12 December 2008

Health comment

In response to Ryan's comment from "Looming Health Care Crisis"

btw, thanks for the comments

1) right; health care is not a public good.

But, there are other market failures besides the markets for non-rival and non-excludable goods, that warrant government intervention, even by conservative economists' standards.

In particular, private information (hard to verify claims) hamper market transactions; whether a worker's back really hurts and he needs time off, or whether a used car salesman's car has some glitch somewhere.

Consider 10 used car salesmen, one of which is selling a lemon, but consumers are unable to discern which. All salesmen then must lower their price below fair value to deal with the customer's belief that their car might be the lemon. Introducing an independent verification service, as long as it costs less than the price reduction from having the lemon in the market, would make the good salesmen better off.

The problem is, with the health market, health diagnosis are not confident enough to stand by a verification service at this point, so the government is legitimately offering to stand in. Even as a minimal interventionist, this makes sense to me. The human body is still too complex relative to an automobile.

As an after thought, another market failure besides 'public goods' and 'private information' that could warrant government intervention in markets is 'market power', when firms don't act as price-takers. In this situation, the firms profit maximizing behavior can differ from a benevolent social planner's behavior when maximizing consumer surplus (total value added by market transactions).

2. I do think adverse selection is a real problem.

Sure, among financially well off individuals, most are buying health insurance as to insure against unforeseen risk. But, in less well off social circles, where money is tight, many individuals prefer to spend their money on more tangible commodities until they begin to believe that they might be afflicted with a very costly condition.

3. Last, health insurance is comparable to car insurance for just the reason you point out, the potential costs on others. Many people would find it objectionable to someone afflicted with a treatable emergency condition to die on the front steps of a hospital. And, since we wouldn't let that happen to those who don't pay for it, we shouldn't make anyone pay for it. Unforeseen emergency afflictions should be covered by a pool of small contribution from all who want to participate.

Sure, if you don't want to be pooled for this cheap coverage, you can wear a tag that either says "don't help me, I chose not to be pooled" or "help me, I don't have the pooled coverage, but I have been verified to have the financial resources to cover this emergency operation up to the level …"
Honestly, it would probably be just as cheap to include both 'those who want pooled coverage on emergencies' and 'those want to forego the coverage', since then insurance companies would not have to do any screening at all. They wouldn't have to worry about any adverse selection (people more prone to emergency, more likely to buy insurance) because everyone would be getting it. So, basically, we'd cover those who don't want it for free.

In conclusion, this is a great and timely conversation. In the wake of a severe implosion to the Republican Party, it reminds us of how the virtues of fiscally conservative policies should be considered in even the most Democratic agendas such as Health Care in order to get the most cost effective coverage with the least government infringement on our individual freedom of choice.


Ryan said...

Thanks for your response. I really appreciate the chance to discuss. If I can impose on your patience to ask a couple follow-up questions:

1. Moral hazard. I agree absolutely. But I don't see why you think this is an argument for government provision of health care, or for more insurance in general. Generally speaking, moral hazard tends to be an argument against these things. Subsidizing health insurance or health care lowers the marginal private cost of malingering, consuming health care, taking less preventative action, etc. So we should have less insurance and less government and a greater proportion of health care expenditures ought to be out-of-pocket

2. Adverse selection. Why would the poor necessarily be buying less health insurance than is optimal in your model? Wealth is a form of self-insurance, so one might think there would be greater demand for insurance by the poor, and more attempts at adverse selection by the rich (since being detected & denied coverage later wouldn't be as much of a utility cost for them -- they can more easily afford it). I would think that if the poor buy less health care and/or less insurance, it's rational, not adverse selection.

More empirically, does this sort of thing happen often? This is a real concern when people tell "adverse selection" stories -- e.g., if the "market for lemons" model were correct, there would be no used car sales between strangers; healthy skepticism (pardon the pun) is warranted here. As I mentioned, there can be quite a lot of positive selection as well -- being a low-risk sort of person is obviously correlated with having a large demand for insurance (i.e., being "risk averse"). Keep in mind how information asymmetry has to work here. Is it actually the case that the average person knows more about his or her real, inherent risks than a health care company employing lots of actuaries & health experts does? I would think that adverse selection might be a lot harder with health than with cars. So it seems like you have to make the argument that, in point of fact, people know a lot about there own inherent risks (as opposed to controllable behavior), that this results in a very strong effect for adverse selection, and that this adverse selection completely swamps all positive selection -- in fact, it also swamps moral hazard, government inefficiency, rent-seeking, etc. Is there really any evidence that this is the case? And even if there were, the problem pointed to here is either (a) people are poor, or (b) poor people don't buy enough catastrophic insurance, not (c) ALL people don't buy enough catastrophic insurance, much less (d) ALL people don't spend enough on health care in general. But general subsidies for health care requires claiming (d).

3. Regarding the claim of external cost from being sick, I wasn't clear before. If person A is sick and person B helps him, this isn't really an externality. Apparently person B wants to help person A when person A is sick: if he doesn't, he should stop. My point is that in order to say this is a cost for person B, you have to say that person B actually doesn't want to give money to A, but is doing so. This implies B is irrational (or has an addiction problem, and not the rational kind) so I don't see why this would call for government action to force A to do something. One might even wonder if the efficient intervention would be to simply prevent B from holding his own money (or at least suggest he stop voting and making gov't policy less rational or even triggering a Condorcet cycle).

Most evidence I've seen, particularly the RAND randomized trials (which seem particularly devastating here) says that more expenditures don't make people healthier. My understanding is that even token co-pays substantially reduce expenditures; if revealed preference tells us people value a procedure or doctor's visit at less than $20, how can we conclude that in reality, we spend way, way, way too little on health care? And let's remember: that's what we need to argue here. 28% of the 2007 federal budget was Medicare and Medicaid. Add in state & local expenditures. Then add the cost of tax-exempting health-care benefits. Is there any evidence or good economic theory, at all, that justifies all this -- that says that left to their own devices, people spend this much less money on health care than they should? Those RAND trials sound like the opposite: we spend way too much. And remember, these programs are subsidizing health expenditures in general, not actual insurance. Adverse selection and insurance arguments are non sequiturs when you're talking about Medicare. Health care when you're old isn't unexpected. The programs we have, or the ones being proposed, aren't about unlikely, catastrophic events. They're not even really insurance. So I think that a lot of these arguments are really nonresponsive. What is the argument and where is the evidence? Why is it so scarce when so much is needed to justify the gov't health care interventions we have, to say nothing of expanding them?

What are the standards of evidence here, anyway? Far too many economists commit the nirvana fallacy or assume people are schizophrenic: selfish consumers but selfless voters (and gov't actors). Surely some public choice analysis is needed before calling for intervention. If we're holding the market up to the standard of perfection, why not the gov't as well? The standards should be higher in the latter case: removing private incentives should make us more skeptical, not less.

Anonymous said...

Keep posting stuff like this i really like it