« | Main | Book Review: Three Empires on the Nile »

June 26, 2009



I don't regard Krugman to be intellectually honest in political matters, so basically I ignore his other qualifications.


That's not true. I should say rather that Krugman, when discussing politics, makes no attempts to distinguish between evidence-backed theory and personal conjecture, so that one shouldn't take anything he says as necessarily a product of intellectually honest contemplation by an expert.


Nathan's criticism is valid, but I'd like to touch on the critique against a public health option for the moment.

If the government would not as competently provide health insurance as the private sector, in what way would that be manifest? In order for a private insurance company to be successful to investors, it has to not only accurately assess the risk involved in insuring people, but it also has to slyly deny claims. The government option would "fail" on both of those counts to the benefit of people who are high risk and who actually need treatment. And isn't that the whole point of the government option, to ensure that even high risk people can get the treatment they need?


The government option would absolutely have to ration care, though they are not eager to say this. Some of the rationing might simply happen because the wait for treatment would be so long that people will die waiting. CERTAINLY, the elderly will suffer the most because they would be at the bottom of the priority list. Since the government option would involve taxpayer subsidy and could thus undercut private insurance and eventually drive them out of business, I fear that in the end the government option would be the only option. Since all of us, if we are lucky, will at some point in our lives be elderly, I think we should all be afraid. Be very, very afraid.

Nathan Smith

Tom is touching on a reason that the government plan might *look* worse, not because of inferior efficiency, but because private insurance companies would find ways of denying claims, and clients, leaving the public sector to take the worst cases. In that case the public option could become a dumping ground for anyone who couldn't get private insurance-- a different kind of outcome which might be good in some ways, although expensive to the taxpayer. If that happened, the government program might look much less efficient than private programs by most statistical measures you could construct, but that would be misleading. Of course, this isn't the outcome private health insurance companies are most worried about.

It's almost inconceivable that "the elderly... will be at the bottom of the priority list," isn't it? The elderly are always at the top of politicians' priority list. Right now, and for many years past, the elderly have been getting big, unsustainable handouts through Medicare and Social Security at everyone else's expense. Also, don't we need to introduce rationing of some kind, somehow, in order to constrain costs?

Not exactly, I guess. We need to introduce rationing to the extent that we continue the third-party payer system. The key is to displace the third-party payer system so that people have to pay most medical expenses directly out of pocket, with insurance covering only catastrophic care. Even then, it might be nice if huge medical expenses were paid for by automatic long-term loans rather than by simple third-party payer setups, to the extent possible. People should have to internalize the costs of medical care, and then the perverse dynamics of health care cost growth will be reined in.


It is true that the current elderly have done well with medicare and social security. That simply can't continue with baby boomers now attaining senior status. Of course, I know that elderly people tend to go and vote, so that gives them more power than those who don't vote, but nevertheless, the simple numbers will tell us that boomers cannot get the benefits that their parents have enjoyed. We all know that rationing would be necessary under a government plan and since elderly people need a lot more medical care than other people, they will sacrifice more than others. And maybe they should. Truthfully, rationing is with us now and will continue to be--it's just that I don't think I want the government making decisions about who lives and who dies. And I think the lack of competition would be a bad thing, though I agree that care needs to be closer to the consumer.


"I don't think I want the government making decisions about who lives and who dies."

Unfortunately, the government may be the least bad option, since market actor behavior is notoriously irrational in health matters. I've never seen an analysis that lead to any believable solution except government rationing, and those were, while sometimes believable, nevertheless deeply worrisome. Maybe some combination of government regulation and private insurance would work, but this seems more likely one more way to avoid rationing by creating large indirect taxes. All so we can spend a fortune on palliative care for the terminally ill.


Of course we all agree that terminally ill people should not have fortunes spent on them--but we don't always know when people ARE terminally ill. In addition, what about operations that significantly improve a persons life, like hip replacement? At what age should these be denied? Bascially, decisions such as these need to be made on a person by person basis with the involvement of the patient and family and doctors. In other words, the more local the better. We'd all like some national fix-it-all solution for health care, but in fact, solutions need to happen locally, in famlies and communities, where decisions can be made humanely and personally instead of bureaucratically. My point is that a national plan with national policies and rules is bound to be brutal and hamfisted. Every health decision comes down to an individual. I think we can save money by putting care closer to the patient and giving incentive to practice preventive care and avoid going to the doctor for trivial reasons, but any solution should allow health care to be mostly controlled locally.


"but any solution should allow health care to be mostly controlled locally."

