The US Election Polling / Data Thread

[octothorp]

Man, I've had poll withdrawal for a couple days, with no meaningful polls coming out. PPP has what I think are the first totally-post-debate polls from a reputable pollster, and there's early signs of an uptick for Clinton. They've got her +4 nationally (which is the same as their last poll in late August when she had a comfortable lead), plus narrow (+2) leads in North Carolina and Florida, and comfortable (+6) leads in Colorado, Pennsylvania, and Virginia.

Possibly more significantly, Rasmussen has a poll taken from Monday to Wednesday which has Clinton up by 1, which is a 6 point swing from a week ago. (538 adjusts their numbers to Clinton by +2).

[Senator Clay Davis]

Debate has helped her without question. With four days of Trump ramping up the misogyny, these might be the lowest deficits he'll see until the next debate. Ugly numbers for him in Michigan and New Hampshire.

FL: Clinton 46, Trump 42, Johnson 7, Stein 1 (Mason-Dixon)
MI: Clinton 42, Trump 35, Johnson 9, Stein 3 (Detroit News)
NH: Clinton 42, Trump 35, Johnson 13, Stein 4 (WBUR/MassINC)

[Drak]

Originally Posted by Senator Clay Davis

Debate has helped her without question. With four days of Trump ramping up the misogyny, these might be the lowest deficits he'll see until the next debate. Ugly numbers for him in Michigan and New Hampshire.

FL: Clinton 46, Trump 42, Johnson 7, Stein 1 (Mason-Dixon)
MI: Clinton 42, Trump 35, Johnson 9, Stein 3 (Detroit News)
NH: Clinton 42, Trump 35, Johnson 13, Stein 4 (WBUR/MassINC)

What I don't get about polls is the myopic tendency for certain voters to suddenly change their opinions on a dime. The next debate if Trump shines people conveniently forget all the other stuff he's said and done, which are alarming foundational flaws in his character. Perhaps I just don't understand the complexities of how polls work.

[troutman]

Clinton has climbed 7.2% in last week on Betting Odds to 70.7%. Trump at 27.2%

https://electionbettingodds.com/week.html

66.3% to 33.6% on 538. NV and FLA have flipped to Clinton, and OH and NC are getting close.

[troutman]

Will there be an SNL bounce (premiere on Saturday)?

Sarah Palin dives in poll ratings as Tina Fey impersonates her on Saturday Night Live

http://www.telegraph.co.uk/news/worl...ight-Live.html

[Parallex]

Originally Posted by Drak

What I don't get about polls is the myopic tendency for certain voters to suddenly change their opinions on a dime. The next debate if Trump shines people conveniently forget all the other stuff he's said and done, which are alarming foundational flaws in his character. Perhaps I just don't understand the complexities of how polls work.

Meh, Candidate X does better in newscycle... 10% of Candidates Y's voters become undecided, 10% of undecideds now support Candidate X. It's not like the people whose opinions are changing are firmly held.

[octothorp]

Originally Posted by Drak

What I don't get about polls is the myopic tendency for certain voters to suddenly change their opinions on a dime. The next debate if Trump shines people conveniently forget all the other stuff he's said and done, which are alarming foundational flaws in his character. Perhaps I just don't understand the complexities of how polls work.

My take is that what we're seeing is not movement toward anyone, it's movement away from whoever is in the spotlight (which looks the same in the polls in a binary race, but is actually different). That's really unusual, and is a reflection of the unlikeability of both of these candidates. So I don't think that Trump having a good debate changes anything. Clinton having an awful debate (or a different bad story) might. Trump has never really gotten any sort of bump (other than a totally average convention bounce) as a result of his own behavior. It would be difficult to say for certain that Clinton has, either (her big convention bump could be attributed more to Trump's Khan tangent).

The tendency is that as the election gets closer, the potential for swing tends to decrease. That may not hold true in this weird year, but typically second and third debates do not produce much movement at all, and while the first debate produces at most a moderate bump that then fades a bit.

My gut feeling is that Clinton might be getting a largish bump that could have significant staying power. But more important than that, I think this debate (and the fall-out) really caps Trump's ceiling; there's there's probably 3-4% who are still vacillating between supporting him or not, but even if he gets all of those supporters, he probably caps out at around 43-44%. Which isn't likely to be enough to win unless third party votes and electoral college votes break in exactly the right way.

[Itse]

Originally Posted by octothorp

typically second and third debates do not produce much movement at all, and while the first debate produces at most a moderate bump that then fades a bit.

