University of California, Riverside

11/13/2024 | Press release | Distributed by Public on 11/13/2024 14:13

Were the 2024 election polls wrong? UCR expert weighs in

Most political polls for the 2024 presidential election saw a close race in the electoral college, and a resounding popular-vote victory for Democrat Kamala Harris. Trump won the electoral college 312 to 226, and the popular vote by more than three million votes. We asked polling and survey expert Andy Crosby with the UCR School of Public Policy to tell us what the polls may have missed, or whether we - the voters - misinterpreted the polls. Crosby writes:

UC Riverside polling expert Andy Crosby

How accurate were the polls in the 2024 election?

Overall, high-quality polls were quite accurate. For example, the final national New York Times/Siena of likely voters conducted October 20-23 showed a tie (48%-48%). As of Monday, November 11, with most votes now tallied, former President Trump holds a 50.2%-48.1% lead nationally - a result which is within that poll's margin of error of 2.2%.

Although one might initially think a final election result differing than the final polls indicates that polls were incorrect, this is why polls publish their margin of error: it tells the reader how much variability we might expect in the result. This year, our final election results largely appear within that margin of error, indicating the polls were in fact quite accurate. We also know that some voters make their final voting decision on election day.

As polls also predicted, race, ethnicity, and gender also played an important dynamic in the final electoral outcomes. Notably, according to current exit poll data, white respondents favored Trump (57%-41%), whereas black respondents heavily favored Harris (85%-13%). Asian respondents also favored Harris (54% Harris, 39% Trump), as did Hispanic respondents (52% Harris, 46% Trump). Gender was also an important factor: men favored Trump (55%-42%), whereas women favored Harris (53-45%).

Results in Swing States

Polls were also accurate across several swing states, including perhaps the most important battleground state, Pennsylvania. The final New York Times/Siena in Pennsylvania conducted October 29-November 2 showed the state as a tie (48%-48%) in its final poll before the election. As of Monday, November 11, Trump leads the state 50.5% to 48.5% with 99% of the vote in. That result is again well within the Times/Siena poll's margin of error of 3.5% (state polls tend to have slightly larger margins of error in part because fewer people are sampled).

Michigan was similar: the final Times/Siena poll conducted October 29-November 2 was 47%-47% with a margin of error of 3.5%. As of Monday (November 11), Trump leads the state 49.7 percent to 48.3 percent with 99 percent of the vote in, again well within the poll's margin of error. Arizona only had approximately 93 percent of its vote counted as of this writing, but the polls again at least initially appear to be accurate: the final Times/Siena poll in Arizona conducted October 25-November 2 showed Trump leading 49%-45% with a 3.5% margin of error, and current election returns show 52.3%-46.6%.

Looking to the Future

Of course, the 2024 election is not the end of electoral polls: we will soon start to see polls regarding soon-to-be President Trump's approval rating and various policies, and of course we are not far from starting to track the 2026 midterm elections. Stay tuned and happy polling!

Share this Article

Media Contacts