Adverse Selection

Health insurers face a problem: people who expect to get sick are more eager to buy insurance. If the insurer prices for average health, the healthiest people find it too expensive and drop out. The pool gets sicker; premiums rise; more healthy people leave. At the limit, only the very sick buy insurance at prices only the very sick would pay.

This is adverse selection: when one side of a transaction has better information, the market attracts disproportionately bad risks. The sellers of lemons know they’re selling lemons. The borrowers who accept high rates know their risk is even higher. The asymmetry of information skews the population toward the party that knows they’re getting the better deal.

Adverse selection differs from moral hazard. Moral hazard is behavior change after protection — the same person acts differently because consequences are reduced. Adverse selection is composition change in who participates — different people join because they know something the other side doesn’t. Moral hazard is about the insured becoming careless; adverse selection is about the careful not becoming insured.

The mechanism generalizes beyond markets. Job openings with low pay attract people who can’t get better offers. Volunteer committees attract those with axes to grind. Social programs attract those who benefit most. Any system where participation is voluntary and information is asymmetric risks selecting the wrong population.

Solutions attack the information asymmetry. Insurers require medical exams. Employers require references. Buyers inspect goods. Signaling and screening mechanisms try to reveal private information before transaction. But complete revelation is often impossible or too costly. Some adverse selection persists.

The deeper lesson: voluntary participation isn’t neutral. It filters. The population that chooses to engage isn’t a random sample of all possible engagers. Ignoring this guarantees surprise at who shows up.