Normalization of deviance
Also known as: Normalisation of deviance
Rule-breaking that goes unpunished quietly becomes the new standard, until it fails.
What it means
Normalization of deviance is the process by which a group gradually comes to treat a departure from its own standards as normal, because the departure has been repeated without visible harm. The mechanism is learning from outcomes rather than from risk: each uneventful violation looks like evidence that the rule was over-cautious, so the shortcut is reclassified as acceptable practice and becomes the baseline from which the next small step is taken. Vaughan traced it through NASA's tolerance of O-ring erosion before Challenger, and it recurs in aviation, clinical checklists, construction, and engineering on-call culture. It is easiest to diagnose in hindsight, and the label can flatter the analyst, since some rules genuinely are too tight and loosening them is adaptation rather than decay. It matters because organizations rarely fail by deciding to be unsafe; they drift there one reasonable-looking exception at a time.
Examples
A ward skips the two-person check on a drug dose one busy night, nothing goes wrong, and within a year the shortcut is simply how the ward works.
On site, workers stop clipping harnesses for quick roof jobs. Nobody falls, supervisors stop mentioning it, and the exception for 'quick' jobs quietly widens until it covers most of them.
A team merges hotfixes straight to production to hit a deadline. It works, so the review step becomes optional, and by the next quarter nobody recalls it was ever mandatory.
First described in Diane Vaughan (1996), Challenger launch decision.