Very interesting. I actually read this as a more nuanced take on the Vox article (which I'd read more as saying "it's not economic anxiety" rather than "it's this other thing than economic anxiety"). The verdict seems to be, we cannot tell whether it's economic anxiety or something else. Anyway, interesting blog, which I've now added to my subscriptions, so thanks for that!
The very same Gallup study which that Vox article relied on to argue that Trump support was correlated with higher, not lower income also found that blue collar and less educated workers were disproportionately likely to vote for Trump, and Vox intentionally omitted that fact. This is particularly dishonest of Vox because the result they did quote was after correcting for the fact that working-class status and education are correlated with income, and they didn't mention that either.
That is, any correlation with income that appeared to be better explained by both being correlated with being working-class or anything else was excluded, and worse still there's no guarantee that the attribution to a particular category (say, class instead of income) is the result of anything other than random chance.