The Norm

Begin with what a norm is, for it sits between two things we have already named and is exactly neither. It is not a law. No legislature passed it, no text records it, no court will hear a suit upon its breach; the state that stands behind the contract and the statute does not stand behind the norm. And it is not a contract. No one signed it, no consideration was exchanged for it, and it binds people who never agreed to it and never could have — for one does not opt into the expectation that one will keep one's word or take one's turn or deal honestly; one is simply held to it by being a member of the community that holds it. A norm is a third kind of rule: a shared understanding of how a person is to act, enforced not by the law's coercion but by the softer and stranger powers of approval and disapproval, reputation, and the conscience that has taken the rule inside.

Douglass North gave economists the frame that makes sense of this. The rules of the game, he said, come in two kinds. There are the formal institutions — the constitutions, the statutes, the property rights, the enforceable contracts — that The Rule examined, written down and backed by the organized force of the state. And there are the informal institutions — the norms, conventions, customs, codes of conduct, and taboos — held in common and enforced by society itself. A functioning economy runs on both, intertwined; but the proportion between them is not what a lawyer might guess. The formal rules are the visible tip. The informal rules are the great submerged mass, governing the vast reach of conduct that no statute touches and no contract specifies — which is to say, as the volume on the contract showed, nearly all of it.

The reason norms are so easily overlooked is precisely that they work so well. We notice a law, for it is written and enforced in the open, with penalties and officers and courts. We do not notice a norm at all, until it is broken. The queue we join without being told, the tip we leave without being billed, the greeting we return, the appointment we keep, the honest weight the grocer gives, the change the stranger hands back — these are the water we swim in, felt not as rules at all but as the way things simply are. And that feeling, that a norm is not a rule but a fact of the world, is the deepest naturalization this table has had to confront. The whole work has labored to show that prices are not facts of nature but verdicts of contests; here, at the foundation, we meet the rules that feel most natural of all, the done things that present themselves as obvious and eternal — and that are, no less than the prices, human constructions, grown and enforced and alterable.

What makes a norm a rule, and not a mere suggestion, is that it is enforced — not by the law, but enforced all the same. A regularity that carries no sanction is a habit or a fashion; a norm carries a sanction, and the sanction is what gives it its grip. The plates that follow will anatomize the enforcement — the conscience within, the watching community without, and the remarkable willingness of people to punish a violator even at a cost to themselves. But the point to fix here is that the bindingness of a norm comes from this enforcement and not from any writing, so that an unwritten rule, socially enforced, can be every bit as real and every bit as constraining as a written one. The norm is not the law's soft shadow. It is an institution in its own right — a structure of shared expectation that shapes what we do as surely as any code, and over a far wider field.

The simplest norms are the conventions, and they are the place to begin because their workings lie open to view. A convention is the solution to a coordination problem — a situation in which each of us wants above all to do the same as the others, and it scarcely matters what that is, only that we agree. The cleanest case is the side of the road. It is of no importance whether a nation drives on the left or the right; what matters, enormously, is that everyone in it drives on the same side. The figure draws the choice. Two drivers approaching each other each pick a side; if they pick the same, they pass in safety; if they pick differently, they collide. There are two ways to get it right and two ways to get it catastrophically wrong, and the whole problem is to land together on one of the two right ones.

Two features of the convention follow from the figure, and they are the signature of this whole class of norm. The first is that it is arbitrary: nothing in the nature of things favored the right side over the left, and a nation that drives on one is in no way wiser than a nation that drives on the other. The second is that, arbitrary in origin, it is absolute in force: once a nation has settled on a side, the rule binds every driver with a completeness no statute could surpass, and binds him not because the law commands it — though the law may also command it — but because everyone else is there, and to depart is to crash. The convention enforces itself. This is what sets it apart from the harder norms the later plates will treat, the norms of honesty and cooperation that run against self-interest and so require a punisher; a convention runs with self-interest once it is established, and asks no enforcement beyond the mutual expectation that sustains it.

Where does such a rule come from, if no one need impose it? Mostly it grows. David Hume gave the classic account, and chose his examples from the heart of the economy: the value of money, the institution of property, the keeping of the rules of justice — none of these, he argued, rests on a contract or a command, for all arise as conventions do, gradually and without agreement, from the repeated experience that it serves each to act as the others act. Two men who row a boat, he wrote, keep time by a convention though they never promised it to each other. Money is accepted because each expects the next to accept it, in the self-fulfilling circle the volume on money called faith. Where no precedent yet exists, Thomas Schelling showed, people coordinate on whatever option is salient — the focal point, the obvious place to meet, the round number, the natural Schelling point on which expectations converge without a word. By precedent and salience and the slow sediment of repeated dealing, conventions crystallize, and once crystallized they are as binding as anything in the social world.

