The Contract

A contract is, at bottom, a promise the law will enforce; and the place to begin is with the strangeness that a promise should bind at all. A promise is nothing but words — a sound in the air, a mark on a page — and it concerns the future, which has not yet happened and over which the speaker has no certain power. How, then, does the saying of words now create an obligation that grips the speaker tomorrow? The older world had an answer: the oath. A promise bound because the one who made it had pledged his honor, sworn before his God or his fellows, and to break it was to be forsworn — dishonored, shamed, perhaps damned, perhaps the object of a wronged party's vengeance. The bond of the word was the swearer's own person, his standing in the eyes of those who mattered to him, and the promise held because the cost of breaking it fell on who he was.

The contract is the descendant of the oath, but the source of its binding force has changed, and the change is everything. A contract binds not because the promisor's honor is staked upon it but because the law will make him perform. Behind the contract stands the whole apparatus of the state that The Rule anatomized: the court that will hear the complaint, the judgment that will issue against the breacher, and, in the last resort, the officer who will seize his goods or garnish his wages to satisfy it. The bond of the word has been socialized — lifted out of the keeping of the individual's honor and placed in the keeping of an institution that will make the promise good whether the promisor, when the time comes, wishes it or not. This is the first thing to fix about the contract, and the thing that makes it an institution rather than a mere agreement: that its force is not in the paper, nor in the characters of the parties, but in the public power that has undertaken to enforce the promises its citizens make to one another.

That promising itself is a human institution, and not a fact of nature, is a point the philosophers reached long before the economists. David Hume called promising an artificial virtue — artificial not in the sense of false, but in the sense of made, a convention humanity adopted because it serves us, not a duty written into the structure of things. There is, on this view, no obligation of promises antecedent to the practice of promising; the practice creates the obligation, and we maintain the practice because a world in which words can bind the future is a world in which we can do vastly more together than a world in which they cannot. The contract takes this useful convention and gives it the further force of law, so that what began as a moral practice sustained by conscience and reputation becomes a legal institution sustained by the courts. At the root of the law of the deal, then, lies an invention — the invention of the binding promise — and like the other inventions of this group it is a thing made to make cooperation possible.

What the socialized promise makes possible, above all, is the cooperation of strangers — and this is why it matters so for the economy. The plates that follow will trace what the contract does and where it fails, but the foundation is here: that by lodging the enforcement of promises in an impersonal institution, a society lets its members rely on one another without the personal knowledge that honor required. I need not know your character, your kin, or your faith to contract with you; I need only know that the same law binds us both. The bond of the word, once the most personal of things, has become the most impersonal — and on that transformation rests the extended market, the market of strangers, that no economy of honor among intimates could ever have reached.

What the binding promise gives us is commitment, and to see why commitment is precious we must look at the problem it solves — the oldest problem of cooperation across time. Suppose two parties could both gain from an exchange in which one performs first and the other repays after: the worker labors and is paid at the month's end, the merchant ships and is settled on delivery, the lender advances and is repaid in a year. The trouble is that the second party, once the first has performed, is tempted to keep the benefit and give nothing back; and the first party, foreseeing this, declines to perform at all. The cooperation that would have enriched them both is strangled by the very rationality of the parties, each reasoning correctly about the other. The figure draws this in its barest form.

The contract is what breaks the trap. By making the second party's promise enforceable — by lowering, through the threat of damages, the payoff to reneging below the payoff to performing — it removes the temptation that strangled the deal, so that the second party reciprocates, the first party trusts, and the cooperation that pure reason forbade becomes the course that pure reason now commends. The contract does not change the parties' selfishness; it changes their incentives, rearranging the payoffs so that keeping the promise is the selfish thing to do. And in doing so it unlocks the entire structure of deferred and conditional exchange on which a developed economy depends: the investment made today against a return promised tomorrow, the credit advanced against repayment, the long employment, the trade across oceans and seasons. The capacity to make a binding promise is the hinge on which intertemporal cooperation turns, and an economy without it would be confined to the hand-to-hand exchange of the marketplace, a poor thing beside what the binding promise makes possible.

Seen from the side of the promisor, this capacity wears a paradoxical aspect, and the paradox is the plate's title. To make a binding promise is to give up a freedom — the freedom to change one's mind, to renege when reneging would profit. Yet the power to surrender that freedom is itself a freedom, and frequently the more valuable one. Thomas Schelling built a science on the observation that the power to constrain another may rest on the power to bind oneself, and Homer had seen it long before: Ulysses, who longed to hear the Sirens yet knew their song was death, had himself lashed to the mast and his oarsmen's ears stopped with wax, and so could do what no free man could — hear the song and live. He bought an experience by surrendering, in advance, his freedom to act on it. The contract is the mast of economic life, the device by which we bind ourselves so that we may safely be relied upon, and so may do together what no party free to renege could ever undertake.

