The Market for 'Lemons': Quality Uncertainty and the Market Mechanism
Source: https://doi.org/10.2307/1879431 ↗
Full text: open-access source ↗
Akerlof's foundational demonstration that information asymmetry alone can destroy a market.
Using the used car trade as his model — where sellers know whether a car is good or a "lemon" but buyers cannot tell — he shows that when quality is unobservable, buyers offer average prices, good sellers exit, and the market spirals toward low quality.
The paper launched the economics of asymmetric information and earned Akerlof the Nobel Prize.
For product direction the implications are immediate: any market where buyers cannot evaluate quality before purchase — SaaS, AI systems, consulting, platform services — is structurally vulnerable to adverse selection.
Read alongside Spence's Job Market Signaling for how costly signals restore separation, Stigler's Economics of Information for the search-cost foundation, and Erlei et al.'s When Life Gives You AI for the contemporary application to AI markets.
Central argument
Akerlof demonstrates that information asymmetry between buyers and sellers can unravel a market entirely, even without fraud or irrationality. In the used car market, sellers possess private knowledge about vehicle quality that buyers lack; unable to distinguish good cars from lemons, buyers offer only a price reflecting average quality. Owners of above-average cars then withdraw rather than sell at a discount, which lowers average quality further, which depresses prices further — a recursive collapse that can drive the market to extinction. The paper establishes adverse selection as a structural property of any market where quality is unobservable before purchase.
Critique
The model assumes a relatively static and homogeneous buyer population with no learning over time — yet repeated interaction, reputation systems, and platform feedback mechanisms can substantially attenuate adverse selection without requiring the costly formal signals Akerlof implies are necessary. The used car market, his empirical anchor, has not collapsed; warranties, dealership reputation, and later services like vehicle history reports all emerged endogenously to solve the problem, suggesting the model better describes an equilibrium tendency than an inevitable outcome. A thoughtful reader might argue the paper overstates market fragility and underweights the institutional creativity that markets generate in response to information gaps.
Why it matters for product
For a CPO, the lemon dynamic maps directly onto any SaaS or AI product sold to buyers who cannot assess quality before committing — enterprise procurement, platform marketplace onboarding, or AI feature launches where outputs are opaque. If your product's true quality is unverifiable at the point of purchase, you are structurally competing against worse products on price rather than value, which invites a race to the bottom in positioning and margin. The actionable implication is that product strategy must treat credible quality signalling — certifications, transparent metrics, free trials with meaningful exposure, case studies with verifiable outcomes — not as marketing decoration but as a structural defence against adverse selection collapsing your segment.