The Royal Society’s results exaggerated the efficacy of mandated lockdowns.
A recent Royal Society report on the Covid-19 pandemic produced by a team led by Mark Walport, a professor of medicine at Imperial College London, captivated the press. The report concludes that non-pharmaceutical interventions (NPIs), including mandatory lockdowns, “can provide powerful, effective and prolonged reductions in viral transmission.” In sharp contrast, our meta-analysis, Did Lockdowns Work? The Verdict on Covid Restrictions, found that NPIs “had a negligible effect on Covid-19 mortality” in early 2020. For example, we found that mandated restrictions reduced mortality by 6,000 to 32,000 deaths during the first Covid-19 wave in Europe. For context, 72,000 flu deaths occur during a typical flu season.
Why do our conclusions differ so significantly from those of the Royal Society? Much of the difference stems from the fact that we are economists who specialize in economic measurement. As economists, we regularly use aggregation theory and index-number theory to measure things, like the effect of Covid-19 lockdowns on mortality. It is important to stress that economics is meant to be about people’s behavior and to measure how people react to changing incentives. Indeed, a sub-field in economics is “behavioral economics,” which incorporates cognitive and social psychology. On the other hand, the Royal Society team is, for the most part, made up of experts from medical and public-health disciplines.
In our meta-analysis, we examined the specific impact of compulsory Covid-19 mandates, whereas the Royal Society evaluated the broad societal response to the Covid-19 pandemic in its entirety. By doing so, the Royal Society’s measurements aggregate the effects of mandatory restrictions and the spontaneous changes in voluntary behavior. Importantly, this exaggerates the effects of mandatory restrictions.
To recognize the magnitude of the measurement error resulting from the Royal Society’s inclusion of voluntary behavioral changes that occurred as a result of the pandemic in its aggregate, consider what happened in early 2020 in Sweden. Despite the absence of stringent regulations, Swedes opted to cancel events like birthday parties, adopted remote working, and practiced physical distancing. Similar behavioral changes were evident in Denmark, where the majority of businesses remained operational throughout the pandemic, but many Danes voluntarily chose to work from home, and the government refrained from imposing stay-at-home directives. These spontaneous, voluntary behavioral adjustments played a significant role in shaping the pandemic’s trajectory and its effect on mortality.
These behavioral adjustments were not the result of mandatory restrictions and should not be included in studies intended to measure the effects of mandates. In consequence, the Royal Society’s results exaggerated the efficacy of mandated lockdowns. Indeed, our meta-analysis shows that, while mandatory restrictions did contribute to saving lives, their impact was minor in comparison to the profound effect of spontaneous, voluntary behavioral changes. Once people sense danger, a defense mechanism kicks in, and in an attempt to avoid or mitigate the costs associated with the perceived danger, people adjust their behavior given the new incentives.
If this weren’t enough, the Royal Society took another methodological tack that we find puzzling. Randomized control trials (RCTs) are the gold standard in scientific research because of their ability to filter out extraneous variables. A notable Cochrane Library meta-analysis of RCTs found that masks offered limited or no benefit in reducing the spread of respiratory viruses, including Covid-19, a conclusion consistent with our findings. But when evaluating with the efficacy of masking, the Royal Society’s report leaned heavily on 46 observational studies that, in contrast to the RCT studies reviewed in the definitive Cochrane study, concluded that mask mandates were efficacious.
Given the Royal Society’s prestige, it’s not surprising that its Covid report received considerable press coverage. The headlines and reporting were uniformly favorable. The message and narrative are clear: Covid non-pharmaceutical interventions, including lockdowns and masking, were efficacious. But that narrative is false.
Steve H. Hanke is a professor of applied economics at the Johns Hopkins University in Baltimore and a senior fellow at the Independent Institute in Oakland, Calif. Jonas Herby is a special adviser at the Centre for Political Studies in Copenhagen. Lars Jonung is professor emeritus at the Knut Wicksell Centre for Financial Studies, Lund University. Hanke, Herby and Jonung are the authors of Did Lockdowns Work? The Verdict on Covid Restrictions, a peer-reviewed book published in June by the Institute of Economic Affairs in London.