AP-NORC: Biden +4%

[TiltsAreUnderrated]

https://apnorc.org/wp-content/uploads/2020/09/topline_september_campaign.pdf

September 11-14
1108 adults
MoE: 4%
Changes with July 16-20 AP-NORC poll.

Biden 44% (-2)
Trump 40% (+4)
Another candidate 7% (-3)
Probably not vote  7% (-1)
Skipped/refused 1% (n/c)
Don't know 0% (but some voters) (n/c)

[Trump v. Wong Kim Ark]

7% wouldn’t vote and another 7% say someone else? Throw it in the trash.

[Fmr. Gov. NickG]

7% wouldn’t vote and another 7% say someone else? Throw it in the trash.

It appears to be a poll of all adults so this isn't that suprising.  If people were being honest, there'd be a lot more "not voting" responses.

[GeorgiaModerate]

Something I noted in the Trump approval thread about this poll:

AP/NORC, Sep. 11-14, 1108 adults (1-month change)

Approve 43 (+8)
Disapprove 56 (-9)

Strongly approve 25 (+7)
Strongly disapprove 44 (-2)

When I see an eye-popping change like that, it's time to take a deeper look.  Let's check the party ID breakdown in this poll:

D 43
R 43
I 15

That's too Republican, since more people in the country identify as D than R.  What was it last month?

D 48
R 35
I 16

OTOH, that one is too Democratic; the D-R gap isn't that wide. 

So they went from a too-D to a too-R sample.  It's not surprising that there was a huge change in approval.

Another way to frame it: they went from D+13 to D+0 samples, which is an R+13 change, and saw an R+17 change in approval.  The other four points are likely to be real movement (Trump's approval has ticked up in other polls recently), or it could be normal variation, or some of both.

[Gass3268]

RCP is figuring how fast they can finally add these polls to their average.

[หมูเด้ง]

RCP is figuring how fast they can finally add these polls to their average.

They didn't with the last one. Anyways, the main takeaway from this poll is that if Republicans get to parity with Democrats in the exit polls, only then will Trump PROBABLY win.

[Alben Barkley]

I don’t even really care about the national polls anymore, although this one is clearly wonky.

Recent state polls by Civiqs, NYT/Siena, ABC/WaPo, and Quinnipiac all are great for Biden. Hell, even Rasmussen has him doing better in Ohio and other states than nationally. So whatever weirdness may or may not be going on in the national polls (which mostly seems to just be an onslaught of low quality polls more than anything real), it doesn’t seem to be reflected at all in the state polls. And if I could pick one to be better for Biden, it would definitely be the state polls.

[CellarDoor]

Seems silly that given the polarization in this country, we'd see a poll this close to the election where the two major candidates only make up only 84% of the vote, but the incumbent being at only 40% is the most telling thing here.

[🐒Gods of Prosperity🔱🐲💸]

removing "probably not vote," that's

Biden 47%
Trump 43%
Other candidate  8%

[GeneralMacArthur]

I wonder why 13% of the country decided to switch from the Democratic Party to the Republican Party in the last month?  After all, that's how the demographic breakdowns changed between this month and last month for this same poll.