from there to wider and wider discretion, until municipalities have the discretion to offer no coverage at all. My point is that there is a tradeoff: the more discretion, the less uniform the coverage. Universal coverage implies a certain degree of uniformity beyond simply paying for insurance.
Mobility barriers based on preexisting condition surcharges should be no higher for municipalities than for employers; the same barriers exist in both cases.
That's a good point. Are there preexisting condition surcharges in the group market, though? To be honest, I don't know what regulations there are on employer-based insurance, but I know the market for it is very different from individual insurance. Is the group market required to provide maternity care, for example? I don't think employers can charge new employees more, but there is usually a fixed enrollment period.
Even if the regulations on town-based insurance were the same as employer-based insurance, there would be other problems relating to pooling risk by geography. For example, it would create incentives for rent-seeking: towns could try to compel their unhealthiest members to move away.
Medicare payments are already doled out far from fairly due to local practice variation; why is a federal payment to municipalities based on demographics less fair?
Because, as far as I know, people still have the option of leaving their town without penalty, whether temporarily or permanently, if they wish to escape local practices. In addition to the unfairness, the federal funding ensures a large degree of centralization: the government will have to decide, for example, whether an aging community will receive more funding than a young, healthy one.
And why point out that government is important; my proposal puts the entire industry completely in government hands; why isn’t that enough?
You were saying maybe the debate should be less about private vs. government control. But I think that part of the debate is very important and probably isn't overemphasized.
I thought your idea was interesting, and I gave them some thought. Sorry if that didn't come through! Geography just strikes me as a poor way to decentralize things for the reasons I've already mentioned. Decentralization itself could create room for some compromise, though I don't think you've really specified what that compromise would look like. The compromise is all about how much coverage the decentralized pools would have to provide, not just whether they would have to provide coverage; their discretion to reject new applicants or charge for preexisting conditions; their source of funding; and whether private insurance would be illegal.
Josh, wide discretion only conflicts fatally with an expansive concept of universal coverage, one that knows exactly who should be treated when and how. Mobility barriers based on pre-existing condition surcharges should be no higher for municipalities than for employers; the same barriers exist in both cases. Medicare payments are already doled out far from fairly due to local practice variation; why is a federal payment to municipalities based on demographics less fair? And why point out that government is important; my proposal puts the entire industry completely in government hands; why isn't that enough?
We could require municipalities to give universal coverage, and require national standards so businesses could avoid differing local regulations.
What businesses?
But we should otherwise give municipalities wide discretion on cost control measures.
The game is entirely what those "national standards" would be. Without strong enough regulations, the concept of universal coverage could be made meaningless by high deductibles or poor coverage. As regulation became stronger, there would be little room for municipalities to go above and beyond, which would make the system very uniform. So allowing wide discretion on cost control measures conflicts with the mandate for universal coverage mentioned earlier.
Municipalities should be empowered to charge new residents higher prices, just as an insurance companies can charge for pre-existing conditions. (Perhaps the districts folks leave should pay districts they go to.)
That would create some barriers to labor mobility; the severity of those barriers would increase with the discretion to charge higher prices to new residents. And again, there is a trade-off: reducing that discretion would lead to people moving to towns with better insurance if they got sick.
While its no one’s first choice, my libertarian friends prefer this proposal to the status quo. Progressives should like its universal coverage, full government control, and strengthened local communities ties.
That's because the most important details are absent. Progressives are most likely imagining strong regulation to ensure universality of coverage. Libertarians are most likely imagining weak regulation. We don't even know if private insurance would be outlawed.
How about equality? Poor individuals tend to cluster together, and small municipalities would be less able to bear the overhead required. Federal funding could help to level the playing field, but we all know it wouldn't be doled out fairly. Geography is a weird way of clumping people to share this particular risk.
Maybe the medical reform debate should be less about government versus private control, and more about centralization versus decentralization.
