2016 Polling Thread

[pbrower2a]

My weekend projection (early morning, 8/27/2016)



New Mexico has been polled for the first time in a long time, and the state is effectively out of range for Donald Trump. I consider at least 272 electoral votes out of range a meaningful contest at this stage in a Presidential race.

8% is not close.

My criterion for "strong" is 8%, twice the margin of error.
My criterion for "weak" is 4%, basically the margin of error.

8% is effectively out of range for peeling off a state from the Other Side with normal campaigning  even at its most intense.  Donald Trump now  needs miracles to have a chance now. After a huge number of statewide polls of states in the previous week, there were comparatively few last week. One poll that I saw turned out to be a fake; it had Hillary Clinton within 2 in Kentucky. As I rejected one commissioned for a labor union in the previous week I rejected one for the Florida Chamber of Commerce.

[Anthony '58]

WPIX-TV in New York is reporting this morning that a "new poll" has Hillary's lead being down to 3 points.

Have been unable to find any further information.

[pbrower2a]

We may be seeing a Trump charge now. What I first thought were anomalous polls may not be so anomalous.

Quote:Monmouth University, Pennsylvania:

http://www.monmouth.edu/polling-institut...PA_083016/

Clinton - 48%
Trump - 40%
Johnson - 6%
Stein - 1%

There  are some polls by an entity called Emerson College, but they are screwy enough to be suspect.

WISCONSIN:

https://law.marquette.edu/poll/2016/08/3...in-voters/

RV:
Clinton - 42%
Trump - 37%


Clinton - 37%
Trump - 32%
Johnson - 11%
Stein - 7%

LV:
Clinton - 45%
Trump - 42%


Clinton - 41%
Trump - 38%
Johnson - 10%
Stein - 4%

http://www.monmouth.edu/polling-institut...WI_083116/

Clinton - 43%
Trump - 38%
Johnson - 7%
Stein - 3%

Your choice on which you go with. Marquette recently had Clinton up by double digits

Some polls from Emerson College:

New York



North Carolina

[img]<br />https://static.wixstatic.com/media/3bebb2_a808fd9437734397aea0d9df48bcd703~mv2.jpg/v1/fill/w_208,h_208,al_c,q_80,usm_0.66_1.00_0.01/3bebb2_a808fd9437734397aea0d9df48bcd703~mv2.jpg[/img]

Michigan

[img]<br />https://static.wixstatic.com/media/3bebb2_627b152f417a4b50809d1d91f6ab0498~mv2.jpg/v1/fill/w_208,h_208,al_c,q_80,usm_0.66_1.00_0.01/3bebb2_627b152f417a4b50809d1d91f6ab0498~mv2.jpg[/img]

Pennsylvania:



Ohio:
[img]<br />https://static.wixstatic.com/media/3bebb2_23699d62d3eb417f9d5d63c3c90f1a44~mv2.jpg/v1/fill/w_457,h_457,al_c,q_90,usm_0.66_1.00_0.01/3bebb2_23699d62d3eb417f9d5d63c3c90f1a44~mv2.jpg[/img]


Arizona, Breitbart/Gravis:

Trump - 44%
Clinton - 40%
Johnson - 8%
Stein - 1%

Poll was of 1,244 likely Arizona voters conducted on August 27th.

http://www.breitbart.com/big-government/...8-stein-1/

Also from Arizona, but nothing from Emerson College:

http://email.connectstrategic.com/t/j-73630F4BD52B12D8

Clinton - 40%
Trump - 39%
Johnson - 7%
Stein - 1%

Conducted August 25-27. 728 LV. Rated C+ on 538.

Useless because it is a tie.



Binary race, Hillary Clinton (D) vs. Donald Trump ®  




Leader up with

60% or more -- saturation 80%
55-59% --     saturation 70%
50-54% --     saturation 60%
46-49%, margin 4% or greater saturation 40%
46-49%, margin 3% or less saturation 20%

(the usual color applies for the partisan leader, but yellow blue to green and red to orange below  

40-45%, margin 4% or greater, saturation 40%
43-45%, margin 3% or less, saturation 20%  





Johnson support:

16%+: 80
13-15: 70
10-12: 60
7-9: 50
4-6: 40
2-3: 30
0-1: 20
Poll w/ no Libertarian number: clear








Small states and districts in area or with shapes that allow confusion:

FL D4/4;4
ME D 10/10; 4
NH D15/8,4

Clinton (D)
Trump ®
Johnson (L)

The nature of this election cycle must change dramatically for Donald Trump to have any chance of victory.

It may have.

[Anthony '58]

If the gap between Johnson and Stein has shrunk to 4 points, then Trump has real chances - but that gap will have to shrink further, if not vanish entirely.

