absentee/early vote thread, part 2

[Ozymandias]

https://projectnewamerica.com/latest/colorado-state-play

[Ozymandias]

https://projectnewamerica.com/latest/florida-state-play

[Ozymandias]

https://projectnewamerica.com/latest/nevada-state-of-play

[Ozymandias]

https://projectnewamerica.com/latest/north-carolina-state-play

[Ozymandias]

https://projectnewamerica.com/latest/ohio-state-play

[Ozymandias]

https://projectnewamerica.com/latest/pennsylvania-state-play

[Ozymandias]

Steve Schale ‏@steveschale  4m4 minutes ago Florida, USA
Low propensity EV voters:
36.8% of all Dems
29.5% of GOP
54.0% of NPA

Dems have roughly 187k advantage with low propensity

Steve Schale ‏@steveschale  2m2 minutes ago Florida, USA
Percentage of FL Hispanic EV that is low propensity:

All: 55%
GOP: 37.4%
DEM: 59.2%
NPA: 66.4%

[Ozymandias]

Won't the ED vote in Florida reverse some of the diversification that's happened in the EV? 

I'm not sure why it is that Steve doesn't seem to think so/be worried about it. One thing that helps is that more than 2/3rds of all votes will already be in by ED.

Also, the low-propensity evidence suggests that Republicans are cannibalizing more of their election-day vote than are Democrats.

[Ozymandias]

Michael McDonald ‏@ElectProject  2m2 minutes ago
Something went very wrong for African-Americans' voting in North Carolina

[Ozymandias]

Can someone explain, why you think Florida is in Hillary's pocket??

2012: D 42.9%, R 39.1%, Other 18.1%,         total Dems +170k
2016: D 39.6%, R 39.1%, Other 18.9%,         total Dems   +32k
(today is not included)

I took date from http://www.electproject.org/

Why is so good? Yes, I know many Dixies switched to Reps. Other is more Clinton-friendly. But how does this data implies that Florida is gone for Trump? Probably slightly lean D, but gone?

No entiendo nada

Robby Mook did say he believes Clinton is ahead by about 170k in FL. This point in 2012, he said the Obama campaign thought they were down 15k. I don't believe it's in the bag for Clinton but it's definitely looking good.

Exactly. The main reason to think FL is in the bag is that the Clinton campaign isn't spending that much time here-- why spend the last day of early voting in Florida fighting so hard over a tossup (at best) like Ohio if you only have a 1 point in Florida?

[Ozymandias]

Can someone explain, why you think Florida is in Hillary's pocket??

2012: D 42.9%, R 39.1%, Other 18.1%,         total Dems +170k
2016: D 39.6%, R 39.1%, Other 18.9%,         total Dems   +32k
(today is not included)

I took date from http://www.electproject.org/

Why is so good? Yes, I know many Dixies switched to Reps. Other is more Clinton-friendly. But how does this data implies that Florida is gone for Trump? Probably slightly lean D, but gone?

No entiendo nada

Robby Mook did say he believes Clinton is ahead by about 170k in FL. This point in 2012, he said the Obama campaign thought they were down 15k. I don't believe it's in the bag for Clinton but it's definitely looking good.

As for the numbers: if Schale's claim is correct and 50 K Rep EV this year were Dem EV in 2012, than the actual Dem advantage in 2012 was really just +70 K.

Add to that the increase in unaffiliated, Hispanic, and female voting-- which is reflected in the HUGE edge Clinton has consistently had in polls of early voters-- and you can very easily see how Clinton got up to a +170K lead... and I think he said that before yesterday's voting?

Lastly, as Schale has noted, the Dems still have a +70K edge in ballots not yet returned, and have turned out significantly more low propensity voters.

Putting it all together (numbers + campaign strategy), and my best guess is that Clinton will win Florida by at least 3 points.

[Ozymandias]

Robby Mook did say he believes Clinton is ahead by about 170k in FL. This point in 2012, he said the Obama campaign thought they were down 15k. I don't believe it's in the bag for Clinton but it's definitely looking good.

He thinks they are making +140k today?

No, I think he was saying that their internal polling + modeling had Clinton up +170K actual votes going into Sat.

[Ozymandias]

From Ralston's latest update: http://www.ktnv.com/news/ralston/the-nevada-early-voting-blog

"Even though most observers, including myself, have used 18 percent of the electorate in 2012 being Hispanic because of exit polls, that is not as foolproof as an actual look at the voter file. The Democrats have always believed it was slightly lower -- 16 percent, perhaps.

