Trump approval ratings thread, 1.4

[pbrower2a]

Old thread here.

We are rapidly approaching 2000 posts, so bring over the posts from the 1.3 thread here upon which you expect to build. I anticipate the 1.3 forum being closed shortly.

Oh, yes -- start posting here if you want responses to your posts.

[pbrower2a]

Blank map for your use.

Without electoral votes and with no distinction for districts:

[pbrower2a]

So probably doesn't mean anything, and I'm cherry picking, but most of Trump's more favorable polls show even(or close to it) party i.d:

Marist -12 approval:
35% - Democrat
27% - Republican
36% - Independent

http://maristpoll.marist.edu/wp-content/uploads/2018/07/NPR_PBS-NewsHour_Marist-Poll_Nature-of-the-Sample-and-Tables_Pres.-Trump-Congress-and-the-Midterm-Elections_July-2018_181807241056.pdf#page=3

Quinnipiac -20 approval:
31% - Democrat
25% - Republican
38% - Independent

https://poll.qu.edu/images/polling/us/us07242018_demos_ufgp12.pdf/

Morning Consult -6 approval:
35% - Democrat
34% - Republican
25% - Independent

https://morningconsult.com/wp-content/uploads/2018/07/180736_crosstabs_POLITICO_v1_DK.pdf

I know we're not supposed to read too much into party i.d, but its pretty obvious some pollsters are seeing very different electorates right now.

 

It could be that Morning Consult recognizes fewer 'independent' voters, probably accepting that people who voted with the GOP in 2010, 2012, 2014, and 2016 despite claiming to be independent might as well be Republicans now.  Quinnipiac sees fewer partisan voters and more independent voters.

It is possible to have different polling results based upon one's screen. "Adults" includes people unlikely to vote. Registered voters will find voting easy if they so wish. Both of these are easy to determine.

So who is a 'likely voter'? Good question. Someone on life support in terminal care is hard to write off even if the person hasn't missed a vote in over 60 years. Habits may not die, but the person will.  In my county, the election officials read the obituaries and disqualify absentee votes from persons who can be identified as decedents. On the other side, what of someone who turns eighteen a couple days before the election, who has not yet registered to vote but shows signs of registering at the first possible moment,  and considers voting just slightly less of a rite of passage than driving a motor vehicle or graduating from high school?

Some people born in as late as 2002 who have never voted will vote in the 2020 election. Nobody can predict who those people are.

If one sees changes in predictors of voting based on how certain demographics will vote, and so far that appears as approval and disapproval, then we may have cause for believing that the political climate has changed. Quinnipiac just had a poll  in which it connected people distrustful of both parties voting largely for Republicans beginning in 2010 and lasting through 2016 -- only to find that such voters have swung sharply against Republicans including the President.  I don't have to discuss why such voters swung sharply R in 2010 or why they have suddenly swung as they have. I do not need to connect such to other demographic characteristics as religion, ethnicity, education, income, or region. A vote is a vote.  Maybe I will do as I am prone -- showing a model of how such changes  expectations of the 2020 election if that alone is applied as a shift in voters from 2016 to 2020.

[pbrower2a]

Vermont (VPR-Vermont PBS)

28% approve
61% disapprove

http://projects.vpr.net/vpr-vermont-pbs-poll

Vermont will be one of the worst two or three states for the President in 2020.

Michigan, Emerson:

You must be logged in to read this quote.

https://www.emerson.edu/sites/default/files/Files/Academics/ecp-pr-mi-7.23.pdf

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:

CT 39
DC 17
DE 39
HI 33
NJ 37
RI 30
NE-01 45
NE-02 38
NE-03 55
NH 39
RI 30
VT 32

Nebraska districts are shown as 1, 2, and 3 from left to right on the map, even if they are geographically 3, 1, and 2 from west to east.

