We have consistently considered safety culture (SC) in the nuclear industry to be one component of a complicated socio-technical system. A systems view provides a powerful mental model for analyzing and understanding organizational behavior.
Our design and explicative efforts began with system dynamics as described by authors such as Peter Senge, focusing on characteristics such as feedback loops and time delays that can affect system behavior and lead to unexpected, non-linear changes in system performance. Later, we expanded our discussion to incorporate the ways systems adapt and evolve over time in response to internal and external pressures. Because they evolve, socio-technical organizations are learning organizations but continuous improvement is not guaranteed; in fact, evolution in response to pressure can lead to poorer performance.
The systems view, system dynamics and their application through computer simulation techniques are incorporated in the NuclearSafetySim management training tool.
A critical, defining activity of any organization is decision making. Decision making determines what will (or will not) be done, by whom, and with what priority and resources. Decision making is directed and constrained by factors including laws, regulations, policies, goals, procedures and resource availability. In addition, decision making is imbued with and reflective of the organization's values, mental models and aspirations, i.e., its culture, including safety culture.
Decision making is intimately related to an organization's financial compensation and incentive program. We've commented on these programs in nuclear and non-nuclear organizations and identified the performance goals for which executives received the largest rewards; often, these were not safety goals.
Decision making is part of the behavior exhibited by senior managers. We expect leaders to model desired behavior and are disappointed when they don't. We have provided examples of good and bad decisions and leader behavior.
Safety Culture Assessment
We have cited NRC Commissioner Apostolakis' observation that “we really care about what people do and maybe not why they do it . . .” We sympathize with that view. If organizations are making correct decisions and getting acceptable performance, the “why” is not immediately important. However, in the longer run, trying to identify the why is essential, both to preserve organizational effectiveness and to provide a management (and mental) model that can be transported elsewhere in a fleet or industry.
What is not useful, and possibly even a disservice, is a feckless organizational SC “analysis” that focuses on a laundry list of attributes or limits remedial actions to retraining, closer oversight and selective punishment. Such approaches ignore systemic factors and cannot provide long-term successful solutions.
We have always been skeptical of the value of SC surveys. Over time, we saw that others shared our view. Currently, broad-scope, in-depth interviews and focus groups are recognized as preferred ways to attempt to gauge an organization's SC and we generally support such approaches.
On a related topic, we were skeptical of the NRC's SC initiatives, which culminated in the SC Policy Statement. As we have seen, this “policy” has led to back door de facto regulation of SC.
References and Examples
We've identified a library of references related to SC. We review the work of leading organizational thinkers, social scientists and management writers, attempt to accurately summarize their work and add value by relating it to our views on SC. We've reported on the contributions of Dekker, Dörner, Hollnagel, Kahneman, Perin, Perrow, Reason, Schein, Taleb, Vaughan, Weick and others.
We've also posted on the travails of organizations that dug themselves into holes that brought their SC into question. Some of these were relatively small potatoes, e.g., Vermont Yankee and EdF, but others were actual disasters, e.g., Massey Energy and BP. We've also covered DOE, especially the Hanford Waste Treatment and Immobilization Plant (aka the Vit plant).
We believe the nuclear industry is generally well-managed by well-intentioned personnel but can be affected by the natural organizational ailments of complacency, normalization of deviation, drift, hubris, incompetence and occasional criminality. Our perspective has evolved as we have learned more about organizations in general and SC in particular. Channeling John Maynard Keynes, we adapt our models when we become aware of new facts or better ways of looking at the data. We hope you continue to follow Safetymatters.