Donald Trump: polls of approval and favorability

[pbrower2a]

Morning Consult's 50-state data, colors signifying 100-disapproval, which I see as the Trump ceiling for 2020 unless things change dramatically.

I had 100-DIS numbers for all fifty states and DC, only to lose them. I'm going to try to show that again.

Without electoral votes and with no distinction for districts:



Basically -- dark blue, Trump wins 55% or more
medium blue, Trump ahead with at least 50% with 100-DIS
pale blue, Trump tied or ahead with 49% or more but less than 53%
white -- Trump tied or behind with 50% or less

Every state in any shade of red is one in which Democrats won a majority of the statewide votes for House candidates in 2018. This is an apples-to-apples comparison because House votes are on issues perceived to be federal matters. Republicans still could win majorities of House delegations in some such states, but they cannot gerrymander out the relevance of the statewide vote on how people perceive the biggest of all federally-elected officials. This may be mere coincidence, but Republicans are in trouble in these states. The President is already toxic, and stands to lose every one of them. Should he pick up two of the states with ten or more electoral votes in this category (most likely among Arizona, Michigan, Pennsylvania, and Wisconsin), then he gets re-elected. But even if Democrats aren't running the calendar down on hi9m in these states. the President is clearly not gaining in them.

Every state in white is a possible loss for the President. Approval is lower than disapproval in these states or he is tied with disapproval just under 50%. The President is at a 48-48 tie in Texas, which is awful for a state in which no Democratic nominee has won in a Presidential election since 1976. Sure, the demographics of Texas are getting uglier for Republicans, but they are getting uglier elsewhere too. Trump may have won Ohio by 8% in 2016, but a favorability poll for Trump was underwater and that same poll had Trump barely behind on a poll asking whether people wanted to re-elect him or wanted someone else. And don't even ask me about a Marist poll from the summer in which Trump support was abysmal.

Utah? Trump is a bad match for Utah culture. His serial divorces do not fit Mormon family values.

With this map I show the most favorable view of the prospect of the 2020 election for Trump -- and it does not look good. At this point, any state in pink looks to have at least a 70% chance of voting for the Democratic nominee in 2020. States in white? 30-70% one way or the other, and practically any one of them (all but Utah has at least 10 electoral votes to offer) is a killer for any chance of a Trump re-election in the event of a Trump loss.

How some of the states will go will reflect who the Democratic nominee is, what campaign strategies the Trump campaign uses, and of course events between now and November 2020. I can predict none of those now. But -- there is no international disaster that has materialized in the loss of an ally, and the economy is not in a meltdown. I do not expect the President to recover from the serious scandals involving his associates. His personality is unlikely to change, and there is no cultural trend (such as a religious revival favoring the Right as in the 1970s) moving political life toward the Right.


Hard to see

CT 40
DC 19
DE 44
HI 37
NH 42
RI 42
VT 34

... Morning Consult has an R bias, but even using its data in the most flattering way (more cautious than anything else on my part) this polling (some of it is interactive, so that is harder to control) it cannot hide that Donald Trump is in deep trouble. He must turn things around (which I cannot see happening), or he and his associates can cheat. He has failed badly to deliver on his promises, and his Administration is an ethical disaster. But it does get all states every month in ways that other pollsters do not.

I use 100-DIS as a ceiling on the assumption that if the undecided voters tend to be on the Right side of the political spectrum, then those will trend strongly (and in my model completely -- that is how cautious my model is, as I thoroughly despise President Trump), Trump picks up the undecided voters.

I see the Democratic nominee, unless that nominee should be severely troubled winning anywhere between 296 (everything in pink and NE-02 because Nebraska itself is surprisingly close) and 312 electoral votes. (No, I don't see Missouri or Utah going for the Democratic nominee even if my model suggests that such is possible. I so discount the chance of Missouri going for Trump that I see Indiana, Kansas, and Montana more likely to go against him).

Eight years ago I would have said with Obama that if he did enough things right he would win re-election. He would have to be successful running against Congress, avoiding scandals, not showing anger at possible voters, and running a competent and spirited campaign. He did that because he had the skill-set with which to do so, even if he faced one of the historically-strongest challengers against a President who still got re-elected. But Obama was in no way erratic, and he had no scandals to dodge. He was also as slick a campaigner in 2012 as in 2008.

Trump? He is doing well for all the problems that he has and lacking the skill set of the usual President who gets re-elected. Every state in any shade in red is one in which Democrats won the majority of the vote for House seats -- in a midterm election in 2018. He practically has to undo the scandals of his administration and his campaign to get re-elected. That does not happen.