Drawing on mixed-methods research, Jet Sanders, Alessia Tosi, Sandra Obradovic, Ilaria Miligi and Liam Delaney found behavioural science to be a divisive topic in UK newspaper articles and on Twitter. They reviewed newspaper and social media discourses on behavioural science in the UK’s COVID-19 response, with a view to identify the role of transparency and trust in science actors in this high-stake context. Based on their findings, they recommend that greater efforts are made to clarify both the function of a behavioural scientist in a policy context and the diversity of approaches taken toward behavioural science to avoid media divisiveness in future emergencies.
During the COVID-19 pandemic, policy choices and the scientific advisors who have informed them were scrutinised by the media and the public in most countries. Notably in the UK, this scrutiny was in part directed at the role for behavioural science in the policy response.
The incorporation of a behavioural science perspective in the COVID-19 response can be seen, for instance, by the inclusion of David Halpern, chief executive of the Behavioural Insights Team in the government’s Scientific Advisory Group for Emergencies (SAGE) and the development of a behavioural advisory group known as the Scientific Pandemic Influenza Group on Behaviours (SPI-B). It is possible that behavioural science was particularly well-represented in the UK because it has been embedded in British policy for longer, and more widely, than in other national systems. The UK Cabinet Office was amongst the first with a dedicated behavioural science unit.
Despite the relatively smooth integration of behavioural science into a number of key policy areas, in March 2020 the role of behavioural scientists in the UK’s COVID-19 response was heavily debated in the media. We noted this as an opportunity to study the publicly perceived barriers and drivers of this new scientific tool for policy making under policy constraints.
Following an analysis of over 650 UK print articles and over 2000 original tweets (plus over 11,000 retweets) for the 24-week period surrounding the first lockdown, our research demonstrates several important findings. First, attention heightened towards behavioural science actors and principles in the lead up to the lockdown decision, and again after the first easing took place (Figure 1). These trends were marked by increasingly divisive sentiment toward their contribution to COVID-19 policies at both timepoints.
Note: Salience is calculated for a 2-week period as a) the normalised keyword frequency (per 10,000 words) multiplied by the proportion of print articles that mention the keyword and b) salience calculated as the proportion of tweets in that 2-week period that mention the keyword. Bold line represents salience in original tweets only; Dotted line represents salience accounting for retweets also. Sentiments are represented in counts of positive (+1 or +2), neutral (0), and negative (−2 or −1) bubbles over time, in green, white and red respectively. The area of the bubbles is proportional to the count of sentiments. Full-colour bubbles represent sentiments in original print or tweets only; shaded-colour bubbles represent sentiments accounting for retweets.
Though both are considered ‘behavioural science’, we identified two distinct clusters of association in social and print media: ‘nudge’ and associated concepts and actors were perceived as more embedded with policy application and most negatively; ‘behaviour change’ and associated concepts and actors were perceived as more distant from policy and most positively.
Note: Salience is calculated for a 2-week period as a) the normalised keyword frequency (per 10,000 words) multiplied by the proportion of print articles that mention the keyword and b) salience calculated as the proportion of tweets in that 2-week period that mention the keyword. (A) behaviour change (concept), (B) behavioural economics, (C) behavioural insights team (named actor), (D) behavioural scientist (unnamed actor), (E) halpern (named actor), (F) michie (named actor), (G) nudge (concept), (H) psychologist (unnamed actor), (I) psychology (discipline), (J) SPI-B (names actor). Bold line represents salience in print media (panel A) or tweets (Panel B); dotted line represents salience including retweets (Panel B). The area of the bubbles is proportional to the count of sentiments (red = −2, −1; white = 0; green = +1, +2) toward the keyword. Full bubbles represent sentiments in original tweets only; shaded bubbles represent sentiments accounting for retweets.
What drove these patterns of sentiment? Using a thematic analysis of the 111 newspaper articles with the most extreme sentiments, we noticed that concepts like ‘nudge’ or ‘behaviour change’ were often mentioned but not always fully explained. The concept of nudge, often portrayed as something risky and not totally transparent, was associated with a manipulative intent, whereas the concept of behaviour change was associated with behavioural interventions and academic rigour.