A certain amount of decentralization could help, but I don't think we should decentralize using geography. Maybe there is another way?
Also, government control is important. Private health insurance companies can take advantage of the bounded rationality of their counterparties and imperfect competition arising from their large provider networks, which can lead to excessively high profits and administrative costs. You could tax them, but what would you do with the money? If you give it to lower-income individuals, you're just subsidizing them again. The only way to overcome this problem is to roll up your sleeves and start a competing company for the benefit of patients and doctors. This could potentially be done in the market: doctors and patients could work out a scheme wherein they share ownership of an insurance company. But if doctors and patients are going to send zillions of proxies to negotiate complicated contracts anyway, why not have the public start it and use government as the forum for those negotiations? That's what the government is there for.
It seems that the critical factor is to have comparisons and an incentive to experiment and learn from others, regardless of how this is achieved. So the main conclusion is that you need several sources of funding that compete with each other. Not too many so that dealing with the multitude of payers causes too much hassle, but more more than a handful.
But if you base this on municipality you create an incentive for municipalities to lower health care costs by handing chronically ill people a one-way ticket to a nice sunny place.
Robin, several posters mentioned that at a municipal level, the administrative overhead and the statistics don't seem to work out. However, maybe at the state level, although the fine-grained feedback on local variation may get blunted. However -- Gwande's article really seems to me to lay blame on local cultural problems rather than a traditional economic incentive problem per say.
On the other hand the blogosphere problem isn't the best place to craft compromise -- maybe having a direct question for discussion at the end of the post would help. Especially on controversial topics. Volokh posters often try and limit the scope of discussion/comments.
What is the timescale of poor health care? What is the timescale of local business growth? If they don't match with the timescale of a politician in power, we can't expect to match incentives.
Good article.
Conceivably to:Treatments that aren't subject to medicare/medicaid price limitations.Treating people who wouldn't have gotten treated before. (Say, cases w/ only small benefits.)Treatments that didn't exist before.
Robin – your blog isn't the venue to determine the political palatableness of a reasonable sideways policy-tug.Josh – yes, yes, your criticism and your proposals are all well and good, but they're as a improbable as Robin's innocuous suggestion. And just because something is illegal, doesn't mean it doesn't frequently occur, more so for illegals acts that are hard to identify and lightly punished.
Note that I said “compensation for individual physician services diminishes every year…”
And yet the price of medicine is going up. Is it new technology? an aging population? drugs? Maybe we're not spending enough on services. I happen to think primary care is especially neglected. But something has to give, since we can't spend 120% of GDP on health care.
@Josh:
Note that I said "compensation for individual physician services diminishes every year...". I'm referring to the government-mediated agencies, Medicare and Medicaid, and their price codes for specific services. For example, I was reimbursed $2600 in 1979 for a certain surgical procedure that I performed, in 1979 dollars. When I retired in 2005, I was being reimbursed $700 for the same procedure, this time in devalued 2005 dollars. The difference in purchasing power is about 8-fold. Note that I could not collect any more than this Medicare allowance to make up the difference, even if the patient volunteered. As Casey Stengel often said: "You can look it up."
Compensation for individual physician services diminishes every year and with each “reform”, perhaps a factor encouraging unnecessary interventions . . . healthcare costs are out of control . . .
I keep hearing stuff like this. Where is all that money going?
Employers can and do consider employee health when hiring, and deciding what wage to offer. They also can and do encourage unhealthy employees to leave.
Are you sure? I'm almost positive both of those actions are illegal. Anyway, a better response than the one I had earlier is that employer-based insurance isn't ideal either.
Since governments are almost all geographically defined, I see no other way to decentralize via government than geographically.
You could pool people together randomly and then have them vote, via proxy if necessary, on a menu of policy choices. Or you could vote on a metric for well-being, propose a few different plans, and then set up a betting market to decide which plans were likely to lead to the highest well-being. The highest-ranking plans would be then administered for several of the random pools. There are lots of other ways to do it.
