Maybe it also helps if I make clear what I’m advocating for. When we do what we now call association studies, we should instead be clear that our goal is to generate evidence for a causal effect. We carefully state our estimand, are explicit about our causal assumptions and…

I was going to say the same thing but about Houston. My prior is that Houston is capturing the large concentration of Black people in the mostly red Southern states. Would be interesting to see how names vary by race and partisanship or geography.

Finally, the last 3!