KY-AARP: Trump +12

[Skill and Chance]

Folks need to seriously worry about Trump voters hanging up on the polls.  Every time I see one of these in a blowout Trump 2016 state, it makes me suspect that all those Biden +5-10 margins we keep seeing in the major Midwest swing states are skewed several % by nonresponse bias. 

[Skill and Chance]

Folks need to seriously worry about Trump voters hanging up on the polls.  Every time I see one of these in a blowout Trump 2016 state, it makes me suspect that all those Biden +5-10 margins we keep seeing in the major Midwest swing states are skewed several % by nonresponse bias. 

This is a common talking point, but there's little to no evidence of an actual bias to that effect.  See for example this 2017 article from Pew Research.

It's an issue primarily at the state level.  At the nationwide presidential PV/GCB level, it appears to be cancelled out by people who speak Spanish at home not answering in the Southwest.  But we've had polling consistently underestimate Republicans in most of the Midwest/Florida and consistently underestimate Democrats in TX/NM/AZ/NV/CA/arguably CO during the Trump era.  I mean, Cruz was up 8 in the RCP average and Clinton was underestimated in CA by more than she was overestimated in the key Midwest states. 

[Skill and Chance]

It actually would not surprise me all that much if Biden got 41% of the vote in Kentucky; that's how much Obama got in 2008, and Biden is a considerably better fit for the state even if it has only trended right since then.

Trump wouldn't only get 53% of the vote, but I wouldn't be too shocked if Biden broke 40. A 57-41 margin like McCain had would make some sense, especially if it does turn out to be another strong year for Dems nationally.

There's a plausible scenario in which Biden improves a point or 2 on Clinton, but even that's unlikely.