Promoting a nationwide collective response: lessons from the Social Activity Measure during the COVID-19 pandemic

Introduction

Alongside testing, healthcare and vaccines, society’s main defence against COVID-19 was a coordinated and collective effort to restrict behaviour in order to reduce the likelihood of infection. This included reductions in social activity and attempts to mitigate the risk of transmission when in company (e.g. maintaining distance, mask-wearing). This research makes use of a unique data-set that measured the social activity of adults in Ireland, including people’s efforts to mitigate transmission risk. It analyses how behaviour evolved over time and how it was linked to people’s perceptions of, and attitudes towards, the pandemic. The findings have implications not only for responses to future pandemics, but for any situations where Government (and society more broadly) faces a threat that demands coordinated, national collective action.

Data and methods 

We analysed data from the Social Activity Measure (SAM), which was collected fortnightly over an 18-month period during the COVID-19 pandemic from January 2021 to June 2022. SAM adapted an established psychological method, the Day Reconstruction Method, to record people’s daily behaviour in detail, together with their background characteristics, perceptions of the pandemic and attitudes towards it. The study was administered anonymously, online, to a nationally representative sample of 1,000 people in Ireland every two weeks. The final data-set consists of 36,000 surveys completed by over 8,000 different adults. 

The data were used as the basis for three types of analysis. First, we examined the trends in behaviour over time and how these related to the epidemiological situation. Second, we analysed specific influences on the riskiness of individual behaviour. Third, we looked at how people perceived the level of risk that they faced. 

Results

There were clear trends in whether individuals had close contacts, visited multiple locations outside their own home, met with other people from outside their household and, when at external locations, took measures designed to mitigate the risk of COVID-19 infection. These behaviours were strongly associated with the contemporaneous number of new daily cases of COVID-19. This link to case numbers was stronger than the association with indicators of more serious disease, such as hospitalisations and deaths, even after the vaccine became available. In general, behaviour changed slowly and cautiously, with no clear step-changes associated with events or policy changes, until the rapid lifting of restrictions in early 2022.

The amount of risk people undertook in their daily behaviour was linked to background characteristics. Working people took more risk than non-working people. Once working status was controlled for, older adults took more risk than younger ones and people with higher socio-economic status took more risk than those with lower socio-economic status. Gender differences were small and there were no consistent regional differences. 

However, overall, these background characteristics were less important than psychological factors. Chief among these was a person’s overall level of worry about COVID-19, which centred on the health of family and friends, as well as broader societal concerns about the healthcare system and the amount of the virus in the community and internationally. The implication is that most people believed that collective efforts to be cautious in behaviour would limit the transmission of the virus and have broad benefit.

While fatigue with sticking to public health guidelines became a factor in people’s behaviour as the pandemic wore on, it was far from the strongest. More important was whether people viewed the public health restrictions as straightforward to follow and whether they saw them as coherent rather than contradictory. These perceptions were, on average, positive and strongly correlated with confidence in the Government. Perceptions of how much others were complying with restrictions mattered too. These factors were much greater influences than people’s perceived likelihood of being caught and fined were they to break restrictions; voluntary willingness to do the right thing mattered more than deterrence.

People perceived more risk the more socially active they were, but perceptions were biased. For instance, people perceived increased risk from meeting someone from another household, but little additional risk from meeting a second, third or fourth person. This pattern was similar for close contacts and visiting locations outside the home. This insensitivity to the frequency of a behaviour once an individual had engaged in it suggests a degree of ‘binary thinking’. That is, risk perceptions reflected a notion that some types of social activities were safe while others were not – it was either okay to do or it wasn’t. Protective behaviours while undertaking activities (e.g. wearing masks, maintaining social distance, cleaning hands) had a much weaker relationship with perceived risk. 

Policy Implications

If circumstances are right, people can cooperate on a massive scale to achieve collective outcomes. Overwhelmingly, cooperation during the pandemic was voluntary, with only a limited role for legal deterrents. Nevertheless, voluntary collective action on this scale needs to be coordinated by the policy and communication that surrounds it.

Willingness to cooperate depends on people perceiving that the actions asked of them will lead to the desired collective outcome. Thus, in emergencies, policymakers need to search for simple rules, whether legally binding or otherwise, that are easy to follow, can be straightforwardly shown to generate the collective benefit if everyone follows the rule, and can be consistently applied and communicated. Straightforward, explicit rules are also easier to self-police and to observe in others, making cooperation more likely. Where this is achieved, the COVID-19 pandemic shows us that the large majority of people are likely to cooperate voluntarily and, to a substantial extent, to self-police the rule. 

The close relationship between behaviour and the COVID-19 case numbers confirms that putting accurate, numeric indications of risk into the public domain can strongly influence the public response. This may occur because members of the public respond to changes in the relevant number, but also because official communications become more urgent and focused when the number changes. In future emergencies, where possible, the publication of a number linked to the scale of a threat is likely to be similarly impactful.

Risk perceptions during the pandemic were biased in predictable ways. Deploying rapid, real-time research to record behaviour and to identify misperceptions can assist policy and communication to counter such biases and to identify specific, simple rules to reduce risk.