IA-Selzer: Trump +7

[Fmr. Gov. NickG]

I’m not really getting the whole “this is Trump’s best poll of the whole cycle” claim.

Trump has had several better polls in Florida than this one.  In particular, there was a WaPo poll in late September showing Trump +4 among LVs.  There have been multiple high quality polls showing the race tied or +1 either way in that state.

This poll is a much smaller and less pivotal state, and still shows a swing toward Biden.  So I don’t understand what the big deal is.

[Fmr. Gov. NickG]

I don’t understand why you’d ever weight the result of a presidential poll by congressional district.  It’s not meaningful division of the populations in this context.  If you want to make sure you have a geographically representative sample, weight by metro area or media market or something similar.  If you want to use the same poll for congressional results, stratify the sample by CD.  Weighting by CD shouldn’t bias the result in any way, but it will slightly increase the margin of error.

[Fmr. Gov. NickG]

I’m not really getting the whole “this is Trump’s best poll of the whole cycle” claim.

Trump has had several better polls in Florida than this one.  In particular, there was a WaPo poll in late September showing Trump +4 among LVs.  There have been multiple high quality polls showing the race tied or +1 either way in that state.

This poll is a much smaller and less pivotal state, and still shows a swing toward Biden.  So I don’t understand what the big deal is.

Biggest reason I say it might be is that it’s coming so late, and echoes a similar poll from Selzer in 2016. Florida has been pretty all over the place this whole time too, but Iowa was looking pretty consistent across the board recently in showing a tight race that tilted Biden if anything until now.

Well now WaPo has Trump +2 just as late.  That poll seems much better for Trump than this one.