« Keynesianism on Crack | Main | In Defense of Fake Social Security Cards »

June 12, 2011



I think HSAs could be expanded to take on this kind of role. I sort of feel like there needs to be bigger incentives to push everyone, including businesses, to move toward HSA-type plans. However, I think it's worthwhile to temper expectations regarding results.

Besides foolish regulations on health provision (which is, I think, much more important in this respect than than insurance regulation), I think the biggest source of market failure in healthcare is the lack of information. The historical studies I've seen indicate we've always been pretty terrible at choosing effective and efficient health care on our own, a product of 1)originally, the fact that medical 'science,' wasn't, and 2)once science started to make real headway, it quickly became too complicated for any layman to keep track. Now, there are lots of other things for which the layman has no real hope of mastery, but markets for the relevant information have risen to meet the need. I have no personally-feasible way of judging the reliability of a car, but I can pay Consumer Reports a small sum to provide me with very good information on the topic. Because of the history of the healthcare market, this kind of service hasn't evolved for health. I think that it would, if we gave it a chance. The problem is that it might take a generation to eventuate, during which things don't seem likely to be *better* than they are now where other actors with more resources (but frequently misaligned incentives) make most of the decisions for us.

But the information does at least exist, and it seems like this is one place where a government council might be of service, attempting to judge to cost benefit relationship of procedures with which to furnish the public consumer. I would support this wholeheartedly if I didn't think it would likely kill/delay the evolution of the private alternative. It might still be worth it, I don't know, but it is a concern.

In any case, I've come to agree with Nathan that employers providing health insurance is (now) a terrible, distorting arrangement.

But back to the debt thing - I think the government should start providing vouchers for currently covered individuals for Medicaid & Medicare HSAs, along with insurance for catastrophic care above the voucher amount, such that the overall level of coverage is unchanged but so that the individual can benefit from cost reduction. This wouldn't save the government any money directly (only the recipients would potentially benefit), but it would support the right incentives in the market. At the same time, start moving future beneficiaries over to HSAs as well, but with the "debt-coverage" phased in as the market starts to provide the necessary kinds of information. So envisioned, it would take at least a generation and probably more than two before everyone is transitioned to a functional and maximally* efficient private health care market.

This probably sounds a lot like the Ryan plan, but I think that's primarily because it's so diferent from others actually on the table, not because it's actually that similar. The critical difference is in ideas about the role of insurance in health care and to whom the near-to-mid-term savings would accrue. That is to say that I think we could do it as a long-term cost fix but in the meantime we shouldn't count on it fixing the budget hemorrhage. On the other hand, I think might would end up producing actual efficiency gains while I think almost all of the Ryan plan's saving would come from forcing reduced consumption. Ryan's plan may still be worth it for other reasons, but I don't think it gets us where we want to go for health care specifically. Not that any of the others do, either**.

The proposal deserves more extensive commentary, but I guess I'm trading sketch for sketch at this point. Also, I should be doing other things.

*For healthcare, I think there's grounds for skepticism that it will ever be as efficient as other markets. I still think we could improve a lot, but just maybe not to the point of the garden variety market.

**Granted, there are other ways of improving efficiency, and I think some of the other plans have a realistic chance of doing that by emulating other countries that have done so. There's even good reasons to go this way (i.e. it's been tried elsewhere and it seems to work fairly well), but I do think that those public systems are 'local maxima' in the possible solution space. We're very far from any maxima (it seems to me), but that also gives us the freedom to go in the direction of the (I believe) global maximum without having to endure anything much worse than what we already have.


I've always kind of wondered... if the US has a fairly mediocre health care system cost-wise and there are other systems around the world that are clearly better, then why don't we try to learn from those systems and emulate them? The main reason you wouldn't is if you were in denial and thought the US system was the best in the world. But for argument's sake, let's suppose some country somewhere has an obviously better system: what's the rationale for not trying to implement that system?

Nathan Smith

It's a myth that "other systems around the world are clearly better." The US health care system costs more, but we tend to do more testing and to be at the cutting edge of medical technology, so we're getting something for what we pay. Also, the US plays a leading role in medical innovation, so to some extent other countries are free riding. When people try to say the US system is so bad, they tend only to cite higher life expectancies in other countries. But that's not a valid indicator, because longevity depends on lifestyle, and health care is partly about improving the quality of life, not just lengthening it. If you start looking at the details, about how cancer, say, is handled, the US system tends to do as well as or better than other systems in terms of quality, albeit at high cost. This is not "denial," it's just taking a sober look at the facts. I don't think the US has a whole lot to learn from other countries when it comes to health care, or at least not from the ones most often cited, western Europe and Canada and Japan. We should develop a new method of our own, by thinking carefully and afresh from market principles.


