Sunday, August 2, 2009

Some simple, and not so simple, economics of health care

This recent column by Paul Krugman claims to make the case that "markets" can't cure health care.

Part of the column is very good. Health care, at least big ticket health care, is not like other goods. This is less true of day-to-day things like glasses and cavities, which is part of why there is not a lot of talk about these markets. But, for big things there are two key issues. First, there is a lot of uncertainty. Big costs - very big costs - can arise suddenly. This suggests the value of insurance, which is of course something that can and does arise in markets. Second, there is an information asymmetry. The doctor typically has more information than the patient about the available treatments. As a result, just as with car repairs, doctors have a financial incentive to mislead and sell more treatment than is necessary. Insurance helps to work against this tendency by limiting what doctors can do via selective reimbursement. As Krugman notes, though, insurers have incentives that push in the opposite direction from physicians, though this point is oversold in Krugman's piece because insurers also have long-run concerns about their reputations, which affect their future profits. So, there is no question that health care is not like bread or even cars. No economist would disagree with this.

The trickier question is what implications this has for policy. If the government provides insurance, it has much the same incentives as private insurers to cut costs. Treatment provided to individuals is as much a "medical cost" to the government as it is to a private insurance company. I think it is reasonable to argue, given the different institutional incentives, that government run insurance is more likely to respond to the situation by approving more treatment and running a deficit. But that is not socially optimal either if it means, as it certainly does, for example, in the case of medicare's payments for treatments near the end of life, that a lot of treatment is provided that does not come even close to passing a social cost-benefit test. Saving administrative costs by failing to limit treatments that do not pass such a test is a fool's bargain.

I guess what puzzles me is that there is not more emphasis on the non-profit sector in the current policy discussions. As I have learned from my colleague Paul Courant, we rely on non-profits to solve similar problems of information asymmetry in the higher education field. Why not in the health field? Non-profits would seem to avoid some of the problems with both for-profit private insurance (your care is not some shareholder's loss in profits) and government insurance (because non-profits face a hard rather than a soft budget constraint).

So, I applaud Krugman's focus on the particular nature of health shocks and on the information asymmetry between patients and providers. One can certainly find discussions in libertarian and conservative quarters that do not face these issues as they should, but such discussions represent the low end not the high end of the debate. Simply pointing out that health care is not the same widgets, though critically important, does not alone make the case for the Obama plan, or for a single payer system, or indeed even for many aspects of the current design of Medicare and Medicaid.