3 forecast modelers predict Trump will win in 2020

[Drunken Irishman]

25. These models are flawed - just look at Trump's approval for proof of this.

With any normal president, absolutely, they would be very difficult to beat in 2020 with an economy that we're seeing right now. If it was Jeb Bush or another Republican president sitting in the Oval, their approval would probably be very close to what Obama was experiencing at the end of his presidency.

But Trump's approval is far below what one would expect for a candidate in his position. And that's where you've got to look. How many presidents, with approval in the low-40s, actually go on to win reelection?

Here's the approval of the last six presidents on election day:

Obama: 50.1% (RCP average) - won 51.1% of the popular vote
Bush: 49.5% (RCP average) - won 50.7% of the popular vote
Clinton: 58% (Gallup) - won 49.2% of the popular vote
H.W. Bush: 34% (Gallup) - won 37% of the popular vote
Reagan: 58% (Gallup) - won 58.8% of the popular vote
Carter: 37% (Gallup) - won 41% of the popular vote

H.W. Bush and Carter lost reelection. They were also the only two candidates with more disapproving of their presidency than approving. Bush had the lowest approval for a winner - but still had a smaller amount of disapproval overall (47%). Clinton and Reagan tied for the highest approval but Clinton actually underperformed his approval significantly - though, that was a three-way race and he won the electoral college by a sizable margin.

Obama saw an increase of +1 from his approval to overall popular vote percentage.
Bush saw an increase of +1.2 from his approval to overall popular vote percentage.
H.W. Bush saw an increase of +3 from his approval to overall popular vote percentage (still lost, tho).
Reagan saw an increase of +.8 from his approval to overall popular vote percentage.
Carter saw an increase of +4 from his approval to overall popular vote percentage.
Clinton saw a decrease of -8.8 from his approval to overall popular vote percentage.

I should point out that some of these polls were taken a week or so before the actual election, whereas Obama and Bush were their average polling number on the day of the election according to RCP.

Just focusing on those two, you can see both actually saw a similar increase in overall popular vote total compared to their approval. This is likely due to a good amount of those who were undecided on whether they approved of Bush/Obama ultimately voting for each on election day. Regardless, the improvement isn't significant - but let's assume Trump sees a +1 increase over his approval rating on election day, 2020.

Looking at his numbers right now, he's currently at 42.8% approval in his average. This is significantly lower than either Bush or Obama on election day and a number like this would spell a good amount of trouble for Trump - as his disapproval 53.4%.

In 2016, Trump lost the popular vote but won the electoral college with 46.1% of the vote. That's probably what his campaign will be gunning for again in 2020 as anything less than that and it likely means he loses Pennsylvania, Michigan and Wisconsin - three states he won by only a few thousand votes combined.

IF the trend of the last two incumbent presidents continues and Trump sees a +1 improvement over his approval, and his approval remains just as low in 2020 as it is today, that would put him at 43.8 - rounded to 44%, or, two points off his 2016 total.

That two-points is not a lot but within the margins of his win last go around? It's devastating. If Trump wins only 44% of the popular vote in 2020, he'll lose.