The Abilene Paradox

Is it a paradox at all? It sits in the same family as the Prisoner's Dilemma — individually reasonable choices producing a collectively terrible outcome — but the mechanism is different in an important way, and the difference is where the interest lies. In the Prisoner's Dilemma, the incentives really are opposed; each prisoner gains from betraying. Fix the information and the trap remains. In Abilene, nothing whatsoever is opposed. Everyone wants the same thing, and communication alone would resolve it completely. The whole disaster is made of misinformation that the participants generated themselves, for free, out of politeness.

How solid is the evidence? Harvey's article is a piece of essayistic management writing built on a family anecdote and two organisational cases, not an experiment, and it should be read as one. What has held up is not Harvey's particular framework but the underlying phenomenon, which was described in social psychology well before him. Floyd Allport introduced pluralistic ignorance in the 1920s; Daniel Katz and Floyd Allport documented it in 1931 among students who each privately disapproved of a campus norm they all believed the others endorsed. Later work found the same pattern in college drinking, where students substantially overestimate how comfortable their peers are with heavy drinking, and in surveys of racial attitudes in the mid-twentieth-century United States, where private opinion had shifted considerably ahead of what people believed their neighbours thought. Harvey's contribution was to give the organisational version a name and a story that people remember.

Pluralistic ignorance or preference falsification? These are two different diagnoses and the distinction changes what you should do. In pluralistic ignorance, people are simply mistaken about what others think, and telling them the truth fixes it. In preference falsification — the economist Timur Kuran's term, from his 1995 book Private Truths, Public Lies — people misrepresent their views because there is a real social cost to honesty, and telling them the truth changes nothing while the cost remains. Harvey's porch is clearly the first kind: the cost was imagined. Many organisational cases are the second kind, where the person who objects to the executive's pet project pays a genuine price. Calling a preference-falsification problem "an Abilene paradox" is comforting and useless, because it implies the fix is a candid conversation when in fact the fix is changing what happens to people who speak.

How would you ever know? The paradox is defined by private states that, by construction, nobody expressed at the time. That makes it awkward to demonstrate. Every account of it, including Harvey's, is reconstructed after the fact from people who have just discovered they wasted an afternoon — and people in that position have an obvious motive to remember themselves as having been sceptical all along. Anonymous polling before the decision is the only way to catch it live, which is part of why anonymous pre-voting is both a good remedy and the main source of decent evidence that the phenomenon is real.

The label is over-applied. Not every bad group decision is a trip to Abilene. The Challenger launch decision of 1986 is often cited as an example and mostly is not one: Thiokol's engineers, Roger Boisjoly among them, objected clearly and in writing, and were overruled. That is a failure to act on stated dissent, which is a different and in some ways worse problem. The Abilene test is strict — the objection has to be unspoken by everyone. If somebody said it out loud and was ignored, you are looking at something else.

July, in the 1970s, Coleman, Texas. Jerry B. Harvey, then a professor of management science at George Washington University, was visiting his wife's parents. The afternoon happened as described: the porch, the fan, the dominoes, the father-in-law's suggestion, the fifty-three miles each way in an unairconditioned Buick, the disappointing cafeteria, and the conversation afterwards in which the whole thing came apart. Harvey has always presented it as a true account of his own family, and it is by some distance the most effective piece of writing in the management literature.

1974. Harvey published "The Abilene Paradox: The Management of Agreement" in Organizational Dynamics, volume 3, issue 1, pages 63–80. The article opens with the drive, then works through two organisational cases and sets out the six symptoms and his diagnosis — action anxiety, negative fantasies, and fear of separation from the group. The central claim is stated flatly: organisations fail at least as often from an inability to manage agreement as from an inability to manage conflict, and the second failure gets all the attention.

The context it landed in. Irving Janis had published Victims of Groupthink two years earlier, in 1972, and the idea that groups make bad decisions was suddenly fashionable. Harvey's article is in part an argument against being absorbed into that fashion. His insistence that the Abilene paradox is not groupthink is the reason the entry is worth having separately, and it is also the reason the two have been conflated ever since.

