I am not a policy wonk

I coined the term ‘Safety Differently’ in 2012. It was the header of an email I sent to a motley group of company representatives—from Laing O’Rourke, Sunstate Cement, Queensland Rail, Origin Energy and others. I had newly arrived in Australia and had been approached by them to help critically examine the sense of safety ‘getting stuck,’ of a pervasive compliance culture that no longer generated much progress.

In the email to invite them to a new round of discussions, I actually called it ‘Safety, Differently,’ with a comma in between—as in: ‘Safety, but then Differently.’ The term stuck, though not the comma. A book of mine called ‘Safety Differently’ was already in the planning (and the editors didn’t like the comma either). Since then, Safety Differently has become that book; it has become a film, a trademark, even a movement. Daniel Hummerdal, back then my graduate student, asked if he could use the term for a new website he was building. Of course, that was definitely the end of the comma, as website names are not punctuation-friendly. Thanks to his early efforts, you’re looking at that website right now. It has been, and is, a great platform for sharing ideas, inspirations, questions and successes.

One of the questions about Safety Differently or Safety II that many people find most pressing is ‘How do we do this?’ It is a fair and important question. It is a problematic one too, particularly when it comes from people who have only just listened to one talk, but who have never read a single word of what I, or Erik Hollnagel, or others in this space, have written. Having listened to a talk on Safety Differently, people can easily become impatient and keen and eager. They want to make the shortcut from excitement to execution. They want to immediately go from being inspired to seeing it implemented. They want to waste no time on (and don’t actually see the need for) any further investment in understanding the intellectual scholarship and moral thinking from which Safety Differently comes.

This is actually not the question

So people ask about Safety Differently ‘Howdo I do this.’ But what they really want an answer to is the question ‘Whatdo I do now?’ What they really want is someone to tell them, because they haven’t taken the time to think it through, to study the ideas further, to show curiosity and discover the difficulties and adaptive triumphs of frontline work for themselves. They just want other people to tell them what to do. That is literally taking a Safety I mindset to a Safety II world. Of course, the ‘how’-to-get-to-Safety-Differently question is increasingly getting answered in the expanding menu of method options—from embedded discovery to micro-experiments, collective improvements, appreciative inquiry and more. But not the ‘what’ question.

Giving you, or anyone, the ‘what’of the procedural steps, milestones and content for the implementation of anything (including Safety II or Safety Differently) would fundamentally negate what Safety Differently is. There is no intellectual shortcut into a simple procedure for the application of Safety Differently. If there was, it wouldn’t be Safety Differently. It would be Safety I. In Safety I, after all, you have to be willing to hand over your brain, your expertise, your experience, to someone else who has already written the solution for you. You don’t have to think, you just receive and apply. Follow the procedure, stick with the rule, do the checklist that someone else has filled with things they believe are important, so that you can see whether you’re on track according to their definition of that ‘track.’

The work of safety, or the safety or work?

Safety I has had at least 80 years to embed itself in how we organize work. Safety I has enjoyed decades to become a handmaiden of what we could call neoliberal organizational governance, in which workers are an interchangeable commodity—to be measured and held accountable for their loss of productive time (through LTI or TRIFR). Safety I offers safety professionals a ‘safe’ retreat behind mountains of bureaucracy and paperwork, so that they may be really busy with—as another student of mine, David Provan would say—the work of safety, rather than the safety of work.

At the same time, managers and boards enjoy the luxury to focus only on the moments of high drama in the Safety I itinerary: the incidents, violations, non-compliances, the drop or rise of injury frequency rates. These can be expressed in superficial and utterly math-illiterate ‘statistics.’ They fit nicely inside the five-minute presentation about ‘safety’ in a quarterly meeting. Safety I has learned to comfortably co-exist with the short-termism driven by competitive profit: show me low numbers of negatives—in fact, show me ever-reducing numbers of negatives—because otherwise I look bad compared to my peers and I might hurt the interest of the most important stakeholders of all: the shareholders.

Safety I, in a sense, has it really easy. It can ignore the subtle textures of everyday work, and the human reasons why it almost always goes right despite the obstacles and difficulties. It can afford to ignore the ingenuity and innovation that produce adaptations and finesses that actually get stuff done on the frontline. Safety I doesn’t have to really invest in understanding work, or what makes work hard.

Safety I instead uses carefully stage-managed ceremonies (such as the recent IOSH joining of the ‘Vision Zero’ campaign at the World Congress in Singapore). It relies on celebrations of conformity and affirmation (an award for the longest incident- and injury-free run in our fleet!). It deploys quiet rituals (a ‘safety moment’ before a meeting, as if saying a prayer; or a checklist-driven toolbox talk without much of substance for the work of the day). It communicates via banner-ready sloganeering (‘work safe, home safe!’) and charged rhetoric (‘if you choose not to follow this rule, you choose not to work here!’ When these things become entrenched in your organization, they are less a guide to action than a signal that debate has now ceased. Unquestioning submission is what matters, not reasoned assent.

