A manager was stuck in his office completing still more paperwork. He knew his crew was changing a motor in plant #7 -- a high-profile event due to its process safety implications. He decided to go out and check if his guys were doing everything correctly to avoid injury. Sure enough, when he got to the plant three guys were finishing their work and he saw that one of them didn’t have his face shield down as they tested for leaks with the lines activated. The manager caught the worker with an apparent violation. This guy’s day was about to get a lot worse.

The manager had gone on a fishing trip for at-risk behavior.  And he “hooked” one of his crew doing “wrong.” He found what he was looking for – at-risk behavior -- and administered discipline, based on this single data point.

Missing safe behaviors

He didn’t consider all the safe behaviors he overlooked. Indeed the crew had confirmed the LOTO tags before unwiring, made sure they had at least three people to lift the motor, and used proper carts and ladders throughout the process. As they were running the motor to test performance levels one guy took a break to get a drink to fight fatigue.  He lifted his face shield to drink and forgot to drop it back as he returned to the job. The manager only targeted the face shield violation and wrote up one of the team.  Did he show that he valued the safe work the team had done? 

No. He swooped in from nowhere, like a spying drone.  He got angry as he scolded the young man who forgot his face shield, and came off as a grump.  Nothing positive to say.  

Worse, consider the longer term implications if this manager had a habit of ”spying.” Do you think the young man or his fellow laborers would want to interact about safety with this manager in the future? He shut down everyone -- and made them more likely to hide their safety errors. This erodes a safety culture – and the damage is inflicted using but a single data point.

Just like any human, his behavior was shaped over time. He had been molded to favor aversive tactics such as scolding, threats, and discipline over more positive ones such as praise.  And his approach was based on a very limited sampling – a single data point, a single incident.

Going negative

This type of fishing reminds me of single data-point management. A manager or supervisor (perhaps even you) is responsible for performance on measures consisting of outcome data (such as injuries), instrumentation data (from machines or via other forms of technology) and the like. You spend your work day looking for the kinds of errors that can hurt these numbers. You don’t want to wait until an incident happens, because it would suck for you and your numbers.  Instead, you engage in walk-arounds to find errors so you can react to them before they conspire to create a negative outcome (such as an injury). These walk-arounds are like fishing trips. You are searching for violations of rules. You’ve got your line out and are trolling for rule-breaking behavior. 

Safety management often resembles blind fishing. We don’t pay attention to thousands of safe behaviors. Instead, we just hunt for violations. We almost feel if we don’t catch something, it isn’t a successful walk-around. So we continue looking for that singular instance in which a worker’s behavior gives us what we think is a clear danger signal. Then we yank on the line to attempt to hook the violation.

Many of us through experience evolve a management behavior that becomes more negative.  Violations first result in scolding, scolding turns into reprimanding, reprimanding to threats of discipline, threats to administering discipline. 

Then we go fishing, looking for opportunities to use unpleasant consequences -- under the illusion that they work. We get good at fishing, throwing out that line, trying to catch bad behavior. We know what to do when we catch one -- we threaten discipline, we do discipline. When we’re out there fishing for risks we fail to see the overwhelming majority of safe behaviors taking place.  We fail to see the variation that exists, the very variation that can help us solve problems and create a safer workplace.  Even worse, we’re not stopping injuries.

Best practices

Don’t be fooled by this illusion. Instead look for variation, discover its source, and manage it.  Better fisherfolk than I use sonar in their boats. Sonar bounces sound waves down toward the lake bottom.  Schools of fish reflect this wave back to the sensors and appear on the sonar screen, giving you a good idea of where fishing may be productive. Modern sonar can locate size and composition for individual fish by bouncing sound waves up to 40 times a second.  Fisherfolk have even figured out that the speed of the sonar pulse (sound) is affected not only by depth but by water salinity (which is a constant on a lake) and temperature.  Add a temperature gauge and you can now tell the distance of the fish from the boat.  Those of you in oil exploration know that sonar has become so advanced that it can detect not only the geography of the ocean floor but also gaseous seeping out of promising wells.  We will do all this sophisticated measuring for pleasure fishing yet we leave our safety measurement in the hands of single data points, looking to hook at-risk behaviors.

This is an excerpt from Dr. Ludwig’s upcoming book, Dysfunctional Practices that Kill Your Safety Culture (and what to do about them).
Visit Safety-Doc.com for more information.