The economy is built through and through on such conventions. Money is one, as we have seen; so are the units in which we reckon — the meter, the measures, the standards and protocols that The Rule traced — and the meanings of the words in which we deal, and the trading hours and customary terms and settled practices of every market and trade. These are the coordination infrastructure of exchange, the shared expectations that let strangers transact without renegotiating from nothing each time, and they are mostly the work of no one, grown rather than made. They show, against the picture of the market as a thing that must be wholly designed and enforced, that a great part of its order arises spontaneously, self-organizing out of mutual expectation. And they carry the table's lesson into its strangest corner: that what feels most fixed and necessary — the medium, the measure, the settled way things are done — is often a mere convention, an arbitrary historical settlement frozen into a rule, which could have set otherwise and feels inevitable only because we have forgotten that it was ever open.

If conventions are the gentlest norms, reciprocity is the deepest, and it may be the oldest social rule our species keeps. Return a kindness; repay a gift; answer help with help and, often enough, harm with harm. The anthropologists find some version of this in every human society, and Marcel Mauss, studying the gift economies that precede and underlie the market, showed that the gift is never truly free: to give is to oblige, to receive is to be bound, and to fail to return is to be shamed. The gift creates a debt that no contract records but that everyone feels, and the cycle of giving and returning weaves the web of obligation on which a society of exchange is built. Long before there was a contract, there was the gift and its return, governed by a rule older than any law.

In its simplest form reciprocity is direct: I help you, and you, in time, help me. This is the logic the contract met under the name of the shadow of the future, and Robert Axelrod gave it its sharpest expression when he set strategies to compete in a repeated prisoner's dilemma and found that the winner was the simplest reciprocator of all. Tit-for-tat — cooperate on the first move, then do whatever the other did last time — beat every cleverer scheme, and it did so by being nice, never defecting first; retaliatory, always answering a defection; forgiving, returning to cooperation the moment the other did; and clear, so that any partner could see at once that cheating did not pay. Reciprocity, formalized as tit-for-tat, sustains cooperation in any relationship expected to last, by making each party's future treatment depend on its present conduct.

But reciprocity reaches further than the bilateral pair, and its longer reach is what makes it the foundation of a whole community's cooperation. The figure shows the extension. A helps B; but the help that returns to A comes not from B at all — it comes from C, who saw or heard of A's good conduct and so is willing to help A in turn. The return travels round the circle rather than straight back, and what carries it is reputation: by helping, A earns a good name, and a good name draws help from others who were never helped themselves. This is indirect reciprocity, and the biologists Nowak and Sigmund, building on Richard Alexander, showed how it can sustain cooperation across an entire community, each person helping those of good repute and withholding from those of bad.

This indirect form is of the first importance, for it is the bridge from the cooperation of pairs to the cooperation of whole societies, and it is, very nearly, what we mean by a moral system. The injunction to do good carries the promise that good will come back — not, as the child supposes, from the very person one helped, but from a world that keeps a rough account of one's conduct and metes out its help and its withholding accordingly. The reputation systems of commerce, the credit ratings and the reviews and the references the contract relied upon, are indirect reciprocity made explicit; and the older, vaguer reputation that attaches to a name in a community is the same mechanism running on gossip and memory. Reciprocity, in its direct and indirect forms together, is the engine that turns the prudence of self-interested parties into the cooperation of a society.

And here we reach the deepest thing reciprocity has to teach, which is its relation to exchange itself. The market transaction is reciprocity made immediate and explicit: I give you the goods, you give me the money, the kindness and its return collapsed into a single balanced act and the debt discharged on the spot. What the gift economy did slowly and diffusely, through obligations that hung in the air and were repaid over years, the market does instantly and precisely, through a price. Exchange is the ancient norm of reciprocity, formalized, accelerated, and stripped of its personal residue — which is why the market could grow out of the gift, and why, beneath the impersonal machinery of price and contract, the old rule still beats: that what is given must be returned, and that the one who takes without returning has broken something the whole order depends upon.

Reputation and the watching community explain why we keep a norm when we are seen. They do not explain the harder and more telling case: why we keep it when we are not. The traveler returns the overpaid change to a shop he will never enter again; the finder mails back the wallet dropped on an empty road; the diner leaves the tip in a town to which he will not return, for a server who will not remember him. In none of these does reputation hang in the balance, for no one will ever know; the shadow of the future falls nowhere. And yet people do these things, reliably, and feel a wrongness when they fail to. The enforcement that reaches into the unobserved cannot be external. It must be within.

It is within, and we call it conscience. Norms, when a society does its work of raising the young, are not merely obeyed but internalized — taken inside, made part of the self, so that to violate them is to suffer not only the disapproval of others but a punishment one administers to oneself. Jon Elster located this enforcement in the moral emotions, and the two that matter most are shame and guilt. Shame is the pain of others' disapproval, real or imagined — the burning face, the wish to disappear, the eyes of the community felt even when the community is absent. Guilt is the pain of one's own judgment — the inward verdict that one has done wrong, which needs no audience and survives in perfect solitude. Both make the violation of a norm costly to the self regardless of detection, and so both reach where reputation cannot: into the dark, where most of conduct actually lives.

