Weekly Update: COVID Death Rates by Partisan Lean & Vaccination Rate

I've been posting weekly looks at the rate of COVID-19 cases & deaths at the county level since the point at which every U.S. adult could theoretically have received 2 COVID vaccination doses nearly a year ago, broken out by partisan lean (i.e, what percent of the vote Donald Trump received in 2020), as well as by the vaccination rate of each county in the U.S. (nonpartisan).

For a long time I used July 1st, 2021 as my start point, but in recent months I decided to back this up to May 1st, 2021 instead. Pinning down an exact date for this is a bit tricky since a) different populations were made eligible at different points in 2021, and b) it takes 3-4 weeks after getting your first vaccination dose before you can get the second one, but May 1st is what I've finally settled on.

As always, here's my methodology:

Remember: "Decile" means 1/10th or 10% of the total population (all 50 states + DC).

IMPORTANT UPDATE: I've decided to stop posting the case rates for awhile for a simple reason: The rise of home-based testing, and the general lack of the results of those tests being officially reported to county/state health departments, has made official case rates kind of meaningless at this point. I may bring case rate tracking back in the future but for now I'm giving them a pass in these updates.

As for death rates, there's still a clear and dramatic correlation between how much of the population has been 2-dose vaccinated and its COVID death rate since last May. The least-vaccinated decile has a death rate 3.7x higher than the most-vaccinated decile. This is actually down somewhat from last week, though it's important to note that the some counties jump up or down between decile brackets from week to week as their vaccination rates pass by each other.

NOTE: The vaccination rate graph includes the Miami-Dade County change I wrote about several weeks ago:

As always, what's even more disturbing is how closely the death rate by partisan lean matches the death rate by vaccination rate; they're nearly mirror images of each other: The death rate since last May in the reddest tenth of the country is 3.3x higer than in the bluest tenth:

Here's scatter-plot versions of the county-level death rates since 5/01/21 by vaccination rate and by 2020 partisan lean:

The graph below shows how the ratio of case and death rates in the reddest and bluest deciles have changed over time. The gap reached a peak of 4.5x higher at the height of the Delta wave last fall before dropping significantly during the Omicron wave...but since then it's flattened out and is now moving back up again. Note that the small bump in March 2022 was due to a one-time data adjustment by the Massachusetts Health Dept.:

The final graph below shows the 14-day moving average of the death rates in the reddest & bluest deciles, dating all the way back to January 1, 2021. As I predicted back in late December, after reaching near parity in late January, the death rate in the reddest parts of the country is again on the rise relative to the bluest parts. Axios recently posted a state-level version of this graph, although in that case they went with states where Trump or Biden won by 15 points or more: