absentee/early vote thread, part 2

[Erich Maria Remarque]

Early Vote in North Carolina Seems Consistent With a Clinton Lead, Nate Cohn - Upshot

[Erich Maria Remarque]

https://twitter.com/FoxNews/status/793828588197863425

You must be logged in to read this quote.

https://twitter.com/electionsmith/status/793833846902894596

[Erich Maria Remarque]

https://twitter.com/FoxNews/status/793828588197863425

Yeah republican's are screwed.

IDK, probably they are, but it is certainly not what this picture tells us.

[Erich Maria Remarque]

I doubt such a large increase in hispanic turnout is a good sign for Trump.

Lets turn it around. Whites is up by 5% points, while Others is down by the same amount.
Of course, it might be a realignment among Whites (more educated or more female etc), but generally the more Whites => the better Trump's doing.

[Erich Maria Remarque]

Even Latino Decisions shows a 40-ish point margin for Clinton. What is the basis for your claim?

http://www.politico.com/states/florida/story/2016/10/clinton-dominating-trump-among-florida-hispanics-in-new-poll-106082

This.

http://www.latinodecisions.com/files/8714/7689/5921/NALEO_Latino_Battleground_States_Toplines.pdf

63 - 23 Clinton, but with a relatively long time in the field.

I suppose you can pick whichever in this case, but LD has been pretty accurate in the past and the TargetSmart poll would lend credence to this defection.

Reading these posts online here of all liberals you would think trump was down with hispanics by 40-50% in FLorida, not the case obviously.

I'd say there is plenty of evidence to support both conclusions. You can assert that "no way he's down by 40 with FL latinos" if you want, but you are by no means on solid ground when you say that, and you should at least acknowledge that.

LD overestimated Obama's margin both times IIRC. (Not 100% sure)

TargetSmart, well...

[Erich Maria Remarque]

Looks like they were a little bit Obama-friendly in '12. Couldn't find '08, though I didn't look hard.

LD 2012 final proj: 75-23
2012 Exit poll final: 71-27

Yeah. Marginwise from Obama +52 to Obama 44%. 8% is not that little. If they have the same error now, it would mean Clinton +32 rather than Clinton +40. A huuuuge difference for Florida.

I don't know though how good they were in Florida in 2012/2008

[Erich Maria Remarque]

"hidden trump supporters"

lol k

laugh all you want but like i said i know many college educated white voters like myself who wont admit in public they support him. On an anonymous board i will but if im asked at work by anyone or in a public setting i say im fed up with the whole system and both parties and will vote for the independent candidate.

Why the hell would people lie to pollsters? This secret Trump voter argument doesn't make sense, because there is no reason for people to hide their intentions.

I am sorry, this just reeks of desperation.

people lie all the time

upshot/nyt said that 16% of the people who SAID they ALREADY voted, actualy didn't. so i don't know why but obviously they lied and 16% is a huuuge number

Is it true? Link?

Probably small sample caused it.

[Erich Maria Remarque]

If you included 2012 numbers for the states we don't have any data for yet this year, we're basically already past 2012's total early vote.

Wow. Is the any estimation of turnout (total, not just EV). I recall a poll from Gallup that predicted smaller turnout than 2012. But it is old.

Intresting...

[Erich Maria Remarque]

538 has her 0.2 up in polls-only. Hardly a big advantage, black turnout is down as well.

538 doesn't take into account early voting.

Nate Cohn said, thar early voting demographis is in line with his/other polls. So if his poll was wrong or if race changed (which it did according to 538), early voting is "good" neither for Clinton or for Trump

[Erich Maria Remarque]

You must be logged in to read this quote.

You must be logged in to read this quote.

Agree

[Erich Maria Remarque]

On Upshot's NC vote tracker, they have over half the total vote as in, and Hillary winning already-submitted votes by 10. That means Trump has to win remaining votes by at least 10 to carry the state. Not to mention 4 more days of early voting at current support numbers (roughly 50/50, if we're being conservative). That means Trump would have to win the election day vote pretty overwhelmingly unless things shift his way in the state right now. So, even if that Trump +7 poll was accurate and that's how people vote for the remainder of the election, he'd still lose the state overall.

Somebody please double-check my math, because it sounds a bit Karl Rovey when I read it back to myself.

It is based on a single Sienna/Upshot poll with Clinton +7. If the poll was right and the race has not changed since    OCT. 20-23, Trump is done.