A single-payer system would not preempt that.

Joyless Moralist

It's hard to be sure, but I'm also not sure that the elderly would be the biggest losers from nationalized health care, except insofar as they tend to need the most care, and thus would be most affected by declines in the overall quality of the health care system in America. I don't know how likely it is that age would become a major factor in determining who gets to be treated. What I suspect is more likely is that major procedures of all kinds (ones that are expensive and that have lower chances of success) would be cut back, along with treatments for non-life-threatening conditions that can be allowed to fester for awhile. Health care would tend to be more focused on putting out fires, with uncomfortable-but-not-serious conditions being neglected.

And the all-around quality of the system would diminish. Medicine would become a less prestigious career, hospitals would have less state-of-the-art equipment, and you'd have a lot more stories of people being put off because their condition didn't seem serious, until a major organ ruptured, or of radical misdiagnoses by doctors who didn't want to spend the money on extra tests. That sort of thing.

Just as with higher education, you have to put a lot of money into medicine in order to have high quality institutions. The government is not going to put as much money into medicine as would come from a privatized system, and lack of competition tends to breed inefficiency. So the overall quality of care will decrease for everyone, though there would probably be less inequalities in how the pain is distributed.

Nathan Smith

Tom says: "A single-payer system would not preempt [local control of health care solutions]." Depends on who the single payer is. If it's the patient, no it wouldn't. If it's the federal government, it would, no? You can't say the solution is local if it's financed from Washington, D.C.

Anyway, ms's recommendation that "solutions need to happen locally, in famlies and communities, where decisions can be made humanely and personally," begs the all-important question of financing. Would solutions also be *financed* locally, i.e., not by federal government programs such as Medicare, and not by insurance companies with national scope? But the problem is that patients often don't have enough money to pay for operations themselves. So what are we really saying here? Wealthy or middle-class individuals, of course, might be able to hoard enough money to cover even catastrophic illnesses and disabilities, or they might be able to borrow enough. Maybe "local control" could mean that "solutions will be made humanely and personally" for the rich, the thrifty and those with good credit scores, while those with less earning capacity or those who have ended up with bad credit scores one way or another will... um... just have to die?

Once you introduce insurance, well, insurers like to diversify their risks and defray their overhead through economies of scale, so it's unlikely that insurance companies will remain local. Also, the nature of medical innovation is inconsistent with localism. Once a medical innovation is made in one place, it is easily available everywhere, and to create proper incentives for medical innovation users of the medical innovation everywhere, or at least over a wide area, need to pay for it.

Finally, determining *how effective different medical procedures are* requires research that is conducted on a large scale. That is, if we want to know whether Procedure X is cost-effective, we need to study large numbers of cases, and adequate sample sizes may not be locally available. More importantly, once you've derived a result about whether Procedure X is cost effective or not, that information can be used everywhere for making decisions.

I guess I'm not saying so much that localizing medicine is a bad idea, as that it's not clear what it entails. If it means that patients should pay for their own health care, the role of insurance needs to be reduced, the government should play little or no role, and we're going to have to be a bit hard-hearted about medical care for those who can't afford it, while accepting that medical innovation and cost effectiveness research will be conducted on a global scale... well, I guess the localism in medical care doesn't seem to emphasize the most salient features or to be the most apt description of such a program. If it means something else, what?

Joyless Moralist's belief that the government would spend less on health care and reduce its quality is supported by the experiences of Canada and Western Europe, but it's not so consistent with my general impression of how government acts. America might be different: we might finance health care at a level of quality comparable to what we have now, at exorbitant cost.


Insurance is for profit. The government is generally not. There are no investors, no CEO's to pay in a single-payer (government) system. The expected efficiency loss in nationalizing insurance would be manifest as 1) less or no risk assessment, and 2) less or no claim denials, which is exactly what the single-payer system is aiming for. The money that would have been spent on risk assessment, claim denial, etc, would instead be absorbed partially by the expected efficiency loss, but not all of it would be, and the left overs would improve the breadth (and maybe the depth) of health care.

What are the merits of universal health coverage? If society spends X amount of dollars raising and educating a person, and the expected productivity of that person is Y over a nominal life-expectancy, then it is worth insuring that parson for a significant percentage of Y with the hope that it will protect that potential productivity. If a person is uninsured, then there is a possibility that all of the future productivity would be lost in the absence of routine health care. It therefore is plausible that universal health coverage is more economically efficient than otherwise, and that it is in the general public's interest to insure all people.

