Influencing safety behaviours in the 'new world'

Psychologist Dr Jennifer Lunt and L’Oréal's global health and safety director, Malc Staves, are urging businesses to consider using behaviour change interventions to mitigate COVID-19 risks. 

As we slowly emerge from lockdown, we are facing a ‘new’ normal, where it looks increasingly likely that we must live and work with COVID-19 risks for many months to come. Throughout the pandemic, we have seen health and safety take a dominant role in national and international discourse and within organisational risk management. Health and safety has been openly acknowledged as integral to keeping a workforce functional. This must be sustained in a complex COVID-19 risk context. 

Implicit in creating the new normal is behaviour change. It's helpful to reflect on what we have learnt so far about behaviour change in OSH generally to determine what is critical to working safely in this new normal.

Change in itself is regarded as one of the best opportunities on which to piggyback further behaviour change. This article represents a summary of this learning. It also invites readers to share their experiences in applying low-cost behaviour change interventions – in particular via this survey. The results will be published on ioshmagazine.com at a later date. 

Human error vs violation

In health and safety, we know that there are two main types of risk-taking behaviours: violations, which are intentional; and human error, which are unintentional. The rise of behavioural insights and nudging over the last decade emphasises that we must consider two types of decision-making systems as mediators of behaviour change. These comprise: 

• System 1: referring to a more instinctive, emotional and automatic decision-making processes; thereby more closely aligned to human error
• System 2: referring to a more deliberate, logical but slower decision-making processes, appearing more like violations. 

Contrasting the UK public’s response early on in the pandemic with what has since followed, helps highlight the distinction between the two systems. Prior to lockdown in mid-March, when social distancing was more reliant on individual discretion, more of the public went to the beaches, parks and mountains than should have done.

At this point, we were perhaps prone to some of the hunches and decision-making biases that characterise system 1. This includes ‘discounting’, whereby we tend to underestimate longer-term risk, or ambiguity bias, where we favour making decisions based on what we know (usual ways of enjoying sunshine) over what we don’t know (which was, back then, the experience of having COVID-19). The more people that did it, especially from among our own ‘herd’ (such as hill walkers or sunbathers), the more it was legitimised through ‘herd mentality’. 

System 2 then seemingly got the upper hand. The imposition of lockdown, daily death tally, array of personal stories of tragedy, severity of symptoms experienced by formerly healthy people along with protracted recovery for many made the health advantages of observing social distancing much more salient, obvious and concrete. At that time, New York Governor Andrew Cuomo bluntly captured the choice as ‘isolation’ or ‘death’ as COVID-19 gripped his city. The choice between the immediate gains (life preservation and being a responsible citizen) and cons (disrupted lives, economic disadvantage) kept the decision firmly within our more rational system 2. It was a no brainer. 

As we slowly emerge from lockdown, the balance between systems 1 and 2 is seemingly in continual flux. Now that the economic consequences are becoming tangible and personal, and pandemic advice more nuanced and context dependent, we are once again more prone to the biases, emotions and hunches that reside within our system 1. 

Nonetheless, our response to the pandemic tells us that that health and safety messaging must appeal to both emotional and intellectual reasoning both to motivate and ensure that risk judgements are reliably informed.

Behaviour change

So what else have we learnt within health and safety in recent years about how we can encourage safety behaviour via these two systems? Relevant learning can be drawn from human factors, safety culture, behavioural safety and behavioural insights work. The latter in particular refers to the use of simple low-cost interventions to affect behaviour change within large populations, primarily but not exclusively through nudging a desired behaviour without restricting choice and without necessarily invoking conscious awareness. 

Behavioural insights research has burgeoned over the last 10 years, with successes in a range of areas such public health, finance and business practices. Its reach eventually extended to health and safety, culminating in publication of the EAST Health and Safety report.

Key lessons learnt from applying relevant behaviour change theory to health and safety, and nudge theory in particular, include the following.

1. Nudges aren’t just subconscious

The most widespread behaviour change approaches used in health and safety has mainly concerned behavioural safety and safety culture, intended to modify behaviour via observation and feedback on the one hand, or reshaping the underlying values on the other. Arguably the mechanisms by which these operate make them more applicable to intentional behaviour, such as violations. 

The advent of ‘nudge’ presented the health and safety practitioners potential new options for changing safety behaviour. The emphasis on the design of the social and physical context so as to nudge more desirable behaviour without invoking conscious awareness implied some relevance to hard-to-resolve human error.

From the nudge narrative that has developed since, it appears misguided to regard nudging as a purely sub-conscious process. For example, a nudge classification developed by Lindhout and Reiners (2017) separates nudges according to whether they are either transparent or non-transparent, and, crucially, automatic or rational. Similarly, nudge-type interventions utilised by the behavioural insights team include examples that encourage deliberate contingency planning via ‘implementation intentions’ (for example, pledges such as ‘if I sit at my desk longer than 50 minutes I will take a short walk’).  

