What to know about exit polls and how they’ve changed for this year’s elections

By Jennifer Agiesta, CNN

(CNN) — For decades, Americans have relied on exit polls to illuminate the story of what happened on election night. By gathering opinions from voters largely after they’ve cast their ballots, exit polls fill in the details that our secret-ballot elections otherwise leave impossible to know.

This year, for the first time since 2016, CNN, ABC, CBS, Fox News, NBC and the Associated Press are working together to produce this critical research, in collaboration with SSRS, a nonpartisan research company that also conducts CNN’s polling. On behalf of the six media organizations, SSRS will conduct The Voter Poll in California, New Jersey, New York City and Virginia to cover the marquee contests on this November’s slate. You’ll see the results here as CNN’s Exit Poll.

Reporting on CNN’s exit poll results will begin at 5 p.m. ET. More complete data will be available in the CNN Election Center after polls close in each location.

Traditionally, exit polling has leaned heavily on in-person interviews of a randomly selected sample of voters at different Election Day polling locations. That remains a key part of the polling this time around. But to include people who vote early or vote by mail, those in-person interviews will be combined with survey results gathered before Election Day to ensure that exit polls reflect the views of the full electorate, regardless of when they vote or how they cast their ballot.

There are also adjustments to how the sample is selected to better reflect the electorate, to be more confident that people included in the poll are in fact voters, and to ensure that the adjustments pollsters make on the back end, known as weighting, result in a demographic profile that reflects the people who actually turned out.

Here’s what to know about how this year’s exit polls are done:

• The surveys in California, New Jersey and Virginia will each include results from about 4,000 voters, while the New York City survey will include about 3,700 voters.
• Voters can take the survey in several ways. Those interviewed before Election Day will take it online or by phone, while those interviewed at their polling place fill out a paper questionnaire. Participants will answer the same questions about their characteristics and opinions whether they take the survey online, on the phone or on paper.
• To ensure that everyone taking the poll is an eligible voter, those who do so preelection are either selected from or matched to a voter file, which is a commercially available list of registered voters. Those who are interviewed in person are not matched in this way but will have demonstrated eligibility by voting at their polling place.
• People who are initially interviewed before they cast their ballots will be recontacted by the researchers on Election Day to confirm whether they voted, how they cast their ballot and for whom.
• Once results are collected, researchers apply weights to align the results to what we know about the full electorate. For example, adjustments are made for demographics such as age, race, gender and education based on what we know about how voters who took the poll differ from those who did not. Researchers also adjust based on their best estimates of turnout and on how many people are expected to vote by mail, in person before Election Day, or in person on Election Day itself. Finally, once votes are counted, the surveys are weighted to match the final vote results to produce the best possible picture of who voted and what they think.

A technical description of the methodology of the poll can be found here.

The-CNN-Wire
™ & © 2025 Cable News Network, Inc., a Warner Bros. Discovery Company. All rights reserved.