This is a dumb, dumb, dumb, dumb analysis.
YOU CANNOT USE COUNTY-LEVEL DATA FOR THIS, because the largest counties contain BOTH the richest AND the poorest places in the nation. nytimes.com/interactive/20…
YOU CANNOT USE COUNTY-LEVEL DATA FOR THIS, because the largest counties contain BOTH the richest AND the poorest places in the nation. nytimes.com/interactive/20…
For example, the notoriously poor and left-behind city of Baltimore has a higher median income than all but three counties in West Virginia. That doesn't mean many Baltimoreans aren't living in dire poverty, it means their incomes are being averaged with wealthy urbanites.
All this analysis is actually showing is that Republican counties are smaller and more economically homogeneous. Because they're smaller, you can pool together all the poor, R-voting areas. But all the poor, D-voting areas are part of larger, indivisible county units.
Seriously, this analysis is amateurish and actively misleading, and it's all in service of the same old dull, false cliche about the Trump-voting Forgotten Man of the Rust Belt. Please don't do this.
To give you a sense of the problem, the first map is all the counties richer than Los Angeles County. The second map is all the counties richer than the Florence neighborhood of South LA. 