[pbrower2a]

Hillary Clinton cannot count on Johnson to gut the Republican support for Donald Trump. Remember -- Donald Trump shows signs of standing for cheap labor, callow treatment of the environment, lax regulation of elite behavior, and of course crony capitalism. He may have said that he is for keeping immigrants out -- but not the ones that he needs to work cheaply for him.

Note well that Republican campaigns often wilt beginning on Labor Day. It's hard to see what Donald Trump has to offer working people.

For a real indication of how Hillary Clinton is doing should Gary Johnson not be a factor, then look at the Senate races.

[pbrower2a]

The polls by Emerson College are English-only, and perhaps even more significantly, landline only.

With landlines one has an obsolescent technology. If something were to take out the landline grid, then it would not be rebuilt except for highly-specialized purposes. Prime example: after the Second Gulf War that took out the Iraqi telephone structure, the Kurds in the autonomous (and effectively independent, anti-fascist part of Iraq) went practically 100% cellular. Such was cheaper, safer, and more reliable than land lines.  Reliance upon demographics wedded to some obsolete technology implies a distorted sample.

Normal consumer use is not one of those specialized uses. Reliance on landlines will soon be as obsolete as using pre-recorded VHS tapes for entertainment. Landline users skew elderly, and hence Republican.People with cellphones only tend to be younger and more Democratic. As early as 2014, nearly half of all households had no landline phones ; as late as 2004, having a landline phone but not a cellular phone was so for nearly all households.





Even more problems accrue to the Emerson College polls. It is amazing that any pollster which heretofore yet has offered no polls of the Presidential race gives seven in one week. These polls show themselves much more pro-Trump than most national tracking polls hold.

[pbrower2a]

Casting out the polls involving Emerson College (English only, landlines only) and adding some other polls for Alaska, Kansas, New Hampshire, West Virginia, and Wisconsin:


Binary race, Hillary Clinton (D) vs. Donald Trump ®  



Leader up with

60% or more -- saturation 80%
55-59% --     saturation 70%
50-54% --     saturation 60%
46-49%, margin 4% or greater saturation 40%
46-49%, margin 3% or less saturation 20%

(the usual color applies for the partisan leader, but yellow blue to green and red to orange below  

40-45%, margin 4% or greater, saturation 40%
43-45%, margin 3% or less, saturation 20%  





Johnson support:

16%+: 80
13-15: 70
10-12: 60
7-9: 50
4-6: 40
2-3: 30
0-1: 20
Poll w/ no Libertarian number: clear








Small states and districts in area or with shapes that allow confusion:

FL D4/4;4
ME D 10/10; 4
NH D11/12,4

Clinton (D)
Trump ®
Johnson (L)

[Anthony '58]

I do find it curious that every poll that includes Gary Johnson and Jill Stein results in Hillary doing worse than in a two-way race.

How is this possible when Gary Johnson, who presumably is siphoning votes from Trump, is well out in front of Jill Stein, who presumably is siphoning votes from Hillary?

[pbrower2a]

The Washington Post has polled all fifty states, missing only the District of Columbia and the separately-voting Congressional Districts of Maine and Nebraska. With such a rich collection of polls, I can begin anew:

Blank map.



OK -- not that new!

This is all by the same pollster even with over 5000 people polled in Texas, and some of the results are counter-intuitive. We can ignore prior controversies from hereon and be stuck with new ones (oh, well!). Counter-intuitive data can be right, and change in the way people show valid perception of the world often begins with counter-intuitive data. In a binary choice:

Hillary Clinton (D) vs. Donald Trump ®:

Blank map.



Tie -- white

60% or more -- saturation 8
55-59.9% -- saturation 6
50-54.9% -- saturation 5
45-49.9%, lead 8% or more -- saturation 4
45-49.9%, lead 4-7.9% -- saturation 3
45-49.9%, lead 1-3.9% -- saturation 2

Any lead with less than 45% will be considered unusable.

Numeric data here:

http://apps.washingtonpost.com/g/page/po...poll/2086/


A few comments:

1. We no longer have the situation in which anyone has a lead with less than 45% of the vote in any state. I will include no such polling results hereon unless the margin is outside the margin of error (4%).

2. I could say such things as "Michigan/Pennsylvania/Wisconsin typically closes late and hard against Republicans in a Presidential year", but I cannot say whether an inverse is true in other states. What you see happening in polls will be what you get.

3. As always I will reject any polls from trade associations, campaigns, political parties, lobbyists, unions, ethnic associations, or advocacy groups. Until recently I might have done so on the principle that 'beggars can't be choosers', but I am not begging any more. I already have data for all 50 states.

4. I was tempted to expect that the Dakotas might trend D because they are reasonably-well-educated states... but Donald Trump is doing well there.

