Why Smart Teams Make Bad Decisions Together

A group of capable people can reach a worse decision together than most of them would reach alone, and the cause is rarely a lack of intelligence. People in a group want to keep the peace. They also want to stay on good terms with whoever has the most power, so they talk about what everyone already believes and keep their doubts to themselves. Leaders who understand that pull can set up their decision meetings so that disagreement is expected, and so that the facts, and not the loudest voice, settle the question.

How Bright People Talked Themselves Into the Bay of Pigs

The psychologist Irving Janis coined a term for this pattern in the early 1970s. He called it groupthink. He built the idea from a study of decisions made by some of the most qualified advisers in the United States government.

In 1961, President Kennedy and his advisers approved a plan for a force of Cuban exiles to land at the Bay of Pigs and overthrow Fidel Castro, on the assumption that the Cuban people would rise up to join them once the landing began. Few of the advisers pressed on that assumption, and the invasion collapsed within days. The men who approved it had years of experience with hard problems. Some of them had serious doubts about the plan. Arthur Schlesinger, one of the advisers, later wrote that he reproached himself for staying nearly silent during those meetings. The group had seemed so sure, and he explained that an objection would have done little except earn him a reputation as a nuisance.

Janis found the same features in other failed decisions, starting with a group's belief that it could not fail, and he saw that members censored their own doubts to protect the group's sense of unity. Each person took the silence of the others as agreement, and nobody went looking for evidence against the plan, because the plan already felt settled.

What Groups Spend Their Time Discussing

A second pattern makes groupthink worse, and researchers have measured it. In the 1980s, Garold Stasser and William Titus ran experiments in which each member of a group received some information that everyone shared and some that only they had. The groups spent most of their discussion on the shared facts, and the information held by one person, which was often the information that pointed to the best choice, came up far less and was easy for the group to dismiss when it did.

Ordinary workplace meetings follow the same pattern for plain reasons. Repeating a fact everyone knows feels safe, because the others nod along. Raising a fact nobody else has heard means standing alone for a minute and possibly being wrong in front of colleagues. The person closest to the problem, the nurse who saw the near miss or the analyst who spotted the odd number, is often the one with the unshared information and the least standing to push it.

Rank adds to the effect, because when the most senior person speaks first, everyone else learns which answer is the safe one, and the discussion that follows tends to circle that answer, though a confident expert lower down can sometimes break the pattern.

Letting Reality Decide

Positive Leadership's approach to communication, called Communication for Alignment, starts with a principle we call Advocate Reality. When the facts do not decide a question, a battle of wills decides it. The person with the most power usually wins that battle. A team that agrees to let reality decide has a way to disagree that does not depend on who outranks whom.

A shared Purpose, Outcomes, and Methods (POM) makes that easier to do. When everyone is aligned on the Purpose and the Outcomes, a disagreement about Methods becomes a question about which route is more likely to reach an Outcome the whole team wants. People can argue about the route without anyone's loyalty being in doubt. A member who says "I think this tactic will fail" is helping the team reach its Outcome.

Advocating reality in a decision meeting starts with sorting what the group knows from what it is assuming. The Bay of Pigs plan depended on an assumption presented as a fact, and one question about the evidence for a popular uprising might have changed the discussion.

The premortem, developed by the psychologist Gary Klein, gives people a safer way to raise doubts. Before committing to a plan, the group imagines that it is a year later and the plan has failed. Each person then writes down the reasons it failed. People who would never attack a plan their boss supports will often list its weak points freely when the failure is treated as a given, because they are explaining a story and nobody is asking them to oppose a colleague.

What Kennedy Changed the Next Year

Janis also studied the Cuban Missile Crisis of 1962, when many of the same advisers faced a far more dangerous decision and handled it far better. Kennedy changed how the group worked by staying away from some of the meetings so his presence would not steer the discussion, and he asked his brother Robert and his aide Ted Sorensen to challenge every proposal. The group also split into smaller sets that worked on competing options before coming back together. The advisers argued for days over a blockade versus air strikes. The crisis ended thirteen days after it began, with the Soviet Union agreeing to remove its missiles from Cuba. Janis credited much of the difference to the changes in process, since the people were largely the same.

A leader can use the same moves in an ordinary meeting. The leader's own view can wait until the others have spoken, which keeps people from guessing which answer is safe. Each person can write down a position before the discussion starts. Doubts that are in writing are harder to abandon. One person can be assigned to argue against the leading option. Rotating that role keeps it from becoming anyone's reputation. When a member has information nobody else has, the leader can ask that person directly and give it the same time the shared facts got.

These practices improve the odds of a good decision. They also show people how the leader treats bad news. A team that watches a leader welcome a hard objection, and then change course because of it, brings the next objection sooner.

The next decision meeting on a leader's calendar is a good place to start, with one change made before it begins, which is deciding who will speak first and making sure it is not the leader.