Context: insurance, risk management, microeconomics
Adverse selection (also known as antiselection) refers to the tendency where people who are more likely to incur a loss to acquire insurance. Thus the pool of risks that are borne by an insurance company increases in riskiness as time goes on.
As a concrete example, consider motor insurance. Suppose that premiums are set assuming that the insured will make a claim once every two years. For bad drivers who make a claim once every year, then the premium charged by the insurance company is less than what they should have paid. Thus these bad drivers are attracted to take out insurance policies at these lower premiums, while better drivers would be drawn away from taking the same policies.
Along with moral hazard, adverse selection is a risk that an insurance company that wants to avoid ruin must manage. Rapid increase in the riskiness of the pool means a rapid increase in both the number and the amount of claims that needs to be paid out, increasing the likelihood of ruin.
The main strategy in combating adverse selection is underwriting. Here the insurance company wants to know as much as possible the riskiness of the insured so that it could decide whether to insure, and at what premiums to charge. The aim is to detect bad risks and charge them higher premiums or to refuse insurance at all.