I am referring to the conclusions section:
> First, people respond to dangers outside their door. When a pandemic rages, people believe in
social distancing regardless of what the government mandates. So, we believe that Allen (2021)
is right, when he concludes, “The ineffectiveness [of lockdowns] stemmed from individual
changes in behavior: either non-compliance or behavior that mimicked lockdowns.”
> Third, even if lockdowns are successful in initially reducing the spread of COVID-19, the
behavioral response may counteract the effect completely, as people respond to the lower risk by
changing behavior
Since it is an obvious consequence of the germ theory of disease that isolation stops the spread of disease, the only real question is if lockdowns are efficient at enforcing isolation (and in what conditions, with what costs etc). At best, the paper concludes that they are not, and that government propaganda about the risks of the disease works as well or better.
> You're attempting to refute a meta-analysis of studies looking at all the data with the example of a single country. That's meaningless. A single data point can falsify a theory but it cannot prove a theory. To prove lockdowns work there has to be a consistent impact of the policy.
I would say the null hypothesis would be that lockdowns do work, by relatively simple, almost mechanical, principles (lockdown forces isolation, isolation means the virus can't physically get from one person to another).
And there is basically no country that hasn't seen significant isolation that did well in the pandemic. Whether that isolation was caused by efective, if perhaps draconic lockdowns (such as in China or Vietnam) or by cultural norms and self preservation (such as in Finland or Norway), this remains true.
The real exceptions are countries that successfully isolated themselves from the outside world, and thereafter isolated only those carrying the virus, through broder closures and strict quarantine requirements plus testing - Taiwan, New Zealand are examples of such.
For this type of response, using broad statistical studies that equate very different ground-level phenomena (lockdowns varied wildly in form and in the degree of compliance, but the paper ignores all that in its quantitative data) to make it look good on a graph are mostly misleading - but what more can you expect from three economists dabbling their hand at epidemiology and sociology?