
Understanding how Citizens Respond to Risk
In the context of the pandemic, we ran a series of experiments to inform our understanding of people’s perceptions of risk, using the findings to develop a framework to guide people on communicating risks to citizens.
The Challenge
All of us are confronted by risks every day. We could be hit by a car when crossing a road. We might take a pill which has potential side-effects. Or we could be weighing up a business decision that looks sensible but might turn sour at some distant point. Decades of behavioural science research have taught us that there are many factors that influence our risk-based judgements. In the context of the COVID-19 pandemic, these became all the more important. So in partnership with FGS Global, the global strategic communications firm, we set out to create an evidence-based framework that could be used by individuals communicating risk to consumers or citizens.
The Approach
We first reviewed the academic literature on the factors that influence how people perceive, understand, and respond to the different risks they face in their everyday lives. We identified four key considerations: 1) conveying risks intuitively; 2) taking emotions and existing beliefs into account; 3) understanding messenger and peer effects; and 4) making it easy for people to take appropriate action.
We then sought to test this framework through a series of studies that put these different principles into effect. To do so, we recruited a representative sample of more than 4,000 participants based in the UK and the US. And we randomly assigned them to treatment groups in order to test the effect of communicating risks in different ways.
In one of the studies, for example, we looked at how easy it was for participants to understand the risks associated with certain activities (such as eating more than 25g of processed meat every day, which some studies have suggested might be associated with developing bowel cancer). Some participants had this risk conveyed to them in a way that is typical in the media: as a relative risk increase (‘the risk increased by a fifth’). Others were randomly assigned to be shown the information conveyed using ‘natural frequencies’. This included two separate images of 100 stickmen. One of these images showed the bowel cancer distribution of the group that eats more than 25g of processed meat each day; while the other displayed the effect upon the group that ate less than 25g each day (see image below).
Figure 1: Risk information conveyed using ‘natural frequencies’

The Result
In the study relating to the communication of the risk of bowel cancer, we found that our participants found it very difficult to understand the risks that they faced when using a ‘relative risk’ frame (e.g. the risk increased by a fifth). Just 3% of participants who saw the risks presented in relative terms answered questions related to the risks correctly, compared to 61.8% of those who saw the risks conveyed using ‘natural frequencies’ (see chart below).
Figure 2: Proportion of participants answering a factual question about the bowel cancer risk correctly under a relative risk frame vs ‘natural frequencies’

This study, together with many others set out in the full report, shows how important it is for anyone communicating risk to be aware of the behavioural factors underpinning how we understand and react to information.
“It was fantastic to work with the CogCo team. The highlight was the use of large scale experiments to demonstrate how people respond to risks when making critical business and life decisions.”
-Graeme Trayner, Partner and Global Head of Insight, FGS Global