Collective Decision-Making
Why Groups Often Know More Than Their Smartest Member — And When They Don't
The terrifying thing about groupthink isn't that everyone agrees — it's that everyone secretly disagrees and says nothing.
The Idea
There's a seductive idea in decision theory called the wisdom of crowds: under the right conditions, the aggregated judgments of many independent people outperform even expert individuals. Francis Galton noticed this in 1907 when a crowd at a county fair estimated the weight of an ox with more accuracy than any single judge. The average guess was almost exactly right. The logic is elegant — individual errors are random and tend to cancel out, leaving something close to truth. But the conditions matter enormously, and this is where the idea gets interesting. The crowd only outperforms the expert when its members are genuinely independent — when they haven't heard each other's answers, haven't been anchored by a confident voice in the room, and aren't nudging themselves toward consensus to avoid social friction. The moment people start influencing one another, errors stop cancelling out. They cascade. A few confident wrong answers corrupt the pool, and the group converges on a fiction everyone privately doubts. This is the dark twin of collective wisdom: herding. It's why financial bubbles form, why unanimous jury verdicts are sometimes the least reliable, and why boardrooms staffed with brilliant people still produce baffling decisions. The mechanism is the same in all cases — information that exists in people's heads never makes it into the shared pool, because the social cost of dissenting feels higher than the cost of staying quiet. Collective intelligence isn't a property of groups. It's a property of how groups are structured to think.
In the World
In the lead-up to the Space Shuttle Challenger disaster in January 1986, engineers at Morton Thiokol had serious concerns about launching in cold temperatures. The rubber O-rings sealing the rocket boosters had never been tested below a certain threshold, and the forecast was brutal. Some engineers knew the risk was real. They said so internally. But the decision-making process had a fatal structural flaw: it asked engineers to prove the launch was unsafe, rather than asking NASA to prove it was safe. Under pressure, with a launch schedule and reputation at stake, the engineers' objections were gradually softened, reframed, and ultimately overridden. The group reached consensus — not because everyone genuinely agreed, but because dissent had become structurally costly. When sociologist Diane Vaughan later studied the disaster exhaustively in her book The Challenger Launch Decision, she described what happened not as recklessness but as 'normalisation of deviance' — a group process in which warning signals were absorbed and reclassified as acceptable risk through repeated exposure. The information existed. The judgment existed. What failed was the architecture of the decision: who spoke first, who held power, whose framing defined the question. Seven people died because the group's collective knowledge never actually became the group's collective decision. The wisdom was present. The structure prevented it from mattering.
Why It Matters
Most of us live and work inside groups — families, teams, committees, communities — and we make important decisions together constantly. The natural instinct is to treat consensus as a sign of good thinking. If everyone agrees, the reasoning must be sound. What decision theory reveals is that consensus is almost meaningless as a quality signal. It tells you about social dynamics far more reliably than it tells you about truth. The real question to ask when a group reaches agreement is: did the dissenting information actually get heard, or did it quietly withdraw from the room? Building better collective decisions isn't really about intelligence or even good values — it's about process design. It means creating conditions where the quieter, less confident, socially riskier opinion still gets aired. It means asking people to commit their views privately before the group discusses. It means nominating a devil's advocate and taking them seriously. For you, today, this might mean paying less attention to what your group decided and more attention to who went quiet just before you reached that decision — and what they might have been about to say.
A Question to Ponder
Think of the last important decision you made as part of a group: was the outcome shaped by the best information in the room, or by who spoke first and with the most confidence?
Get a new one of these every morning.
Start learning with Thinkable