You are buying a used car, and the seller knows something you do not: whether it is sound, or a lemon. He will not tell you — or rather, he will tell you it is sound whether it is or not, so his word is worth nothing. Unable to tell the good car from the bad, you offer a price somewhere in the middle, fair, you think, for the average. But the man with the genuinely good car will not sell it at a price meant for the average, and withdraws; and now the average is worse, and your fair price falls, and the next-best car withdraws in turn. The market, left to itself, threatens to unwind to the worst car on the lot — undone not by greed or scarcity, but by a single fact: that one of you knows, and the other does not.
The situation is more general than the lot of used cars on which it is most often staged. It arises whenever one party to a bargain knows something material that the other does not and cannot cheaply discover — the condition the economists call asymmetric information. The seller of a house knows of the damp behind the new plaster; the applicant for life insurance knows of the cough he has not mentioned; the borrower knows whether he means to repay; the worker knows whether he will trouble to work. In each case one side can see a quality that bears directly on the value of the deal, and the other must act in the dark, pricing what it cannot inspect and trusting what it cannot verify. The used car is merely the cleanest illustration of a predicament that runs through half the transactions of economic life: that we are forever buying things whose worth is plain to the seller and hidden from ourselves.
George Akerlof, in a paper of 1970 that a doubtful journal twice rejected before it became among the most cited in the discipline, traced what such hiddenness does to a market. Suppose cars are either good or bad — lemons, in the American idiom — and that sellers know which they hold while buyers know only the proportions. A buyer facing a car that might be either will pay at most its average worth, weighing the good against the bad. But that average price is an insult to the good car and a windfall to the lemon, so the owners of good cars, unwilling to sell below their worth, keep them; the cars left for sale are therefore worse than the original mix, the rational buyer lowers his offer to match, and at the new price a further tranche of the better cars withdraws. The reasoning repeats, each round driving out the best of what remains, until in the limit the market clears only lemons, or fails to clear at all. Akerlof named the engine of this collapse adverse selection: the systematic tendency of the bad to crowd out the good wherever quality is private and price cannot tell them apart.
What makes the result so unsettling is that it requires no villain. There is no monopolist throttling supply, no cartel fixing terms, no pollution fouling a neighbor's air; there need not even be a liar, for the mechanism works though every seller is scrupulously honest, since honesty cannot help a good seller who has no way to prove that his word is true and the lemon-seller's false. The market fails not because anyone misbehaves but because of the structure of what is known — an inequality of information, and nothing more. And the failure is a genuine loss, not a mere redistribution: the good cars exist, and there are buyers who would gladly pay what they are worth, so that trades both sides would welcome simply do not happen, the gains from them destroyed by the impossibility of telling which car is which. A market can leave willing buyers and willing sellers apart, with value on the table that no one can reach, for want of nothing but the means to verify.
Economists knew the pattern at once, for it was an old one in new dress. Four centuries before Akerlof, the same logic had been seen in money: when full-weight coins circulate beside clipped and debased ones that the law compels all to accept at equal value, the sound coins are hoarded or melted and the debased do the work of exchange — the principle that bears Thomas Gresham's name, turning on the very mechanism of the lemon, since the holder of the good keeps it and spends the bad. Once the structure is seen, the lemon haunts every market where quality hides. Health insurers offering one premium to all draw the sick and repel the well, so the premium climbs and the well depart, in the spiral that empties a voluntary pool from its healthy end. Lenders who cannot sort safe borrowers from reckless post a rate the safe find dear and the reckless cheap, and keep the reckless. Employers who cannot see diligence, and pay all alike, give their best reason to leave. Wherever the careful cannot set themselves apart from the careless, the careless inherit the market.