An affair of the gut

This explains why Safety I never bothers to write any serious literature when it changes its program or the concomitant slogans—as it does often and without compunction (say, from ‘Zero Harm,’ to ‘Work Safe, Home Safe, Everyday’). It can do this because Safety I does not rest explicitly on any elaborated philosophical system. It never was given the intellectual underpinnings by any system builder, such as Marx or Freud, or by any major critical intellect, such as Mill, de Tocqueville, or Bonhoeffer. It doesn’t even have the scatterbrained brilliance of a Nietzsche or Kierkegaard. Unlike anything in science, the rightness of Safety I does not depend on the empirical demonstrability or theoretical justification of the propositions advanced in its name. Safety I is “true” insofar as it musters emotional appeal and helps fulfill the aims of the unholy mixture of interests it represents. Safety I is perhaps an affair of the gut more than the brain. It offers what Robert Paxton called ‘mobilizing passions’—not always overtly argued propositions, but rousing and available for leveraging and harnassing nonetheless.

Safety II actually does rest on an elaborated philosophical system—stemming from decades of scholarship in human factors, sociology, psychology, cognitive systems engineering, organization and other sciences; from centuries of enlightenment and the emancipation of human thought; and from millennia of moral philosophy and the development of restorative practices and accountability. Safety Differently replaces control with curiosity; it replaces prescription with participation; instructions with involvement; deference with diversity. Scholarship is appearing all the time: it probably gets written quicker than you can keep up reading it. This creates the increasingly elaborate and solid basis for empowering and enabling critical and innovative thinking from the bottom up.

Escape from freedom

So I am not a policy wonk. I am not in the business of writing procedures for the implementation of Safety Differently. I do not provide checklists that will assess the restorative nature of the just culture your organization may have. I cannot give you a course of treatment interventions to reduce the risk of your organization drifting into failure. For sure, I have written some books (like The Field Guide) that are more bulletized than others (like The Safety Anarchist). This way I hope to provide different ways and levels of access into the scholarship that drives the movement.

But giving you implementation procedures and checklists from the top-down would be the diametrical opposite of Safety Differently. To ask for the checklist or policy guide on what to do to implement Safety Differently, is to regress to a Safety I mindset. When you do this, in the words of Erich Fromm, you ‘escape from freedom.’ Fromm, a German social scientist of the Frankfurt School writing in the early 1940’s, argued influentially that freedom can be so frightening, so effortful, that many people seek the comfort of submission—submission to someone else who tells them what to do, submission to someone who tells them how the world works and what their place is in it, submission to a rule, a procedure, a checklist that takes the cognitive and moral load off their shoulders and supposedly does the hard yards for them. Given the time and place he was writing in, Fromm would know. So when you escape from freedom, you deny your professionalism, your insights, your innovations and adaptations, your understanding of the messy details of work, of the nuances and subtleties of what it means to get stuff done despite the rules, the supervision, the tools, the organization, the pressures, the resource limitations and goal conflicts. You deny yourself.

Safety Differently decentralizes and devolves decision power about how to do things safely to expertise, to the sharp ends of your organization. It asks how things should be done of people who do them every day. Safety Differently enables people, and sees them as the resource to harness. With it comes a commitment to the ‘view from below;’ a call to put justice over power; a humility and curiosity to discover how the world looks from the point of view of other people; and the self-discipline to halt judgment and develop explanations for why they do what they do. The scholarship underneath Safety Differently helps you see (and argue) why all of this matters. And as Nietzsche reminded us, if you know ‘why,’ you can bear almost any ‘how.’

Postscript: In an affirmation of how the implementation of Safety Differently rests fundamentally with those who do it, please have a look at the recent work by Mitchell Services, including their short film which you can view from this link: Mitchell-Services.

You can also read the Woolworths Experiment, to learn how a randomized controlled Safety Differently trial got a huge organization to change the way it does safety.



  1. Sean Walker Reply

    Good article as usual Sidney. In Ireland at the moment we are midway through a ‘Safety Week’ in the Construction industry which has essentially alot of talks on works at height, plant safety and positive mental health (which is a great reminder to all in this industry) and many more. Alot of H&S professionals are giddy this week with joy, ‘making a difference’ and spreading the safety l feel good factor.

