What Democrats’ panic over young voters misses

Vox/Joan Wong; Getty Images

If a spate of surveys over the last four months are to be believed, young voters aren’t just dissatisfied with Joe Biden — they’re switching to supporting Donald Trump.

Is Donald Trump on track to win an unprecedented share of the youth vote?

Months of polling, and thousands of words of political writing based on that polling, seem to suggest just that. If those polls are to be believed, President Joe Biden isn’t just in trouble with dissatisfied young voters. He’s facing the possibility Trump could do better with young voters than any other Republican candidate of the modern era.

A recent warning sign came from USA Today/Suffolk University’s poll of registered voters on the first of the month: “A fraying coalition: Black, Hispanic, young voters abandon Biden as election year begins,” read the headline of the accompanying piece. Under the hood, the numbers look dire: “Among voters under 35, a generation largely at odds with the GOP on issues such as abortion access and climate change, Trump now leads 37%-33%,” USA Today notes. The poll of 1,000 registered voters was conducted in the last week of the year, December 26–29.

That finding echoes the December results from the New York Times/Siena College poll that also triggered a fierce debate among political strategists, pundits, and pollsters over just how much to believe the poll’s findings among subsets of American voters, like Black, Latino, and young voters. In that December survey, the weak support for Biden from young people registered as a Trump lead. Among registered voters between the ages of 18–29, Trump actually led Biden 49 to 43 percent.

As Nate Cohn, the Times’s chief political analyst, explained, a near-even split between Trump and Biden among young voters has been the “basic story about young voters … in nearly every major survey” by the end of 2023. The Times’s poll of battleground states from earlier in the year showed a similar conclusion: Biden was up just 1 point over Trump (47 to 46 percent) among 18- to 29-year-old registered voters. And more recently, a Quinnipiac poll of registered voters in Pennsylvania found Biden up 5 points over Trump with voters under 35; compare that to the 16-point advantage Biden had on the eve of the 2020 election in Quinnipiac’s own polling.

These numbers mark a shocking shift from 2020, when Biden handily defeated Trump among young voters, according to various exit polls and post-election verified voter surveys. AP VoteCast, for example, found Biden holding a 25-point advantage over Trump among young voters; Pew’s survey, the Catalist research firm’s estimates, and national exit polls, meanwhile, all reported a 24-point win. That lead among young voters was crucial to Biden’s victory. Unprecedented young voter turnout in 2020 helped offset Trump’s slight advantages among voters over the age of 50.

The youth defections have only amped Democratic anxiety ahead of a likely Biden-Trump rematch in 2024, but the shift is so large, and shocking, that some are exploring alternative explanations. After all, a Republican hasn’t won a majority of the youth cohort since the era of Ronald Reagan. And Biden’s numbers look like they are in freefall: How could Biden go from beating Trump by 21 points before the 2022 midterms and 10 points during summer 2023 to losing young people to Trump now?

Something seems amiss. In an effort to break through the topline numbers, I talked to a range of pollsters from across the political spectrum to come up with a few theories that explain these drastic (and varying shifts). They largely described three theories.

The polls are right: Biden has a major problem that could turn catastrophic. Young people are pissed at him for his approach to Israel and climate change, down on the state of the economy and the country more generally, and concerned about his age. And no amount of nuance or cope can wish away the fact that young voters are dropping Biden for Trump and other third-party candidates.

The polls are noisy: Biden faces a problem, but not one that should send Democrats into a tailspin. The campaigns haven’t really swung into action, polls this far out are noisy, and third-party support is always overestimated in the polls.

The polls are missing something: This is a complex case for scrutiny that holds the polls are capturing some dissatisfaction with Biden, but they are overstating Trump’s edge because of how polls are conducted now. It’s harder for polls and pollsters to reach young people specifically and to accurately capture their sentiments. And the respondents feeding these polls are even less likely to be representative of the rest of their cohort than of other generations.

No theory alone explains these Biden numbers fully. But together they suggest caution at drawing sweeping conclusions about young people — and at sensationalized poll-driven media coverage.

