Donald Trump: polls of approval and favorability

[pbrower2a]

Lightest shades are for a raw total of votes that allows for a win with a margin of 5% or less; middle shades are for totals with allow for wins with 6 to 10% margins; deepest shades are for vote percentages that allow wins of 10% or more. Numbers are for the projected vote for Trump.   

100-DIS gives a reasonable ceiling for Trump in 2020 -- at least the most suitable one that I can think of. He may have won Iowa by nearly ten points, but he would be crushed there if an election were to be held there today. The recent poll suggests that the President has disappointed Iowa voters very much.

An assumption of a close race for President depends upon most states being close to their Cook PVI ratings. So how is that going?

Use green for a poll that diverges in favor of Trump's likely opponent, and orange for polls that  that diverge to the favor of Trump. Use light shades for divergence of 4% or less, medium for 5% to 8%, and dark shades for 9% or more.  For example, I show a ceiling of 48% for Trump in Alabama in a state that usually gives the average Republican a 14% edge in a 50-50 election. Trump would be reasonably expected to win Alabama by about 14% in a normal electon, but my estimate (100-DIS)  suggests that Trump would fare worse by about 12% for the average Republican in a 50-50 election nationwide.  
 
   


Eighteen states with recent polls... slightly more than one third of all states, If you can see a 50-50 election for the Presidency in 2020, then you see something that I don't see. I cannot predict that the President's biggest losses of support are in the Midwest or South.

The 2020  election will not be a 50-50 contest or even close. Even the historically 'average' opponent will defeat President Trump decisively.

[pbrower2a]

Missouri, PPP. Trump approval 48%, disapproval 47% -- slight edge for now for President Trump.

The "Show Me" state has some other issues with the GOP:

Mitch McConnell 23-54, 23 not sure
DREAM Act 61-27
Obamacare repeal, 45-37... keep but improve, 54 and repeal-and-replace 37

...Senator McCaskill has a slight but indecisive edge against her most likely Republican opponent. Add the usual 6% gain from an early match-up for an incumbent, and she wins.  

https://www.publicpolicypolling.com/wp-c...ry2018.pdf

My map may overstate the chance of Trump winning Missouri in 2020. It is intended to be charitable to any incumbent who has even the slightest edge in approval polling. My seat-of-the-pants prediction for Missouri in 2020 now has Missouri as an extreme tossup. If Missouri indicates anything, it is that the 2020 election will look much like the 2008 election in result.

Now for a real shocker -- Georgia! The state has typically been about even in approval and disapproval for Trump. That is over. This is an outlier... but when we see lots of other outliers, maybe we have to treat the outliers as norms. Trump is awful, and I pay less attention to the 36% approval than to the 59% disapproval.

This may reflect President Trump's recent use of the word $#!+hole to describe some countries whose peoples are melanin-rich. There are of course many melanin-rich people in Georgia, and racist talk that one might expect from a KKK leader or a neo-Nazi might be unsettling to many white people, too.   

Quote:Shock Poll From The Atlanta Journal Constitution/University of Georgia’s School of Public and International Affairs:

Georgia:

36.7% Approve
58.7% Disapprove

It's hard to believe, but in view of Alabama and Tennessee showing Trump about even, and Georgia differing from those two states by having one giant metro area instead of several smaller ones (Greater Memphis spills into Arkansas and Mississippi), I can almost think this reasonable. I must accept it.  The deviation from the assumption of a 50-50 election nationwide is about the same as for Alabama and Tennessee. Atlanta vs. Memphis/Nashville/Knoxville or vs. Birmingham/Mobile/Montgomery/Huntsville? That is the demographic difference between Georgia and either Alabama or Tennessee. Arguably Missouri, too, as St. Louis and Kansas City metro areas spill into Illinois and Kansas,  neither of which are swing states, respectively.

..........


Here is the 2017 map of Cook PVI.  With a 50-50 split of the popular vote, a PVI of D+6 (Connecticut) means that the Democrat can expect to get 56% of the popular vote from Connecticut and that a PVI of R+9 (Indiana) indicates that one can reasonably expect the Republican to win 59% of the popular vote in Indiana. Districts of Maine and Nebraska are shown as their Congressional districts vote for members of Congress. DC? I'm guessing that the Democrat is going to win about 90% of the vote in just about any election, so that is about D+40.  

Color and intensity will indicate the variance from a tie (ties will be in white) with  in a 50-50 election with blue for an R lean and red for a D lean. Numbers will be shown except in individual districts

int      var
2        1-4%
3        5-8%
5        9-12%
7        13-19%
9        20% or more







DC -- way out of reach for any Republican.

ME-01 D+8
ME-02 R+2

NE-01 R+11
NE-02 R+4
NE-03 R+27

(data from Wikipedia, map mine)
 I use 100-DIS as a reasonable ceiling for the Trump vote in 2020. Thus:




Lightest shades are for a raw total of votes that allows for a win with a margin of 5% or less; middle shades are for totals with allow for wins with 6 to 10% margins; deepest shades are for vote percentages that allow wins of 10% or more. Numbers are for the projected vote for Trump.   

100-DIS gives a reasonable ceiling for Trump in 2020 -- at least the most suitable one that I can think of. An assumption of a close race for President depends upon most states being close to their Cook PVI ratings. So how is that going?

Use green for a poll that diverges in favor of Trump's likely opponent, and orange for polls that  that diverge to the favor of Trump. Use light shades for divergence of 4% or less, medium for 5% to 8%, and dark shades for 9% or more.  For example, I show a ceiling of 48% for Trump in Alabama in a state that usually gives the average Republican a 14% edge in a 50-50 election. Trump would be reasonably expected to win Alabama by about 14% in a normal election, but my estimate (100-DIS)  suggests that Trump would fare worse by about 12% for the average Republican in a 50-50 election nationwide.  
 