I don't have a dog in this fight, but by all accounts, the system we have right now isn't that great. For that reason, a certain amount of experimentation is prudent, and one of the possible experiments is providing a public insurance option. That does not seem to me to be a terribly profound change, but it's something we could try for a couple terms.

Nathan Smith

If we tried it, it would be hard to back out of it, since government programs create their own constituencies. And it's virtually certain that it would be a fiscal drain, which is something we very very very very very much can't afford right now.

Please Congress. Kill it.

Nathan Smith

The fix I would suggest is:

(a) tax employer health benefits like normal income, in an effort to destroy the third-party payer system;

(b) create Medicare private accounts, so that the elderly can keep what they don't spend on non-exceptional health care;

(c) create a system of government loans which uninsured people who need/want medical care which cost effectiveness research shows is worthwhile can access. Later, garnish wages, seize houses and other assets, and renegotiate only if they really have nothing left;

We need to stop insulating people from the costs of their health care. That's the only way to fix the system.


In an ideal world, certain preventative care should be subsidized in some means-tested way, because people use too little, frequently for irrational reasons. This bit of nanny-statism is advised only because the same people being subsidized are also the ones who will show up in emergency rooms later on the public dime anyway. Care beyond this is when we want to get into Nathan's (c). I'm not sure I'm all on board with (a) and (b), but clearly we need something like them.


Maybe there should be two classes of health care, 1) accidental care, and 2) lifestyle care. For the first one, presumably the injury is/was unavoidable, and so it's in our interest to provide care in all cases. For the second one, the person made the choice to have a certain lifestyle, and the injuries resulting from that lifestyle should be insured/paid for by that individual only.


It seems like it would be difficult to distinguish between 1) and 2) in many if not most cases. For example, a diabetic has complications related to their disease. Do we deny because many diabetics become so because of health choices, or do we try to get a council of doctors together to try and decide if it was primarily due to genetic factors? And are the diabetic's complications are due to poor management of the disease, or bad luck? I'm not sure there'd be an easy way to address these kinds of questions.

Nathan Smith

re: "1) accidental care... For the first one, presumably the injury is/was unavoidable, and so it's in our interest to provide care in all cases."

Even if you could distinguish "accidental care" from "lifestyle care," which would hardly be feasible in practice (though of course we already try to some extent), it does not follow that it would be in "our" interest to provide care in all cases, regardless of whether the "our" refers to the societal interest including or excluding that of the patient. If a life-saving operation costing $1 million will be required in an *ex ante* improbable situation, it might be in a patient's interest *not* to secure insurance against this risk, because the value of the quality-adjusted life years gained times the probability that the life-threatening condition will occur is less than the cost of the premiums. If it is not in the patient's individual interest to provide insurance, it might still be in society's interest (including the patient's) that the operation be paid for, if the patient's survival benefits others (e.g., through the consumer surplus they enjoy by trading with him, or through his public-interested and charitable activities, including support for his own children and family). Alternatively, the patient's survival might create negative externalities for society, e.g., if the patient is an elderly person collecting Social Security checks (putting to one side the benefits loved ones get from seeing that person-- though if they are the beneficiaries, why shouldn't they pay?). If we take the patient's failure to provide insurance as evidence that the patient did not regard the quality-adjusted life years as worth the cost of the premiums, and if the externalities from the patient's survival are zero or negative, we might still want to subsidize the operation out of charity or some kind of moral conviction. But we shouldn't say "it is in our interest" to do so.

re: "certain preventative care should be subsidized in some means-tested way, because people use too little, frequently for irrational reasons."

How do we know this? I know this is often claimed, but is the research to back it up really there? It would be quite difficult research to do. I've seen research pointing in the opposite direction too, e.g., that routine doctor visits are useless. It's certainly plausible that people have the opposite problem, namely that there is widespread "hypochondria," over-consumption of medical care by people who imagine themselves sick when they're just fine.

All in all, I think the hypothesis that health care is just another market and supply and demand can get price and quantity right if regulation and weird indirect payment systems is much stronger than people give it credit for.


Let's say you are mortally wounded, but there is health care that would almost assuredly save your life: how much are you willing to pay for that care? I imagine most people would pay everything they're worth and more for that care. Now let's say that the facilities and doctors that could save your life are in extremely high demand: how much should the hospital charge you? Should the care go to the highest bidder? What if the other people demanding care are only trying to get cosmetic surgery (or something similarly less urgent), but they have/are willing to pay more money? Is it in society's interest to let someone die because they cannot afford health care?

