n o ren
Human Performance & Leadership

More Safety Checks, Faster Disasters

In 1986, NASA’s Challenger crew survived a pre‑launch test that caught a faulty O‑ring, yet the shuttle still exploded minutes later.

Leaders love checklists because they promise certainty, but each added layer of approval can paradoxically accelerate catastrophic error. The first safety sign‑off creates a “completion bias” – the team feels the problem is solved and stops looking for hidden flaws.

When the next manager receives the cleared list, the implicit signal is “go ahead,” so dissent is muted and the clock ticks faster toward execution. This dynamic was starkly illustrated during the Challenger launch: engineers at Morton Thiokol warned about O‑ring erosion in cold weather, but senior officials, having received a clean “go” from lower‑level safety reviews, pressed on, interpreting the sign‑off as a green light rather than a pause.

The result was a tragic explosion that could have been averted if the final decision had required a fresh, independent risk assessment rather than relying on prior approvals. The lesson is that each successive sign‑off can become a shortcut to confirmation, not a barrier to failure.

A sign‑off is a decision point, not a safety guarantee.
Completion bias makes teams treat a cleared checklist as a problem solved.
Independent re‑evaluation after a sign‑off restores critical scrutiny.
Visible dissent after a sign‑off signals hidden risk, not disloyalty.
The faster a team moves past a sign‑off, the higher the probability of an undiscovered failure.
Embedding a “second‑look” pause after any approval cuts the speed‑risk trade‑off dramatically.

Ignoring the completion bias lets a single missed fault cascade into a full‑scale catastrophe.

Over‑reliance on layered approvals erodes team psychological safety, silencing critical voices when they matter most.

1
Open the last three project post‑mortems, count how many contain “final sign‑off” as the decisive factor; note if any mention “no further review.”
2
In today’s next decision meeting, ask each participant to state one “unknown” they still have after the checklist is completed; tally the responses.

The phenomenon traces back to cognitive psychology research on “goal completion” effects, where achieving a subgoal reduces vigilance for subsequent errors. Daniel Kahneman and Amos Tversky described this as “premature closure” in judgment, a bias that intensifies when procedural milestones are celebrated.

While additional reviews can slow progress, the cost of a single missed failure often dwarfs the marginal delay. In high‑stakes domains like aerospace, the “fast‑fail” mantra is inverted: deliberate slowing after each checkpoint yields net speed by preventing costly rework.