Studies Can’t Prove Lockdowns Didn’t Work - Or That They Did

As government-imposed pandemic lockdowns begin to wind down, policymakers, analysts and citizens turn their attention to the question of how well states and political leaders balanced their costs and benefits.

This topic was the focus of a virtual event, “Are Lockdowns Effective? How to Measure the Impact of COVID-19 Policies,” sponsored by the Mackinac Center for Public Policy. It featured economists Dave Hebert of Aquinas College and Michael Makovi of Northwood University.

In his review of lockdowns, Hebert used a cost-benefit analysis to estimate the optimal balance between their benefits (fewer cases of illness and fewer fatalities) and their costs, both in economic terms and in personal and societal ones.

The ideal lockdown would deliver the most protection at the least cost. This, Hebert said, would appear to call for a lockdown to be at roughly 40% of its full force, depending on how the concept of a lockdown is defined. As he added, we can’t “pretend to know what a 40% lockdown would look like.”

The least costly response, he said, would not involve government mandates. Rather, it would rely on individuals to regulate their own behavior voluntarily, using measures such as mask-wearing, careful use of sanitation, social distancing and other measures. A culture of personal responsibility would lower the pandemic risks while also weakening the case for very stringent government regulations.

Overbearing government lockdowns and more social isolation, he said, imposes a large number of costs on individuals, including mental health problems, intensified loneliness, increased anxiety and more domestic conflict. Herbert said these costs are “grossly under-acknowledged,” and disproportionately harm women. The gains achieved by women in the last 50 years, he said, are “slowly being eroded due to strictness of the government lockdowns.”

Although empirical models can be useful in investigating the pandemic, Michael Makovi advised everyone to approach their conclusions with skepticism. The important question for an analytical model, he said, is whether its predictions are correct.

Makovi compared states with high levels of restrictions to those with low levels. He looked at policies such as closing schools, canceling public events and issuing stay-at-home mandates. He used four measures of health outcomes: total deaths per 100,000 people, excess deaths per 100,000 (total deaths less the typical average number for the time period studied), COVID deaths per 100,000, and COVID cases per 100,000.

His conclusion was that over a single month, random chance alone could explain the difference in COVID death numbers between states with more restrictive policies and those with less restrictive ones. And when he extended the analysis to two-month periods, the results became confused in ways that made it impossible to draw any conclusion.

Makovi also said that epidemiologists were aware, even before the current pandemic, that lockdown policies could be ineffective or even do more harm than good. He cited a 2006 study from the University of Pittsburgh Medical Center. A team of scholars there found that quarantines are ineffective because they only delay the onset of widespread immunity.

Epidemiologists, he said, have also recognized that lockdowns might be pursued for harmful reasons. He cited a 2019 report from Johns Hopkins University, which noted, “Implementation of some NPIs [non-pharmaceutical interventions], such as travel restrictions and quarantine, might be pursued for social or political purposes by political leaders, rather than pursued because of public health evidence.”