Thoughts on risk and public safety
With colleagues Tim Gill and Bernard Spiegal I have been working on ‘A short guide to implementing risk benefit assessment’ on behalf of the sponsors who are Play Wales, Play Scotland, PlayBoard Northern Ireland, Play England and other members of the Play Safety Forum. This guide has now been completed and is available here.
The following is an article which appeared in the Activity Industries Mutual (AIM) Newsletter in autumn 2013, written by D. J. Ball and L. Ball-King, and which we reproduce here with permission of AIM.1
We are fortunate in being able to devote some of our work time to things in which we enjoy participating, namely, adventure activities. Consequently, over the last few years, we have had the rare luxury of being able to research the conduct of risk assessment and how it is applied and interpreted, not just in shops and factories, the finance sector, and the courts, but also adventure pursuits. This has identified a series of what we believe to be deep issues affecting a surprisingly broad swath of life beyond the conventional factory fence.
First, we would like to draw attention to continuing disconnects over the meaning of that, for us, crucial word – ‘risk.’ We define risk as the likelihood that some particular thing will happen as a result of exposure to some hazard. So, for instance, in the aftermath of some accident, the immediate question which springs to mind is ‘What was the prior risk of that event?’ leading to the follow-up question which is ‘Was that level of risk acceptable in the circumstances?’ Furthermore, we are not thinking here about the subjective risk (‘it looks dangerous’), but the objective risk (‘it might look dangerous but what is the reality?’).
Experiences in court, however, have shown that expert witnesses and enforcement agencies are not consistently applying this approach. Instead, the logic sometimes followed is what Professor Ragnar Löfstedt has called a hazard-based approach. Basically, this seeks to identify hazards and then to eliminate or control them. In contrast, a risk-based approach would identify hazards, assess the risk, and then decide what control is warranted in the circumstances. Which approach is correct? Mark Hoban, Minister for Employment, summarised the position in February 2013 when, in the context of the government’s reforms of health and safety including the Löfstedt review, he said: We need to implement the remaining measures in both reports as well as continue to identify areas where further reforms are needed to create a modern, simplified, risk-based framework for health and safety in Great Britain.”2
We take it as unassailable that the risk-based approach is fundamental to the national risk control philosophy. It is, after all, enshrined in the requirement of the Health and Safety at Work etc Act to do what is reasonable practicable. Second, assessing risk of some hazardous activity is, contrary to advice often given, not easy. As Professor John Adams has said “It’s not rocket science. It’s much harder.” We agree, and recent research supports this contention. Risk assessors, especially when parachuted in from some external setting, find it very difficult to quantify consistently the risk of harm from activities and situations. The primary means of quantifying risk include use of accident statistics (if they exist) and personal experience through immersion in the sector of interest. Thus, if you are an outsider, you will be in a weak position to assess risk. This is one of the reasons why AALA has been held in such high regard. As initially set up, it was staffed by sector experts who had been immersed in the industry throughout their lives. The danger was, when AALA was subsumed by the HSE, that a hazard-based factory culture of risk control would take over which was also comparatively uninformed about actual risk levels and more driven by perception. If your experience derives from shopping malls and office environments, the natural world can appear an horrendously dangerous place.
Third, we have observed a tendency in litigious settings for the prosecution, and their experts, to demand evidence of proper governance and appropriate management systems. The question, seldom asked, is ‘What is proper governance and a proper management system in the circumstances?’ You don’t need to be an expert in management to know that there are different management styles and one size does not best fit all circumstances.
Over the decades, health and safety has developed an increasingly autocratic, command-and-control
structure with numerous written protocols to be followed. While this may work in some situations the downside is that it shifts responsibility away from front-line professionals. We think that what is most effective in managing the safety of adventure participation is the training and experience of leaders who need more autonomy than might be expected of a factory environment.
