Harvest of Lemons

SYNOPSIS: Krugman explains the work of three Nobel Prize winners in the economics of incomplete information

What does a Princeton degree have in common with a peacock's tail? Both are ornaments that demonstrate their possessor's quality. And the winners of this year's Nobel Memorial Prize in Economic Science showed, among other things, why it may be worth acquiring such ornaments, even at considerable cost.

One of the downsides of this column is how often I must argue against really bad ideas. So it's a nice change of pace, especially in these fraught times, to write about the really good ideas of this year's laureates. Recalling the wonderful papers of George Akerlof, A. Michael Spence and Joseph Stiglitz reminds me why I became an economist in the first place.

All three men focused on the complications a market economy faces when information is "asymmetric" — that is, when sellers know something buyers do not, or vice versa.

Mr. Akerlof started the field with his classic paper "The Market for Lemons." He pointed out that sellers of used cars — and many other items — are often better informed than potential buyers about the quality of those items. This means that the selection of goods actually made available is biased toward low quality: the owner of a trouble-free car is less likely to sell than the owner of a lemon. This in turn means that the buyers of used cars demand a substantial discount, which further discourages owners of good cars from reselling them. Rational buyer suspicion therefore prevents deals that would have benefited both buyers and sellers: the invisible hand drops the ball.

So what's the answer? Loosely speaking, Mr. Stiglitz looked at how those with information can be persuaded to reveal it. For example, insurance companies use deductibles to screen clients: those who know themselves to be good risks are willing to accept policies with larger deductibles. But this screening comes at a cost: in order to get a good rate, an insurance buyer must accept incomplete protection.

Mr. Spence looked instead at how those with inside information can credibly convey it. For example, a student who knows himself to be smart and motivated can signal those qualities to potential employers by graduating summa cum laude at an elite school; the demonstration of his quality may be more important than what, if anything, he learned. (In "Liar's Poker," Michael Lewis offered a meaner but similar analysis: he declared that would-be investment bankers studied economics in order to demonstrate their willingness to engage in boring, humiliating activities.)

About those peacocks: evolutionary theorists believe that one of the main reasons for extravagant displays like the peacock's tail is that they demonstrate a male's fitness to skeptical females. The tail is, of course, a handicap in other matters; that's why it's a credible signal. Biologists spent years arguing bitterly about whether the "handicap theory," introduced in the mid-1970's, made sense; they could have saved themselves considerable effort if they had realized that their theory was identical to the theory of "market signaling" Mr. Spence had introduced to economics some years earlier.

This is all fun stuff, but does it have any policy relevance? You bet it does. To take only one example, asymmetric information is a key reason why it's important to police insider trading: potential stock investors will stay on the sidelines if they suspect that low prices, instead of representing buying opportunities, usually reflect bad news they haven't heard about.

Or to take a more pointed example: the theory of asymmetric information tells us why the Bush administration's plan for prescription drug coverage under Medicare — remember that? — wouldn't have worked. That plan was based on the claim that coverage could be provided on the cheap by subsidizing insurance companies. But people are likely to know more than the insurance companies that cover them about their future drug costs; this puts retirees seeking drug insurance in the same position as people trying to sell used cars. Insurance companies would set rates high, denying coverage to most people, out of the rational suspicion that their clients would consist disproportionately of "lemons" — people with high drug expenses.

Of course, I don't expect politicians and lobbyists to understand such arguments; as Upton Sinclair said, it's difficult to get a man to understand something when his salary depends on his not understanding it. But there will be plenty of occasions for cynicism in future columns; for now, let me simply celebrate an inspiring economics Nobel.

Originally published in The New York Times, 10.14.01