1984 onwards. A training film based on the story, produced for corporate audiences, carried the idea far beyond people who read journals; it became one of the most widely used management training films of its era and was remade in the 1990s. "Are we on the road to Abilene?" entered boardroom vocabulary as a way of raising the question without accusing anybody.

1988. Harvey collected and extended the argument in The Abilene Paradox and Other Meditations on Management, pushing it into harder territory: organisational collusion, the treatment of whistleblowers, and what he considered the ethical weight of going along with something you believe to be wrong. The later essays are less comfortable than the original article, and deliberately so — Harvey's view was that the polite silence on the porch and the polite silence around serious misconduct are the same behaviour at different stakes.

The lasting contribution is a distinction that changes what you do in a room.

Before Harvey, the standard diagnosis of a bad group decision was that dissent had been suppressed, and the standard fix was to protect dissent. That fix is correct — for the cases it fits. Harvey identified a large and previously unnamed category where it does nothing at all, because there was no dissent to protect. Everybody was already on the same side. What failed was not the group's tolerance for disagreement but its ability to discover an agreement it already had.

Once you have the distinction, you can see how much of ordinary organisational machinery is built to solve the wrong one of the two problems. Devil's advocates, red teams and "any objections?" all assume the risk is a majority steamrolling a minority. None of them help a room where every person is privately hoping somebody else will say it.

The second contribution is the reframing of silence. It is normal to treat a quiet room as a settled room, and the quieter it is the more settled it seems. The Abilene paradox says the opposite can be true: unanimity that arrives quickly, with no friction and no questions, is not evidence of agreement. It is evidence of nothing at all, because agreement and universal private objection produce exactly the same sound.

There is also a small, useful piece of self-knowledge in it, which is probably why the story has travelled so well. Most people, reading about the drive to Abilene, recognise themselves immediately — not as the victim of a group that went somewhere silly, but as one of the four who said "sounds good to me". The cost of being the first to speak is consistently overestimated, and the relief of everyone else in the room is consistently underestimated. Those two errors, running in every head at once, are enough to move a group anywhere.

Meetings that approve things nobody believes in. The commonest modern form is a project sign-off where each attendee assumes the enthusiasm is real and located elsewhere in the room — with the sponsor, or with the executive, or with "the business". The tell is a decision that passes with no questions and is privately grumbled about within the hour. Where the pattern is chronic, the cause is usually structural rather than personal: the meeting is designed for ratification, so the only available inputs are assent and awkwardness.

Anonymous pre-voting. The most direct remedy in current practice, and a straight implementation of Harvey's prescription. Before any discussion, each participant privately records their position — a number, a confidence level, a would-you-fund-this — and the distribution is revealed all at once. Nobody is identifiable, nobody has to go first, and the distribution frequently looks nothing like the room. It is the same logic as blind bidding: you cannot anchor on a signal you have not seen.

Premortems. Gary Klein's technique, popularised by Daniel Kahneman: before committing, the group is told to imagine it is a year later and the project has failed badly, and each person writes down why. This works on Abilene for a reason worth noticing. It does not ask anyone to object — objecting is what people will not do — it asks them to explain a failure that has been stipulated. The social cost of being the difficult one is removed by the framing, and the private doubts come out as an assignment rather than as dissent.

"Disagree and commit." Widely adopted in technology companies, and a genuine partial fix, since it makes disagreement a normal contribution rather than an act of obstruction and promises the objector that raising it will not stall the work. Its failure mode is exactly the one to watch: in practice the phrase is frequently deployed at the commit end, to close a discussion, rather than at the disagree end, to open one. Used that way it becomes a faster road to Abilene with better vocabulary.

Remote and asynchronous work. Written, asynchronous decision-making cuts both ways here. A comment thread makes silence even cheaper and even more ambiguous than silence in a room — a document with no comments looks endorsed. But written-proposal cultures, where reviewers are required to record a position rather than merely attend, remove the guessing that the paradox runs on. The mechanism is the same as the anonymous poll: the failure mode is inferred consent, so any process that makes each person state something is an improvement.

Safety-critical work. Aviation crew resource management and the surgical safety checklist both, in part, exist to break the assumption that a junior person's silence means agreement. Requiring each team member to speak — name, role, concerns — before an operation or a departure is not ceremony. It is the recognition that a person who has said nothing has not been consulted.