The May 2017 issue of NVVK Info (the magazine of the Dutch Society of Safety Science) features an opinion piece about the Certification on the Safety Culture Ladder.

You can read the original Dutch version here, or check the English translation below.

A Ladder on Quick Sand - A Critical View at the Safety Culture Ladder

The safety industry has brought a large number of ineffective and contra-productive initiatives into this world. The problem is that many of those initiatives are generally well intended, sound intuitively well, seem to have effect, and most of all: make us feel good. What these initiatives often lack, however, is a sound theoretical basis and there is insufficient attention for adverse side effects.

The relatively recent certification on levels of the Safety Culture Ladder absolutely fits this category. Thinking critically, the Safety Culture Ladder-circus is based on a couple of Myths. I will discuss five of those below. First, a bit of background for the uninitiated.

What is this Safety Culture Ladder?

The origins of the Safety Culture Ladder (‘Veiligheidsladder’) are in the accreditation system for railway contractors of the Dutch railway infrastructure manager, ProRail. On one hand, this was intended to stimulate safe work by contractors. On the other hand, the certification was also a tool in the awarding of jobs to contractors.

How does this work? The basis is the known safety culture/maturity ladder model. This “ranges the safety awareness and behaviour on five levels”. The more an organisation integrates safety responsibility, reflection and investments in the way they conduct business, the higher their score. The higher the level the organisation is rated on, the higher the likelihood that they are awarded a job they have signed up for. The ‘Safety Awareness Certificate’ is valid for a period of three years max. Every year the holder of a certificate will be audited to check if the organisation still lives up to the criteria attached to the certification.

The Dutch organisation for standardization, NEN, took over the Safety Culture Ladder in June 2016. [7]

Myth 1: Safety Culture is measurable

One major problem of ‘safety culture’ is that many people think that it is a ‘thing’. It is not. Apart from fungi and yeast, you will not find culture in nature. It is simply impossible to collect and measure all culture of an organisation. Objective, scientific measurement of culture is in fact impossible. To worsen the matter of subjectivity: there are numerous definitions of safety culture [1, 5]. In general, these mention things like common or shared values, norms and behavioural patterns within an organisation or part of the organisation.

(Safety) culture is extremely complex. We can describe some of its characteristics, but providing a complete picture is not possible. Todd Conklin [3] compares measuring of complexity with trying to measure a river. We can measure the level of water, how fast it flows, the pollution and temperature of the water, but none of this really describes the river, not even combined. At best, we describe a part of the river at some moment in time.

We see the same with so-called culture assessments and measurements. A couple of questionnaires are sent out, to get an impression how people within the organisation think about certain things. Sometimes additional observations on location are done. Afterwards this information is summarized into a short profile. This is like pointing towards the box at the right hand side of the Wikipedia page when trying to describe the Rhine.

A so-called safety culture assessment is nothing more than a snap shot. Making this snap shot can be compared to anthropologists who travel to some tribe in the rain forest. There they take a couple of Polaroid pictures. Then they travel back home, where they study these pictures safely within the walls of the university and try to determine the tribe’s culture based on these snapshots. Proper culture research is intensive and hard work on location. Researchers spend a lot of time (often years) on location and become more or less part of the community they are researching.

Myth 2: Safety Culture is makeable

Culture emerges from the interaction between people. It is a social construct. From these interactions, certain patterns in thinking and behaviour emerge spontaneously at some point in time, and at some point, these patterns will start to affect the interactions. A simple definition of culture is “The way we do things around here”. Countless factors determine the way people do things and how they interact. Research and experience teaches us that some factors can have certain consequences.

We can make an effort to enhance of weaken these factors in an attempt to get the desired effect and to influence culture through this. We can facilitate growth of certain phenomena within the organisation, but there is no guarantee that these will develop in the way we wish. We cannot design culture on a drawing board and then implement it with a five-point plan. Culture is not a mechanic phenomenon that is makeable and manageable to our whims. We do not live and work in a laboratory with controlled circumstances, after all, but in an inherently messy and complex world with a multitude of influences, goals and uncontrollable or unexpected factors.

Myth 3: There is ONE best Safety Culture

Culture helped to bring us where we are today. When our ancestors lived on the savannah with numerous natural threats, their interactions contributed to the survival of the individual and the community. These interactions and the culture that emerged from this depended on the circumstances. The culture of a desert tribe is very different from the culture of a tribe living in the rain forest. Culture helps handling specific problems in certain circumstances.

Therefore, it is nonsense to try to copy what companies like DuPont do. What is right for them is, may be entirely useless for other companies. Surely, one should see what works for others and if you can learn from that, and possibly implement it in a form that suits your organisation. Copying blindly, however, because they have such a ‘great’ safety culture and that ‘must’ be ‘good’ for us also, is a fallacy and a way that is bound to fail or lead to adverse results in many cases.

The credo One Size Fits No One applies also to safety. Therefore, it is an illusion to describe an ideal safety culture in some standard, or to define levels of safety culture. It has to be tailor-made for each organisation. There may be similarities in the characteristics, but every culture will be unique.

