Reason and Resilience
James Reason's Legacy: Have you made a slip, lapse or mistake in your interpretation?
There are very few professionals that will not have been taught about James Reason’s work so in honour of him, his work and his passing away this week, it is important that his legacy is shared and that the interpretations of his work are correct.
Many have not engaged with his original writing and have been exposed to it second or third hand. These interpretations include the claim that the dominant root cause of incidents is human error and the incorrect use of his work on incident causation (swiss cheese model) to claim that it is violations and compliance failures that lead to barriers and safeguards failing. His work on error types used to classify causes in taxonomies then go no further to assess the underlying latent conditions!
This is not what he intended, and it is not factually correct.
His work fundamentally challenged traditional safety paradigms by shifting focus from individual blame to systemic factors.
Human Error: He introduced error classifications as slips, lapses, mistakes and violations and how they emerge from the normal cognitive processes of humans rather than moral failures in intent, as bad apples or stupid people. He shows that errors are inevitable through skill, rule and knowledge-based performance of people moderated by their sense-making processes and the systemic environment they are in.
Managing the Risks of Organizational Accidents: He developed his accident causation model (named the swiss cheese by colleagues) showing that for an accident to happen, there had to be alignment of
(i) latent organisational failures (which he later called conditions as they were never deemed failures but normal),
(ii) local workplace conditions and
(iii) active human errors.
For these active human errors, he noted that people ‘triggered’ these underlying and inadequate conditions, which resulted in people suffering the consequences rather than causing them.
Echoing Jens Rasmussen his peer, he said “Incidents are as a result of unkind environments, triggered by people but not caused by people”.
Just Culture: He introduced Just Culture principles promoting system error reduction with error management over elimination and providing the distinction between blameless errors versus reckless violations. We see this now with the modern concept of setting up systems to ‘fail safely’, ‘error proof work systems’ and to act with the belief that incidents WILL happen and when they do, the question is not who to blame but how will you ensure people are not hurt.
Above all, further work on his just culture principles has brought this further into ultra safe industries to protect reporting cultures and employee safety engagement with evolved iterations to avoid being ‘just cultured’ through a process rather than driving Reason’s original mindset which was about looking to the system to identify the factors that influenced the error.
Metrics: He argued against "body count" safety metrics, stating that "Low injury rates can create false security - catastrophic risks persist even when minor incidents decline". He showed how production pressures erode safety defenses regardless of good injury statistics.
He also challenged the use of injury metrics by promoting the use of hazard reporting and near misses as work system weaknesses worthy of taking action on.
Further work, he then critiqued traditional incident causation models for overlooking socio-technical system interactions and failing to address migration of work systems towards danger.
He proposed safety resilience rather than compliance-based metrics. We can see the trajectory of this work in contemporary discussions of drift, variability of normal work and resilience engineering.
Ultimately, he provided the basis for key shifts that have yet to see consistently prevalent around how safety is taught or done.
To move from human error as the cause and conclusion of incidents towards error being a symptom of work system flaws
To move from injury rate tracking as the primary measure of safety to the consistent analysis of latent conditions, error traps and strength of barriers
To move from the belief that incidents are caused in a linear manner towards understanding that incidents happen due to complex system interactions combining both proximate and remote factors in time and place to when an incident occurs.
Reason was a legend – and one that provided us a foundation not only to understand safety better but to treat workers better in how we learn and improve our work systems. His analysis moves us upstream from looking at workers actions to organisational decision making and system design.
A key question to ask ourselves is whether we as professionals have the skills, courage and mindset to accept that this has to happen to make work safer.
Maeve O'Loughlin