And here the price tells its usual lie. A price is a verdict on a hidden contest, mistaken for a fact of nature — and the lemon price is a verdict of exactly this kind. The depressed sum a good used car commands, the steep premium the insurer demands, the high rate the unknown borrower must pay, the wage that fails to reward the able: each presents itself as the natural price of quality under uncertainty, as though uncertainty were a fact of the world and the discount it imposes a law of nature. But the discount is nothing of the kind. It is the price the good is forced to accept because it cannot prove that it is good — the toll levied not by nature but by the distribution of knowledge, by the brute circumstance that the seller can see what the buyer cannot. Change what can be known and shown, and the price moves; the good car's worth has not altered, only its ability to demonstrate it. What looks like the market's valuation of quality is in truth its valuation of provable quality, and the gap between the two is the measure of the hidden contest.
It will be objected that the markets do not, in fact, collapse — that used cars are bought and sold every day, that insurers and lenders stay in business, and that the lemon model therefore describes a catastrophe the real economy mostly avoids. The objection is correct, and it is the hinge on which the rest of this volume turns. For markets are not passive before their own asymmetry; they breed devices to defeat it — the warranty a sound seller can afford and a shyster cannot, the brand that stakes its reputation on consistency, the independent inspection, the dealer who lives on repeat custom, the law that compels disclosure — and these do much to hold the lemon at bay. Yet three things should be said before the objection is allowed to dissolve the result. The first is that every such device is costly, a real expenditure undertaken for no purpose but to fight an asymmetry a world of common knowledge would never impose — a standing tax on trade, invisible because we take it for granted. The second is that the devices are imperfect, narrowing the gap without closing it, so that some good cars still go unsold, some risks uninsured, some markets half-collapsed at the edges. And the third, which matters most, is that the cure no more belongs to nature than the disease: whether a market drowns in its lemons or is rescued by its signals, the outcome that results — the price, the terms, who trades and who is shut out — is the work of an information structure and the institutions built around it, and not a fact of the world at all.
This is what the lemon establishes for everything that follows. Information is not a free and even background against which exchange takes place, but a scarce and unequally held thing that enters into every price — an asset to the one who has it, a hazard to the one who lacks it, and a prize fought over by both. The buyer's ignorance is not a private misfortune; it is a structural feature of the market, as consequential for what things cost as any scarcity of goods. And the lemon works a quiet revolution in the very idea of quality, for in a market where quality cannot be seen, the price attaches not to quality but to what can be proven, and the unprovable good — however genuine — is bought and sold as though it were bad. The whole apparatus this volume will examine, the signals and screens and seals and mandates, is in the end an apparatus for making quality provable, for letting the good say credibly what it is. The lemon is the wound; the rest of the volume is the suture.
For this is where the matter comes to rest, in the barest of all the volume's contests. A price is a verdict on a hidden contest, mistaken for a fact of nature; and in the market for lemons the contest is stripped to its elements — the seller's private knowledge set against the buyer's exposure, with nothing else in play. The verdict it returns, the lemon price, is read as the worth of a used car in an uncertain world, as though the uncertainty were written into the nature of things; but the uncertainty is only a distribution of knowledge, a fact about who has been told what, and a warranty or a certificate or a law could redraw it tomorrow. To call the lemon price natural is to forget the simplest thing about it: that somewhere a person knows what another does not, and that this, and not the order of nature, is what sets the terms on which they meet. The seller knows. The buyer fears. The price is the distance between them, and the distance is no law of nature.
In a market where quality cannot be proven, the price is paid not for what a thing is worth but for what its seller can show — and the good, unable to show it, is sold as though it were bad.
So the volume begins where its whole argument is in miniature. An inequality of knowledge, alone — no monopoly, no spillover, no malice — is enough to bend a market and, at the limit, to break it; and the price that emerges from the wreckage is no fact of nature but the residue of who could know and show what. Everything that follows is an answer to the lemon. The signal will let the good seller spend his way to credibility; the screen will let the buyer set a test the types sort themselves by; the seal will lend a stranger's trust; the mandate will compel the truth into the open. Each is a way of making the hidden visible, of letting quality speak where it could not, of redrawing the distribution of knowledge that the lemon price merely records. The seller knows; the buyer fears; and the work of every institution in this volume is to close, or to bridge, or at least to price, the distance between the two. The argument remains open.