This inverts the picture of freedom we are apt to carry into economics, and the inversion matters for everything that follows. We tend to equate freedom with the keeping-open of options and to see every promise as a cage. But the party who can only ever choose, and never commit, is shut out of every cooperation that asks one side to rely on the other's future conduct — which is most of economic life. The borrower who might repudiate at will cannot borrow; the firm that might always renege cannot raise capital or hire on long terms; the nation that might always retreat cannot deter. To bind the future is to make the future bankable, and the contract is the institution through which we do it. The plates that follow will examine the machinery by which the law makes promises bind, the gaps that machinery cannot reach, and the power that hides within the form of free agreement; but the foundation is this — that the binding promise is less a constraint upon freedom than the instrument of a deeper freedom, the freedom to give one's word and have it count.

Not every promise is a contract, and the law has always needed some way to mark the promises it will enforce off from the promises it will not. The common law's chief device is the doctrine of consideration, which holds that a promise binds only when it is part of a bargain — given in exchange for something, the consideration, which is the price of the promise and the thing the promisor gets in return. A promise for which nothing is given back — the gift promise, the gratuitous undertaking to hand over a sum or render a service for nothing — is, as a rule, not enforceable, however solemnly it was made. The law will compel you to keep a promise you sold; it will leave you free to break one you merely gave. To bind the law to your word, you must have been paid for it.

The rationale is not that the gift promise is less sincere, but that the law of contract is the law of the market, and its proper business is exchange. The bargain requirement performs several offices at once. It marks the serious from the casual, for a promise made as part of a deal, with something staked on each side, is unlikely to be the idle word of a moment. It separates the domain of exchange, where mutual benefit is the rule and enforcement the natural support, from the domain of gift and affection, where generosity is its own warrant and the law's coercion would be an intrusion. And it leaves evidence, for a bargain involves an exchange that can be seen and proved, where a bare promise leaves only the contested memory of words. Before consideration did this work, formalities did it — above all the seal, the wax and the signature that marked a promise as binding regardless of any exchange. Lon Fuller saw that such formalities serve three functions that the law still needs: they are evidentiary, proving a promise was made; cautionary, making the promisor pause and feel the weight of binding himself; and channeling, signaling that this utterance, unlike the day's other words, is meant to be a legal act.

Yet the bargain principle, for all its sense, can work injustice, and the law has grown a second principle beside it to catch what it misses. Sometimes a person stakes much on a promise that was freely given — gives up a job, spends a sum, alters his life in reliance on an assurance for which he paid nothing — and to let the promisor walk away unscathed offends justice plainly. The doctrine of promissory estoppel answers this: a promise that foreseeably induces reasonable reliance may bind, even without consideration, so far as is needed to prevent injustice. With it, reliance joins bargain as a ground of obligation, and the enforceable promise is no longer only the promise that was bought, but also the promise that was trusted. The history of contract, as Atiyah told it, is in part the rising and falling rivalry of these two principles, the bargain and the reliance, each claiming to say why a promise should bind.

Beneath the doctrines lies the question they answer without ever quite settling: why does the law enforce promises at all? Three answers contend. The will theory says the law enforces a promise because the promisor willed it so — bound himself by his own free choice — and the law does no more than hold him to his self-imposed obligation, out of respect for his autonomy. The reliance theory says the law enforces a promise to protect the one who relied on it, sparing him the harm of a trust betrayed. And the economic theory says the law enforces promises because doing so sustains the institution of binding promise on which all intertemporal cooperation depends — that we enforce promises not chiefly to honor wills or shield reliers but because a world where words can bind is a world in which we can all do more. The law's actual rules are never purely the issue of any one of these; they are a shifting compound of all three. And the proportions it strikes — how far it enforces to honor the will, how far to protect reliance, how far to serve the general welfare — are a choice, remade in every generation, about what the binding of the word is finally for. The boundary of the enforceable promise is drawn, not found.

A promise binds, the plates so far have said; but what does the law actually do when a promise is broken? The answer reveals more about the nature of the obligation than any amount of talk about its solemnity. The law's usual response to breach is not to compel the breacher to perform but to make him pay — expectation damages, a sum of money calculated to put the promisee where performance would have left him, to give him the benefit of his bargain in cash since he cannot have it in kind. The law could, in principle, order specific performance, commanding the breacher to do the very thing he promised, and for unique things — a particular parcel of land, an irreplaceable painting — it sometimes does. But for most contracts it does not; it sets a price on breach and leaves the breacher free to pay it. To enforce a contract, in the ordinary case, is not to insist that it be kept but to compensate the disappointed party when it is not.

Oliver Wendell Holmes drew the deflationary conclusion with a bluntness that still startles: the duty to keep a contract at common law, he wrote, means a prediction that you must pay damages if you do not keep it, and nothing else. On this view the contractual obligation is not a command to perform but an option — perform, or pay the promisee's loss — and the promisor who breaches and pays has done all the law asks of him. The figure shows why an economist would see in this not a defect but a virtue. Let the promisee value performance at some amount, and let the promisor's cost of performing rise — because, say, a third party has appeared who wants the promised goods more. So long as the cost of performing stays below the promisee's value, performance creates value and the promisor performs. But once the cost of performing exceeds the value to the promisee, performance destroys value, and the efficient course is to breach.