How conscience comes to be is no mystery, though it is easy to mistake. The norms a child meets are at first wholly external — the parent's frown, the teacher's reproof, the playmate's scorn — and through the long apprenticeship of growing up these external sanctions are taken in, until the parent's disapproving voice becomes the inner voice that disapproves when the parent is gone. The social becomes psychic; the rule enforced from without becomes the rule enforced from within. And there is reason to think the very capacity for this internalization was favored by evolution, for Robert Frank argued that the moral emotions function as commitment devices: the person who genuinely feels guilt at cheating is, for that reason, a person who can be trusted not to cheat, and so a person others will deal with, cooperate with, and prefer — so that a conscience, far from a handicap in the struggle of life, is an asset that draws the cooperation on which thriving depends. The honest man's honesty advertises itself, and is repaid.

The economic significance of this is hard to overstate, though it appears in no ledger. A market runs, to a degree we scarcely notice, on internalized honesty — on the countless dealings in which a party could cheat undetected and simply does not, because cheating is not the kind of thing he does. Strip this away, and imagine a world in which everyone cheated whenever cheating would pay and go unseen, and the result is the lemons market everywhere at once: every transaction policed, every quality verified, every promise distrusted, the costs of suspicion eating the gains of exchange until little exchange is left. The conscience is, in cold economic terms, the cheapest enforcement mechanism ever devised, for it requires no watchers, no courts, no surveillance; it works in the dark, for free, in the only place that the expensive machinery of external enforcement can never reach. A society blessed with widely internalized norms of honesty is rich before it produces anything, for it can transact at a fraction of the cost of a society that must watch its every member.

Conscience reaches the unobserved, but much of conduct is observed, and there a second layer of enforcement operates — external, social, and wielded by the community without any help from the state. Its instruments are humble and familiar. The mildest is approval and its withdrawal: the warm word and the cold one, the nod and the snub, the esteem we extend to those who keep the norms and withhold from those who break them. Because we are a species that craves the regard of others and shrinks from their contempt, this costs the community almost nothing to administer and yet bites deep. Above it stands reputation: the violator's name is marked, and others grow wary of trusting, dealing with, or standing beside him — the loss of all the future cooperation that a good name would have drawn, which the volumes on the contract and on reciprocity showed to be a genuine and heavy cost.

Between the snub and the spoiled name runs the mechanism that makes reputation possible at all, and it is one the respectable affect to despise: gossip. Gossip is the community's information system, the means by which a violation in one corner becomes known in all the others, so that reputations can be assigned and the watching eyes can see even what they did not witness. Anthropologists who have studied it closely find it anything but idle — it is the registry of the unwritten law, the distributed record of who has dealt well and who has cheated, without which no reputational sanction could function, for one cannot withhold trust from a cheat one has never heard of. What the court record and the credit bureau are to formal enforcement, gossip is to informal: the memory of the community, kept in common and consulted before every dealing.

At the summit of the social sanctions stands the heaviest of all: exclusion. To be shunned, boycotted, cast out — denied the cooperation, the company, and the standing of the group — is for a social animal a kind of death, and the threat of it has disciplined human conduct more thoroughly than any prison. The merchant guilds that Avner Greif studied enforced honest dealing across continents by the threat of collective boycott; the close community enforces its norms by the threat of the shut door; and the lesser exclusions of daily life, the un-invitation and the quiet dropping, enforce the smaller norms. These sanctions are graded, from the flicker of disapproval to total banishment, and they are decentralized — administered by no authority but by the diffuse, spontaneous reactions of everyone at once, which is what makes them at once so cheap and, under the right conditions, so overwhelming.

The phrase "under the right conditions" carries the weight, for the social sanctions have a precise domain in which they are mighty and outside which they fail. They work where conduct is observable, where dealings are repeated, and where the community is dense enough that gossip reaches everywhere and no one can escape his name. They are the enforcement of the village, the guild, the trade, the close community — and within such a world they govern nearly everything, leaving the law little to do. But they depend utterly on being seen and remembered, and so they reinforce the three-layered structure of enforcement this volume has been assembling: the conscience within, reaching the unobserved; the community without, reaching the observed and remembered; and, beyond both, the formal law of The Rule, reaching where neither conscience nor community can — among the strangers of the anonymous society, where the watching eyes are blind and the spoiled name does not follow. A society leans on all three, in proportions that shift with its scale, and the next plates will show what happens when the work of enforcement grows too great even for the costly punishers who give the social sanctions their final teeth.

The social sanctions of the last plate all depend on someone willing to administer them — to shun the cheat, to spread the word, to refuse the unfair deal, to spend the effort that punishment requires. And here a puzzle opens that threatens the whole structure. Punishing a violator is costly to the punisher: it takes effort, invites retaliation, and yields him no private gain, since the benefit of a well-enforced norm — a more honest community — flows to everyone alike. The rational self-interested party should therefore decline to punish and hope that others will, free-riding on their costly enforcement exactly as the free-rider rode on others' contributions in the first place. But if everyone reasons so, no one punishes, the cheats go unanswered, and the norm collapses. This is the second-order free-rider problem, and on paper it should unravel cooperation as surely as the first.

