By Tim Harford Posted Saturday, April 29, 2006, at 8:17 AM ET
I recently did something that is, in theory, most unwise: I bought a secondhand car. Since economists hate to compromise between safety and style, it was a Volvo. You'd think I would know better. The American subtitle of my book is, "Why you can never buy a decent used car."
In 1966 an assistant economics professor, George Akerlof, tried to explain why this is so in a working paper called "The Market for 'Lemons.' " His basic insight was simple: If somebody who has plenty of experience driving a particular car is keen to sell it to you, why should you be so keen to buy it?
Akerlof showed that insight could have dramatic consequences. Buyers' perfectly sensible fears of being ripped off could, in principle, wipe out the entire used-car market: There would be no price that a rational seller would offer that was low enough to make the sale. The deeper the discount, the more the buyer would be sure that the car was a terrible lemon.
More plausibly, only the market for cheap, shoddy used cars would survive. A person with a good car would hold onto it because he couldn't prove it was good and so wouldn't expect an attractive offer for it. And if the good cars aren't put up for sale, the lemons will be what is left. This is a problem not just for buyers, but for sellers, too, who wish they could be trusted.
Many people recognize that they are reluctant to sell their own car if it's running well. That is informal evidence for Akerlof's thesis, since if buyers were able to appreciate the qualities of each car, good cars would fetch more money and it should be no more attractive to sell a cheap lemon than an expensive peach.
But the used-car market hasn't disappeared, and so economists have debated how much Akerlof's model really explains the market. For instance, there's been a lively controversy as to whether pickup trucks that are sold secondhand have higher maintenance bills than pickup trucks of the same age that are not sold. (If they do, that's evidence in favor of Akerlof's "lemons" model.)
But that is not really the point, because used cars were just the beginning for Akerlof. His neat little paper was turned down by two top journals because they couldn't see past the trivia of his example. He recalls that a third, the Journal of Political Economy, had a better reason for rejecting him: The paper couldn't be true, because if it were true then economics would be turned on its head.
The Journal of Political Economy was half right. Akerlof did turn economics on its head—and eventually received the Nobel Prize for doing so—not by documenting the travails of used-car buyers and sellers, but by showing how corrosive a little bit of inside information can be to all sorts of markets. Insurance, including health insurance, is one possible casualty. If you eat well, exercise, and have parents who lived to 100, you might expect a big discount from your health insurer. If you do not get the discount, you may decide to take your chances without health insurance, especially if you are not getting a subsidy from an employer. And if, as a good health risk, you decide not to insure yourself, the average cost of premiums will have to rise. Other people may decide to join you. The same goes for loans and mortgages, too: Some people simply cannot get a loan, no matter how much interest they offer to pay. Progressively higher interest rates scare off increasing numbers of creditworthy borrowers, leaving only the bad risks.
Then there's the market for jobs. How many of your colleagues are lemons? If you're competent but can't prove it to your boss, you may prefer to be a freelancer. If other competent workers think that way, it may explain why you think your colleagues are idiots and they think the same about you.
These problems can be solved, but at a cost. Mortgage companies can cut out high-risk borrowers, but at the cost of refusing credit to whole neighborhoods. Employees spend years acquiring qualifications with little value other than proving that they're smart and work hard. And used-car buyers will look for trusted sellers, even if that raises the cost of doing business. I bought my Volvo from my brother-in-law because I thought that would lower the risk of being sold a lemon.
Not only are the solutions costly, they don't always work. My Volvo lasted less than a fortnight before the clutch burned out: a bitter little lemon after all.
The Undercover Economist appears on Saturdays in the Financial Times Magazine.
At least on the Fray, the consensus appears to be that used-car shopping is a rational activity—provided a buyer makes an adequate investment in information before purchasing. Fray Editor wonders if we're only seeing half the picture. An '86 Chevy Nova is certainly hard to sell at any price, no matter how well it runs. –G.A.
Remarks from the Fray:
I suppose at the time the Lemon Theory was written, it could be considered true that the only used cars available were ones that weren't fit to drive any more, so they were sold or traded in. What the lemon theory fails to take into account nowadays, is how many people consider cars akin to fashion accessories and trade out of them when something else catches their eye. Not to mention the cars that are sold from dealers lots that were given to salespeople to drive, and have very few miles on them. That's how we got our used car--three years in, and no problems so far. I'm sure it's fair to say that there are more lemons than not on dealer's lots, but if you know what you're looking at, there are great deals to be had. The Undercover Economist did nothing wrong by trying to buy used--his mistake was buying from a family member--that pretty much assures you of getting screwed.
[B]uying a low mileage one or two year old used car, after somebody else has eaten the depreciation, is the smartest thing you can do financially. […] The day a new car is driven out of the dealer it's market value drops by several thousand dollars, and a year or so later by almost as much. The most sensible financial decision is to buy a low mileage used car. Many dealers offer warranties on used cars they sell. Without reading his book, Harford sounds to me like he's spouting theory based on a crackpot economic model rather than the real world.
[F]ew sellers actually know how long their car will last. It's a guessing game. A good buyer can inform himself, and stands a good chance of finding a car which only needs some cheap repairs, but which the seller thinks is on its last legs.
Even a little knowledge can upset the Dr. Akerlof's lemon thesis. In fact, even without knowledge on the part of the buyer, the fact that the seller can't predict the future of his car pretty much destroys the thesis. The connection between the seller's evaluation of his car and it's eventual lifetime is tenuous at best.
When [the author turns his] attention to health insurance, however, they hit on a really important fundamental problem. This is an area where "inside information" is routinely gathered and used to decide who gets what product and at what price. Anyone who has a "pre-existing condition" is either excluded or charged out the wazoo. Ultimately, everyone who is likely to be unprofitable from the insuror's point of view (including everyone over 65, unless they can afford a huge premium) is dumped onto the public half of the health care market, where you and I pay for their care.
Armchair economists love to defend our expensive, exclusive health care "system" as a "free market" and decry attempts to rectify it as "socialized medicine." I am pointing out that the more expensive half of our present system is precisely that - socialized medicine, paid for by the taxpayer.
Isn't the point of the article (and the basis for the Nobel Prize) really that since the lemon theory obviously fails, other factors must be at work? In particular, one of the most important factors is inside information (e.g. watch out for a particular year BMW 740 - bad block casting, certain years of Ford 3.6L V6 - continual head gasket failure, there's a huge market for GM LS1/2/6 engines, etc.).
In the economists' dream world of uniform consumer knowledge and rational purchasing and selling decisions, not only are hideous boxes like Volvos considered stylish (the S40 is an exception), but the lemon theory should hold.
Since by definition inside information is not owned by everyone (though not necesarily unavailable to everyone) people will make purchasing and selling decisions that given their level of knowledge make sense - including value judgments based on stylishness, reliability, economy, family situation, etc. on which there is not universal agreement.
Those with additional useful knowledge are able to use that knowledge to extract the remaining value of a vehicle sold by someone who lacks that particular knowledge or doesn't share a value judgment on style/performance/utility.