   


Twenty states with recent polls... now two-fifths (40%) of all states, and 237 electoral votes. If you can see a 50-50 election for the Presidency in 2020, then you see something that I don't see. I cannot predict that the President's biggest losses of support are in the Midwest or South.

The 2020  election will not be a 50-50 contest or even close. Even the historically 'average' opponent will defeat President Trump decisively.

Of course I would love to see polls of Arizona, Florida*, Georgia, Ohio, and Pennsylvania... and maybe Texas. Trump stands to lose states that Republicans simply do not lose in Presidential elections. Note that 100-DIS is an estimate of the ceiling for the President in a re-election bid...
it is tough to win a state in which one's disapproval rating is above 50%, although Obama did come back from a poll of 43-51 in 2010 in Ohio, only to subsequently win the state. But Obama is one of the slickest campaigners in American history. Trump is not Obama, which is a severe understatement.  

*A Florida poll from October was awful for Trump.

[pbrower2a]

The old map (but updated!) showing approvals for all 50 states, but nothing to do with Cook PVI:

This approval map shows  electoral votes to the states on the approval map.



Trump approval, net positive

55% or higher
50-54%
44-49%

Ties are in white.

Trump approval, net negative

46-49%
40-45%
39% or lower

But raw disapproval numbers appear instead  of electoral votes here:




Disapproval (net negative for Trump) :

55% or higher
50-54%
44-49%

Ties are in white.

(net positive for Trump)
46-49%  
41-45%
40% or lower

The near-ties in Alabama, Arkansas, Missouri, Mississippi; and Tennessee;  the edge that I see against Trump in Louisiana; and his abysmal approval ratings in Georgia all suggest a collapse of Trump support within the Mountain and Deep South. This is before I discuss such states as Florida, North Carolina, and Virginia  that are or have been marginally "Southern" in their politics. This is more recent (no earlier than October) than what I see for thirty states for which I have not shown contrasts between 'normal' results in a 50-50 election according to their Cook

*((((((((( CV&U

(sorry, the cat jumped upon my keyboard)

PVI ratings in some more recent exercises.  This almost suggests a reversion to the situation with Carter in 1976, except that the demographic patterns of voting suggest something different.

The election of Senator Doug Jones suggests that a pattern of educated white suburbanites drifting away from the GOP in 2006, 2008, and 2012 in the North and West that did not happen in the South in those elections may be happening in the South. Jones did well among the presumable 'high-information' voters in such places as Huntsville and Tuscaloosa in December. To be sure, that is but one election, and any trend can reverse quickly.

Meanwhile, things are not going well for Trump 'Up North'. One poll of Iowa showed disapproval for Trump at 60% and another showed disapproval at 51%; were I to make a guess than to rigidly follow my model (which I must do for the sake of consistency) his disapproval in Iowa is probably about 55%, which suggests that Trump would get no more than about 45% of the vote in Iowa. That is about what John McCain got in 2008. Michigan and New Hampshire, the two closest states for Trump in 2016, show intense disapproval for President Trump.

Sure, I'd like more data. Texas, which is marginally "Southern" because it straddles cultural regions of America and is not a region in itself despite its size, might confirm or deny my conjectures. I have seen polls of Texas in which Trump is underwater, and I would not be surprised to see such again.  I still want to see new polls of Ohio and Pennsylvania, for which I have nothing, and for Arizona and Nevada, for which polling is old. Not that there are any big electoral prizes in the states of the High Plains and Badlands from Montana to Oklahoma, I have a big hole for recent polling in such states.

Would you be surprised to see South Carolina underwater for Trump at this stage? I wouldn't.

OK, you say -- maybe the Georgia poll is an exaggeration. Tough luck on that. It follows a shocking statement by the President with racist overtones about some countries and people from them. But in my experience in watching elections, shocks stick. Never mind that President Trump could hardly do worse with the black vote in 2020 than he did in 2016 -- the poll shows white people turning against him. Racism is a dying phenomenon in America, and the President should have kept his mouth shut in the presence of people who could confirm the "$#!+hole" remark. He can't quote lie his way out of it.  

At this point I see the 2020 election looking like one in which the Democrat has a chance of winning every state that any Democratic nominee for President since 1976, which implies a vote for even an average Democratic nominee that combines Carter and Obama wins while picking off Arizona.

[pbrower2a]

CBS News' unique take:

Trump's first year

Quote:One year into Donald Trump's presidency, Americans feel more positive about the economy but not as good about the state of the country overall -- and the latter is closely tied to views of the president.

By a two to one margin, more say that the country is doing well economically than that it isn't. But three in four Americans say the country is divided, six in 10 don't have much confidence in the U.S. political system and six in 10 say racial tensions have increased. The president's strongest backers believe things are going well, but his opponents -- who have grown increasingly opposed to the president over the year -- say things are not. Overall, the number of Americans who say having Donald Trump as president makes them feel "pessimistic" is higher than it was a year ago.

So what are "believers", "conditionals", "curious", and "resisters"?

Quote:A year ago, this study began analyzing four groups: the strongest of Trump backers (whom the study labeled "believers"); another set of those who support the president on the condition that he delivers what they want (the "conditionals"); a group opposing the president for now but willing to back him if things change, (the curious) and those who are firmly opposed (whom the study labels the "resisters.")

