The polls have shown Donald Trump with an edge for eight straight months, but there’s a sign his advantage might not be quite as stable as it looks: His lead is built on gains among voters who aren’t paying close attention to politics, who don’t follow traditional news and who don’t regularly vote.

Disengaged voters on the periphery of the electorate are driving the polling results — and the story line — about the election.

President Biden has actually led the last three New York Times/Siena national polls among those who voted in the 2020 election, even as he has trailed among registered voters overall. And looking back over the last few years, almost all of Trump’s gains came from these less engaged voters.

Importantly, these low-turnout voters are often from Democratic constituencies. Many back Democratic candidates for U.S. Senate. But in our polling, Biden wins just three-quarters of Democratic-leaning voters who didn’t vote in the last cycle, even as almost all high-turnout Democratic-leaners continue to support him.

This trend illustrates the disconnect between Trump’s lead in the polls and Democratic victories in lower-turnout special elections. And it helps explain Trump’s gains among young and nonwhite voters, who tend to be among the least engaged.

Trump’s dependence on these voters could make the race more volatile soon. As voters tune in over the next six months, there’s a chance that disengaged but traditionally Democratic voters could revert to their usual partisan leanings. Alternately, they might stay home, which could also help Biden.

How they’re different

It’s not just that less engaged voters are paying less attention. The Times/Siena data suggests that they have distinct political views, and that they get their political information from different sources.

In the battleground states, Democratic-leaning irregular voters are far less likely to identify as liberal. They’re less likely to talk about abortion and democracy and more likely to worry about the economy. They overwhelmingly believe the economy is “poor” or “only fair,” while most of their high-turnout counterparts say it is “good” or “excellent.”

One important factor might be media consumption. While Biden holds nearly all of his support from voters who consume traditional mainstream media — national newspapers, television networks and the like — the disengaged are far likelier to report getting their news from social media. Biden defectors are concentrated in this group. (A TikTok analysis found nearly twice as many pro-Trump posts on the platform as pro-Biden ones since November.)

Low-turnout voters also pose a challenge for pollsters. While millions of them will undoubtedly turn out this November, no one knows just how many of them will ultimately show up — let alone exactly which ones will do so. This is always a challenge for pollsters. But in this cycle, if enough of them stay home, Biden could do much better on Election Day than it appears in the polls.

Who will ultimately vote?

If there are two consecutive elections with the same level of turnout, you might assume that the same people are voting each time. That’s not the way it works.

There’s a lot more churn in the electorate than people realize. Even if the turnout stays the same, millions of prior voters will stay home and be replaced by millions who stayed home last time.

Historically, around 25 percent of presidential election voters do not have a record of voting in the previous presidential election. This is partly because of newly registered voters, who usually vote in the next election (and who may have previously voted in a different state). But it’s also because around 30 percent to 40 percent of previous registrants who skipped the last election ultimately show up for the next one.

There are good reasons to expect fewer voters in 2024 than in recent cycles, as the 2020 election was the highest-turnout election in a century. But if you think that means that there won’t be many new voters, you’re already wrong: In fact, 10 percent of registered 2020 nonvoters already turned out and voted in the relatively low-turnout 2022 midterms. The usual churn is already at work.

Still, Trump’s strength among nonvoters means the exact number of new voters could be decisive. And exactly which new voters show up could also be pivotal. In recent years, Democrats have benefited from a hidden turnout advantage — a tendency for Democratic-leaners who vote to be more anti-Trump than those who stay home.

With that history in mind, Democrats can hope that November’s election will draw a disproportionately anti-Trump group of irregular voters to the polls. There were signs of this yet again in The Times’s recent battleground polls.

Of course, it’s unlikely that disengaged, irregular voters have already formed solid plans about November. There’s plenty of time for them to make up or change their minds about whom they might vote for — and about whether they’ll vote at all.