Adverse Selection

In 1970, George Akerlof published "The Market for Lemons," a thirteen-page paper that nearly didn't get published — three journals rejected it as trivial — and that won him the Nobel Prize in Economics in 2001. The core argument: when one side of a transaction knows more than the other, the market doesn't just underperform. It collapses. Not slowly. Systematically. The mechanism is elegant, brutal, and everywhere once you see it.

Start with used cars. A seller knows whether their car is reliable or a lemon. A buyer cannot tell the difference. Because the buyer cannot distinguish good cars from bad, they price every car as if it might be a lemon — offering a blended price that averages the value of good cars and bad ones. That blended price is too low for sellers of good cars. They withdraw. Now the remaining pool is worse. Buyers adjust downward. More good sellers leave. The average quality drops again. Buyers adjust again. The spiral continues until only lemons remain — or the market disappears entirely. The information gap didn't just create inefficiency. It destroyed the market from the inside out.

Insurance is the textbook case. People who know they are sick buy health insurance at higher rates than people who know they are healthy. The insurer cannot perfectly distinguish between the two groups. So premiums reflect the average risk of the insured pool. Healthy people look at the premium, decide it's too expensive relative to their actual risk, and drop out. The remaining pool is sicker. Premiums rise. More healthy people leave. The pool gets sicker. Premiums rise again. Actuaries call this a death spiral, and it is not metaphorical — it is the mathematical consequence of adverse selection left unchecked. The Affordable Care Act's individual mandate was a direct structural response: force healthy people into the pool to prevent the spiral from initiating.

The hiring market runs the same logic. A company with high turnover loses its best employees first — they have the most options. The remaining workforce is, on average, less capable. Prospective hires observe the turnover rate and infer that something is wrong. The best candidates avoid the company. The company hires from a degraded applicant pool. Performance drops. More good employees leave. The talent market has adversely selected against the company, and the company's own churn rate is the signal that accelerates the selection. Akerlof's insight applied to labour: when the best options can leave and the worst options cannot, every exit makes the average worse.

Akerlof's solution — and the solution that markets have independently evolved across every domain — is signalling. A warranty on a used car signals that the seller believes the car won't break. A college degree signals that the candidate invested four years and significant resources into capability development. Equity vesting schedules signal that the company expects the employee to find the next four years worth staying for. The signal works because it is costly to fake: a seller of a lemon cannot profitably offer a warranty, a low-capability candidate cannot easily obtain a degree from a selective institution, and a company with a toxic culture cannot retain employees through vesting alone. Michael Spence formalised this in his signalling theory, sharing the 2001 Nobel with Akerlof. The signal doesn't transmit information directly. It transmits the willingness to bear a cost that only a high-quality party would accept — and that willingness is the information.

Venture capital is an adverse selection arena that most participants don't recognise as one. The best startups — the ones with the strongest traction, the clearest vision, the most experienced teams — have the most fundraising options. They can choose their investors. The startups that cannot choose their investors are the ones that no preferred investor wanted. An investor who wins a deal should always ask: why did I win? If the answer is "because no one else offered," the investor has been adversely selected. The best VCs counter this with their own signalling: brand reputation, value-add services, portfolio network effects. The signal says "choosing us is itself evidence of quality, because only founders with options would choose a fund this selective." The signal filters the pool before adverse selection can degrade it.