This article originally appeared in UnHerd.
By Eric Kaufmann, June 24, 2025
In a New York Times interview back in March, centrist Democratic pollster David Shor told Ezra Klein that, with the exception of minority women, young people preferred Donald Trump to Kamala Harris. “I find this […] shocking,” Klein replied. For years, he said, Republicans had been ignoring young people and concentrating on the senior vote. Yet “75-year-old white men supported Kamala Harris at a significantly higher rate than 20-year-old white men.”
Writing in the Atlantic on Friday, prominent author Jean Twenge — whose work has informed that of Jonathan Haidt — claimed that Shor is wrong, and that young people are actually pretty liberal.
The truth is that both Shor and Twenge are wide of the mark. Rather than being Right- or Left-wing, it is best to think of young people as more polarised than older age groups on culture war questions. There is also a significant opinion gap between elite and non-elite young people. With questions that offer response categories on a 5- or 7-point scale from “strongly oppose” to “strongly support” (for instance, that the US is a racist country), young people are more likely than their elders to select “strongly” options in either direction. Likewise, European youth incline towards the populist Right or Left rather than mainstream parties, even as the median young voter leans to the Left of their elders.
A common denominator between Shor and Twenge is that they both are basing their results on a single survey. Yet survey research is in crisis as attention spans and civic-mindedness wither. Only 1% of people answer polls. Within that percentage, young people are especially difficult to reach; among them, non-university students are even more elusive. As a result, the voting pattern of young Americans fluctuates a great deal from survey to survey. According to three prominent election studies scholars, “given the challenges in polling young voters, the best advice is to be skeptical of any one polling source and instead attempt to see what a collection of high-quality surveys have to say.”
Youth who self-select into surveys or find themselves contacted online seem to be more liberal, which skews polls. Consider that in surveys, the LGBT share of 18-20 year-olds in Britain is 25%; but on the census, just 7.6% of 18 to 24-year-olds identified this way.
The more a sample approximates the methods of an official census or vote, the better for corralling recalcitrant voters, who tend to be younger, less educated and less likely to follow politics. The exit poll thus has a major advantage over surveys. By approaching people randomly at polling stations, it catches people who ordinarily would not answer a poll or be recruited online for surveys. It is also geographically representative. No wonder exit polls tend to find a higher share of young Trump voters than many surveys.
We only get an exit poll once every two or four years, and the questions are limited, so we must rely on surveys. Among them, those that capture less educated voters tell a more representative story about young people. For instance, liberals have outnumbered conservatives by about 3:1 since 2016 at the elite Andover Phillips Academy, with little change over time. Likewise, the 50,000-strong Foundation for Individual Rights and Expression (FIRE) annual student surveys which over-index on elite universities find that the liberal-to-conservative share has been consistent since 2020 at around 5:2.
On the other hand, the Higher Education Research Institute (HERI) annual freshman survey of around 25,000 incoming students, which extends to lower-ranked colleges, finds a 3:2 liberal-to-conservative tilt with a 10 to 15-point Rightward shift between 2021 and 2024. Echoing the FIRE-HERI gap, academic work which tracks people over time finds that young respondents lacking a degree move to the cultural Right much more than graduates as they age.
Finally, young people have definitively moved Right on certain issues, such as trans. YouGov’s tracker reveals that Britons under 25 have shifted almost 30 points more conservative since August 2022 on the question of whether there should be separate bathrooms for men and women rather than unisex only (the trans-activist position). This is a bigger change than among older age groups.
Young people are evidently polarising between a woke segment and an anti-woke group that has grown in size since 2021. Formerly moderate or apolitical, often lower-status young people have moved Right, with this phenomenon apparently encompassing both men and women, which has slightly narrowed the youth gender gap. Generation Z, then, is at once more woke and more anti-woke than the rest of society, producing political and cultural consequences that are harder to pin down than Twenge and Shor might suggest.
Eric Kaufmann is Professor of Politics at the University of Buckingham and author of Taboo: How Making Race Sacred Led to a Cultural Revolution (Forum Press, 4 July) and a senior fellow at the Macdonald-Laurier Institute.
