The Wealth Theory of Vaccine Efficacy Confounding Gets a Boost
The Vaccine Wars Part XLVIII
Pun intended, of course.
A couple of months ago, I ran some analyses on U.S. countywide data carefully put together from several sources. What I found was that
There were small and inconsistent correlations that pointed to Vaccine Efficacy (VE).
These correlations matched perfectly with county wealth differentials.
The county-level mortality data from 2019 was highly predictive of the same data from 2021, just slightly scaled.
In other words, the mild VE noted in U.S. statistics, that does not seem to match results from elsewhere in the world, much less global summaries, seems to be entirely explained by wealth effects that do not otherwise fall into variable categories we usually examine for confounding effects. And that makes sense when you think about the way Mendellian genetics studies rarely find substantial effects without aggregating hundreds or thousands of genes. Wealth improves lives in innumerable ways.
I was recently made aware of this study (Rancourt et al, 2022) recently posted to preprint with the following meandering title:
I was planning to discuss the data, but instead I will simply point out that it falls in line with the "VE as wealth effects" hypothesis, and point readers to my friend Joel's discussion:
This spurred in me the idea of reworking the U.S. county data along economic strata. That's generally the way we tease out Simpson's paradoxes. While I suspect that I already largely know what the results will look like, it's important to do the work to see.
An unexpected second part of this story emerged, and shed a lot of light on what is going on.