Training as a Control Mechanism

One of the focal points of any solution-set for better safety is always training, and risk-management is no exception, though it has the unique recognition that training is a control mechanism and not something in and of itself any different from a physical control. Yet given the supreme importance attached to training, our direct experience has been that we have never yet dealt with a company that can produce a valid set of training records on demand, either because they cannot be validated, or in some cases because the data simply doesn’t exist. What they can do is produce plenty of spreadsheets, though none of them are related or even passable as proof of anything.

We call this basic problem with proving training metrics evidence a company relies upon the “safety ACE.” That is, expanded, the “safety Ass-Covering Effect.” And companies that rely upon the safety ACE invariably have a challenge when a disaster strikes, forcing them to create evidence to support the low-quality data they have available. At best they can provide anecdotal evidence, and at worst they intentionally manufacture supports for claims of training that cannot otherwise be relied upon to have been true. The argument against such a statement is the existence of certificates, but the reality is that the assumption of validity is the problem. Is the certificate real, proved and validated to the present workplace environment?

Training is one of the most costly elements of any safety program, and the most useless of most, since training tends to be reactive or regulatory rather than productively defined, and training tends to be inconsistent by the very fact it ends up applied in bulk. After a serious accident, how many times has a review of training data shown that while all employees had some hare-brained course or other, not a single one had any training that could have turned a serious accident into something mundane? How many times has an investigation revealed that the nine apprentices were waiting on training that got pushed back by some horrifically inappropriate budget gaffe that resulted from mass-applying training the last accident improperly defined as critical? How many times has a deep investigation shown that the workers affected had never heard of half the safety protocols they were supposed to follow, because no one had ever cracked the miraculous five-hundred page manual for operations?

Training is almost always governed by perceived business needs rather than real business needs. After an embarrassing accident, you can always see a company scramble to “refresh the employees on safety protocols.” These knee-jerk reactions happen because the perceived business need is to rectify the effect of the accident, and almost never in deep analysis does the reaction address the underlying cause. What is the point, after all, of spending a hundred thousand dollars refreshing all employees on the use of harnesses, even to the point of showing them how to put them on to prevent high-fall deaths if a proper investigation observes the dead worker wilfully ignored the rules? Wilful ignorance as a cause for violation of basic safety protocols is no reason to show people who already wear their harness how to put it on. Harsh though it may be, that is money ill-spent.

Understanding cause is the main problem. What is good for a business, what it needs, is to understand the cause of an effect, not react to an effect. It is no more suitable than treating a symptom of an illness rather than the illness itself.

Training has to be viewed as a control mechanism to provide effective suppression of negative outcomes when risks are encountered, or to provide risk avoidance opportunities. That means training has to be focused, has to have measureable effect, and has to address actual underlying causes. Otherwise, you are training at a high cost to scattered effect, and you can never measure it against the outcomes you experience.

What is most peculiar about training is that traditional safety considers it an activity despite all the evidence that activity counts have no impact on bottom line results. Traditional safety would ask, “How many people were trained in harness use and maintenance?” Risk-management would ask, “How many people who required the training got it in a timely fashion, and based upon recent incidents how likely is it a refresher would reduce those incidents?” Traditional safety uses counts and has no metric to impart a value return indicator; risk-management applies training as a control, and consequently measures against actual outcomes.

Traditional safety cannot apply training as a control because it maintains an ignorance of the causes that motivate focused training. The answer to why training is done is almost always, “because it is regulated, or required.” Asking even the one further question (“Why is it required?”) produces a baffling silence in almost every circumstance.

Training is a control mechanism, and in risk-management it produces results by being focused, measurable, and reinforcing of productivity.