On paper. In the laboratory, something else happens, and its discovery is among the most consequential in modern economics. Ernst Fehr and Simon Gächter set people to play a public-goods game over many rounds, each deciding how much to contribute to a common pool that benefited all. The figure shows what they found. Without any means of punishment, contributions begin at a moderate level and then decay, round by round, toward nothing, as the cooperators tire of being exploited by the free-riders and stop contributing themselves — cooperation unraveling just as the model predicts. But when the players are given the option to punish, at a cost to themselves, those who contributed little, everything changes: contributions begin at the same moderate level and then climb, round by round, toward near-complete cooperation.

The starkest demonstration of the same disposition is the ultimatum game, and it is worth stating in full because it refutes so cleanly the picture of the purely self-interested agent. One player is given a sum and proposes how to split it with a second; the second either accepts, and both keep their shares, or rejects, and both get nothing. Self-interest is unambiguous here: the responder should accept any positive offer, since any amount beats nothing, and the proposer, foreseeing this, should offer the smallest possible crumb. What happens is the opposite. Responders routinely reject offers they find insulting — a fifth of the pot, a tenth — choosing nothing over a pittance, paying with their own loss to punish a proposer who tried to take too much. And proposers, knowing this, mostly offer something close to half. People enforce a norm of fair division at a cost to themselves, and the enforcement is so reliable that it reshapes the offers made in its shadow.

Why people punish at a cost is a question with answers at several depths. The proximate answer is emotion: anger and indignation at the cheat and the chiseler, which make punishing feel satisfying enough to be worth its price — the moral emotions, once again, doing the work that calculation will not. The ultimate answer, more contested, is evolutionary: Samuel Bowles and Herbert Gintis argued that a disposition to costly punishment — strong reciprocity, they called it — could have spread by cultural group selection, since groups well supplied with such punishers sustained more cooperation and prevailed over groups without them, even though within any group the punishers bore a cost the free-riders escaped. However it arose, its presence is not in doubt, for it appears across cultures, survives when the stakes are raised, and resists every attempt to explain it away. And its significance for the table is profound: it means that the cooperation on which every economy rests cannot be derived from self-interest alone, however far-sighted, but depends at its foundation on a willingness to enforce fairness that is not itself self-interested — the punisher in us, on whom the whole order of norms, and the cooperation it sustains, finally leans.

We have seen what norms are, how they bind, and how they are enforced; the question that remains is where they come from, and the answer is the strangest in the volume. A few norms are made — proclaimed by a leader, instilled by a campaign, left behind as the residue of a law. But the great body of them is made by no one. They emerge, the way a path is worn across a field by feet that never agreed to make it: each walker takes the way the others have taken because it is easier and because the others are taking it, and every passage wears the path a little deeper, until what began as the chance track of a few has hardened into the only way across. No surveyor laid it; no authority decreed it; and yet it governs every crossing. So it is with the conventions of property and dealing that David Hume watched arise without contract or command, with the spontaneous order Friedrich Hayek placed at the center of his thought, and with the self-enforcing regularities of cooperation and exchange that Robert Sugden showed could grow up out of nothing but repeated interaction. Norms are the result of human action but not of human design.

That emergence is not guaranteed, however, and the reason is the tipping dynamic the figure draws. Because a norm is self-reinforcing — the more who keep it, the stronger the pressure on each to keep it too, partly because one wishes to do as others do and partly because the sanctions fall on the deviant — its fate hangs on a threshold. The line below maps the fraction of a community that follows the norm, from none to all, and marks the three points that matter. At either end sits a stable resting place: a world where no one follows, and a world where all do, each self-perpetuating. Between them lies an unstable point, the tipping point, and it divides the line into two countries. Begin to the left of it, with too few adherents, and the norm decays toward extinction, each defection making the next easier until it fades back into the grass. Begin to the right of it, with enough, and the norm spreads toward universality, each new adherent drawing in the next until it governs all.

The tipping line explains what the simple image of the worn path cannot: why some norms catch and spread to every foot while others, walked by too few, fade away. A norm must reach its critical mass to survive, and below it the most sensible rule will die unobserved, for there are too few keeping it to make keeping it worthwhile. This is the mathematics of the fad and the riot and the bank run as much as of the norm — Thomas Schelling's tipping and Mark Granovetter's thresholds, the cascade that waits on a critical mass — and it governs the birth of the unwritten rules as surely as it governs their death.

