In a variety of industries, the number of people getting fatally injured has not reduced significantly for a number of years (Dekker & Pitzer, 2016; Loud, 2016; Manuele, 2013). This plateauing of workplace deaths suggests that the strategies to achieve progress in preventing major accidents are providing diminishing returns. And so calls for new approaches to safety management have grown (Dekker & Pitzer, 2016; Hollnagel, 2014; Hollnagel, Woods, & Leveson, 2006).

Safety Differently grew from these calls, as researchers began to realize that traditional approaches were growing increasingly inadequate to deal with the complex realities of today’s work (Dekker, 2015; Hollnagel, 2014; Hollnagel, Woods, & Leveson, 2006). Safety Differently seeks to develop a more proactive, productive, and inclusive approach to safety management.

Safety - the focus on negatives

Traditional safety management emphasizes identifying and eliminating or reducing negatives -- accidents, risks, hazards, “human error”, and “unsafe acts” (Hollnagel, 2014).

Eliminating negatives is a worthy goal. But the sole focus on negatives has consequences. Hollnagel (2014) argues that focusing on negatives not only leads to misunderstanding of what it takes to make an organization successful, but also to misunderstanding the causes of failure.

For example, believing that “human error” or “unsafe acts” are the primary cause of most accidents may stem from this negative focus (Besnard & Hollnagel, 2014). In fact, the processes leading to “errors” and “faulty” decision-making are present because they are adaptive (Hollnagel, 2009; Gigerenzer, 2008). When it comes to people, the conditions preceding success and failure are the same and one cannot fully understand how humans fail without understanding how they succeed (Woods et. al., 2010).

How safety is defined

Currently there is no agreed upon definition of “safety” within the safety profession. But the profession is heavily focused on the reduction of negatives, such as accidents. For example, safety performance is commonly measured by injury rates, and defined by the absence of negatives.

By contrast, Safety Differently defines safety as a capacity to be successful in varying conditions. Safety is not as a goal unto itself, but an enabling objective within the organization. No organization exists with the sole goal to be safe. Safety enables the organization to achieve its fundamental mission.

Still, achieving a successful outcome itself is not sufficient. Organizations can push margins to achieve success in a way that makes failure more likely (Patterson & Wears, 2016). A safety management system should be to help organizations achieve success as conditions change.

Insufficient resources to meet demands is a key threat to achieving sustainable success. Workers are forced to adapt to create success. This creates significant variability in work processes as resource availability, work demands, and the abilities of workers to adapt all change over time (Hollnagel, 2009). This variability mostly goes unnoticed because workers are so good at adapting (Woods & Hollnagel, 2006). But sometimes performance adjustments that workers make fail, and an accident may result.

Normal work is a central area of focus in Safety Differently. By studying what conditions enable or constrain normal work, we can help workers overcome constraints and get work done. When work becomes difficult, you often find adaptation, short cuts, and “errors.” Increasing worker capacity to handle these situations not only decreases probability of failure, but increases the ability of your workers to be successful. This naturally aligns the goals of safety with other organizational goals (Hollnagel, 2014).

People are the solution

Focusing on eliminating negative events results in treating people as a problem or hazard that must be controlled (Besnard & Hollnagel, 2014; Dekker, 2015). The assumption: our system is reasonably safe enough as designed, but the actions of people disrupt the system design, triggering accidents (Dekker, 2014b).

But organizations are never perfect. Insufficient resources to meet the demands is common. Human adaptions fill the gaps to keep things running. Only humans are able to creatively overcome organizational complexity (Woods et. al., 2011). We need human adaptation to succeed.

Safety Differently notes that people are the solution to enable or facilitate. Adapting to changing circumstances is a strength organizations need to leverage on the complex problems they face. Workers at all levels regularly deal with these complex problems. They are in a unique position to identify innovative solutions, or at least identify solutions that won’t work. As Hummerdal (2015) notes, “organizations are filled with people whose capacity goes above and beyond the roles and responsibilities that we have assigned them.”

Safety is an ethical responsibility

The explosion of safety as a bureaucracy is a disturbing trend in safety management (Dekker, 2014a). Safety practice is dominated by regulatory requirements. Ideas and input from workers (with the curious exception of those who have been involved in accidents) is much less common. The result: what regulators believe is “safe” is more important than what workers think. So we spend much of our time ensuring workers comply with regulations (Dekker, 2015).