Here's what the data shows on early/mail vote:

2012
- 11 percent Latino
- 77 percent white

2016
-- 13. 5 percent Latino
--77 percent white

The more ominous news for Republicans is how the majority Hispanic precincts looked between the two cycles: Huge increases except for one. Here is the actual data:"



More examples in blog.

[Ozymandias]

daniel a. smith ‏@electionsmith  38m38 minutes ago
Total black turnout in Florida in 2016 will be higher than 2012.  Mark it.
At 777k as of yesterday. Souls to polls today. VBM & ED to come

daniel a. smith ‏@electionsmith  35m35 minutes ago
As a point of reference, in 2012 in Florida, the share of turnout was:
B14%/H12.5%/W68.4%/A1.5%

daniel a. smith ‏@electionsmith  2m2 minutes ago
In Florida in 2012, 44% of whites waited to vote until ED; 50.5% of Hispanics; 35.2% of blacks.

[Ozymandias]

Taniel ‏@Taniel  5m5 minutes ago
On Florida early voting: "911,000 Hispanics have voted—more than a third of whom did not vote in 2012" http://www.miamiherald.com/news/politics-government/election/article112958953.html …

[Ozymandias]

daniel a. smith ‏@electionsmith  9m9 minutes ago
In FL, 36% of 907k Hispanics who have voted didn't vote by any method in 2012. That's 12 points higher than whites

daniel a. smith ‏@electionsmith  6m6 minutes ago
I don't draw on exit polls except when I'm criticizing them.
My numbers are all from the DSDE's and 67 county SOE's official voter files

[Ozymandias]

https://www.washingtonpost.com/graphics/politics/2016-election/where-early-votes-were-cast/

Lazaro Gamio ‏@LazaroGamio  1h1 hour ago
Here's an early voting cartogram @KevinUhrm and I made using the invaluable data from @ElectProject:

[Ozymandias]

Nate Cohn ‏@Nate_Cohn  1m1 minute ago Manhattan, NY
Clinton 48.8, Trump 47.7 in the final North Carolina tracker, which now adds the data from our last poll

"Here’s a breakdown of those estimates for the two major-party candidates:

IN RAW VOTES   AS A PCT.
Clinton   Trump   Clinton   Trump
Already voted   1,610,000   1,418,000   51.9%   45.7%
Yet to vote      685,000   825,000           42.8%   51.6%
Total estimates   2,295,000   2,243,000   48.8%   47.7%"

http://www.nytimes.com/interactive/2016/upshot/north-carolina-early-vote-tracker.html

(Note NY Times free for next two days so you don't have to worry about burning up monthly click allotment.)

[Ozymandias]

http://steveschale.com/blog/2016/11/7/1-more-day-we-can-do-this.html

"Total Ballots cast: 6,419,154
Democrats: 2,558.072 (39.85%)
Republicans: 2,470,823 (38.49%)
NPA: 1,390,259 (21.66%)
Total Margin: DEM +1.36%"

"Right now, I think about 67% percent of the likely electorate has voted. Late last week, I predicted 70%, but to be fair, I had it at 9.2m turnout late last week. At that turnout, EV ended up at 69.7%, so that prediction was almost OK. 9.5 million is where I do think it lands – that is roughly 08 level turnout (exact 08 would be 9.55m). Given the sheer numbers of low propensity voters, it could go beyond that, but honestly, I would be surprised (albeit pleasantly)."

"But two things emerged last week. One, this low propensity Hispanic thing became a thing. While Trump folks argued that Trump would turn out low propensity voters, we'd see slight edges for Democrats in this category. What became clear over days last week, this was a Hispanic deal, and as week 2 of early voting took hold, so did this surge. As of Saturday, Democrats had an egde of more than 175K low propensity voters.

Secondly, we began to see the edging upward of NPA voters. I had projected NPA at 21% of the electorate, but it will probably land closer to 23. And it is really diverse, running an average of four points more diverse than the electorate as a whole."

[Ozymandias]

https://twitter.com/reidepstein/status/795646505029877760

ouch.