Expectation damages are what make the private decision to breach track the social interest. Because the breacher must pay the promisee's full loss, he keeps the gain from breaching only when that gain exceeds the loss he must cover — which is exactly the case in which breach moves the goods to someone who values them more, leaving the promisee whole and the world richer by the difference. The seller who has promised widgets to one buyer for a hundred, and is offered a hundred and thirty by another who needs them more, ought to breach, pay the first buyer the ten he lost, and let the widgets go where they are worth most. The law's preference for damages over forced performance, on this account, is not a weakness in its respect for promises but a mechanism for keeping resources flowing to their highest use, with the promisee protected by the damages and no one made worse off.

Against this stands an older and stubborner intuition, that a promise is meant to bind and not to be priced. To treat one's word as an option, kept or broken as the ledger dictates, is — the moralists of contract insist — to drain it of the very quality that made it a promise; the seller who breaches has not merely reallocated goods efficiently but broken his word, and the cash he pays does not undo the breach so much as buy his way out of it. Nor is the economic story as clean as it looks, for expectation damages routinely undercompensate: they rarely cover the cost and delay of litigation, the losses too diffuse to prove, the value to the promisee that no court can price, or the trust that breach destroys. If damages fall short of the true loss, then breach the law calls efficient may not be, and the promisor who breaches captures a gain that the promisee, undercompensated, has partly paid for. The dispute is not merely philosophical, for the two views recommend different remedies — the economic, damages and the liberty to breach; the moral, performance and the duty to keep — and the law's choice between them settles what a contract most deeply is. Where the remedy is damages, the contract is an option; where it is performance, the contract is a bond. The bindingness of the binding promise turns out to be not a fact about promises but a decision about remedies.

We come now to the fact from which this volume took its question, and around which the law of the deal is truly organized: that no contract can foresee everything, and every contract is therefore incomplete. The reason is not laziness or haste but the structure of the world and the mind. The futures that may befall a contract are too many to enumerate — the price that triples, the supplier that fails, the war, the flood, the regulation, the thousand contingencies no one thought of until they arrived. The human mind, as Herbert Simon insisted, is boundedly rational, able to foresee and to reckon with only a fraction of what may come. Even the contingencies one could foresee are mostly too costly to address, the lawyer's hours and the negotiator's patience exhausted long before the rare cases are reached. And even those one troubles to specify may prove unenforceable, for a court can act only on what it can verify, and much that the parties themselves can see — the effort given, the quality intended, the spirit of the thing — no court can confirm. The complete contingent contract of ideal theory, which names an act for every state of the world, is a fiction as thorough as the frictionless market, and impossible for the same kind of reason.

It follows that incompleteness is not merely a failing to be regretted but, within limits, a choice to be made well, and the figure shows the choosing. Each additional contingency a contract addresses yields some value — the protection it affords, the dispute it forecloses — but the value falls as one proceeds, for the likeliest and weightiest cases are specified first and the rest grow ever rarer and slighter. Each addition also costs something — the foreseeing, the drafting, the haggling — and the cost rises, for the contingencies that remain are the subtle and improbable ones, hardest to imagine and to pin in words. The rational contract is written up to the point where the marginal value of specifying one more thing just meets the marginal cost, and there it stops, leaving everything beyond deliberately blank. The gap is not an oversight; it is the region the parties wisely chose not to fill, because filling it would have cost more than it was worth.

The consequence is that a contract is never the whole of the parties' relationship but only its visible skeleton, and that something must govern them in all the situations the document left blank — which is to say, in most situations that will actually arise. What governs them there is the subject of the plates immediately following. Part of it is the law itself, which supplies default terms where the parties were silent, reads in an obligation of good faith, and asks what reasonable parties would have intended — a vast unwritten contract beneath the written one. Part of it is trust and reputation and the value of the continuing relationship, which lead parties to deal fairly in the gaps without any compulsion to do so. And part of it, as The Firm showed, is ownership — the residual control that decides the unwritten when nothing else does. The contract proper is only the first and smallest of these; the others do the larger share of the work.

The gap that incompleteness opens is, moreover, not merely a blank to be filled but a danger to be managed, for it is exactly where opportunism lives. The Firm traced the holdup that the unspecified situation invites — the party who exploits the silence, who reads the gap in his own favor when the other is locked in and cannot easily walk away. Incompleteness and opportunism together make the central problem of contracting that Williamson placed at the heart of his economics: that because we cannot specify everything, and because some will exploit what we did not specify, the design of contracts and of the institutions around them must aim less at the impossible completeness than at safeguarding the parties through the gaps that must remain. The contract is necessarily a sketch; the art is in how the blank spaces are governed, and the gravest error in thinking about contracts is to mistake the skeleton for the body — to suppose that what the parties owe each other is exhausted by what they wrote, when what they wrote was never meant, and never able, to be complete.

