Skip to Content, Skip to Navigation

Safeguard OSH Solutions - Thomson Reuters

Safeguard OSH Solutions - Thomson Reuters

Safeguard Magazine

Seductive or sensible?

RICHARD COLEMAN and KAREN TEO review the two most commonly used safety maturity models and explain which one they prefer to use.

The reason some safety people don’t get invited to parties is because of how much they enjoy debating the relative merits of risk assessment types, zero harm, behavioural safety and safety culture maturity models. In the spirit of fostering one of these debates, we’re going to look at two commonly used maturity models and how they can be used in a pragmatic and positive way.

We would encourage you to think of H&S maturity models as like a jacket; the one you would choose for Sydney in October is not the same as the one you’d choose for Invercargill in August. If you stick to the same jacket, it’s likely that you’ll be uncomfortable in one of those places. They are both “good” jackets but in a specific context one is “better” than the other.

We believe that maturity models are a useful way of thinking about the effectiveness of health and safety management. This is not a universally held view. Dr Rob Long, one of Australia’s most prominent H&S thinkers, describes the use of one safety culture model as “Like finding meaning in the entrails of a goat, the curves … are defended like sacred objects in the religion of safety fundamentalism. My god, they are even sacralised with predictive value.”1

On this last point, we agree with Rob. Safety maturity models are not truths delivered from on high on tablets of stone. They are a tool – useful when applied appropriately and with an understanding of their limitations. We start from a position of open-eyed utility-seeking, broadly supporting not opposing them. We’ve internalised the words of the statistician George Box: “All models are wrong, but some are useful”.

We do not know of any large New Zealand or Australian-based business that has not at some point or is not currently using a maturity model to help them think about what the next component might be in their safety management.

Over the last two years we have delivered cultural change projects in companies with more than 140,000 employees. In each of these businesses, the issue of organisational maturity was central to the change at hand. In the early stages of these projects we often hear “We’re at ‘dependent’ on the Bradley curve” or “We’re at ‘calculative’ on the Hudson ladder”. These statements are made in different ways with different intents, but rarely with a deep and considered understanding of the pros and cons of these maturity models and the challenges inherent in their use.

A recent paper identified over 40 maturity models in use2. However, the comments from my clients are informative because they reflect the two primary ways of assessing H&S maturity: the Dupont Bradley curve model with its strongly behaviourist underpinnings; and the Hudson model with its social science heritage. Both models give you a lens into an organisation, but they are not the same lens and they seek out and therefore see different things. Because they see different things, they lead you to different solutions and different strategies.

Author / Model NameElementsProsCons
HudsonPathologicalGrounded in theory.Complex language – needs interpretation.
 ReactiveOrganisationally rich.
 CalculativeWidely professionally understood. Generative closely aligned to SafetyII thinking.Seen as the “expert” model.
 ProactiveTends to produce highly variable results across the hierarchy of an organisation.
Dupont Bradley curveReactiveSimple.Limited depth – person centred.
 DependentBacked by solid professional support and resources.Aligned to behavioural approaches.


This is a proprietary model developed by DuPont to enable strategy setting and safety programme design. It has four levels, each of which is associated with a hypothesised reduction in injury rates. Dupont describe the levels as follows:

  • • 
    Reactive – People do not take responsibility. They believe that safety is more a matter of luck than good management, and that over time “incidents will happen.” And over time they do.
  • • 
    Dependent – People see safety as a matter of following rules that someone else makes. Incident rates decrease, and management believes that safety could be managed “if only people would follow the rules.”
  • • 
    Independent – Individuals take responsibility for themselves. People believe that safety is personal, and that they can make a difference with their own actions. This reduces incidents further.
  • • 
    Interdependent – Teams of employees feel ownership for safety and take responsibility for themselves and others. People do not accept low standards and risk-taking. They actively converse with others to understand their point of view. They believe that true improvement can only be achieved as a group and that zero injuries is an attainable goal.

There are many conference papers, infographics and presentations about the Bradley curve, but there is very little peer-reviewed, published research into the model or its effectiveness. However, this lack of a solid research base has not affected its uptake with business.

As an approach to improving safety management and outcomes, the Bradley curve is seductive. It is easy to understand, couches the issues in simple people-based behaviours, and suggests a straightforward and linear path to improvement.


However, if you peel back the onion, it becomes clear that the Bradley curve is a child of behaviour-based safety management, which is essentially an extension of applied behaviour analysis or behaviour modification. It traces its intellectual line-age back to Pavlov and his dogs. It argues that humans, like those dogs, respond to stimulus – positive or negative – that if couched correctly will result in the required behaviour change.

We have scientific, philosophical and practical concerns with this thinking, one of which is that in practice this is always aimed at the front one or two layers of an organisation. Chief executives and board, planners, project managers and design engineers are never subject to the same scrutiny, reward or punishment for their behaviour because it’s hard to see their “safety behaviour”. Their behaviour is buried in strategic plans, in capital allocation, in choices that are opaque, but which ultimately lead to a system of work that frontline staff are expected to implement.

