Kamala Harris faced off against Donald Trump in the 2024 US election. Illustration / Phil Welch
Kamala Harris faced off against Donald Trump in the 2024 US election. Illustration / Phil Welch
United States President Donald Trump’s victory last year was fuelled by gains among Hispanic, Asian, black and young voters, eroding key pillars of the Democratic Party’s coalition, according to an in-depth study of the 2024 election released today.
The Pew Research Centre’s validated voter survey found Trump had a turnouttailwind, retaining more of his 2020 supporters while also winning new voters and those who voted in 2024 but not in 2020.
Pew surveyed a random sample of nearly 9000 adult US citizens, matching respondents to official voter registration records to determine whether they voted.
The findings confirm many patterns seen in exit polls and other data sources while revealing new details about turnout and the make-up of each party’s voters.
The survey found Trump won white voters by 55% to 43% over Democrat Kamala Harris, identical to his 2020 margin over Joe Biden and similar to his 2016 margin over Hillary Clinton.
But nearly half of Hispanic voters supported Trump last year (48%), up from 36% in 2020 and 28% in 2016.
Network exit polls found Trump winning a similar 46% of Hispanic voters last year, the highest support for a Republican candidate in exit polling since at least 1980.
Trump’s support among Asian voters grew from 30% in 2020 to 40% last year, according to the Pew survey.
His support among black voters grew from 8% to 15%, including an increase of 12% to 21% among black men.
While low in absolute terms, Trump’s support among black voters is the highest for a Republican in decades, according to exit polls.
Republican presidential nominee, former US President Donald Trump arrives to speak during an election night event in West Palm Beach, Florida. Photo / Getty Images
While Trump has long appealed strongly to white voters without college degrees, his 2024 support among Hispanic and black voters increased among university graduates as well as those with some college or less education.
Pew noted that Trump’s gains among these groups “were largely the result of differences in which voters turned out in the 2020 and 2024 elections”, with a relatively small share of voters switching which party’s candidate they supported.
Younger voters have consistently favoured Democrats over the past two decades, but Trump made headway among this group as well.
He won 39% of voters under age 30, up from 35% in 2020 and 28% in 2016.
Trump made similar gains among voters ages 30 to 49 and 50 to 64, while maintaining a slight edge with seniors, but not improving on his previous margins.
Harris’ loss in November reignited a familiar debate among Democrats: Was their defeat primarily driven by shortcomings in voter persuasion or voter turnout?
This question, and the lessons that are drawn from it, speak to whether the party needs to change its stances to appeal to more voters or prioritise motivating its base in future elections.
While Pew’s latest report doesn’t provide a definitive answer, it offers some compelling insights.
At first glance, the data points to turnout as Harris’ main challenge. Trump managed to retain 85% of his 2020 voters, with just 11% opting out of voting and only 3% switching to Harris.
In contrast, a smaller 79% of Biden’s 2020 voters turned out and voted for Harris, while 15% did not vote and 5% crossed over to Trump.
A closer look complicates the picture. Each election has a significant amount of churn, and Harris suffered from many Biden voters staying home.
But an analysis of the Pew data by the Washington Post suggests that she gained most of those votes back from people who turned out in 2024 after not voting in 2020 or who were too young to vote then.
Moreover, the 5% of Biden voters who switched to Trump had an outsize impact. Each of these roughly four million voters not only subtracted from Harris’ total but added to Trump’s, effectively counting twice on the margin.
“The distinction between turnout and persuasion is blurry,” Pew senior survey adviser Scott Keeter wrote in an email.
The turnout drop-off among Biden voters “seems like a turnout problem, but we know that a lot of those Biden drop-offs were people who did not favour Harris (based on our interviews with non-voters). Those were effectively switchers who just didn’t make the journey all the way to Trump.”
Supporters react to election results during an election night event for US Vice-President and Democratic presidential candidate Kamala Harris at Howard University in Washington, DC, on November 5, 2024. Photo / AFP
In addition to shedding fewer 2020 supporters, Trump won people who didn’t vote in 2020 but turned out in 2024 by 54% to 42% over Harris.
Keeter wrote that many polls find “a shift in the Republican direction across the electorate between 2020 and 2024. So, the pool of people who did not vote in 2020 had probably become less Democratic-leaning than it was in 2020.”
Indeed, higher turnout probably would not have helped Harris.
Pew’s data shows that among those who didn’t vote in 2024, Trump held a slight edge: 44% to Harris’ 40%.
In other words, boosting turnout could have slightly worsened her position. In Pew’s model simulating full turnout, Trump’s margin would have remained essentially unchanged.
Taken together, these findings suggest that the story of Harris’ loss isn’t simply about turnout or persuasion - it’s about both.
She faced a meaningful defection from the Democratic base that was partially masked by gains from new voters, and she struggled to inspire the kind of enthusiasm needed to keep marginal supporters engaged. For Democrats, the lesson may not lie in choosing between persuasion and turnout, but in recognising that electoral success demands strength on both fronts.
The 2024 Pew Research Centre survey was conducted on November 12-16 among 8942 US citizens 18 years or older, including 7100 validated voters. All respondents are members of the Centre’s American Trends Panel, an ongoing survey panel recruited through random sampling of US households. The margin of sampling error for results among all validated voters is plus or minus 1.5% and is larger for subgroups.