1. The polls are right

This is the case for Democrats to panic. Under this theory, the big, high-quality national polls are capturing large defections from Biden over the last few months, picking up after the October 7 Hamas attack on Israel and Israel’s military response in Gaza. In turn, the polls are showing big swings toward Trump and third-party candidates, showing not just frustration, but anger at Biden.

It’s no surprise that Biden’s support would have declined since 2020. He was never young people’s preferred candidate in the 2020 primaries; young Americans were reluctant to support him; and though 2020 saw a record number of young people turn out, they still tended to be voting against Trump, rather than for Biden, in the general election. Even before Israel’s war in Gaza, Biden’s low approval numbers from younger Americans were frequently making headlines (including here at Vox). In the minds of many young Americans (and interest groups), he wasn’t progressive enough, including by not acting boldly enough on student loan cancellation, climate change, and other priorities for the political left.

This decline accelerated after the Hamas war’s escalation. Cohn, at the Times, notes this in explaining the December poll’s findings: “The young Biden [2020] voters with anti-Israel views are the likeliest to report switching to Mr. Trump … It’s possible that the kinds of young voters opposed to Israel already opposed Mr. Biden back before the war. That can’t be ruled out. But it’s still evidence that opposition to the war itself is probably contributing to Mr. Biden’s unusual weakness among young voters.”

Combined with the still sour mood many Americans, including younger ones, feel about the economy, and general frustrations with Biden’s age, these factors offer a simple explanation. “One of the reasons you’re seeing Biden’s numbers be so low, particularly with younger voters and with minorities, is he’s got a policy problem with them,” Republican pollster Amanda Iovino told me. The simplest explanation is often the closest to the truth, she said. “Sometimes we get a little bit too in the weeds on ‘is our methodology right’ to confirm our suspicions that these folks will eventually go back to voting Democrat when faced with a choice.”

Under the theory, Biden either needs to enact immediate big changes in his platform and his campaign or face losing young voters en masse.

2. The polls are noisy

This is the case for patience. Using this lens, pollsters caution it’s still way too early for polls to be predictive of the final vote choices, that the campaigns have barely started to rev up, that voters are still not really thinking about the election, and that the poll numbers showing advantage to Trump and swings to third-party candidates should be reasons to scrutinize narratives built around those results.

“The most important thing to remember is that we are still 10 months out from the election, and people who don’t live and breathe politics are really not thinking that much about it,” Natalie Jackson, a longtime pollster and the vice president of the public opinion research firm GQR, told me. “Voters are kind of mad and cranky about pretty much everything right now, and we’re seeing that across all the data. That extends to the current administration.”

Jackson said that the most obvious explanation of the polls right now is that asking people to decide between Trump and Biden is not being interpreted as a choice between the two, but as a gauge of satisfaction with the country’s state. And those most likely to answer in that way are low-information and low-propensity voters.

“These low-information people are always the last to consolidate and decide who they’re going to vote for,” she said. “And young people are very frequently low-information voters. They turn out less often. They’re less likely to pay attention. It’s easy to forget that because we see so much coverage of young people protesting and bringing up certain issues. But by and large, they are a low-information group that is not going to solidify until much, much closer to the election.”

There’s also the issue of young people just not really being as engaged in national political news as they have historically been. Daniel Cox, a pollster at the American Enterprise Institute and the director of the AEI Survey Center on American Life, cautioned about this phenomenon when the first wave of young voter polling discourse was picking up in November. Young people already tend to not follow the news as closely as older Americans. But according to Gallup’s research, 2023 marked a unique time of youth disengagement. Only 9 percent of 18- to 29-year-olds were following national politics very closely, down from the average of 16 percent of them over the last 22 years.

“People are historically disengaged, and that raises two questions. One is ‘okay, if they feel this way now, how will they feel after the campaign is going in earnest and Trump’s craziness in some of his speeches is brought to the forefront as the campaigns go to great lengths to differentiate the candidates from each other?’” Cox told me.

Cox’s other question: “Is it going to impact people who are more marginally paying attention more than other folks? We may have a snapshot of what’s happening now, but it’s very likely that this group who’s sort of paying less attention will shift more than other folks.”