Over the long run, the norms that emerge are sifted by something like selection. Those that serve their communities — that sustain cooperation, coordinate behavior, solve the recurring problems of living together — tend to spread and endure; those that do not tend to die out, taking their groups down with them or being abandoned for something better. The students of cultural evolution, Robert Boyd and Peter Richerson foremost among them, have shown how norms are transmitted and selected across generations like a second inheritance, faster and more flexible than the genetic one. And Samuel Bowles and Herbert Gintis argued that the cooperative norms in particular may have spread by selection among groups, the well-normed prevailing over the ill-normed across the long contest of human history. On this reading the norms we keep are evolved solutions to the games our ancestors played without end — the coordination games that left us conventions, the cooperation games that left us reciprocity and fairness — the sediment of what worked, carried forward in the customs of the living.

Two consequences follow for the table, and they pull against each other in a way the volume will not resolve. The first is humbling: because norms are grown and selected rather than designed, they may carry a wisdom no designer possesses, encoding in their wordless way the lessons of more trials than any planner could run — which is Hayek's case for treating the inherited rules with a certain reverence, and for doubting the reformer who would sweep them away in the name of a reason that sees less than it thinks. The second is unsettling, and it is the table's own: because norms are grown and not given, they are contingent, the frozen accidents of which paths happened to be walked and which groups happened to prevail. A different norm could have caught; the tipping could have gone the other way; the done thing that feels eternal is the survivor of a contest that might have ended otherwise. The deepest rules of economic life, no less than the prices the table began with, were not handed down by nature. They were worn into us by use, and what use wore in, use can wear away.

We have anatomized the norms; now we must price them, or at least show why they are beyond price. The volume on the contract found trust filling the gaps that no clause could close, in the particular relationship between two parties who deal again and again. Raise that finding from the pair to the society, and it becomes one of the largest facts in all of economics. Kenneth Arrow stated it plainly: trust, he said, is an important lubricant of a social system, extremely efficient, saving a great deal of the trouble it would take to verify everything one is told. Where people can rely on one another's word, exchange flows; where they cannot, it seizes. And what people rely on, when they trust, is precisely the dense web of norms this volume has described — the expectation that others will deal honestly, keep their promises, and return a kindness rather than exploit it.

The figure shows the mechanism by which trust does its economic work. As the trust in a society rises along the horizontal, two things move together. The cost of transacting falls, for where one can take a word on faith there is less need to verify, monitor, guard, and litigate — the whole apparatus of suspicion that the volumes on asymmetry and the contract showed to be so expensive grows lighter as trust grows. And the volume of beneficial exchange rises, for as the cost of each dealing falls, dealings that were not worth their friction become worth doing, and cooperation widens to fill the space that distrust had kept empty. The two curves cross the page in opposite directions, and the gap that opens between them — low cost, high exchange — is the dividend that trust pays.

This asset has a name. Following James Coleman and made famous by Robert Putnam, economists and sociologists call it social capital — the networks, norms of reciprocity, and trust that let a community cooperate, and that raise its productivity as surely as a stock of machines or a fund of skills. Social capital behaves like capital in that it is productive and it accumulates; it is unlike the capital of the earlier volumes in that no one owns it, for it lives in the relations between people rather than in any of them, and it is built not by saving and investing but by the slow practice of dealing well and keeping faith. A society well supplied with it can do cheaply what a society lacking it cannot do at all: trust a stranger, enforce a bargain by reputation, build an institution larger than a family.

That last difference is the one Francis Fukuyama placed at the center of his study of trust, and it carries the argument to the scale of nations. Societies differ enormously in the radius of their trust — in how far the expectation of honest dealing reaches beyond the circle of kin. Where trust extends widely, to strangers and not merely to family, a society can build the large impersonal organizations that modern production requires, can staff them with the unrelated, and can transact across the distances that wealth demands. Where trust is confined to the family or the clan, the firm struggles to grow beyond the household, the costly scaffolding of formal enforcement must do what trust will not, and the gains of wide cooperation go unreaped. The wealth of nations, on this view, rests in no small part on a thing that no nation can purchase and few can quickly build: the diffuse, generalized trust that is the dividend of good norms widely kept.

So the unwritten rules, which the early plates showed binding and constraining us, show here their other and kindlier face. They are not only the chains of conformity; they are the foundation of cooperation, the social capital that makes wide and cheap exchange possible, the lubricant without which the gleaming machinery of prices and contracts would grind to a halt. Beneath all the priced wealth the table has anatomized lies this unpriced wealth, the stock of trust and honest dealing that never appears on a balance sheet and may be worth more than everything that does. And it returns us, by a long road, to the thing this whole work has sought — for the numéraire of honesty is not merely a standard against which to read the verdicts hidden in prices; it is itself the most valuable asset a society can hold, the keystone norm of honest dealing on which the dividend of trust is paid.

The volume on the commons posed a tragedy: a resource held by all and owned by none — the pasture, the fishery, the forest, the air — is used by each to exhaustion, since each takes the whole of his private gain while the cost of depletion falls on all. And it offered the two remedies that economic thought has long allowed. One may privatize the commons, carving it into owned parcels so that prices and the owner's self-interest do the conserving, which is the market's answer. Or one may regulate it, setting a distant authority over it to ration and prohibit, which is the state's. The whole debate over the things we share has run, for a century, between these two poles: make it private, or make it public; give it to the market, or give it to the state.