IMO Trump can't keep losing 1-on-1 battles to Clinton this bad. Being a winner is such a big part of his image. I also think this might somewhat of a cumulative issue. One time is one time. Losing bad every time is a much bigger deal.

[Drak]

Clinton climbing again as per 538

http://projects.fivethirtyeight.com/...tion-forecast/

Before the debate they were almost tied.

[Senator Clay Davis]

Interesting round of polls, with Hillary now holding a bigger lead in Florida than Pennsylvania. And perhaps the most interesting poll of the election, the New Mexico poll. Could Trump's supporters go strategic and try and give Johnson the win? Neither Hillary or Trump getting 270 is the same as Trump winning.

CO: Clinton 49, Trump 38, Johnson 7, Stein 3 (Monmouth)
FL: Clinton 46, Trump 41, Johnson 5, Stein 2 (Q)
NM: Clinton 35, Trump 31, Johnson 24, Stein 2 (Albuquerque Journal)
OH: Trump 47, Clinton 42, Johnson 6, Stein 1 (Q)
PA: Clinton 45, Trump 41, Johnson 5, Stein 2 (Q)
NC: Clinton 46, Trump 43, Johnson 7 (Q); Clinton 44, Trump 43, Johnson 6 (Bloomberg)
VA: Clinton 42, Trump 35, Johnson 12, Stein 1 (Q)

[GGG]

I don't know if house Republicans take Trump over Johnson in the tied scenario and if the democrats can't win given the house make up they would likely back Johnson over Trump.

Also you could run into the scenario where the loyalty of your electoral college individuals are tested. I had read somewhere that not all electoral college voters are actually obligated to vote with the state result. If it only took one or two defectors in the college it could be very interesting.

[troutman]

This looks familiar:


only difference today is Ohio, Iowa and NC?

[octothorp]

Two polls (Monmouth, Franklin & Marshall), find Clinton up 10 and 9 points in Pennsylvania. +6 in North Carolina according to Elon. Clinton +6 nationally.

Another awful day of polls for Trump (maybe the worst yet on swing-state numbers), and we're still not at a point where the full impact of his awful week is being captured.

[Senator Clay Davis]

As Nate pointed out she has won all but one swing state poll since the debate. Florida slipping away from Trump is the big one though, that's the election ender. So I wonder if we see the Hillary campaign maybe shift some resources from PA and CO, which again appear pretty safe, and go for the kill in Florida.

[octothorp]

Or you try and pull Trump in as many directions as possible (with particular focus on senate-race states... Pennsylvania, North Carolina, Nevada, New Hampshire, Florida) with the upside that you increase the likelihood of a landslide and powerful mandate.

[bob-loblaw]

Are the undecided numbers finally starting to drop?

[afc wimbledon]

Originally Posted by bob-loblaw

Are the undecided numbers finally starting to drop?

A couple of serious republican pundits are saying this is the election 'breaking' for Clinton, short of some massively damaging completely new scandal this probably it for Trump

[octothorp]

That's a good question and is kinda difficult to read across different pollsters who have different methodologies in selecting between 'undecided likely voters' and unlikely voters. However, for most of the election, declared support for either Clinton or Trump has held at around 80-81% total. But that number has been gradually rising, and in the current 538 aggregate, is at 84% (44% Clinton, 40% Trump), which is the highest it's been. Trump's popular support hasn't dropped significantly since the debate, it's just that Clinton's has risen, presumably due to drawing in undecideds and third-party voters. She's about as high right now as she was at her post convention peak (but Trump isn't as low as he was at that point).
So yeah, it does look like former undecideds are starting to declare for Clinton, but it's not happening at a rapid speed.

(This is kinda grouping third-party and undecided support together which isn't really fair, but there's reason to think that some of those third-party supporters will 'decide' on a major candidate in the last month, as that's what the current movement and movement in past elections looks like.)

[ernie]

Haven't looked at the polls but is the third party support starting to shift back to major parties as it becomes more apparent that the third parties don't have complete platforms or leaders worth voting for? Clinton was taking the brunt of that third party bleed so migration back to the old standbys only helps her.

[octothorp]

Originally Posted by ernie

Haven't looked at the polls but is the third party support starting to shift back to major parties as it becomes more apparent that the third parties don't have complete platforms or leaders worth voting for? Clinton was taking the brunt of that third party bleed so migration back to the old standbys only helps her.

Slightly. Johnson and Stein have lost roughly 2% since a peak in early September (they were at ~12% combined, and are now around 10%). Some of that looked like movement toward Trump in September, but looks like movement more towards Clinton in the last week.