The limitation of the approach is that it focuses squarely on the issues that can be modified in the moment by employees. Culturally, it only considers the behaviours of employees, largely ignoring issues of leadership, systems, processes, and values. Strategies that fall out of the Bradley model tend to be focused on the behaviours of frontline staff; even when they are about leaders, they are about leaders influencing the frontline.


In their 2006 paper Dianne Parker, Matthew Lawrie and Patrick Hudson3 proposed a five-tier structure for considering the organisational maturity of an organisation with specific reference to how safety was managed:

  • • 
    Pathological: Who cares about safety as long as we are not caught?
  • • 
    Reactive: Safety is important: we do a lot every time we have an accident.
  • • 
    Calculative: We have systems in place to manage all hazards.
  • • 
    Proactive: We try to anticipate safety problems before they arise.
  • • 
    Generative: HSE is how we do business round here.

The authors identify two aspects of organisations – concrete (management systems) and abstract (attitudes and behaviours) and they provide a structure for assessing where an organisation sits against the tiers for various elements of the systems and of behaviour.

We adhere to a philosophy of safety which comes largely from management and organisational behaviour theory. With respect to safety, the logic of the culture change approach is that the organisation’s basic values or assumptions about safety broadly influence the level of effort and the specific plans and initiatives used by that organisation to manage safety. In turn, these activities serve to shape the perceptions held by employees at all levels regarding the importance of safety and their expectations regarding the importance of leaders thinking strategically about work design, organisational objectives, safe work practices, hazard control, incident reporting, and so on.

One of the key strengths of the Hudson model is that it is organisationally rich and, by extension, somewhat messy. We prefer models that consider the social messiness of work over those that encourage the user to go straight to the behaviour of people at the pointy end of the organisation. We unashamedly champion taking an organisational lens to understand safety, not the narrow individual behavioural lens.


These tools are primarily used as a basis of assessment and as a guide for strategy. The message behind these models is that organisations can be different over time; that how they are and how they can be is both assessable and modifiable. There is an attractive optimism inherent in that thinking.

In our consulting work, we’ve built on the Hudson model to deep dive into systems and processes, leadership, organisational behaviours and values, and we assess them against the Pathological – Generative scale. In our view, this is a more granular model that addresses the key issues that drive safety outcomes. We conduct online surveys, focus groups and interviews to build a picture of how the organisation is. We take the outcomes as indicative of the organisation, not as a base truth.

Religious adherence to the purity of the model (as criticised by Rob Long) is not particularly useful and as a result, we would never argue that our model is “more right” than another model, but it works for our purposes and helps us to develop appropriate interventions.

In using a maturity model, there are a number of issues to take into account:

  • • 
    All models are wrong; therefore, all assessments are wrong.
  • • 
    Regardless of the rigour, or how careful you are to build a picture of an organisation, it is always a sample, a moment in time, a picture built on shaky inconsistent perceptions.
  • • 
    So it’s OK to hold ratings loosely because they are a fabrication.
  • • 
    How you assess is more important than the rating.

If your purpose is improvement, then a wide and deep assessment of what everyone in the organisation feels and understands about health and safety is likely to influence those feelings in a positive way – people like to be asked. (Having experts and managers in a room, isolated from the workforce, says something on its own about the maturity of the culture.)

The two most powerful ways that models can be used is in qualitative data collection – collect the stories that illustrate the culture – and in using the rating as a discussion starter.

What you do with the rating is more important than how you assess.


Some organisations undertake assessments of culture and then continue to behave as before. An assessment that doesn’t lead to change is rarely a good use of time and resources. It is likely to drive disenchantment and to set back efforts to improve. Do an assessment, share the results and make organisational changes as a result.

Things change, sometimes by design but often by chance and circumstance. Organisations can change rapidly, and in most industries the pace of change is increasing. Even if you haven’t consciously set out to intervene in your company’s culture, it is likely that change – in people, in your market, in products and services, in technology, in customers and suppliers – will have changed your business, and your culture will have changed in response.

Finally, regardless of the model that you use to assess your culture, the path to improvement doesn’t have to follow some pre-ordained plan. Each organisation is unique and the path to improvement ought to be unique. Take guidance from a model, but as with Google Maps, there’s always another route. In our view bespoke solutions work best.

Richard Coleman is chief executive and Karen Teo is a business analyst/intern with The Interchange in Melbourne. Richard is a keynote speaker at the 2018 Safeguard conference, 30-31 May.


  1. *Richard Coleman is chief executive and Karen Teo is a business analyst/intern with The Interchange in Melbourne. Richard is a keynote speaker at the 2018 Safeguard conference, 30-31 May.
  2. 1Dr Rob Long Nonsense Curves and Pyramids – SafetyRisk.Net Blog November 10, 2014
  3. 2A Pinto et al Maturity models and safety culture: A critical review Science 105 (2018) 192–211
  4. 3D. Parker et al. A framework for understanding the development of organisational safety culture Safety Science 44 (2006) 551–562
comments powered by Disqus

From Safeguard Magazine

Table of Contents