You may have seen this New York Times graph a couple weeks ago:
This graph, based on New York Times/Siena polls from August 2024, was accompanied by a story about why Trump was doing so well among young men.
I saw some people complain on X/Twitter that the real story seemed to be young women, who are particularly pro-Harris.
But the real story is something different: This extraordinary gender gap does not replicate in other surveys. Nor do those surveys show that young men are leaning toward Trump.
First, here is a similar graph from an August 2024 YouGov survey (N=5,057) that I was involved in fielding:
You can see the dramatic difference with the New York Times/Siena graph. The gender gaps are smaller in every age group. And although young women are still the most pro-Harris group, young men are also pro-Harris. Indeed, although the New York Times/Siena surveys show almost no age differences among men, the YouGov survey shows a fairly typical difference, with younger men more Democratic than older men.
A second survey, this one by the Pew Research Center (N=9,270), shows this pattern:
In this survey, the gender gap is pretty similar in each age group. There is no unusually large gap among the youngest age group, unlike in the New York Times graph. The Pew Research Center data also shows that, in general, younger voters are more pro-Harris than older voters. This is true among both men and women. Of course, the exact estimates differ from the YouGov survey, particularly among the 30-44 age group.
Here is a third data point. In this same time period, NBC News conducted a survey among only 18- to 29-year-olds (the survey was fielded by Survey Monkey). Here is what they report about the gender gap:
Now, in a Harris-Trump race, the gender gap among Gen Z voters is significant. Young women said they’re going to vote for Harris for president by 30 points. Young men also said they favor Harris – but by only 4 points over Trump.
The NBC survey shows a larger gender gap than YouGov or Pew, but not nearly as large as the New York Times. They find that young men are a pro-Harris group – unlike the New York Times survey – but not as pro-Harris as YouGov and Pew surveys show.
Why do these polls get such different results? Of course, we can’t discount the effect of random variation due to sampling. Even in a big sample like the Pew Research Center poll, the margins of error for subgroups can be large. Pew estimates that the margin of error for 18- to 29-year-old men for a statistic like “the percent supporting Trump” is 8 percentage points.
But these varying results could also be the result of differences in polling methodologies – what political analyst Michael Podhorzer calls the “margin of pollster.” Although these polls were fielded at roughly the same time, they differed in how the surveys were conducted (phone vs. online), how the samples were built, how the samples were weighted, etc.
Ultimately, no survey is magic. And there’s no way to know which survey’s results are “correct.” But the range of estimates in these recent surveys means that an extraordinary Gen Z gender gap in the 2024 election is hardly a certainty.