2. ‘How’ rather than ‘why’ classifications 

Various methods for classifying nudges have been developed over the years. ‘MINDSPACE’ is an earlier example that appeared to group nudges according either the type of decision-making bias they target (‘norms’ or ‘ego’) or the nudge ‘method’ used (incentive).

Simpler groupings have more recently emerged that focus more on nudge ‘methods’. This includes the Behavioural Insights 'EAST' classification steering selection of methods that are easy, attractive, social and timely. Similarly, and by no means exclusive to nudge, Michie et al’s (2014) behaviour change wheel separates behaviour change interventions according to their ‘function’ (such as coercion, incentivisation, persuasion, education, etc).

Despite this focus on how nudges methods work, there could still be merit in detailing to the health and safety practitioner the nature of the decision-making biases that nudges are designed to target. This could (a) explain why employees don’t follow the rules despite safety awareness raising and (b) educate health and safety decision makers on some of the biases they may be subject to when interpreting the level of engagement within their workforce. For example, ‘loss aversion’ can be used to explain why stop techniques can fail to work. The certain losses of stopping production could easily be perceived as outweighing the uncertain gain of preventing an accident that might or might not happen.

Similarly, explaining to senior managers accountable for health and safety how confirmation bias works (sub-consciously selecting information that verifies an existing point of view) might encourage them to actively seek bad news as a way of staying responsive to risk. 

3. Nudging one-by-one?

Guidance on nudge design advocates identifying a single behaviour to target and then working backwards (for example identify determinants via research and consultation, select change techniques, choice of media, etc). For the health and safety practitioner, this creates clarity over how to design a more precise nudge-type intervention. However, we know that risk taking can be a reflection of wider safety culture, and therefore occur across a range of behaviours. To focus on discrete behaviours or even ‘spill-over’ behaviours’ may overlook shared underlying root causes and be unnecessarily inefficient as a result. 

4. It’s not just effectiveness that matters

Formal guidance on selecting the most appropriate behaviour change intervention recognises that is not just effectiveness that is important, but that affordability, practicality, acceptability (usability, ethic), side-effects (unintended consequences) and equity (will reduce harm to a similar extent for all recipients) should also feature in behavioural intervention design decision (Michie et al 2014). 

5. Sludge or nudge?

Nudging has often been criticised as manipulative, especially where it is ambiguous as to whether it is in the recipient’s best interests. Nudge misuse has more recently labelled as ‘sludge’, referring to behavioural interventions that cause friction in some way, for example, by being onerous, confusing, or against the persons best interest. An example of the latter includes the gambling apps that require football fans to make a wager in order to view a football game on line Arguably, safety has had an easier time in fending off sludge. Nudging someone to be safer can only ever be in their best interest, meaning that nudging safety comfortably fulfils ‘Rawls Publicity Principle’ of being to the general public’s advantage.  

Other more specific lessons learnt from applying nudge-type interventions to health and safety are captured in the table below and organised according to the BIT’s EAST framework. Examples of their relevance to COVID-19 risk control are provided in italics.

Striking lessons

The rise of sludge in particular reinforces the importance of testing nudge-type interventions prior to mass roll out, for potential unanticipated consequences. For example, the widely cited automatic pension enrolment schemes once touted for their success in enabling employees to plan for their future has received some criticism for denying employees opportunity to seek alternatives (Bask 2019).

For such reasons, we urge companies to test nudge ideas’ acceptability and potential. However, when conducting research for this article, what became blatantly apparent was the paucity of nudge reporting and testing with health and safety relative to other areas, especially within peer reviewed publications and from within the UK. If we are truly to understand the potential of nudge to countering human error in particular, we as a health and safety community need to be more transparent in the detail by we report nudge-type interventions and measures we take to test them. This need not necessarily be full blown randomised control trials, running some focus groups to sense-check the acceptability of ideas would be a good start.

With this in mind, we are running a survey to facilitate more shared learning on how nudges have been used in health and safety, and similarly collect views on how nudge-type interventions can help mitigate COVID-19 risks going forward. In particular we are interested in ‘upwards’ nudge-type interventions that have been used to convince senior leadership and board executives about the importance of health and safety. If you would like to participate in this survey please access the link here. We will report the findings in a subsequent article on ioshmagazine.com. 

Finally, one striking lesson learnt from this pandemic is the need to re-evaluate the societal significance of now essential jobs that were formerly regarded as low skilled and poorly rewarded. The reporting of human capital metrics (such as leading indicators such as safety culture survey ratings) has been championed in recent years as a critical mechanism for reinforcing the health and safety business case and engendering awareness about why workforce sustainability is critical to an organisation’s overall sustainability. As an example of a nudge-type intervention, incentivising more widespread corporate human capital reporting could help drive a more compassion-based view of the jobs and workers that matter that is attractive to investors, c-suite executives and employees alike.