So on the weekend thread I was contemplating the house districts, states, and open senate seats that were the opposite of the variables that the Upshot correlated with Trump voters.
I took a look at four of them this morning, and I think its simpler than that.
I made four lists states with a higher percentage of 1968 Wallace voters than the national average, states with less than 5% voting for John Anderson in 1980, and states with more than 10% American ancestry in their 2000 census forms. There were 14 states on each list. I also copied the top 10 states for mobile home usage linked to by ChadmanFL.
11 states were on at least three of the lists 7 on all 4 (AL, AR, KY, LA, MS, NC, SC); 3 missing mobile home residents (GA, OK, TN); and WV, which was not on the Wallace list.
I don't need some sort of statistical analysis to say what I think is an obvious answer: Trump will underperform Romney less in Southern/Appalachian states that voted for Romney in 2012.
The good news for Democrats is that, if you look at the list of competitive seats from the DKE ratings, theres only one seat (WV-02) that falls in this list of states. In the Senate, there are 10 seats the Republicans are defending that are Likely R in the DKE ratings, and only one (NC) is on this list of states.
This may not be just a wave election, it could be a tsunami.