100-Disapproval

55% or higher dark blue
50% to 54% or higher but not tied medium blue
50% or higher but positive pale blue

ties white

45% or higher and negative pale red
40% to 44% medium red
under 40% deep red

States and districts hard to see:

CT 41
DC 20
DE 43
HI 36
NH 49
NJ 37
RI 30
NE-01 55
NE-02 46
NE-03 66
RI 30
VT 36


Nebraska districts are shown as 1, 2, and 3 from left to right on the map, even if they are geographically 3, 1, and 2 from west to east.

*With the explicit question of whether the President should or should not be re-elected, or 100-DIS if such is all that is available:


Re-elect/do not re-elect if known; 100-DIS otherwise

100-DIS

55% or higher dark blue
50% to 54% or higher but not tied medium blue
50% or higher but positive pale blue

ties white

45% or higher and negative pale red (or 55% do-not-reelect or higher)
40% to 44% medium red (or 50 to 54% do-not-reelect or higher)
under 40% deep red (or 50% or less do-not-reelect if do-not re-elect if do-not-reelect is higher than reelect)
Ties for elect and re-elect are also in white.

States and districts hard to see:

CT 41
DC 20
DE 43
FL 37-54
HI 36
NH 49
NJ 37
RI 30
NE-01 55
NE-02 46
NE-03 66
RI 30
VT 36


Nebraska districts are shown as 1, 2, and 3 from left to right on the map, even if they are geographically 3, 1, and 2 from west to east.


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

[pbrower2a]

Read down to see the relevant material at https://www.nbcnews.com/politics/first-read/cagle-vs-kemp-headlines-runoff-day-georgia-n893971:


Polling of a significant part of the electorate: people who hold both Parties in contempt:

So how are these kinds of voters breaking now? Well, our latest NBC/WSJ poll finds that Democrats are over-performing among voters who hold negative views of both parties (representing 13 percent of the sample). Here’s the past and current congressional preference among these voters:

- 2010 merged NBC/WSJ poll: 49 percent GOP, 23 percent DEM (R+26)
- 2014 merged NBC/WSJ poll: 51 percent GOP, 24 percent DEM (R+27)
- 2018 merged NBC/WSJ poll (through June): 50 percent DEM, 36 percent GOP (D+14)
- 2018 NBC/WSJ poll from July: 55 percent DEM, 25 percent GOP (D+30)

What’s more in our current poll, these voters disproportionately are down on Trump (68 percent disapprove of his job, versus 52 percent of all voters), and they are enthusiastic about the upcoming midterms (63 percent of them have high interest, versus 55 percent of all voters who say this).

Democratic candidates are increasingly seen as being out of the mainstream

Those numbers above are good news for Democrats in the NBC/WSJ poll. Here’s some bad news, however: Democratic candidates for Congress are increasingly seen as out of the mainstream — a change from 2012 and 2016.

According to our poll, 33 percent of voters view Democratic congressional candidates as in the mainstream, versus 56 percent who say they are out of step. That’s essentially the same score that GOP congressional candidates get — 33 percent mainstream, 57 percent out of step.

But the 33 percent viewing Democratic candidates in the mainstream is a drop of 15 points from 2016 and 12 points from 2012 (while the GOP numbers have been pretty flat).

Comment: this suggests a Blue Wave in 2018. That is a 57% swing in a significant part of the electorate, probably comparable to the swing between the near-landslide victory for Barack Obama and the rise of the Tea Party to political influence.

It also suggests that Donald Trump will lose, possibly in a landslide, taking a bunch of Republican Senators (echo of the 2014 election that gave Republicans control of the Senate) with him.
[/quote]

from the previous month:

You must be logged in to read this quote.

So let's see how a swing of 57% of 18% of the electorate  shifts the vote from  2016 to 2020.  That is 10.2% of the vote! That is without discussing demographic shifts (such as an increasing size of the Hispanic population in a state like Texas) or shifts in the vote in certain demographics (I can only speculate that farmers and ranchers will turn against Trump should his tariffs and trade war hurt farmers and ranchers, especially those who depend upon revenue from exports).  

The 10.2% shift that this change suggests flips every state in blue -- including Iowa.  That comes to 161 electoral votes!

