top of page
Search

Local Rationality in Incident Investigation: Why Decisions Made Sense at the Time

5 days ago
8 min read

What seems unreasonable in hindsight may have made perfect sense to the person making the decision.


After an incident, it is often surprisingly easy to identify what someone should have done differently. A warning should have been recognised, a procedure should have been followed, or a different decision should have been made. With the outcome now known, the alternative course of action can seem obvious. Yet this perspective overlooks something fundamental: the people involved did not know what investigators know afterwards.


They made decisions in a particular situation, using the information available to them and influenced by their experience, expectations, and understanding of what was happening. Their choices may have appeared reasonable at the time, even when those same choices contributed to an unwanted outcome.


This is the central idea behind Local Rationality, an important concept within Human Factors and incident investigation. It encourages investigators to understand actions from the perspective of those involved, rather than judging them solely against what became apparent afterwards. The purpose is not to justify every decision, but to understand why people acted as they did. That distinction can fundamentally change the quality of an investigation.


Industrial operator considering the operational situation at a control station, with colleagues and refinery equipment in the background.

What Is Local Rationality?

Local Rationality is the idea that people's actions and decisions generally make sense to them within the circumstances in which they occur. Their behaviour reflects how they understand the situation, what they believe is happening, the information they have available, and the goals they are trying to achieve.


This does not mean that every decision is objectively correct. People can misunderstand situations, overlook information, or make choices that later prove unsuccessful. Local Rationality recognises that these decisions are not made with complete knowledge or from the detached perspective available to an investigator.


Consider an operator who decides to continue running a piece of equipment after receiving an unusual indication. Following an incident, it becomes clear that stopping the equipment would have prevented further damage. The decision to continue may therefore appear careless or difficult to understand. However, the operator may have encountered similar indications before without any actual equipment problem.


Production may have been stable, other readings may have appeared normal, and previous experience may have suggested that the indication was unreliable.

None of these circumstances automatically makes the decision appropriate. They do, however, help explain why continuing operations may have seemed reasonable at the time. The operator was responding to the situation as they understood it, not to how it would later be reconstructed.


This is the essential contribution of Local Rationality. It shifts the investigation from evaluating a decision solely against its outcome to understanding the conditions, knowledge, and expectations that shaped it.


Hindsight Changes How We See Decisions

Hindsight is one of the most powerful influences on how incidents are interpreted. Once the outcome is known, earlier events tend to appear more predictable than they actually were. Warning signs become clearer, missed opportunities become more obvious, and alternative decisions appear easier to recognise.


An investigator reviewing an incident may have access to equipment records, witness accounts, photographs, procedures and technical analysis. Together, these sources create a picture that nobody involved in the original situation necessarily possessed. Information from different locations and moments is brought together into a single reconstruction, making relationships visible that were not apparent at the time.


This creates a subtle risk. Investigators can begin to assume that the people involved should have recognised the same relationships. What appears obvious in the reconstructed sequence is treated as though it should also have been obvious during the unfolding event.


Imagine a maintenance technician who continues a task after noticing an unfamiliar sound. Following an equipment failure, investigators identify that sound as an early indication of a developing mechanical problem. The connection may be convincing when examined alongside subsequent evidence, but the technician did not have access to it. At the time, the sound may have seemed similar to ordinary operating noise or may not have suggested an immediate threat.


Local Rationality encourages investigators to preserve the distinction between the information available at the time of the event and the understanding developed afterwards. Without that distinction, an investigation can unintentionally replace the experience of those involved with the much clearer perspective of hindsight.


The investigator knows how the story ended. The people involved were still living through it.


Understanding the Situation as People Experienced It

Human behaviour is shaped by more than individual knowledge or competence. The environment in which a decision is made influences what people notice, how they interpret information and which actions appear possible or appropriate.


A task that seems straightforward when described in a procedure may be considerably more complicated in practice. Equipment conditions may vary, information may be incomplete, interruptions may occur, and several operational priorities may compete for attention. People must continually interpret these circumstances and adjust their actions accordingly.


These adjustments are not necessarily signs of poor performance. They are often part of how everyday work is successfully completed. Employees routinely make decisions about priorities, timing, available resources, and the practical application of procedures. Many of these decisions are unremarkable because the work is completed without incident. The same adaptations can become the subject of considerable scrutiny when an unwanted outcome occurs.


Consider a team that regularly uses an alternative approach to complete a task because the prescribed sequence creates delays under certain operating conditions. Over time, the adaptation may become familiar and accepted. It may be used by experienced colleagues, discussed openly and regarded as a practical response to the realities of the job.


