The less voters knew, the more they liked Trump in 2024. Not Anymore
The least-engaged Americans have swung 25 points against him since 2024 — about twice the shift among everyone else. Trump has flattened the engagement gap.
Two notes: First, this is a special article with insights from Strength In Number’s recent polling with Verasight. We just finished writing our February poll, and will publish the results in two weeks. (Please pray to the news gods that our questions stay relevant until then.)
Second, I’ll be recording the weekly Strength In Numbers Podcast live on Substack with David Nir today (Thursday) at 2:00 PM Eastern. Subscribers can ask questions during the interactive portion of the taping, so join us via the Substack app or at the Strength In Numbers website.
In 2024, the voters who knew the least about politics were some of Donald Trump’s strongest supporters. One pre-election poll found Americans who didn’t consume any news at all said they’d vote for him over Kamala Harris by a 20-point margin, 60%1 to 40%.
Today, the president’s support among low-knowledge voters has cratered to just 43%, according to a new analysis of data from our January Strength In Numbers/Verasight poll. The share of 2024 voters who now disapprove of the president is well over 55%.
According to our poll, low-knowledge voters backed Trump by a net margin of 11 points in 2024. Now, however, the same low-knowledge voters say they disapprove of the president by 13 points — a 25-point shift away from the president.
High-knowledge voters were roughly evenly split in 2024 (voting for Harris by 2 points, per self-reports in our data), and have moved against Trump at a softer rate, to -14. The two groups are compared in the following chart:
This isn’t the first time I’ve found this pattern. In May 2025, I reported a similar result using data from YouGov on self-reported news attention. Then, Trump’s approval had fallen 33 points among people who said they paid the least attention to the news — roughly twice the drop among the most attentive Americans.
Voters who knew the least about politics were the ones most favorable to Trump in 2024. Not anymore; low-knowledge voters are now just as likely as high-knowledge voters to oppose his presidency. With fresh polling data, Strength In Numbers can report on other questions that distinguish these two groups. We can gain a better understanding of why they have changed their minds—and what this means for 2026.
This article is free to read, but it was not free to produce. I spent a couple of days crunching raw data to run these survey crosstabs, perfecting the charts, and writing, editing, and re-writing this analysis — and that’s not counting the time and resources that went into the poll itself.
If you find this kind of work valuable, please consider becoming a paid subscriber to Strength In Numbers. Your support is what makes it possible. Paid members also get access to all my Tuesday Deep Dives and support a growing suite of interactives at projects.gelliottmorris.com.
If a paid membership is not in your budget, the best thing you can do to support this business is to share this article across social media.
I. Low-info voters are more punishing of incumbents when conditions are poor
In our poll, we asked respondents to answer two factual questions about politics today — which party controls the U.S. House and which party controls the U.S. Senate — in order to compare the attitudes of high- and low-knowledge voters. We also asked people to tell us their approval of Trump’s presidency overall and for key issues, and whether they voted in 2024 (and if so, who they voted for). We call someone “high knowledge” if they know which party controls both the House and Senate today (about 75% of 2024 voters), and “low-knowledge” if they got either one of those questions wrong (about 25% of 2024 voters).
Based on their 2024 voting behavior, you might expect low-knowledge voters to be more conservative across the board. However, that’s not what the data show.
When you break out Trump’s approval on specific issues, the high- and low-knowledge groups look remarkably similar on almost everything. Net approval of Trump’s handling of jobs and the economy (-19 vs. -21), trade (-19 vs. -22), foreign policy (-15 vs. -18), immigration (-9 vs. -9), health care (-30 vs. -27), and government funding (-20 vs. -21) barely differ between the two groups.
The one exception is on prices. Low-knowledge respondents disapprove of Trump’s handling of prices and inflation by 40 percentage points in our poll, compared to -30 among high-knowledge adults. That 10-point gap is statistically significant.
