NYT LIVE POLL THREAD 2

[Crumpets]

To preserve conversations, I split the old thread a few pages back from the end to make this new thread. -Mod.

https://www.nytimes.com/interactive/2018/upshot/elections-polls.html

****

Nate Silver made an observation a few days ago that is very true: https://twitter.com/NateSilver538/status/1049694810314031104.

Recent polls have been very extreme. Either the Democrat gets really bad news (TN-SEN, ND-SEN, TX-31), or the Republican is getting crushed (PA-1, VA-10). Even with the generic ballot, I have seen some polls that show a close race and others that show a double-digit Democrat lead. It's kind of hard to know what the national environment really looks like at this point.

Or maybe politics has become so polarized by geography that the national environment is no longer helpful in analyzing local races. I think the idea of a uniform national swing might be well past its prime.

IIRC, after the 2014 midterms, Silver was saying that inconsistent polls actually produce more accurate polling averages, since they tend to be less influenced by herding.

[BudgieForce]

I know I shouldnt pay much attention to this right now, but MN-8 just got like 20 republicans in a row back to back. Ive never seen anything like that happen.

[GeorgiaModerate]

Nate Silver made an observation a few days ago that is very true: https://twitter.com/NateSilver538/status/1049694810314031104.

Recent polls have been very extreme. Either the Democrat gets really bad news (TN-SEN, ND-SEN, TX-31), or the Republican is getting crushed (PA-1, VA-10). Even with the generic ballot, I have seen some polls that show a close race and others that show a double-digit Democrat lead. It's kind of hard to know what the national environment really looks like at this point.

Or maybe politics has become so polarized by geography that the national environment is no longer helpful in analyzing local races. I think the idea of a uniform national swing might be well past its prime.

"All politics is local" may be becoming even more true.

[I spent the winter writing songs about getting better]

Nate Silver made an observation a few days ago that is very true: https://twitter.com/NateSilver538/status/1049694810314031104.

Recent polls have been very extreme. Either the Democrat gets really bad news (TN-SEN, ND-SEN, TX-31), or the Republican is getting crushed (PA-1, VA-10). Even with the generic ballot, I have seen some polls that show a close race and others that show a double-digit Democrat lead. It's kind of hard to know what the national environment really looks like at this point.

Or maybe politics has become so polarized by geography that the national environment is no longer helpful in analyzing local races. I think the idea of a uniform national swing might be well past its prime.

https://twitter.com/NateSilver538/status/1049694810314031104

[Chancellor Tanterterg]

The other thing is that this whole notion of Wallace being an awful candidate has been completely overblown.  I mean, he's probably not the best we could've run here, but he's still been a solid B-list recruit.  

I know Republicans like to pretend that he's a closet anti-Semite or whatever, but speaking as someone who is both Jewish and has been among the more vocal posters on Atlas about criticizing candidates for anti-Semitism and Jew-baiting, I think it was a mistake to go with that line of attack in this district.  It's like how DeSantis tried to paint Gillum as an anti-Semite when Gillum has done far more outreach to Florida's Jewish community since the primary that DeSantis has during the whole campaign.  Any Jewish voter who hadn't already made up their mind to oppose Wallace no matter what is going to be able to call BS on the attempts to depict him as an anti-Semite and it'll backfire for Fitzpatrick (if it hasn't already).  This is also not a district where using Wallace's wealth to try to depict him as Mr. Burns is necessarily going to be terribly effective.  If I were Fitzpatrick, I'd try to paint Wallace as an elitist carpet-bagger, but as far as I can tell, they haven't really done as much as would expect with those lines of attack.  

So where does that leave Fitzpatrick?  Well, he seems to have a decent relationship with labor, but the unions aren't going to actually work to bail him out either.  I don't think he has much else going for him this cycle tbh.  He clearly can't count on the Bucks County machine any more than local Republicans could count on the DelCo machine last November (and that was an absolute bloodbath).  Honestly, I've had this down for a while as not only a more likely Democratic pickup than CW would suggest, but as one where I think the Democrat will win by a lot more than people expect (I could see it being somewhere between 7-12%, probably more like 7-9%).

[Xing]

I know I shouldnt pay much attention to this right now, but MN-8 just got like 20 republicans in a row back to back. Ive never seen anything like that happen.

If you mean without any calls that apparently weren't answered in between, that's probably just a hiccup that happens sometimes when the site is loading. At least, it's happened for me several times.

[Florida Man for Crime]

"All politics is local" may be becoming even more true.

It is the opposite, and the fact that there is increasing geographical polarization is the proof of it.

A 2 party system tends towards roughly 50% of the people being on one side, and roughly 50% being on the other side. If politics were local, then politics would polarize within each local area so that roughly half the voters took one side on the various local issues, and the other half took the other side on those same local issues. And in other local areas, it would also polarize based around local issues so that about half the people in each area were on either side of the various local issues. This would lead to roughly even splits across different local areas. That is why there was less geographical polarization in the mid/late 20th century than now, because politics was more localized then than now.

Instead of politics polarizing intra-regionally, it is polarizing inter-regionally. It's us - (the people here) against them (those other people over there on the other side of the country that are making everything go to hell). So homogeneous geographical voting is an indication that all politics is increasingly national, not that all politics is local.

[Fmr. Gov. NickG]

I wonder if they really just decided to re-poll Minnesota because the response rate there is so good.

[BudgieForce]

Probably has nothing to do with anything, but interesting none the less.

The first MN-8 poll had 60% cellphone, 40% landline. This current one is 24% Cellphone, 76% landline.

[DataGuy]

Looking good for Stauber so far in MN-8. He's up 50-34 with 100 responses in. At the same point in their first poll of this point, he was down a few points.

