Published
Methodology
Comparative case study of nine boards, drawing on minutes, observation, and 38 director interviews
Co-Authors
Prof. Helena Marsh
Paper Length
24 pages
Cite as
Ellery, J., & Marsh, H. (2024). Consensus and its costs: Evidence from nine boards. Corporate Governance Review, 31(2), 118–142.
Copy citation
Boards are designed to reach agreement, and are seldom assessed on the quality of the disagreement that preceded it. This paper compares nine boards over four years and links early consensus to weaker risk anticipation and slower correction.
What we measured
Across nine boards we coded four years of minutes, observed 61 meetings, and interviewed 38 directors. For every material decision we recorded how long dissent stayed on the record, who raised it, and whether the board returned to the question once outcomes were known.
The pattern was consistent. Boards that reached agreement fastest were not better informed; they were better rehearsed. Papers arrived pre-aligned, objections were resolved before the meeting, and the meeting itself became a ratification exercise with a minute-taker in the room.
“
Early agreement is not evidence of clarity. It is often evidence that the difficult question was settled somewhere the record cannot reach.
The cost of pre-alignment
Boards with the shortest dissent windows anticipated fewer of the risks that later materialised, and took a median of eleven weeks longer to correct a decision once it had gone wrong. Agreement made the first decision cheap and every revision expensive.
What changed the pattern
Three boards reversed it, and none did so through culture work. They changed the paperwork: a standing record of the strongest argument against each recommendation, a named director responsible for stating it, and a scheduled return to the decision at six months. Small procedural changes, durable behavioural effect.



