One of the primary foundations of aviation safety management systems (SMS) is the organizational-wide, focused approach towards proactive risk management.
Risk management is a complex process that requires certain tools, behaviors, and conditions to be done effectively.
For example, effective aviation risk management requires:
In so many words, what risk management boils down to is analyzing and breaking down hazards and risks in order to apply “controls” to them. “Controls” are simply ways of being prepared for when hazards manifest themselves and adversely affect operations.
The long-term goal of hazard safety risk analysis in aviation SMS is to better understand the relationship between certain actions/choices and results in a given set of circumstances, in order to practice better predictive risk management.
This is where hindsight bias comes into play – and it is no small issue.
Hindsight bias is analyzing a safety issue from the vantage point of the results of individuals’ actions during the issue-event. What this always entails is judging a choice as good or bad based on whether the outcome was favorable or unfavorable.
Ideally, employees’ actions during an issue-event should be analyzed "real-time", which means that actions are looked at about the environment and choices at the time the action was made.
The line between “outcome” and "real-time" is extremely delicate, which is why quality aviation hazard analysis is truly a developed skill.
We can recognize hindsight bias because it generally includes phrases or assumptions like “should have,” “bad choice,” or “naturally lead to…” Basically, phrases that apply judgment to an individual’s actions almost always tend to come from a place of hindsight bias. Such a practice can skew the results of aviation hazard risk analysis.
In the aviation safety industry, risk, hazards, and choices interact in extremely complex ways that often produce totally unexpected results. Rational, “good” decisions can produce negative outcomes. Irrational, “poor” decisions can lead to favorable outcomes.
And what we might generally call “luck” is, in safety risk management, interactions between choices and results that we don’t properly understand.
This is why Hazard Analysis that exists without hindsight bias sticks as close as possible to the relationship between
And when we talk about a “choice,” hindsight bias in a hazard risk analysis will usually focus on the “quality” of the choice/result (i.e. bad, poor, etc.). Effective hazard risk analysis focuses on the facts rather than the “quality” of the choice.
Generally, we only consider hindsight bias as existing in the event of negative safety outcomes. This is an extremely dangerous assumption. Hindsight bias can happen just as easily in safety issues with favorable outcomes. In fact, I would go so far as to say the most dangerous hindsight bias actually occurs in favorable outcomes – and it’s a secret bias that is too common in the aviation industry.
For one, it’s easier to practice hindsight bias in favorable outcomes for several reasons:
Because it is easier to practice hindsight bias, also easier to overlook favorable outcomes – oversights that could potentially cause unfavorable outcomes in similar future events. As humans, we are prone to accepting a past success as a future guarantee – but there are no future guarantees in aviation safety.
Hazard risk analysis is practiced to reduce risk and mitigate the impact of hazards, both of which rely on reducing “likelihood” percentages rather than complete elimination. A guarantee is a type of elimination because it assumes an outcome cannot happen under certain choices. Any assumption of past success equals future guarantees will stunt proactive risk management.
Aviation safety management must keep in mind that though favorable outcomes don’t have the obvious “problem” that unfavorable outcomes do, both favorable and unfavorable outcomes require equal due diligence in the hazard risk analysis process.
Hazard risk analysis in aviation SMS is a long-term proactive risk management strategy. It’s important because it gives references and points of comparison for analysis of current issues.
When a hazard risk register has grown sufficiently large, an aviation safety manager will be able to see more precise relationship trends between choices and outcomes in his/her organization’s environment.
There are amazing benefits of having sophisticated data mining capabilities integrated with proactive hazard identification and safety risk analysis tools. Integrated hazard registers make it so much easier for a safety manager to categorize risks and outcomes based on how similar their relationship is to historical issues.
Without such SMS database tools or a large register to work with, safety managers are stuck relying more on their intuition and personal abilities (rather than historical precedent), which can be a slippery slope to hindsight bias.
While I have harped on making no judgments during the hazard risk analysis process for reported safety issues, we are all human, and certain decisions just seem plain bad or good. It’s impossible to escape that – how shall we say it – human limitation. Certainly, some behaviors should be rewarded or, sometimes, punished for their direct relation to certain outcomes.
However, the point I am making is that the act of rewarding or punishing, feeling one way or another about actions, and forming judgments should remain outside the documented risk analysis process.
Last updated November 2024.