Normalization of deviance
Normalization of deviance is a sociological concept introduced by Diane Vaughan to describe the gradual process by which unacceptable practices or conditions become accepted as normal over time. The term gained prominence through Vaughan's analysis of the Space Shuttle Challenger disaster, in which the progressive acceptance of O-ring erosion on earlier shuttle flights created the conditions for catastrophic failure. What was initially an anomalous and alarming finding — rubber O-rings burning through — was gradually redefined as an acceptable engineering uncertainty as repeated flights succeeded despite the anomaly.
The mechanism is insidious because it operates without conscious decision. No one chooses to accept greater risk. Each step is small, each justification is plausible, and the cumulative effect is a fundamental shift in what counts as normal. The organization drifts into failure not through a single catastrophic choice but through the slow erosion of safety margins that were never explicitly abandoned. This is why normalization of deviance is particularly dangerous in high-reliability organizations: the very success of the safety systems creates the conditions for their own undermining.
Normalization of deviance connects to institutional blindness, success trap, and path dependence. It is a form of institutional blindness in which the institution's own success screens out the signals of its own degradation. It is a success trap because past success — flights that survived despite O-ring erosion — becomes the justification for continued risk-taking. And it is path-dependent because each normalization step makes the next step easier, locking the organization into a trajectory of increasing risk.
Normalization of deviance is the most important safety concept that no organization believes applies to them. Every organization that has normalized deviance believes it has merely adapted to reality. The Challenger engineers genuinely believed the O-ring erosion was acceptable because they had constructed a framework of justification around it — a framework that was rational, detailed, and wrong. The problem is not that people become careless. It is that carelessness becomes carefulness, and danger becomes mere uncertainty, one small step at a time.