Is Trump really heading for re-election?

The Electoral College based on states where Trump's disapproval exceeds 50 percent (SourceL MorningConsult) 

The Electoral College based on states where Trump's disapproval exceeds 50 percent (SourceL MorningConsult) 

Somebody was bound to write the contrarian take on Donald Trump's political prospects, and Democratic political consultant Doug Sosnik stepped up to the plate in the Washington Post over the weekend:

...despite dismal poll numbers, Trump enters the contest with a job approval rating that is certainly at least marginally better than what the current national polls would suggest. Throughout the 2016 election, most analysts tracked the national polling, which failed to capture Trump’s strength in key battleground states. Current surveys continue to understate his support. Many national polls survey all eligible voters, rather than registered or likely voters, which can underestimate Trump, and some voters may be reluctant to admit that they are pro-Trump at all. Add to that the fact that Trump effectively demonstrated during the 2016 campaign that he is capable of expanding his support by effectively demonizing his opponents.
— https://www.washingtonpost.com/opinions/trump-is-on-track-to-win-reelection/2017/10/06/91cd2af0-aa15-11e7-850e-2bdd1236be5d_story.html?utm_term=.7257d36fe074

Admittedly, if the excerpt above underwhelms, you're not alone. When I was alerted to Sosnik's column, I was briefly excited by the idea that there's an argument to be made, despite all appearances, that Trump not only knows what he's doing, but that it's succeeding.  Nothing better than an astute, unexpected and counterintuitive take that forces the reader to question his or her assumptions and view issues from a different perspective.  Better still, it's by a Democrat from whom one might expect the opposite argument.  This column, despite the promise, was not that. 

Which is not to say I am certain Trump will not be reelected in 2020 -- given my errant prognostications of late, it would border on criminally stupid to assert as much.  Who knows what 9/11 style exigency awaits us?  Who knows the extent to which Bernie Sanders and his supporters are willing to sacrifice electability for their revolution? To paraphrase Rummy, unknown unknowns abound.  

But there are ways to look at Trump's political health at this particular juncture that led me to wonder whether Sosnik and others aren't being a little coy, if not disingenuous.  Sure, Sosnik's point about 2016 smashing political norms, but he doesn't persuade me that Trump has upended basic arithmetic in the process.  

The pollster Morning Consult has released the findings of a presidential approval survey drawn from interview with 472,032 registered voters across each state and Washington, D.C., from Trump’s Jan. 20 inauguration to Sept. 26. That's one helluva dataset.  

What does it show?  

While Trump's approval has declined in each of the fifty states, the MC survey allows us to track how Trump is doing in the traditional swing states that flipped the election, as well as purple states where Democrats are trawling for upsets.  Across both categories, the news isn't good for the White House.  

To give you the most conservative interpretation of the data, I have identified only those states where Trump's disapproval exceeds 50 percent.  A more traditional metric would be to look at states where he is underwater, i.e. disapproval exceeds approval, which would take in North Carolina and Ohio (both Obama states in 2008) netting 33 additional electoral votes.  In addition, Trump is only a fraction of a decimal point above water in two other states -- Florida and Georgia -- with 45 more.  

Important context.  

According to Gallup, Obama did not win a single state where his disapproval rating exceeded 50 percent in 2012. The highest negative number in a state Obama went on to win was 47.3 percent in New Hampshire.  Aside from NH, Virginia was the only "underwater" state carried by the Democrats in 2012, although both Pennsylvania and Ohio were line-ball.  

Sosnik may be right. Trump may romp home in 2020.  So much water, so many bridges.  But let's not bamboozle ourselves into believing Trump's apparent unpopularity is a clever ruse.  If an election were held today -- and that's all polls can help us with -- he would lose in a popular vote and electoral college landslide of such a scale people will be forced to wonder whether the party he has hijacked can even survive him.