Nate Cohn

Nate Cohn @Nate_Cohn

@PoliticalKiwi yup — PolitiTweet.org

Nate Cohn @Nate_Cohn

There's also a bunch of new, cheap online state polls with, at best, no track record. So far, they tilt to the left--both in 2020 and in a comparable 2018 effort. So that's a little troubling, though they're easy enough to ignore if you're disciplined — PolitiTweet.org

Nate Cohn @Nate_Cohn

Education weighting still a big deal, and half of state polls still aren't doing it. Times/Siena polls from last October were Biden+6 (v. Biden+2 IRL) without education weighting: PA Biden+8 (+3) WI Biden+6 (+3) MI Biden+4 (0) AZ Biden+6 (+5) FL Biden+6 (+2) NC Biden+2 (Trump+2) — PolitiTweet.org

Nate Cohn @Nate_Cohn

Are the state polls better than they were four years ago? Maybe, maybe not https://t.co/4tR3XZMboR — PolitiTweet.org

Nate Cohn @Nate_Cohn

And yes the poll is anonymous. I can't see any information on the respondent — PolitiTweet.org

Nate Cohn @Nate_Cohn

Take my 10 question poll! It's not a scientific poll, of course, but it has a purpose Anyone's eligible, feel free to share https://t.co/Xttm3HyfRZ — PolitiTweet.org

Nate Cohn @Nate_Cohn

Take this 10 question poll! It's unscientific, yes, but it does have a purpose. Feel free to share https://t.co/DE2rgEjVRy — PolitiTweet.org

Nate Cohn @Nate_Cohn

The Emerson poll is a bit better, to be fair. It includes more Texans than my Twitter feed. The landline sample is probably better. But like, fundamentally idk if it really counts as a 'poll,' even if the result is plausible enough because I've weighted it to Trump+9 — PolitiTweet.org

Nate Cohn @Nate_Cohn

Let's suppose that right now I polled my Twitter feed on its 2020 vote preference, 2016 recalled vote, and some demographics. Let's say I then it to '16 results in TX (trump+9) and TX demographics. I don't really believe that's fundamentally much different than an Emerson poll. — PolitiTweet.org

Nate Cohn @Nate_Cohn

Let's suppose that right now, I polled my Twitter feed on its 2020 vote preference, 2016 recalled vote, and some demographics. Let's say I then it to '16 results in TX (trump+9) and TX demographics. I don't really believe that's fundamentally any different than an Emerson poll. — PolitiTweet.org

Nate Cohn @Nate_Cohn

smh stop these polls are not worth your time https://t.co/HaYc3y9Jom — PolitiTweet.org

Nate Cohn @Nate_Cohn

Folks, a majority of the respondents to this Emerson poll of Texas were reached on MTurk. Let's not fill up my Twitter feed with further discussion of it — PolitiTweet.org

Nate Cohn @Nate_Cohn

RT @MULawPoll: New @MULawPoll will be released tomorrow, May 12. We will post a video conversation of @MikeGousha and @PollsAndVotes at 1… — PolitiTweet.org

Nate Cohn @Nate_Cohn

RT @jmartNYT: “The F.B.I. and the Department of Homeland Security are preparing to issue a warning that China’s most skilled hackers and sp… — PolitiTweet.org

Nate Cohn @Nate_Cohn

RT @isaiahthomas: Bring back our Sonics!!! — PolitiTweet.org

Nate Cohn @Nate_Cohn

RT @maggieNYT: Network the president has encouraged folks to watch repeatedly during the pandemic > — PolitiTweet.org

Nate Cohn @Nate_Cohn

RT @tackettdc: "The decision to shelve detailed advice from the nation’s top disease control experts for reopening communities during the c… — PolitiTweet.org

Nate Cohn @Nate_Cohn

RT @AlecMacGillis: Hard to overstate how much Germany has surpassed just about everyone. It has a quarter of the deaths of other major Euro… — PolitiTweet.org

Nate Cohn @Nate_Cohn

RT @skoczela: Our new poll for @WBUR of Mass. front line nurses is now posted. Stories throughout the day. On personal protective equipme… — PolitiTweet.org

Nate Cohn @Nate_Cohn

@PollsAndVotes do those post-election jumps represent same-day registrants or pre-election registration? — PolitiTweet.org

Nate Cohn @Nate_Cohn

RT @clairecm: We wondered whether lockdowns, with all the extra housework and child care, meant men were doing more of it -- or if women we… — PolitiTweet.org

Nate Cohn @Nate_Cohn

Monmouth shows Biden up by 9, 50 to 41, adding to a growing body of state polls suggesting that Biden's lead has been growing over the last few weeks https://t.co/t2gHWE5GG7 — PolitiTweet.org

Nate Cohn @Nate_Cohn

Monmouth gives Biden a 9 point lead, 50-41--adding to a growing body of state polls suggesting that Biden's lead has been growing over the last few weeks some confirmation to the idea that Biden's lead https://t.co/DdMtC2PxZs — PolitiTweet.org

Nate Cohn @Nate_Cohn

RT @nickconfessore: The NYT now has 6 million subscribers. This support means more to us than ever, as the pandemic kills what's left of a… — PolitiTweet.org

Nate Cohn @Nate_Cohn

It's been a long time since we've had a live interview, high-quality national poll. The state polls sure seem like they've drifted toward Biden over the last three weeks; will be interesting to see if there's confirmation or cold water here https://t.co/5ZYEr1SJcv — PolitiTweet.org

Nate Cohn @Nate_Cohn

This isn't a state-only poll challenge. But there could easily be state-level variance in the way this plays out, and I'd guess that it's easier to conceive of some relatively large outcomes in a state or two even as the nation overall seems relative normal — PolitiTweet.org

Nate Cohn @Nate_Cohn

Another is the big question about COVID turnout. I think turnout's always a challenge, but there's no way around the possibility that the pandemic has some big and tough to predict consequences for both the level and pattern of turnout — PolitiTweet.org

Nate Cohn @Nate_Cohn

Another is the growing use of recalled 2016 vote to ensure that mediocre samples wind up in the ballpark. This is a fine way to reduce error--i mean, you could probably get in the ballpark by weighting a sample of NYC to clinton/trump nationally. But bias is a very real risk here — PolitiTweet.org

Nate Cohn @Nate_Cohn

One is the explosion of new, online state polls. There were fairly rare in 16/18, making up like 10% of state pollsters. Over the last month, they're half of the pollsters. This is a pretty easy risk to guard against, but there is some risk here — PolitiTweet.org

Nate Cohn @Nate_Cohn

Three years ago, I wrote this summary of state polling issues in '16. I think if you go down the list, you'd probably conclude that most of them are less serious--in some way or another--than they were in '16. There are some new issues to mull, though https://t.co/HzwTpfq75Y — PolitiTweet.org