[Erich Maria Remarque]

On Upshot's NC vote tracker, they have over half the total vote as in, and Hillary winning already-submitted votes by 10. That means Trump has to win remaining votes by at least 10 to carry the state. Not to mention 4 more days of early voting at current support numbers (roughly 50/50, if we're being conservative). That means Trump would have to win the election day vote pretty overwhelmingly unless things shift his way in the state right now. So, even if that Trump +7 poll was accurate and that's how people vote for the remainder of the election, he'd still lose the state overall.

Somebody please double-check my math, because it sounds a bit Karl Rovey when I read it back to myself.

It is based on a single Sienna/Upshot poll with Clinton +7. If the poll was right and the race has not changed since    OCT. 20-23, Trump is done.

I believe that Sienna is recontacting the poll's respondents to measure whether they have voted early or not and who they voted for, and is then making a projection for the rest of the state.

They are not recontacting, but looking at NC statistics. NC is one of few states, who gives all statistic with names etc.
But as I said, it depends on initial poll results. If results were wrong, or race shifted, then you cannot longer apply it. They use EV statistics only as a LV screen [and probably demographigs breakdown? IDK].

Yup.

http://www.nytimes.com/2016/11/04/upshot/donald-trump-cant-count-on-those-missing-white-voters.html?_r=0

It's telling when a candidate's supporters have been reduced to almost solely relying on strange, unproven & unlikely phenomena or straight up electoral miracles to justify their assertions that their candidate will beat the odds and pull out a win.

This article has nothing to do with Shy Theory. Like not at all.

[Erich Maria Remarque]

Ralston answered my question! If you comment he'll answer it.

He "dissed" Nate's model a bit

[Erich Maria Remarque]

Buhhhh I was told that WI would go for Trump because working white class whites!!!!

Yeah. Working white class whites that are registred Democrates

This article has nothing to do with Shy Theory. Like not at all.

Uhm? I read his link before posting and I'm well aware of what it was about. Did you not consider why I responded to his post while using two different terms? If I was solely referring to the shy voter theory I'd have responded to Ebsy's post.

Sheesh LittleBig, maybe next time be a little slower to pounce would ya?

I like you and you are so mean

[Erich Maria Remarque]

Jon Ralston did a live chat today about NV early voting and what's wrong with NV polling

https://www.youtube.com/watch?v=hiKfDL_Mlho


Basically polling sucks because of underpolling young transient and Hispanic voters, early voting good for Clinton and Mastro

Is there EV data over age, gender etc?

[Erich Maria Remarque]

Steve Schale:

"Not insignificant for HRC:
Non-white share of FL early voters thru yesterday: 31.3
Among NPA: 35
In other words NPA more diverse than state."

Schale just said on Maddow "We'll know pretty early on Tuesday who won Florida."

Considering early voting is huge this year, yup.

Do you in USA count EV before the polls are closed?

[Erich Maria Remarque]

Steve Schale:

"Not insignificant for HRC:
Non-white share of FL early voters thru yesterday: 31.3
Among NPA: 35
In other words NPA more diverse than state."

Schale just said on Maddow "We'll know pretty early on Tuesday who won Florida."

Considering early voting is huge this year, yup.

Do you in USA count EV before the polls are closed?

They are released quickly when polls close.

In Sweden they start to count EV after the polls close. But in Sweden you may change your mind. Your latest vote counts. Probably that's why. My bad.

[Erich Maria Remarque]

Democrats have closed the gap again in Florida yesterday and the difference is now less than 2k from 12k

But they won it 43-40 in 2012...

With Dixiecrats in the mix. Many of them have switched reg since 2012.

you don't know that for a fact

Actually, we do. Schale found 50,000 Republican early voters in 2016 who were registered as Democrats in 2012.

50k is nothing.
And how much switched back?

According to Upshot/Sienna poll. Trump is doing great among white uneducated registred Dems.

You must be logged in to read this quote.

Unfortunately, they don't specify what is numbers among all registred Dems/Reps. But they do specify the numbers among self-id D/R.
Dems: Clinton 86, Trump  7
Reps:  Trump  86, Clinton 5

Upshot mentions though that % of registred D > % self-id D (some D are saying they are R probably), which implyes, that Trump is probably doing even better among registred Dems than among self-id Dems.