Overall, the movement we have seen over the year is a slow shift away from Mr. Trump, and we have that movement across the four groups: the believers, the conditionals, the curious and the resisters.

My comment: nobody else seems to use this nomenclature, but this method says something.




https://www.cbsnews.com/news/nation-trac...in-office/

Quote:The CBS News 2018 Nation Tracker is conducted by YouGov using a nationally representative sample of 2,164 U.S. adults between January 10–12, 2018. All respondents were recontacted January 12–13, 2018 for two follow up questions about the President's recent comments on immigration, and 1,654 responded. The margin of error was +/-  2.6.

[pbrower2a]

Trump approval rating in West Virginia (arguably one of the strongest R states in Presidential elections now):


Approve: 51 (+3)
Disapprove: 48

Down from +21 in January 2017.


 Tax Bill approval ("will it help average West Virginians?")
Help - 38
Hurt - 27
No effect - 19

http://www.wsaz.com/content/news/WVa--469737723.html


The biggest deviation from the assumption that the 2020 election will be a 50-50 election that I have sen yet.




Color and intensity will indicate the variance from a tie (ties will be in white) with  in a 50-50 election with blue for an R lean and red for a D lean. Numbers will be shown except in individual districts

int      var
2        1-4%
3        5-8%
5        9-12%
7        13-19%
9        20% or more







DC -- way out of reach for any Republican.

ME-01 D+8
ME-02 R+2

NE-01 R+11
NE-02 R+4
NE-03 R+27

(data from Wikipedia, map mine)
 I use 100-DIS as a reasonable ceiling for the Trump vote in 2020. Thus:


 

Lightest shades are for a raw total of votes that allows for a win with a margin of 5% or less; middle shades are for totals with allow for wins with 6 to 10% margins; deepest shades are for vote percentages that allow wins of 10% or more. Numbers are for the projected vote for Trump.   

100-DIS gives a reasonable ceiling for Trump in 2020 -- at least the most suitable one that I can think of. An assumption of a close race for President depends upon most states being close to their Cook PVI ratings. So how is that going?

Use green for a poll that diverges in favor of Trump's likely opponent, and orange for polls that  that diverge to the favor of Trump. Use light shades for divergence of 4% or less, medium for 5% to 8%, and dark shades for 9% or more.  For example, I show a ceiling of 48% for Trump in Alabama in a state that usually gives the average Republican a 14% edge in a 50-50 election. Trump would be reasonably expected to win Alabama by about 14% in a normal election, but my estimate (100-DIS)  suggests that Trump would fare worse by about 12% for the average Republican in a 50-50 election nationwide.  
 
   


Twenty-three states with recent polls... now nearly half (46%) of all states, If you can see a 50-50 election for the Presidency in 2020, then you see something that I don't see.

The 2020  election will not be a 50-50 contest or even close. Even the historically 'average' opponent will defeat President Trump decisively.

Of course I would love to see polls of Arizona, Florida*, Georgia, Ohio, and Pennsylvania... and maybe Texas. Trump stands to lose states that Republicans simply do not lose in Presidential elections. Note that 100-DIS is an estimate of the ceiling for the President in a re-election bid...

it is tough to win a state in which one's disapproval rating is above 50%, although Obama did come back from a poll of 43-51 in 2010 in Ohio, only to subsequently win the state. But Obama is one of the slickest campaigners in American history. Trump is not Obama, which is a severe understatement.

[pbrower2a]

OHIO 43/52

http://www.cleveland.com/open/index.ssf/...y_nar.html
[/quote]

The poll was commissioned by the 1984 Society, a nonprofit, bipartisan group of former Ohio Senate employees and senators who wanted good independent polling for the political community. It was conducted via phone interviews by Fallon Research with a sample size of 801 and a margin of error of 3.5 percentage points.

.....

Quote:OHIO STATEWIDE SURVEY
1/16/18 - 1/19/2018 N=801, +/- 3.46% General Election Voters
(percentages may not add up to 100% due to rounding)

Generally speaking, would you say that Ohio is going in the right direction or has it gotten off onto the wrong track?

54.2% Right direction 24.2 Wrong track 5.3 Mixed/both (volunteered) 16.3 Unsure/no answer
Generally speaking, do you approve or disapprove of the job being done by President Donald Trump? If approve or disapprove, interviewer follow-up: Would you say that you strongly [approve/disapprove], or just somewhat [approve/disapprove]?
43.4% TOTAL APPROVE
31.3 Strongly approve 12.1 Somewhat approve
51.6% TOTAL DISAPPROVE
11.1 Somewhat disapprove 40.5 Strongly disapprove 4.9% Unsure/no answer

Generally speaking, do you approve or disapprove of the job being done by Ohio Governor John Kasich? If approve or disapprove, interviewer follow-up: Would you say that you strongly [approve/disapprove], or just somewhat [approve/disapprove]?
57.2% TOTAL APPROVE
20.9 Strongly approve 36.3 Somewhat approve
29% TOTAL DISAPPROVE
15.6 Somewhat disapprove 13.4 Strongly disapprove 13.9% Unsure/no answer

Ohio has offered no poll since the 2016 election, and it is the bellwether state. It looks as if it will go against Donald Trump in 2020. Note that my estimate for Trump is a ceiling based solely upon disapproval, which is a conservative estimate.

Color and intensity will indicate the variance from a tie (ties will be in white) with in a 50-50 election with blue for an R lean and red for a D lean. Numbers will be shown except in individual districts

int var
2 1-4%
3 5-8%
5 9-12%
7 13-19%
9 20% or more







DC -- way out of reach for any Republican.

