@merrymouse
Aye - a working paper, not peer reviewed.
Claims to be systematic review but the exclusion criteria is strange and will certainly bias towards the null - for example they've disregarded any study which includes a modelled counterfactual (which most epidemiological studies do)
This means the review is not representative of research on lockdowns, and basically has removed any research which is likely to demonstrate benefits of lockdown on infection rates/death.
One really notable error (which draws into question the agenda here) is that the quote the Oxford Stringency Index a fair bit, but when it comes to it, they exclude the paper this quote came from in their meta analysis. This paper found a huge benefit of lockdown on COVID deaths.
Finally no robust risk of bias/quality assessment done, which is the gold standard of systematic reviews methodology. Presumably this was omitted because many of the papers backing their conclusions are too low quality as to be used to infer causality.
Other things revealing their bias - they include a paper by Chisdza et al, stating it found no benefit for lockdowns, when this isn't what their analysis showed. The authors have got in touch stating this. This is true for several other quoted papers.
It's strange that @Flyonawalk who quoted more or less talking about the paper didn't pick up on any on this, even though the prenseters referenced it.
Conclusion? This is a perfect teaching example of how to not conduct a systematic review, and how if you cherry pick evidence and misinterpret it, you make "the science" say anything you want it to.