>Why doesn't it? Is there some evidence that it does or doesn't.... or is this more of a claim?
For the difference to be explained solely by there being more old people because old people are more vulnerable to COVID, I would interpret that to mean that any particular individual is not more or less likely to die depending on whether they live in California or Florida and the overall difference is determined by the difference in age distribution. But, as a limiting case, if all COVID deaths were among elderly people, then a 100%+ higher overall death rate in Florida which has only a 43% larger share of its population who are elderly compared to California would require that deaths per capita among the elderly population would have to be ~40% higher in Florida than in California (2 / 1.43). It's possible that my intuition is making an error here, but I think that the situation with only elderly people dying is the upper bound on how much effect the age distribution can directly have. So, something further is necessary to explain the additional (at least) 40%.
> Is there direct evidence that it's masks and not the density and quantity of old people?
I don't know whether masks explain the cited difference (I haven't seen convincing evidence but I also haven't really examined the matter), just that age distribution cannot explain the entirety of the difference. At least some age groups must have been more likely to die due to COVID-19 if they lived in Florida than in California during the cited period. It's certainly possible that some portion of that additional difference can be explained through downstream effects of the age distribution, but demonstrating that would require some additional evidence beyond the bare fact of the age distribution differences.
My understanding is there were also other policy differences between the two states besides mask requirements, but I do not have any specific reason to believe them to have had a significant (or any) effect, just like I don't have any specific reason to believe that downstream consequences of the age distribution to have had a significant (or any) effect.
> I guess it's subjective levels or risk. Is two percent lower infection rate worth it to require everyone to wear masks?
I couldn't say, although it's also worth remembering that the positivity rate data in question doesn't tell you anything about the infection rate. It doesn't even tell you what proportion of people who got tested received positive results, since the data in question is about samples rather than patients.
For the difference to be explained solely by there being more old people because old people are more vulnerable to COVID, I would interpret that to mean that any particular individual is not more or less likely to die depending on whether they live in California or Florida and the overall difference is determined by the difference in age distribution. But, as a limiting case, if all COVID deaths were among elderly people, then a 100%+ higher overall death rate in Florida which has only a 43% larger share of its population who are elderly compared to California would require that deaths per capita among the elderly population would have to be ~40% higher in Florida than in California (2 / 1.43). It's possible that my intuition is making an error here, but I think that the situation with only elderly people dying is the upper bound on how much effect the age distribution can directly have. So, something further is necessary to explain the additional (at least) 40%.
> Is there direct evidence that it's masks and not the density and quantity of old people?
I don't know whether masks explain the cited difference (I haven't seen convincing evidence but I also haven't really examined the matter), just that age distribution cannot explain the entirety of the difference. At least some age groups must have been more likely to die due to COVID-19 if they lived in Florida than in California during the cited period. It's certainly possible that some portion of that additional difference can be explained through downstream effects of the age distribution, but demonstrating that would require some additional evidence beyond the bare fact of the age distribution differences.
My understanding is there were also other policy differences between the two states besides mask requirements, but I do not have any specific reason to believe them to have had a significant (or any) effect, just like I don't have any specific reason to believe that downstream consequences of the age distribution to have had a significant (or any) effect.
> I guess it's subjective levels or risk. Is two percent lower infection rate worth it to require everyone to wear masks?
I couldn't say, although it's also worth remembering that the positivity rate data in question doesn't tell you anything about the infection rate. It doesn't even tell you what proportion of people who got tested received positive results, since the data in question is about samples rather than patients.