What are we going to do about fatalities?

26 August 2016

“For every complex problem there is an answer that is clear, simple and wrong.” – H.L. Mencken

In the past few years, a disturbing trend has appeared in workplace injury data across several industries and countries: the number of non-fatal injuries has continued to fall while the number of deaths and serious injuries has remained the same or increased. This alarming trend has led many to question some of the most long-held assumptions about the causes and remedies of workplace injuries.

For decades, vast improvements in our approach to safety have led to significant breakthroughs in safety results. In general, workers are safer now than at any time in recent history. However, what has worked in the past is not working to eliminate fatalities and other catastrophic failures. We are killing too many people while they are at work trying to make a living.

Our lack of progress with serious injuries and fatalities is not due to a lack of commitment. Even in tough times, companies continue to focus on safety, spending millions. However, all this time and attention is not eliminating the most serious types of injuries. We need a breakthrough. To accomplish this, we will need to examine the models and assumptions the drive our work. We need to develop new models, test them for their effectiveness, and quickly improve them.

The Safety Pyramid

One of the most ubiquitous safety models across industries, nations and cultures has been the Safety Pyramid, originally proposed by H.W. Heinrich over 80 years ago. The pyramid model’s early success was due to the elegant display of simple statistical relationships between the types and severity of incidents and injuries—from near misses to fatalities. Although the model has been seriously critiqued by many researchers over the years, it still persists as the primary mental model for explaining all kinds of safety problems, from manufacturing to health care, and from airlines to process industries to construction sites.

To this day, the Pyramid Model suggests basic theoretical assumptions inherited from Heinrich’s original research. First, the model proposes that the probability of serious or fatal accidents (at the top of the pyramid) is correlated with less severe injuries or near misses (at the bottom). If true, then a remedy for serious accidents would be to reduce the number of less severe incidents. Up until recently, safety leaders have interpreted spikes in near misses or lost work days as an indicator that a more serious injury is more likely to occur. However, recent data across several types of industries contradicts this prediction. In fact, fatalities have remained constant, or, in some places, have increased, while minor injuries have continued to decline.

A second assumption very popular in safety centers on the mechanism of accident causation. Again, we see Heinrich having a significant influence in shaping our thinking in this area. In his Domino Theory, Heinrich argued that injuries resulted from accidents; accidents from unsafe acts which in turn occurred from the faults of people which had their origin in the social environment. Injuries could be best prevented by stopping accidents from happening. As the immediate cause of accidents was unsafe acts then eliminating them was the most effective focus of injury prevention programs.

Together, these two assumptions have dominated the thinking of managers and safety professionals across the globe. The combined impact of probabilistic analysis and simple causality has been useful as a broad metaphor and has helped create broad improvements in safety across industries and geographies. Recently, however, the limitations of this approach have begun to outweigh its advantages. We cannot continue our attempts to address serious injuries by addressing all incidents equally. We must discover and then focus on the critical pathways that lead to failure, and to serious injury.

In the second edition of Safety Management: A Human Approach, the late and venerable Dan Petersen proposed that the types of accidents resulting in temporary disabilities are quite different from those resulting in permanent total disabilities or fatalities. “There are different sets of circumstances [and causes] surrounding severity.” Since Peterson’s remarks, there has been more and more focus on the various causal pathways that lead to fatalities.

Looking for New Models for Safety

While our current thinking in safety was adequate for the first half of the 20th century, the increasingly complex and incomprehensible socio-technical environments that developed in the last half and beyond require more intricate and more powerful methodologies. We don’t want to throw away the good work that we have done. However, we must transcend our long standing models recognizing they do not fully match the complexity of our current work environments. Fortunately, quite a lot of research and application work has already occurred, which will replace some of our most basic assumptions with improved models.

Some new thinking coming on line in safety with emphasis in fatal and catastrophic event prevention are built on the following perspectives:

  • People are not a problem to control but a solution to harness
  • Safety does not just occur because people follow rules. It occurs because people adapt
  • Safety is not just the absence of negative outcomes; it is the presence of positive capabilities
  • Safety is not something you have but something you do. People create safety

In the traditional approach to safety the common understanding follows a mechanical or Newtonian worldview – a linear model where cause and effect is visible and wherein the system can be broken down into its parts and rearranged again into a whole. This model is the way that most organizations describe safety. Although it is simple, coherent and intuitive, it ignores or denies human agency, values, creativity and evolution. Likewise, for complex systems, the linear model falls short.

The new models for safety will transcend the limitations of the traditional approach and will be more holistic in nature. In the remaining parts of this paper, we will suggest two areas of focus that are key to enriching our models for how to address the challenges of safety, especially the need to address fatalities and the infrequent, but catastrophic failures. We suggest that both complex systems theory and approaches to leadership point us toward a rich new set of tools and distinctions that will be necessary for making the next leap in safety performance.

Complex Systems

Complexity science is the scientific study of complex systems. This includes both human and nonhuman systems, but here we are mainly interested in human, or sociotechnical systems. For lots of reasons, project work is becoming more complex, and this is the primary reason our conventional models for managing safety are not as useful as they once were.