I can think of a few explanations for this. First, as I have discussed in this newsletter previously, low-engagement voters over the last decade have been particularly pessimistic about prices and economic mobility. They were likelier to vote for Trump in 2024, I have argued, not because they were more conservative than other voters, but because they were more anti-incumbent. Now that Trump’s in power, we should expect this group to punish him more, too.
Second, demographics: The people in our poll who are most exposed to price increases are exactly the people in the low-knowledge group. Low-knowledge adults skew lower-income, younger, are less politically engaged, and are less educated. These groups spend a larger share of their income on groceries and essentials — and thus would be more likely to balk at price increases from things like tariffs and labor shortages. If your monthly grocery bill increases by $100 but you make $100,000, you might not really notice the difference. But for someone making $35,000 a year, an extra $100 a month is ~5% of their after-tax income.
Finally, our survey suggests there may be a compounding effect from health care costs. In our poll, 19% of low-knowledge adults report losing coverage or facing premium increases since the enhanced ACA subsidies expired at the end of 2025, compared to 11% for high-knowledge adults. Low-knowledge respondents drove Trump’s victory, but are now feeling the squeeze from his policies.
II. Low-knowledge voters are more elastic
So low-knowledge voters are more upset about prices. But that may not fully explain why they’ve moved twice as much against Trump as everyone else. The bigger reason is that their opinions were softer to begin with.
John Zaller’s The Nature and Origins of Mass Opinion (1992) is the best framework here. In his model of public opinion, people can resist political messaging only to the extent they have the context to evaluate counterarguments. High-knowledge voters have pre-existing ideological filters through which they can sort new information based on whether it affirms their existing worldview. Low-knowledge voters, on the other hand, lack those filters and absorb information less selectively, which means their opinions move more easily in whatever direction conditions push.
This shows up in our data. Among high-knowledge adults, 24% say they strongly approve of Trump’s job as president, while 50% strongly disapprove. So in total, 74% of highly knowledgeable Americans have “strong” feelings about the president. Now compare that to low-knowledge adults: just 15% strongly approve of Trump and 43% strongly disapprove — a total of 58%. The “soft” opinions of low-info voters are easier to shift.
III. How Trump won in 2024 (and how Democrats can win in 2026)
Over the long term, this all means voters who are habitually disengaged are more likely to vote based on the general conditions and direction of the country. When consumer sentiment is as low as it is now, that means incumbent parties are in danger.
This pattern explains a lot about what happened in 2024. Trump won low-knowledge voters without them knowing much about what he planned to do. They were unhappy with Biden, unhappy with prices, and voted accordingly. They weren’t making an ideological commitment to Trumpism — most aren’t even ideological at all. Now that conditions point the other way, Trump’s in trouble. If conditions keep deteriorating, these voters will keep moving against the president.
Trump won in 2024 in large part because ~one-quarter of the electorate wasn’t paying enough attention to his promises to know much about what he’d do as president. Now that they are seeing the results — especially on prices — they are just as anti-Trump as voters who spend all day consuming political news.
If Republicans aren’t winning high-knowledge respondents because of the negative news, and they aren’t winning low-knowledge respondents because of conditions... then they’re not winning elections.
Related Articles
What the Democrats can actually learn from the 1992 election
New poll: Trump loses ground on immigration as Democrats lead generic ballot by 8
If you’re a frequent reader of Strength In Numbers, I’m confident you will get a lot of value out of a paid subscription. You’ll get access to all of my analysis, including weekly Tuesday Deep Dives, and you’ll be supporting independent data journalism.
Ignoring respondents who said they were “not sure” who they’d vote for.
I think the other key feature of low-knowledge voters is that they are prone to magical thinking. They not only lack knowledge of the candidates’ policy positions; they also lack knowledge of how the government and the economy work, and thus they think that a new President can come in, wave a magic wand, and solve all their problems. Trump indulged this belief in the magical powers of the President in his 2024 campaign, making preposterous promises that knowledgeable people knew he could not fulfill. These low knowledge voters will thus usually oppose incumbents, because incumbents will never fulfill their fanciful ideas of what a President can actually achieve.
Another GEM of a post from G.E.M!