ME-01 D+8
ME-02 R+2

NE-01 R+11
NE-02 R+4
NE-03 R+27

(data from Wikipedia, map mine)
I use 100-DIS as a reasonable ceiling for the Trump vote in 2020. Thus:




Lightest shades are for a raw total of votes that allows for a win with a margin of 5% or less; middle shades are for totals with allow for wins with 6 to 10% margins; deepest shades are for vote percentages that allow wins of 10% or more. Numbers are for the projected vote for Trump.

Note -- if Trump is underwater in the polling, then the results come out in pink.

100-DIS gives a reasonable ceiling for Trump in 2020 -- at least the most suitable one that I can think of. An assumption of a close race for President depends upon most states being close to their Cook PVI ratings. So how is that going?

Use green for a poll that diverges in favor of Trump's likely opponent, and orange for polls that that diverge to the favor of Trump. Use light shades for divergence of 4% or less, medium for 5% to 8%, and dark shades for 9% or more. For example, I show a ceiling of 48% for Trump in Alabama in a state that usually gives the average Republican a 14% edge in a 50-50 election. Trump would be reasonably expected to win Alabama by about 14% in a normal election, but my estimate (100-DIS) suggests that Trump would fare worse by about 12% for the average Republican in a 50-50 election nationwide.




Fully half (25) states now show recent polls... If you can see a 50-50 election for the Presidency in 2020, then you see something that I don't see. Wishful thinking and two dollars + tax will get you a two-dollar cup of coffee. Two dollars + tax will get you the same cup of coffee even without wishful thinking.

The 2020 election will not be a 50-50 contest or even close. Even the historically 'average' opponent will defeat President Trump decisively.

Of course I would love to see polls of Arizona, Florida*, Georgia, Ohio, and Pennsylvania... and maybe Texas. Trump stands to lose states that Republicans simply do not lose in Presidential elections. Note that 100-DIS is an estimate of the ceiling for the President in a re-election bid...

it is tough to win a state in which one's disapproval rating is above 50%.

*A Florida poll from October was awful for Trump.

[pbrower2a]

At this point i cannot see Donald Trump winning re-election except through

(1) a division within the Democratic party
(2) a rigged election, or
(3) miracles for Donald Trump that violate about everything that makes statistical sense.

[pbrower2a]

Gallup data from all polls in 2017 (average assumed in mid-July) in the form of a map:



*Approval lower than disapproval in this state

for barely-legible numbers for DC and some states -- CT 37 DC 11 DE 42 HI 40 MD 35 RI 38

Lightest shades are for a raw total of votes that allows for a win with a margin of 5% or less; middle shades are for totals with allow for wins with 6 to 10% margins; deepest shades are for vote percentages that allow wins of 10% or more. Numbers are for the projected ceiling for the vote for Donald Trump in 2020, which is  100-DIS. Disapproval is far stickier downward, and 100-DIS is far sticker upward than approval is either way. Yes, if you support the President, it really is that bad and you might as well get used to that. Yes, disapproval can increase, and that will mess up campaigning and canvassing for your side. What was true for Obama in 2012 will also be true for Trump; the numbers for Trump are far worse on the whole for him than they were for Obama eight years ago.

Here is more recent data. Note that I add back a poll of Florida from October, as this is newer than the Gallup data.
 


DC -- way out of reach for any Republican.

ME-01 D+8
ME-02 R+2

NE-01 R+11
NE-02 R+4
NE-03 R+27

(data from Wikipedia, map mine)
I use 100-DIS as a reasonable ceiling for the Trump vote in 2020. Thus:







Now here is polling from October or later (I am adding a poll of Florida because this is newer than the data whose age averages about six months now). Lightest shades are for a raw total of votes that allows for a win with a margin of 5% or less; middle shades are for totals with allow for wins with 6 to 10% margins; deepest shades are for vote percentages that allow wins of 10% or more. Numbers are for the projected vote for Trump.

Note -- if Trump is underwater in the polling, then the results come out in pink.      

Note that polls in the Mountain (Appalachia and the Ozarks) and Deep South (Louisiana through Georgia) all show drops in approval ratings and rises in disapproval ratings for the President from the average of 2017. I say nothing about South Carolina because I have no recent poll of the state; Virginia has more in common with Michigan in politics than it has with any other former Confederate state except perhaps Florida; Florida has huge numbers of Yankee retirees who brought their politics with them from Up North and people of Central American and Caribbean origin; Texas straddles the Deep South, the Midwest, and the Southwest -- and it is difficult to determine what region five of its six largest cities are in (OK, El Paso is clearly Southwestern, being closer to San Diego than to Beaumont or Texarkana). I do not have a recent approval poll of Texas not from an advocacy group).

[pbrower2a]

The Gallup data and source?

Trump Approval: State By State for 2017

State: Approve-Disapprove

National: 38-56

West Virginia: 61-35
North Dakota: 57-39
Wyoming: 57-36
South Dakota: 54-42
Alabama: 53-42
Oklahoma: 53-40
Idaho: 53-42
Montana: 52-45
Kentucky: 51-45
Tennessee: 50-44
Arkansas: 50-44
Alaska: 50-44
Nebraska: 49-47
Louisiana: 49-45
Kansas: 48-47
South Carolina: 48-46
Mississippi: 48-46
Utah: 48-47
Missouri: 47-48
Ohio: 45-50
Indiana: 44-51
Iowa: 43-52
Maine: 42-55
New Hampshire: 42-56
Pennsylvania: 42-53
Nevada: 42-53
Wisconsin: 41-55
Georgia: 41-53
Florida: 41-53
Arizona: 41-53
Michigan: 40-55
North Carolina: 40-54
Texas: 39-54
Minnesota: 37-58
Virginia: 37-57
Colorado: 37-59
Delaware: 36-58
Oregon: 36-59
New Mexico: 35-59
New Jersey: 34-61
Washington: 34-60
Illinois: 33-62
Rhode Island: 32-62
Connecticut: 31-63
New York: 30-63
Maryland: 30-65
California: 29-65
Hawaii: 29-60
Massachusetts: 27-68
Vermont: 26-69
District of Columbia: 6-88

Source

[Eric the Green]

A small TRUMP BUMP is happening. Better economic news and a reasonable state of the union speech might have helped, although the Nunes memo is causing more trouble.


