The cool thing is, if he's right (& the polls appear to be confirming it so far) it means the polling is busted in exactly the way people suspected it was, & the "historic realignment" stuff is not necessarily balderdash, but it's at least partly an artifact of partisan nonresponse
Which way do people suspect polling is busted? I’d like to think it overestimates Trump’s support, but polling in 2020 actually underestimated it. Midterm 2022 polling was closer than most years.
Not sure! Probably a little in Biden's favor but the degree is basically impossible to tell. But "Trump making huge gains with nonwhite voters & voters under 35" seems a lot less likely than "those demographics being particularly unwilling to respond to pollsters unless they're sickos"
Yeah, I’m very suspicious of the polls showing a close race with young people. Though some show 3rd party candidates doing well with young people and if RFK Jr (YGTBFKM) is pulling more of them from Biden than from Trump?
Also some polls have shown Biden ahead in the 65+ age group. Huh.
I can see seniors being for Biden if AARP has been writing about social security and Medicare, idk?
Plus those on the Gen X cusp are aging into retirement, so that should move it somewhat left, but 50-64 was for trump iirc…ahh they vote dem in midterms, so republicans aren’t motivated…excellent
I remember polls were at least busted enough for pundits to say D women had moved on from dobbs by Nov 2022 and I never saw much attempt to reconcile that
Pollsters usually over sample those groups and then correct for it mathematically when they aggregate. That’s the part that Nate Silver is *somewhat* good at explaining.
But click through to see the survey questions, which have a huge impact, and read the methodology statement by the polling firm.
Which is to say that I am also extremely sus at those polls, and trump being up by some insane amount, like +15 in Nevada, and it may very well be a math error, but it’s probably not a sampling error.
