Is the concept of safety culture still a viable tool?

Safety culture, a derivative of organizational culture, was a useful reframe to lower accidents and eliminating fatalities.

Organizational culture was front and center in the curriculum when I was getting my masters in Organizational Development. I remember Dave Jamison, my organizational systems professor saying, “You can’t change an organization by changing the individual. You have to change the whole system, which means changing the culture.”

In that moment light bulbs flashed and I realized that the programs I had been designing targeted controlling the individual whether through incentives, training. or rules and discipline. Teaching leaders how to change a culture seemed to be the answer. So, I went on to teach safety leadership courses, designed employee empowerment programs, and the first culture assessment for safety using an open systems model. All of these tools are still in use and have been successful, but not because we “changed the culture,” I’ve discovered.

After 25 years, the evidence has mounted showing that the success of the safety programs wasn’t due to the culture workshops or assessments. It was due partially to the empowered safety teams that were able to get employees involved in the safety improvement efforts. However, the safety team efforts were not successful after the leader sponsoring the programs left. The verdict is in: successful change efforts are directly linked to the leader sponsoring it. When she or he is gone, so is the movement, unless an equally strong sponsor takes their place.

The concept of culture and systems has failed to provide practical insight to manage the complexities of human interaction and relationships. I say this because so many of the solutions to manage operational risk focus on linear and static controls.

For example, I have heard that Deepwater Horizon disaster analysts concluded that operators were handicapped by poorly defined criteria for making that decision and poorly defined roles and responsibilities between BP and Transocean.  However, an insider said the breakdown in communication was ultimately between a high-ranking manager in charge of the operation and headquarters. The financial stakes were too high for this manger to risk being seen as losing control of the situation.

When viewed from a systems and culture perspective both these findings make sense. Culturally, the leader’s behavior sets the stage for willingness to share bad news. The systems, rewards, governance structures, roles and responsibilities, and supply chain partnerships, all seem to have contributed to the disaster. This led logically to solutions like strengthening oversight of high-risk operations and emphasizing a clear definition of the criteria for making decisions, where the information is obtained, and who is responsible to make the decisions and take action.

Problem solved? I don’t think so. Charting of decision-making roles, information mapping, and contractual definitions (for which there is little patience or willingness to finance) does not lead to this vision of open communication across functions. Anyone who has worked on a team or led one knows trust and open communication is one of the most fragile conditions to maintain in the ordinary work environment. Nevertheless, it can be achieved, but not without a leader who supports it and models it.

All of this leads into a very troublesome area for leadership development. Managing the anxiety of change and uncertainty isn’t something we spend much time talking about during the business day, yet it may be one of the most important tasks of a leader. Its uncontrolled presence can cause people to withhold valuable information, drop out of engagement, or sabotage a project. According to McKinsey, executives attributed 75% of the cause for failure in mergers and acquisitions to culture and leadership issues. Studies keep coming up with the same failure rates over and over.

Can we find a more viable framework to help leader’s understand their role? What if we called it helping people hold the tension of fear and anxiety rather than managing the culture? We can’t eliminate tension and conflict from the workplace, neither would we want to. Humans will either grow and innovate when challenged or they will withdraw. Once again, the leader plays a key role in the choices people make. Not because a leader is responsible for people’s choices, but because they create a relationship that supports taking the risks necessary to growth and innovation.

In 1999, Amy Edmondson’s research on psychological safety began to provide language for one of the most difficult aspects of human nature to accept in a work environment where one is expected to be competitive, self-sufficient, and independent. She began to find evidence that people could not be innovative, creative, communicative, or self-confident unless they felt psychologically safe.

Psychological safety describes an individual’s perceptions about the consequences of an interaction with a group or individual in their work environment. We predict other’s reactions based on our beliefs about how they will respond when we ask a question, seek feedback, report a mistake, or propose a new idea are based on experience (Edmondson 2002). We ask ourselves, “Will this action result in embarrassment or harm to myself or others?” The answer to that question can vary radically from workgroup to workgroup.

Psychological safety is very relevant to accident prevention. We want people to speak up about drift, weak signals, and stopping unsafe work or a colleague. Creating the psychological safety to speak up is only part of enabling teams to notice and speak up about hazards. Learning to recognize, communicate, and take correct action on hazards as a team also requires coordination and some structure to ensure that learning takes place and guide subsequent action. This is an iterative process in which actions are taken, reflected upon, and modified. Mistakes must be tolerated to allow for continuous learning.

What is the leader’s role in creating psychological safety? Since Edmondson describes it is a set of intangible interpersonal beliefs and predictions, it doesn’t present a managerial lever for action anymore than safety culture does. There are actions leaders can take to build psychological safety, but it cannot be mandated or altered directly.

This brings us to the issue of leadership development in this area. Typically they focus on emotional intelligence factors such as empathy and self-awareness, which are necessary to ones ability to notice the absence of psychological safety, and its benefits. This is usually accompanied by a set of leadership behaviors that have been documented to improve employee engagement such as clarifying goals, listening, asking questions, explaining the why behind change, respect contributions, and show you care about employee’s well being.

The 2-3 day leadership development retreat has so often fallen short in this area because the full learning cycle cannot take place in such a short timeu. There is potentially a much better way to go. I discovered it in partnership with a client who told me he wanted every member of the management team to become an extension of the safety team. We laughed that we would make a face mask of the HSE leader and hand them out to make our point. But that wouldn’t turn the managers into HSE leaders any more than sending them to a workshop.

Instead we invited 10 of his function leaders to a series of workshops on how to facilitate safety conversations and learning sessions. They learned how to encourage participation, listen, encourage, give positive and corrective feedback, and ask questions. Then they went into the field and had 1-1 conversations to find out what people were working on and what they needed. So, we combined learning to be a facilitator with listening and responding to employee’s input. Slowly the plant transformed its safety and quality.

What we discovered was that it is true that what turns a person into a leader is the iterative learning that each success and failure creates. Practice and repetition create the concertmaster. Leadership is no less of an art. So when are we going to start giving it the proper time and resources.

Edmonson, Amy (2002). Managing the risk of learning: Psychological safety in work teams, West, M. (Ed) International Handbook of Organizational Teamwork, London: Blackwell.

Rosa Carrillo may be reached for consultation or presentations at

©2017 Rosa Antonia Carrillo


  1. Martin Nelson on May 9, 2019 at 9:22 am

    I’m enjoying and learning so much from your work, Rosa.

    Here’s a question that comes up for me working with public sector employees. You observe that “mistakes must be tolerated to allow for continuous learning.”

    What do you advise for a public sector organization where mistakes often make media headlines and learning is less valued that bad publicity avoidance?

    • Rosa Carrillo on May 9, 2019 at 9:56 am

      Excellent question Martin. In the USA our punitive approach to incidents that are visible to the public is driven primarily by lawsuits. I don’t see that changing anytime soon. We’ve witnessed the bankruptcy of PG&E after the wild fires. I would say that the change begins with internal incidents. How we react to minor incidents, how people feel about working in the organization, and the level of psychological safety affects performance. This in turn will lower public incidents. In addition a leader who refuses to punish people for a mistake (does not intentional harm) only creates fear and raises the likelihood of incidents.

Leave a Comment

This site uses Akismet to reduce spam. Learn how your comment data is processed.