Elinor Ostrom won a Nobel Prize for showing that this exhausts neither the possibilities nor the facts. Going out into the world to look, she found commons that had been governed sustainably for centuries by neither market nor state, but by the communities that depended on them — the high pastures of the Swiss Alps, the irrigation networks of Spain, the village forests of Japan, the inshore fisheries of Maine — each managed not by private owners and not by government officials but by the users themselves, under rules of their own making, monitored and enforced among themselves. The tragedy of the commons, she demonstrated, is not a law of nature but a possibility that communities have repeatedly escaped, and the means of their escape was neither the price nor the command but the norm.

She did more than find the cases; she distilled what made them work, and the distillation reads as a catalogue of everything this volume has described. The enduring commons had boundaries clear enough that everyone knew who might use them and who might not; rules fitted to local conditions rather than imposed from afar; the participation of the users in making the rules they would have to keep, so that the rules were theirs and were felt as legitimate; monitoring, often by the users watching one another; and — the punisher's signature — graduated sanctions, mild for a first transgression and mounting for the stubborn, applied by the community to its own. These are the norms and the social sanctions of the earlier plates, organized into a working government of the commons. Self-governance by norm is not a sentimental hope; it is a documented and reproducible institution, with conditions that can be stated and met.

What Ostrom's discovery restores to economics is a third pillar of governance that the discipline had let fall. Beside the market, which governs by exchange and price, and the state, which governs by law and command, stands the community, which governs by norm and mutual obligation — the realm of cooperation that is neither bought nor commanded, which Samuel Bowles and Herbert Gintis insisted belongs in the picture and which Raghuram Rajan named the third pillar. An economy is held up by all three, and leans too far when it forgets any one: the society that knows only the market and the state will dissolve its working communities in the name of efficiency or order, and discover too late that it has destroyed the cheapest and most legitimate governor it had.

Yet the third pillar is a pillar and not the whole building, and Ostrom was the first to mark its limits. Community governance works under the same conditions that make any norm strong — a bounded membership, a stable population, the capacity to watch and to sanction, the long shadow of repeated dealing — and it falters where those conditions fail: for the large, the diffuse, and above all the global commons. The atmosphere has no bounded community to govern it, no users who watch one another, no shared norms reaching across the whole of humanity; and so the greatest commons of all remains the hardest case, beyond the reach of the village solution and waiting still on the cooperation of strangers who do not know one another's names. The community can govern what it can see and sanction. Past that horizon, the old problems return, and the work of governing what we share passes back to the market, the state, and whatever fragile cooperation the species can contrive among those it cannot watch.

The last plate but one showed markets resting upon norms — trust the lubricant, honesty the foundation. This plate shows the darker half of the relation, in which the market does not rest upon the norm but dissolves it. The intuition to be overturned is the most natural in economics: that incentives work in one direction, so that to want more of a behavior is to pay for it, and the more one pays the more one gets. Against this stands one of the most instructive findings in the modern literature, and it begins with a daycare.

Uri Gneezy and Aldo Rustichini studied a set of Israeli daycare centers troubled by parents who came late to collect their children, leaving the staff to wait. The centers introduced a fine for lateness, expecting it to deter. Lateness rose. It rose because the fine had changed the meaning of being late: what had been a small moral failing — an imposition on the staff, a thing one felt ashamed to do — became a service with a posted price, which a parent was now entitled to purchase. The norm that had restrained lateness was a sense of obligation; the fine replaced it with a fee, and once lateness could be paid for, there was no longer anything wrong with it. Worse still, when the centers later removed the fine, lateness did not return to its old level: the norm, once dissolved, did not reassemble. The price had not added a deterrent to the norm; it had driven the norm out and taken its place, and proved a weaker master than the shame it had banished.

The figure generalizes the lesson, and it is the shape of the curve that matters. With no payment at all, the behavior stands high, sustained by the norm or the intrinsic motive that moves people to do a thing for its own sake. Introduce a small payment, and the behavior does not rise but falls, for the payment dissolves the norm — reframing the act as a transaction — while being far too small to call forth as a transaction what the norm had called forth as a duty. Only when the payment grows large does the behavior climb again, the price at last buying what the principle used to give. The relation between incentive and effort is not the rising line of intuition but a valley: a small reward is worse than none, and the dashed line marks how far below the unpaid level the lightly paid behavior sinks. Gneezy and Rustichini drew the moral in four words — pay enough, or don't pay at all — but the deeper moral is that the middle case, the small price laid upon a norm, can be the worst of every world.