These thoughts lead us to a number of considerations:
• it appears imperative to us that any new and future regulatory body for adventure activities be populated by persons with contributory expertise and thus drawn from the adventure sector
• that a risk-based philosophy must be pursued
• that it not be pretended that risk assessment is easy (that’s one reason why you need in-house experts)
• that there is a fundamental difference between managing risk in factories and the natural environment, the latter being undertaken in order to realise benefits3 and which implies the need to trade-off risk of injury against gains in things like health, enjoyment etc
• in seeking support or advice via external agencies it is important to ensure that they have the relevant insight into your sector
1. http://www.activitiesindustrymutual.co.uk/sites/default/files/uploads/misc/On%20Target%20Autumn%202013.pdf).
2. https://www.gov.uk/government/uploads/system/uploads/attachment_
data/file/92634/progress-report-health-safety-reforms-feb-2013.pdf
3. See article by James Willis, On Target, Autumn 2012.
Risk assessment is used to hugely-beneficial effect in many industries, ranging from off-shore oil and gas to nuclear and transportation. It is also used to plan for health epidemics, food safety and flooding. Generally the methods deployed are highly sophisticated, science-based and provide useful information for decision makers.
In recent years risk assessment has also been applied to public space and public activities, often using, by necessity, much simpler protocols such as HSE’s ‘Five steps to risk assessment’ (http://www.hse.gov.uk/pubns/indg163.pdf), or devices known as risk matrices of the type shown below which may range from 2×2 to even 10×10 cells.
These matrices, and the way they are used, are a source of worry for some eminent risk practitioners. For one thing they are usually qualitative in that there are no numerical values on either axis and both risk and consequence are subjectively-defined. What, for instance, does it mean to say something has a ‘medium’ likelihood? And in the case of public space is it the likelihood of something happening to one individual member of the public on a single visit, or on multiple visits over a year, or is it the likelihood that any member of the public might come to grief during, say, the next month? There are numerous possibilities but most matrices seem to leave this question unanswered despite having massive implications for their meaning.
Another technical anxiety is that users quite often number the cells along each axis from, say, one to three in the case of the matrix shown, and then multiply them together to generate a score for each cell. In this way the top right cell in the matrix shown would score 9, and the bottom left 1. This is troubling for several reasons. For one thing the axes are qualitative and labelling them from 1 to 3 produces ordinal numbers only (as in first, second, third etc), and not cardinal numbers. It never has made sense to multiply ordinal numbers together.
Another matter is that when numbers are multiplied, however ill-advisedly, they are sometimes said to be used as a means of prioritising and the aim of many risk assessors appears to be to shift hazards falling in the red-coloured boxes to the green ones. The requirement under the law, however, and which is deeply rational, is that control measures should be implemented if they are reasonably practicable. This is a much more sophisticated concept and is not replaceable by fiddling with coloured squares on grids. It could mean that hazards in higher-scoring boxes should not be subject to further control measures because there are none which are reasonably practicable.
An altogether different issue comes up when considering the risk assessment of public space and public activities, and this relates to the reason for their provision which is, obviously, their benefits. Public facilities, such as city squares, tree-lined streets, ponds and fountains, riverside walks, parks, playgrounds, forests, church yards, sites of cultural heritage, national parks, village fêtes, carnivals and sports of whatever kind produce benefits of all kinds ranging from health (physical, emotional and social) to intellectual stimulation, beauty and tranquillity. The problem is that these benefits get no mention in standard risk assessment protocols. It is as if the decision about whether, say, a public activity should be permitted or curtailed could be made purely by thinking about its inherent dangers and without any need to consider the benefits of the activity. This would be an irrational decision process. There is almost always some sort of trade-off between control measures and benefits when dealing with public places and public activities and it is essential that this be recognised, not ignored.
In effect, what has to be done is to weigh the benefits of, say, an unspoilt riverside walk and its risks, and make a decision on whether it should be left as it is or modified. This approach, of weighing risks and benefits, is what is increasingly being referred to as risk-benefit assessment (RBA). If you try to make this decision without thinking about either one of these commodities, the risks or the benefits, I for one could not conceive what was going on in the risk assessor’s head. But I suspect these for me unfathomable thought processes are quite common given that standard risk assessment protocols only refer to one side of this equation, leaving it open to ignore or forget the other.
Collectively these issues suggest that risk assessment as used in public life may require radical surgery.