Myth 4: Safety Culture is homogenous

Often you hear that a company has ‘a safety culture’. This suggests that the culture within that company is homogenous. That one will find the same elements everywhere within that organisation. Everyone should be able to confirm from personal experience that this rarely will be the case. Within an organisation, stuff works differently at the accounting department than at the maintenance crew, and sales will probably differ hugely from production.

These differences are necessary to do the work of these different departments in an effective and efficient way. Accounting may be more precise and formal, governed by the strict rules of bookkeeping and their closeness to top management, while the maintenance people have to solve problems quickly when something happens. They will have a greater focus on improvisation. Sales will have their main attention on customers and may promise they sky, even if production can hardly make these promises true.

It is problematic to talk about one organisational culture. Even if there are many similarities between the units of an organisation, there will be always countless nuances and differences. Therefore, it is highly unlikely that the entire organisation will be on the same level; usually an organisation will be on several levels at the same time. Sometimes depending on the part of the organisation, sometimes depending on the subject.

Assuming that it is possible to measure “something”, then the result of a certification will always indicate some kind of ‘average level’ without a real meaning. What if most departments do rather well, but maintenance systematically screws up? It reminds me of this story of a person with a healthy average body temperature of 37 degrees Celsius. Regrettably, he did not survive with his head in the stove and his feet in the fridge...

By the way, you cannot separate an organisational culture from national/regional cultures. They may be much stronger than an organisational culture anyway. I just have to think of my own experiences during my stint in offshore with the major differences between Dutch and British employees, for example with regard to the need for procedures and risk appetite.

Myth 5: Certification is a Good Thing to do

During recent decades, some people have come to regard certification as a solution for most problems. We see them popping up for all kinds of subjects on various levels: personal, companies, systems and products. Certification does have a few positive effects, but also significant negative side effects. One of these is that many will do what is necessary to pass the requirements for certification and not any more than that. One will choose the way of least resistance (‘learn for the exam’), instead of doing what is really necessary to live up to the underlying goals of the scheme. This is a typical example of means that become the end. Certification should never be a goal in itself; it is merely a tool.

We see some reversed logic at work here. The thought is obviously that a good company should be able to get a great score without much effort. The selection of companies, however, is done on basis of certificates and the certificate does not show you whether you get the top of the class, a company that barely passed, or a company that managed to fool the system. I have heard stories of cases where an audit resulted in Level 4, because auditors had been convinced without even observing operators out in the field. That way contractors are selected by excluding others and seemingly you can document that this happened with due diligence.

I fear that the Safety Culture Ladder will be just another victim of the ISO-illness. Rubbish will be produced in a highly controlled manner, while everyone lives in the illusion of delivering quality. No, that is not just a start-up problem of the Safety Culture Ladder scheme; this is an inherent problem with all kinds of certification.

And we haven’t even mentioned the enormous incentives that will undermine any good intentions. After all, a requirement like “if you reach this level on the Safety Culture Ladder, then we will award you with a job worth millions” a major stimulus! What you measure is what you get. If you require your contractors to have Zero LTIs, you can be pretty sure that they will have pleasingly low accident levels. If you require you contractors to reach a certain level on some certification scheme, they damn well will be at least on that level. I presume that if you would require them to prove that they employ at least 1% Martians, a good deal of contractors will deliver that proof. [2]

And so

A friend asked me recently if the Safety Culture Ladder was representation of reality. Of course it is not. It is a model. As George Box said “All models are wrong”. And then he added “...but some are useful”.

The Safety Culture Ladder can definitely have added value when used properly. It can show you in a very simple way where opportunities with regard to safety (culture) can be found. One of the ‘fathers’ of the culture ladder, Mark Fleming, called it at the time the Safety Culture Maturity Model. He did however caution “the model is provided to illustrate the concept and it is not intended to be used as a diagnostic instrument”. [4]

The model has seen a metamorphosis from a descriptive/educational model, to a tool for self-evaluation (Hearts & Minds), to an audit tool, and now even a way to certify and award jobs. This is a clear example of how originally good and useful ideas are hijacked, twisted and abused until they do not contribute anymore and rather facilitate the opposite. A pity.

Will this improve safety? Maybe, a bit. Just like ISO 9000 has improved quality. Researchers do not fully agree about that, by the way and the so-called evidence for ISO’s positive effects are rather anecdotal than thoroughly scientific. [2] And just as with ISO, we have to ask ourselves seriously whether it will be worth the amount of paper work and adverse side effects? I am very sceptical! Regrettably, I fear that just like in the case of ISO many organisations not really have a choice, for that the vested interests have already grown too big.


[1] Antonsen, S. (2009) Safety Culture: Theory, Method and Improvement. Farnham: Ashgate.

[2] Busch, C. (2016) Safety Myth 101, Mysen: Mind The Risk

[3] Conklin, T. (2016) Pre-Accident Investigations: Better Questions. Boca Raton: CRC Press

[4] Fleming, M. (2001) Safety Culture Maturity Model. Norwich: HSE Books

[5] Guldenmund, F.W. (2010) Understanding and Exploring Safety Culture. Oisterwijk: BOXPress

[6] Hendriqson, E., Schuler, B., Winsen, R. van, Dekker, S. (2014) The constitution and effects of safety culture as an object in the discourse of accident prevention: A Foucauldian approach. Safety Science 70 (2014) 465–476