Human error, human performance and investigations

‘Human error’ is a central concept in ergonomics and human factors. You may have read reports along the lines of ‘80% of accidents are due to human error‘. The language that you use when discussing human error says a great deal about how you view human behaviour and the causes of accidents.

My experience visiting different organisations as a UK HSE Inspector was that the traditional view of human error in the oil, gas and chemical industries went something like this – “the facility would be safe if it wasn’t for all the people”. This view is still relatively common in many industries. In order to assess whether your organisation takes this approach to human error, ask yourself whether any of the following look familiar:

People are seen as hazards, a threat, a liability
People are the weakest link
The less of them the better
People need to be closely managed
We need to protect the system from people
Things sometimes go wrong
Less thinking, more procedures
People are told to “be careful out there”
Investigations tend to assign blame
Investigations focus on the last person to touch the equipment
Human behaviour is seen as binary: humans either fail, or they succeed
‘Human error’ is the conclusion of an investigation.

As a result of this approach, people are blamed, warning signs and near-misses are often not reported (perhaps even covered-up) and the true lessons from incidents are not learnt. Individuals who are involved in accidents are seen as ‘bad apples’ and removing them from the organisation is assumed to be the solution. And yet, accidents continue. . .

The late Prof. Trevor Kletz contributed to a CSB video on the 2005 explosion and fires at the BP Texas City refinery, in which he stated that “For a long time, people were saying that most accidents were due to human error. And this is true in a sense, but it’s not very helpful. It’s a bit like saying that falls are due to gravity” (Anatomy of a Disaster, CSB, 2008).

Is the focus on behaviours of individuals (particularly those at the sharp end or the coal face) not having the impact that you had hoped? Has your accident rate plateaued? It may be that ‘human error’ is a symptom of more fundamental problems in the organisation.

A new way of thinking about human error would be to turn this traditional approach on its head and take the view that “the facility wouldn’t be safe if it wasn’t for all the people”. If your organisation subscribes to this new perspective, the following statements would be key principles:

People create safety and success
Things usually go right, because of people
People adapt to imperfect systems and match the conditions
Human performance is not simply ‘fail/succeed’, but is variable
We need to support people and make their life easier
Identifying ‘human error’ is just the start of an incident investigation
Human error is not a cause, but a symptom of failure
Investigations aim to understand why people behave (and think) the way they do
Investigations ask ‘why things made sense’ to people at the time of an event
Human performance may be influenced by people distant in time and space
‘Human error’ is influenced by features of People, Work and Organisations.

In this new way of thinking about human error, near-misses are more likely to get reported, investigations uncover the true causes of events and appropriate lessons are more likely to be learned.

Sidney Dekker summarised this new approach very well when he said that:

“Human error is not an explanation for failure, but instead demands an explanation” (Dekker, 2002).

In my inspections and investigations as a regulator, I often referred to a UK HSE publication ‘Successful health and safety management‘, (second edition, 1997), known to many by its publication code HSG65. This document emphasized the importance of management and organisational factors in both preventing incidents and in their investigation. One of my favourite principles outlined in this publication is that the majority of accidents and incidents are not caused by ‘careless workers’ but by failures in control, which are the responsibility of management.

You’ll notice the new view of human error in some recent investigations, for example:

“The disaster at Texas City had organizational causes, which extended beyond the ISOM unit, embedded in the BP refinery’s history and culture” (CSB Report into BP Texas City, March 2007, p.139).

However, despite the importance of wider organisational factors being considered in these high-profile investigations, it’s not uncommon to see references to the old view of human error.

‘Human error’ in disguise

Some organisations have shifted away from using the term ‘human error’. Instead, they conclude that incidents and accidents are a result of someone losing situational awareness, that there was a loss of Crew Resource Management, a lack of mindfulness or attention, poor judgement, or perhaps that an operator was complacent or made a poor decision. But these are no better – they are just new labels for human error!

Each of these labels is no more an explanation of an incident than saying that it was simply due to human error.

In order to provide an explanation for human error (or human behaviour more generally), we need to understand the factors that lie behind the label. Whether a finding of ‘human error’ refers to a front-line worker, a supervisor, a manager, designer or engineer, it is of little value unless it is followed with an explanation of ‘why’ someone did what they did.

The principles that describe the new view of human error should be mandatory reading for all supervisors, managers and incident investigators. By changing the language that is used to describe human behaviour and human performance, our ‘investigation mindset’ changes and we start to see behaviours from a different perspective.