Generative HSE
Safety beliefs, standards, leadership & performance


Effective Safety Leadership – The Vexed Questions of Measurement

Professor Erik Hollnagel made the interesting observation that safety is one of the few disciplines that defines itself by the absence of what it is not, in this case no accidents. This immediately gives rise to the vexed question what does a leader use to measure safety effectiveness, when by definition what is being measured is a nil outcome? How do you know if over time the safety performance of your organisation is improving or not? Many organisations rely heavily on incident frequency rates or lag indicators to determine whether or not safety is improving. The obvious problem is that this type of measure by definition captures the data after the event, whether it is a minor injury or fatality or high potential incident that could have led to catastrophic loss. The measure tells us very little about the factors that have contributed to or detracted from sustainable safety performance. 

Is an organisation’s recordable injury rate low due to:

  1. The risks attributed to incidents are being effectively managed?
  2. Good fortune a number of close calls, which have gone unreported, have not resulted in anyone being injured
  3. Results are being artificially manipulated through the efficient return of injured people to work?
  4. You fill in the blank? 

What has been described is not ground breaking news it has been well understood for a long time by leaders. The issue is what do you supplement lag measures with? The focus on positive or lead indicators is a great start, such as number of safety conversations a leader has undertaken or the ratio between planned and unplanned maintenance but the problem remains how do you know that these activities/interventions are connected to the perceived improvements in an organisation’s safety performance? Another of way of asking the question is to consider what are the safety related antecedents and how do they vary depending upon the specific risk context an organisation is working in?
A helpful meta-analysis of what safety and organisational constructs predict safety performance was undertaken by Nahrgang, Morgeson and Hofmann for the 22nd Annual Conference of the Society for Industrial and Organizational Psychology in New York in 2007. Their search at the time identified 300 research articles that dealt with the topic. Through the use of a rigorous selection criterion they reduced the meta-analysis to 59 research articles.
They found some very interesting correlations that could be reliably used to determine what activities are more likely to lead to improvements in safety performance. They concluded, “we found that leadership and commitment have the strongest relationships with accidents and injuries, whereas safety climate and leadership have the strongest relationships with positive safety behaviours.” They also found that, “factors that lead to a safe working environment also produce other organisational benefits.”
A group of Nordic OHS researchers, led by Pete Kines have worked for the past decade to better understand this relationship between safety climate, leadership and safety performance. They have identified seven dimensions that enable organisations to predict safety motivation, perceived safety level and self-rated safety behaviours, thereby providing a proactive means of measuring safety performance.
Taking this research and applying it to measuring effective safety leadership there are a number of positive lead indicators that could be developed to better understand if what a leader is doing is actively contributing or not their organisation’s safety performance.
The research suggests that a good place to start is through measuring the work group’s perception of:

  • The leader’s commitment and priority they place on safety
  • The leader’s competency in managing safety
  • How the leader empowers their work group to work safely
  • The leader’s approach to safety justice
  • The level of trust in the organisation’s safety management system
  • Safety communication and learning

It might be time to make the shift from lag to measuring the real effectiveness of leaders behaviours in some of the above areas in order to achieve sustainable safety performance.
As the old adage goes “what gets measured gets done.”
For those organisations that have made the shift it might be worth considering how do these positive lead results really figure in day-to-day operational decision making? Even when the shift is underway often the lag indicator are the real decision making drives, pause for a moment and think about what consumes conversations and what gets rewarded in your work place?