U.S. COVID-19 shutdowns have saved far more lives than they cost

As daily confirmed cases of COVID-19 race upwards in the United States, many states have acted to stop, slow down or reverse the “re-openings” that have occurred. The dreaded word “shutdown” is being heard again, and even the conservative governor of Texas has talked of re-imposing one “as a last resort.” But as state and local authorities respond to the surge of new infections, we can expect insistent claims that the shutdowns last spring not only had dire economic consequences, but actually cost Americans more lives and life years than the pandemic could ever have. This viewpoint, however, is not only unsupported by actual evidence but is sharply contradicted by that evidence. It is important that policymakers understand why.

The most sophisticated presentation to date of this “cure worse than disease” argument was probably the widely circulated and quoted article titled “The COVID-19 Shutdown Will Cost Americans Millions of Years of Life” [1]. Its authors deserve credit for trying to quantify the adverse health consequences of the recent shutdowns. That said, their analysis is immensely misleading. Their article is replete with statistics, but one fundamental quantity is conspicuously absent: the number of lives or life-years that have been saved by the shutdowns. In effect, the authors performed a cost-benefit analysis that treats the benefits as nonexistent. The only data about COVID-19 that enter their calculations concern the deaths that occurred despite the shutdowns. The authors thus counted the failures of the shutdowns against the virus but not the successes.

But how many lives were actually saved by the shutdowns? Here a credible analysis starts with the epidemiological study in Nature [2] that estimated that U.S. shutdowns have prevented 60 million COVID-19 infections. Given a U.S. population of 330 million, 60 million infections corresponds to 18% of the population. That percentage hardly seems outlandish if we consider some early outbreaks of COVID-19 such as the Washington state nursing home that was an early “hot spot” in which 57 of the 89 residents (64%) tested positive for COVID-19, as did 100 of the 175 (57%) participants at the Biogen conference in Boston.

How many deaths might have occurred because of 60 million infections? While there is uncertainty about the death-rate-per COVID-19 infection, an especially reliable estimate arises in New York City, where 21,940 people died from COVID-19 by June 30, while random testing showed that 1.814 million residents (21.6% of the population) had experienced the virus. Those two numbers yield a death rate of 1.21% (i.e., 21,940/1,814,000), under which 60 million COVID-19 cases would generate 725,400 deaths. That this number is plausible is suggested by another study in Nature [3], which estimated that shutdowns in Western Europe saved 3.1 million lives.

Skeptics might respond that results based on epidemiological models are inherently suspect. For that reason, it is important to consider an illuminating real-life “natural experiment” in Scandinavia. All four Scandinavian countries are prosperous and have excellent healthcare systems. Of those, Sweden rejected shutdowns but by no means ignored COVID-19: it issued public warnings, closed down universities and high schools, and urged its citizens to voluntarily take precautions. Denmark, Finland and Norway went further and implemented lockdowns. As of this writing, Sweden’s rate of deaths per million people is at least seven times greater than the overall rate of the other three – five times that of Denmark, nine times that of Finland and 11 times that of Norway.

But what of the deaths not caused by the disease that can be attributed to the shutdowns? It is correctly noted that economic hardship can cause suicides, while the declines in non-COVID medical treatment during the shutdowns could increase deaths from heart disease, diabetes and cancer. But one must move beyond simply identifying the indirect harms from the shutdown to actually estimating their true death toll. Data from the Centers for Disease Control and Prevention (CDC) for March 15 to May 1 indicate that while the spring shutdowns were in effect, there were at most 24,900 excess U.S. deaths beyond those tied to the virus. The phrase “at most” is appropriate because experts believe that thousands of deaths attributed to influenza and pneumonia were actually caused by COVID-19. (These CDC data, which are considered accurate eight weeks after an observation period, were retrieved on July 5.) While all such deaths are tragic, this upper limit of 24,900 is a factor of 29 below the 725,000 U.S. deaths the shutdowns may well have averted.

The current spike in COVID-19 cases may be only a prelude to a huge resurgence of the virus in the fall. For that reason, it would be greatly unfortunate if policymakers believed that a strong response would inevitably be counterproductive. That belief could cause large numbers of needless deaths, and there is no serious evidence on its behalf.

References

1. S.W. Atlas, J.R. Birge, R.L. Keeney and A. Lipton, 2020, “The COVID-19 shutdown will cost Americans millions of years of life,” The Hill, May 25, https://thehill.com/opinion/healthcare/499394-the-covid-19-shutdown-will-cost-americans-millions-of-years-of-life.
2. S. Hsiang et al., 2020, “The effect of large-scale anti-contagion policies on the COVID-19 pandemic,” Nature, June 8, https://www.nature.com/articles/s41586-020-2404-8.
3. S. Flaxman et al., 2020, “Estimating the effects of non-pharmaceutical interventions on COVID-19 in Europe,” Nature, June 8, https://www.nature.com/articles/s41586-020-2405-7.