Our lockdowns are not deadlier than the disease

Our lockdowns are not deadlier than the disease
© Getty Images

Four scholars argued in a recent op-ed for The Hill that the deaths and “accumulated years of life lost” caused by the “near complete economic shutdown” in response to the COVID-19 outbreak are greater than the lives saved. The lead author, Dr. Scott AtlasScott AtlasFauci defends Birx: 'She had to live in the White House' UPDATED: McEnany, Fox News talks on pause Birx: Someone was delivering a 'parallel set of data' on coronavirus to Trump MORE, a senior fellow at Stanford University’s Hoover Institution and a retired neuroradiologist, has made similar arguments in other media outlets for two months.

The Hill piece appeared impressive. It had links to medical and academic journals and seemed to provide hard numbers. We eagerly set out to validate their work.

The authors said that the shutdown was costing more than $1 trillion a month, and would kill 7,200 of us each month from the economic effects alone. But as we examined their assumptions, the model collapsed like a house of cards.


They got the $1 trillion figure from a “back-of-the-envelope” model (in the words of the cited study’s authors) by Makridis and Hartley, based upon a Department of Labor survey of percentage of specific industries that are “digitalized.” Even though social distance regulations differed by state, there was no adjustment for differences in the Makridis-Hartley model: Why did Alabama have a 5 percent GDP decline when it had no shelter-in-place at the time of analysis? This may be why that study’s two authors warned that their model should be “considered the upper bound of estimates.” 

U.S. monthly GDP is $1.78 trillion, so the $1 trillion GDP reduction would represent a 56 percent contraction of the economy in those months, an indefensible number when unemployment nationally was only 14.7 percent at the end of May and there was a $2.4 trillion stimulus package. And most studies show that the majority of jobs lost were not in high-paying industries that contribute the most to GDP.

The core of their economic impact analysis is, thus, wrong.

They insist that the $1 trillion monthly contraction would result in 7,200 deaths per month, based upon a 1994 New England Journal of Medicine editorial. The analysis cited in the 1994 editorial was done in 1989, validated with national data on income and mortality statistics of “white family members aged 25-64” in 1959 (with a top income bracket >$10,000) and 1980. So much has changed in our health care system – life expectancy, distribution of income and the economy as a whole – that the only excuse for using such ancient data would be if it were the only such study. 

Fortunately, more rigorous recent data show that there is an inverse relationship between economic downturns, such as recessions, and mortality: “Recessions coincide with lower mortality, while economic expansions coincide with higher mortality, so that mortality fluctuates with the business cycle, procyclically.” As it turns out, when everyone is suffering together from an economic downturn, deaths go down. 


So, the two biggest numbers in the column are shown to be false.

The authors also predicted that more than 40 percent of workers would not return to employment, based on 1979-1981 data from unemployment insurance recipients in Missouri and Pennsylvania. How is this applicable across the entire nation during a 2020 pandemic? And not returning to one’s original job does not mean not finding another job. 

Their claim that “(e)mergency stroke evaluations are down 40 percent” based on a study by Kansagra et. al. was true, but only for a two-week period. Imaging quickly recovered so applying this 40 percent reduction across the entire isolation period is a misuse of the data. And even the assumption that each missed imaging appointment equals a missed stroke is ludicrous. Only some stroke evaluations identify a recent stroke or prevent an incipient one, and anyone with significant signs of stroke (paralysis, aphasia) would certainly have been imaged.

Their assertion that 50 percent of cancer patients are missing treatments is from a United Press International story about COVID-19 in China, and that story itself never states that 50 percent of treatments were missed.  

The statement that 150,000 cases of cancer were not being diagnosed is extrapolated from a study in the Netherlands, and those data do not indicate that the cancer was never diagnosed, but that diagnostics were likely “postponed.” Their statement that two-thirds to three-quarters of “routine cancer screening” in the U.S. are not getting done was another misinterpretation. 

The report actually stated: “Tests for Leukemia and Multiple Myeloma Down Roughly 20%.” The only screening reduced by two-thirds was PAP smears. PAP smears are done annually for early detection of pre-cancerous cells, and there is no evidence that a one to two months delay in this annual screening has any impact on mortality. And their study showing delay in screening increases mortality? It was a Canadian study that showed there was no association between treatment delay and outcomes in men.

So don’t believe the hype: There was never a near-complete economic shutdown. 

The economy did not contract 56 percent per month. Data from the 1950 and 1980s should not be applied to a 2020 pandemic, especially when recent studies show the opposite effect. 

In an opinion column on March 26, citing the early fatality-rate estimates of another researcher, Dr. John Ioannidis, Dr. Atlas wrote: “There is massive uncertainty, but using Ioannidis’ mid-range fatality rate, this virus could cause about 10,000 deaths in the United States overall, overall [Sic], a number that would not be extraordinary news in the total of flu-like deaths every season.”

We think the predictions in his more recent model will prove to be equally valid.

Tracy Mayne, PhD, is a former director of HIV epidemiology and surveillance at the NYC Department of Health and Mental Hygiene, and has spent more than 20 years doing epidemiology and health economic research in the health care and pharmaceutical industries. Jeremy Mayer is an associate professor at the Schar School of Policy and Government at George Mason University.