5. Texas is a gigantic surprise. Of course a 1% lead there by a Democratic nominee is both counter-intuitive and practically insignificant. So basically, don't make a bet that Hillary Clinton will win Texas unless it be a long shot.

6. The only real chance that I see for a Trump pick up from any Obama state from 2012 is Iowa. Usually Iowa votes much like Wisconsin, but this time it seems to be voting more like Nebraska or South Dakota.

7. The small margin (by usual standards) for Nebraska suggests that the Second Congressional District will be in play. Maine's Second Congressional District will not be a quick call, but I can reasonably expect that Hillary Clinton will get all four electoral votes from Maine.

8. Mississippi close? What is going on there?

9. Wyoming looks like the best state for Trump, and Maryland looks like the best state for Clinton.

Binary here so far.

[pbrower2a]

Now -- the tricky one, the three-way or four-way map.

Hillary Clinton (D) vs. Donald Trump ® vs. Gary Johnson (L):



I'm going with saturation for the raw vote for the leader. The percentage (3 for 30-39, 4 for ro-49, 5 for 50-59, 6 for 60-69...) will be the number for the saturation.

No internal number will be shown for any nominee who has at least 60% of the raw vote or has a lead of at least 8%. and at least 40% of the raw vote.  Otherwise I will show

the leader by color (white for a tie), the margin for the leader, and the amount for Johnson (maybe McMullen added should he become relevant).  

Note: Gary Johnson is in second place, above Donald Trump, in New Mexico. Jill Stein is at 10 (just under Johnson) in New Hampshire.

Numeric data here:

http://apps.washingtonpost.com/g/page/po...poll/2086/

[Eric the Green]

I'm not clear on what those numbers mean. Stein has 11% in Texas, and 1% in NH although you say she has 10%??

[pbrower2a]

On the upper (binary) map, the number is the electoral votes that a state has. On the three-way map, a number is not shown if the leader is above 50% or above 40% with a lead of at least 8% (4% is the margin of error, and 8% is tough to undo). I'm not showing Jill Stein... but I am showing by how much Hillary Clinton or Donald Trump leads the other. Following the slash I have the support for Johnson.

[pbrower2a]

The Washington Post has polled all fifty states, missing only the District of Columbia and the separately-voting Congressional Districts of Maine and Nebraska. With such a rich collection of polls, I can begin anew/ The binary map: Hillary Clinton (D) vs. Donald Trump ®:

Binary map.



Tie -- white

60% or more -- saturation 8
55-59.9%        --  saturation 6
50-54.9%        --  saturation 5
45-49.9%, lead 8% or more -- saturation 4
45-49.9%, lead 4-7.9% -- saturation 3
45-49.9%, lead 1-3.9%  -- saturation 2

Any lead with less than 45% will be considered unusable.  I might show the poll but not use it.

Three-way polling (I am not showing Jill Stein for now)



I'm going with saturation for the raw vote for the leader. The percentage (3 for 30-39, 4 for 40-49, 5 for 50-59, 6 for 60-69...) will be the number for the saturation.

No internal number will be shown for any nominee who has at least 50% of the raw vote or has a lead of at least 8%. and at least 40% of the raw vote.  Otherwise I will show the leader by color (white for a tie), the margin for the leader, and the amount for Johnson (maybe McMullen added should he become relevant).  

Note: Gary Johnson is in second place, above Donald Trump, in New Mexico. Jill Stein is at 10 (just under Johnson) in New Hampshire.

[pbrower2a]

Old polls compiled by PPP on a group that wishes to get the ninth Justice on the US Supreme Court.

Pennsylvania

47% Clinton (D)
42% Trump ®

http://weneednine.org/wp-content/uploads...lines1.pdf

New Hampshire

46% Clinton (D)
41% Trump ®

http://weneednine.org/wp-content/uploads...lines1.pdf

Iowa

45% Clinton (D)
43% Trump ®

http://weneednine.org/wp-content/uploads...lines1.pdf

What may be even more telling is that the gambit of Senate Majority Leader Mitch McConnell to ensure that Barack Obama does not get his appointee to the Supreme Court ratified no matter what does not help two Republican Senators get re-elected... and a third one, long-time incumbent Chuck Grassley, could be hurt should the Democrats succeed in using that against him.

Now a big and current poll of Florida by PPP for its own sake:

Trump 44
Clinton 43
Johnson 5
Stein 1
McMullin 1

Clinton 47
Trump 46

http://www.publicpolicypolling.com/main/...orida.html

Colorado President by Magellan Strategies on 2016-08-31[/url]

Summary: D: 41%, R: 36%, I: 16%, U: 7%

Poll Source URL: Full Poll Details

This is from a Republican-leaning pollster. No binary choice.




