Behavior-based safety pitfalls and pointers
Still, many questions remain about behavior-based safety. This article will answer some of the most common questions many people have about behavior-based safety as well as more detailed questions you might have once you’re involved in the practice.
Just what is behavior-based safety?It is performance management that workgroups can carry out for themselves. To manage their own performance, workgroups measure and track the rate at which they perform critical, identified at-risk behaviors. Those are the task-related observable acts that expose the workforce to injury. Using their accumulating performance data, the groups then do problem solving and action planning to reduce their exposure levels.
Which behaviors should a site measure?The inventory of task-related behaviors that are critical to safety performance emerges from careful analysis of the past 2-3 years of a site’s data, including its incident reports.
What is the manager’s role?Managers have several crucial functions in behavior-based safety. They support and foster ongoing wage-roll input, integrate supervisors into the effort, and allocate resources to remedy barriers identified by the behavior-based safety effort.
Why is it necessary to engage and involve wage-roll personnel so centrally in behavior-based safety?Wage-roll personnel are closest to the daily factors of safety at a site and are most at risk for injury. For the same reason they are integral to quality improvement; behavior-based safety depends on them to identify and correct barriers to continuous improvement.
Behavior-Based Safety Provides Managers with leading indicators of the same kind and reliability they require for the rest of their functions.
Why does industry need behavior-based safety?Traditional approaches cannot achieve continuous performance improvement. The most significant exposure to injury occurs during the performance of site-specific at-risk behaviors. The traditional approaches miss this critical window of improvement opportunity by relying on injury rates or by trusting in motivational and consciousness-raising efforts. Industry needs behavior-based safety because it provides managers with leading indicators for safety of the same kind and reliability they require for the rest of their functions.
What is wrong with relying on injury rates?Companies will always need to know their injury rates because whether or not someone is injured is the bottom line in safety management. Injury rates are important descriptive statistics, but they are not of prescriptive value. Prescriptive measures are leading process indicators of performance, and that is precisely what managers need in any field of their endeavor, including safety. (The analogy to quality is that injuries are "defects," and trying to steer safety by injury rates is a form of defect management. The work of Deming, Juran, Crosby, and others has demonstrated the unsuitability of defect management. The same holds true for safety.)
What is wrong with motivational and awareness programs?Nothing, provided the user recognizes that they do not address the root causes of injuries and illnesses. Once again, the comparison with productivity and quality is instructive. No forward-looking company today puts its trust in something as vague as motivation or awareness to assure production and quality targets.
At leading companies, awareness and motivation are grounded and supported by production and quality systems that channel, direct, and harmonize group behaviors toward measurement, problem solving, and action planning to assure continuous improvement. Lacking such systems, no level of sheer awareness is going to produce more and better widgets continuously. The same is true of safety performance.
Safe Behavior SpecificsBy putting the focus on workforce behaviors, doesn’t the behavior-based approach encourage managers and supervisors to blame employees for incidents?
At first, some managers and supervisors may be confused on this point. Managers at all levels must realize that the identified at-risk behaviors are system driven: these are behaviors that the site either condones or rewards outright.
Therefore, although the behavior in question is the behavior of someone, the individual in question is conforming to what the plant management system supports. System flaws must be corrected by management. Fault-finding and scapegoating are counterproductive. This is one of the central messages that behavior-based safety has for management teams.
By putting the emphasis on the employee-driven component of safety initiatives, doesn’t the behavior-based approach give hard-pressed managers a ready excuse to ignore safety?
Once again, some managers may think so at first. Downsizing has left many sites with fewer people to perform the same, or more, tasks. However, as with quality management, the point of behavior-based safety is to use wage-roll input better, not to cancel management responsibility for safety.
If anything, in a certain way, managers are more involved in safety than before, but now their involvement is more upstream and efficient. Instead of being "responsible" for accidents‹something that is beyond their direct control -they are now responsible for an accident prevention process. Management activities can now be driven by upstream process data.
Aren’t workers anxious or worried about peer-to-peer observation and feedback on their safety-related behaviors?
At first most people in all regions and industries express some anxiety about being routinely observed and measured on their safety performance. In short order, however, the majority of workers at a site see that the data from ongoing observations is used strictly as information for their safety and benefit. Most people also learn that most of the time they perform most of their site’s critical behaviors well.
Instead of being "responsible" for accidents, behavior-based safety makes managers responsible for an accident prevention process.
The upshot is that after several weeks of peer-to-peer observation and feedback, most people welcome the input they receive about how they currently put themselves at risk, and precisely which changes they need to make to reduce their exposure to injury. The fact that the feedback is very practical is central to this outcome. The observer never says vague, unhelpful things like: Look alive, be alert, pay attention, etc. Instead the feedback is about specific task-related behaviors such as eyes-on-path, line-of-fire, 3-point-contact.
These days, many electric utility companies have well-established "buddy systems" or "extra eyes" programs. Since those workers are used to being observed for task-related behaviors, they typically welcome the behavior-based approach to their more general safety-related behaviors. Also at chemical companies used to very rigorous sampling for quality, the workforce often sees the immediate connection between quality sampling and behavioral observation.
Does observation have to be peer-to-peer? Why can’t supervisors do the observation and feedback? A note is in order here. At some sites, supervisors are able to perform observation and feedback with no adverse effect on the behavior-based safety process as a whole. It is the experience of the author and his associates, however, that even at sites where there is a high level of cooperation between wage-roll and supervisory personnel, it is most effective to launch a behavior-based initiative with peer-to-peer observation.
The input of wage-roll personnel is crucial and there is no better way to assure their ownership than to train them from the outset to do the data collection, problem solving, and action planning for performance improvement.
At sites where adversarial relations between levels are the norm, it is all the more important to lead with wage-roll buy-in and maximum involvement.
Do the observers have to use an inventory of objectively defined behaviors? Why can’t they simply make notes on what they see?
Observers need to work from an accurate inventory of their site’s critical at-risk behaviors because only then do their peers know that: ·
- The observation process is free of bias and personalities. ·
- The feedback they receive will actually prevent injuries. ·
- The data collected for analysis and problem solving will be a useful guide for action planning.
Does the site really have to enter the observation data in a database for analysis and problem solving? Why can’t they rely on approximate tallies and counts made by hand?
It is most practical to use computer software to track behavioral data because of the sheer volume of data gathered in this approach. Take the case of a 'simple’ site with a shop floor workforce of 100, making a fairly limited range of intermediate extruded aluminum rod for manufacturers. In the course of a month, observers will typically gather over 1,000 pieces of data distributed across 6-7 variables such as shift, location, etc.
In addition there are comments on antecedents and consequences identified as relevant to at-risk behaviors. Data of this range and depth is too complex to analyze effectively without computer assistance.