Actually that can be misleading, since you really need to drill down into red and blue counties at least. For example, my state of WA is "blue" but divided into distinct red and blue sections. East of the Cascade mountains is very red. So we had our initial wave here in the Seattle area, and since then have kept it pretty well suppressed. East of the mountains they were hit hard this summer. But looking at the state as a whole you'd think the whole state was hit hard in the summer.
I realize that, but there is a limit to how finely I can divide the statistics.
If someone wants to do an electoral district by electoral district (county by county if they are REAL gluttons for punishment) analysis for the entire United States of America and then compare that analysis with a similar electoral district by electoral district analysis for every other country in the world (or even just the "G-8 + China" group plus "Europe", they have my total blessing.
However, eventually reach the point where you are trying to answer the "
If there are 10,000,000 long-haul truck drivers,
and if 300 out of every million long-haul truck drivers have an accident in a year,
and if 5% of the population is Mennonite,
then how many Mennonite long-haul truck drivers will have accidents each year?" question WITHOUT giving any consideration to the fact that Mennonites do not drive long-haul trucks.
Mind you, I rather suspect that within a couple of years we will see some Masters and Doctorate level theses coming out that DO break the analysis down to the County level (at least for an individual state) However, that those theses are likely to prove is something like
After controlling for all other variables including age, sex, education, existing rates of disease, and income, for "State A" which had 90% of its population in 10% of its area and an overall population density of 100 / sq.mi. and for "State B" which had its population evenly distributed but which also had an overall population density of 100 / sq. mi. there was a difference in infection rate that was statistically significant but which did NOT have any correlation to the states' overall population density.
HOWEVER when population density per county was examined, the correlation between population density and infection rate became stronger.
AND when population density within local sub-regions of counties was examined, the correlation between population density and infection rate became even stronger.
CONCLUSION, the closer together people live, the more likely it is that an infectious disease will be spread amongst them.
That paper should be sufficient to fulfill the requirements for the Master or Doctorate level degree it was filed in support of.
It will also rate a "Well Duuu
uhhhh!" from any reasonably intelligent and educated person.