[Florida Man for Crime]

Probably has nothing to do with anything, but interesting none the less.

The first MN-8 poll had 60% cellphone, 40% landline. This current one is 24% Cellphone, 76% landline.

I have noticed in previous polls sometimes they have started off calling all landlines (or all cell) in different districts. When they start with all landlines, Republicans usually start off overperfroming. When all cell, Dems usually start overperforming.

[BudgieForce]

Probably has nothing to do with anything, but interesting none the less.

The first MN-8 poll had 60% cellphone, 40% landline. This current one is 24% Cellphone, 76% landline.

I have noticed in previous polls sometimes they have started off calling all landlines (or all cell) in different districts. When they start with all landlines, Republicans usually start off overperfroming. When all cell, Dems usually start overperforming.

I was thinking it'd be something like that. Because if Stauber is really up double digits, democrats really are in trouble.

[GeorgiaModerate]

Probably has nothing to do with anything, but interesting none the less.

The first MN-8 poll had 60% cellphone, 40% landline. This current one is 24% Cellphone, 76% landline.

I have noticed in previous polls sometimes they have started off calling all landlines (or all cell) in different districts. When they start with all landlines, Republicans usually start off overperfroming. When all cell, Dems usually start overperforming.

That would actually account, at least in part, for the big swings in some of the polls from the early results (of course, normal sample variation is a factor too.)

[Bidenworth2020]

It is pretty clear that rn MN-08 is too high for pubs, the gcb has them doing better than Trump.

[Xing]

Probably has nothing to do with anything, but interesting none the less.

The first MN-8 poll had 60% cellphone, 40% landline. This current one is 24% Cellphone, 76% landline.

I have noticed in previous polls sometimes they have started off calling all landlines (or all cell) in different districts. When they start with all landlines, Republicans usually start off overperfroming. When all cell, Dems usually start overperforming.

I was thinking it'd be something like that. Because if Stauber is really up double digits, democrats really are in trouble.

There's no way that Trump's approval rating is +22 there, though. It's too early to be drawing conclusions.

[Florida Man for Crime]

Probably has nothing to do with anything, but interesting none the less.

The first MN-8 poll had 60% cellphone, 40% landline. This current one is 24% Cellphone, 76% landline.

I have noticed in previous polls sometimes they have started off calling all landlines (or all cell) in different districts. When they start with all landlines, Republicans usually start off overperfroming. When all cell, Dems usually start overperforming.

I was thinking it'd be something like that. Because if Stauber is really up double digits, democrats really are in trouble.

The Trump approval will probably be a pretty decent comparison point between this MN-08 poll and the early one. In the earlier one, Trump was -1 approval. As the poll gets more responses, it will probably go somewhere to around the area of -1 Trump approval. Currently it has +22 Trump approval, and relative to that, Radinovich is actually overperforming (so far).

[BudgieForce]

Yeah, im just gonna come back tomorrow. The current poll has Trump with a +26 approval rating and an R+20 generic ballot.

I know its early but its still hard to look past how strange these early results are.

[Xing]

Watch NYT get Wallace +10 or something.

That doesn't sound entirely crazy or implausible, TBH. Remember, NJ-03 is literally right next door. Whatever trends are happening in the NJ suburbs, similar trends will be going on in the PA suburbs.

Even Andy Kim's TV ads - insofar as they are anti-Republican in general, will be hurting Fitzpatrick in PA-01 as well, since voters are seeing them across the Philadelphia media market. Every time voters in PA-01 see a Kim ad about MacArthur repealing Obamacare, that is a reminder about how Republicans tried to repeal Obamacare, and that reminder hurts Fitzpatrick as well. So to some extent anyway, he will have interstate coattails. Probably not a huge effect, but media markets tend to swing together for this reason.

Well played. It looks like this could be like KS-03, where fundamentals are really too much for the incumbent to overcome, and people write off the challenger because they're a "weak candidate", which turns out to be nonsense.

[BudgieForce]

Oh god, why did I check again!?

LOL.

[Pericles]

If NY-24 was polled, would it have the same phenomenon? Katko has been favored the whole cycle and I don't think anyone has paid his race much attention, but Clinton did win his district by just under 4%, so by that alone he seems very vulnerable. And Obama won it twice by landslide margins. Correct me if I'm wrong, but it isn't a very Hispanic heavy district unlike say TX-23, FL-26 and CA-21.

[I spent the winter writing songs about getting better]

That MN-08 geographic distribution looks pretty skewed toward Republican areas.

[Dr. Arch]

That MN-08 geographic distribution looks pretty skewed toward Republican areas.

Yep, almost nothing from Duluth and the NE corner.

[Xing]

That MN-08 geographic distribution looks pretty skewed toward Republican areas.

I'm guessing that they're trying to reach the rural areas first in this poll, and then later on they'll call more people from places like Duluth. That's basically the only way to explain Trump being at +34 here (which actually doesn't make Stauber +26 look that good.)

[Ebsy]

They had a very hard time meeting quotas when they polled MN-08 last time, so as people mentioned, they are almost certainly calling rural voters first.

[Florida Man for Crime]

If NY-24 was polled, would it have the same phenomenon? Katko has been favored the whole cycle and I don't think anyone has paid his race much attention, but Clinton did win his district by just under 4%, so by that alone he seems very vulnerable. And Obama won it twice by landslide margins. Correct me if I'm wrong, but it isn't a very Hispanic heavy district unlike say TX-23, FL-26 and CA-21.

Could be, but it is not a pure-ish suburban district in the same sense as PA-01. It is a mix of rural, suburban, and urban.