The price/value of something in a free market is set by supply and demand. In this case, the something in question is a person's health. The demand for health care is potentially limitless, and the supply meager in comparison. This sets the price pretty high, as evidenced by the fact that the US spends more on health care per capita than any other nation in the world, even nations with universal health coverage. It's debatable whether or not the US also has the highest-quality health care. In light of these facts, I have serious doubts about the free market's ability to judiciously determine the price-point of individual lives.

It seems to me that there is not only a rational argument in favor of a single-payer system, but there is also empirical evidence that such a system offers better care for more people for less money.


Regarding accidental care vs life-style care, I merely meant to draw a distinction between immediate events beyond one's control and all other events. So if a kid breaks his arm playing sports, the care to set and cast the arm would be paid for by the public, but if the kid later develops arthritis in the arm, then he has to cover any attendant care himself. The diabetes example that Nato gave would fall into the latter category. Anything immediate and urgent that had no obvious bodily predictors would be covered by the public.

Nathan Smith

Tom writes as if the United States were an example of a free-market health care system, so that by comparing its performance to that of health care systems abroad we could judge how well a free-market health care system works. But it isn't. Entry into the medical profession is heavily regulated, with the result that many functions are performed by very over-qualified people. In a free-market system little or no official licensing would be needed to practice medicine, but private certification and reputation systems would emerge to measure doctor quality. Or again, in a free-market system, the tax system would be neutral as to whether insurance would be provided (if at all) through employers, by individuals themselves, or by some other means. Instead, our tax system heavily favors a weird employer-based health insurance system that emerged during World War II, when wage controls caused employers to develop roundabout ways to compensate workers. There's no real economic rationale for health care to be provided through employers, and as people become more professionally mobile this system is becoming ever more inconvenient, but the tax system locks it in. Or again, in a pure free-market system, hospital emergency rooms could turn away indigent patients who had no insurance and couldn't pay; in our system they cannot. This helps to preserve a third-party payer system, and so, please note, it is NOT "supply and demand" which set the price of health care high, because the price consumers pay for health care is typically far lower than the price producers charge, with health insurance and/or government paying the difference. If people had to pay the full costs of health care, they would probably consume less.

Tom writes: "The demand for health care is potentially limitless, and the supply meager in comparison." This is true of any market. Economists call it "scarcity."

To the questions Tom starts with: If I were mortally wounded, I would be willing at present to give my full net worth and more if I could borrow it; but if I were richer I might not be willing. For example, if I had $50 million, and it would cost $40 million to keep me alive, I would refuse treatment and leave my money to charities or other organizations that could put it to better use than saving a human life. To spend a fortune on one's own preservation when money could save so many more lives if spent on bednets or AIDS drugs in Africa strikes me as immoral. But of course, a selfish "rational agent" would spend anything he had to keep himself alive. However, that is not a good way to measure the value of keeping someone alive, because if you give someone a bribe to die they won't get the chance to enjoy the bribe. A better way to get a measure of the value of a life is to see how much people are willing to pay to avoid a *small probability* of death, because since they expect to stay alive, the money they save is of normal value to them.

Is it in society's interest to let someone die because they can't afford health care? I could take a moment to object to the phrase "society's interest," which seems to suggest an adding up of utilities which sound economic theory forbids. But that's not really needed because by any reasonable definition of society's interest the answer is: Yes, sometimes. Take an extreme case: suppose it takes 5% of GDP to keep Jimmy Rutgers of Belle View, California alive. Should we raise 5% of GDP and spend it on the exorbitant treatment? If so, what about 10%? 50%? Where do you draw the line?

Should care go to the highest bidder? Yes, probably so. I don't think it would often happen that someone who wants cosmetic surgery would outbid someone with a life-threatening illness, particularly not if institutional arrangements were set up so that the person saving his life could credibly commit to pay back whatever was spent on him if he possible could. But also, that might be the right decision. Suppose a Hollywood actress who gets cosmetic surgery can save her looks for a while and make movies that will delight and inspire millions, to the tune of, say, $100 million at the box office. Or an old man who's led a good life but whose loved ones are mostly dead can overcome a fatal cancer and survive for a couple more painful years before succumbing to old age. Which is worth more? I don't think you can necessarily devise any hard-and-fast rules about which kinds of treatment are worthwhile that would be wiser than the market.


