Intended for healthcare professionals


The UK government’s attempt to frighten people into covid protective behaviours was at odds with its scientific advice

BMJ 2023; 380 doi: (Published 21 March 2023) Cite this as: BMJ 2023;380:p652
  1. Stephen Reicher, professor1,
  2. John Drury, professor of social psychology2,
  3. Susan Michie, professor of health psychology and director3,
  4. Robert West, professor4
  1. 1School of Psychology and Neuroscience, University of St. Andrews
  2. 2School of Psychology, University of Sussex
  3. 3Centre for Behaviour Change, University College London
  4. 4Department of Behavioural Science and Health, University College London

The scientific literature shows that fear is generally an ineffective way of persuading people to engage in health protective behaviours

Among the many revelations from recent leaked UK Government WhatsApp messages published by The Daily Telegraph is the fact that Matt Hancock, former health secretary, proposed using fear in order to get the public to comply with covid-19 restrictions. In December 2020, he wrote that “we must frighten the pants off everyone with the new strain.” In January 2021, Hancock discussed how to increase levels of mask wearing and worried that minor interventions would have little impact. The cabinet secretary, Simon Case replied: “I think that is exactly right. Small stuff looks ridiculous. Ramping up messaging—the fear/guilt factor vital.1

These revelations seem to confirm what many had been arguing for two years: that the UK government aimed to control the public through use of fear.2 We will leave aside the ethical and political dimensions of this argument for now in order to concentrate on the science. Both the government and The Telegraph accept the effectiveness of fear appeals—the one to extol it, the other to decry it. Both presuppose that fear is an effective way of controlling people. However, the scientific literature tells a very different story. It shows that frightening people is generally an ineffective way of persuading them to engage in health protective behaviours.3

It is true that people have to perceive that there is threat before they will take protective actions in a pandemic,4 including the covid-19 pandemic.5 So it is critical that people have a realistic understanding of what confronts them. Where they underestimate the risk, that needs to be tackled.

However, if threat communications are necessary to produce protective behaviours, they are not sufficient. Just telling people they face danger—just like not telling them—leaves them helpless to deal with it. It is only if you ensure that they also know what to do in order to stay safe, and also that they have the resources to do it, that you empower them to overcome the dangers they face. Such a combined approach has been repeatedly shown to be effective.678 Moreover, in acting to control danger—fear and anxiety is reduced.9

When Hancock and Case advocated scare tactics they were going against the scientific advice they had been given. They were not, as some have suggested,1011 in lockstep with their scientific advisors.

In March 2020, before vaccines and effective treatments were available, the behavioural science advisory group to the UK Government, SPI-B, was asked to report on “Options for increasing adherence to social distancing measures.”12 The group evaluated 10 possible options. One of these was persuasion. Here, SPI-B concluded that: “the perceived level of personal threat needs to be increased among those who are complacent, using hard-hitting emotional messaging. To be effective this must also empower people by making clear the actions they can take to reduce the threat.

As can be seen, this translates the findings of the literature into the covid domain: make sure that people have a realistic appraisal of the risks they face and that they are empowered to address those risks (which is why, elsewhere in the report, SPI-B advocated supporting communities—especially the more marginalised and deprived—and providing the resources they need to follow the rules). To suggest that this is, in any sense, seeking to control people, is nonsense. Not only is it based on sound science; arguably it is plain common sense.

This emphasis on empowerment is even clearer when one looks across the corpus of SPI-B reports. It reflected a conception of the public as an asset rather than an impediment in the pandemic. The advice was to engage with the public and focus on supporting them in doing the right thing rather than assume they need frightening and coercing in order to stop them from doing the wrong thing. This is particularly clear in another report of 3 April 2020 on “harnessing behavioural science to maintain social distancing”13 (subsequently published as a journal article).14 Among the key principles set out in the paper were the need to avoid authoritarian messaging based on coercion, an emphasis on enabling behaviour rather than the use of punishment or castigation, and the need to engage with communities in order to co-design interventions with them as opposed to imposing interventions upon them.

It is in this context that the comments by Hancock and Case attain a larger significance. Not only do they promote a naïve and mistaken strategy of fear, they also convey a wider attitude of contempt towards the public, suggesting that they cannot be reasoned with and can only be scared into doing the right thing. This contempt is a defining feature that is found throughout the leaked Whatsapp messages: contempt for the teaching profession who are described as “absolute arses” who “just hate work”; contempt for the police, described as “the plod” whose concerns about repressive enforcement of covid rules was airily dismissed with the comment that they “have been given their marching orders”; contempt for travellers placed in quarantine whose discomfort was described as “hilarious”—and, of course, contempt for the general public, not only in words, but also in deeds—for instance, Hancock getting tests couriered for a colleagues’ son even as he refused testing to those entering care homes from the community for want of adequate supplies.15

The reason why this matters is not just because of what it says about the attitudes of our elected representatives towards their electorate, but because it fundamentally compromised the government’s covid strategy. To take one glaring example, time and time again, SPI-B argued for financial support so that people could self-isolate.16 The government provided only the most minimal resources in response; and the reason became clear when Matt Hancock told a House of Commons Inquiry that, were more money available people would simply “game the system.”’17

Similar points could be made about the government’s use of fines and other punishments to secure compliance, despite advice that this could alienate the public—especially those who lacked the resources to comply.18 The same could be said about the government’s failure to heed the advice to engage with communities—especially those with a troubled history with authority—in order to increase vaccination rates19; and much more besides. The common theme is that the UK government’s contempt for, and distrust of, the public led to a systematic set of failures to engage with the public in confronting covid. In effect the government spurned the most valuable resource they had for dealing with the pandemic. This is the most fundamental revelation of the WhatsApp scandal. It is arguably a failure as bad as the failures over personal protective equipment and Test and Trace. It did not come from the government conspiring with behavioural scientists in SPI-B. It came from them ignoring their advisors.


  • Competing interests: The authors were all participants in SPI-B. SM and SR are members of independent SAGE.

  • Provenance and peer review: not commissioned, not peer reviewed.