    They convey all of the safety l messages, ZERO this ZERO that and work safe home safe rhetoric. Do the wooden tops in our industry think this a game changer? Speaking to workers today, they know its all a corporate safety function with pictures taken with banners and safety l slogans that will furnish their own organisational websites, LinkedIn and trade magazines. (I’ll probably do myself out of a job saying that) and the safety performance bonuses are looking good for Christmas.

    I have been met with very strong resistance in trying to introduce ‘Gemba Walks’ I’ve resisted calling them Safety Gemba Walks. I walk and engage with workers on site in what they are doing and look at improving the works from a safety perspective, this where the safety week should focus it’s resources not on talks and photo shoots. But nobody can see the benefit of engaging the workforce in I think alot of my profession don’t like getting down and dirty at the grass roots level.

    This year I have been introduced to Safety differently and Safety ll. I lecture on a diploma health and safety course and I’m introducing the students to Safety differently and Safety ll. The feedback I’m getting from the students is very positive, I’m not overly dissing safety l but showing them a different way of managing work with regard to Safety. To close out I’m a believer who has seen what safety l is (zero accidents) shh….. you didn’t see the guy in the canteen with the broken arm. My advice to anyone who is struggling with where to start on the safety differently path, begin engaging with workers in areas of your processes that need to be challenged and then simply watch this space as it grows across your facility. Don’t over think the change process.

    Seán A Walker

  2. charlestortise Reply

    Dear Professor Dekker, possibly the problem is more fundamental than getting acceptance of a different approach to safety or a movement from Safety I to Safety II or blaming people more politely in a Just Culture? The foundations of many organisations is a management style for all they do, not just safety, that can trace its lineage to the epitome of industrial organisation of the 1800s railway or railroad companies. Being among the first to combine mechanization with both employees within the company and the public outside who might engage the service offered or just interact with the passing locomotives and therefore be at risk of harm they were required to reduce that risk and demonstrate an ability to control their operation. The problem was it was on a scale and span in both geographical scale and time that meant the usual method of control for businesses at the time, oversight by the owner and his representatives was difficult if not impossible if there was not a development in the means of control or introduction of a new skill set: management. As Erik Hollnagel describes the first railway fatality in the UK was a government minister mown down during a demonstration in the 1830s. In the US in the 1850s a fatal crash inspired Daniel McCullum to devise a method of organization initially copied from the military and then developed by his company and the equally successfully Pennsylvanian Railroad and used to describe businesses to this day. But what few are aware of is the set of rules he devised to use with the hierarchical structure that divided functions into areas of responsibility horizontally and grades of authority vertically and the strictures for being able to find and report on who was at fault so the delinquent could be corrected. Copied by other American businesses the model was further ‘enhanced’ by Frederick Taylor to improve efficiency and by Henry Ford with the combination of mechanic and machine and production lines and economies of scale. Efficiency and productivity became the goals of management in the West but in Japan post WWII quality was the underpinning attribute and they were able to learn how to organize management around this and by doing so engaged everyone, the reason being they listened to Dr Deming first and then worked out the methodologies and tools later, which have been slavishly copied to try and secure the same developments. But first they got rid of the management style that was retained by the West in great part and still underpins what is practised today. If anyone wants a way to introduce safer work they should perhaps consider how to do better work. As Myron Tribus described the workers work within the system, the manager’s role is to improve the system, constantly with their help. Instead of optimising the system of work much of traditional management’s effort is focused on correction of the worker and applying extrinsic motivation to ‘improve’ their performance and control their behaviour. A Deming Approach to Safety would apply a System of Profound Knowledge to learn about the variation within that system and apply knowledge to managing that variation. The methods are available and guidance can be given to understand and appreciate this but the change needs to come from a desire to transform, any work on safety management, even that as enlightened as safety differently is equivalent to tampering. The methods that would deliver a Deming Approach require application and sustained effort but the benefits are proven, please do ask. In every example getting worker buy in is easy, their jobs depend upon it it is those who stand to lose in terms of perceived reputation and levels of control who are invariably the barrier.

  3. Euricério Filho Reply

    Great article as all the other ones I have read since I first got to this website. I trully believe we are reaching a new era for Safety but people must leave behind old paradigms and beliefs then embrace the new. Here in Brazil the companies are struggling to achieve good results in Safety but they are all still doing the same old things. Hope we have people here at the same tune as you so they can plant the seeds we need. I am doing my part. Cheers man!