Has there ever existed a truly free market, Nathan? There are degrees of freedom, and the US system is a number of degrees more free than the systems I was comparing it to. I suppose no conclusion can be drawn from this, however, due to the heterogeneous nature of the systems in question? How are we to learn from countries that have universal health coverage if you will not allow a comparison? Shall we just assume that they have nothing to teach us?

'it is NOT "supply and demand" which set the price of health care'

So health care is artificially high due to various confounding factors that produce a spurious relationship between supply and demand. That is what you're claiming. We cannot know to what extent the price of health care is determined by the relationship between supply and demand due to these confounding factors, which are a result of a market that is less than free. This reasoning indicates that you think there is potentially a "health care bubble" which would pop (to our benefit?) if we did ... ? Limited/eliminated the confounding factors? That's a bad assumption, because the factors could have a negative, not a positive, correlation with price. And actually, that's the crux of my argument for public insurance: health cost will go down the less free the insurance market is. A free market solution for health insurance is bad because ultimately the incentives are perverse. That is the rational argument I've been lamely trying to make. Free market principles are ideal when the industries involved have agreeable incentives, but they are not ideal when the incentives are perverse.

Nathan Smith

re: "This reasoning indicates that you think there is potentially a "health care bubble" which would pop (to our benefit?) if we did ... ? Limited/eliminated the confounding factors? That's a bad assumption, because the factors could have a negative, not a positive, correlation with price."

Health savings accounts, taxing employer health care benefits, etc., are ways to shift from a third-party payer system to a system where the customer had to pay most of the costs directly. If people were not insulated from the costs, they would be more likely to take them into account, and consume less health care, try more self-treatment, etc. We can predict that health care costs would fall, because the marginal price of health care for individuals would rise, and people always respond to higher prices for something by consuming less of it.


When you pay a dollar to an insurance company, in an ideal world, that dollar would go towards paying for a treatment. But the business model for insurance companies necessitates them pocketing as many of those dollars as possible, in essence, providing no service. A single-payer government system has no such requirement.

Imagine a small village that pooled together resources to afford a health treatment for one of its members. There would be a sort of tax on each member, and the taxes collected would only be used on health care (perhaps only for extreme situations). That is what a single-payer system is like. Now imagine instead a monarch collecting the taxes, taxing people more for being old, overweight, disabled, etc, choosing who gets to benefit from the taxes, denying certain hard cases from benefits, and ultimately pocketing the money that he doesn't spend on the people. That is the insurance market.

Nathan Smith

Tom is trying to argue that a single-payer system has inherent advantages over an insurance system. His arguments don't work.

re: "the business model for insurance companies necessitates them pocketing as many of those dollars as possible, in essence, providing no service."

In any industry, private businesses try to maximize profit. But competition restrains them. So for private insurance. Of course, insurance *is* unusual in that some people will get nothing for their money, while some will get far more than they paid for. But insurance companies are bound by contract to provide service when specified situations arise. They will, of course, have to design the contract so that the premiums they pay exceed the liabilities they incur, or they will go out of business, and they will try to design them so that this excess is as large as possible. But if they get too greedy customers will go to their competitors.

Tom seems to think that "claims denials" are a special problem with the insurance market that a single-payer system would avoid. Yet he makes the parenthetical suggestion that health taxes would be spent "perhaps only for extreme cases." This is essential, because without something like this there's no cost control mechanism at all. But what is to be done for cases that are, for whatever reason, deemed not "extreme?" Some form of claims denial or rationing must occur, perhaps in the form of long queues for services, if costs are to be controlled. A single payer system doesn't offer some magical escape from this necessity.

Tom is not making the subtle "moral hazard and adverse selection" type arguments that are sometimes made against markets in health care. His real point seems to be that we could save money if insurance companies weren't making profits. By the same logic, we should nationalize the restaurant industry: we could all get our food cheaper if those restaurateurs weren't making profits at our expense!

Please note that what is currently called "health insurance" is, for the most part, not insurance at all. True health *insurance* would cover only unusually large expenses, e.g., medical costs in excess of a few thousand dollars per year. What we have is simply *cost insulation,* a service that subsidizes quite ordinary and predictable expenses. It is as if we were to set up a system of "hunger insurance," whereby everyone paid a fixed rate and then could eat at any restaurant they liked, with a small co-pay. Of course, in that case, we would all eat more luxuriously, and the system would become exorbitantly expensive. Ordinary medical expenses should be paid neither by the state nor by insurance, but should simply be paid out of pocket.

The comments to this entry are closed.

My Photo

Only use a payday cash advance as a last resort.


Blog powered by Typepad