2016 result among states decided by 10% or less:

[/quote]

8% or more -- saturation 7
4% to 7.99% -- saturation 5
1.5% to 2.99% -- saturation 3
under 1.5% -- saturation 2

States in gray look too far away to be affected by a shift in votes from 2016 to polling in 2018.

(The next-closest states were South Carolina and Alaska, which would still go for Trump -- and everything in any shade of red would be a strong win for the Democrat at 10.1% or higher). The result:

8% or more -- saturation 7 (9 if over 10%)
4% to 7.99% -- saturation 5
1.5% to 2.99% -- saturation 3
under 1.5% -- saturation 2

the Democratic nominee 413
Trump 125

Even without Texas going to the Democrat, Trump loses much as the elder Bush did in 1992.

[pbrower2a]

Marist just tweeted that approval ratings for the President are execrable in Minnesota (which he barely lost), Michigan (which he barely won), and Wisconsin (which he barely won). In each, the President's approval is in the mid-30s.

[Calthrina950]

Marist just tweeted that approval ratings for the President are execrable in Minnesota (which he barely lost), Michigan (which he barely won), and Wisconsin (which he barely won). In each, the President's approval is in the mid-30s.

It is amazing how rapidly the Midwestern states have turned against Trump. The tax cuts and his failure to reverse outsourcing are probably the main reasons why, and I think the tariffs will hurt him further. I still think Trump has a 50-50 chance in winning Wisconsin and Michigan in 2020, but it's going to be tough.

[here2view]

Marist just tweeted that approval ratings for the President are execrable in Minnesota (which he barely lost), Michigan (which he barely won), and Wisconsin (which he barely won). In each, the President's approval is in the mid-30s.

It is amazing how rapidly the Midwestern states have turned against Trump. The tax cuts and his failure to reverse outsourcing are probably the main reasons why, and I think the tariffs will hurt him further. I still think Trump has a 50-50 chance in winning Wisconsin and Michigan in 2020, but it's going to be tough.

He has a better chance at winning Wisconsin than Michigan. His margin of victory was smaller in Michigan and it voted more so for Obama in 2012 than Wisconsin.

They're both tossups IMO but he's more likely to win Wisconsin or Pennsylvania than Michigan. If you put a gun to my head I think he loses all three.

[Pericles]

There's not much difference between 0.8% and 0.2%.

[Inmate Trump]

Marist just tweeted that approval ratings for the President are execrable in Minnesota (which he barely lost), Michigan (which he barely won), and Wisconsin (which he barely won). In each, the President's approval is in the mid-30s.

It is amazing how rapidly the Midwestern states have turned against Trump. The tax cuts and his failure to reverse outsourcing are probably the main reasons why, and I think the tariffs will hurt him further. I still think Trump has a 50-50 chance in winning Wisconsin and Michigan in 2020, but it's going to be tough.

He has a better chance at winning Wisconsin than Michigan. His margin of victory was smaller in Michigan and it voted more so for Obama in 2012 than Wisconsin.

They're both tossups IMO but he's more likely to win Wisconsin or Pennsylvania than Michigan. If you put a gun to my head I think he loses all three.

There are still plenty of morons allowed to vote who think Obama/Hillary destroyed their livelihoods or whatever.  These people, coupled with Russia probably being much more heavily involved in the 2020 election than anywhere near the level they were in 2016, could easily reelect him.

[IceSpear]

There's not much difference between 0.8% and 0.2%.

Did you forget where you are? This is a forum that assigns meaning to 0.1% changes in approval rating.

[Virginiá]

It was kind of thrilling to see if he would reach the depths of Nixon and 2nd term Bush, but his approvals are so stable that it's become very boring to watch this. It also seems unprecedented for how stable his numbers are. I mean, even the strong disapproval has been locked in more or less.

Considering Trump chews through what should be presidency-ending scandals every week, it's fair to say that one of the only plausible ways to make these numbers budge is a recession.

[pbrower2a]

Marist just tweeted that approval ratings for the President are execrable in Minnesota (which he barely lost), Michigan (which he barely won), and Wisconsin (which he barely won). In each, the President's approval is in the mid-30s.

