What Evidence Demonstrates that Lockdowns didn’t Help Stop the Pandemic?

On the one hand, just look around us. Whether locked down or “freed up”, we didn’t see a significant reduction in the spread of COVID-19 for nearly two years. But that doesn’t necessarily make a slam-dunk case immediately that lockdown policies were counterproductive. 

Fortunately, we don’t have to worry about that either, because the results are in, and lockdowns didn’t work.

Johns-Hopkins University researchers performed what’s called a “meta-analysis” of 34 studies on lockdowns (prior to the Delta wave), and their conclusion was that “lockdowns have had little to no public health effects, [but] they have imposed enormous economic and social costs where they have been adopted.”

It’s notable that this is a working paper and is not as of March 2022 in the peer-reviewed literature yet, but the results are still striking. Here are a few key points from the text:

• The meta-analysis took into account lockdown stringency index studies, shelter-in-place order (SIPO) studies, and specific non-pharmaceutical intervention (NPI) studies.
• While epidemiological simulations suggested that lockdowns would have a good effect, the actual results as transpired in 2020 suggest that the simulations were way off.
• None of the three groups suggest that lockdowns had a significant effect on COVID-19 mortality. At best, shelter-in-place orders might have reduced mortality by a mere 3%.
• Part of why that reduction of 3% is so inconsequential is that some studies even show increased mortality rates as a result of lockdown policies, on the order of 1-3%. Hence, assigning much significance to the policies as a help against COVID-19 does not seem warranted.
• The unintended consequences of lockdown policies, both economically and socially, outweigh the barely perceivable benefits, if any can be found at all.

Clearly, the JHU paper is not the final word, it suggests strongly that finding significant observable effects on mortality from lockdown-style COVID policies is almost impossible.