Diane-Chadwick Jones on the importance of trust as a leading safety indicator.
CREDIT: Hernan Schmidt/Alamy Stock Photo
The quest continues for reliable leading safety indicators to measure the activities that support safety performance. There is a huge effort in organisations to generate metrics, but little evidence to show that these statistically link to safety performance.
Some leading indicators introduce a heavy burden of bureaucracy in the organisation. We see so many resources expended on activity-related metrics such as leadership time in the field, or the number of general safety conversations. It is soul-destroying to spend valuable time to comply with targets on these metrics, when we know that the large volume of information gathered means that analysis and follow-up actions are impossible. And, even if these metrics are related to the quality of the activities, they are susceptible to the Hawthorne Effect, where the act of measurement artificially improves the metric.
Indeed, companies continue to be frustrated and puzzled by having incidents at sites that they thought were their safest, or repeat incidents where they assumed that remediation actions had been effective.
If only we could know what is really going on so that we can find those difficult-to-see risks in our ever-changing workplaces!
A way to identify a useful leading indicator that correlates to safety performance is to look at standard employee engagement surveys that many companies complete. These surveys usually have questions on trust, care, speak-up and listening, which have a proven link to safety performance. They are also a good reflection of peoples’ views since they are anonymous and routine, and therefore less prone to the Hawthorne Effect. Companies can easily analyse this data at a site level to identify locations that are deteriorating. And the insights are easy to gather, with a few minutes of effort from employees. Additional safety culture surveys are not necessary.
This relationship has been established over the past 20 years in academic literature, with studies performed by several research teams. Originally it was thought to be about leader role-modelling and workers “trying harder”, but when we look closer, we can see that the link reflects how leaders set the tone for the workforce to speak-up about problems, providing a superior level of intelligence and information on the operation to the management – exactly what leading indicators are intended to do.
And it is not only safety that links to trust, but efficiency also. A study of 50 sites in BP showed that the higher the level of trust there was, the lower the level of spills and injuries, and the higher the level of addressing concerns about procedures, the higher the plant productivity. This was published in a peer-reviewed paper in Safety Science: Empirical validation of operating discipline as a leading indicator of safety outputs and plant performance. Other studies on trust and safety performance are cited in this paper.
We could live in a world where we think that the workplace setup is ideal – perfect procedures, perfect equipment, perfect maintenance, perfect training – “work-as-imagined” – and then point to the actions of individuals as the cause of incidents. But that approach is not working out for us, since we discipline and fire those involved, preventing us from obtaining insights about the reality of the situation that influenced people, the “work-as-done”. This is what gives us repeat incidents, and the appearance of “everything is fine” at a site. People are reluctant, maybe even scared, to speak about mistakes or what made sense to them at the time.
When trust is low, we should be concerned that we are not getting an accurate picture of the state of our operation
Building trust is the game-changer. It happens through fostering a culture of care where workers are invited to speak-up about difficulties, are listened to, and are kept up to date on the progress of actions to fix issues. These are the fundamentals of discovering more about risks in the workplace.
When trust is low, we should be concerned that we are not getting an accurate picture of the state of our operation. Not only is this reflected in the willingness to speak to managers and leaders about operational problems, but it defeats many of our other indicators that we rely on.
Maintenance backlogs may be managed to give an enhanced picture to a management team that reacts badly to backlogs. Walkabouts and observations may be crammed into the end of month in order to satisfy targets. People who do not trust leaders to respond helpfully will keep their heads down and struggle on with defective equipment and processes that would otherwise be reported.
The key is to respond in a supportive way when things do not go to plan. You might think that this is too high a mountain to climb, but building knowledge in the organisation about the ever-present gap between ‘work-as-imagined’ and ‘work-as-done’, changes the understanding of why incidents happen. Rather than attribution to the actions of individuals on the day, the focus is on the issues in the workplace setup. This unlocks feedback from the frontline, enabling speak-up about problems, with leaders being able to listen, act and follow-up.
And if you’re thinking: “I’ll never get leaders to respond supportively to bad news”, then change the site visit questions from, “What is the risk?” or “Are you following the procedure?” to “What is making the work difficult?” or “What issues have happened in the past?” or “What advice would you give to someone who is new to the site doing this task?”. These questions enable us to better see the reality of work.
Consistent actions and communications are vital, as well as skills development for leaders so that there is role-modelling of a positive response to bad news and engagement with the workforce at sites. It is possible to set up the expected culture. A good example is bp’s safety leadership principles, available online here.
These principles, based on academic studies and workforce input, cover the main components that enable trust. They are designed to guide the ways of working to drive a robust, consistent safety culture, including a culture of care and looking at the quality of systems and procedures.
Of course, it does not make sense to use solely trust as a leading indicator, but we can avoid a cottage industry of measurement, especially where there are targets for the number of activities. And there can be powerful leading indicators relating to findings and action close-out from walk-throughs of safety critical tasks, such as the Energy Institute’s Task Improvement Process. Independent assessment of the quality of incident investigations has also been found in the energy industry to improve the strength of actions to reduce repeat incidents. Companies are also using leading metrics that already exist and relate to how well the workplace system is enabling people to do their work, such as the speed of updates to procedures, staff turnover and procurement backlogs.
Therefore, looking at trust is not the only answer to understanding future safety performance, but it is a strong and reputable indicator. So strong in fact, that it is a super-indicator – when the level of trust is high it actually makes the other systems and indicators work effectively, as well as being an indicator of risk identification and mitigation.
So, what are you waiting for? Go harvest the existing data from your employee attitude surveys and look at the year-on-year trends to identify issues and at the same time introduce education on ‘work-as-imagined’/’work-as-done’ and building a culture of care.
A bit odd to publish an article based on a paper that is 5 years old where most of the authors worked or still work for BP and written at a time when BP was still reeling from having major disasters or just about avoiding same every 5 years – Grangemouth 2000, Texas City 2005, Deepwater Horizon 2010. Please can we have some comment about current issues, with less risk that the author(s) are presenting a biased narrative.
Poor Leadership, in my opinion, has had serious consequences at BP, not just in safety but in the performance for the company as a whole. You don’t need to be an academic to realise that if people you work with trust you / management, there will be better outcomes not just safety but in productivity and performance
Diane
I love your article and agree wholeheartedly with your approach. My question is, can we have two sets of metrics? One that is visible to the management team to provide assurance of compliance and conformance to standards and other which is visible to all with the aim of measuring organisational culture centred around trust as is suggested in your article.
Thank you
Paul
Excellent article, thanks! There really are only few validated Safety Proactive Output methods and metrics in the world. That is why most companies use leading input metrics. One validated output metric is Safety Index, based on standardized checklists combined with Tally-scoring method. Example of these are TR method for construction industry (https://www.constructionnews.co.uk/sections/news/no-accident-finns-know-the-score-17-02-2005/ or https://pubmed.ncbi.nlm.nih.gov/10440543/) and Elmeric method for manufacturing industry (https://www.sciencedirect.com/science/article/abs/pii/S0925753512002706)