The gap that every contract must leave has to be governed by something, and the first of the things that govern it is the law itself. Where the parties have fallen silent — which is to say, across most of the situations they will actually meet — the law does not throw up its hands but supplies a term, reading into the agreement the provision the parties did not write. These supplied provisions are the default rules of contract, and they make up the great bulk of contract law. They are to be distinguished from the mandatory rules, the terms the law imposes whatever the parties may say — the prohibition of unconscionable bargains, the consumer protections that cannot be waived — which we shall meet at the volume's limit. The default rule is gentler: it governs unless the parties have agreed otherwise, an off-the-shelf term they are free to replace but, in the ordinary case, do not.

The law fills the gap by several instruments. It implies terms — some implied in fact, as what the parties must have intended though they did not say it; some implied in law, as what justice or policy requires whatever they intended, like the warranty that goods be fit for their ordinary use. It reads into every contract a duty of good faith and fair dealing, an obligation not to exploit the silences opportunistically nor to defeat the other's reasonable expectations — which is, in effect, the law's general weapon against the holdup that the gaps invite. And where no more definite term will serve, it imports the standard of the reasonable: the reasonable price where none was fixed, the reasonable time where none was set, the reasonable efforts the party must make. The reasonable is the law's all-purpose gap-filler, the device by which it pours a community's settled expectations into the blank the parties left. And when a dispute finally arises in a gap, it is a court that decides — interpreting, implying, applying good faith and reason — so that the contract's incompleteness is resolved, in the last resort, by the ex post discretion of a judge. What The Firm called residual control, the right to decide what the contract did not, the law keeps a share of for itself.

Why the law sets its defaults where it does is a question with more than one good answer. The natural thought is that defaults should be majoritarian — set to what most parties would have wanted, the term the typical pair would have chosen had they bothered to negotiate — so that the law's standard deal saves the parties the cost of writing the common terms themselves. But Ian Ayres and Robert Gertner showed that the law sometimes does better to set a penalty default, a term the parties do not want, precisely in order to force the better-informed party to speak. If one side knows something the other does not, a default set against the knowing party gives it a reason to reveal what it knows by contracting around the default — and the revelation may benefit everyone. The default rule, on this view, is not merely a gap-filler but an instrument, a lever the law can pull to shape what the parties disclose and negotiate. Either way, the setting of defaults is a choice, and a consequential one, about what the law supplies when the parties say nothing.

The upshot is that the contract which actually governs the parties is far larger than the document they signed, and most of it was written not by them but by the law. The page bears the few terms that distinguish this deal from the standard one — the price, the quantity, the dates, the particular promises; everything else, the great unwritten remainder, is supplied by the default rules, the implied terms, the duty of good faith, and the standard of the reasonable that the law reads in. This rescues the contract from its own incompleteness, for if the parties had to write every term there could be almost no contracts at all; but it should temper the picture of the contract as a pure creature of the parties' will. The will draws the few lines that mark this deal as theirs; the law draws all the rest, against a background the parties did not choose and mostly never see. Yet the law is only the first and crudest of the gap-fillers, for courts are slow and costly and can act only on what they can verify, and much of what passes between contracting parties is beyond a court's reach entirely. For the subtle, the ongoing, and the unverifiable, the gap is filled not by the law at all, but by something the next plate must take up: the trust and reputation that govern the relationship the contract only partly describes.

The law fills the gaps it can reach, but it cannot reach far, and the larger share of the work of holding a contract together is done by something the law books scarcely mention: trust, reputation, and the value of the continuing relationship. The point was made unforgettably by Stewart Macaulay, who in the early nineteen-sixties went and asked businessmen how they actually used contracts, and found that they used them remarkably little. They planned their dealings loosely, left much unspecified, and when disputes arose they settled them by negotiation and forbearance rather than by the terms of any document; lawsuits were rare and faintly disreputable, a last resort that usually ended the relationship it was meant to vindicate. The formal contract sat in a drawer, drawn up and forgotten; what governed the parties day to day was the expectation of continued dealing and the simple understanding that a person's word was his bond. The law of contract and the practice of contracting, Macaulay showed, were two different things.

Ian Macneil built the observation into a theory. The classical law imagined the contract as discrete — a single, self-contained exchange, planned in advance and enforced at law — but most real contracts, he argued, are relational: embedded in ongoing relationships and governed by the norms of those relationships, by flexibility and reciprocity and the expectation of a shared future, far more than by their written terms. The long-term supply agreement, the employment, the franchise, the joint venture — these are not transactions but relationships, and the document that nominally governs them is at most a frame within which the relationship does its work. The contract, on this view, is not the paper but the relation; and the relation is held together not by the threat of the court but by the parties' stake in its continuance.

The mechanism by which that stake enforces cooperation is the shadow of the future, and the figure draws it. In any relationship expected to continue, a party who is tempted to cheat must weigh the gain from cheating once against the value of all the future dealings that cheating would destroy — for the other party, discovering the betrayal, will deal with him no more. Cheating yields a fixed one-time gain that scarcely depends on how the future is valued. Cooperating yields a stream of future benefit whose worth rises the more the future is weighted. Where the future weighs little, the one-time gain wins and cooperation breaks down; but where the future weighs heavily enough — past the threshold the figure marks — the value of the continuing relationship exceeds the gain from any single betrayal, and honesty becomes simply the profitable course.

