Chapter 18 Asymmetric Information 351
©2014 Pearson Education, Inc.
David Genesove (1993) observed this difference in adverse selection at wholesale used–car auctions. Both
used–car dealers, who sell only used cars, and new-car dealers, who sell both new and used cars, attend
these auctions.1 Dealers trade because they receive too many or too few used cars in trade-in, have the
wrong mix of cars for their local retail customers, or are trying to unload lemons.
Retail customers, who have time to inspect and test-drive cars, are more likely to spot a lemon than a
dealer at one of these auctions. At the auctions, potential buyers have only a few minutes to inspect a car
and cannot drive it, so they don’t know if the car is a lemon. After these quick visual inspections, bidding
takes place. The owner then accepts or rejects the highest bid.
New–car and used–car dealers differ substantially in their likelihood of selling a given car. There are three
principal differences between new– and used–car dealers. First, new–car dealers tend to sell to final customers
only recent-vintage used cars that are close substitutes for new cars. Nearly 60 percent of the used cars
sold by new-car dealers are no more than four years old, compared to only 30 percent for used–car dealers.
Second, new-car dealers obtain a higher share of their used cars through trade–ins: 69 percent versus 26
percent. Third, new-car dealers obtain more used cars in trade–ins. Used–car dealers are relatively small: 85
percent obtain fewer than 100 cars in trade–ins and only 2.5 percent obtain 300–600 cars a year. In contrast,
only 7 percent of new–car dealers obtain fewer than 100 cars, 33 percent receive 300–600, and 14 percent
obtain over 600 trade-ins a year.
Because new–car dealers have more trade–ins and sell relatively few older cars, they are more likely to
want to sell their older cars at the wholesale auctions. Indeed, new–car dealers are twice as likely to sell a
six–year–old trade-in at wholesale. Because of this greater tendency to unload all old cars, they are less
likely to engage in adverse selection and keep the higher-quality, old cars.
Thus if the seller is known to be a new-car dealer, we would expect buyers to be willing to pay more for
their older cars. Indeed, new-car dealers receive a premium of 17 percent ($400) over other dealers when
they sell six-year old or older cars. Thus buyers use the identity of the seller as an imperfect signal of
quality.
Similarly, buyers apparently view the number of previous owners as a signal about whether the car is a
lemon. A one-owner car sells for 9 percent ($330) more than multiple-owner cars.
Thus by using information about the number of former owners and the current owner, potential buyers
adjust their estimates of the quality of a car. Although these signals decrease the probability of paying
too much for a lemon, they do not eliminate that possibility.
1. If buyers had more time, or if the auctioneers would certify quality, the asymmetry could be
eliminated. Why doesn’t that happen?
2. Why are extended warranties on used cars sold for an additional fee rather than building the cost into
the price and including them on all sales?
Good Customers Only2
Brad Anderson, chief executive officer of Best Buy Co., embraced a heretical notion for a retailer recently.
He wanted to separate the “angels” among his 1.5 million daily customers from the “devils.”
1 David Genesove, “Adverse Selection in the Wholesale Used Car Market,” Journal of Political Economy, 101(4),
August 1993:644–665.
2Gary McWilliams, “Minding the Store Analyzing Customers, Best Buy Decides Not All Are Welcome,” Wall Street Journal,
November 8, 2004; Page A1.