I suspect there’s a bit more to this one. My immediate question was “do these populations differ?” because you could imagine some confounding effect where counties which rejected mask-wearing have something else in common, like say not following other best practices.<p>Skimming the paper (<a href="https://www.cdc.gov/mmwr/volumes/69/wr/pdfs/mm6947e2-H.pdf" rel="nofollow">https://www.cdc.gov/mmwr/volumes/69/wr/pdfs/mm6947e2-H.pdf</a>), check out the chart on page 4; seems like when you look at the trend lines, the counties that didn’t enact mask restrictions had a lower rate of increase in infections prior to those restrictions going into effect, and the rate of increase didn’t change after. Furthermore, there was a really big spike around the time that the restrictions were put in place in the counties that enacted mandates. That spike didn’t show up in the counties where restrictions were not put in place.<p>I’m not sure to what extent regression to the mean applies in the context of infectious diseases; you could imagine some effects like holiday week where folks increase their risk for a short time and then stop again, causing the infection rate to drop again. I wrote that before checking the dates, and indeed the spike was around 4th July, so there’s a fairly good reason to believe that at least some of the reported post-intervention drop in infection rate was just due to people reducing their mixing/socializing back to pre-holiday levels. Though the rate started increasing before Jul4 so this can’t be the full explanation.<p>I support wearing masks, but this particular study doesn’t seem like strong evidence.