In Safety Differently, the organization is refocused. Safety is an ethical responsibility to those who do the organization’s risky work. The organization becomes far more interested in the realities of normal work and creating a safety management system that facilitates work rather than constrains it. Safety is a service the organization provides to its employees. Compliance matters, but balance is needed. The concerns of workers should balance the concerns of the regulatory agency.

Doing Safety Differently

Practicing Safety Differently starts by defining “safety.” Many organizations say safety is not defined by the absence of accidents, yet still measure how safe they are by the number of accidents they have. Workforce creativity and innovation may be valued, yet the safety management system focuses on rote compliance with regulations and procedures. Talking with employees about how they define safety in practice can yield interesting data about the values of the organization. 

Rather than wait for an accident to occur, improve normal work. Conduct daily debriefing sessions for workers. Management site-walk observations or asking workers how work really gets done (Gemba walks) are useful (Gesinger, 2016). Do this in the spirit of curiosity and learning, not coaching for compliance. Training on how to interact with workers, including humble inquiry skills, may be needed.

A clearer understanding about what it takes to get work done in your organization is the result. Treat all findings as symptoms, understanding that people adapt their behavior to the conditions they are in (Woods et. al., 2011). Knowing the conditions that influence workers enables you to look for things that make work difficult. Then find ways to make it easier to get work done successfully, even as conditions change. Examples:

  • If work processes depend on certain tools, how can we ensure that those tools are readily available in foreseeable conditions? What conditions cause those tools to be unavailable? What do we need to do to identify and manage those conditions?
  • If workers deal with complicated or unrealistic procedures, how can those procedures be changed to be more workable? Can workers develop their own procedures for a task that might be better? Is a procedure even necessary?
  • If work is sensitive to time pressure, how can we provide slack to allow workers to achieve all necessary goals? An organizational response to these pressures is preferable to relying solely on individual responses.

Understanding and improving normal work processes is central to improving safety performance. Safety Differently sees improving work as a means to improve safety performance. With Safety Differently, rather than constraining work processes, we are facilitators of work.


  1. Besnard, D. & Hollnagel, E. (2014). I want to believe: Some myths about the management of industrial safety. Cognition, Technology, & Work. 16(1), 13-23. doi: 10.1007/s10111- 012-0237-4.
  2. Dekker, S. (2014a). The bureaucratization of safety. Safety Science. 70, 348-357. doi: 10.1016/j.ssci.2014.07.015
  3. Dekker, S. (2014b). The Field Guide to Understanding ‘Human Error’ (3rd Ed.). Boca Raton, FL: CRC Press.
  4. Dekker, S. (2015). Safety Differently: Human Factors for a New Era (2nd Ed.). Boca Raton, FL: CRC Press.
  5. Dekker, S. & Pitzer, C. (2016). Examining the asymptote in safety progress: A literature review. Journal of Occupational Safety and Ergonomics. 22(1), 57-65. doi: 10/1080/10803548.2015.1112103.
  6. Gigerenzer, G. (2008). Gut Feelings: The Intelligence of the Unconscious. London, UK: Penguin Books.
  7. Gesinger, S. (2016). Experiental learning: Using Gemba walks to connect with employees. Professional Safety. 61(2), 33-37.
  8. Hollnagel, E. (2009). The ETTO Principle - Efficiency-Thoroughness Trade-Off: Why things that go right sometimes go wrong. Boca Raton, FL: CRC Press.
  9. Hollnagel, E. (2014). Safety-I and Safety-II: The Past and Future of Safety Management. Boca Raton, FL: CRC Press.
  10. Hollnagel, E., Woods, D. D., & Leveson, N. (2006). Resilience Engineering: Concepts and Precepts. Boca Raton, FL: CRC Press.
  11. Hummerdal, D. (2015, September 22). People are the solution. Retrieved from
  12. Loud, J. (2016). Major risk: Moving from symptoms to systems thinking. Professional Safety. 61(10), 50-56.
  13. Manuele, F. A. (2013). Preventing serious injuries and fatalities: Time for a sociotechnical model for an operational risk management system. Professional Safety. 58(5), 51-60.
  14. Patterson, M. D. & Wears, R. L. (2015). Resilience and precarious success. Reliability Engineering & System Safety. 141, 45-53. doi: 10.1016/j.ress.2015.03.014.
  15. Woods, D. D., Dekker, S., Cook, R. Johannesen, L., & Sarter, N. (2010). Behind Human Error (2nd Ed.). Boca Raton, FL: CRC Press.
  16. Woods, D. D. & Hollnagel, E. (2006). Joint Cognitive Systems: Patterns in Cognitive Systems Engineering. Boca Raton, FL: CRC Press.