Conformity

Conformity is the tendency to voice or follow a group’s judgment even when it conflicts with one’s own private view. In the book’s account, it is not simply a failure to think: people may see the answer clearly and still go along because disagreeing is socially harder than changing what they say. The distinction matters for collective intelligence. Diversity contributes not only by bringing different perspectives into a group, but by making it easier for members to state what they actually believe.

Conformity and Groupthink

Conformity is related to groupthink, but the two operate differently. Groupthink can make dissent seem unlikely in the first place: an assumed consensus gains force as members’ cohesiveness turns its appearance into something real, and doubts are dismissed or rationalized away. Conformity, by contrast, describes a person changing an expressed opinion not because the group has persuaded them, but because challenging it feels more difficult than going along. Both can close discussion rather than open minds, but conformity can occur even when a person privately recognizes that the group is wrong.

That difference makes the social cost visible in the individual’s conduct. A group does not need to censor an objection explicitly. It can be enough for a person to feel alone in holding it. The resulting agreement may look like a shared judgment while concealing private disagreement; the appearance of unanimity then deprives the group of information it needs to make a sound decision.

Asch’s Line-Judgment Experiment

The classic demonstration came from Solomon Asch’s experiment with groups of seven to nine people. Each group included one unwitting subject and several experimenter’s confederates. The subject sat at the end of the row, and everyone was asked aloud to identify which of three lines matched a line printed on a white card. For the first two of twelve cards, the group gave the same, correct answer. Beginning with the third, the confederates deliberately chose lines that were plainly not the same length as the displayed line.

The subject therefore had to answer after hearing a public consensus contradict a visible fact. Some subjects shifted their heads, stood to inspect the lines more closely, or joked nervously about whether they were seeing things. Their uncertainty was not evidence that the lines had become hard to compare; it showed the strain of facing a group whose confident answers conflicted with their own perception.

A significant number nevertheless joined the false majority. Most subjects gave their real opinion most of the time, but 70 percent changed their answer at least once, and a third went along with the group at least half the time. In interviews afterward, most emphasized their wish not to stand out. They did not say they had come to believe that the unequal lines matched; they had said so to avoid being the lone dissenter. The experiment thus separates public compliance from private conviction.

The Effect of an Ally

Asch’s experiments also showed that conformity was not inevitable. In one variation, a confederate chose the line that actually matched the card instead of echoing the false majority. This gave the unwitting subject even a single ally, and the rate of conformity fell sharply. With someone else willing to state the same judgment, subjects became more willing to say what they thought.

The result gives diversity a practical function beyond the addition of new ideas. A different voice can alter the conditions under which everyone else speaks: it reduces the isolation that makes silence or false agreement attractive. Independence of judgment is essential to wise group decisions, yet difficult to preserve; a group’s composition can help protect it by ensuring that dissent does not appear to belong to one person alone.

Sequential Decisions and Private Information

The same vulnerability to other people’s expressed choices becomes more consequential when decisions are made in sequence. In a classroom experiment devised by economists Angela Hung and Charles Plott, participants had to infer which of two urns had been chosen. Urn A held twice as many light marbles as dark ones, while Urn B held twice as many dark marbles as light ones. Volunteers drew in order, kept the color of their own marble private, and announced only their guess. A correct answer earned a couple of dollars. Each participant thus had private evidence from a draw and public evidence from the guesses already announced.

When a participant drew a light marble but had heard three people guess Urn B, most guessed Urn B as well. That response was individually rational given the apparent evidence from the group, but it allowed earlier answers to outweigh later participants’ private information. Information cascades began in 78 percent of the trials. The experiment illustrates how sequential choices can make people converge without adding their separate evidence to the collective judgment.

Hung and Plott changed the incentive: instead of being paid for an individually correct answer, students were paid according to whether the group’s answer, determined by majority vote, was correct. Because the reward depended on collective accuracy, participants had reason to attend more closely to their own information and less to the guesses already made. Individual private evidence was imperfect, so relying on it could mean making a wrong guess; nevertheless, the group became significantly more accurate than the cascading groups. Making decision-makers more independent improved the group’s result. The practical implication is that organizations can improve judgment by having members offer their assessments simultaneously, or close to it, rather than letting each person’s answer be shaped by those who spoke first.

Incentives, Cooperation, and Truthful Reporting

Conformity belongs to a wider problem of how social and material incentives shape what people contribute. In public-goods experiments by Ernst Fehr and Simon Gächter, four players each began with twenty tokens and played four rounds. A token placed in the common pot cost its contributor one token, returned 0.4 tokens to that player, and gave 0.4 tokens to each of the other three players. The group gained 1.6 tokens for each contribution. If everyone kept their tokens, each left with twenty; if everyone contributed all of them, each left with thirty-two. Yet the individually attractive strategy was to contribute nothing and benefit from everyone else’s contributions. Most participants initially put in about half their tokens, but contributions fell as players saw others free ride. By the final round, 70 to 80 percent were free riding, leaving the group poorer than it could have been.

The example extends the question from whether people dare to disagree to whether a group’s rules make cooperation worthwhile. People’s willingness to contribute depends partly on what others appear to be doing. When observed behavior signals that others are taking advantage, cooperation erodes; a group’s visible choices can therefore shape the choices that follow.

Workplace performance systems can also reward distorted reports rather than candid effort. In the early 1950s, sociologist Donald Roy worked as a lathe operator in a machine shop where workers were paid by the piece. Their rate rose after they crossed successive output targets and stopped rising after a final cap. If they worked too quickly, the company might raise the hurdles, so workers restricted their output and focused on manipulating the rate system rather than producing as much as they could. Roy called his account “Goldbricking in a Machine Shop.” The incentive scheme turned restraint into a rational response to the risk that greater productivity would bring a higher target.

A related problem arises when corporate bonuses depend on surpassing targets. Managers have reason to set targets low by understating prospects, then to meet those targets by methods that can include accounting maneuvers that improve current results at the expense of the future. Under such arrangements, apparent agreement with organizational expectations may hide information about performance and prospects that decision-makers need. Across the experiments and workplace examples, the recurring issue is not simply that people follow others: it is that the conditions of decision and reward can suppress independent evidence or make misleading behavior pay. A group’s judgment is more dependable when people can state their views without standing alone and when its rules reward the information and cooperation the group actually needs.