If an incident later occurs during that activity, the decision to depart from the procedure may appear to be an obvious error. Yet understanding why the adaptation developed may reveal important information about how the work was actually organised and performed.


Local Rationality does not require investigators to accept the adaptation as safe. It requires them to understand how it became a reasonable option to those involved. Only then can the organisation properly examine the relationship between the behaviour, the working conditions and the incident.


Why Behavioural Labels Can Hide Local Rationality

Incident investigations sometimes describe human actions using terms such as carelessness, complacency, poor judgement or lack of awareness. These expressions can appear to provide explanations, but they often say more about how behaviour is judged afterwards than about why it occurred.


Describing an operator as complacent, for example, does not explain what made the situation appear sufficiently safe to continue working. Labelling a decision as poor judgement does not establish what information was considered, what alternatives were perceived, or how previous experience influenced the choice.


Such labels can create an illusion of understanding. They make behaviour seem like a personal characteristic rather than something that emerged from a particular situation.

Local Rationality offers a different perspective. Instead of assuming that an apparently inappropriate action reflects an undesirable attitude, it encourages investigators to recognise that the action may have been consistent with the person's understanding of the circumstances.


This is particularly relevant when examining decisions that appear to contradict established procedures or safety expectations. The existence of a clear written requirement does not demonstrate that the person recognised its relevance in the moment. Nor does it establish whether the requirement was workable under the prevailing conditions or consistently reinforced through everyday practice.


An investigation that stops at a behavioural label may therefore miss important influences on performance. It may identify an individual action without understanding the circumstances that made the action more likely. Recognising Local Rationality helps preserve the distinction between describing behaviour and explaining it.


Local Rationality Does Not Eliminate Accountability

One concern sometimes raised about Local Rationality is that explaining why a decision made sense to someone could be interpreted as excusing the decision. This concern reflects an important distinction between understanding behaviour and evaluating responsibility.


A decision may have appeared reasonable to the person making it, while still being inconsistent with professional obligations, established standards, or the responsibilities associated with their role. Understanding the person's perspective does not remove those considerations.


Equally, the fact that a decision resulted in harm does not automatically establish that it was unreasonable when made. An unfavourable outcome and an unacceptable decision are not necessarily the same thing.


This distinction connects Local Rationality with the principles of Just Culture. Fair accountability requires attention to what people knew, what they could reasonably have anticipated and what alternatives were available to them. It also requires consideration of the working conditions and organisational expectations that influenced their choices.


The investigation and the accountability process may have different purposes. The investigation seeks to understand how the incident became possible, while the organisation may separately need to determine whether conduct met appropriate standards. Both can benefit from a credible understanding of the original circumstances.


Local Rationality therefore does not suggest that responsibility should disappear. It suggests that judgements about responsibility become better informed when behaviour is understood before it is evaluated.


From Understanding Decisions to Organisational Learning

The value of Local Rationality extends beyond explaining individual decisions. When investigators understand why an action appeared reasonable, they may discover conditions that influence behaviour more widely across the organisation.


An operator's response to an unreliable alarm may reflect a long-standing technical problem. A technician's decision to adapt a procedure may reveal a mismatch between prescribed work and operational reality. A supervisor's acceptance of a deviation may reflect expectations that have developed gradually through production pressures and established practices.


These circumstances are not necessarily unique to the individuals involved in an incident. Other employees working under similar conditions may interpret situations and make decisions in much the same way. Understanding Local Rationality can therefore reveal opportunities for organisational improvement that would remain hidden if the analysis focused solely on the individual action.


This also explains its relevance to methodologies such as Tripod Beta. Identifying an Immediate Cause may establish which action or condition contributed directly to a barrier failure. Understanding the circumstances that shaped that action can help reveal meaningful Preconditions and, where appropriate, the organisational influences behind them.


The contribution of Local Rationality is not a separate investigative procedure. It is a way of interpreting human behaviour that helps prevent hindsight from distorting the analysis.


Ultimately, incident investigations should do more than establish that someone made a decision which, knowing the outcome, appears incorrect. They should develop an understanding of how that decision emerged from the situation as it was experienced.

That understanding may challenge initial assumptions, reveal weaknesses in organisational arrangements and lead to more meaningful learning. It may also support a fairer assessment of individual responsibility. The essential point is that human behaviour cannot be fully understood from the outcome alone.


To understand why people acted as they did, we must first understand the world as they saw it.


Continue Exploring Human Factors and Incident Investigation

Local Rationality provides an important perspective for understanding human performance, but it is only one aspect of a broader approach to incident investigation. Developing investigation capability requires the ability to examine evidence, interpret behaviour in context and recognise how organisational conditions influence everyday work.


Comments

bottom of page