What sustains the bilateral relationship, reputation sustains across a whole market. A party who cheats does not only forfeit the relationship he betrayed; he acquires a name, and in a trade where dealings are observed and talked of, a spoiled name closes other doors as well. The medieval merchants whom Avner Greif studied enforced honesty across great distances not by any court but by a network of reputation, in which the trader who cheated one was boycotted by all; the modern firm guards its brand for the same reason, because a reputation for fair dealing is an asset that cheating destroys. Reputation is the multilateral shadow of the future, and it lets trust extend even to parties who may never deal again, so long as each must deal with someone who will know. On this foundation of relationship and reputation rests the trust that is the true lubricant of exchange — the willingness to rely on another's performance without, or well beyond, any legal compulsion — and that some have counted, under the name of social capital, among the deepest determinants of which economies prosper.

None of this makes the law superfluous. It is the backstop that makes trust safe to extend, the recourse when relationships fail or when the parties are strangers with no shared future to protect; the market of strangers that the first plate described could not exist without it. But the daily operation of the law of the deal is, it turns out, mostly not law at all. The gaps that no clause could close — the unverifiable quality, the unforeseen contingency, the adjustment that fairness requires — are filled not by the court but by the relationship and the reputation that ride upon it. The deepest enforcement of promises runs not through the sheriff but through the simple, powerful fact that we must deal with one another again. The contract is the frame and the rarely-summoned guarantor; trust is what holds the ordinary economy together.

The plates so far have treated the contract as a meeting of two wills, a bargain struck between parties who each had something to offer and something to withhold. A great share of the contracts that actually govern modern life are nothing of the kind. They are contracts of adhesion: one party — the bank, the insurer, the employer, the platform — writes every term in advance, the same terms for everyone, and offers them take-it-or-leave-it; the other party does not negotiate but merely adheres, signing the form or clicking the box or, having no real alternative, simply using the service and being deemed to have agreed. The terms arrive in fine print that no one reads and that studies confirm almost no one reads — the arbitration clause, the liability waiver, the fee schedule, the license to one's data — and that the adhering party could not have changed had he read every word. The freedom of the adhering party is the freedom to accept the terms entire or to do without the thing, and where the thing is a necessity and the terms are much the same wherever he turns, doing without is no freedom at all.

To call this agreement, and to enforce it as the free concurrence of two minds, strains the very concept on which contract is supposed to rest. The doctrine of contract presupposes mutual assent, a meeting of the minds upon the terms; but in the adhesion contract there is no meeting and, as to the terms, scarcely any minds — one party dictates and the other submits, and the law treats the submission as assent because it must have some workable rule and cannot inquire into each transaction's reality. The fiction is convenient and largely unavoidable; a commerce of standard forms could not function if every term required real negotiation and real consent. But it is a fiction, and Friedrich Kessler named it as such eighty years ago: the standard form, he saw, is an instrument of the powerful, a way of legislating private law for one's customers under the forms of contract, and the language of free agreement disguises what is in substance the unilateral imposition of terms.

What makes this more than a technical curiosity is the use to which the rhetoric of free agreement has been put. Freedom of contract sounds like the very voice of liberty — the right of each to make whatever bargain he chooses, and of the law to hold him to it — and within its proper sphere it is. But it has repeatedly been wielded to defeat the protections that the weaker party most needs. In the era that takes its name from Lochner v. New York, the courts struck down maximum-hours laws, minimum-wage laws, and the early statutes of industrial safety, on the ground that such laws interfered with the worker's and the employer's sacred liberty of contract — as though the worker who must labor to eat and the employer who may hire a hundred others were equal parties freely meeting in the market. The realist critique, which the law absorbed only slowly and against resistance, is that formal freedom of contract among unequal parties produces substantive unfreedom: that to treat the strong and the weak as free and equal contractors is precisely to ratify the power of the strong, and that the form of consent then launders the substance of command. The employment contract that The Wage and The Firm examined, the lease, the loan, the terms of service — these are the places where the contract is most plainly a vessel of power, and least plainly a meeting of minds.

The law has not, in the end, enforced adhesion contracts entirely as written, and the next plate takes up the limits it has come to impose — the refusal to enforce the unconscionable term, the consumer protections that cannot be waived, the construction of ambiguities against the party who drafted them. These are the law's admission that agreement to dictated terms is suspect, and that the floor of fairness must sometimes override the form of consent. But the deeper lesson of the adhesion contract is that the contract, which the earlier plates showed to be the instrument of cooperation across time, is also the instrument through which economic power is exercised and, worse, legitimized. Where the line falls between free agreement and dictated terms — which contracts the law enforces as written, and which it polices, and which protective statutes it allows to stand — is not a question contract doctrine can answer from within. It is a political question, a recurring verdict on how much the form of consent may be permitted to clothe the substance of power.

