This is a review article of the original paper by George Akerlof on the above mentioned topic.

There exist markets in which buyers judge the quality of prospective purchases from market statistic. This enables incentive for the sellers to put out poor quality commodities for sale since the return for good quality accrue to the entire group as opposed to any individual seller. To this end, there seem to be a reduction in both the relative quality of goods as well as the market size. For instance, suppose there are four kinds of cars – new, old (used), good or bad (a lemon) car. A new car maybe any of good or a lemon and so may a used car. Individuals buy cars without knowing if they would buy a good car or a lemon, but then they have some level of expectations for either good or a lemon. These expectations are revealed by the respective probabilities and 1-q respectively. After a while using the car, the owner assigns new levels of probabilities to the car’s status (good or a lemon) which is more accurate than the previous probabilities. The discrepancy in these expectations create an asymmetry since the sellers are more knowledgeable about the cars’ quality than the buyers. Both quality of cars – good or lemon would sell for the same price since the buyers cannot differentiate between either, thus being advantageous to sell a lemon at the price of a good car. After the lemon is sold, a new car would be bought at a newer probability of being good and a corresponding probability -q of being a lemon.

The market for lemons could be reflected in the cost of dishonesty, the employment of minorities, insurance markets, and even credit markets in under-developed countries.

In analysing Gresham’s law, provision is made for bad cars driving out good cars since the buyers cannot differentiate between either. This though is not applicable for money since those who acquire money can tell the difference between both. In a thread of good to bad cars, bad cars could drive out relatively bad cars, which could drive out the medium cars and then the relatively good ones till the good ones could be out. This would mean there would not be a market for cars. The demand for used cars could depend by assumption, on two factors which include price, p and average quality, u while the supply and average quality would depend on price such that at equilibrium, supply would equal demand [(p) = D(p, u(p))]. A fall in the price would reduce the average quality of the used car, that there is possibility that no good is traded at any price.

In insurance, those who insure themselves would be those who are increasingly sure that they would need an insurance policy even as the price level rises. Patient-related factors like the doctors’ sympathy with older patients make it much easier for applicants of insurance policies to assess the risks involved than the insurance firms, thus allowing the average medical condition of insurance policy seekers deteriorate as price levels rises. This would imply that older patients would be less likely to be given an insurance policy at any price, thus an associated problem of adverse selection. The employment of minorities is also a reflection of the lemons principle. This is analogous to the explanation in signalling and indices. People from minority races may not get employment not for reasons of prejudice but for the purpose of profit maximization by employers. The argument is that race may bear an infringement on the applicant’s social background, quality of schooling and overall capabilities. Quality schooling could serve as a substitute to this as an indicator when the students are graded. An untrained worker may possess natural talents, but these must be certified by an educational system for employers to value them. To this end, the certifying educational institution must possess credibility since unreliability of the institution is correlated with the economic possibility of graduating students. Similarly, employers may not employ applicants from minorities because of the difficulty in ascertaining those with good job qualifications from those with the bad qualifications, such that the reward for work in slum schools seem to accrue to the entirety in the group, rather to the individual.

The underlying model could also be of importance in reviewing the cost of dishonesty. Suppose we have a market in which goods are sold honestly or dishonestly, quality maybe represented or not. For the consumers, the basic problem would be to identify quality, whereas the likelihood of people offering low quality or inferior goods would seem to drive the market out of existence as earlier argued in the case of the automobile market. This is what represents the major cost of dishonesty. The cost of dishonesty would thus be predicated by the existence of bad dealers who would rather pawn out bad wares [though at ruling market prices] as opposed to potential buyers and sellers of good quality products. This action pushes out legitimate business. The cost of dishonesty is thus measured not only as the amount by which the purchaser is cheated but also by the loss incurred from driving out legitimate business.

Quality uncertainty is counteracted upon by some institutions some of which include guarantees, brand-name good and even business chains. Guarantees are given to ensure the buyer of some form of quality over the purchased good, such that the seller bears risk over the product’s quality. Brand names also indicate quality but the consumer reserves the right to retaliation by curtailing future purchases if quality is not met. Business chains also ensure quality to some extent. Consumers would rather patronize an already known business outfit than wanting to try out new ones, especially where quality has been established in the past.

The above article is a review piece of the article below. Ideas contained therein are also credited to the authors cited within in base article.

Akerlof, George A. (1970). The Market for "Lemons": Quality Uncertainty and the Market Mechanism. The Quarterly Journal of Economics, 84(3); 488-500