It is amazing how rapidly the Midwestern states have turned against Trump. The tax cuts and his failure to reverse outsourcing are probably the main reasons why, and I think the tariffs will hurt him further. I still think Trump has a 50-50 chance in winning Wisconsin and Michigan in 2020, but it's going to be tough.

He has a better chance at winning Wisconsin than Michigan. His margin of victory was smaller in Michigan and it voted more so for Obama in 2012 than Wisconsin.

They're both tossups IMO but he's more likely to win Wisconsin or Pennsylvania than Michigan. If you put a gun to my head I think he loses all three.

Disapproval in all three states is above 50%, so winning any one of them will be tough. The difference between the three states in current approval ratings is insignificant.

President Trump wins these states together or loses them together. Note well that he is doing badly in Arizona, Florida, and Ohio, states more sympathetic to Republicans.

I am holding off on showing the polling results here in the event that Marist asks the question "Will you vote to re-elect President Trump or will you vote against re-electing him in 2020?", as Marist did for Arizona, Florida, and Ohio. The answer came to 35-57 in Arizona, 37-54 in Florida, and 34-58 in Ohio.  I'd like to see a similar question answered for in the Cold, Wet North.  Heck, I far prefer this to my 100-DIS number,

[The Mikado]

It was kind of thrilling to see if he would reach the depths of Nixon and 2nd term Bush, but his approvals are so stable that it's become very boring to watch this. It also seems unprecedented for how stable his numbers are. I mean, even the strong disapproval has been locked in more or less.

Considering Trump chews through what should be presidency-ending scandals every week, it's fair to say that one of the only plausible ways to make these numbers budge is a recession.

Honestly, I think they could end up budging due to Trump interacting with a Democratic Congress, but I think it might help him more than hurt him. Giving him Pelosi-Schumer to play off of rather than unified GOP control which just highlights how incompetent and ineffectual he is.

[Virginiá]

Honestly, I think they could end up budging due to Trump interacting with a Democratic Congress, but I think it might help him more than hurt him. Giving him Pelosi-Schumer to play off of rather than unified GOP control which just highlights how incompetent and ineffectual he is.

I've been wondering a lot about this, and I'm still not sure. On one hand, yes, he can blame stuff on Democrats in Congress, sure. But on the other hand, there are probably going to be a lot of investigations into his administration. Knowing what we know about Trump, it's very possible that scandals could end up offsetting any gains he may have from the blame game with Congress. But I'm also just not sure how effective it is to try and pass the buck to Congress. Americans don't seem to care about anyone but the president. I imagine that is why cable news isn't actually about news anymore, but rather 24/7 coverage of what the president says and does.

[ProudModerate2]

... Marist asks the question "Will you vote to re-elect President Trump or will you vote against re-electing him in 2020?", as Marist did for Arizona, Florida, and Ohio. The answer came to 35-57 in Arizona, 37-54 in Florida, and 34-58 in Ohio.

Wow.
I must have missed those numbers when they were first released.
Are the numbers really that high? That's huge!

[💥💥 brandon bro (he/him/his)]

Honestly, I think they could end up budging due to Trump interacting with a Democratic Congress, but I think it might help him more than hurt him. Giving him Pelosi-Schumer to play off of rather than unified GOP control which just highlights how incompetent and ineffectual he is.

I've been wondering a lot about this, and I'm still not sure. On one hand, yes, he can blame stuff on Democrats in Congress, sure. But on the other hand, there are probably going to be a lot of investigations into his administration. Knowing what we know about Trump, it's very possible that scandals could end up offsetting any gains he may have from the blame game with Congress. But I'm also just not sure how effective it is to try and pass the buck to Congress. Americans don't seem to care about anyone but the president. I imagine that is why cable news isn't actually about news anymore, but rather 24/7 coverage of what the president says and does.

I think investigations and cooperation are mutually exclusive. If Congressional Democrats investigate, then I can't see someone as transactional and thin-skinned as Trump reaching out to cooperate. Given what the energy in the Democratic party looks like right now, I can't see them simultaneously cooperating and investigating, and they certainly are more interested in investigating.

