Or by the time it hit rural areas, we'd also gotten better at testing for it. Perhaps the IFR is the same across both demographics, but better testing means a higher proportion of cases were detected later in the pandemic.
Using cases, it comes out 344 Trump, 194 Biden. Using deaths, it comes out 354 Trump, 184 Biden.
The big difference this time is largely rural states that are getting hit much harder than the rest of the states. North Dakota in this allocation gets 35 cases EV and 46 deaths EV, and South Dakota gets 30 and 36. Montana get 21 and 23.
For comparison to some high population states, California is 4 and 4, New York 4 and 3, Texas 8 and 10, Florida 6 and 10.
Now I'd be curious to see something that looks at various sliding windows on the data and sees how to allocations change, but that would be a bit more elaborate than I want to do.
However, if we go by cumulative per capita COVID deaths, then the result would have been 228 Trump, 310 Biden.
Why the big swing? I think it is because the big early outbreaks that swept through retirement and nursing homes before we got better at treating it were in big urban areas. By the time it got big in rural areas, we'd gotten a lot better at dealing with it.
It would be interesting to do a similar calculation, but using say the rates over say the last three months instead of cumulative from the start. If someone wants to do that one, a good place to get the COVID data is here [1]. They have links to download the data for all the graphs as CSV files.
[1] http://91-divoc.com/pages/covid-visualization/