How is trump running so well when he is doing worse with major demographics compared to 2016

Politics & Elections
1

so hear me out. my understanding is predictions have trump at about a 1 in 3 chance of winning. that’s about the odds he had in 2016.

But several voter demographic groups have reduced support for him.

  • college educated men and women. I think trump is doing about 20 points worse with this group than he was in 2016. with the educated women I think they went from preferring the Democrat by 7 points in 2016 to nearly a 25 point margin. college educated men went from supporting trump by 13 points to now being even or slightly dem I believe. keep on mind college educated white voters make up roughly 40% of all voters who show up to vote, so not a small demographic.

  • the elderly. isn’t trump doing worse among the elderly now due to coronavirus? my understanding is his support among them has dropped a bit. Even if he is still winning them, he isn’t winning them by as much.

  • high school educated women. trump is winning high school educated men by about the same margin he did last time, but high school women went from nearly a 30 point margin for trump down to maybe a 10 point margin for Trump.

  • aging demographics. older voters are more republican. the silent generation is more conservative than boomers, boomers more republican than Gen X who are more republican than millennials, who I think are more republican than Gen Z.

In the last 4 years roughly 15 million kids turned 18, while about 11 million Americans died. The vast majority of those Americans who died were silent generation.

Plus the share of voters who are college educated whites and latinos has grown about 4% of the total electorate, while the share of the electorate who are high school educated whites has dropped by about the same %. I forget the exact number of these though.

I guess if someone can help me understand the math I’d appreciate it.

College educated whites - ~40% of the electorate
High school educated whites - ~18% of the electorate
Both supporting him less than in 2016.

elderly turning on him

High school educated whites shrinking as a % of the total electorate.

~10 million silent generation dead while ~15 million kids gained the right to vote (granted youth turnout is low, but in presidential years I think its almost 50% for youth).

With all these trends going on, how is Trump doing as well in 2020 as he did in 2016?

If Trump is losing votes with all these groups, and the electorate has changed a bit to replace older voters, then doesn’t he need to do better with other voters to make up for it? But who is left?

High school educated men are about the same as they were in 2016, about 45-50 points for Trump.

I don’t think Trump is doing better with blacks or latinos.

Doesn’t he need to do better with ‘someone’ to have the same odds he did in 2016? Who are that someone? because its not college educated whites, high school educated white women or the elderly, hes doing worse with them than he did in 2016.

2

Perry Bacon Jr. partly addresses your questions in this piece. In sum, more college-educated whites than you might think (especially in the South, but also in certain northern suburbs*) are still Trumpy, as are more Hispanics than you might think (including in the Southwest and Florida).

(*e.g. Milwaukee‘s, though the most recent polling shows them breaking more for Biden.)

3

As has been discussed in another thread, you also have to be careful not to confuse a forecast made now, based 40 to 50% on factors other than polling, knowing that much often changes over the homestretch, with an election eve forecast.

Biden’s polling is better than Clinton’s was and much more consistent. If these are the polls on election eve then the forecast will be higher Biden probability than it is today. “If” and if likely voter models are not off in this very atypical year.

4

Also as I understand it, it comes down to Electoral College math. Solid blue states don’t hurt trump any more if he loses even more support there; solid red states will stay red even if he loses some support he had in 2016. So 40+ states don’t really enter into the equation; it comes down to a battle between 6-8 swing states. trump lost the popular vote by almost 3 MILLION votes in 2016 and still won; theoretically he could be down by 5 million and still win the electoral college.

5

White voters without college degrees are still the single biggest block of voters and millions of them didn’t vote in 2016. They remain a big potential haul for The Donald in November even while he has fallen in almost every other demographic except Hispanic voters, where he’s doing curiously better than he did four years ago (don’t ask me to explain that one).

Bottom line, though, is if the yoots come out at rates equal to or greater than they did in '18, Trump is toast.

6

The forecasters themselves are saying Trump is doing much worse than last time. The thing is, a 30% chance two months before the election is very different from a 30% chance on Election Day. Nate Silver, for example, has said that if this were Election Day, with the same polls as now, Trump’s chances would be only around 10%.

7

I find it surprising the model figures that there could be that big a change between now and Election Day. Trump’s approval numbers seem like they’ve been stable for over a year now, based on my memory of what 538 posts. Looking back at the actual numbers, Trump has actually been in basically the same range since about March of 2017. He had one small improvement in his numbers lasting a few weeks in March of this year when his disapproval number went down to 50% and approval up to 45%, but other than that he’s been in the same small range since about 2 months after he took office. I’d be genuinely surprised if after almost 4 years of almost constant disapprovals in the low 50s and approvals in the low 40s that things will change much in the next 7 weeks.

8

Apologies for the double post. After putting two and two together, my guess is Trump’s very small improvement in his numbers in March this year was a rally around the flag response when Covid-19 was first becoming a thing, before Trump started saying wacky things like telling people to not wear masks or to drink bleach.

9

Make that a triple post.

Trump is now down to a 25% overall chance according to 538’s model. In addition, I find it interesting that Trump’s chances of winning PA are at 27% while Biden’s chances of winning TX are at 31%. Trump’s chances in the other “blue wall” states are even lower than the 27% in PA. This means that if we are to take seriously the possibility that Trump can win one or more blue wall states*, we should take just as seriously the possibility that Biden can win Texas :smiley:.

*. Assuming that the four blue wall states are linked, and that if Biden wins the state with the lowest probability of victory from PA, MI, WI, and MN that he wins the other three as well.

10

FWIW I do take seriously the possibility that Biden can win Texas. It’s just that in the case he does it does not effect the outcome. He’s won enough other states by then. A huge landslide is possible. The scenarios that have Trump winning Pennsylvania though? Other states are likely lost then too and Trump is likely victorious. Any chance of that at all makes me nervous.

11

You didn’t mention black voters. The last polls I saw had Trump at 19% with Black voters. Outlier polls have his support closer to 30%. Even 19% is slightly more than double what he got in 2016.

Also, it appears that many of the people who voted Trump jn 2016 but shifted to Democrats in 2018 are now going back to Trump:

12

You do realize that an article and poll from 2019 may be a bit less than very informative?

Meanwhile more recent polling has good news and bad news for Biden.

91% of Clinton voters intend on voting for Biden while only 85% of Trump voters intend to vote for Trump again, and Biden is the strong preference over Trump among those who voted other party last time or who had not voted.

That’s the good news. The bad news?

13

I’d also appreciate a link to those recent polls running 19 to 30% of Black support. I can find one at 14 but per 538 it’s more like 10 overall which is no change.

14

Articles like this really worry me:
https://www.cnn.com/2020/09/11/politics/donald-trump-joe-biden-electoral-college/index.html

Please dear Lord let this snippet from the article be true:

…is the movement effectively a dead-cat bounce rather than a sign of an actual increased chance for Trump to beat former Vice President Joe Biden on November 3?

15

That’s CNN hoping for a horse-race.
We shouldn’t be complacent, ever, until this is all said and done…but I wouldn’t freak out (yet). If I were you, I’d spend more time over at 538. They have less of a vested interest in a horse-race.

16

Yeah, thanks for that. I just checked 538 and right now they’re predicting a 75% chance for Biden to win, so that’s good. Still not too reassuring, in light of what happened in 2016, so that’s bad. But I’ve also seen opinion pieces saying these polls are more trustworthy than 2016 polls because of reasons, so that’s… good? But I’m concerned about trump and his henchmens’ plans to fubar up the election, so that’s bad…