Independents are going to lap up cooperations if, by some miracle, they ever happen. I'm curious to see how the base would go along with hypothetical cooperation though. The two animating ideologies in the party right now are defense of Trump and attacking liberalism (or, really, just liberals). Which side would win out in this cataclysmic battle of hatefulness?

[Tartarus Sauce]

Honestly, I think they could end up budging due to Trump interacting with a Democratic Congress, but I think it might help him more than hurt him. Giving him Pelosi-Schumer to play off of rather than unified GOP control which just highlights how incompetent and ineffectual he is.

I've been wondering a lot about this, and I'm still not sure. On one hand, yes, he can blame stuff on Democrats in Congress, sure. But on the other hand, there are probably going to be a lot of investigations into his administration. Knowing what we know about Trump, it's very possible that scandals could end up offsetting any gains he may have from the blame game with Congress. But I'm also just not sure how effective it is to try and pass the buck to Congress. Americans don't seem to care about anyone but the president. I imagine that is why cable news isn't actually about news anymore, but rather 24/7 coverage of what the president says and does.

In addition to the litany of investigations Democrats will be bulldozing into the administration, it will be interesting to see how Republicans react if they lose badly in the midterms. Short-term power for public embarrassment and pain is the Faustian bargain Republicans have made with Trump. What about when the power part of the equation is removed and they’re only left with pain? I have to imagine some of the Republicans who are less ride-or-die for Trump and more just partisan loyalists won’t be enthused with his antics having generated a massive liberal backlash. Plus, it cuts pretty deep into his image as a “winner.”

I’m not saying Republicans will abanadon him en-masse, but I doubt he’ll have the rock-solid consolidation of Republicans he currently possesses in the wake of a massive Democratic tsunami. He’d definitely drop back down under 40.

Also, sweet Jesus we talk about his approvals a lot.

[pbrower2a]

Marist just tweeted that approval ratings for the President are execrable in Minnesota (which he barely lost), Michigan (which he barely won), and Wisconsin (which he barely won). In each, the President's approval is in the mid-30s.

In case you wonder why I was so slow to put up the polling numbers for these three states, Marist asked a question even more telling than whether people approve or disapprove of the President's performance. The response is even more devastating than my 100-DIS model, which may be a bit conservative in estimating the chance of a Trump loss. Approval of the President is horrid in these three states, but even if the President won the decided on approval vs. disapproval he would lose all three states --   if barely.



These are devastating, suggesting a loss analogous to that of Carter in 1980. Of course, Carter actually won Minnesota in 1980, but then Minnesota was the most Democratic state in the Union as shown in the solitary win by Mondale in 1984. Minnesota has drifted away from being the strongest Democratic state, and I can easily imagine it being slightly more R than the national average in a Democratic blowout.  
  




As in 2008, 2012, and 2016 these states were very close to each other (differing by a margin of about 2%), and will likely be so again in 2020. I still have but three states for which I have the answer to the question of re-election or bring in someone else... but that is enough to suggest that President Trump will lose to a Democrat who will win at least as decisively as Obama in 2008. Sure, I would love to see this question asked of voters in Iowa, New Hampshire, Nevada, and Pennsylvania. Have patience, folks.  

 California, PPIC  -- I found the disapproval numbers from mid-July, and for likely voters approval is 34 and disapproval is 64.

http://www.ppic.org/wp-content/uploads/ppic-statewide-survey-july-2018.pdf

Old (mid-July), but necessary.

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:

CT 39
DC 17
DE 39
HI 33
NJ 37
RI 30
NE-01 45
NE-02 38
NE-03 55
NH 39
RI 30
VT 32

Nebraska districts are shown as 1, 2, and 3 from left to right on the map, even if they are geographically 3, 1, and 2 from west to east.