A correction: The US doesn't have a "fairly mediocre health care system cost-wise." On a cost basis, it's the worst by a surprisingly wide margin. On a cost for measured results it also seems to be the worst on any metric I've seen. Basically the US is always a giant outlier. It may be that we consume much more in terms of tests or whatever, but it would seem that this isn't leading to better health outcomes in a statistically noticeable way, so this would seem to be just another sign of inefficiency.

But I don't think that leads directly to emulating other countries' systems. I think they definitely have things to teach us, but I also believe there are other solutions available to us that would achieve the same goals while harnessing the power of the market.

Having said that, the conservative solution would seem to be to follow empirical success rather than attempting to make the market match economic theories.

Nathan Smith

re: "this isn't leading to better health outcomes in a statistically noticeable way"

If by "health outcomes" and "statistically noticeable" Nato means that only crude health outcomes like longevity should be included, and statistical tests should be naive and not control for other influences on health outcomes, this statement is true. But you can't just look at macro health outcomes and judge a health care system by that. Lifestyle, cultural, and genetic variables are independent of the health care system but affect health outcomes a lot. A more appropriate method of measuring the quality of a health care system is by looking at how well it does treating individual diseases. As Timothy Noah concedes:

"It's a fair point. Granted, nobody seeking entrance to a Cuban, Canadian, or English hospital would likely seek help from an American politician. But there's a reason that foreign potentates turn up at New York's Memorial Sloan-Kettering Cancer Center and Houston's M.D. Anderson Cancer Center on a regular basis. The international nature of the medical-student body at Harvard and Johns Hopkins is not a figment of your imagination. The United States is pre-eminent in the treatment of cancer and in most other high-end medicine, too. Michael Moore's stubborn refusal to acknowledge this fairly obvious reality undermines his argument for national health care in Sicko."

There's also the fact that in places with socialized medicine you have to wait, sometimes a long time, for nonemergency services, which obviously reduces the quality of life, even if it doesn't shorten your life. So, again, it's just not the case that foreign health care systems unambiguously do a better job. If you think foreign health care systems deliver the same thing for less money, you're just wrong. They deliver something different, and in important ways clearly worse, albeit while taking a much smaller bite out of GDP. You may think that's a trade-off worth taking, but it *is* a trade-off.

Now, in this case, it is NOT true that our health care system is market-oriented and we can deduce that what Arnold Kling calls "premium medicine" is worth it because people buy it, by "revealed preference." That's because our health care system insulates people from the costs of their own choices through misnamed "insurance" policies and, in the later stages of life, through Medicare. Probably what people would really want, if they had to internalize all the costs themselves, would be somewhere in between.


I thought, after I wrote the comment, that there were an especially large number of 'seems' qualifiers this time - totally appropriate because it's such a complicated subject that it's hard to feel like one has scratched the surface in any particular respect.

Regarding waiting: there doesn't appear to be significantly less waiting in the US than anywhere else, unless one is fairly wealthy, in which case the US has a pretty large advantage. Certainly my mother had to wait almost ten years to finally get surgery on her cervical vertebrae from her HMO and I've heard a number of similar anecdotes, so I can see how the statistics which show comparable wait times could be accurate.

Finally, I'm not sure if the comment about whether the healthcare system is market-oriented or not is to my address, and if so, what exactly you mean to say. If not, then I think we agree.

Nathan Smith

Let's put it this way. If you understand how markets work and how bureaucracies work and what's wrong with central planning, your instinct will be to dismiss the claim that universal health care in foreign countries delivers just as good service at a much lower cost than the American system as implausible. The facts do not contradict your knee-jerk reaction. There are plenty of advantages to the US system. Certainly it's plausible that the relative benefits are not worth the relative costs, but attempts to PROVE that foreign health care systems work better (in effect, that universal health care is Pareto-superior) are always based on confusion and/or obfuscation.


The problem is that I don't see anything that looks like a well-functioning market in the US. I see a bunch of rent seeking, poorly-designed regulation and perverse incentives. If it worked anything like an efficient market then I would be utterly shocked and totally suspicious of the numerical comparisons, but as it is, it seems like a bunch of mediocre systems being compared to a really bad one. I do not imagine the bureaucratic systems of other countries to be paragons of efficiency or anything like the best that can be done. Thus I support us trying to design an intelligent market-based solution rather than just copying theirs. But I do that because I have a theoretical presumption that a market could be designed such that US healthcare would be the best in the world in efficiency, not because I think our current mangled market actually works better than their intact central planning.


