Parkinson, in 1957, imagined a committee with three items: a nuclear reactor, a bicycle shed, and the coffee budget.

The reactor is approved in two and a half minutes. Nobody understands it, nobody wants to admit that, and the plans are too complex to argue with. The bicycle shed takes forty-five minutes, because everybody knows what a shed is and can have a view on the roofing material. The coffee budget takes an hour and fifteen, because everyone drinks coffee.

Attention goes where competence to comment goes, and that has almost nothing to do with what is at stake.

Why it is not stupidity

Each person is behaving reasonably. Contributing on the reactor risks exposing that you do not follow it. Contributing on the shed is safe, demonstrates engagement, and is genuinely within your judgment. The incentive to participate is highest exactly where the value of participation is lowest.

It follows that a silent room on a large decision is a signal rather than agreement. Silence usually means the group cannot evaluate the item, which is the moment to slow down rather than to record approval.

What helps

Put money and time on every item. An agenda that says "reactor — $4M, 20 years" and "shed — $3,000" makes the mismatch visible in the room.

Decide the large things in a different forum. Small groups with the relevant knowledge, in writing, before the meeting. Group discussion is a poor instrument for anything most of the group cannot assess.

Treat unanimous speed as a warning. The fastest agreement in the meeting is usually the item nobody understood.

Give the trivial item away. Assign it to one person with a budget. Most bikeshedding is the group having no other outlet for wanting to contribute.

This is the sibling of Parkinson's law and the two travel together: work expands to fill the time, and the expansion collects around whatever is easiest to have an opinion about.