Date Approve Disapprove No opinion
%
%
%
2018
2018 J.29-F. 4 40 57 3
2018 Jan 22-28 38 57 5
2018 Jan 15-21 36 59 5
2018 Jan 8-14 38 57 5
2018 Jan 1-7 37 58 4
http://news.gallup.com/poll/203198/presi...trump.aspx

[pbrower2a]

Ideally I would have a logarithmic scale. But the President has lately touted the value of the stock market,  and it won't be long before people ask "What have you done lately?"

[pbrower2a]

A week after the SOTU speech, Quinnipiac has Trump at 40 approve and 55 disapprove. For him that is an improvement. Gal;lup, which has had similar results, seems to go along.

It is still awful for an incumbent President at this stage.

[pbrower2a]

On the Big Parade planned for November 11, 2018:

Quote:As the Pentagon continues to work on parade options to present to President Donald Trump, an overwhelming number of Military Times readers have weighed in: Don’t have one.

The informal poll was launched Wednesday after news reports that Trump had requested a military parade and that the Pentagon was working on parade options for him.

As of Thursday afternoon, more than 51,000 readers had responded. The majority, 89 percent, responded “No, It’s a waste of money and troops are too busy.”

The other 11 percent responded “Yes, it’s a great opportunity to show off U.S. military might.”
On Thursday, Pentagon press secretary Dana White said any parade plans were in the very beginning stages, and that the Pentagon had tapped the Army to lead the effort.

“We are looking at several different options right now,” White said. “The Army is the executive agent. But we don’t have those options yet. Its still in nascent stages and when we have those options we’ll provide that to the White House and the president will decide.”

https://www.militarytimes.com/news/your-...no-parade/

[pbrower2a]

Updates for Florida and Iowa; first  poll of Oregon (positive  or negative ratings); favorability  polls of Illinois and Texas:


Cook PVI ratings:

Color and intensity will indicate the variance from a tie (ties will be in white) with  in a 50-50 election with blue for an R lean and red for a D lean. Numbers will be shown except in individual districts

int      var
2        1-4%
3        5-8%
5        9-12%
7        13-19%
9        20% or more







DC -- way out of reach for any Republican.

ME-01 D+8
ME-02 R+2

NE-01 R+11
NE-02 R+4
NE-03 R+27

(data from Wikipedia, map mine)

Here  is the Gallup data for 2017. Figuring that this is an average from early February to late December, I will have to assume an average date of July 15 or so for the polling data. This is now rather old data, and in some cases obsolete. For example, I see Trump support cratering in the Mountain and Deep South. the Mountain South and Deep South are going back to a populist phase (the South has typically oscillated between the  two) or whether The Donald is beginning to appear as a bad match for either part of the South.  This data (or later polling) is not  intended to show anything other than how support appears at some time or at an average of times. As a general rule, new polling supplants even better old  polling.

So here  is the Gallup polling with a number  of 100-DIS reflecting what I consider the ceiling for Donald Trump. This is lenient to the extent that I assume that he can pick up most undecided voters but recognizes that undoing disapproval at any stage requires miracles. By definition, miracles are unpredictable.

 Gallup data from all polls in 2017 (average assumed  in mid-July):



*Approval lower than disapproval in this state

for barely-legible numbers for DC and some states -- CT 37 DC 11 DE 42 HI 40 MD 35 RI 38

Lightest shades are for a raw total of votes that allows for a win with a margin of 5% or less; middle shades are for totals with allow for wins with 6 to 10% margins; deepest shades are for vote percentages that allow wins of 10% or more. Numbers are for the projected vote for Trump.

This is polling from October or later, and I will be adding a poll of Florida from October because it is newer than the Gallup polling data.

 I use 100-DIS as a reasonable ceiling for the Trump vote in 2020. Thus:




Lightest shades are for a raw total of votes that allows for a win with a margin of 5% or less; middle shades are for totals with allow for wins with 6 to 10% margins; deepest shades are for vote percentages that allow wins of 10% or more. Numbers are for the projected vote for Trump.

Note -- if Trump is underwater in the polling, then the results come out in pink.    

Variation from PVI (polls from October 2017 and later):

[Eric the Green]

Trump Makes Gains Ahead of Midterms but Risks Remain for GOP
Political experts say rise in net approval ratings among voters reflects in part a solid U.S. economy

BY CAMERON EASLEY
February 13, 2018

https://morningconsult.com/2018/02/13/tr...tate-poll/

While Trump’s net approval rating remained 7 points underwater in January — with 44 percent of registered voters approving of his job performance and 51 percent disapproving — the number represents a 2-point bump compared with his rating in September and a 4-point bump from October and November.

Much of that gain was driven by an increase of enthusiastic support from Republican voters: Forty-eight percent of those voters in January said they strongly approve of Trump, compared with 43 percent in September.