The freedom of contract, however grand its rhetoric, has always stopped somewhere, and the places where it stops are worth a plate of their own, for they mark the moral perimeter of the market. Not every agreement the parties might wish to make is one the law will enforce. The figure arranges the possibilities along a spectrum, from the bargains the law enforces as written, through the bargains it enforces only subject to terms it will not let the parties waive, to the bargains it refuses to enforce at all — the void and the forbidden, the things that may not be made the subject of a contract however willing those who would make it.

The limits fall into kinds, each protecting something the bare enforcement of agreements would endanger. Some protect the conditions of free agreement itself: the contract signed under duress is voidable, the consent procured by fraud is no consent, for an agreement extracted by force or deceit is not the free act the law means to honor. Some protect against exploitation even where consent was formally given: the doctrine of unconscionability lets a court refuse to enforce a term so grossly one-sided, so oppressively imposed on a party with no real choice, that to enforce it would make the law complicit in the oppression. Some protect third parties and the public: the contract to commit a crime, the agreement that injures those who never signed it, the bargain in restraint of trade that The Power condemned — these the law holds void because their enforcement would harm the wider order the contract is supposed to serve. And some protect what must not be alienated at all if persons are to remain persons and citizens citizens.

This last category is the deepest, and the most revealing. There are things one cannot contract away however willing, because to allow their sale would undo the very status that makes the seller free to sell. One cannot bind oneself into slavery; the law will not enforce a contract of self-enslavement, for liberty is inalienable, and a freedom that could be sold would be no freedom at all. One cannot sell one's vote, for a democracy in which suffrage had a price would cease to be one. In most places one cannot sell one's organs, one's children, oneself. These inalienabilities mark the goods a society holds too fundamental to be subject to exchange, and beyond them lies the territory The Rule mapped under the name of the repugnant — the things that perhaps could be bought and sold, but that a society judges should not be, because the market is thought to corrupt them in the very act of pricing them, as Sandel and Satz argued of the goods that money ought not to buy.

What the limits share is that they are judgments of value, and because they are judgments of value the boundary moves. Surrogacy and sexual labor, the sale of organs and of blood and of risk, the trade in personal data, the waiver of rights once thought unwaivable — each is drawn differently in different places and contested everywhere, enforced as ordinary commerce here and forbidden as repugnant there, the line shifting with a society's sense of what dignity requires and what the market would spoil. The limit of contract is not a fact of nature awaiting discovery but a verdict a society passes on itself; and in deciding what may not be bought or promised, however freely, it decides in part what kind of persons, and what kind of polity, it means to be. The freedom of contract is real and precious within its sphere, but the drawing of the sphere's edge — the marking-off of the coerced, the exploited, the inalienable, and the repugnant — is among the most consequential moral and political acts a legal order performs, and the least reducible to the logic of agreement it bounds.

A different difficulty besets the contract when the parties know different things, and it is distinct from the difficulty The Firm met in its agent. There the problem was hidden action — the effort a principal could not observe; here it is hidden knowledge — the quality a buyer cannot see but the seller can. The trouble this makes is severe. When buyers cannot tell the good goods from the bad, they will pay only what the average is worth; at that price the sellers of genuinely good goods, unwilling to sell quality for a mediocre return, withdraw; their withdrawal lowers the average quality that remains, which lowers the price buyers will pay, which drives out the next tier of sellers — and the market unravels toward the lemons, in the spiral George Akerlof made famous. Asymmetric information, left to itself, can hollow out an exchange that would have benefited everyone, and the bare contract of a price for a good is powerless against it.

The contract escapes this fate by carrying devices whose whole purpose is to make hidden quality credible, and the warranty is the purest of them. A warranty is a promise to make good any failure of the goods — to repair, replace, or refund — and its beauty as a signal is that it costs a seller exactly in proportion to how often his goods fail. The figure shows why this makes it credible. To the maker of good goods, which rarely fail, even a generous warranty is cheap, for few claims will ever be made upon it; to the maker of bad goods, which fail often, the same warranty is ruinous. There is therefore a range of warranty — generous enough that the maker of bad goods cannot afford to offer it, yet still affordable to the maker of good goods — within which only the honest seller will go.

Because the warranty is costly in a way that tracks the truth, it conveys what no mere assertion could. Any seller may say his goods are good; only the seller whose goods truly are good can profitably back the claim with a promise that failure will be made expensive to him. The buyer, knowing this, believes the warranty though he would not believe the words, and the honest seller, by binding himself to a cost he can bear and his rival cannot, separates himself from the dishonest. The same logic, traced by Michael Spence for the labor market, runs through a family of contractual devices. There is the signal that only the high type finds worth sending. There is the screening menu, in which the uninformed party offers a choice of terms — a high deductible at a low premium, a low deductible at a high one — designed so that each type sorts itself by what it chooses. There is the brand, an asset built at long expense that the sale of bad goods would destroy, and so a standing hostage to quality. There is the bond posted and forfeited on cheating. Each converts the unverifiable into the credible by the same alchemy: it makes honesty cheaper than deceit.