100-Disapproval

55% or higher dark blue
50% to 54% or higher but not tied medium blue
50% or higher but positive pale blue

ties white

45% or higher and negative pale red
40% to 44% medium red
under 40% deep red

States and districts hard to see:

CT 41
DC 20
DE 43
HI 36
NH 49
NJ 37
RI 30
NE-01 55
NE-02 46
NE-03 66
RI 30
VT 36


Nebraska districts are shown as 1, 2, and 3 from left to right on the map, even if they are geographically 3, 1, and 2 from west to east.

*With the explicit question of whether the President should or should not be re-elected, or 100-DIS if such is all that is available:


Re-elect/do not re-elect if known; 100-DIS otherwise

100-DIS

55% or higher dark blue
50% to 54% or higher but not tied medium blue
50% or higher but positive pale blue

ties white

45% or higher and negative pale red (or 55% do-not-reelect or higher)
40% to 44% medium red (or 50 to 54% do-not-reelect or higher)
under 40% deep red (or 50% or less do-not-reelect if do-not re-elect if do-not-reelect is higher than reelect)
Ties for elect and re-elect are also in white.

States and districts hard to see:

CT 41
DC 20
DE 43
FL 37-54*
HI 36
MI 28-62*
NH 49
NJ 37
RI 30
NE-01 55
NE-02 46
NE-03 66
RI 30
VT 36


Nebraska districts are shown as 1, 2, and 3 from left to right on the map, even if they are geographically 3, 1, and 2 from west to east.


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

[KingSweden]

I’d be leery of the “deserves reelection” vs “give someone else a chance” numbers if only because someone else could be Generic R, Generic D, or even Santa Claus. It’s a bad number in context, obviously, but it isn’t as informative as we think IMO

[pbrower2a]

... Marist asks the question "Will you vote to re-elect President Trump or will you vote against re-electing him in 2020?", as Marist did for Arizona, Florida, and Ohio. The answer came to 35-57 in Arizona, 37-54 in Florida, and 34-58 in Ohio.

Wow.
I must have missed those numbers when they were first released.
Are the numbers really that high? That's huge!

They are that bad for the President, and the overt question is better than my 100-DIS model in which the incumbent has a chance of winning to the extent that he can get a plurality by winning over the undecided vote.  Even more definitive results are available if one has a match-up, but the Democrats have yet to nominate someone -- and it is far from certain whom they will nominate.

Answers to the binary choice 're-elect or do not re-elect' suggest that approval is even more telling than whether the President has simply offended too many people with incompetence or with offense to the sensibilities of too many people.

I have figured that he has been clever enough to load distress upon people who were never going to vote for him while appealing to the subliminal values of those who can still support him.

So do enough people want to re-elect him? This question assumes that the incumbent President is the dominant factor in deciding whether to vote for him in 2020. As I see it, Democrats are in so powerful a position against Trump that they could win with a mirror-image of him, someone similarly corrupt, intellectually-hollow, unprepared for the office, and pathologically narcissistic -- who upon taking office exploits ethnic resentments and uses the government largely to reward special interests on the Left. I do not see that happening. The Democratic nomination will have people who have political experience and records. Maybe they nominate someone bland and lacking in sizzle -- let us say a new Al Gore.  But that will be fine if things go well; that nominee will win the Presidency once and be re-elected if most things go right.

[pbrower2a]

Honestly, I think they could end up budging due to Trump interacting with a Democratic Congress, but I think it might help him more than hurt him. Giving him Pelosi-Schumer to play off of rather than unified GOP control which just highlights how incompetent and ineffectual he is.

I've been wondering a lot about this, and I'm still not sure. On one hand, yes, he can blame stuff on Democrats in Congress, sure. But on the other hand, there are probably going to be a lot of investigations into his administration. Knowing what we know about Trump, it's very possible that scandals could end up offsetting any gains he may have from the blame game with Congress. But I'm also just not sure how effective it is to try and pass the buck to Congress. Americans don't seem to care about anyone but the president. I imagine that is why cable news isn't actually about news anymore, but rather 24/7 coverage of what the president says and does.

I know of a state governor (Jerry Brown, D-CA) who took a 180-degree turn on big government and taxes once a referendum demonstrated which way the wind was blowing.  But this said, the young Jerry Brown (who is easy to confuse with the old Jerry Brown, who coincidentally is the same person) didn't have reeking scandals and had not displayed anywhere near the incompetence or ultra-partisanship that Donald Trump has shown. He got re-elected.