Some experts suggest Trump’s improved standing with the public, and Republican voters in particular, was partly due to rosy economic indicators during this period — from stock market gains to strong employment figures — and Republican lawmakers’ overhaul of the country’s tax code, despite the public’s mixed response to certain parts of the legislation.

The improvement in Trump’s approval rating since the fall “most fundamentally reflects the economy,” Allan Lichtman, a political historian at American University, said in a Feb. 4 interview. A secondary factor is that Trump “did finally get something done in Congress with the tax bill.”

Daniel Hopkins, a political science professor at the University of Pennsylvania, said the passage of the tax bill has eased GOP voters’ concerns “that the Trump administration wouldn’t be able to advance a traditional GOP agenda.”

“Any lingering worry that Trump wouldn’t be loyal to the GOP’s priorities is likely to be history,” he said in a Feb. 5 email.

The improvement in Trump’s popularity doesn’t mean Republican lawmakers shouldn’t be concerned about the midterm elections, Lichtman noted.

“I don’t think there’s much that can happen between now and November that’s going to dampen the enthusiasm of the anti-Trump vote,” he said.



See all approval and disapproval % in all 51 states at the website.

[pbrower2a]

More on an Iowa poll because it could have put my model in question.

Quote:Trump approval: 44/51 (-7)

In the 2020 election, do you anticipate you will definitely vote for President Trump, consider voting for another candidate, or definitely vote for another candidate?
(Based on likely voters in the 2020 general election; n=656. MoE = ± 3.8 percentage points.)

26% Definitely Trump
20% Consider voting for another candidate
48% Definitely vote for another candidate
5% Not sure


The Iowa Poll, conducted Jan. 28-31 for the Des Moines Register and Mediacom by Selzer & Co. of Des Moines, is based on telephone interviews with 801 Iowans ages 18 or older. Interviewers with Quantel Research contacted households with randomly selected landline and cell phone numbers supplied by Survey Sampling International. Interviews were administered in English. Responses were adjusted by age and sex to reflect the general population based on recent census data.

Questions based on the sample of 801 Iowa adults have a maximum margin of error of plus or minus 3.5 percentage points. This means that if this survey were repeated using the same questions and the same methodology, 19 times out of 20, the findings would not vary from the true population value by more than plus or minus 3.5 percentage points. Results based on smaller samples of respondents — such as by gender or age — have a larger margin of error.

https://www.desmoinesregister.com/story/...316063002/

It does not change the map, but I see that my assumption that "disapproval is a way of saying I will not vote for him" 51% disapprove, and 48%  'will definitely vote for another candidate'. Iowa is close to the national mean, and it looks as if 2016 was a fluke win for Trump in Iowa (as in Michigan, Pennsylvania, and Wisconsin as well). It would not take many of the 'I would consider voting for another candidate' to give the Democratic nominee a majority in an essentially binary election, or the plurality in the event that there should be a significant challenger to Donald Trump on the Right side of the political spectrum.

Guessing how this relates to several thresholds of the potential vote,

1. 26%  of the electorate is already certain that it will vote for Trump in 2020.

26% will definitely vote for Trump, which is an absolute floor.

38% of Iowans voted for Goldwater in 1964, and 40.5% of Iowa voters went for McGovern in 1972. That is a reasonable floor for a politician rejected nationwide, as that is close to the percentage of the national vote that those two got. But Trump is an incumbent. The two biggest failures as President as shown by their electoral defeats in binary races (Ross Perot probably handed the 1992 election to Bill Clinton on silver platter) were Hoover in 1932 (who got close to 40% of the vote) and Carter in 1980 (who got 38% of the binary vote, but there was John Anderson getting 8% of the Iowa vote). Almost any Presidential nominee will get about 38% of the vote nationwide even after a poor performance as President or being a weak challenger to an incumbent who has been somewhat successful.

44% is his approval.  That has been as low as 39% with this pollster.

46% either will definitely vote for Trump or might. That obviously will not be enough.

100-DIS is my estimate of a ceiling for the President: 49%. Sure, he can win at that level (or at 46%) if the left side of the electorate has a strong third-party or independent candidate. I cannot yet rule such out, but I see no indication of such already. It is far more likely that someone will challenge Donald Trump as a 'genuine conservative' or someone shorn of demagoguery and ethnic bigotry. Yes, one can have conservatism without homophobia, misogyny, racism, religious bigotry, and anti-intellectualism, but Republicans have no viable alternatives to Trump should he run in 2020. 2024? Too late for this time.  My model crudely predicts that the Democratic nominee will get at least 51% of the Iowa vote base3d on the assumption that those who disapprove will vote for someone else.

This is not to say that the President will not do worse in Iowa. If 20% would consider voting for someone else, then I would have to split that at best (for a Democrat) about 3 to 1, as that part of the electorate is clearly to the right of center. But if even 75% of those considering voting for someone else still go to Trump, then the President ends up with about 41% of the Iowa vote.... which is just a bit better than Carter did in 1980 in his troubled bid for re-election.

[pbrower2a]

Gallup tracker --  it looks as if we are back to the "old Trump".

[pbrower2a]

Updates involving Florida and South Carolina:

South Carolina, for which I have seen no polls for nearly a year:

Quote:South Carolina: Winthrop University, Feb 17-25

Approve 42  (no change from last Spring)
Disapprove 50 (+3)

Approval ratings for Governor Henry McMaster ® 47-25
Senator Lindsey Graham ® 38-48
Senator Tim Scott ® 53-28
Congress of the United States (majority R in both Houses)

11-78 overall; 7-88 Democrats, 20-69 Republicans

This is consistent with the most recent polls showing support of President Trump in near collapse in the  Deep and Mountain South.


