Intermediatemicroeconomics
Asymmetric Information: Lemons, Adverse Selection, and Signaling
Overview
Asymmetric information — when one party to a transaction knows more than the other — can unravel markets entirely. George Akerlof's 'market for lemons' showed how hidden quality drives good used cars off the market; insurance markets suffer adverse selection as high-risk buyers flock to generous policies; and hidden actions create moral hazard. This lesson explains these three problems, then shows the market's defenses: signaling, screening, warranties, and reputation.
Learning objectives
- Define asymmetric information and distinguish hidden characteristics from hidden actions
- Explain Akerlof's lemons model and the adverse-selection death spiral
- Describe adverse selection in insurance markets
- Define moral hazard and give principal-agent examples
- Explain signaling, screening, and other market responses
Hidden characteristics vs. hidden actions
Asymmetric information means one side knows something relevant the other does not. Hidden characteristics (adverse selection): the seller knows the used car's true quality; the insurance applicant knows their health risks; the job candidate knows their work ethic — before the contract is signed. Hidden actions (moral hazard): the insured driver takes less care; the hired manager slacks off; the borrower gambles with the bank's money — after the contract is signed.
The distinction organizes the economics of information. Adverse-selection problems are about sorting: attracting the wrong types. Moral-hazard problems are about incentives: encouraging the wrong behavior. Both can destroy gains from trade, and the remedies differ — disclosure and screening for hidden types, incentive contracts and monitoring for hidden actions.
The market for lemons
Akerlof's 1970 model: used cars come in good and bad ('lemon') quality, sellers know which, buyers do not. Buyers will pay only the average expected value — but at that price, owners of good cars (worth more than average) withdraw, leaving a worse pool. Buyers lower their offer, more good cars exit, and the spiral can end with only lemons traded, or no market at all.
The logic generalizes wherever quality is hidden: credit markets (lenders fear the riskiest borrowers self-select), labor markets (firms fear the least capable apply hardest), even academic publishing. Markets do not merely misprice under asymmetric information — they can shrink or vanish, a far deeper failure than a wrong price. Akerlof's insight helped earn the 2001 Nobel Prize shared with Spence and Stiglitz.
Adverse selection in insurance
Insurance is adverse selection's natural habitat. A generous health policy at an average premium attracts the sick (who gain most) and repels the healthy (who overpay) — the pool worsens, premiums rise, more healthy people exit: the death spiral. Insurers respond with the standard defenses: medical underwriting, waiting periods, deductibles and copays (which screen out low-risk-averse buyers less), and risk-rated premiums.
This is the economic logic behind insurance mandates: the Affordable Care Act's individual mandate (and similar schemes worldwide) forces the healthy into the pool, stabilizing premiums. Without compulsion or clever screening, voluntary insurance markets for high-risk events chronically unravel — which is why flood, crop, and long-term-care insurance often need government backing.
Moral hazard and the principal-agent problem
Moral hazard arises when being insured — or unmonitored — changes behavior: the comprehensively insured driver parks carelessly, the salaried manager coasts, the bailed-out bank takes wild risks. The principal-agent problem is the general form: a principal (shareholder, insurer, employer) hires an agent (manager, policyholder, worker) whose effort is unobservable, so contracts must align incentives.
Solutions tie payoffs to observable outcomes: deductibles make the insured share losses; performance pay and stock options link managers to firm value; efficiency wages make job loss costly. Each trades off risk-sharing against incentives — full insurance kills incentives, full incentives kill insurance. Optimal contracts balance the two, which is why real insurance always has deductibles and real jobs mix salaries with bonuses.
Signaling: the informed reveal themselves
The informed side can sometimes credibly reveal quality through signaling — taking a costly action that low-quality types would not mimic. Spence's job-market model: education signals ability if schooling is costlier (in effort) for low-ability workers; the degree separates types even if it teaches nothing. Warranties signal product reliability: only a manufacturer confident in quality can afford generous guarantees.
For a signal to work, it must be differentially costly — cheap talk ('trust me, it's a great car') convinces no one. Advertising can signal quality too: only firms expecting repeat purchases waste money on brand-building. Signaling equilibria are often wasteful (burning resources to prove type), but they beat market collapse — the relevant comparison.
