As a single data point: despite a universal mask mandate that was enacted early on (May 6th 2020) along with a very high compliance, Massachusetts had the 3rd highest COVID death rate in the US (as of Aug 9):
A little over 20% of Massachusetts deaths occurred before May 6th, 2020. Another 20% occurred in the 3-4 weeks immediately ofter which are likely largely attributable to people who got COVID before early May.
If you compare Massachusetts deaths after that point to other states they come out somewhere near the middle.
Not significantly, it came into force May 6th indoors and outside if you couldn't maintain 6', but most cities in eastern MA had 100% outdoor mandates too. I think it was Nov 6th when they made it 100% outdoor statewide regardless of distancing. The mask mandate wasn't lifted until May of this year.
That's not how this works. You can only compare with the future that wasn't. People may decide all by themselves that wearing a mask is a good idea, mandate or not, and then there is the point that not all masks are created equal, which is the subject of this thread.
Not without carefully controlling for all kinds of variables, demographics, geography, internal and external connectivity and so on. No two areas are closely alike in all those respects.
Do you similarly feel that comparisons across different populations are invalid for arguing the affirmative?
e.g. arguments like "we should try masking / UBI / $15 minimum wage / free healthcare / universal pre-k because it worked well in ${location}" are also "not how this works" because we can "only compare with the future that wasn't?"
Do such arguments need to control for the infinite number of confounding variables before having any value?
https://www.statista.com/statistics/1109011/coronavirus-covi...