An Empirical Test of the Lemons Model
George Akerlof famously pointed out that there are two possible outcomes when sellers have better information than buyers. The first is that bad products (or lemons) will drive down the market and force high-quality products out. This is because buyers will be unable to distinguish between good and bad products and sellers are incentivized to sell bad products instead of good, driving average quality down. The second possibility is that the market may develop responses to counteract this information asymmetry. This usually involves some (costly) way of signaling quality. This could take the form of warranties, rating systems, branding, etc.
In this study by Eric Bond, the lemon model is tested in the used pickup truck market. Bond hypothesizes that if the lemon model holds this market is likely to be flooded with “high maintenance trucks” that have become costly to maintain (assuming there are not strong counteracting quality signaling measures in this market).
If this hypothesis is true then a sample of trucks that have been traded in the used truck market should have higher maintenance costs than a sample of similar trucks that have not been traded.
Looking at a 1977 truck inventory and use survey (census data from trucks registered at the DMV in all fifty states), Bond determined that (after age and lifetime mileage are controlled for), there is no difference between the two samples of trucks. This rejects the initial hypothesis, but leads to two possible explanations: either there is some quality signaling mechanism in this market, or buyers are able to obtain enough information on their own to overcome the asymmetry.
While Bond does not explore either of these two ideas he does offer a third possibility. Suppose that the initial hypothesis is true that trucks are sold on the secondary market when they become too expensive to maintain. This can still be true if the cost of ownership varies for different types of owners. For example, there may be “high maintenance” truck owners, who are not as knowledgeable about their vehicle, and for whom it is more expensive to maintain. These are the likely sellers in the used truck market. There may also be a set of “low maintenance” truck owners who are more knowledgeable and are able to make repairs themselves. These are the likely buyers in the used truck market. This would result in a transfer of ownership from “high maintenance” owners to “low maintenance” owners. This suggests that the used truck market is simply a transfer of ownership from users of low value to users of high value.
While additional research is needed to verify any of these explanations, this market still serves as an interesting case study of a potential lemons market.
http://www.jstor.org/stable/1810022?seq=1#page_scan_tab_contents