Free market analysis usually starts with supply and demand. For everyday items like utensils and furniture, it's easy for individuals to predict and control their demand. Healthcare is different. Ideally, a person would be completely healthy their whole life until they passed away quietly and suddenly at a ripe old age. In this ideal scenario, a person would not demand/consume any healthcare. Even when people could benefit from healthcare, there is a general fear of hospitals, medicines, surgeries, vaccines, needles, etc, that causes many people to not seek healthcare. It's impossible to know for most people what kinds of healthcare they will need in their lives. Some people will need a lot of healthcare and others will need a little. Since nobody really knows how much healthcare they will need in their lives, the financially safest way to play it is to distribute the burden of healthcare costs evenly across the population. The financially riskiest way to play it on an individual level is to let people pay for healthcare by themselves when they need it. The shared burden strategy distributes risk and minimizes individual hardships, while the individual burden strategy is maybe more fair in a natural selection sense.

It seems to me that healthcare is demanded out of necessity first and foremost; people in general do not intend to consume it frivolously. Now, it's possible that healthcare professionals provide/recommend services frivolously because it simply means more profit for them, just like an auto mechanic frivolously recommending various services and repairs to your car to improve the bottom-line. What is an almost completely ignorant consumer supposed to do in such a situation? If an auto mechanic recommends X, Y, and Z to fix your car, if you don't take the advice, the worst case scenario is your car will need some other kind of repair in the future (or maybe you'll need a new car). If a healthcare provider recommends X, Y, and Z to potentially save your life, how many people are really going to go against that advice regardless of whether or not it's sound?

So there are two main issues here: 1) how much of the cost, if any, do we want to distribute among the population, and 2) how can we minimize frivolous procedures and maximize consumer choice without hurting the quality of healthcare much? Laissez faire capitalism and government single-payer are the two extremes that address the first question. The answer to the second question is not obvious. Maybe some kind of consumer healthcare advocate or standards body is needed. I don't know, but it's clear to me that the costs associated with the first issue are dwarfed by the costs associated with the second issue.

Since this issue is so complex, it would be better to experiment with different systems at the state or regional level than at the federal level. Let states try out their favorite ideas, and let's see which ones are worth keeping and which are worth discarding.

Nathan Smith

Tom is trying to argue that free market analysis is somehow inapplicable to health care. He does not seem to want to make the more modest point that the health care market has some special features that make the free market paradigm harder to apply. He seems to want to say that the framework of supply and demand applies only when "individuals can predict and control their demand" and that "health care is different" because "in the ideal scenario, a person would not demand/consume any healthcare at all."

In an ideal world, I would not demand/consume any mosquito repellent at all, because there would be no mosquitoes. I would never repair my car, because it would not break down. I would not need sunblock, because the sun wouldn't have the kind of ultraviolet rays that burn me. I wouldn't need to pay for gas to heat my home, or electricity to run the A/C to cool it, because the temperature would always be comfortable. I would not need to buy light, because the nights would not be dark but only pleasantly dim, providing enough light to see by. There are all sorts of things which in an ideal world one would not need to consume. That criterion has nothing whatsoever to do with whether free market analysis is applicable.

I cannot predict how often I will need to employ the services of a mechanic to repair my car. I cannot predict how many movies will come out this summer that I consider worth seeing. I cannot predict how many incoming calls I will get and whether my budget will suffice to pay for the minutes. I cannot predict when I will need an umbrella, or a snow shovel to dig out my car after a blizzard. I cannot predict when my iPod will die and I'll need another one, when pipes in my house will start leaking, when a tree branch will fall on my house. The free market is not at a loss to deal with any of these cases. Most of these eventualities I just cover out of pocket. For the larger unknowns there is insurance, e.g., car insurance, homeowner's insurance. Private insurance markets do have some special problems, but they work reasonably efficiently. Health care is an exception, but that's not because it's a problem that markets are inherently incapable of dealing with, but because of the way the sector has been distorted by government policy.