Approval




55% or higher dark blue
50-54% medium blue
less than 50% but above disapproval pale blue
even white
47% to 50% but below disapproval pale red
42% to 46% medium red
under 42% deep red

100-Disapproval




55% or higher dark red
50% to 54% or higher but not tied medium red
50% or higher but negative pale red
ties white
45% or higher and positive pale blue
40% to 44% medium blue
under 40% deep blue


Nothing from before November. Polls from Alabama, New Jersey, and Virginia are exit polls from 2017 elections. 

...at this point (as I just got a poll of South Carolina) I could see President Trump losing almost as many 'former secessionist' states in a re-election bid as Jimmy Carter did. Sure, I make this conjecture two and a half years before the 2020 election, and Donald Trump could hardly be more different from Jimmy Carter. Carter won all but one former Confederate state in 1976 (the exception was Virginia), and it is mere coincidence that Donald Trump won all but one former Confederate state in 2016 (the exception again is Virginia). In 1980 Jimmy Carter lost every former Confederate state except Georgia, his home state. At this point I see a possibility of Donald Trump losing almost as badly in the South. Southerners do not like obnoxious d@mnyankees, and Donald Trump is about as obnoxious a d@mnyankee as there is. I lived for seventeen years in Texas and nearly two in Arkansas, so I know about what I speak. They just didn't get to see him often enough to make that judgment until he got elected.

In the three states (New York, New Jersey, and Connecticut) in which he is best known he got some of his worst results in the Presidential election. But it is enough to say that the 42% approval in Florida and the 40% approval in Michigan indicate that he will lose in his re-election bid unless he has some miracles working for him between now and November 2020.

Cook PVI against which I make comparisons to the baseline of a 50-50 election for the Presidency, a reasonable assumption since 2000:
 
Color and intensity will indicate the variance from a tie (ties will be in white) with  in a 50-50 election with blue for an R lean and red for a D lean. Numbers will be shown except in individual districts

int      var
2        1-4%
3        5-8%
5        9-12%
7        13-19%
9        20% or more



Cook PVI assumes a 50-50 Presidential election, reasonable since 2000 because except for the 2008 Presidential election all such exception, those five all were basically even for almost the entire electoral season. One can use polling to predict whether the next Presidential election will be a 50-50 proposition, and if not, how far the likely reality diverges from that assumption.

For DC (not measured) and Congressional districts that vote independently of states, I have common sense for Dee Cee and the congressional votes for those districts.

DC -- way out of reach for any Republican.

ME-01 D+8
ME-02 R+2

NE-01 R+11
NE-02 R+4
NE-03 R+27

(data from Wikipedia, map mine)

............

Variation from PVI (polls from October 2017 and later):



Orange implies that President Trump projects to do better than Cook PVI based on 100-DIS. In Minnesota I have a 49-47 poll with which to work, and people in that thread tell me that the pollster who got those results is suspect. So the President is doing 2% better in California in accordance with 100-DIS than Cook PVI suggests. Not significant, obviously, because that is the difference between losing the Golden State 60-40 instead of 62-38.

This is likely the last that you will see of my  analysis of polling based on deviation from Cook PVI. I think we can be assured that President Trump is doing worse, in general in  polling, than something consistent with a 50-50 split of the popular vote.

Oh well, I have at least made this analysis more concise. With South Carolina offering a poll, I now have every state in the former Confederacy. My interpretation of data from the South is that President Trump is beginning to seem like an obnoxious d@mnyankee.

The sea of green suggests at this time that Donald Trump will face an electorate that will split against him. Green, and even very dark shades of green, do not indicate that the President will lose any particular state. But look at all states that President Trump won by 10% or less, and except for Pennsylvania  (no recent polling) and Texas -- the measure of 100-disapproval suggests that he will lose. Disapproval is stickier than approval. A hint: Obama won only one state in which his disapproval polling ever went above 50% (he got to 51% in Ohio, even if such is 'just barely') before the 2012, and he barely won Ohio. Obama was a far better campaigner than Trump and has been far less erratic in his behavior. Based on this data I predict that even a mediocre Democratic nominee wins at least a near-landslide against Donald Trump. To be sure, the optimal predictor is a person-to-person match, but because about a dozen Democrats have a chance of being the 2020 nominee for President such is unavailable.

[pbrower2a]

Now at 31 states, with the addition of Nebraska and Rhode Island:


Nebraska may have only five electoral votes, and except for one instance all of its electoral votes (Second Congressional District, basically Greater Omaha inside Nebraska in 2008) have gone for the Republican nominee for President after the LBJ landslide of 1964. But Nebraska can be a big political story in 2020 -- and even 2018.

The incumbent Governor and Senator Deb Fischer (who flipped a US Senate seat from D to R in 2012, not a good year for Republicans) have their work cut out to get re-elected.

Approval, Governor Pete Ricketts (R, incumbent) 37-40; deserves to be re-elected 39-42  
Approval, Senator Deb Fischer (R, incumbent) 34-42; deserves to be re-elected 34-42

Trump approval overall, 46-44

Source

But the three districts vote separately as in only one other state (Maine)

NE-01 45-46 (eastern Nebraska except for greater Omaha. including Lincoln)
NE-02 38-54 (greater Omaha within Nebraska)
NE-03 55-34 (central and western Nebraska, including Scottsbluff and Grand Island)

With the recognition that this poll has large numbers of undecided, perhaps because Nebraska has never been a pivotal state in any Presidential election or in the composition of Congress, it may not be the sort of state that pays much attention to elections of any kind.

