Back to the Main ET Page

The Market for Lemons

  Arvind Panagariya

   IN MY write-up on the Economics Nobel laureates last year, I began by stating that if you wanted to predict future laureates, your best bet was to look at the list of the recipients of the John Bates Clark Medal, awarded every other year to the best American economist below 40 years of age. This year, I successfully applied this model, predicting all three Nobel laureates correctly and surprising a figure no less than Professor Assar Lindbeck, the chairman of the Nobel Committee from 1980 to 1994. Lindbeck and I happened to be together in Manila around the time the prize was announced.

 Daniel McFadden, the Clark Medal recipient in 1975, was awarded the Nobel Prize last year. Martin Feldstein, the 1977 recipient of the medal, has not appeared prominently on the recent lists of potential Nobel laureates. Lindbeck, who knew the winners in advance, also seemed to confirm that my other favourite for the prize, Jagdish Bhagwati, was not the winner this year.

  Therefore, I chose to put my money on Joseph Stiglitz and Michael Spence, winners of Clark Medal in 1979 and 1981, respectively, and both pioneers in the field of asymmetric information. But they had to be paired with George Akerlof, who is credited with founding the field of the economics of asymmetric information.

  In its citation, the Swedish Academy rightly describes Akerlof’s 1970 essay, “The Market for Lemons” as “the single most important study in the economics of information.” In it, Akerlof offers the revolutionary idea that when agents on one side of a transaction are better informed, some markets may entirely fail to emerge.

  To explain how, consider the market for used cars. Suppose these cars can be divided into two identical types: high quality and low quality. Buyers value high-quality cars at $10,000 and low-quality cars (lemons) at $8,000. Sellers value the same cars at $9000 and $7000, respectively. As long as both buyers and sellers are able to distinguish between the two types of cars, mutually beneficial transactions take place at prices between $9,000 and $10,000 for high-quality cars and between $7,000 and $8,000 for low-quality cars.

  Suppose, however, that the agents are “asymmetrically” informed such that sellers know the type of each car sold while buyers only know the proportion of high-quality cars in the total number of cars on the market. Assuming this proportion to be 1/4th, buyers value each car at (1/4)(10,000)+(3/4)(8,000) = $8,500. But this being less than $9,000, demanded by sellers, no high-quality cars are sold. The market is left with “lemons” only.

  Spence demonstrates that under certain circumstances, informed agents may be able to “signal” the buyers of the quality of their products and, thus, ensure that high-quality cars stay on the market. To see how, suppose a typical seller can add features such as new paint, tyres and brakes to the car that are viewed by the buyer as signalling high quality. By assumption, the cost of adding these features rises more sharply for low-quality cars than for high-quality cars. In particular, suppose the cost is $1,000 for high-quality cars and $4000 for low quality cars. Assuming the buyer attaches no extra value to these features, he is still willing to pay $10,000 for the car. This is just enough to cover the minimum price of $10,000 (= $9,000+$1,000) demanded by the seller and the car is sold on the market.

 Can the seller of the low-quality car take advantage of the signal to pass a “lemon” as a high-quality car? For the specific set of numbers chosen, the answer is in the negative. If he adds the features, he must be paid $11,000 (=$7000+$4,000) to be compensated fully. But this is more than $10,000 the buyer is willing to pay for high-quality cars. The seller chooses not to add the features and, instead, lets the low quality of his car be revealed. Both types of cars are sold in the market.

  Stiglitz offers an alternative solution in which uninformed agents manage to separate the two types of sellers via two types of contracts. Thus, suppose the low-quality car is known to breakdown once in the first year while the high-quality car is not. Under one contract, the buyer offers to pay $5,000 upfront plus $5000 at the end of the year if the car does not breakdown (ignore interest costs for simplicity). Under the second contract, he offers to pay $8,000 upfront plus $1,000 at the end of the year if the car does not breakdown. Sellers of high-quality cars take the first contract while sellers of low-quality car take the second one. Once again, both types of cars are sold.

  Applications of these basic ideas are numerous. Akerlof himself applied the “lemons” idea to explaining his observation during a visit to the Indian Statistical Institute in 1967-68 that moneylenders in rural India were able to charge interest rates twice those prevailing in large cities. He argued that anyone trying to arbitrage between the two rates without knowing the creditworthiness of borrowers was doomed to the adverse selection problem whereby he would attract only those borrowers likely to default (“lemons”). Additional questions tackled by these ideas relate to the IT bubble and insurance contracts.

  Akerlof is regarded among the most original economists. He outlines his approach to economics in “An Economic Theorist’s Book of Tales” (Cambridge University Press, 1984). “Economic theorists, like French chefs in regard to food, have developed stylised models whose ingredients are limited by some unwritten rules,” he writes. “Just as traditional French cooking does not use seaweed or raw fish, so neo-classical models do not make assumptions derived from psychology, anthropology, or sociology. I disagree with any rules that limit the nature of the ingredients in economic models.”

  Economic Times, October 16 2001