The behavior of sociotechnical systems cannot be explained or predicted based on analysis of performance of the parts. As our work environment becomes more complex, complexity theory tells us that we must avoid simplistic solutions to safety, such as pointing to “behavior” or “culture” or “rules and procedures” in isolation from each other and as the answer to safety problems. Our new models must be integrated, and more holistic.

In complex sociotechnical systems, safety outcomes are seen as emergent properties of the system, not as the result of simple “root” causes, or component failure. Small changes and variations in conditions can have disproportionately large effects. Cause-effect relations are complex and non-linear; the system is more than just the sum of its parts. Because variability and adaptation are necessary for system health, variability can cascade through the system and can combine in unexpected ways. Sometimes variability is the source of innovative behavior, a really good thing, and sometimes it is the source of catastrophic failure, a very bad thing. When certain kinds of variability interact with other hidden system features, system breakdowns catch us by surprise (EUROCONTROL 2013). This is at the source of fatal and catastrophic events.

The current models in safety, by definition, could not detect the difference between good and bad variability, and thus could not make necessary corrections. Fortunately, there are a host of new concepts, models and practices that are now being introduced to help cut through complexity and provide better safety solutions.

Role of Leadership

Having new models to understand accident causation inside of complex sociotechnical systems is necessary, but is only part of the solution. Because the elimination of fatal and catastrophic events is a critical issue for business, the work should be treated as a total business transformation, not an isolated initiative delegated down the ranks. Bringing in a new model of safety is not a technical fix; it is an adaptive challenge, requiring a reexamination of beliefs, values and mindsets. It is a challenge for which bold leadership is needed.

As with any transformational effort, the work starts with a clear vision, created and shared as an expression of evolutionary leadership. The vision will include some of the best of the past, but will also transcend the past by putting it in a new context. The time is ripe for a new vision of safety that addresses the problem of fatal and catastrophic events, but also goes beyond it to embrace a whole safety solution. A new vision of safety can bring together many of the fragmented pieces in the world of safety, and breathe new life into both business and safety.

To be clear, we are talking about a paradigm change for safety, and we are not naïve about the magnitude of the change. It will not happen overnight. However, in our experience, there are a critical few practices that leaders need to put in place to support the kind of transformation we are pointing to. Once the vision is clear, and people are committed to the journey, we believe that building trust and capability building will create the necessary and sufficient conditions for the emergence of a new safety space.

Next Steps

As someone wise once said in response to the question “how do you eat an elephant?” The answer: “One bite at a time!” Here are a few suggestions based on our experience that could help you take the next step or maybe even the first step toward updating your approach to safety and addressing fatalities and major accidents at your company.

  • Don’t throw the baby out with the bathwater. A lot of what you’re already doing is fine, and you should retain it.
  • Do a self-examination of the current model in use at your company. Are you accepting the default safety models as true without reflection? Are you looking at new models, but don’t know what to do with them? Have you embraced some new thinking, but have not operationalized it? Just knowing where you’re at can be useful in planning out the next step.
  • Consider the extent to which your approach to safety is holistic, makes sense, and supports your business, or if it seems piecemeal, fragmented, and seems to conflict with business goals. If it’s the latter, you most likely have hidden risks that are not being addressed.
  • Take a pulse on your readiness to change. Many companies want the quick fix to safety and this will point them toward the traditional models. However, quick fixes and easy answers leave the big risks unaddressed. Solving the deeper systematic safety issues requires leadership commitment, resolve and some facility with the new tools of safety.
  • The next time an incident occurs at your company, observe the automatic response. This will tell you more about what models are embedded in your culture than from reading corporate safety statements or speeches at annual events. These underlying models become habits and habits aren’t easy to change. However, if we want to address the big risks, both habits and the assumptions that lie beneath them must change.

There is a lot of interest around the world to create the next paradigm in safety. Both complexity theory and new approaches to leadership have contributed to our emerging understanding of what is next. A few companies are beginning to take steps toward replacing outmoded models with new and better ones in service of creating workplaces where both safety and business performance are enhanced.

Ray Master is the Director of Loss Prevention/Safety Consulting with Construction Risk Partners. He is accountable for the design and delivery of the firm’s safety consulting services. The offering aspires to challenge conventional thinking in construction safety in providing both safety management system optimization and safety culture/performance enhancement. His has over 25 years of experience in construction safety span a wide range of industries to include heavy, power, process, high-rise buildings, oil/gas, marine, transportation, hazardous material clean-up and emergency response.

Rick Strycker is the co-founder of WhyNot Partnering, a global company committed to creating workplaces where people thrive, offering support for enabling Why-Based Organizations and Why-Based Safety. Rick has been working with organizations for 30 years to produce extraordinary results. He has developed unique approaches to leadership, safety, and high performance all over the world mainly in high risk industries.

Ray Master

Ray Master is the Director of Loss Prevention and Risk Control Services with 25 years...

What are we going to do about fatalities?