Why a price should dissolve a norm rather than reinforce it is now fairly understood, and the mechanisms all turn on meaning. A payment reframes an act, moving it from the domain of duty, gift, or honor into the domain of the market, where the norms of duty, gift, and honor no longer apply and only the price speaks. It can undercut the intrinsic satisfaction of doing a thing freely, as the students of motivation have shown, so that what was done for its own sake is now done only for the money or not at all. And it can signal distrust — the fine that says we expect you to shirk, the monitor that says we do not take your word — corroding the very goodwill on which cooperation depended. In each case the market does not layer an incentive atop the norm; it changes what kind of thing the act is, and in changing it, destroys the norm that governed it.

Set beside the earlier plate, this one completes the account of how markets and norms entangle, and the account has three strands rather than one. Markets depend on norms, as the lubricant of trust made plain. Markets can erode norms, as the crowding-out makes plain here. And markets can, in their gentler moods, create norms — the punctuality, the fair dealing, the tolerance that commerce among strangers is sometimes said to teach. The market is not a separate sphere walled off from the moral order, as the model's clean machinery suggests, but is embedded in that order — the word is Mark Granovetter's, and the warning behind it Karl Polanyi's — drawing on its norms, sometimes nourishing them, and sometimes consuming them. Michael Sandel pressed the consequence furthest: that there are goods whose nature is altered by being priced and relations the market degrades by entering, so that the question of what may be bought and sold is not only a question of efficiency but of what each thing is and how it ought to be held.

The volume has so far shown the kindly face of the norm — cooperation, trust, the social capital that lubricates exchange and the community that governs the commons — and it would be a sentimental and dangerous account that stopped there. For the machinery this volume has described is neutral as to what it enforces. The shared expectation, the watching community, the graduated sanction, the internalized conscience — these sustain a norm of honesty exactly as they sustain a norm of corruption, a norm of fair dealing exactly as a norm of discrimination, a norm of courage exactly as a code of silence that shields the wrongdoer. A norm is only a rule the community keeps and enforces upon itself; whether the rule is good or vile, the keeping and the enforcing work the same. And so the very strength that makes a good norm a blessing makes a bad norm a curse that is exceptionally hard to lift.

The figure shows why, and it is the dark twin of the tipping line drawn earlier. Along the horizontal runs the fraction of a community that keeps some good norm — honest dealing, say, in a place where corruption is otherwise the rule — and up the vertical runs the payoff each person receives. There are two stable resting places. At the right sits the good equilibrium, where nearly all keep the norm and all are prosperous, the high payoff of a society that has solved its problem of cooperation. At the left sits the trap: a low fraction keeping the norm, a low payoff for everyone, a world of mutual cheating in which the honest are fleeced and the corrupt are merely normal. Between them, the open circle marks the unstable tipping point that must be crossed to get from the one to the other. And the cruelty of the picture is that a society can be locked at the left-hand point — everyone worse off than they would be at the right — with no individual able to move it.

The mechanism of the lock is the same self-enforcement that made good norms durable, now working for ill. Each person keeps to the bad norm because everyone else does and the deviant is punished: the official who refuses the bribe is frozen out, the merchant who alone deals honestly is undercut and ruined, the one who breaks the code of silence is destroyed by those it protects. No one will be the first to leave, for the first to leave bears the whole cost of deviance while the norm rolls on unmoved — the collective-action problem of the contract and the commons, returned in its bleakest form. And to this is often added a subtler trap still, which Timur Kuran named preference falsification: under the pressure of the norm, people misrepresent what they truly think, professing in public a support they do not feel in private, so that each comes to believe — falsely — that all the others endorse the very rule that all of them privately curse. The norm then stands not on real conviction but on a shared illusion of conviction, a pluralistic ignorance in which everyone conforms to please a majority that does not exist.

This dark side also revises the kindly accounts that preceded it. The social capital of the eighth plate, it turns out, can be dark capital: the dense networks and fierce norms that let a community cooperate are the same that let a mafia, a cartel, or a closed and bigoted guild cooperate, and a society can be rich in a trust that binds its members the more tightly to a collective harm. The conformity that coordinates can also oppress, and John Stuart Mill thought this social tyranny more penetrating than the political kind, since it reaches past the public acts that law can touch into the private life that only opinion governs, punishing not the crime but the difference, and silencing not the dangerous but the merely deviant — including the dissenter and the innovator on whom all improvement depends. Norms stabilize the good and the bad with equal force, and suppress, along with the harmful deviation, the beneficial one.

We asked, at the opening of this volume, which of the market's rules are written and which are merely understood, and the long answer the plates have given is that the understood are the deeper and the larger. The conventions that let strangers coordinate, the reciprocity that turns prudence into cooperation, the conscience that keeps a rule unobserved, the sanctions and the costly punishers that enforce what no court will hear, the trust that is a society's hidden capital, the community that governs by norm what neither market nor state governs well — these are the informal institutions, and they do the greater part of the work of holding an economy together. The formal rules of The Rule are the visible writing; the norms are the understood foundation on which the writing rests. And the dark eleventh plate forbade us to sentimentalize them, for the same machinery that sustains honesty sustains corruption, and a society can be trapped by its unwritten rules as cruelly as it is served by them.

