WHS Institute / Systems That Actually Work

Investigate for cause, not for fault

The purpose of looking into an event is to change the future, and blame does not do that.

An investigation aimed at fault terminates the moment a responsible individual is identified, because that is what it was looking for. An investigation aimed at cause treats that identification as the beginning: this person did this thing, so what made that action seem reasonable at the time, what information did they have, what pressures were operating, what would a competent person in that position have been likely to do, and what else in the system permitted the consequence. The second line of enquiry is longer, less satisfying to close, and vastly more useful, because it produces findings that can be acted on.

The distinction shows up in the recommendations. Fault focused investigations produce retraining, counselling and reminders, all of which target the individual who has now been made hyper aware of the specific event and is the least likely person in the organisation to repeat it. Cause focused investigations produce changes to equipment, layout, sequencing, resourcing and decision rights, which affect everyone including the people who have not had the event yet. If a set of investigation reports consistently produces the first kind of recommendation, the method is generating comfort rather than knowledge.

Timing matters too. The pressure to conclude quickly and demonstrate responsiveness works against the depth required, and early narratives are sticky. Once a story about what happened has been circulated it becomes the frame through which subsequent information is interpreted, and contradicting evidence gets filed as noise. Keeping the question genuinely open for longer than is comfortable, and involving people who do the work rather than only those who supervise it, is the main defence against arriving at a tidy answer that happens to be wrong.