For a state that is usually reliably Republican in its voting patterns, this is a very bad sign for Republicans in 2018 and for President Trump in 2020. Not very often does Nebraska say much about the national political scene, but on the whole this is a very bad sign for Republicans. Of courser, NE-03 is one of the surest electoral votes for a Republican, and if a Republican nominee for President got only 20 or fewer electoral votes, he would get NE-03. But when NE-01 and the state at large are shaky for the President, let alone the incumbent Governor and Senator up for re-election in 2018, the GOP is in really bad shape nationwide.

I do not believe that Nebraska's political culture has undergone a transformation; I simply see Donald Trump as a great disappointment in the Cornhusker State.  

Note that I have needed to alter the legend to accommodate a bare lead with 45%.
 
Approval:




55% or higher dark blue
50-54% medium blue
less than 50% but above disapproval pale blue
even white
46% to 50% but below disapproval pale red
42% to 45% medium red
under 42% deep red

States and districts hard to see:

RI 30
NE-01 45
NE-02 38
NE-03 55

100-Disapproval




55% or higher dark blue
50% to 54% or higher but not tied medium blue
50% or higher but negative pale blue
ties white
45% or higher and positive pale red
40% to 44% medium red
under 40% deep red

States and districts hard to see:

 RI 30
NE-01 55
NE-02 46
NE-03 66


Nothing from before November. Polls from Alabama, New Jersey, and Virginia are exit polls from 2017 elections.  






States hard to see:

RI 30


Nothing from before November. Polls from Alabama, New Jersey, and Virginia are exit polls from 2017 elections.

[pbrower2a]

a poll on issues if not on the President

March 6, 2018 - U.S. Voters Oppose Steel, Aluminum Tariffs, Quinnipiac University National Poll Finds; Voters Oppose Armed Teachers, Back Armed Security 6-1


PDF format
Additional Trend Information
Sample and Methodology detail

American voters oppose 50 - 31 percent tariffs on steel and aluminum, and disagree 64 - 28 percent with President Donald Trump's claim that a trade war would be good for the U.S. and easily won, according to a Quinnipiac University National Poll released today.

Every listed party, gender, education, age and racial group oppose steel and aluminum tariffs, except Republicans, who support tariffs by a lackluster 58 - 20 percent and white voters with no college degree, who are divided with 42 percent supporting tariffs and 40 percent opposed, the independent Quinnipiac (KWIN-uh-pe-ack) University Poll finds.

American voters oppose these tariffs 59 - 29 percent, if these tariffs raise the cost of goods they buy. The tariffs will be good for American jobs, 26 percent say, while 36 percent say tariffs will be bad for jobs and 24 percent say the tariffs will have no impact on jobs.

American voters disapprove 54 - 34 percent of the way President Trump is handling trade. Only Republicans and white voters with no college degree approve.

That National Rifle Assn. (NRA) has too much influence over politicians, 60 percent of American voters say, while 7 percent say it has too little influence and 26 percent say it has the right amount of influence.

Republicans in Congress are afraid of the NRA, voters say 54 - 42 percent.

Democrats in Congress are afraid of the NRA, voters say 49 - 45 percent.

President Trump is not afraid of the NRA, voters say 65 - 31 percent.

"Tariff, smariff, say voters who see punishing other countries on imports will do more harm at home," said Tim Malloy, assistant director of the Quinnipiac University Poll.

"Who's afraid of the NRA? 535 Senators and Congress people. "Who's not afraid of the NRA? The man with the most to lose to one of Washington's most powerful lobbies, President Donald Trump."

American voters oppose 58 - 40 percent allowing teachers and school officials to carry guns on school grounds.

But voters support 82 - 14 percent having armed security officers in schools.

American voters say 81 - 14 percent it's a "good thing" that some private companies now require people to be 21 years old or older to purchase a gun. Support is 65 percent or higher among every listed group. If a company says it will no longer sell assault weapons, 33 percent of voters are more likely to shop there, as 12 percent are less likely and 53 percent say this will have no effect.

If Congress does not pass stricter gun laws, 57 percent of voters say Republicans would be more responsible, as 18 percent say Democrats would be more responsible.

Support for gun control remains strong as American voters support 63 - 32 percent stricter gun laws in the U.S., compared to 66 - 31 percent in a February 20 Quinnipiac University National Poll.

Voter opinions on other gun issues are:

• Support 61 - 35 percent a nationwide ban on the sale of assault weapons;
  
• A 48 - 48 percent split on a ban on the sale of all semi-automatic rifles;
  
• Support 63 - 34 percent a nationwide ban on the sale of high-capacity magazines holding more than 10 rounds;
  
• Support 78 - 20 percent requiring individuals to be 21 years old or older to buy a gun;
  
• Support 89 - 8 percent allowing police or family members to petition a judge to remove guns from a person who may be at risk of violent behavior;
  
• Support 91 - 6 percent banning possession of a gun by an individual with a restraining order against them for stalking or domestic, sexual or repeat violence.
  

"The outcry on guns continues and voters say it's on you, Republicans, to make the change," Malloy said.

From March 3 - 5, Quinnipiac University surveyed 1,122 voters nationwide with a margin of error of +/- 3.5 percentage points, including the design effect. Live interviewers call landlines and cell phones.

The Quinnipiac University Poll, directed by Douglas Schwartz, Ph.D., conducts nationwide public opinion surveys, and statewide polls in Pennsylvania, New York, New Jersey, Connecticut, Florida, Ohio, Virginia, Iowa and Colorado as a public service and for research.

https://poll.qu.edu/national/release-det...aseID=2525

Call (203) 582-5201, or follow us on Twitter @QuinnipiacPoll.