The committee of one

A common piece of workplace folklore says that decisions should be made by groups because groups are wiser than individuals. The research is less flattering and more specific.

Groups are better at some tasks and worse at others, and the dividing line is reasonably clear. For problems with a verifiable answer — estimating a quantity, spotting an error — collecting independent judgements and combining them beats almost any individual. For problems requiring a choice between options with no checkable answer, groups often do worse than their best member, because discussion converges on whoever speaks most confidently rather than whoever is most likely to be right.

The practical consequence is that the method should follow the problem. Independent estimates, written before discussion, then averaged, work well. Open discussion followed by a vote works badly, because the vote is taken after everyone has heard the loudest person.

There is a second finding that matters more in schools. A group that knows it will have to produce a single answer begins converging almost immediately, often within the first two minutes, long before the alternatives have been examined. Groups told to produce three options and rank them explore substantially more ground, and the ranking can be done in the last five minutes.

None of this says committees are useless. It says that the shape of the meeting decides most of what the meeting produces, and the shape is usually inherited rather than chosen.