Thread by @charles_gaba

📣 Head's Up: This week I'll be re-running my partisan COVID death rate estimates again, but with *age* factored in.

Should be interesting to see how this compares w/my previous "just shift the results red 20% to account for GOP voters being older" spitball.

Note: The data will be as official/concrete as possible (based on data from the CDC, Census Bureau, FEC, etc), but the one factor I'm debating is which age-based 2020 exit poll source to use.

Pew has a good *national* study but I need it for each state. Suggestions welcome.

For instance, in Alabama, here's NY Times, CBS, ABC & CNN.

Hmmm...The NY Times seems to be the odd man out here. On the other hand, only 5% of all U.S. COVID deaths to date are among those under 45, so I can probably combined the first 5 age brackets anyway.

I was thrown by the NY Times saying that 52% of *18-24 yr olds* voted for Trump, but then I remembered that this was Alabama so I suppose that's possible.

UPDATE: Hmmm...OK, for Alabama specifically, I previously estimated that perhaps ~4,500 more Trump voters have died of COVID since 11/04/20 to date than Biden voters, assuming a flat 1.2x GOP lean.

When broken out by age group, it comes in at around ~5,400 more.

Unfortunately, I don't think there's exit polling at the county level (certainly not for smaller counties) & even if there was the sample size would be too tiny to make it usable, so I'll have to apply the statewide estimates across each state's counties. Not ideal but (shrug).

Anyway, I hope to have all 50 states + DC done early this week. Stay tuned!

D'oh!! OK, it looks like the reason ABC/CBS/NBC/CNN are all so similar is because...they all used the same pooled exit polling data.
s.abcnews.com/assets/dtci/el…