The Censure — The Norm, Plate vi

The community needs no court to make a violation costly. A frown, a cold word, a name passed quietly along, and at the last the shut door: these are the sanctions of the unwritten law, familiar enough to escape notice until they fall upon us. The mildest withdraw esteem; the strongest withdraw the cooperation on which a livelihood depends. Yet even the frown has someone behind it, and the shut door has someone who must give up the advantage of opening it. An unwritten rule acquires force when people expect a breach to become known, expect others to respond, and have reason to believe that someone will undertake the response. The censure belongs to all three expectations together. A community may agree perfectly about what ought to be done and still leave the person who breaks its rule entirely undisturbed.¹

1. Adapted from the earlier Sanction and Punisher, with the Margin Shame plan. Their movement from disapproval to exclusion and then to the cost of enforcement supplies this plate’s base. These are distinct earlier treatments, preserved in their own editions.

Reputation gives the sanction a reach beyond the original witness, provided that a report is believed and that other people act upon it. Gossip is the community’s information system: the means by which a violation in one corner becomes known in another. Its power rests on the quality of the information it carries as well as the severity of the response it can summon. Consider an invented calculation in which cheating brings an immediate gain of four, while detection and a credible boycott together bring a discounted loss of ten in future dealings. If the probability of that consequence is one half, cheating has an expected net return of minus one; if it is one fifth, the return is two. The threatened loss has stayed the same, while the rule’s grip has changed. False accusations introduce another cost, borne by the innocent and by anyone who comes to distrust the community’s reports. A widely repeated accusation and a well-founded judgment can travel through the same channels.

The cost of making that judgment and acting on it creates a second collective problem inside the first. Suppose, in another invented example, that one person can undertake an enforcement action costing two and thereby secure a benefit of one for each of ten people, including himself. The group gains eight in total, but the enforcer loses one; each of the other nine gains one. A person concerned solely with these material returns would prefer that someone else do the work, and would decline to do it himself even if no one else volunteered. The same difficulty attends the boycott that sacrifices a profitable customer, the accusation that invites retaliation, and the time spent finding out what actually happened. Jon Elster pressed the question behind these costs: who sanctions the person who fails to sanction? A further expectation of disapproval can make enforcement attractive, but its own enforcement must also be explained. Adding another unwritten rule does not by itself supply another willing enforcer.²

2. Jon Elster, Social Norms and Economic Theory (1989), pp. 99–105, especially the complete discussion of sanctions and meta-norms at pp. 104–105 and note 4. Both numerical examples are invented. The first assumes a risk-neutral choice and a present-value loss additional to the gain: 4 − 10p. The second assumes a single effective action with no other reward or cost: the enforcer receives 1 − 2; the ten-person total is 10 − 2. Neither is an estimate of actual punishment.

⁂

People sometimes bear that cost even when the experiment removes the prospect of a material return. Ernst Fehr and Simon Gächter gave students a public-goods task in which each could keep an endowment or contribute to a project benefiting everyone in a group of four. In one condition they could also pay to reduce another member’s earnings: each unit spent on punishment removed three from its recipient. The experiment ran for six periods in each condition, with the groups rearranged so that, within a condition, no participant met the same person twice; earlier individual decisions were concealed from new partners. Punishment therefore could not earn the punisher repayment from the person punished or a useful personal reputation. It nevertheless occurred, predominantly against those contributing less than their companions, and contributions were higher when punishment was available. The finding gives the willingness to enforce a place in the explanation of cooperation. It establishes that willingness in this setting, while leaving open how it is learned, why it varies, and what people elsewhere will spend to defend a different rule.³

3. Ernst Fehr and Simon Gächter, Altruistic Punishment in Humans (2002), pp. 137–140, including Methods. The study involved 240 university students and two six-period conditions. Its matching restriction applies within a treatment; the histories withheld from new partners prevented reputation formation. The authors’ separate responses to hypothetical scenarios support an interpretation involving anger, without experimentally isolating anger as the cause of punishment.

A continuing community can also organize the work so that enforcement asks less of any one person. Its members may share the watching, recognize a monitor’s contribution, or agree on a response whose modest first step leaves room for correction. In Elinor Ostrom’s account of enduring common-resource institutions, monitors are users or answerable to them; sanctions begin mildly and grow when violations recur; disputes have accessible local means of resolution. Such arrangements give the rule a memory and an interpreter, while making the exercise of judgment visible to those who bear its costs. They can join informal expectations to explicit procedures, and some depend on a wider government’s recognition of the users’ authority. Their lesson reaches beyond a romance of spontaneous harmony: self-government has work to do, positions to fill and mistakes to correct. The court’s absence creates a demand for those functions wherever the community undertakes to perform them itself.⁴

4. Elinor Ostrom, Beyond Markets and States, Nobel lecture (2009), §4.E, pp. 421–423. Her design principles describe institutional regularities associated with long survival, with substantial variation in the specific rules. They include user-accountable monitoring, graduated sanctions, local conflict resolution and recognition of the right to organize. These institutions are not all wholly unwritten or wholly separate from the state.

The ability to enforce a rule leaves the value of the rule itself to be judged. The same willingness to withhold cooperation can protect an honest transaction, shelter an incumbent from competition, or silence someone whose account would embarrass the group. Even a useful rule can be enforced at excessive cost. Ostrom’s laboratory discussion records a setting in which sanctioning raised gross benefits while its expense substantially reduced net returns; the punishment must enter the ledger beside the cooperation it induces. Elster makes the broader distinction between explaining a norm’s force and establishing its social usefulness. A practice can bind a group closely while imposing losses outside it, and a benefit that happens to follow from a rule does not explain the mechanism by which that rule survives. Censure therefore needs a judgment about the act condemned, a judgment about the response, and some way of hearing those for whom the community’s agreement supplies no protection.⁵

5. Ostrom, Beyond Markets and States, §5.A, pp. 423–425, distinguishes gross gains from the costs of sanctioning and reports improved returns under a further design permitting communication and a chosen sanctioning system. Elster, Social Norms and Economic Theory, pp. 107–114, separates benefits within a group, effects on others, and a mechanism explaining a norm’s persistence. The applications here are the plate’s argument.

What makes the unwritten rule bite is thus the community’s capacity to turn an understood obligation into a consequence that someone is prepared to administer. Regard supplies a reward worth keeping; shared information gives a good name a history; the prospect of lost cooperation gives a breach a future cost. Conscience, indignation, reciprocal support and organized monitoring can sustain the people who impose that cost, in different combinations and under different conditions. None is supplied merely by calling a practice a norm. The rule is strongest where its members can recognize a violation, trust the account of it, and undertake a proportionate response without abandoning the purposes for which they live and work together. The shut door is the visible end of the process. Behind it stands a social relation that has made the door valuable, given its closure a meaning, and persuaded someone to bear the cost of keeping it closed.