Nate Cohn

Nate Cohn @Nate_Cohn

So I think this has some potential to explain some things. If this is generalizable (it may not be), over-representation of more regular voters would tend to overestimate Dems in the Midwest and underestimate them in the Sun Belt (prior version of this tweet was backward/deleted) — PolitiTweet.org

Nate Cohn @Nate_Cohn

Anyway, I think the tendency for high-turnout/highly educated voters to respond to polls helps explain some of the geographic patterns we see in poll bias, and I wonder whether the particular enthusiasm of the 2018-surge resistance vote may be a particular challenge right now — PolitiTweet.org

Nate Cohn @Nate_Cohn

We do account for this both on the front-end and by weighting. It's a major difference between us and other pollsters, and I'd say our more surprising results generally line up with this explanation (though it's messy to try and directly untangle the effect of our choice here) — PolitiTweet.org

Nate Cohn @Nate_Cohn

The pattern is pretty easily explained by demographics: nonvoters in the Midwest are overwhelmingly whites without a degree; in the Sun Belt, they're disproportionately Hispanic — PolitiTweet.org

Nate Cohn @Nate_Cohn

So I think this has some potential to explain some things. If this is generalizable (it may not be), the over-representation of more regular voters would tend to underestimate Dems in Sun Belt (among RVs) and help them in the Midwest — PolitiTweet.org

Nate Cohn @Nate_Cohn

Here's Trump approval in AZ/FL by vote history: Voted 2016 and 2018: 50-49 Voted 2016 not 2018: 44-54 Didn't vote 2016, voted 2018: 44-53 Neither: 40-53 — PolitiTweet.org

Nate Cohn @Nate_Cohn

To my knowledge, no public pollster adjusts for this--either in terms of sampling or weighting. Now, here's Trump approval in PA/WI/MI, by vote history: Voted 2016 and 2018: 47-51 Voted 2016 not 2018: 55-43 Didn't vote 2016, voted 2018: 38-61 Neither: 52-45 — PolitiTweet.org

Nate Cohn @Nate_Cohn

In our polls from last month, here's the completion percentage by vote history. Voted 2016 and 2018: 2.3% Voted 2016 not 2018: 1.3% Didn't vote 2016, voted 2018: 2.3% Didn't vote 2018 or 2016: 1.1% — PolitiTweet.org

Nate Cohn @Nate_Cohn

Here's a pattern that I think may have some consequences for polling in 2020, at least for now: response rate by vote history. — PolitiTweet.org

Nate Cohn @Nate_Cohn

@clairecm There's also a slice of Clinton voters who basically thought Trump was unqualified and risky last time, but have basically come around with the economy doing well and without obvious catastrophe at home or abroad — PolitiTweet.org

Nate Cohn @Nate_Cohn

@clairecm A lot of the explanations, enjoyably, feel fairly consistent with some classic explanations for election outcomes, like economic growth, or intentionally voting for divided government in midterm elections — PolitiTweet.org

Nate Cohn @Nate_Cohn

@clairecm Trump-Dem '18 voters backed Trump over all his Dem rivals in our poll, and it's a big reason why Dems don't match their midterm success in the Rust Belt in our poll (Trump '16, stay home '18 is another big factor in the Rust Belt, though not necessarily nationally) — PolitiTweet.org

Nate Cohn @Nate_Cohn

My colleague @clairecm talked to voters who said they voted Trump in '16, Dem in '18, and now say they'll vote Trump again in 2020 in our Times/Siena polls from last month. They're fascinating interviews. https://t.co/ND9Ma991xf — PolitiTweet.org

Nate Cohn @Nate_Cohn

@WhiskeyFDC i'm not sure whether it is different. perhaps they're top-tier talents as well, or perhaps patrick isn't actually top-tier. but the belief in his talent is different, in that it precedes his decision to run — PolitiTweet.org

Nate Cohn @Nate_Cohn

@WhiskeyFDC no, steyer isn't taken seriously. there are a lot of people who think patrick is a top-tier talent worth consideration, and that would translate into much more news coverage — PolitiTweet.org

Nate Cohn @Nate_Cohn

I do think he really might have jumped up to 5%+ in the polls and qualified to the Dec. debate just off of a week or two of sustained coverage, sort of like a Thompson or Clark in '08/04. This seems like it could be an afterthought with impeachment, tho — PolitiTweet.org

Nate Cohn @Nate_Cohn

And the impeachment hearings are going to get in the way of any effort to capitalize on his announcement with big free media coverage, which he really might have gotten otherwise https://t.co/KUUf2b18Sa — PolitiTweet.org

Nate Cohn @Nate_Cohn

RT @jmartNYT: NEWS: @DevalPatrick told senior Democrats in phone calls today that he's running for president, per two sources. Will start… — PolitiTweet.org

Nate Cohn @Nate_Cohn

RT @samdolnick: 🚨Here we go! The Daily team is launching a new evening podcast to catch you up on impeachment news. It’s called The Latest… — PolitiTweet.org

Nate Cohn @Nate_Cohn

@PatrickRuffini @jon_m_rob we have trump at 76-16 v. biden in precincts where he exceeded 75 percent, so i think we're in the ballpark here at least — PolitiTweet.org

Nate Cohn @Nate_Cohn

It's less than a year to go until the presidential election, and there's no way around concluding that a startling number of state pollsters continue to make the methodological choices that were widely acknowledged as responsible for underestimating Trump in '16 — PolitiTweet.org

Nate Cohn @Nate_Cohn

Meanwhile, Dems lead in the version of Georgia that has more college grads than Silicon Valley https://t.co/kVclMq8evz — PolitiTweet.org

Nate Cohn @Nate_Cohn

@RachelBitecofer there are few to no apples-to-apples comparisons in the battlegrounds, which is why, strictly based on these data, one would think the EC-PC gap could be poised to grow. AZ02 may be closest, shows no shift. And I have no data assistants. — PolitiTweet.org

Nate Cohn @Nate_Cohn

Live look in at the popular vote-electoral college gap (there is no shift in party registration in the top states, even in AZ) https://t.co/DL8gicmwLk — PolitiTweet.org

Nate Cohn @Nate_Cohn

RT @JakeSherman: HILLARY CLINTON, on BBC about 2020: "I, as I say, never, never, never say never,. I will certainly tell you, I’m under eno… — PolitiTweet.org

Nate Cohn @Nate_Cohn

@PollsterPatrick @MonmouthPoll the trendline is relevant too imo — PolitiTweet.org

Nate Cohn @Nate_Cohn

Didn't think it would be long before a result like this after we had Pete jumping to 18 https://t.co/KBy7zD8o30 — PolitiTweet.org

Nate Cohn @Nate_Cohn

our data is inconsistent with those findings, to be clear — PolitiTweet.org

Nate Cohn @Nate_Cohn

So maybe there's a Trump rebellion. Or maybe it's tough to reach rural white voters and, even when they do, may be getting a fairly non-Trump group. Can't prove it either way, but it's a big gap with my priors, fits a persistent bias in polling, and our data is inconsistent — PolitiTweet.org

Nate Cohn @Nate_Cohn

(IA-4 neighbors Nebraska and voted about the same for Trump, though much better for Obama) — PolitiTweet.org