Inferences travel up a hierarchy while the evidence for them does not, so senior levels receive conclusions they cannot check.
Often misattributed to James D. Thompson, whose contribution is buffering the technical core. It is why senior decisions rest on summaries whose original evidence nobody upstream can now retrieve.
In practice
- A ward's staffing concern becomes "pressures noted" in the divisional report and "no material issues" in the board pack, and every step was written honestly.
- A board approves a supplier rated amber for risk, and nobody in the room could say which of the eleven underlying checks produced the rating.
- Possible explanations
- The explanation is summarising: each level passes up a conclusion rather than the raw material, because the material is too voluminous to transmit, so uncertainty is stripped out at every step and seniors receive confidence they cannot audit.
- In its favour
- The mechanism is well documented in organisational research and in accident inquiries, where reconstructed decision chains repeatedly show warnings that were qualified at the bottom and firm by the top.
- Against it
- It describes a tendency rather than a rate, nobody has measured how much uncertainty is lost per level, and it is often misattributed to Thompson, whose contribution was buffering the technical core.
- Standing
- Often misattributed to James D. Thompson, whose contribution is buffering the technical core.
- Subject
- Organizations & Bureaucracy
- Essay
- How organisations behave when nobody is steering