Your team agrees with you. That should worry you.

Everyone nodded in the meeting. The decision was taken, the actions were agreed and the room moved on. Two weeks later nothing has happened, and you hear in the corridor that half the room never agreed in the first place.

Most leaders have lived through some version of this. The research on group pressure explains why it happens, and, more usefully, how little it takes to stop it.

We adjust to the room

People adjust what they say to fit the group, even when little is at stake. In a 2026 study, 360 Americans took part in a group chat with three "colleagues", who were in fact AI, about issues such as gun ownership, immigration and vaccine mandates. Half were placed in a chat that leaned liberal and half in one that leaned conservative. Over five rounds of messages, neutral positions faded and people's replies increasingly matched the room's view. Those who expressed views against the room reported more irritation and anger afterwards.

The researchers were careful about what this shows. They measured what people wrote, not what they privately believed. As they put it, "apparent agreement in a conversation may give an incomplete picture of the views people privately hold." For a leader, that sentence is the whole problem.

The pattern is not new. In Solomon Asch's experiments in the 1950s, people were asked which of three lines matched another in length, an easy question with an obvious answer. When a group of actors all gave the same wrong answer first, the real participant went along with them on about a third of the questions, and about three quarters went along at least once.

One ally breaks the pull

The same experiments show how fragile that pressure is. When one other person in the group gave the right answer, going along with the wrong majority fell from about a third to about 5%. When people wrote their answers down privately instead of saying them aloud, it fell to about 12%.

That is the most practical finding in this area. As I read it, group pressure does not need to be overcome with courage or training. It needs one dissenting voice, or a way for people to give their view before they know what everyone else thinks.

Authority still works

Pressure from above adds a second force. Stanley Milgram's experiments in the 1960s, in which volunteers were told to give what they believed were increasingly painful electric shocks to another person, are among the most famous in psychology. They have been repeated with modern safeguards.

In 2009 an American replication stopped the experiment at the point where most people hesitate, and found that 70% of volunteers were prepared to continue, close to the 82.5% Milgram found. In a 2025 Polish version with 80 volunteers, 88% followed every instruction from a female professor and 90% from a male one. The difference was not significant. Authority, the authors concluded, "knows no gender".

But people do not simply follow

The famous experiments are often summarised as proof that people blindly follow. The detail tells a different story.

In Asch's experiments, most answers were independent: participants gave the right answer on about two thirds of the questions, and a quarter never went along with the group at all. In the obedience studies, researchers who have re-examined the evidence argue that people continued because they believed in the scientific purpose the experimenter stood for, not simply because they were told to. They describe this as "engaged followership", and note that participants almost always refused when given a bare order to continue.

My reading is that people follow leaders whose purpose they share, and resist leaders who simply give orders. That is a more hopeful conclusion, and a more demanding one.

The anxious follow the bad news

Group pressure does not land equally on everyone. In a small 2026 study of 44 young adults, people with high social anxiety rated social scenes again after seeing what a supposed group of peers thought. They moved their ratings towards the group when the group was more negative, but largely ignored the group when it was more positive.

The study is small and needs repeating. If it holds, it suggests that the most anxious people in a team may be the quickest to adopt the group's worries, and the slowest to adopt its confidence.

Groups talk about what they already share

A further problem sits underneath all of this. Since the 1980s, studies of group decisions have found that groups spend most of their time discussing information everyone already has, and too little on what only one or two members know. When the best decision depends on that unshared information, groups often miss it. A team can be unanimous and still wrong, because the person who knew something different never said it.

What this is costing you

None of these studies was run in a boardroom, so applying them to leadership is my own step. The patterns will be familiar, though:

• the meeting where the most senior person speaks first and everyone else agrees;
• the decision that is "agreed" in the room and quietly resisted afterwards;
• the analyst who knew the numbers were wrong but did not want to be the only one to say so;
• the risk that everyone saw and nobody raised.

The cost is not visible on the day. It shows up later as decisions that do not happen, problems that were known but never voiced, and leaders who believe they have more support than they do.

Questions worth asking

A few honest questions show how much of this is happening around you.

• When did you last see someone disagree with the most senior person in the room, and what happened next?
• In your meetings, do people give their views before or after you have given yours?
• When someone raises an objection, how many others join in, and how quickly?
• What does someone on your team know that has never been said in a meeting?

What you can do this week

Everything in this section is a suggestion of mine, drawn from the research rather than tested by it. Each one takes minutes and costs nothing.

Suggestion

If you lead a team

• Collect views in writing first. Before a decision is discussed, ask everyone to write down their view and their reasons, then read them before anyone speaks.
• Speak last. Give your own view after the others have given theirs, not before.
• Thank the first dissenter. When someone disagrees, thank them in front of the room. One voice makes it far easier for others to follow.
• Ask for what is not shared. Ask each person, "What do you know about this that the rest of us might not?"
• Explain the purpose, not just the instruction. People follow a purpose they share more readily than an order.

Suggestion

For yourself

• Write your view before the meeting. Note what you think and why before you hear anyone else, then compare it with what you end up saying.
• Be the ally. When someone raises a point you agree with, say so out loud. You may be the one voice that makes it safe.
• Notice the pull of the room. When you find yourself agreeing quickly, ask whether you would have said the same thing first.

Suggestion

At home

• Choose privately, then vote. When the family picks a film or a meal, ask each child to choose privately before anyone says what they want.
• Make disagreement normal. When a child disagrees with the rest of the family, thank them for saying so, even if the decision goes the other way.

Sources

• Mitsui, C., & Kawamura, Y. (2026). People shift political expression to fit the room in small online group chats: Evidence from AI-simulated chat experiments. Computers in Human Behavior: Artificial Humans. doi:10.1016/j.chbah.2026.100331
• Asch, S. E. (1956). Studies of independence and conformity: I. A minority of one against a unanimous majority. Psychological Monographs: General and Applied, 70(9), 1 to 70.
• Burger, J. M. (2009). Replicating Milgram: Would people still obey today? American Psychologist, 64(1), 1 to 11.
• Grzyb, T., Dolinski, D., & Cantarero, K. (2025). Authority knows no gender: Gender effects in exerting obedience in Milgram's experiment. Social Psychology, 56(2), 85 to 97. doi:10.1027/1864-9335/a000575
• Reicher, S. D., Haslam, S. A., & Smith, J. R. (2012). Working toward the experimenter: Reconceptualizing obedience within the Milgram paradigm as identification-based followership. Perspectives on Psychological Science, 7(4), 315 to 324.
• Cui, L., Gao, C., Yan, X., & Dong, X. (2026). Aberrant emotional conformity among individuals with social anxiety. Biological Psychology. doi:10.1016/j.biopsycho.2026.109276
• Stasser, G., & Titus, W. (1985). Pooling of unshared information in group decision making: Biased information sampling during discussion. Journal of Personality and Social Psychology, 48(6), 1467 to 1478.