Tom says that "Since nobody really knows how much healthcare they will need in their lives, the financially safest way to play it is to distribute the burden of healthcare costs evenly across the population." This is a complete nonsequitur. Even if our only goal is to minimize risk, there is no reason that the burden should be distributed "evenly across the population." You could distribute the burden any which way and achieve the same minimization of variability in the costs faced by the individual. You might put all the burden on the rich. You might make predictions about who is most likely to be sick and place the heaviest burdens on them. Tom's suggested policy, taken literally, would be to socialize all costs and then finance them with a poll tax. Nothing wrong with proposing that. But his suggestion that that is the unique way of minimizing risk is spurious. And would it really be fair? Suppose Bill Gates has to have surgery for $1,000,000. Do we really want to let the state pick up the tab and then charge every American a fraction of a penny-- including homeless beggars for whom this might actually be a hardship? Or do we say, Bill Gates can afford it, let him pay for the surgery himself? The answer to all this, of course, is that yes, it's efficient to socialize medical risk to some extent, though probably a lot less than we currently do, but this could, and probably should, be mediated through private and more or less unregulated insurance markets. It's just a fallacy to think that because there is risk, the mechanism of private contract can't deal with it.

Tom tries to make a distinction between "necessity" and "frivolous" consumption. An argument along these lines could be applied to anything, not just health care. How much do people eat, drive, sleep, read, from necessity, and how much of it is frivolous? The truth is that with health care as with other goods, people have choices. What if Treatment A costs $500 and gives you a 99.0% chance of survival, and Treatment B costs $500,000 and gives you a 99.9% chance of survival? Is Treatment B a necessity? If it costs $1 million to keep a very old, very sick person alive for a few more months with cutting-edge technology, is that a necessity? What if my ulcer, or my arthritis, is not life-threatening, but limits my movement and keeps me in a constant state of mild pain? Is treatment "frivolous" because it is not necessary for survival? Tom's necessity/frivolous distinction just confuses the issue, it should be dropped. We're faced with the usual economic problem: people have things they want, but resources are scarce. Health care is no different, really, from anything else in this respect.

There are some peculiarities of health care markets, to be sure, but free market analysis, supply and demand, are still pretty much applicable. Tom's argument for why we should throw that conceptual framework out the window and start from *ad hoc* new principles is not convincing.


Hmm... I don't know if Nathan is being deliberately obtuse here or not.

Most things in life are luxuries, not necessities. Is healthcare a luxury? If that's the way you want to look at it, then fine, treat it like any other good or service. If you asked someone bleeding to death what their opinion was, they would probably say that healthcare is a necessity, albeit one that nobody ever knows in what way they will need it. You don't need mosquito repellent, you don't need car repairs, etc, etc. Heck, many people will never even really need more than a trivial amount of healthcare. Is this too nuanced a position to take?

Your third paragraph is actually wrong. You can predict fairly accurately (not 100%, but much better than a coin flip) most of the examples Nathan gives as unpredictable. The probability of certain events happening comes from some distribution or other which is usually non-uniform and has a well-estimated (even to a layperson) mean and variance. Even if you assumed the distribution was always Gaussian, the penalty for model mismatch is not going to hurt your predictions that much. But all of this is beside the point: if your iPod dies, 1) it's not problematic to replace it, and 2) you don't even need to replace it; if your kidneys start failing, your options are 1) die or 2) get treatment. Again, maybe this position is too nuanced.

@ fourth paragraph. Um... if you wanted to minimize the maximum impact of healthcare costs to the individual, you'd want to distribute the costs uniformly; a uniform distribution would be the solution to this minimax problem. It's not progressive, though, so maybe Nathan is thinking about putting more of the burden on the rich and less on the poor? The private sector insurance market does distribute financial risks, but I don't think it would necessarily do it better than a public option. I've made this point before, but I don't want an insurance company to be efficient at what it does (i.e. taking in as much money as possible while paying out as little money as possible). Governments are less efficient than private companies and they don't have a profit motive. That seems pretty ideal to me for this kind of service.

@ fifth paragraph. I think you're restating my point here and yet you're somehow missing the point. Cost-wise, it's not very important whether or not your eating and sleeping is frivolous, but if you really wanted to know, it wouldn't be difficult for the average person to determine whether or not they were frivolously eating and sleeping. Healthcare, on the other hand, is many orders of magnitude more costly than the vast majority of personal expenses, and it's also incredibly difficult for an average person to determine whether or not they really need it. So it would be nice if there was a disinterested third party that could help healthcare consumers decide whether or not they really need that $1,000,000 procedure or whether they can get by with the cheaper $10,000 procedure instead. Nathan says healthcare is no different from anything else, but it's actually several orders of magnitude different in cost and importance to the individual.