What this reveals about the contract is that much of its apparent complexity is not the record of the exchange at all but machinery built around the information the parties do not share. The warranties and guarantees, the contingent terms, the menus and the deductibles, the elaborate representations and the bonds — these are the contract's response to the simple, stubborn fact that one party knows what the other cannot check, and that without some device to make knowledge credible the exchange would fail. The contract is thus shaped, in its very form, by the structure of information in the transaction, and to read a contract's terms is often to read a map of who knew what. The general anatomy of that information — the hidden action and the hidden knowledge that shadow every bargain, and the failures they breed — belongs to The Asymmetry, the element this plate can only gesture toward. Here it is enough to have seen that the contract is not merely the record of a deal but a structure built to make a promise believed by someone who cannot verify it — and that this, as much as the binding of the future, is among the quiet labors the institution performs.

The contract idea did not stay within commerce. In the hands of the political philosophers it was raised to its highest power and made to bear the weight of the whole political order, in the family of theories that go by the name of the social contract. Their question is the question of legitimacy: by what right does the state command, and why are we bound to obey? And their answer is the answer of contract: that political authority rests, at bottom, on an agreement — that we are bound to the polity because we have, in some sense, consented to it. The boldness of this is easy to miss from a distance of centuries. It located the ground of authority not in the will of God or the right of conquest or the order of nature, but in the agreement of the governed — and in doing so it made the legitimacy of every state a matter of whether it could be justified to those it ruled.

Thomas Hobbes drew the starkest version, and the figure draws Hobbes. Imagine, he said, a state of nature with no common power to keep men in awe: each, rightly fearing every other, strikes first, and the result is a war of all against all, in which life is solitary, poor, nasty, brutish, and short. The structure is the one this volume opened with, now written across a whole society — a trap in which each does worse by the very prudence that leads all to defect.

The escape, for Hobbes, is the social contract: men agree, each with all, to lay down their natural liberty and submit to a sovereign who will enforce the peace, and it is this agreement — hypothetical, but binding in its logic — that grounds the sovereign's authority and the subject's duty to obey. Later hands softened and redirected the idea without abandoning its form. John Locke gave the state of nature a law and rights of its own, and made the contract establish a government whose authority was conditional on its protecting life, liberty, and property, forfeit if it betrayed the trust — and so grounded both limited government and the right of revolution. Jean-Jacques Rousseau, declaring that man is born free yet everywhere is in chains, made the contract the union of citizens under a general will, and so grounded popular sovereignty and self-rule. And in our own age John Rawls remade the contract as a thought experiment, asking what principles of justice free and equal persons would choose from behind a veil of ignorance that hid their place in society — turning the contract from a story about the origin of the state into a test of the justice of its arrangements. Through all these transformations the core held: that the legitimacy of the political order is to be measured by its agreement with the will of those who live under it, a thought to which The Tax appealed when it grounded the fiscal claim in consent.

And yet the social contract, as a contract, is a fiction, and the fiction is the same one the eighth plate found in the standard form, written now across a nation. We never signed the contract that binds us; we were born into the state, never asked, and cannot in any real sense leave it. David Hume pressed the objection with finality: to say a citizen tacitly consents by remaining in his country is like saying a man carried sleeping aboard a ship consents to the voyage because he does not leap into the sea. The consent that supposedly binds us is hypothetical or tacit — a consent imagined, not given — and most people obey from habit and interest, not from any agreement they ever made. This does not destroy the idea, but it changes what the idea is for. The social contract was never the record of a historical bargain; it is a standing question, posed in the form of a hypothetical agreement — the question whether the order we inhabit is one that free and equal people could accept, whether the state can justify itself to those it binds. That, and not the false claim that we agreed, is the contract's deepest gift to political thought.

The metaphor must be handled with care, for the polity is not in truth a deal. One cannot negotiate the clauses, withhold one's signature, or take one's custom elsewhere; the parties do not pre-exist the society that supposedly made them; and our obligations to one another may rest on more than consent — on fairness, on community, on duties we did not choose. Pushed too hard, the contract model of the state misleads, dressing in the language of agreement a bond that agreement only partly explains. But handled with care it remains the most powerful instrument we have for thinking about legitimacy, and it inherits, at the scale of the polity, exactly the three features this volume has traced in the smallest commercial promise. Its logic is the logic of enforcement that makes cooperation possible; its consent is the fiction that the forms of agreement throw over a relation of power; and its incompleteness is the constitutional truth that no founding document can foresee the futures a people will meet, so that the social contract, like every contract, is a sketch forever being filled in by interpretation, convention, and trust. The contract, carried to its height, turns the whole political order from a fact we must endure into a question we may always press — which is the work this table means everywhere to do.

We began with a gap — the gap between the deed done and the return awaited, into which one party must step on the strength of another's word — and we end having seen what it takes to make that step safe. The contract is the institution of the enforceable promise: the device by which a society lets its members bind the future, make their words count, and cooperate across time with strangers they have no other reason to trust. It is the mast to which we lash ourselves so that we may be relied upon, the bond of the word lifted out of the keeping of honor and placed in the keeping of the law. And it is, on the showing of these plates, at once indispensable, incomplete, and contested — the three things that must be held together if the contract is to be understood at all.

