35 Percent of Young Trump Voters Say an Online Creator Changed Their Mind

Young Harris voters are in the same feeds at the same rate, but only 19 percent report being persuaded by an influencer.

In early 2020, Michael Bloomberg’s presidential campaign started paying influencers to post memes for him, and my phone started ringing.

Reporters wanted to know what Facebook was doing about it. Was this advertising? Would it show up in the Ad Library? Were we going to label it? The honest answer was that we couldn’t see most of it. The money moved off-platform, between a campaign and a creator, and the post that landed in someone’s feed looked like every other post that creator made. We scrambled anyway, because the thing we were most afraid of was being the next headline. What we came up with was a CrowdTangle dashboard so researchers and journalists could track the accounts themselves. It was a workaround, but we didn’t have any better ideas.

Six years later, online creators are a huge part of the campaign story – and we still struggle with disclosure. Both the right and the left are engaging influencers more and more. I wanted to know more, though, about how different types of voters were consuming this content, whether it was persuading them, and what kinds of disclosure they expected.

So I put a set of questions on the Rainey Center’s July omnibus: 1,017 registered voters, online panel, fielded July 14–15, weighted to race, 2024 vote, age, education, and gender, ±3.5 points. Full top-lines and cross-tabs here.

Thirty-five percent of registered voters say they get news or political information from influencers, creators, or podcasters rather than news organizations at least somewhat often. Among 18-to-34-year-olds, it’s 57 percent. Among voters 65 and older, 14 percent.

This is not new news; what I was curious about was to dig into the age-by-2024-vote crosstabs. This is where things got interesting.

Young Trump voters and young Harris voters consume creator content at essentially the same rate. Sixty-two percent of Trump voters under 35 get news or political information from creators at least somewhat often. For Harris voters under 35, it’s 60 percent.

Then the two groups split when asked if a creator they originally followed for something unrelated to politics, like fitness or cooking or sports, has changed their mind about a political or social issue. Nationally, only 13 percent of voters said they were persuaded. But, for 18 to 34-year-old Trump voters, 35 percent said they’d been persuaded. However, among young Harris voters, the same question returns 19 percent. Same exposure, roughly double the reported effect.

It isn’t a coalition trait either. Inside the Trump coalition the number falls off a cliff by age: 35 percent at 18–34, 11 percent at 35–49, 14 percent at 50–64, 4 percent at 65 and over. Inside the Harris coalition it’s nearly flat: 19, 16, 17, 9.

Two other findings point the same direction. Asked how much they trust political information from influencers compared with national news organizations, 39 percent of Trump voters under 35 say “about the same,” the highest share of any group in the poll, and only 11 percent say they trust influencers much less, against 33 percent nationally. They aren’t saying creators are better. They’re saying there’s no separate tier. That matters for the second reading above: if you don’t rank influencers below news organizations to begin with, admitting one changed your mind costs you nothing. There’s no status gap to cross. Older voters and Harris voters, who both still rank traditional news well above creators, are the ones for whom that admission would mean something.

A small qualitative study backs this up. Semafor reported on how Navigator Research ran a focus group of 26 voters ages 18 to 44 in battleground states last month, and when they asked people what "the media" meant to them, respondents named cable news and social media in the same breath. They didn't distinguish between the two, and they rated both as equally biased. That's not a representative sample the way my poll is, but it's a useful sanity check on the same idea: for a lot of younger voters, the hierarchy between "news" and "whatever's in my feed" was never built in the first place. There's nothing to flatten.

They aren’t naive about money, though. Twenty-one percent of young Trump voters name secretly paid political influencers as their top worry about 2026, more than double the rate among Trump voters 35 to 49. Young Harris voters name it at the same rate. Next week I’ll get into what that shared worry runs into: what kind of disclosure this same group actually wants, and why the thing campaigns need most from influencers is the hardest thing to make transparent.

Now, one thing to note about these questions. Nobody reliably knows what changed their mind. That isn’t a knock on the respondents; it’s how minds work. We reconstruct the story afterward, and we pick a cause that fits how we already understand ourselves.

So the question I asked doesn’t measure persuasion. It measures whether someone is aware of having been persuaded and willing to say so out loud to a survey panel.

That leaves two readings, and I can’t rule either one out with this data.

The first is that young Trump voters really are more permeable to creator influence than young Harris voters getting the same content.

The second is that they’re equally permeable and more comfortable saying so. If your side has spent a decade arguing that legacy media is captured and independent voices are the honest ones, then “a podcaster changed my mind” is a normal thing to admit. On the other side, where the story is that the other guys are the ones getting manipulated, the same experience may not get named the same way.

My read: it’s probably some of both, and the second one is doing more work than most people will want to credit. Either reading is a problem for how we currently talk about this. If it’s the first, one coalition has a persuasion channel the other doesn’t, and it’s largely unregulated. If it’s the second, every survey and focus group measuring creator influence is undercounting it on one side, and the left should read its own numbers here as a measurement artifact rather than good news.

There’s a version of this where I keep thinking back to 2016. Then it was Russians posing as Americans inside Facebook groups. What’s changed is the cost of doing it. An AI-generated creator, with a consistent face and voice and posting history, is now cheap to run at scale, and it doesn’t need to be run by a campaign or even a foreign government. It could just be someone monetizing an audience. If voters already can’t reliably identify what moved them among real people they chose to follow, I don’t think we know yet what happens when the “person” isn’t real at all.

One last thing. I follow a lot of creators: Aaron Parnas, Jessica Yellin, plenty of others across Instagram, TikTok, and Substack. Some of it is professional and some of it is that I like it. If you asked me whether any of them has changed my mind about something, I’d have to think hard, and I’m not confident I’d get the answer right. We’re about to spend two years arguing over rules for a channel where the people inside it, including me, can’t fully tell what it’s doing to them.

---

Will you help me get to SXSW?

I submitted a fireside chat about my new book, Disrupting Politics, and public voting is now open. Your vote could help send me to Austin this March.

If you’ve enjoyed my work, I’d be incredibly grateful for your support.

---