Nate Cohn

Nate Cohn @Nate_Cohn

@billscher (and while i didn't check on this poll, historically in our data whites who won't say they're white are very R) — PolitiTweet.org

Nate Cohn @Nate_Cohn

@billscher similar pattern in 2020. interestingly, the people who are registered as black but who told us something else (whether it's 'dk', some other race, or hispanic-black) are still like 85-10 D so i don't worry about it — PolitiTweet.org

Nate Cohn @Nate_Cohn

@billscher the black share of the electorate is >27% using the state records. it's 24% who then told us they were black in the IRL interview — PolitiTweet.org

Nate Cohn @Nate_Cohn

Have response rates shifted? Or is this related to the same 'oh weird Dems doing so bad in blue states' thing, where there's just some pretty extreme variation in Democratic enthusiasm that happens to be cancelling out nationwide? Something to check if I get a chance. — PolitiTweet.org

Nate Cohn @Nate_Cohn

One thing I find especially interesting is the disparity between the national and state response rates, even the national and district polls just a week apart. Nonresponse seemed pretty in-check -- and nothing like 2020 -- in all three of our national polls. — PolitiTweet.org

Nate Cohn @Nate_Cohn

A Worrisome Pattern Re-emerges in Seeking Response From Republicans https://t.co/NOgaNqT9cp — PolitiTweet.org

Nate Cohn @Nate_Cohn

I was pretty paranoid about this and there’s definitely room for confusion. The voter file we used for polling still characterizes the 2010 CDs as the districts, with 2020 as “proposed.” — PolitiTweet.org

Nate Cohn @Nate_Cohn

And I don't really see any evidence that the 'cheap' polls have fixed the Trump nonresponse problem, either. Just as an example: the last Trafalgar poll in PA had Fetterman leading among whites and doing the same in-and-out of the Philadelphia DMA (ie: leading outside of Philly) — PolitiTweet.org

Nate Cohn @Nate_Cohn

The cheap polls flooding the market right now... do not have good and representative samples. They could easily come closer to the final results! But not because they've figured out how to sample the population especially well https://t.co/OdNvk8pPxy — PolitiTweet.org

Nate Cohn @Nate_Cohn

White reg. Dems were 28% more likely to respond than white reg. Reps. That's nothing like our national polling and quite a bit like 2020, when party quota/weighting was insufficient. That's a troubling sign for nonresponse bias https://t.co/pOmfH1hLBX — PolitiTweet.org

Nate Cohn @Nate_Cohn

@LPDonovan notice the generic ballot too — PolitiTweet.org

Nate Cohn @Nate_Cohn

And worth flagging that, unfortunately, this PA survey was largely conducted before the debate last week. So there's certainly a chance there's been movement since — PolitiTweet.org

Nate Cohn @Nate_Cohn

To me, the most interesting dimension of the poll: Dems running an avg of 8 points ahead of Senate control preference (R+4 on average). Illustrates key dynamic of the race -- a favorable environment for Rs v. bad candidates -- and helps square with the national picture — PolitiTweet.org

Nate Cohn @Nate_Cohn

Newsletter subscribers will learn a little bit about the asterisk and hear a few musings about the the gap between these polls and the averages, which does seem driven by a flood of partisan/IVR polls (right or wrong) https://t.co/vfLG2cOZH5 — PolitiTweet.org

Nate Cohn @Nate_Cohn

Our final* polls: the Senate is close, with some decent results for Democrats https://t.co/mEEUngwujY — PolitiTweet.org

Nate Cohn @Nate_Cohn

RT @ShaneGoldmacher: NEW: Perhaps no place in the nation offers a more symbolic and consequential test of whether Democrats can claw back s… — PolitiTweet.org

Nate Cohn @Nate_Cohn

@davidshor @billscher well, try to — PolitiTweet.org

Nate Cohn @Nate_Cohn

RT @UrsulaPerano: Dressed as the most terrifying thing I could think of ahead of Election Day: the NYT needle Happy Halloween, nerds https… — PolitiTweet.org

Nate Cohn @Nate_Cohn

@SeanTrende i think that's fair, but rightly or wrongly i can't say that was my prior coming out of 2020 given the strength of the pres ~ house relationship — PolitiTweet.org

Nate Cohn @Nate_Cohn

@Redistrict (I say after '20 bc incumbents really struggled to outrun Trump/Biden in '20. A guy like Cartwright would have gone down in '20 in an R+4 environment. Maybe he will in the end, but the polling doesn't seem to show such a tight relationship between pres and ballot so far) — PolitiTweet.org

Nate Cohn @Nate_Cohn

(I say after '20 bc incumbents really struggled to outrun Trump/Biden in '20. A guy like Cartwright would have gone down in '20 in an R+4 environment. Maybe he will in the end, but the polling doesn't seem to show such a tight relationship between pres and ballot this time) — PolitiTweet.org

Nate Cohn @Nate_Cohn

An additional layer--which @Redistrict may disagree w, DK--is that it does seem like, in general, the incumbents are holding up better than I would have guessed after '20 A lot of the D weakness is open seat https://t.co/OIYM7yTmyK — PolitiTweet.org

Nate Cohn @Nate_Cohn

@tbonier @NateSilver538 because my baseline isn't just the 2020 result. our reporting is that cartwright/davids are both modestly ahead. — PolitiTweet.org

Nate Cohn @Nate_Cohn

@SpecialPuppy1 @NateSilver538 maybe +4. it's not my ex ante expectation but everyone seems to think cartwright is holding on — PolitiTweet.org

Nate Cohn @Nate_Cohn

@TomScherschel @NateSilver538 no, i don't think that's the headline? if that was the headline it was changed to that and changed back before i saw it (the current headline is the one i saw pre-pub) — PolitiTweet.org

Nate Cohn @Nate_Cohn

@tbonier @NateSilver538 i think they're consistent with a GOP takeover -- which is all the story notes. as i've said above, i do think they're better than i expected based on our last poll. i don't think KS3 is consistent with a GOP takeover, but i think it's an outlier — PolitiTweet.org

Nate Cohn @Nate_Cohn

@NateSilver538 now if you want to be clever and say that the seahwks point differential over all four games was net-positive, go for it. we just happen to have other data that suggests it's an outlier -- maybe josh allen was injured in that one blowout game — PolitiTweet.org

Nate Cohn @Nate_Cohn

@NateSilver538 is it fresh evidence that they're not favored to win the super bowl? on balance, yeah, i think so. — PolitiTweet.org

Nate Cohn @Nate_Cohn

@NateSilver538 if the seahawks go 1-3 v. the bills, but did a bit better than expected and win one game in a blowout, i think 'seahawks do better than expected' is a perfectly great analytical take (including in my newsletter)! — PolitiTweet.org

Nate Cohn @Nate_Cohn

@NateSilver538 'good for democrats' != 'hold the house' i do think they're good for democrats v. expectations, as this thread notes up high and as i wrote in the newsletter. but that's... not the necessarily the most important metric for judging this — PolitiTweet.org