Everyone agrees that safety culture is important, but there is no clear agreement on what safety culture actually is. In this section I hope to provide a bit of background on how the idea developed, and the different things that are meant when people talk about safety culture.
This story begins in the 1970’s, before we even had a label for safety culture. At that time researchers were very interested in man-made disasters, and were describing accidents as sociological processes rather than technical sequences of events.
Around the same time the concept of organisational climate began to develop. This would later become fashionable with the label corporate culture. Organisational culture tried to explain differences within companies or within branches of the same company, mainly geared towards explaining the amazing success of certain companies.
The first person to talk directly about safety climate was Dov Zohar, who in 1980 published a paper proposing a questionnaire for measuring safety climate in factories. The factors Zohar was measuring were:
Perceived importance of safety training programs
Perceived management attitudes toward safety
Perceived effects of safe conduct on promotion
Perceived level of risk at the work place
Perceived effects of the required work pace on safety
Perceived status of the safety officer
Perceived effects of safe conduct on social status; and
Perceived status of the safety committee.
Zohar was able to show that these factors were consistent, in the sense that the variation within each factory was smaller the variation between factors. He wasn’t able to show that the factors directly predicted safety, because he didn’t have reliable ways of measuring safety. This is a problem which has plagued safety culture research ever since. Safety culture surveys are definitely measuring something real, but we don’t know exactly what the measurements are telling us.
Zohar’s paper didn’t attract a lot of attention until the first International Atomic Energy Agency report into Chernobyl talked about safety culture. The report drew a clear distinction between merely carrying out duties, and carrying out duties “correctly, alertly, with full knowledge and with due regard for safety consequences”. Bingo. This is a problem that we can all empathise with. There’s a real difference between safety as a rigorous intellectual investigation to understand and deal with hazards, and safety as a check-the-box comply-with-the-regulation exercise. That difference is safety culture.
Let’s deal with some language issues to try to make things clear.
Issue Number 1: Don’t worry about the difference between safety culture and safety climate. Some organisations and documents claim to know the difference, but they can’t agree on what that difference is. For all intents and purposes safety culture and safety climate are inter-changable.
Issue Number 2: What we actually care about is the way people behave. We can’t control or predict behaviour directly, which is why we are interested in beliefs and attitudes. When some people talk about culture, they are talking about the behaviour, “the way things are done around here.” When other people talk about culture, they are talking about the underlying beliefs – norms, heroes, symbols.
The clearest way to talk about it is to recognise both. So when we talk about culture, I’m discussing a set of shared beliefs which have behavioural consequences.
Issue Number 3: It should be obvious that all organisations have culture. It is particularly important for safety, because safety activities are open loop. The only guarantee we have that they are done properly, is that they are done diligently by competent people. It doesn’t really make sense, then, to talk about an absent safety culture. To say that an organisation has no safety culture is really just a shorthand for saying that the beliefs in the organisation tend to lead to behaviours which tend to be bad for safety.
Poor safety culture is easy to recognise in hindsight. Some big accident investigations such as the Kings Cross Underground Fire, Piper Alpha, Continental Express Flight 2574, Potters Bar and Challenger have identified safety culture as a key factor. What we’d really like to know, though, is whether we can spot safety culture problems ahead of time. If we can find problems, we’d also like to know what we can do about it.
When we try to measure or influence culture we need to be mindful that awareness and attitude can have a strong influence over whether our organisational systems work, but they aren’t a substitute for those organisational systems. Equipment design, maintenance, procedures, conditions, communication, training, workload and pressure are all systemic factors that are going to influence safety performance. We care about culture because it is going to influence our ability to improve these factors.
The following list of cultural issues is compiled from a number of sources. These are linked to in the following notes drawing on James Reason, Nick Pidgeon, James Roughton, Andrew Hopkins and Sidney Dekker. These are five dimensions where beliefs have an important impact on behaviours, which in turn have an important impact on safety.
