For pollsters, the 2024 presidential election marked progress in accurately forecasting results with Donald Trump leading the ticket—but challenges remain. After notable missteps in 2016 and 2020, pollsters adjusted strategies to capture Trump’s elusive base more effectively, though not without lingering inaccuracies.
Correcting Course After Past Failures
Pollsters struggled in 2016 and 2020 to account for the underrepresentation of Trump voters, a group that tends to distrust institutions, including pollsters. GOP pollster Whit Ayres noted that many Trump supporters refused to participate in surveys when contacted by outlets like The New York Times or The Washington Post.
This year, researchers adopted new methods, including:
- Adjusting modeling of the likely electorate.
- Weighting non-college-educated voters more heavily.
- Expanding outreach to underrepresented groups, especially Republican-leaning demographics.
These adjustments yielded improved results, with final polling aggregates largely predicting the correct popular vote margins and swing state outcomes within the margin of error.
Surprising Strength in Swing States
Polls showed the race tightening in swing states during the final weeks, with Trump holding comfortable leads in Arizona and Nevada. While final results aligned with poll aggregates, Trump still overperformed by approximately 3 percentage points, possibly due to late-deciding voters.
Democratic pollster Paul Maslin highlighted that undecided voters, who typically split against Republicans, skewed toward Trump this time. GOP pollster Brent Buchanan attributed this shift to late momentum, aided by a high-profile endorsement from Joe Rogan, which energized Trump’s base in the campaign’s closing days.
Persistent Challenges
Despite progress, pollsters faced high-profile errors. Legendary pollster Ann Selzer, long associated with accurate forecasts, ended her relationship with the Des Moines Register after a late poll incorrectly showed Kamala Harris leading Trump in Iowa by 3 points. Trump ultimately won Iowa by over 13 points, exposing continued difficulties in predicting his support.
Maslin noted that issues tied to Trump voters’ mistrust of institutions remain unresolved. While some pollsters are optimistic these challenges will subside in future elections without Trump, others caution against complacency, emphasizing the ongoing need for innovation in polling methodology.
A Turning Point for Polling?
The 2024 election demonstrated progress in understanding the “Trump effect” in polling, but it remains unclear whether lessons learned this cycle will fully apply to future contests. With Trump sweeping all battleground states and outperforming polls, questions linger about whether polling can reliably capture shifts in voter sentiment during high-stakes elections.
COMMENTS