Wednesday, March 19, 2014

Safety II – Re-defining Safety (and the way you look at it)

One of the big problems with the safety profession is how difficult it is to define what we’re trying to achieve. Numerous definitions of safety exist (we’ve even talked about a few here and here), but these definitions often seem to describe more about the worldview of the authors (e.g. those that describe safety as the absence of accidents versus those that describe it as an acceptable risk level, or some variation thereof) than give us guidance on what we, as a profession, are trying to achieve, much less how we achieve any goal.

Even still, one common theme exists amongst all such definitions – they are focused on a negative. Safety is almost always defined by the absence of something – absence of accidents, absence of risk, absence of hazards, absence of harm, etc. By defining safety in this way the implication is that our job is to remove, to constrain. Typically getting into the system, finding what’s broken, and either fixing it or replacing it is our way of achieving this. This process applies whether the broken component is a piece of equipment or a person (i.e. human error). The further implication here is that behavior of both people and parts is bi-modal – either it works or it doesn’t. And the things that cause the component (either the person or the part) to work are different than the things that cause the component to not work. Therefore, our job as safety professionals is to prevent the causes of the bad behavior and to constrain the good behavior to keep it from turning into bad behavior through procedures, training, work-rules, etc. This basic approach to safety management has been termed “Safety I” by safety researcher Professor Erik Hollnagel and others.

However, although this line of bimodal thinking applies to parts, it doesn’t work well with people. Consider the following points:
  • The underlying “causes” of either “safe” or “unsafe” human behavior are the same – human capacity for adaptation and innovation.
  • Adaptability and innovation are not inherently bad. They are at least partly (if not entirely) responsible for all of the advances in human civilization. But in some contexts they contribute to accidents.
  • Workers don’t usually come to work to get hurt or killed (although they sometimes do). So, deep down, everyone cares about safety.
  • Workers combine their innate capacity to innovate and adapt to balance often competing goals (production, safety, quality, etc.) in ways that are hard to predict, but are easy to see (assuming you are looking for it).

If this is all true, what if instead of trying to constrain behavior and get rid of all the “unsafe” behaviors and “human errors” we instead accepted that human performance will always have a variability to it and that this isn’t always bad. In fact, if you think about it, most of the time nothing goes wrong. Why is that? If we’re honest with ourselves we’ll admit that it’s because of the ability of the workers to create safety in spite of the imperfect systems that they operate in.

So if we use the Safety I line of thinking, trying to constrain behavior through the use of procedures, training, and bureaucracy, we may actually constrain the very thing that’s creating safety in our organizations.

What if, instead of using only the Safety I line of thinking, we complemented Safety I with a new paradigm that Hollnagel has termed Safety II? Safety II suggests that instead of defining safety as the absence of a negative, instead we should define safety by our ability to achieve success under varying conditions, or, safety should be defined not by the number of accidents, but by the number of successes. Think of it this way, if the roots of success and failure are similar, then only looking at the rare times that accidents happen only tells us a small part of the story.

Safety II is still a relatively new concept, so new that models and mechanisms for implementing Safety II concepts are rare. However, some immediate suggestions from Safety II are apparent:
  • Since workers are one of the primary sources of safety in the organization, make sure you open channels to learn and share experiences. Get out from behind your desk and go observe how work is getting done. Be careful to avoid the Safety I mindset while doing this, of looking for unsafe behaviors. Just observe the innovation and creativity and ask how you can enhance it to increase the likelihood of success.
  • Dedicate some of the investigation resources to investigation how and why things succeed. Sure we need to keep investigating incidents, but because success and failure spring from the same sources, there’s learning to be had from successful jobs. Get out there and find it!
  • While conducting incident investigations, don’t just focus on what went wrong. Instead, identify how things normally work and why that normally succeeds. Then determine why this time was different.


Safety II presents numerous other opportunities for the safety profession (including the potential to eliminate the conflict between safety and production). Let us know your thoughts on the opportunities and challenges of this new paradigm in safety.

No comments:

Post a Comment