The normalization of risk

When nothing happened last time becomes the hazard

The normalization of risk
Nikhil Rattan

“Nothing happened last time.”

 It is a phrase heard in workplaces more often than we might like to admit. A worker takes a shortcut and finishes the job without consequence. A piece of equipment continues to operate despite a known problem. A procedure is skipped because it seems unnecessary. Nothing happens, so the same decision is made again. Eventually, what was once considered an exception begins to feel normal. One of the challenges with managing workplace risk is that unsafe decisions do not always result in harm. In fact, most of the time they do not. This creates an interesting problem. When someone deviates from a procedure and completes the task successfully, the experience can reinforce the belief that the deviation was acceptable. Repeat that experience enough times, and the perception of risk begins to change.

This phenomenon was famously explored by sociologist Diane Vaughan following the 1986 Space Shuttle Challenger disaster. Vaughan described it as the normalization of deviance. Her research examined how departures from expected standards can gradually become accepted within an organization when they repeatedly occur without serious consequences. As Vaughan observed, you succeed by risking a little, you don t fail, so each time you risk a little more. We often think incidents teach organizations lessons. But the absence of an incident teaches lessons too, and sometimes they are the wrong ones. Imagine a worker who bypasses a machine guard to clear a blockage more quickly. The task is completed and nobody is harmed. The next time the machine jams, bypassing the guard becomes easier to justify. Soon others may begin doing the same thing. At no point did anyone formally decide that the practice was safe. It simply became familiar.

When experience changes our perception of risk

Experienced workers are often among an organization’s greatest safety asset. They understand equipment, recognize subtle changes and possess knowledge that cannot easily be captured in a procedure. But experience can also influence how risk is perceived. Performing a hazardous task hundreds of times successfully can create confidence in our ability to control the outcome. The worker may know the hazard is present, but years of successful exposure can send a powerful competing message: ‘I have done this before, and nothing happened.’ This is what makes normalized risk so difficult to recognize. From inside an organization, the behaviour may no longer look unusual. It can simply become the way we do things here. The question for safety leaders should therefore not only be whether workers understand the hazard. We should also ask whether repeated exposure has changed how seriously that hazard is perceived. But organizations also need to pay attention to gradual changes in how work is performed. What shortcuts have quietly become accepted? Which procedures are routinely worked around? What equipment problems have people simply learned to live with? These conversations should begin with curiosity rather than blame. Understanding why people adapt their work can reveal problems with equipment, procedures, staffing, workload or production expectations that might otherwise remain hidden. What appears to be an individual decision may be telling us something important about the system in which that decision was made. Perhaps one of the most dangerous conclusions we can draw in safety is that because something worked yesterday, it must be safe today.

Past success is not always evidence that risk is being effectively controlled. Sometimes it simply means that the conditions required for serious harm have not yet come together. Strong safety cultures recognize this distinction. They question practices that have become comfortable, pay attention to gradual changes in how work gets done and remain curious even when performance appears good. Nothing happened last time should never end the conversation. Sometimes, it is exactly where the conversation needs to begin.