Screening, reputation, and regulation
The uninformed side fights back with screening: designing choices that induce self-revelation. Insurers offer menus of deductibles — the cautious pick full coverage, the daring pick high deductibles — sorting risk types without observing them. Employers use probation periods and interviews; lenders use collateral requirements and credit scores.
Reputation sustains quality where transactions repeat: sellers who cheat lose future business, so the threat of lost reputation disciplines hidden actions — the logic behind brands, platforms' review systems, and professional licensing. Where markets still fail, regulation mandates disclosure: securities laws, food labeling, and lemon laws for cars force revelation the market cannot elicit. Information economics thus spans from eBay ratings to the SEC — all institutions for making hidden things visible.
Akerlof (lemons/adverse selection), Spence (signaling), and Stiglitz (screening and information in markets) shared the 2001 Nobel Prize for founding the economics of asymmetric information.
Key points
- Asymmetric information: one party knows more — hidden characteristics or hidden actions
- Akerlof's lemons: hidden quality can shrink or destroy markets via adverse selection
- Insurance death spirals motivate mandates, underwriting, and deductibles
- Moral hazard: insurance/monitoring gaps change behavior after contracting
- Signaling (warranties, education) credibly reveals hidden quality at a cost
- Screening (menus, probation), reputation, and disclosure regulation complete the toolkit
Common misconceptions
✓ Truth: It can destroy markets entirely. When buyers cannot distinguish quality, good sellers exit and trade collapses to lemons or nothing — a qualitative failure, not a pricing error.
✓ Truth: Signaling models show education can raise wages without raising productivity — but that is a statement about one mechanism, not proof schooling teaches nothing. Most evidence says education both builds skills and signals them.
✓ Truth: It is about incentives facing normal people: anyone takes less care when someone else bears the cost. Seat-belted drivers, insured homeowners, and tenured managers all respond to changed incentives — predictably, not immorally.
Practice questions
Q1. In Akerlof's lemons model, the market unravels because:
- sellers are irrational
- buyers offer average quality prices, driving good cars out and worsening the pool
- the government bans used-car sales
- cars depreciate too fast
Unable to verify quality, buyers pay expected-average value. Owners of above-average cars withdraw, the pool's average falls, offers fall further — the adverse-selection spiral that can leave only lemons or no market.
Q2. An insurance 'death spiral' is caused by:
- moral hazard among doctors
- adverse selection: rising premiums drive out healthy buyers, worsening the risk pool
- government price controls
- too many insurers competing
At average-based premiums the sickest buy first; their costs raise premiums, the healthiest exit, the pool worsens further. Each round of exits raises the next premium — the classic adverse-selection unraveling.
Q3. Which is an example of moral hazard rather than adverse selection?
- A lemon-seller hiding defects before sale
- An insured driver taking less care after buying full coverage
- A sick person buying generous health insurance
- A low-ability worker getting a degree
Moral hazard is hidden action after contracting: coverage changes the driver's behavior. The other options involve hidden characteristics known before the transaction — adverse selection or signaling.
Q4. For education to work as a signal of ability in Spence's model, schooling must be:
- free for everyone
- more costly (in effort) for low-ability workers than for high-ability workers
- unrelated to wages
- mandatory by law
Differential cost is what makes the signal credible: if degrees were equally easy for all, low-ability workers would mimic them and the signal would collapse. The cost gap sustains the separating equilibrium.
Click a question to reveal the answer and explanation.
Frequently asked questions
What is the difference between adverse selection and moral hazard?
Adverse selection is hidden characteristics before contracting — the wrong types select into the deal (lemons sold, sick insured). Moral hazard is hidden actions after contracting — behavior changes because risk is shifted (careless driving once insured).
How does a warranty solve the lemons problem?
A warranty is a differentially costly signal: a seller of a reliable product can offer generous coverage cheaply, while a lemon-seller cannot afford it. Buyers rationally infer quality from the warranty offered, restoring trade.
Why do insurers use deductibles?
Deductibles serve two purposes: they screen risk types (cautious buyers accept low deductibles, risk-tolerant ones choose high) and they curb moral hazard by making the insured share each loss, preserving incentives for care.
What is the principal-agent problem?
A principal hires an agent whose effort or actions are unobservable — shareholders vs. managers, employers vs. workers, insurers vs. policyholders. Contracts must balance incentives against risk-sharing, typically via performance pay, monitoring, or efficiency wages.