The concept of asafetyculture, as an element of the overall organizational culture, has become a prominent part of the research and practice of safety in the workplace. Indeed, we chose the title of my series for ISHN, “Positive Safety Cultures,” for this very reason. Increasingly, companies are conducting so-called “culture assessments” to get a handle on how their employees perceive the organizational commitment and approach to safety.
What are safety culture assessments?You are probably well familiar with such surveys. If you are not, even a cursory scan of the Internet finds many safety culture assessment tools, which the owners will cheerfully sell you! Most of them are online instruments that ask a variety of questions about leadership, policies and procedures, working conditions, constraints and pressures that work against safety, etc. The resulting data pinpoint areas where the culture is and is not supportive of safe work. The goal is then to keep doing the things we are doing right, and correct the things we are doing wrong.
Why are they important?The concept of organizational culture broadly, or safety culture more specifically, remains somewhat ill-defined. Most definitions emphasize that culture represents a high-level, sum-total of attitudes, beliefs, norms, and behaviors. In these terms, culture specifies “how things work around here.”
There has always been a bit of a “chicken-and-egg” problem with culture and its measurement. Which comes first, the behaviors and attitudes, or the culture?
Does the culture create the attitudes and behaviors that are consistent with it, or is culture simply a descriptive summary statement of the attitudes and behaviors exhibited by workers in the organization?
There is also the question of explicit and implicit culture.Explicitculture reflects what the organization says it is about, and what can be seen in the policies, procedures, mission/vision/values statements of the organization, as well as in daily interactions with others.
Implicitculture reflects more subtle inferences that employees draw from their experience in the organization. Some of the unwritten, implicit norms, values, etc. may be entirely consistent with the explicit ones. Others may not be captured in any explicit statements (“They don’t tell you this, but here’s how you should handle this kind of conflict here...”). Still others may be at variance with the explicit culture (“They say they are for safety first, but the reality is, production is king....”).
Important procedural questionsAs with any survey, it is critical that procedures be put in place to help ensure that the data gathered meet certain criteria â€” that they will be from a large and representative sample, will be unbiased, and will be interpretable. The response rate must be reasonably high, all stakeholder groups you want to hear from must be fairly represented in the sample (e.g., top management, department heads, hourly workers), and you must have the opportunity for some focus-group follow-up to help make sense of responses (“75 percent of respondents felt their supervisors do not fully support safety initiatives. Why do you think that might be?”).
What to do with unexpected results“Follow up!” Suppose “survey says” most employees believe people are sometimes subtly encouraged to cut corners to reach production goals. Suppose the explicit culture says that is not so. What does the organization then do?
This is the Achilles heel of many culture assessments. We get a broad message that something is wrong, but we don’t then clarify and directly address the problem. How many organizational surveys have you seen which show that “communication is lacking”? How many times have you seen that result, year after year, without any obvious and visible follow-up to specify just what the problem is, and then to fix it?
There is little or no value in verifying, cycle after cycle, that “our safety meetings are ineffective” or “our safety signage is inadequate.” Identifying such gaps with no follow-up action sends the message, fairly or unfairly, that “they don’t care.” One of my universal bits of advice to companies going into a culture assessment: “Don’t bother asking if you don’t intend to understand the feedback and then do something with the information.”
When you do follow up and act on information from a survey, make sure it’s communicated widely. I know several cases in which management did take constructive remedial actions, but in such a low-key, off-radar way, the perception of “they don’t care” was not redressed. The contribution was made but no credits were deposited in the bank of employee goodwill.
What does it take to build/rebuild a culture?There are “moment of truth” examples that have a huge impact on the creation or recreation of a culture.
In one case, a high-level manager unintentionally violated a very important plant safety procedure. Confronted (positively) by an hourly lead man who had been encouraged to believe that “there would be one set of safety rules in this company,” the manager, who was three or four levels above the worker, agreed that if he had seen someone do what he just did, that employee would be written up. The manager then encouraged the safety-oriented lead man to “write him up” and send a safety-violation report upstairs, which the employee did, in a professional and responsible way.
Do you think that story said something to the people in that plant (and beyond) about their commitment to safety?
When the gaps between safety values and safety realities have been identified and leadership works consistently to close those gaps, substantial building or rebuilding of a safety culture can be accomplished in a couple of years. It is impossible to give an exact timeline â€” there are just too many variables in play â€” but under optimal circumstances, 18-24 months is not an unrealistic timeline for a substantial shift in the safety culture to take place. I emphasize again the role of consistency. If management sends mixed messages or does anything else to foster mistrust, the budding culture-change effort can wither on the vine.