President Trump has so pervasively offended so many liberals on core values that Democratic leaders of the House or Senate that Democratic leadership can intensify the waiting game on him, cooperating with him only on something obvious or yielding on things fully symbolic in return for huge concessions on things that he has supported. He will have to make compromises on an appointment to the Supreme Court in the event of a vacancy. If he wants an infrastructure program it won't have privatization to monopolistic gougers as an objective.  He will be swatted down if he makes outrageous statements, so he might do that less often.

Enough damage has been done to him as President, most of it his own doing, that without change in his policies he is doomed to defeat in an electoral landslide analogous to that that ended the Presidencies of Hoover and Carter at one term. With concessions to liberal sensibilities he sets himself up to defeat by the Real Thing, and I don't refer to a tourist trap on Interstate 10 in southeastern Arizona.   

The question might change from how awful he is to whether he is right for America, which is an improvement.  People who usually don't vote for Republicans vote in Michigan and Wisconsin, and demographics catch up with him in Arizona and Florida. If he is that big a disappointment in Ohio, just think what that means for Pennsylvania. In view of Wisconsin and the fact that Iowa usually votes in tandem with Wisconsin, I cannot see the President winning Iowa. Minnesota is out of the question. Sure, I would love to see Marist ask the question to voters in Iowa, Nevada, New Hampshire, North Carolina, and Pennsylvania. But I am not in the decision-making process at Marist. For all I know they may never ask this question again on a poll -- or, at a lesser level, they might next pick Illinois, Massachusetts, and New York. Or perhaps Quinnipiac or PPP starts asking this question.

I  think this question more telling than my 100-DIS model, and I show my results both ways for the sake of consistency. 100-DIS is still more favorable to the President, and it is what I am stuck with for 44 states and the District of Columbia as well as the independently-voting Congressional districts of Maine and Nebraska -- at least for now.   

[pbrower2a]

I’d be leery of the “deserves reelection” vs “give someone else a chance” numbers if only because someone else could be Generic R, Generic D, or even Santa Claus. It’s a bad number in context, obviously, but it isn’t as informative as we think IMO

You are welcome to believe that. I could think of an irrelevant question for 2020 -- that if the 22nd Amendment were to disappear, would you vote for Donald Trump or Barack Obama if you had the choice. We will not have that choice.

I have so far used the 100-DIS model on the ground that those who disapprove of the President will not vote for him. They might vote for the Democrat -- or if they are too 'conservative' to vote for the Democrat, then they might vote for some conservative or free-market purist, which could include someone running a vanity campaign that basically says "my business expertise is more relevant to running this country well than is that of Donald Trump". That could be someone in high technology, oil, agribusiness, investment banking, or heavy equipment.  Donald Trump is simply a landlord, and he knows little about business enterprise outside his milieu.

I almost expect a strong Third Party or independent challenge to Donald Trump because he offends so many conservative interests.

To be sure, we should always be aware of our assumptions. We largely assume:

1. That Donald Trump will be the nominee for President -- that he will not decide not to run, that he will not die in office, and that he will not removed for diminished capacity (as after entering an irreversible coma).

2. That we will not have a military coup. Sure, we have never had one. But Seven Days in May is becoming much more plausible with this President, if for very different reasons. Doddering old leader? Check. Highly unpopular leader? Check. "Too liberal"? The opposite. "Too squeamish about taking harsh measures"? Exactly the opposite.

I would not rule out that the military would turn on him rather than soil itself in aggressive war against Iran or Venezuela. Yes, the military has not intervened in the government at any time  in American history, and 240 years of civilian control of the military will come to an end only under exceptional circumstances.

So imagine that you are a four-star general and you must choose between overthrowing Donald Trump or becoming complicit in war crimes. You don't know what you would do? Even I can't speak for myself.

3. That the elections of 2020 will not be rigged. Watch the elections this November and see how they go before you answer whether the elections of 2020 will be rigged or won't.

