Nate Cohn

Nate Cohn @Nate_Cohn

Biden by 10 in Pennsylvania and 4 in Florida, per new CNN/SSRS poll on TV — PolitiTweet.org

Nate Cohn @Nate_Cohn

@jbreezy1512 One note, given the Monmouth poll today: there is no LV/RV gap in the survey, perhaps in part because we do not give any bonus to having 'already voted' compared to saying you're 'almost certain to vote,' to avoid biasing the survey toward early voters — PolitiTweet.org

Nate Cohn @Nate_Cohn

The big :eyes must go to the Senate races. --Perdue net-favorability rating tanked over the last month, while Ossoff remains at +5 --Warnock has best favs in the survey, and runoff lead isn't due to favorability. He's up 49/45 if you allocate undecideds by presidential vote — PolitiTweet.org

Nate Cohn @Nate_Cohn

I should note, btw, that this thread is merely describing why advanced-vote based estimates have panned out well to date. There's no guarantee these conditions hold in the future, including this year. I don't *think* a D+4 electorate in NV will vote T, but we've it before — PolitiTweet.org

Nate Cohn @Nate_Cohn

@ElectProject is this a joke? — PolitiTweet.org

Nate Cohn @Nate_Cohn

RT @dparkermontana: .@Nate_Cohn .@baseballot .@LarrySabato .@gelliottmorris We have a poll out in Montana. Results can be found here, as w… — PolitiTweet.org

Nate Cohn @Nate_Cohn

@mikes2start yeah probably not a fun job rn. — PolitiTweet.org

Nate Cohn @Nate_Cohn

South Carolina, on the other hand, is turning out to have exceptionally productivity--Michigan-level bad. It's very strange how some of these files are just dead. — PolitiTweet.org

Nate Cohn @Nate_Cohn

Biden+12 in ABC/Post, 54-42 It's not outlier: that's basically the post-Trump COVID diagnosis average of national surveys at this points. https://t.co/9khete83F6 — PolitiTweet.org

Nate Cohn @Nate_Cohn

Something I didn't realize: Elyse Galvin (D) has led Don Young (R) in led every poll of AK-AL. They're mainly Dem/progressive group sponsored, but still kind of interesting https://t.co/bh38WuAvTx — PolitiTweet.org

Nate Cohn @Nate_Cohn

The economy is the *one* issue that voters think the president has handled really well, and here they forfeit a golden opportunity to note only turn the conversation back to it, but add a popular bipartisan accomplishment. It's wild, at least from a political standpoint — PolitiTweet.org

Nate Cohn @Nate_Cohn

RT @FoxReports: Rick Scott says he has tested positive in interview on Fox News. “I was tested yesterday I think for the 6th time and I t… — PolitiTweet.org

Nate Cohn @Nate_Cohn

RT @dfriedman33: There were just 32 Covid cases reported in DC Thursday. Trump and Melania aren’t part of that. But the scale shows that th… — PolitiTweet.org

Nate Cohn @Nate_Cohn

And surprisingly, the section about non-condemning white supremacists didn't make the segment at all — PolitiTweet.org

Nate Cohn @Nate_Cohn

@HotlineJosh yeah, though that's not a low turnout special — PolitiTweet.org

Nate Cohn @Nate_Cohn

In TX and Florida, their pro-GOP House effect among Hispanic voters has canceled it out. But I couldn't come up with any reason why this poll wasn't going to be for Biden — PolitiTweet.org

Nate Cohn @Nate_Cohn

@tysonbrody a dem senate would probably be ignoring it tho — PolitiTweet.org

Nate Cohn @Nate_Cohn

Morning Consult has rehashed their 2016 shy Trump study and find serious evidence of social desirability bias, though weighting on self-reported 2016 past vote sure seems like an issue for this particular experiment? https://t.co/bmWthNwKsV — PolitiTweet.org

Nate Cohn @Nate_Cohn

It is plausible on paper, but pbly not if the decisive Biden gains in Texas are coming from suburban whites. Maybe that's not what's going on here--maybe it's new reg.--but if it is, I don't see how Biden doesn't also get *enough* padding in northern burbs to survive there — PolitiTweet.org

Nate Cohn @Nate_Cohn

Obviously it sucks to get something like this wrong, but it wound up pretty interesting to scrutinize this tiny group of respondents. Their own explanations are interesting, too. — 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

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

Biden by 3 in NC and 10 in MN, per CNN/SSRS. Nothing for Biden to complain about here. https://t.co/tnBWsbL4Xt — PolitiTweet.org

Nate Cohn @Nate_Cohn

So far, uncertainty debates has been the key front of the 2020 model wars. And I have to say that I'm pretty sympathetic to the 538 side of the case there. But the fundamentals play a big role in the models as well, and I'm a little farther from 538 than I thought I was on this — PolitiTweet.org

Nate Cohn @Nate_Cohn

Now, there are caveats. In the COVID area, we've always go to be extra cautious about vote method. Maybe this is all election day vote, and maybe the mail vote isn't counted and it'll be good for Kennedy--just making up an example. — PolitiTweet.org

Nate Cohn @Nate_Cohn

The real issue, though, is that they just haven't said how they make these decisions, and it is pretty obvious that there's not a long of consistency, whether in terms of the cutoff dates or the qualifying pollsters. It is hard to take seriously as a result — PolitiTweet.org

Nate Cohn @Nate_Cohn

It's also possible there's something in between: the decided a long time ago that the online polls need to be judged on a case by case basis, because of varying quality, and the media sponsored ones are presumptively in, while bias weighs on which ones are out — PolitiTweet.org

Nate Cohn @Nate_Cohn

RT @davidshor: @jon_m_rob *whispers* "73% of 18-34 year olds who didn't vote in 2016 identify as moderate or conservative" — PolitiTweet.org

Nate Cohn @Nate_Cohn

RT @ForecasterEnten: On TV, not everything is as it appears to be... https://t.co/JJIAR4aOlh — PolitiTweet.org

Nate Cohn @Nate_Cohn

One major reason is a change in the composition of the recent polls. There's only one live interview survey among the last *50* recorded by 538. — PolitiTweet.org