But of course, one poll is just one poll.

[Erich Maria Remarque]

Nevada
- Dems ahead 29K, larger than 1 week ago (4 years ago, they were ahead 38K)

My impression was that her lead was better than 4 years ago? According to this Nevada guy?

Or does he includes big increase of NPA voters that more hispanic&yonger&urban than 2012?

Statewide, white share of EIP turnout fell Wed-->Thurs, 62.1-->60.8
Hispanic share of EIP turnout roughly the same Wed-->Thurs, ~16.5%

How did 2012's demographics of EV? I suppose it was more white?
It is hard to understnad without the context.

[Erich Maria Remarque]

Nevada
- Dems ahead 29K, larger than 1 week ago (4 years ago, they were ahead 38K)

My impression was that her lead was better than 4 years ago? According to this Nevada guy?

No clue where those numbers come from. Jon has them, as of this morning:

Statewide: Dems +37,700 (Let's say [TRIGGER WARNING] after the other cows come home, it's 37,000) In 2012 at this time, it was 41,000, and there were 200,000 fewer active voters.

Yeah, CNN... Probably they use old data or something.

[Erich Maria Remarque]

CNN is wildly incompetent and always searching for a neutral narrative. PPP are partisans, but they're good at what they do and aren't actively tipping the scales.

Reporting numbers =  always searching for a neutral narrative? OK, all right!

But yeah, they are incompetent.

on a conference call with nothing to say so doing back of envelope calcs.

NV. (according to sos as of yesterday's numbers)
Total voted:
Clark (Vegas): 199k Dems 139k Rep 94k ind
NV totas: 262k Dems 222k Rep 131 ind

Do you know how many UPA voted early in 2012? UPA should be Hillary friendly.

Obviously, Indie/UNA hold the key
Past 4 elections indies have gone (according to CNN exit):
2012: 43% Dem 50% Rep (BO v MR)
2008: 54% Dem  41% Rep (BO v JM)
2004: 54% Dem  42% Rep (JK v GWB)

Looks to me like, as it stands, Rep have to pick up a lot of votes or Ind/UNA have to break almost 60/40 to Trump. 94k Total ind/UNA have voted in Clark or about 72%. So is it plausible in any world that indie/UNAs from Vegas go for Trump 56/44 (you have to pull about 5 or 6 %, maybe higher for GJ and others, but those are rough numbers.) That's my calcs if the vote were counted EOD yesterday.
(Please, someone check my math, I'm doing this fast.)

Intresting.

Though self-reported Indie from CNN is not the same as UNA.
We'll probably also see more uneducated white Dems voting for Trump and educated white Reps voting for Hillary. Question is which share will be bigger

[Erich Maria Remarque]

Another positive for D

John Harwood:

"Robby Mook notes on an early vote press call that Obama lost white millennials in 2012 -- a demo Clinton is winning right now, he says."

Eh... It does not sound plausible at all. Or? I though that she was underperforming among them nationally. If so (not 100% sure), I doubt she'll do better in NV.

[Erich Maria Remarque]

Another positive for D

John Harwood:

"Robby Mook notes on an early vote press call that Obama lost white millennials in 2012 -- a demo Clinton is winning right now, he says."

Eh... It does not sound plausible at all. Or? I though that she was underperforming among them nationally. If so (not 100% sure), I doubt she'll do better in NV.

White male millennials were 54-41 Romney and females 49-48 according to exit polls, but millennials are only 55% white and will I think drop below 50% by 2020.

Wow, thanks!

And didn't know that just 55% millennials are white

[Erich Maria Remarque]

WOW...

Is it a good news for Hillary? Explain for us who don't know much about VA 

Steve Schale

Keep getting asked about NPA versus FL electorate as a whole.
Its more diverse:
All voters:  68w - 12.2b - 14.4h
NPA: 64.2w - 6.3b - 19.9h

How does it compared to 2012?

[Erich Maria Remarque]

No, as I said many times (more then >3 times, for sure), this model is based on one poll only, but is using EV data as some kind of LV screen. So it is alread brought from +7 to +6. By 1%, which is (oh the irony) is a house effect of Sienna poll according to Nate Silver's model. And if the race have tightened (and it is indeed true), the estimation are wrong as well. But sigh...

#EducationInAmericaSucks
https://twitter.com/Nate_Cohn/status/794694744303632385