Daniel Berman ‏@DanielBerman2  5m5 minutes ago
Worth noting, party ID in Ohio is determined by which primary you took part in. So GOP identification might be inflated this year by Kasich

[Ozymandias]

Here's a summary of all the polling data of early voters in Florida:

FLORIDA polling of early voting summary

Presented from most-to-least percent of Early Voters (EV), with absolute number of EV (which can be used as a proxy for MOE) in parentheses:

Opinion Savvy: 71% (607 EV)-- Clinton +8

Quinnipiac: 65% (575 EV)-- Clinton +6 (48-42)

YouGov: 63% (748 EV)-- Clinton +10 (51-41)

Opinion Savvy: 55% (333 EV)-- Clinton +9 (53-44)

Emerson: 51% (257 EV)-- Clinton +15 (57-42)

TargetSmart: 43% (311 EV)-- Clinton +18 (55-37)

Quinnipiac: 42% (263 EV)-- Clinton +6 (48-42)

NBC: 36% (356 EV)-- Clinton +17 (54-37)

FAU: 26% (226 EV)-- Clinton +12 (54-41)

Based on the most recent polling, and how well the Dems did in early voting this past weekend, I think it's safe to say that Clinton is heading into election day with almost a 10-point lead among the ~76% of the population that has now voted.

[Ozymandias]

Based on the most recent polling, and how well the Dems did in early voting this past weekend, I think it's safe to say that Clinton is heading into election day with almost a 10-point lead among the ~76% of the population that has now voted.

Wow. If I'm doing my math right, Trump would have to win ED vote 68-32 to pull even

Oh yeah, I should probably add a caveat: I used total 2012 votes (8.5M) to calculate the percentage that has already voted, but Michael McDonald tweeted  that he thinks turnout could be as high as 9.5M this year-- in which case only 68% of the electorate has already voted.

[Ozymandias]

NORTH CAROLINA polling of early voting summary

Presented from most-to-least percent of Early Voters (EV), with absolute number of EV (which can be used as a proxy for MOE) in parentheses:

Quinnipiac: 67% (583 EV)-- Clinton +12 (52-40)

Siena/NYT: 63% (504 EV)-- Clinton +9 (49-40)

Quinnipiac: 44% (265 EV)-- Clinton +22 (58-36)

Emerson: 32% (209 EV)-- Clinton +22 (59-37)

Siena/NYT: 31% (492 EV)-- Clinton +15 (53-38)

YouGov: 29% (288 EV)-- Clinton +6 (51-45)

Marist/NBC: 29% (295 EV)-- Clinton +28 (61-33)

PPP: 19% (166 EV)-- Clinton +26 (63-37)

Quinnipiac: 15% (105 EV)-- Clinton +28 (62-34)


So much like Florida, Clinton seems to have built a huge lead in the very early voting that seems to have dropped down to high single digits among the ~66% of voters that voted by the end of in-person EV (turnout estimate of 4.7 M taken from Upshot).

[Ozymandias]

Quinnipiac: 67% (583 EV)-- Clinton +12 (52-40)

Siena/NYT: 63% (504 EV)-- Clinton +9 (49-40)

So much like Florida, Clinton seems to have built a huge lead in the very early voting that seems to have dropped down to high single digits among the ~66% of voters that voted by the end of in-person EV (turnout estimate of 4.7 M taken from Upshot).

Note that the Upshot's early voting tracker model (http://www.nytimes.com/interactive/2016/upshot/north-carolina-early-vote-tracker.html) currently predicts a slightly lower Clinton EV edge of 6%, because they're basing their results off the +9 edge they found in their most recent poll, which is a little lower than the +12 edge Quinnipiac found in their larger sample.

[Ozymandias]

NORTH CAROLINA polling of early voting summary

Presented from most-to-least percent of Early Voters (EV), with absolute number of EV (which can be used as a proxy for MOE) in parentheses:

Quinnipiac: 67% (583 EV)-- Clinton +12 (52-40)

Siena/NYT: 63% (504 EV)-- Clinton +9 (49-40)

Quinnipiac: 44% (265 EV)-- Clinton +22 (58-36)

Emerson: 32% (209 EV)-- Clinton +22 (59-37)

Siena/NYT: 31% (492 EV)-- Clinton +15 (53-38)

YouGov: 29% (288 EV)-- Clinton +6 (51-45)

Marist/NBC: 29% (295 EV)-- Clinton +28 (61-33)

PPP: 19% (166 EV)-- Clinton +26 (63-37)

Quinnipiac: 15% (105 EV)-- Clinton +28 (62-34)


So much like Florida, Clinton seems to have built a huge lead in the very early voting that seems to have dropped down to high single digits among the ~66% of voters that voted by the end of in-person EV (turnout estimate of 4.7 M taken from Upshot).

Ok you are just making up numbers. Troll. There is no way every single poll in both states show her with a double digit lead.

If you look closely, some of them are single-digit leads.

Yes, the YouGov poll was only +6, but since it was sandwiched by 6 double-digit polls (five of which were 20+ leads), I considered it an outlier.

And the Siena/NYT polling of EVs dropping down to +9 is why I guessed that the current lead is high single digits.