This is the twenty-sixth element, and the last, and it is fitting that the table should close here, at the rules that no one wrote. For the whole of this work has been a single sustained act of denaturalizing — of taking the things that present themselves as facts of nature, the wage and the rent and the rate, the price that seems as given as the weather, and showing each to be a made thing: a verdict on a hidden contest, an institution with a history, a choice that could have been otherwise. And beneath every made thing the table has examined lies this deepest made thing of all. The norm is the object that feels most like nature — the done thing, the way things simply are, the rule so woven into us that we mistake it for the grain of the world — and it is, of everything the work has touched, the most thoroughly our own doing.

Once this is seen, the whole edifice the table anatomized is revealed to be standing on it. The money in every pocket is accepted because each of us expects the next to accept it — a convention, as the volume on money found, faith resting on faith. The contract in every drawer is kept, in the main, not for fear of the court but because keeping one's word is the done thing, as the volume on the contract confessed at its end. The property is respected, the queue is honored, the debt is repaid, the weight is given honest, the market clears at dawn — all because a dense and mostly unspoken web of shared expectation tells each of us how the others will behave and what will befall us if we behave otherwise. The invisible hand, whose workings this table set out to make visible, turns out itself to rest on something more invisible still: an invisible substructure of norms, without which the hand would have nothing to grasp and the prices nothing to mean.

And so we come, by the longest road, to the thing this entire work has been seeking. From the first page it has sought a numéraire of honesty — a true standard against which the verdicts hidden inside prices might at last be read, a unit of honest reckoning to set against the dishonest reckonings that prices so often are. We have looked for it through twenty-six elements, and we find it here, at the foundation, and find that it was never a thing to be discovered outside the system of prices but a norm to be kept within it. The numéraire of honesty is the keystone norm itself: the unwritten rule that we will not cheat the count — in our dealings, in our promises, in our reckonings, and in our thinking. It cannot be minted or measured or bought; it can only be kept, and taught, and enforced among ourselves, in the slow unpurchasable way of all norms. The standard the work demanded turns out to be a standard the work can only commend, for it is not a fact to be established but a commitment to be made, the keystone norm of honest dealing on which the trustworthiness of every other reckoning depends.

This volume is also the first of the last family, the Behaviors, and in closing the table it opens that family with the observation on which the family is built: that the agent of economic life is not the solitary calculator of his own advantage whom the model has so long assumed. The punisher who pays from his own pocket to see unfairness answered, the conscience that keeps the rule in the dark where no one will ever know, the reciprocator who returns a kindness without reckoning the return, the ordinary person who does the done thing because it is done — none of these is the creature of pure self-interest the equations describe. They are creatures shot through with norms, with fairness and conscience and the need to belong woven into the very preferences the model takes as given. The plates that were to follow — the Choice and the ways it departs from the rational ideal, the Bias that bends it, the Game that strangers play against one another — would develop this agent in his full and puzzling reality. The Norm has done only the foundational thing: it has shown, beneath the calculation, the creature of rules.

For this last element reaches back through the whole table, and gathers it. It is the informal twin of The Rule, the unwritten institutions standing beside the written. It is the hand-back for the handshake of The Contract, the trust and reputation that fill the gaps no clause can close. It is the deep structure beneath the faith of The Money, the convention that lends the worthless token its worth. It is the third way of The Commons, the community that governs by norm what the market would exhaust and the state would bungle. It is the foundation of the fairness that troubled The Wage and haunted the division of The Surplus, the unwritten sense of the just share that no productivity can fully explain. And it is the soil in which the strategies of The Game take root, the shared expectations that turn a tangle of self-interested moves into the patterned cooperation of a society. The norms are everywhere beneath the table, and in ending on them the work does not so much conclude as return to its own foundations.

So the table ends where it had to end, at the rules that were never written, and finds even these to be no exception to its one persistent finding. The norms, like the prices the work began with, are not given by nature. They are grown and evolved, enforced and internalized, and they are contingent through and through — the survivors of contests we have forgotten were ever held, the done things that could have been done otherwise and may yet be done otherwise still. To see them as made is not to wish them away, for a society without norms is a society without trust, cooperation, or honesty, which is to say no society at all; the unwritten rules are not a prison to be escaped but the ground on which any common life must stand. It is, rather, to hold them as ours — to keep the good ones knowingly rather than blindly, and to know that the bad ones, however eternal they feel, are traps that human beings built and that human beings could, acting together, unbuild. This has been the argument of the whole work, pursued through twenty-six elements and rendered here at its foundation: that the economy is not a piece of nature to be suffered but a human construction to be understood, judged, and where it fails us, remade. Its deepest rules are unwritten, and they hold up everything else. They are also our own work, and they remain open — as this argument, having reached its last element, remains open — to revision by the hands that made them. The argument remains open.