Why people follow a leader who is merely certain

Nine people, an ambiguous situation and four minutes to settle it. One of them says we go left, without qualification or hedging of any kind. Everyone goes left. Afterwards nobody in the room can reconstruct why that person was treated as knowing something the rest did not, and none of them asked at the time.

What everyone sees

Groups are assumed to follow expertise, so the confident person must have been read as competent. Competence was not observable in four minutes. Confidence was, and it functions as a claim about the state of the evidence that costs nothing to make and is not checked until the outcome arrives.

What is actually happening

Anderson, Brion, Moore and Kennedy tested a status-enhancement account of overconfidence and found that people who overestimated their own abilities were accorded higher status by their groups, with the effect running through the confident behaviours peers could observe rather than through any actual advantage in performance. Sah, Moore and MacCoun examined advisers directly and found that confidently stated advice was rated more credible and followed more often than accurately hedged advice, and that advisers who expressed appropriate uncertainty lost influence even when their calibration was better. The two results describe a stable market failure in group judgement. Confidence is rewarded on delivery and accuracy is settled later, usually with no record connecting the two.

Why it stays hidden

The mechanism hides because outcomes are noisy enough to protect it. When the confident call is right it confirms the follow; when it is wrong the situation was difficult. Calibration would require tracking stated certainty against results across many decisions, which almost no group does and no meeting minute supports.