Stian Antonsen, from SINTEF and NTNU had an interesting presentation at our HSE Seminar at Hell (which is a place near Trondheim in case someone wonders), May 2016. Here is a brief overview from his presentation where he discussed safety culture from a general, theoretical point of view with many examples from practice and from accidents.

Since the 1980s culture has been seen as one way to ‘excellence’, and looking at it from the other side, culture has been seen as a cause for accidents. This has led to a kind of parable and easy explanation that things went wrong because of the culture.

The Chernobyl accident was the first where culture was pointed out as a contributing causal factor. The culture there was among others characterised by the mission to produce electricity, no matter what. Also it was difficult to question decisions of a superior. Several accidents in the years to follow saw culture mentioned as a causal factor: Challenger, Clapham Junction, and Piper Alpha. The report of the latter even chose to express its view on the culture through normative language.

In Norway the term HMS kultur (HSE culture) has even found its way into legislation for the oil and gas sector. §15 of the Rammeforskriften says that (my approximate translation):

A good culture for health, environment and safety that includes all phases and areas of activity shall be fronted through continuous work to reduce risk and improvement of health, environment and safety. 

This can be problematic, because how can one enforce and comply with something as ‘vague’ as culture? (the regulator has published a guidance document on the subject that can be downloaded for free).

It’s organisational culture one works on, anyway. There is no such thing as safety culture - that is merely a sticker we stick on elements of organisational culture related to safety.

What is culture? One often used simplified definition says that culture is "the way things are done around here" which points towards some important elements of culture. There is a practice (‘the way things’), behaviour (‘are done’) and it happens within a group (‘around here’). It’s interesting that behaviour is both a cause and a consequence of culture. Experiences are important - they affect the practice that evolves.

Culture is something that is found between the lines and a very powerful social mechanism. Culture is a frame of reference, it’s like a pair of glasses that one sees through and that therefore affects how you look at things.

Some seem to think that culture and attitudes are almost synonymous. While ‘attitudes’ are found in the heads of people, one finds that ‘culture’ is something that happens between people. Culture can get us to abandon our own attitudes… (see for example the Milgram experiments). Often one sees that so-called culture improvement actions are directed against attitudes through awareness campaigns and the like. Attitudes are important, but the possibility to affect through campaigns, and the effect of campaigns is very limited. One should rather use effort for other actions.

Culture and Power is a subject that is relatively little discussed. Perrow has not discussed culture explicitly in his work because he found that when you look into accidents you will usually find pressure connected to power in the causal chains (e.g. Challenger accident). The dynamics of power are important for culture(s), in a group and between groups.

Can culture be used as a tool for change? There are "culture optimists" around who actually think that culture is a tool for change. Stian is very sceptical about this. As he said: It’s not possible to turn culture a bit up or down … Quick-fix management literature gives this impression that often borders on brain washing. Where goes the line between organisational development and manipulation? It’s like some organisations move the border from controlling behaviour to wanting to control "hearts and minds".

There are some serious limitations with culture as a tool for change:

1. Organisations will included several different cultural units.
2. Culture is a side-effect of social interaction. Culture changes through interaction and NOT through attitude campaigns, flaming speeches and noble visions. These can give short-lived positive feelings, but there is always a Monday…
3. Culture is not something that can be managed. But one can affect culture’s ”conditions for growth”.
4. You cannot decide what culture looks like beforehand. The big One Size Fits All programs have only small likelihood of creating real change. Rather use resources for many small drops close to people’s everyday work over a longer period instead of a HUGE program. (and neither should you lat consultants run the process, do things yourself) Culture changes are often set in motions Top Down, but in general these things work Bottom Up.

The objective should not be to create a common culture. There will always be different cultures and the differences are actually necessary. An objective should be to facilitate for a common language, communication and understanding between groups. And it’s essential that what is said and that what is done are aligned: Walk The Talk.

Stian has written a very fine book about safety culture: Safety Culture: Theory, Method and Improvement.