There's a lot of health care talk going on over at the Washington Monthly, and Kevin Drum wonders whether preserving insurance-companie competition in health care, albeit under strict government regulation, would be a good idea on the merits:

[W]hat if we had, say, ten or twenty healthcare providers, all offering different plans? They're still tightly regulated, and there are minimums that all of them have to offer, but they're paid enough that they can afford to offer more than just the minimum. ... They all have an incentive to cut costs because their government reimbursement is a set amount, but they also have an incentive to offer state-of-the-art services in order to attract more customers than their competitors.

That sure sounds nice, but I think David Himmelstein, in a piece long, long ago in the American Prospect, gave a better picture of what would probably happen under tightly regulated competition. The insurance companies wouldn't compete by offering better services; they'd compete by trying their damnedest to pawn off the sick people on somebody else:

Managed Competition would be further undermined by insurers' efforts to attract healthy enrollees, so-called risk selection. Since 10% of the population consumes 72% of health care, the easiest way for insurers/HMOs to undercut their competitors' prices is to quietly avoid enrolling sick people in the first place, and drive away the chronically ill by offering unsatisfactory care. Immense financial reward accrues to insurers that successfully avoid risk, assuring extraordinary efforts to circumvent regulatory bans on risk selection.

Although managed competition in principle requires open enrollment, such requirements for open enrollment [are easily skirted]. Place sign-up offices on upper floors of buildings with malfunctioning elevators. Refuse contracts to providers convenient to neighborhoods with high rates of HIV (an example of medical redlining). Structure salary scales to assure a high turnover among physicians; the longer they're in practice, the more sick patients they accumulate. Assure the easy availability of services for the worried well, and inconvenience for those with expensive chronic illnesses.

In Medicare's HMO Demonstration Project regulatory oversight did not avert even flagrant abuses. Predictably, the HIPCs' efforts to adjust the capitation fee for predictors of health risk will be no match for the creative and subtle means devised by unscrupulous insurers/HMOs to avoid the sick. Millions will be spend on consultants who assist in targeting the most lucrative sub-markets. In a competitive environment, insurers that lower their costs by effectively dodging health problems are sure to succeed, those that tackle them are likely to fail.

That seems virtually impossible to get around, no matter how hard you try to stop it. The Washington Monthly health plan under discussion would presumably offer insurance companies greater compensation for picking up sicker patients. Himmelstein notes that that didn't work with HMOs, and it's fairly clear why. The main problem with these much-heralded "risk-adjusted vouchers", I think, is that different markets tend to price different health risks at different rates, so there would still be a slew of attempts to game the system. A given insurance company would go after only those customers whose risk it prices lower than the government does. This is all just a glorified form of gambling, not health care. Beat the Feds, win a prize. Now perhaps with enough ingenuity you can chase down all these dodges and abuses with a bunch of patchwork regulations, but at some point you have to wonder whether the benefits of managed competition are worth the costs. My guess is "no," although I'll admit that I don't know for certain.

(OK, edited: Didn't have the case against risk-adjusted vouchers quite right the first time around. But I think it's correct now.)