And finally, the performance of third-party candidates in these polls should also be cause for caution. Every pollster I spoke to told me that pre-election polls tend to overestimate the effect and support of third-party candidates. “I have no doubt that that will be the case. These candidates are not young, and they are also kind of fringy in their own ways,” Cox said. “The other thing is people recognize that these candidates don’t have a shot, and so you’re kind of throwing your vote away.”

3. The polls are missing — or misreading — something

The third theory for understanding these polls is also the most complicated one. Looking under the hood at who is getting polled, how they are being polled, and how many people are being surveyed reveals a much less dramatic shift in the way young people are feeling.

The simplest case against sensationalism comes from understanding the sample sizes of many of these polls. Jackson and Cox are quick to point out that many of the headlines hyping up the discontent of young voters are reporting on numbers in the crosstabs of polls, which usually represent a smaller number of respondents with larger room for error. The USA Today/Suffolk poll, for example, was of 1,000 registered voters; the conclusion about Trump leading Biden was drawn from the responses of 238 people under the age of 35.

“We’re often talking about 150 people, maybe 200 people, in some of these surveys. And so the margin of error is quite large in the samples,” Cox said. “We should expect greater fluctuation when we’re looking at relatively small subgroups. And so, pollsters should make note of that when they’re reporting this, and political journalists, as well, should say, ‘Well, we know these are relatively small samples and we expect there to be some fluctuation.’”

Then it’s important to understand who is being reached in the polls: Are they registered voters or likely voters? Are they adults between the ages of 18 and 29, or the larger 18 to 34 demographic? Differences emerge depending on which definition of young voter a pollster uses, and whether they are registered voters or likely voters, Adam Carlson, a former Democratic political pollster, told me. The Trump/Biden split is much closer if looking at registered voters and those under the age of 35; Biden does much better if you look at likely voters and those under the age of 30.

Suzanne Cordeiro / AFP via Getty Images
Young supporters of former President Donald Trump hold “Trump 2024” signs at a 2024 election campaign rally in Waco, Texas, in March 2023.

This far out from the race, it’s also possible that the kind of young people giving their responses to pollsters are simply more amped up about how they feel about Biden specifically. Sometimes called “expressive responding,” this theory proposes that the kind of people responding to surveys may be trying to communicate an ideological or partisan loyalty in their survey response. For disengaged young voters, responding to polls functions as a way of expressing their frustration with the incumbent, Biden, rather than thinking about a binary choice, Carlson said.

And adding to this lens is a final problem from the way the polls are being conducted. In November, the Financial Times data reporter John Burn-Murdoch analyzed a series of polls to compare the way voters across age cohorts feel about Trump versus Biden depending on if they were polled via an online poll or a telephone survey. Breaking out the results by method reveals that the huge decline in young voters’ support for Biden shows up exclusively in telephone polls. Online polls, meanwhile, show little change compared to the way young people voted in 2020. “The age-group swings we’re currently seeing in telephone polling would be completely unprecedented,” Burn-Murdoch concludes.

Asking other pollsters about this phenomenon yielded a mixed bag of responses. No one dismissed these concerns — but they specified that it doesn’t tell the whole story. Carlson’s own analysis of these swings leads him to believe that there is a reason for caution because of the effect of non-response bias. “The phone polling is where we’re seeing the biggest jumps right now from 2020 results,” Carlson said. “They are seeing way bigger swings toward Trump from Biden than online polls.” He cited results from YouGov’s polls with Yahoo News and CBS News as examples of online polls that aren’t showing a drastic swing toward Trump.

Jackson also added a bit of caution to just restricting this scrutiny to a phone versus online poll dynamic. She noted that firms make different decisions about how to pick their samples: picking from a registered voter pool first or matching up respondents to the registered voter list after running a poll; trying to update your sample as you get more responses or letting a poll just run with the original sample you chose. “We’re at a place in surveying where really, the advantages of one mode versus the other, have, for accuracy, disappeared. They all have their biases. What we really have to focus on is reaching people where they are, where they’re most likely to answer,” Jackson said.

And at the same time, another caveat: Almost all public polls right now, including the much-discussed New York Times and USA Today polls, are showing results from all adults or all registered voters. Few firms are trying to say whether these are people who are likely to vote. “And since young people are one of the least likely groups to turn out, they’re going to be really heavily affected by not screening on being likely to vote,” Jackson said.