Differences between clusters are further heightened by perceptions of behaviour change and psychology as enablers of citizen choice (e.g. handwashing, social distancing), whilst negative and divisive sentiments were associated with behavioural science when applied to more politicised restrictions of citizen choice (e.g. lockdown, rules of social isolation). However, we also observed negative sentiment toward nudge for not being restrictive enough, so this polarity does not seem to explain the divisive debate entirely. Another contrast between these clusters of actors and concepts is their perceived embeddedness vs. independence from political, as opposed to public, needs. In other words, a question that is reflected by the media (and public) is to what extent behavioural scientists were seen as working for the public good, instead of biasing the selection of evidence to suit these political needs?
In addition, behavioural science as embedded in the COVID-19 policy response was heavily criticised by the media for lack of transparent practices. In contrast, when individual behavioural scientists discussed behavioural research as a tool to facilitate public involvement and transparency, its use was associated with positive sentiment.
While our results are not conclusive about the impact of this confusion for ongoing trust in behavioural science approaches in the context of public policy, we can conclude that it was a significant source of enduring negative sentiment toward behavioural science and behavioural scientists during this period.
We therefore summarise the following recommendations on transparent communication for future behavioural policy making and their immediate use for shaping communication around the behavioural COVID-19 policy measures, based on our analyses.
Discuss heterogeneity of the field internally and describe its differences externally. With discussions unfolding over the lockdown, we captured a high degree of heterogeneity discipline terms, representation of distinct perspectives and streams of research, with different levels of readiness for policy input. In addition we observed terminological confusion. For this basis we suggest to:
Offer continuous clarification of (behavioural) science terms and origins. Political philosophical tradition of libertarian paternalism and role of nudge (‘soft approach’) are often misunderstood. Addressing the historical origins and prospects of their work might be helpful in resolving confusion and clarifying distinctions between distinctive streams of thought.
Address heterogeneous (scientific) approaches and readiness. Describe the multidisciplinary approaches with a scientific perspective (psychology, behavioural economics etc) and the readiness of behavioural science to contribute to emergency situations (how systematic is the evidence; how representative of the situation addressed, see here; and here for examples of more detailed work on this).
Increase transparency on the role of (behavioural) science in policy. Even if choice processes behind individual policy choices cannot always be fully elaborated in real-time to the public, we suggest that transparency on generalised processes could aid in perceived trust and trustworthiness.
Distinguish roles. In the short term, attempt to preserve differences between policy choices and scientific evidence which supports (or does not support) it. Evidence the rigour and quality of the scientific evidence. Where does science stop and policy (or politics?) start? How do citizens identify the difference?
Address scientific independence. Mention the process of its contributions transparently, and where possible increase efforts to remain alert to independence from political processes.
Embed tools for transparent policy-making. On a longer term, we recommend that further efforts are made to clarify the ethical features of different behavioural policy tools, to embed such tools in day to day practice, and to justify policy choices where suitable.
Track choices. Keep track of choice making in behavioural units following an ethical framework. Consider developing a blockchain of science into policy.
Allow for decision tracing. Publish decision frameworks or ‘in principle routes’ of choice making as (behavioural) science advice.
Frame in line with public understanding. Our analysis shows a substantial body of public opinion expressed concerns that behavioural science, such as its use for manipulation and/or bypassing citizen autonomy. This concern is not new but seemed to heighten due to the high stake context.
Explain its contributions beyond common sense. Address the extent to which behavioural science research is seen as a valuable input beyond lay intuitions about human behaviour.
Highlight its efforts to enable citizen choice. Where this is aligned, address its benefits for enabling citizen choice where possible and if its use is for restrictive purposes offer more extensive insight on specifically why this is necessary (or how this is aligned with the broader public choice).
Monitor, adapt and address arising issues. Periodic monitoring of public confusions, conflations and sentiments can allow for emerging sciences to assess and adapt their approaches quickly. These could be addressed internally and externally and corrected by key public figures in the field in high stake contexts when opportunities arise.
Although these insights are based on one science during one crisis in one country, we expect that for emerging science to aid policy making under emergency conditions these insights will be of value.
The above draws on the authors’ published work in Frontiers of Psychology.
Jet Sanders is Assistant Professor in Behavioural Science in the Psychological and Behavioural Science department at the LSE.
Alessia Tosi is an independent data scientist.
Sandra Obradovic is a Lecturer in Psychology, Faculty of Arts & Social Sciences at The Open University.
Ilaria Miligi is a research assistant in the Psychological and Behavioural Science Department at LSE.
Liam Delaney is Professor and head of the Psychological and Behavioural Science Department at LSE.