The Washington PostDemocracy Dies in Darkness

Senior citizens have emerged as a key 2020 demographic. But why?

President Trump speaks at a White House event on protecting seniors with diabetes on Tuesday. (Evan Vucci/AP)

It is very hard for President Trump to be subtle when he’s making his political appeals. The prepared remarks he has given for events often incorporate subtle jabs at his opponents and plenty of self-promotion, but Trump will invariably introduce his own often hyperbolic asides making his points more explicit.

During an event in the White House Rose Garden on Tuesday focused on health-care costs, Trump switched from his prepared comments to various off-the-cuff asides over and over again.

“Today, I’m proud to announce that we have reached a breakthrough agreement to dramatically slash the out-of-pocket costs of insulin,” Trump said, reading from his script.

He looked up.

“You know what’s happened to insulin over the years, right? Through the roof, insulin,” he added extemporaneously. “So many people, so necessary.”

Back to the prepared remarks, picking up mid-sentence: “— for hundreds and thousands of seniors enrolled in Medicare.”

Off-the-cuff: “That’s a big deal.”

Back and forth between what he was supposed to say and what he wanted to say for the next few minutes. At one point, he thanked an administration official twice, once when he thought to do it and again, later, when it was in his prepared comments. The obvious point of the event, though, was that Trump was delivering for senior citizens in advance of the 2020 general election.

“The press won’t even cover it,” he continued, “but they’ll cover things that are unimportant. But this is a big day for seniors. This is a tremendous saving.”

He listed some of the risks posed by diabetes and reiterated the target cost.

“Nobody’s seen anything like this for a long time,” Trump said, again speaking off the cuff. “Sleepy Joe can’t do this,” he added, referring to former vice president Joe Biden, his likely general election opponent. “That I can tell you. In fact, it was his problem with Obamacare that caused part of your problem.”

A few seconds later, he again bashed Obamacare when it came up in his prepared comments.

Those familiar with recent political history might wonder why a Republican president was so focused on explicitly appealing to senior citizens. Sure, older Americans vote more heavily for a variety of reasons, but they also tend to lean Republican. Is this just Trump shoring up his base?

Well, no. As we reported in early April, voters older than 65 have consistently indicated that they prefer Biden to Trump in the 2020 election. An average of polls at that point showed that Biden was doing 16 points better with voters in that age group than Hillary Clinton did in 2016 exit polls. That is a huge shift — one that, if it held, would make Trump’s reelection all but impossible.

The question, then, is why. Contributors to The Monkey Cage analyzed polling data to show that the shift isn’t a function of Trump’s recent performance and isn’t necessarily a function of factors like Clinton suffering with the demographic because of latent sexism. It may be a function, in part, of the perception that Biden is less liberal than Clinton was.

There’s a related possibility raised by reporting from The Washington Post’s Jenna Johnson: Were senior voters in 2016 turned off by both candidates, opting to stay home instead of casting a ballot? This, it turns out, is a hard question to answer. The available evidence, though, suggests that wasn’t the case.

The best analysis of the composition of the 2016 electorate comes from the Pew Research Center. Pew’s researchers compared polling of voters with the actual voter rolls, allowing them to develop an accurate picture of how Americans voted that maps to demographic categories.

Two things quickly jump out from that analysis. The first is that a very small percentage of those who didn’t vote in 2016 were 65 or older — only 9 percent, according to the analysis. It also discovered that about two-thirds of Trump’s support came from voters 50 and up, including about a third who were of retirement age.

Older voters were a smaller part of Clinton’s electorate (about a quarter of it), but there weren’t that many nonvoters — about 4 percent of the total pool of voters — who also fell into that group.

Polling from shortly before the election also generally matched the actual outcome nationally. A poll conducted by YouGov for the Economist that ended the day before Election Day found that Trump had a 19-point lead with voters older than 65, while he trailed with those under 30 by 30 points. Pew’s analysis indicates that Trump did, in fact, lose younger voters by 30 points — but only won those 65 and older by nine points.

In other words, Clinton did better than the poll would have suggested. This may be a function of margins of error. It may also be a function of undecided voters deciding on Clinton more heavily than Trump. But it doesn’t really bolster any idea that senior citizens decided not to vote in lieu of voting for Clinton against Trump.

At least, not at the last minute. The poll was of likely voters; if people had already decided against supporting Clinton and told the pollster that they wouldn’t vote, their responses wouldn’t have been included.

This question of the voting preferences of people who didn’t like either candidate has come up before. In 2016, voters who didn’t like Trump or Clinton nonetheless voted more for him than for her, according to exit polling — enough in some states to have handed him the election. In polls of the 2020 race, voters who don’t like either candidate are more likely to say they prefer Biden, and by a wider margin.

You will notice that in the Quinnipiac University and Fox News polls cited above, the percentage of people who dislike both candidates is smaller than the 2016 exit polls showed in that race. That alone mutes some of Biden’s advantage.

You will also notice big gaps in the pie charts, reflecting people who didn’t say they’d prefer Trump, Biden or a third-party candidate. What fills in those gaps? In part, people who say they won’t vote, indicated with black slices on the charts below.

That’s the group we’re considering in our assessment of how senior citizens will vote. Those nonvoters don’t show up in exit polls because they didn’t vote, so it’s hard to establish who they were in 2016 beyond an analysis like Pew’s.

In neither the Quinnipiac nor the Fox poll were there enough people in the “don’t like either” group to break out demographically. In 2016 exit polling, though, the age distribution of that group mirrored the electorate overall.

The Fox News poll is informative in one way, though. On the question of health care, voters 45 and older have more confidence in Biden than in Trump by an 11-point margin, lower than his 17-point advantage overall. In an October 2016 poll, the network’s pollsters found that Clinton had the advantage on that issue overall by only two points more than Biden’s advantage with seniors now.

That certainly helps explain Trump’s event on insulin costs — and his focus on Biden as part of it.