The first dimension is the priority safety is given in the organisation. What we’re looking for is a realistic appreciation that safety is good business. This could be because the organisation knows how much accidents and incidents truly cost, and wants to minimise this cost. It could be that they know that one big accident would be a business disaster, and want to manage that risk. It could be that safety is what distinguishes them from their competitors, and they see it as a competitive edge. The common factor in all of these attitudes is that the are realistic. What we don’t want to see is slogans such as “safety is our number one priority”. Ok, if you’re a safety regulator, maybe that’s true. For anyone else, it’s a lie, and it’s not a lie that your employees are going to believe.
I’ve got two benchmarks for measuring the priority placed on safety: status, and time. Very low status is obvious from organisational charts. If your head of safety works for your head of quality and compliance who works for your head of HR, that’s a really bad sign. If someone at director level has safety as their primary responsibility, that’s much better, so we move on to part two, which is to ask people how much status safety personnel have. It’s often the case that safety has a high position in the structure, but low status in reality. Time is a real giveaway when people claim to care about safety. How often do senior staff send someone to deputise for them at safety functions? Do they cancel attendance at safety training at the last minute, or duck out to take important phone calls? The most precious commodities a senior manager has are time and attention, so how they spend those things is evidence of their priorities.
The second dimension is how self-conscious the organisation is. This ties together two ideas that James Reason discusses – a positive culture is one that is informed, and to be informed you must be reporting. How well do individuals understand the hazards that they are dealing with? How well does the organisation understand the strengths and weaknesses of its own safety management? The key benchmark here is how people react when issues are raised. If the immediate response is to challenge the person raising the issue, or dismiss their concern, this is a bad sign. In a positive safety culture, most issues will be resolved without recourse to a formal reporting system. That doesn’t mean that the reporting system is empty though. People will be using the system to get advice or to report experiences that may be helpful to others.
The third dimension is how highly the organisation thinks of itself. When you ask an organisation if they manage safety well, it’s a bad sign if they give an unqualified “yes”. It’s also a bad sign if the organisation gives an unqualified “no”. The ideal, of course, is an organisation that has confidence it is doing fairly well, but always believes that they can and should be doing better.
For the fourth dimension, we have the organisation’s approach to competence. A positive culture encourages and rewards expertise. You can see this most easily in the way training budgets are managed. An average organisation will pay for training where there is a clearly identified need and benefit. A poor organisation will limit training even where there is a clear need, usually by giving reasons relating to short-term pressure on time or budgets. A good organisation will encourage participation in conferences and standards working groups – they value their staff taking a leading role in the engineering community. One very simple policy I’ve seen recently, is paying staff a bonus for achieving chartered engineer status. It works as more than a motivating tool – it’s also a direct recognition that individual professional development is good for the company.
The fifth dimension in safety culture is belief in the importance of rules and compliance. This is an area where there isn’t an obvious good or bad end of the scale. In many respects, bureaucracy is a necessary transition point from a poor safety culture to a good safety culture. In a poor culture, rules aren’t followed. A culture of compliance is an improvement, but it has a lot of negative connotations. Yes, you want people to follow rules, but it’s much better if they understand the reasons for the rules, and participate in making the rules. A genuinely good culture is flexible. It is resilient to pressure or disruption.
There are more aspects to safety culture that I haven’t talked about here. There’s probably a whole podcast episode about what people in an organisation mean when they talk about safety, in particular the distinction between system safety and occupational health and safety. There’s another whole episode of material about reporting systems, and the concept of a just culture.
To finish though, let’s return to why safety culture matters. Most the safety improvements that we are likely to make in an organisation are structural. This involves changing roles and responsibilities, introducing or improving reporting systems, changing procedures for safety analysis or project review, all that sort of thing. These changes will only be improvements if we understand the culture we’re trying to work with. Recognising the culture, and preparing the culture for the structural change are very important.
On the other hand, treating safety culture as an abstract thing to be measured and improved can be very counter-productive. The five issues I talked about here are beyond individuals. That means that you can’t train your workforce to believe these things. You can’t encourage them through briefings and newsletters. Attitudes and beliefs have a real impact on behaviour – the reverse is also true. Changing the attitudes and beliefs of others probably requires understanding and changing your own behaviour.