Nathan Smith

re: "Most things in life are luxuries, not necessities."

No, sorry, not the case. False. Baloney. Erase. (a) Lots of things are necessities. (b) There's never a clear line between what's a necessity and what's a luxury.

re: "Is healthcare a luxury? If that's the way you want to look at it, then fine, treat it like any other good or service."

Stop. The reason to treat health care (more or less) like other goods and services (I'm not denying that it has some peculiarities but far less than Tom suggests) is NOT that it is a luxury. Is food a luxury? Presumably not since we would die without it. Should we nationalize grocery stores?

re: "if you really wanted to know, it wouldn't be difficult for the average person to determine whether or not they were frivolously eating and sleeping."

How would you know? What would that even mean? And concerning health care, Tom hasn't addressed my examples in the last post:

"The truth is that with health care as with other goods, people have choices. What if Treatment A costs $500 and gives you a 99.0% chance of survival, and Treatment B costs $500,000 and gives you a 99.9% chance of survival? Is Treatment B a necessity? If it costs $1 million to keep a very old, very sick person alive for a few more months with cutting-edge technology, is that a necessity? What if my ulcer, or my arthritis, is not life-threatening, but limits my movement and keeps me in a constant state of mild pain? Is treatment 'frivolous' because it is not necessary for survival? Tom's necessity/frivolous distinction just confuses the issue, it should be dropped."

A disinterested third party could make some arbitrary choice about whether to finance the $1,000,000 treatment or not, but they couldn't really decide whether it is a "necessity" because that term is not fundamentally meaningful outside of particular social contexts and conventions. At best, the third party could ascertain that the $1,000,000 treatment would raise survival probability by, say, 0.5%, but does that make it a "necessity," or not? Tom's point is not "too nuanced;" it's too crude, it's too un-nuanced. He's trying to put a lot of weight on a concept that falls apart the moment you look at it closely. He's begging all the major questions.

I think Tom exaggerates when he says that "what an insurance company does" is to "take in as much money as possible while paying out as little as possible." Certainly there's a temptation there, and the legal system has a tricky task here in outsmarting insurance companies trying to write contracts in their own favor at customers' expense. But what a health insurance company does is take monthly premiums and pay for unexpected medical costs (and routine preventative care). Of course, they do try pretty hard to anticipate the likely medical costs of different kinds of customers and set premiums accordingly, but what's wrong with that? Why should a poor but healthy young person pay large sums out of pocket when his health risks are slight, in order to finance care for a wealthy old sick guy? And another advantage of private health insurance markets is that it gives people an incentive to stay fit and healthy, so as to keep their premiums down, whereas if you socialize all the costs, that incentive goes away.

I think a system of private health insurance could work pretty well if it were delinked from employment and regulated in a manner that prioritized freedom of contract while bearing in mind people's limited ability to understand complex contracts. What I'm suggesting in the post, though, is that maybe we don't even need to insure away so much risk, maybe we can handle a lot of it through debt. That wouldn't work for old people, but for people under, say, 40, or even 50, maybe people could pay for medical care the way they pay for college: let the government give them access to guaranteed loans with lenient repayment terms, but let them take the costs themselves. Of course if they have insurance, fine, but to the extent that we feel the need to pay for health care for the uninsured poor, let that take the form of debt.


I have to agree with Tom's point that most things are not necessities. I know that when I was a teenager I was constantly calculating how much it would cost for me to live based on various scenarios. While I don't think I had it all quite down, I will say that I did get into the range where I could live on a minimum-wage part-time job, have a (tiny) entertainment budget and still save ~$50 a month*. This would have involved sharing a room with friends, eating a carefully-circumscribed diet, and so on. In my actual life now, I spend almost an order of magnitude more money on just the mortgage plus HOA for our 660sq.ft. condo, and a whole lot more money on other stuff. I would say that about 90% of the money we spend is on non-necessities.

I would also like to agree with Tom in a general way about the difficulty of determining the 'necessity' of procedures. Nathan seems to complicate his point, but doesn't refute it as far as I can see. We still need that trusted, disinterested third party that has far more resources than the individual for determining what kind of care is called for. Sure, present different options to the individual and do whatever you please to arrange for the payment, but currently people just don't have the information necessary to enable them to make confident decisions in an arena that could hardly be more of a necessity.

*Always assuming I didn't get sick or seriously injured

The comments to this entry are closed.

My Photo

Only use a payday cash advance as a last resort.


Blog powered by Typepad