The Cummings effect: politics, trust, and behaviours during the COVID-19 pandemic

On May 22, 2020, The Guardian and Daily Mirror newspapers in the UK published details of how Dominic Cummings, senior aide to the British prime minister, had broken lockdown rules by travelling 420 km to a family estate with his wife (who had suspected COVID-19) and child. Although some other officials and senior figures had also broken the lockdown rules, this transgression was the first to not immediately be followed by an apology and resignation. The event prompted media condemnation, with concerns about transparency, accountability, and equality,¹ and many scientists spoke out about the effect of Cummings’ actions and the UK Government’s defence of Cummings in undermining essential public health messaging.^1,2

It is only now, however, with the benefit of hindsight provided by systematic data, that we can see these negative effects in stark detail. New analyses of 220 755 surveys from 40 597 individuals in England, Scotland, and Wales, completed between April 24 and June 11, 2020, as part of University College London’s COVID-19 Social Study, show that these events undermined confidence in the government to handle the pandemic specifically.

We report the change in ratings of confidence in the government to handle the pandemic from the baseline on April 24, 2020 (appendix p 2). Participants from England answered about central government, and participants in Scotland and Wales answered about their own devolved governments. Confidence was measured on a scale from 1 (not at all) to 7 (completely). The sample was well stratified across sociodemographic factors and weighted to population proportions for core demographics (appendix p 9). Starting on May 22, 2020, there was a clear decrease in confidence in England, a decline that continued over the following days. Analyses of data from Google Trends showed that public searches of Dominic Cummings’ name peaked 3 days later (May 25, 2020; appendix p 1) when he gave a televised statement. This peak coincided with the steepest decline in confidence in government (appendix p 2).

To ascertain whether this decrease in confidence was as a result of the Cummings events (a Cummings effect), we carried out analyses using two types of comparisons. First, we compared the responses for people living in England to those of people living in the devolved nations of Scotland and Wales who were asked to rate their confidence in their own devolved governments. There was no evidence of a similar large decrease in confidence in the governments of the devolved nations either descriptively (appendix p 2) or statistically (appendix p 5) during the 3 weeks following May 22, 2020. Second, using data from questions identical in format to those about confidence in government, we compared confidence in the health service to cope with the pandemic, and confidence that access to essentials (eg, food and medication) would be maintained during the same time period. There was no evidence of a decrease in confidence in the health system or confidence in acquiring essentials during the same time period, either in descriptive data or when applying statistical tests (appendix pp 2, 5), further showing that the change in confidence in the government was a considerable departure from the weeks preceding the Cummings events.

Public trust in the government’s ability to manage the pandemic is crucial as this trust underpins public attitudes and behaviours at a precarious time for public health. Our data show how closely public confidence is related to government announcements regarding COVID-19. After an initial increase in public confidence in the ability of the government to handle the pandemic well between March 21 and March 23, 2020, as lockdown came in, the government’s announcement on May 10, 2020, that society would begin to reopen in England through a staged series of lockdown easing measures as part of a new COVID-19 alert level system was followed by a decrease in confidence (appendix p 2). Leaders of devolved governments in Scotland and Wales who expressed concern that these measures were risky and premature and who did not change lockdown measures or messaging did not see any clear decreases in confidence from their public.³ Data show that confidence stabilised and even improved slightly in England in the fortnight following these events, until the Cummings effect.

This finding is echoed by data from weekly political surveys, which show that confidence decreased with these announcements but then remained stable for 2 weeks until the Cummings events, when confidence suddenly decreased further (appendix p 1).

Another reason for concern is that trust is related to people’s willingness to follow rules and guidelines, both generally and during the COVID-19 pandemic,^4,5 which is fundamental to the control of infection and mortality. There had already been a gradual decrease in public adherence to guidelines before the publicity about Cummings’ actions on May 22, but the difference in this decline between England and Wales and Scotland grew in the 3 weeks following (May 22–June 11, 2020; appendix p 7).

Although, as of June 17, 2020, more than a month has passed since the Cummings events, data show there has been no recovery in confidence in the government, with confidence in England remaining low and gaps between confidence in England and confidence in devolved nations growing (appendix p 7). Trust in government decisions and actions relating to the management of COVID-19 is a major challenge worldwide, and these data show the negative and lasting consequences that political decisions can have for public trust and the risks to behaviours.

We declare no competing interests. The COVID-19 Social Study was funded by the Nuffield Foundation [WEL/FR-000022583], but the views expressed in this Correspondence are those of the authors and are not necessarily those of the Nuffield Foundation. The study was also financially supported by the MARCH Mental Health Network, UK Research and Innovation [ES/S002588/1], and the Wellcome Trust [221400/Z/20/Z]. DF was funded by the Wellcome Trust [205407/Z/16/Z]. We are grateful for the support of several organisations in recruitment efforts, including the UKRI Mental Health Networks, Find Out Now, UCL BioResource, SEO Works, FieldworkHub, and Optimal Workshop. The study was also supported by HealthWise Wales (the Health and Care Research Wales initiative, which is led by Cardiff University, Cardiff, UK, in collaboration with SAIL, Swansea University, Swansea, UK). The funders of the study had no role in study design, data collection, data analysis, data interpretation, or writing of the report.

The corresponding author had full access to all the data in the study and had final responsibility for the decision to submit for publication.