Slate Magazine's Timothy Noah came out in favor of socializing medicine a few days ago, citing this analysis by Paul Krugman and Robin Wells as a piece that "ought to be read, I'm not kidding, by every American". So I read it.
Krugman's tidy thesis relies on the established fact of adverse selection. The problem, as stated in the original paper, goes something like this: Imagine a market for used cars in which every car is either 'new' and worth $100, or a 'lemon' and worth $50. Only the seller knows what kind of car it is -- the buyer has to guess. Now, at what price should cars be sold? Akerlof cleverly realized that at any price less than $100, the new car owner isn't getting his money's worth, but that at any price over $50, someone buying a 'lemon' will be cheated. As long as information is asymmetric, only lemons will sell -- and they'll only sell at lemon-prices. This is a problem for used-car dealers, who can't offer variety, and a problem for new-car owners, who don't have liquidity (it's also a problem for economists who convince themselves that it's all theoretical).
But the 'lemons' theory isn't "Assuming a variation in quality, markets will be inefficient." It actually states that "Assuming a variation in quality, the reason markets are inefficient is that one side has more information than the other." That's actually quite refreshing, because it offers two solutions: 1) Inform both sides. 2) Use financial engineering to make them the same side.
1) Is the reason Carfax could afford all those obnoxious ads. If the 'information gap' between a buyer and a seller gets narrowed to zero, 'lemons' and new cars can both trade at fair prices, because there are now two markets: Instead of a monolithic bunch of used cars, there are discrete markets for discrete quality levels.
2) Doesn't happen much, but probably could: Instead of simply swapping a car for cash and calling it a deal, buyers could buy an allegedly new car only if the previous owner paid to have it insured. This puts the financial burden of a 'lemon' back on the person who has the information -- the used-car buyer is fully hedged, since any 'lemon'-related costs will be recouped by the seller. Alternatively, reputation and culture can accomplish the same thing: If lying about the state of your car branded you a liar, the cost of dishonesty might be high enough for well-informed sellers to be completely upfront about the state of their goods.
Back to Krugman: He posits a 'lemon' kind of market. If health insurance is offered at a fixed price, the only people with an incentive to buy it are the ones who expect to get sick. For everyone else, it's a bad deal. Of course, if only secretly-sick people buy health insurance policies, the health insurance companies need to ratchet up rates to compensate. The higher their rates, the sicker their customers, until they give up. Or until they try options 1) or 2).
1) As applied to health insurance would be to let customers share their health data with potential insurers, so both parties were equally informed. Short of a Dreaded Lump or Disturbing Ache that the patient hasn't shared with doctors, it would be pretty difficult for one side to deceive the other if they're all looking at the same information. This, of course, would raise the price of insurance for potentially sick people and lower it for healthy people (which sounds bad until you think of it in terms of any other market: Charging for gas by the gallon raises the cost of gas for people who use their Hummer for 50-mile commutes, and lowers it for people who bike). It's pretty fair for people to pay as much for healthcare as their healthcare costs providers -- we can't effectively prioritize unless the cost to us represents our cost to others.
2) Is also easy enough to map onto the healthcare market. Not only can health insurance companies purchase reinsurance (which, because reputations are centralized and information is shared, doesn't have the same 'lemon' issues as the individual market), but they can offer bonuses to customers who follow the right health guidelines. This keeps customers from 'lemonizing' themselves after they've gotten a good deal on an insurance policy. Culture is already starting to make inroads on unhealthy behavior: From indie documentaries to nightly sitcoms, thinness is second only to bored cynicism in our virtue pantheon.
What's truly irksome about Krugman's essay, though, is that it ignores performing the same kind of analysis on its own ideal. If healthcare is socialized, "Adverse selection" is simply replaced by "Moral hazard": If people don't suffer for bad health decisions, they're going to make more of them. Being healthy is its own reward, of course, but that doesn't free us to ignore financial constraints: There are identifiable rewards to healthy food, exercise, and avoiding stupidly risky behavior. If we banish those incentives, we're going to be worse off.