Discussions of healthcare reform infrequently look at the question of who will provide the care. Robin Hanson has, in several earlier posts, pointed out that allied-health entities (physician assistants, nurses, midwives, etc.) deliver equal or superior care to that of physicians. These studies involve mostly uncomplicated health issues. For more serious problems, one prefers an especially intelligent, especially well-trained, especially skilled attendant. What will attract such individuals to the field of medicine? One powerful intangible attractant is respect for the occupation, which is rapidly disappearing, and for disciples of Hanson, has vanished. Money comes to mind, yet most participants in this blog, and indeed the general public, feel that doctors are overpaid. Compensation for individual physician services diminishes every year and with each "reform", perhaps a factor encouraging unnecessary interventions. In my college class, those who completed accounting and began practice as a CPA at age 22, working 40-hour weeks, out-earned their classmates who opted to be surgeons until the surgeon reached about age 55, working 60-80-hour weeks, including nights, weekends, and holidays. What would encourage someone with the requisite potential to see the choice of medicine as having the greatest utility? I readily understand that healthcare costs are out of control. At the same time, experienced doctors are dropping out at unprecedented rates, and about half of all US medical school applicants are accepted, down from about 1-in-20 in my day. I'm out of the game as a provider, but I dread to think that as a consumer, my surgery will be performed by the low bidder.
[Note our new threaded comment feature!]
El Paso now spends half of what McAllen spends without running afoul of federal rules on what is acceptable coverage. So clearly there is a lot of scope for local feedback to reduce spending even with current levels of spending discretion.
Josh, yes of course there is a tradeoff between discretion and central rules. Employers can and do consider employee health when hiring, and deciding what wage to offer. They also can and do encourage unhealthy employees to leave. Since governments are almost all geographically defined, I see no other way to decentralize via government than geographically. There is the fully private option, but this was intended as a political compromise, and that private option does not seem an acceptable compromise to most. Yes of course this proposal isn't fully fleshed out. If someone were interested in doing so, I'd be happy to help.
Josh, wide discretion only conflicts fatally with an expansive concept of universal coverage, one that knows exactly who should be treated when and how.
That's no discretion! You can go from there to wider and wider discretion, until municipalities have the discretion to offer no coverage at all. My point is that there is a tradeoff: the more discretion, the less uniform the coverage. Universal coverage implies a certain degree of uniformity beyond simply paying for insurance.
Mobility barriers based on preexisting condition surcharges should be no higher for municipalities than for employers; the same barriers exist in both cases.
That's a good point. Are there preexisting condition surcharges in the group market, though? To be honest, I don't know what regulations there are on employer-based insurance, but I know the market for it is very different from individual insurance. Is the group market required to provide maternity care, for example? I don't think employers can charge new employees more, but there is usually a fixed enrollment period.
Even if the regulations on town-based insurance were the same as employer-based insurance, there would be other problems relating to pooling risk by geography. For example, it would create incentives for rent-seeking: towns could try to compel their unhealthiest members to move away.
Medicare payments are already doled out far from fairly due to local practice variation; why is a federal payment to municipalities based on demographics less fair?
Because, as far as I know, people still have the option of leaving their town without penalty, whether temporarily or permanently, if they wish to escape local practices. In addition to the unfairness, the federal funding ensures a large degree of centralization: the government will have to decide, for example, whether an aging community will receive more funding than a young, healthy one.
And why point out that government is important; my proposal puts the entire industry completely in government hands; why isn’t that enough?
You were saying maybe the debate should be less about private vs. government control. But I think that part of the debate is very important and probably isn't overemphasized.
I thought your idea was interesting, and I gave them some thought. Sorry if that didn't come through! Geography just strikes me as a poor way to decentralize things for the reasons I've already mentioned. Decentralization itself could create room for some compromise, though I don't think you've really specified what that compromise would look like. The compromise is all about how much coverage the decentralized pools would have to provide, not just whether they would have to provide coverage; their discretion to reject new applicants or charge for preexisting conditions; their source of funding; and whether private insurance would be illegal.
