Asymmetric Information

Information is important in economics and, most often, we have assumed that the agents have perfect information. That is hardly a reasonable assumption. For instance, usually a seller knows more about a product than the buyer does, and a worker knows her skills better than the employer does. We will now look at some implications of the problem of asymmetric information. There are two important subcategories:

• Adverse selection. Depending on the fact that one side in a contractual agreement, the buyers or the sellers, have information that the other part does not have, only some buyers or sellers will want to enter into the contract. Only the ones that will profit the most from the contract will do so. Moreover, those are, typically, the ones the other part wants to avoid.
• Moral hazard. Sometimes, one’s counterpart cannot check whether one fulfills one’s obligations after having agreed on a contract. One may them be tempted to exploit the other’s lack of knowledge. Note the difference between the two types: Adverse selection is about what happens before the agreement has been made. Moral hazard is about what happens after it has been made.

Adverse selection

We will give two classical examples of adverse selection: the market for insurance and the market for used cars. Note, however, that the concept is possible to apply on many types of goods and services.

Insurance

The price of insurance largely depends on the probability that the insurance firm will have to pay, for instance, on the probability that your bike is stolen. If there is a high probability, the price of insurance will also be high. Different people differ in how well they keep after their belongings, and the risk that a careless person will get her bike stolen is much higher than that a carful person will get hers stolen. However, the insurance firm cannot see a difference between careless and careful people, and therefore charges them the same price corresponding to an average of the risks. This, however, makes the insurance a good deal for the careless people, but it might make it too expensive for the careful people. They will probably not lose their bikes anyway. Then only the careless people remain; the ones that constitute a high risk for the insurer. When the insurer realizes that all people buying insurance are high-risk people, they will have to increase the price even more. The high-risk people will then have pushed the low-risk people out of the market, even though the latter might be fully willing to pay for insurance.

Used Cars

Suppose there are 100 used cars in a market, and that they are of two different levels of quality: Half of them are of high quality (H-cars) and half are of low quality (L-cars). The sellers want at least 50,000 for an L-car and at least 100,000 for an H-car, whereas the buyers are prepared to pay at most 60,000 for an L-car and 110,000 for an H-car. There are consequently possibilities for trades that are beneficial for both sides. If there had been two submarkets, one for L-cars and one for H-cars, people could have negotiated prices between 50,000 and 60,000 for L-cars and between 100,000 and 110,000 for H-cars. However, if they are sold in the same market, the buyers cannot tell them apart. Neither can she ask the sellers, as all sellers would say that their car is an Hcar. If the chance that she will get an L- or an H-car is 50% each, the buyer could think of this as a lottery. Suppose, for simplicity, that the buyer is risk neutral (so that she does not demand a risk premium for taking a risk). She would then be willing to pay the expected value of the car, i.e.

She will then maximally offer 85,000. However, at that price no seller is prepared to sell an H-car. Their lowest price for an H-car is 100,000. Consequently, they withdraw the H-cars from the market and only sell L-cars. Then, however, the probability of getting an L-car is no longer 50%, but instead 100%. Since the buyer realizes this, she is prepared to pay a maximum of 60,000 for a used car, and the L-cars have pushed the H-cars out of the market. This outcome is not efficient, since there are cars that the buyers have a higher valuation for than the sellers do, but that are not traded.

Signaling and How to Reduce Problems with Adverse Selection

There are several ways to reduce problems with adverse selection:

• Legislation. For instance, one could demand that sellers have to reveal the ingredients of (food) products. Thereby, buyers gain more information and we get less asymmetric information.
• Demand more information. Insurers often demand, for instance, a medical examination before selling insurance.
• A firm could acquire a reputation for quality. The cost of selling an L-car as if it was an H-car, i.e. lying about the product, would then be too high, since that would damage the reputation. Therefore, the customers know that all the seller’s cars are H-cars.
• One could also offer a warranty for the cars. Since the probability that an L-car will break down is much higher than that an H-car will do that, a seller of L-cars cannot offer the guarantee. Thereby the sellers sort themselves into two groups, and for L-cars and one for H-cars.

The last two examples are variants of signaling. The idea with signaling is that the agents themselves signal to which group they belong. It is, of course, not enough that they say that they belong to a certain group. It must be a signal that the low-quality group cannot afford, so that truth telling is optimal.

For more help in Adverse selection click the button below to submit your homework assignment