Indispensable, because without the binding promise the whole structure of deferred and conditional exchange on which a developed economy rests would simply fall away, leaving only the hand-to-hand commerce of the marketplace. The first plates traced this: how the enforceable promise solves the oldest problem of cooperation across time, the problem of the party who must go first; how the power to commit, far from being a mere constraint, is the instrument of a deeper freedom; how the law's machinery of consideration and remedy gives the promise its grip, though even there — in the choice of damages over performance — the bindingness we imagine fixed turns out to be a thing the law constructs and could construct otherwise.

Incomplete, because no contract can foresee everything, and every contract is therefore a sketch, a core of express terms surrounded by a penumbra deliberately left blank. This was the volume's hinge, the fact from which its question grew, and it forced the central discovery: that what governs the parties in all the situations their words did not reach must come from somewhere else. Part of it comes from the law, which fills the silence with default rules, implied terms, and the standard of the reasonable — a vast unwritten contract beneath the written one, most of it composed by the state rather than the parties. But the larger part comes from no law at all. It comes from trust and reputation and the value of the continuing relationship — from the shadow of the future, under which honesty is simply the profitable course, and the cheat is known and shunned. The deepest finding of this volume is the inversion this implies: that the law of the deal is, in its daily operation, mostly not law, that the written floats always on a sea of the unwritten, and that the contract is in the end less the sheriff's creature than the quiet product of the fact that we must deal with one another again.

And contested, because the form of free agreement may carry a substance of power, and the reach of contract is bounded by judgments no logic of agreement can supply. The plates on the adhesion contract and the limit showed both edges of this: how the standard form and the rhetoric of freedom of contract can launder the dictation of terms by the strong to the weak, so that consent becomes a fiction and the contract a vessel of power; and how every society draws a perimeter around what may be made the subject of a bargain at all, refusing the coerced, the unconscionable, the inalienable, and the repugnant — a perimeter that is a portrait of its values and a verdict it passes on itself. The contract coping with the hidden knowledge of the parties, through the warranty and the signal and the screen, showed a gentler face of the same truth: that much of what a contract contains is not the exchange but the machinery built around what the parties cannot see. And the social contract, the idea raised to its political height, showed the contract's logic, its fiction of consent, and its incompleteness all written across a whole polity — turning the legitimacy of the state itself into a question of what free and equal people could have agreed to.

With this volume the first of the table's six families is complete, and it is worth pausing on what the completed group amounts to. The Institutions are the made things within which exchange occurs, and there are five of them. The Money is the medium — the common measure and means in which the deals are struck and the debts denominated. The Tax is the public claim — the share the polity takes from the exchange to sustain the common life that makes exchange possible. The Rule is the design — the property, the enforcement, the regulation, the very architecture that a market requires and that is built rather than found. The Firm is the planned interior — the islands of conscious coordination, of command in place of price, that dot the market's sea. And The Contract is the binding promise — the institution that lets exchange reach across time, that makes the future bankable and the stranger trustworthy. Money, tax, rule, firm, and contract: these are not additions to the market but the frame within which every market price is struck, and the first family of the table delivers, complete, the lesson the whole work was built to teach. The market is not a state of nature. It is an institutional achievement — conducted in a money we minted, under rules we wrote, taxed by a state we constituted, organized into firms we built, and bound by contracts we agreed to enforce — and every price that emerges from it is a verdict reached within that made and alterable frame.

So the threads run outward, as they have run throughout. To The Firm, which is itself but a nexus of contracts, and which answered the incompleteness this volume anatomized with the device of ownership where the contract answered it with defaults and trust — the two halves of a single response to a single gap. To The Rule, for whom the enforcement of promises is one of the load-bearing pillars on which any market stands. To The Tax, whose fiscal contract is the social contract rendered in money. To The Money, the medium in which the deal is denominated and itself, at bottom, a promise. To The Power and The Wage, where the contract is most nakedly an instrument of command. To The Surplus, since every bargain is a division of the gains from trade, and the terms of a contract are a verdict on who keeps them. To The Rate, the price of the very time across which the contract reaches. And forward to The Asymmetry, which will take up in full the hidden knowledge and hidden action that the warranty and the agent could only begin to meet.

The contract is the most ordinary of things — a deal, an agreement, a handshake, a signature — and we are tempted to see in it nothing made at all, only two parties doing what parties have always done. This volume has tried to show the institution beneath the ordinary thing: the enforceable promise, constructed by law, performing the profound work of binding the future and extending cooperation across time; never complete, and so forever finished by the law's defaults and the parties' trust; and never innocent of power, its freedom often formal and its limits drawn by contested judgments of value. Like the money it is denominated in, the rule that enforces it, the tax that shares in it, and the firm that is woven from it, the contract is a thing a community made to make a certain cooperation possible — and a thing that could be made, and may yet make, otherwise. The argument remains open.