  4. Verschueren Frank Reply

    I dont see why Safety I (not cynically reduced to the ZERO targetting campaigns) couldnt be combined with Safety Ii
    So looking what is good but also what goes or went wrong

  5. Verschueren Frank Reply

    Why cant we look at the good and the bad and combine the advantages of Saferty Ii with the advantages of Safety I

  6. garyswong Reply

    I sometimes will use cooking as an analogy. When talking to workers, I’ll ask: Do you want to be a Recipe Follower or a Chef? Recipes are written to be easily repeated. Anyone can follow. Practice and expertise increase success. And you get a standardized result. There are lots of recipes in Safety-I.
    What happens though if you don’t have all the recipe ingredients at hand? Or someone above demands you must cut the baking time in half? As a Recipe Follower you would be confused, stymied, even paralyzed. A Chef, however, would accept the challenge and adapt to the unexpected conditions.
    A Chef doesn’t follow a cookbook but knows the art and principles of cooking. Samin Nosrat in her book “Salt Fat, Acid, Heat” explains how salt enhances flavour, fat amplifies flavour and makes appealing textures possible, acid brightens and balances, and heat determines the texture of food.
    Safety Differently isn’t a cookbook but a new view of perceiving people as resources, the capacity to change and the margin of manoeuvre to make adjustments.
    I believe we want workers to be both Recipe Followers and Chefs. The key is understanding when the current situation calls for abandoning the recipe and putting on the chef’s hat.

    1. Tristan Casey Reply

      I am working on a model that does exactly this and linking through to behaviour via self regulation. The core idea is to recognise the work situation and deploy corresponding performance enabling strategies to maintain control over work and achieve peak performance. A balance between safety I and safety II underpins this model. We are working with Curtin University to develop it.

  7. Jonathan Lincolne Reply

    Whatever philosophy (or lack thereof) that underpins Safety-I, I see it as dehumanising and the polar opposite to any human-centred practice. I still believe it is best defined as bureaucracy. The eminent German sociologist Max Weber wrote that bureaucracy is a system of administration distinguished by 1. a clear hierarchy of authority, 2. the rigid division of labour, 3. written and inflexible rules, regulations, and procedures and 4. That once instituted is difficult to dislodge or change. Weber, while able to clearly articulate the benefits of imposed regulations, also saw unfettered bureaucracy as a threat to individual freedom. He wrote that “bureaucratization of human life can trap individuals in an impersonal iron cage of rule-based, rational control”. He said bureaucracy led to the great depersonalization of work and the worker.

    I believe that if Max Weber was to take a look at Safety-I in practice, he would similarly define it as “unfettered bureaucracy” and, I believe, would express his concerns that it had gone from being a source of safety improvement and enhancement to a hinderance. Seeing its impacts for the last 20 years I am of the opinion that Safety-I has become the single biggest hinderance safety innovation and performance we face today.

  8. Rob Long Reply

    As policy embeds some degree social politic and ethic, we need to be clear about what people are asking about this ‘new way’ of doing safety. If we retreat into the same kind of positivism and empiricism, and the same discourse as previous, then there will be nothing different regardless of badging and marketing. If the question is about practice, then there is much that can be practiced to put ‘rubber on the road’ in seeking to humanise the way people tackle risk. This requires a more sophisticated sense of ontology than currently exists in the ‘movement’ particularly as it remains anchored to ‘safety’.

  9. Tim Griffiths Reply

    I read ‘The Field Guide’, having an interest in Human Performance and Safety. I am a Train Driver in the UK.
    I made some notes as I read it, did some analysis of the operation I work in. Guess what? We are ultra safe, but old view/safety I. So much so it smacks you in the face. Drivers make a mistake, they will probably get a disciplinary and certainly retraining and re-assessment. Then someone different has the same type of incident next month. They put up a posters, give out booklets on how not to screw up. But they want more. Well actually less. Less operating incidents because they affect punctuality.
    But the scope is there in our system for new view/safety II. They just need to investigate differently and not just call it human error. Routine assessments need to be about finding out how we as Drivers do normal work not just ticking boxes on an ipad with access to the Safety Management System.
    The time an energy goes into finding out where it goes wrong. The regulator says accuracy rate achieved by train drivers in the U.K. at stopping correctly at red signals is 99.99996%. Everyone studies the 0.00004%.
    It’s interesting is when people at the sharp end start coming out with the new view/safety II ideas. Understanding the goal conflict as train drivers (be safe, be on time, be on the PA talking to the passengers). Sharing privately the near misses in the absence of a proper anonymous reporting system.
    So I think perhaps Safety I is suited to the office bound Safety manager. Give someone a PC and they make a poster and a spreadsheet with some statistics. There isn’t a Dell in the train cab, so we have to learn the risks of the system, the work-around, manage the daily frustrations (always late on the trip before your break or the last trip of the shift) and finish the design(use your schedule card to cover the door release buttons on the side where the platform isn’t). So at the sharp end we have to do new view/Safety II. It’s self preservation.

Leave a Reply to garyswong Cancel reply

Your email address will not be published. Required fields are marked *