Nate Cohn

Nate Cohn @Nate_Cohn

RT @Redistrict: Which brings us to our next @CookPolitical rating change... https://t.co/LJkI9sPjnA — PolitiTweet.org

Nate Cohn @Nate_Cohn

In North Carolina, we also had our first 8 voters who said they have already voted by mail (though on further questioning, they hadn't all mailed them back--which is something to keep in mind if you want to use this in your LV screen, fellow pollsters) — PolitiTweet.org

Nate Cohn @Nate_Cohn

Most of the other national polls we've gotten have actually looked pretty ok for Trump, though they're both fewer in number and worse in quality than the recent state polling, which has generally looked quite bad for Trump — PolitiTweet.org

Nate Cohn @Nate_Cohn

With that in mind, the Data For Progress and Navigator polls showing Biden+11 are nice data points, consistent with or better than their prior results for Biden nationwide — PolitiTweet.org

Nate Cohn @Nate_Cohn

I love all the state polls we're getting, though one unintended consequence of so few national polls is that we have lost our grip on the trendline: there aren't many state polls with a long track record in the same geography, making it harder to see if things are changing — PolitiTweet.org

Nate Cohn @Nate_Cohn

@kwcollins does not include leaners — PolitiTweet.org

Nate Cohn @Nate_Cohn

I'll note that the president's party has often enjoyed a bit of party ID bump in prior reelection campaigns. Granted, those candidates generally led! But it's possible the underlying mechanism--attachment to the president leading to partisan realignment--is similar — PolitiTweet.org

Nate Cohn @Nate_Cohn

I'll posit one simple hypothesis: as we move into year 4 of Trump, the Trump vote (and esp wwc Trump vote) increasingly identifies as Republican. Democrats, OTOH, have morphed into a broad anti-Trump coalition with less identity-based attachment than they've had — PolitiTweet.org

Nate Cohn @Nate_Cohn

It also comes as Republicans can point to some genuine strength in voter registration trends in states like PA and FL. There are a lot of factors at play there, between switching DINOs, new registrants, and potential declines in Dem new reg due to COVID. But it's interesting — PolitiTweet.org

Nate Cohn @Nate_Cohn

One thing that's interesting in this wave of polling: the GOP has the party ID edge in every state, including in places with a significant Dem registration advantage and where we've found Dems doing better in party ID in the past. — PolitiTweet.org

Nate Cohn @Nate_Cohn

Doesn't mean we're wrong, and we're not exactly alone there, either. At least three other firms have a 9-10 pt lead for Biden. And the trendline--a shift toward Biden since Summer--could be noteworthy along with similar shifts in YouGov, OHPI as well. — PolitiTweet.org

Nate Cohn @Nate_Cohn

The next biggest thing that stands out to me is our Arizona number. Still a chance it's just noise, but it seems like we might have a bit of a Biden house effect in AZ? I feel like it's been better for Biden than average in all three of our statewide polls there. — PolitiTweet.org

Nate Cohn @Nate_Cohn

That said, Maine has a lot of ingredients for a Biden surge. It's one of the oldest, whitest states in the country--two groups where he's excelled relative to Clinton in national and state polls. It voted for Obama by 15, so there are plenty of swingy voters to get back here. — PolitiTweet.org

Nate Cohn @Nate_Cohn

The Maine ballot tests used ranked choice, and it generally padded the Democratic lead by a point: Maine Senate initial preference: Gideon 44 Collins 40 Lind 2 Savage 2 *deleted a prior version of this tweet that reversed Collins/Gideon. my apologies — PolitiTweet.org

Nate Cohn @Nate_Cohn

For me, the biggest thing that stands out is the Maine result. It's a huge swing over '16. That said, it's not really alone here: the last two Quinnipiac polls were Biden+15/21; the Hart/Fabrizio/AARP poll in ME was Biden+14, right after the RNC. — PolitiTweet.org

Nate Cohn @Nate_Cohn

The Maine ballot tests used ranked choice, and it generally padded the Democratic lead by a point: Maine Senate initial preference: Collins 44 Gideon 40 Lind 2 Savage 2 — PolitiTweet.org

Nate Cohn @Nate_Cohn

Biden leads by 2 points in ME-02, 47 to 45. In the U.S. House race there, Golden (D) leads Crafts (R) by 19, 56 to 37. The ME-02 House race is considered a tossup by most rating outlets, though that may not last long. — PolitiTweet.org

Nate Cohn @Nate_Cohn

Democrats hold a modest lead in several closely contested races for U.S. Senate, as Biden leads as well AZ Sen.: Kelly+8, 50-42 AZ Pres: Biden+9, 50-41 ME Sen: Gideon+5, 49-44 ME Pres: Biden+17, 55-37 NC Sen: Cunningham+5, 42-37 NC Pres: Biden+1, 45-44 https://t.co/2WLjpoF49N — PolitiTweet.org

Nate Cohn @Nate_Cohn

Democrats hold a modest lead in several closely contested races for U.S. Senate, with Joe Biden ahead as well Ariz.: Kelly+8, 50-42; Biden+9, 50-41 Maine: Gideon+5, 49-44; Biden+17, 55-37 N.C.: Cunningham+5, 42-37; Biden+1, 45-44 — PolitiTweet.org

Nate Cohn @Nate_Cohn

As it is, we usually do weight on some kind of interaction between party reg and *something,* whether it's race, region, primary vote. That's actually useful. Partyx16 would basically just be to clean up my interactions (ba dum ba) — PolitiTweet.org

Nate Cohn @Nate_Cohn

Maybe we'll start weighting on 2016 vote x party reg, just so that there are fewer excuses to complain about the crosstabs — PolitiTweet.org

Nate Cohn @Nate_Cohn

So while in our last poll, we had a pretty reg. GOP set of voters and a super reg. Dem set of nonvoters, this time we have the opposite. But either way, the partisan composition of LVs is the same at R + >4 — PolitiTweet.org

Nate Cohn @Nate_Cohn

Cutting in the other direction, just to keep you on your toes: a Clinton edge in self-reported vote in Arizona! Our last poll there had a self-reported edge of Trump+12. (this is mainly due to changes, clearly just noise, in the party reg. of 2016 voters v nonvoters) — PolitiTweet.org

Nate Cohn @Nate_Cohn

Here's an interesting tidbit: a GOP party ID edge in all three states (something we've never had in many previous NC polls, and way better for GOP in ME02 than our final 2018 poll) — PolitiTweet.org

Nate Cohn @Nate_Cohn

And we did full RCV up and down the ballot, which was interesting. More on that later. https://t.co/XdH1sdgvPX — PolitiTweet.org

Nate Cohn @Nate_Cohn

And this is properly stratified and weighted by CD, so it's not like looking at a crosstab; it is equivalent to an ME-2 poll. — PolitiTweet.org

Nate Cohn @Nate_Cohn

We also have an N=440 oversample in Me-02, including the House race there — PolitiTweet.org

Nate Cohn @Nate_Cohn

Today's polls show Biden with a narrow lead in the Sun Belt. Tomorrow's polls: NYT/Siena in Arizona, Maine and North Carolina https://t.co/5xtdrqcEKm — PolitiTweet.org

Nate Cohn @Nate_Cohn

@NateSilver538 thank you, too kind! — PolitiTweet.org

Nate Cohn @Nate_Cohn

@RachelBitecofer @cityofprogress @TravisKniffin @Redistrict loll incredible! i thought there was a 99.95 percent chance of a magnificent, reality tv-style performance and i was right yet again! (as an aside: i have made no prediction, let alone for a white working class revolution) — PolitiTweet.org