Hillary Clinton (D) vs. Donald Trump ®:





Tie -- white

60% or more -- saturation 8
55-59.9% -- saturation 6
50-54.9% -- saturation 5
45-49.9%, lead 8% or more -- saturation 4
45-49.9%, lead 4-7.9% -- saturation 3
45-49.9%, lead 1-3.9% -- saturation 2

Any lead with less than 45% will be considered unusable.




The three-way map:

Hillary Clinton (D) vs. Donald Trump ® vs. Gary Johnson (L):




I'm going with saturation for the raw vote for the leader. The percentage (3 for 30-39, 4 for ro-49, 5 for 50-59, 6 for 60-69...) will be the number for the saturation.

No internal number will be shown for any nominee who has at least 60% of the raw vote or has a lead of at least 8%. and at least 40% of the raw vote. Otherwise I will show

the leader by color (white for a tie), the margin for the leader, and the amount for Johnson (maybe McMullen added should he become relevant).

Note: Gary Johnson is in second place, above Donald Trump, in New Mexico. Jill Stein is at 10 (just under Johnson) in New Hampshire.

[pbrower2a]

Foreign policy is an essential area for the President. Aleppo is of course a war zone in Syria, and someone running for President who does not know that fails on that matter.

But Donald Trump has about everything wrong on foreign policy, anyway. I would never have expected to have gone to the right of the Republican nominee for President without having become a reactionary on any issue. That says a lot about Donald Trump.

In any event the default for American foreign policy is practically that of Ronald Reagan (and Obama's foreign policy is practically what Ronald Reagan got away with); it's now hard to see, except that anti-Communism has become moot, between Ronald Reagan and Barack Obama. Whether one likes Reagan foreign policy or not, Hillary Clinton has been riding it... and it is all that is now available as a sensible foreign policy. Hillary Clinton now owns Obama foreign policy for better or worse -- and at this stage, clearly to the benefit of her campaign. It is now practically inconceivable that the foreign policy of Barack Obama can fail before the middle of November.

On foreign policy, Hillary Clinton can hit Donald Trump from the Right.

[pbrower2a]

A critical demographic (Latinos) shows how it projects to vote in battleground states:


AZ: Clinton 71, Trump 17
CO: Clinton 69, Trump 17
FL: Clinton 63, Trump 27
NC: Clinton 73, Trump 14
NV: Clinton 71, Trump 14
OH: Clinton 61, Trump 22
VA: Clinton 70, Trump 20

400 Latino voters were polled in each state

http://www.latinodecisions.com/blog/2016...decisions/

In Florida, Cuban-Americans go for Trump -- but Puerto Ricans go heavily for Clinton.

...Not one of these states could go for Hillary Clinton without the Hispanic vote.

[pbrower2a]

States for which I have new polls include Arizona, California, Florida, Georgia, Michigan, Nevada, New Hampshire, Ohio, and Utah.


Hillary Clinton (D) vs. Donald Trump ®:




Tie -- white

60% or more -- saturation 8
55-59.9% -- saturation 6
50-54.9% -- saturation 5
45-49.9%, lead 8% or more -- saturation 4
45-49.9%, lead 4-7.9% -- saturation 3
45-49.9%, lead 1-3.9% -- saturation 2

Any lead with less than 45% will be considered unusable.




The three-way map:

Hillary Clinton (D) vs. Donald Trump ® vs. Gary Johnson (L):




I'm going with saturation for the raw vote for the leader. The percentage (3 for 30-39, 4 for ro-49, 5 for 50-59, 6 for 60-69...) will be the number for the saturation.

No internal number will be shown for any nominee who has at least 60% of the raw vote or has a lead of at least 8%. and at least 40% of the raw vote. Otherwise I will show

the leader by color (white for a tie), the margin for the leader, and the amount for Johnson (maybe McMullen added should he become relevant).

Note: Gary Johnson is in second place, above Donald Trump, in New Mexico. Jill Stein is at 10 (just under Johnson) in New Hampshire.

[pbrower2a]

In case anyone has any doubt -- I do not believe that Texas will be in play.

Polarization between the states may be shrinking some... one must admit that Barack Obama could certainly divide regions at the polls in the last two Presidential elections.

[pbrower2a]

The 7-point lead in a poll in Ohio looks like an anomaly, but it is outside ther margin of error -- well outside.

It could also be that the more that voters in a state get to know Donald Trump, the less that they like him. The more that he campaigns, the more people recognize how awful he is.

It is he and not the Republican Party, which seems to be getting an advantage in most Senate races that earlier seemed up for grabs.

[Eric the Green]

The states most likely to flip from R to D are Wisconsin, Illinois, Indiana (Bayh running again) and Pennsylvania. New Hampshire is also possible. Right now, PA looks dicey. Nevada is the state most likely to flip from D to R.