Josh, wide discretion only conflicts fatally with an expansive concept of universal coverage, one that knows exactly who should be treated when and how. Mobility barriers based on pre-existing condition surcharges should be no higher for municipalities than for employers; the same barriers exist in both cases. Medicare payments are already doled out far from fairly due to local practice variation; why is a federal payment to municipalities based on demographics less fair? And why point out that government is important; my proposal puts the entire industry completely in government hands; why isn't that enough?
We could require municipalities to give universal coverage, and require national standards so businesses could avoid differing local regulations.
What businesses?
But we should otherwise give municipalities wide discretion on cost control measures.
The game is entirely what those "national standards" would be. Without strong enough regulations, the concept of universal coverage could be made meaningless by high deductibles or poor coverage. As regulation became stronger, there would be little room for municipalities to go above and beyond, which would make the system very uniform. So allowing wide discretion on cost control measures conflicts with the mandate for universal coverage mentioned earlier.
Municipalities should be empowered to charge new residents higher prices, just as an insurance companies can charge for pre-existing conditions. (Perhaps the districts folks leave should pay districts they go to.)
That would create some barriers to labor mobility; the severity of those barriers would increase with the discretion to charge higher prices to new residents. And again, there is a trade-off: reducing that discretion would lead to people moving to towns with better insurance if they got sick.
While its no one’s first choice, my libertarian friends prefer this proposal to the status quo. Progressives should like its universal coverage, full government control, and strengthened local communities ties.
That's because the most important details are absent. Progressives are most likely imagining strong regulation to ensure universality of coverage. Libertarians are most likely imagining weak regulation. We don't even know if private insurance would be outlawed.
How about equality? Poor individuals tend to cluster together, and small municipalities would be less able to bear the overhead required. Federal funding could help to level the playing field, but we all know it wouldn't be doled out fairly. Geography is a weird way of clumping people to share this particular risk.
Maybe the medical reform debate should be less about government versus private control, and more about centralization versus decentralization.
A certain amount of decentralization could help, but I don't think we should decentralize using geography. Maybe there is another way?
Also, government control is important. Private health insurance companies can take advantage of the bounded rationality of their counterparties and imperfect competition arising from their large provider networks, which can lead to excessively high profits and administrative costs. You could tax them, but what would you do with the money? If you give it to lower-income individuals, you're just subsidizing them again. The only way to overcome this problem is to roll up your sleeves and start a competing company for the benefit of patients and doctors. This could potentially be done in the market: doctors and patients could work out a scheme wherein they share ownership of an insurance company. But if doctors and patients are going to send zillions of proxies to negotiate complicated contracts anyway, why not have the public start it and use government as the forum for those negotiations? That's what the government is there for.
It seems that the critical factor is to have comparisons and an incentive to experiment and learn from others, regardless of how this is achieved. So the main conclusion is that you need several sources of funding that compete with each other. Not too many so that dealing with the multitude of payers causes too much hassle, but more more than a handful.
But if you base this on municipality you create an incentive for municipalities to lower health care costs by handing chronically ill people a one-way ticket to a nice sunny place.
Robin, several posters mentioned that at a municipal level, the administrative overhead and the statistics don't seem to work out. However, maybe at the state level, although the fine-grained feedback on local variation may get blunted. However -- Gwande's article really seems to me to lay blame on local cultural problems rather than a traditional economic incentive problem per say.
On the other hand the blogosphere problem isn't the best place to craft compromise -- maybe having a direct question for discussion at the end of the post would help. Especially on controversial topics. Volokh posters often try and limit the scope of discussion/comments.
What is the timescale of poor health care? What is the timescale of local business growth? If they don't match with the timescale of a politician in power, we can't expect to match incentives.