James Loud was vociferous with his words in his recent Professional Safety magazine article that looked into possible reasons that the U.S. and Canada are experiencing higher risk of serious incidents at worksites, despite all the progress we have all made in reducing the number of incidents overall.

I have been posting an in-depth look at Loud’s article, which delved into the current trends and realities about the quality of the incidents and their relation to the quantity of incidents, then used three well-known case studies of how meaningless the quantity of incidents truly are to a safety culture; then he looked into the work of three safety pioneers and how their theories have developed great strides in overall safety but how sticking to those theories in a dynamically changing workforce has not been able to address the reality of incidents that result in serious injury or death.

Missing the Mark

The work of the three pioneers – Frederick Taylor, B.F. Skinner and Herbert Heinrich – have shaped safety of the last several decades, but reliance on their theories have led to safety protocols not adequately meeting the needs of the current workforce, which is not populated with workers who do the same routine work and same routine actions over and over day after day like these pioneers were experiencing in their day.

Most notably has been the work of Heinrich, who claimed that 88 percent of incidents were a result of unsafe actions, 10 percent by unsafe work conditions and 2 percent were true “accidents” that were unavoidable. This translated into the Incident Pyramid, which claimed that of any 330 incidents, 300 would not result in injury, 29 would result in minor injury and one would  result in significant injury or death.

Our interpretation of his work is to ignore the percentages – if we lessen the number of incidents overall, we will lower the serious injury incident numbers by a similar rate because of the theory that the percentages would remain constant. What we have found though, through lots of data over the decades, is that isn’t true. While we have greatly reduced the numbers of incidents in general, the rate in which serious injuries occur has not dropped by a similar amount, meaning that any incident that happens on a worksite now has a greater probability of being a serious incident than 40 and 50 years ago.

And why have we not been able to reduce the rate of serious incidents by an equal amount to the overall incident rate? Loud theorizes there are four misconceptions we are working from in our safety protocols, based mainly on the work of Taylor, Skinner and Heinrich.

#1: Unsafe acts lead to incidents; minor injuries predict serious ones.

This is all about Heinrich’s pyramid. Even today, many people who are safety officers or prominent managers in companies like to blindly quote Heinrich’s number than 88 percent of all incidents (some even go so high as 95 percent) are due to unsafe acts by workers.

Having that bias already embedded makes incident investigation tough. A company will generally end an investigation as soon as it notes an unsafe act by a worker and will close the case right there, even if the company may have some responsibility. An unsafe act may be a cause of an incident, but often it may not be the “root” cause, and the company for many legal and public-relations reasons may not want to know or uncover the real truth – that if a “root” cause isn’t addressed, then the incidents will continue to happen and it will be impossible to get to zero incidents.

The other point is that minor injuries tend to foreshadow major ones later. However, there has been no causal evidence in recent research to suggest that is the case. This is not a cumulative phenomenon, where factors that play into minor incidents necessarily accumulate and create a far more serious problem later if not addressed. Each incident seems to more or less occur in their own vacuum, even if two incidents seem to come from the same risk factors and one is more serious – they should be addressed as completely unrelated incidents, but often they are not.

#2: Safety means workers under control.

Yes and no. This is also based largely on the theories of Taylor and Heinrich, who both claimed that the best way to keep workers safe is to give them specifics on how to do their jobs, step by step, and demand they do those steps that same way every single time. And in a world where it’s believed that nearly nine in 10 incidents come from workers acting unsafely, it’s no wonder this belief has had so much hold on safety.

But as was mentioned before, with such a bias already ingrained in those who tend to investigate incidents, management or company accountability for incidents is much lower than it probably should be. Often, companies tend to overlook some of their own errors in engineering, operations, design, and other factors that contribute to accidents, even those in which a worker acts unsafely. After all, if engineering or operations isn’t addressed, an incident could occur even when the worker was acting safely – and then what? The worker-centered protocol gets changed, instead of the company looking honestly in the mirror and realizing  that it needs to change?

#3: If you’re compliant, you’re safe.

There are two types of compliance – compliant work conditions according to OSHA and ANSI standards, and compliant workers who do all the things they are supposed to do in a safe way according to company protocols and government standards and industry best practices.

Because many companies re so focused on worker-centered safety (relying on Taylor and Heinrich), the goal for many officers in safety has been to enforce compliance with safety protocols among workers and rewarding or punishing safe or unsafe behaviors. This has been the standard over the last 30 years, after about a decade of companies working with the new OSHA agency to bring their worksites up to safeyt compliance with the working conditions.

However, even with compliant conditions, these safety engineers weren’t seeing dramatic improvements in their incident records, so they went in the other direction and started working on the workers themselves. This has helped drive down incident numbers, but not addressing compliance in both actions and conditions leads to still plenty of risk of trouble. In other words, be careful how you define “compliance.”

#4: Safety is a low injury or incident rate.

If the number of incidents were a guide for safety, then Deepwater Horizon would never have happened. It’s a fallacy that a safe facility is necessarily one that hasn’t had a number of incidents in a quarter or year or even a week. If that is the true goal, to get to zero or a very low number of incidents, there are ways to do it and “work around” the system to win safety awards:

  • Do nothing. Have workers twiddle their thumbs on the job. There is risk with virtually every action on a worksite; that is the reality.
  • “Tweak” the data. If a person only needs a bandage, maybe the company policy is not to report that incident. If the person is a consultant or contractor, maybe the company doesn’t count that incident even if it happens on the company’s site. And so forth.
  • Non-reporting. Of course, the one sure way to get zero incidents is to not report anything at all. And there are instances where companies who focus so much on the goose egg that they would go so far as to scare or intimidate workers into not reporting incidents.

Of course, incidents that are not reported, no matter how minor, are missed opportunities to learn and develop new ideas in order to gain more conditional and behavioral compliance to gain more safety for all workers.

Now that Loud exposed these misconceptions and at least gets the reader to think critically about safety,  my next post will get to what is beyond these issues and what can we, as safety professionals, do moving forward to decrease the numbers of serious incidents and deaths on worksites.