Don't worry, surely some statistical prediction machine will easily replace them. You just need an intern to speak english to a computer, right?

How is that what OP is implying?

It's likely buried in many appendices. Made one such analysis for an appendix recently and it's mostly this. Less issue-specific knowledge and interest, less formal education and, of course, the good old gender gap.

It’s been a very helpful paper for me personally. Part of me still thinks that if you don’t have a somewhat credible causal design, you just shouldn’t go the quantitative route at all, but this approach is a good middle ground.

Nice but how about these 10 different methods from machine learning instead? 🙃

It might be my own bias, but I think many of the individual-level dynamics of the rise of the far right are strikingly similar across Western countries.