4. That the Democrat will not be exposed in having done something discreditable -- insider trading, having sex with minors, being involved in a business failure the result of malfeasance, having a dishonorable discharge from the military, or having a criminal record. That is clearly in the category of 'unforeseen events'  that have nothing to do with polling.

5. That we can derive any conclusions from polling. By November 2010 we could see Obama within easy reach of winning re-election even if his Party was severely defeated in the midterm election -- it would take a spirited campaign by him and competent strategy as a candidate to turn approval in the 45-47% range into either a bare majority or even a plurality. It is a reasonable assumption that a spirited campaign and good campaign strategy were good for turning something like 45% approval into 51% of the vote. Obama ran a competent enough with which to win despite a disapproval rating in the mid 40s around September 1 (just after the Republican national convention, I guess).  44% + 7% = 51%, so it looks as if he did what he needed.  

I look at recent polling numbers for Donald Trump, and he will be lucky to get 46% of the popular vote. Sure, he won with 46% of the popular vote because Hillary Clinton ran up the vote totals in places like California and New York -- but just look at the polling for Arizona, Florida, Michigan, Minnesota, Ohio, and Wisconsin. Those look really bad. The poll of Pennsylvania had Trump disapproval in the mid-fifties.

That disapproval of a nominee is in the 50s may not assure that that nominee will lose 54-46 to the opponent. Maybe it will be more like 49-46-5, which only looks sort of close.

6. That Democrats will not be facing a strong left-wing alternative that guts their support. This is the most likely thing to go wrong for the Democratic nominee for President. I cannot yet rule it out.

But I can't completely rule out such things as an invasion from outer space, a zombie apocalypse, the Coming of a Messiah, or some event that stirs up some expressions of patriotism that pull the President out of the polling doldrums, either.  

[BlueSwan]

I’d be leery of the “deserves reelection” vs “give someone else a chance” numbers if only because someone else could be Generic R, Generic D, or even Santa Claus. It’s a bad number in context, obviously, but it isn’t as informative as we think IMO

This. And we can add general "favourability" as well, which proved a useless metric in the 2016 election.

Approval is a bit more useful, but even that ignores that the election in reality is a binary choice between two candidates.

Frankly, I love polls as much as anybody else, but basically polls don't start to give a solid indication of much until both parties have picked their candidates. But at least we know that the majority disapproves of the terrible president, so in theory, a democrat should be favoured if the candidate is electable.

[pbrower2a]

I’d be leery of the “deserves reelection” vs “give someone else a chance” numbers if only because someone else could be Generic R, Generic D, or even Santa Claus. It’s a bad number in context, obviously, but it isn’t as informative as we think IMO

This. And we can add general "favourability" as well, which proved a useless metric in the 2016 election.

Approval is a bit more useful, but even that ignores that the election in reality is a binary choice between two candidates.

Frankly, I love polls as much as anybody else, but basically polls don't start to give a solid indication of much until both parties have picked their candidates. But at least we know that the majority disapproves of the terrible president, so in theory, a democrat should be favoured if the candidate is electable.

I found favorability (do you like the President or some other elected official?) relevant only very early in his term of office. Favorability polls occasionally pop up, and I now ignore them except to compare them to something else. Initially there were plenty of favorability polls, but those have largely been abandoned. Performance matters now, and people like what they have seen or don't.

It's hard to see whether events cause changes in polling or whether the results will stick. Obama approval ratings skyrocketed for a short time after Seal Team 6 whacked Osama bin Laden, but that uptick didn't last long enough to allow the President to get anything like a 55-45 split of the vote. It is pointless to predict how an event or economic news (which could include the prices of commodity futures) could affect polls until we see polls with connections to demographics. Obama did badly among farmers and ranchers despite doing nothing bad to them. We shall see about Donald Trump, whose trade war can do more harm to anyone whose income depends heavily upon export prices. For good reason we rarely see polls of Kansas, Nebraska, or the Dakotas. But I don't give quantitative predictions. 

Demographics may not be causes, but they are certainly mirrors.