WASHINGTON (AP) — Elections are often boiled down to who won and lost, but it’s just as important to understand who voted and how, what motivated voters, and how they feel about important issues facing the country.
At The Associated Press, The AP Voter Poll by SSRS is a key part of answering those questions, explaining the voting coalitions that shaped the race for congressional control and decided key races for U.S. Senate and governor.
Launched in 2025, The AP Voter Poll is an extensive survey of verified voters that aims to tell the story behind election results. It blends the methodologies from AP VoteCast, which was conducted in general elections from 2018-2024, and traditional exit polls. This involves interviewing voters by phone and online in all 50 states in the days leading up to the election, and collecting additional data from in-person interviews at polling places in 15 states on Election Day.
The result is a detailed snapshot of the American electorate that helps explain how major groups voted, the issues they cared about, how they felt about the candidates and why they voted the way they did.
An initial set of results from the survey are first available at 5 p.m. ET on Tuesday, Nov. 3, with additional state-level and national findings that will be published throughout the night.
The Voter Poll is conducted by SSRS, a nonpartisan research firm, for The Associated Press and ABC News, CBS News, CNN, Fox News Media, NBC News and MS Now.
The AP Voter Poll uses a combination of methods to reach voters no matter how they cast a ballot, whether that’s by mail or in-person on Election Day.
Reflecting the complexity of American elections, The AP Voter Poll has two separate components. One part is a comprehensive survey conducted in all 50 states from Oct. 22 - Nov. 3. The other is an in-person survey at polling places in 15 key states on Nov. 3 that adds depth to those states’ findings.
First, The AP Voter Poll interviews approximately 75,000 voters in the lead-up to Election Day, using mailed survey invitations, phone and text interviews, and samples drawn from probability and nonprobability-based opinion panels. This is similar to the methodology used in AP VoteCast.
To find people to participate in the survey, The AP Voter Poll draws a random sample of registered voters from state voter files. Those voters are then contacted through the mail and invited to take the survey online or by phone. Some voters are also contacted by phone if they do not respond to the initial invitation to take part in the survey.
Other voters are recruited through the probability-based SSRS Opinion Panel, which matches each respondent to the voter file.
The AP Voter Poll also conducts interviews with voters recruited through the Verasight Verified Panel, where some participants are not selected randomly but are instead recruited through internet advertising or other methods. The respondents’ voter registration status is verified separately and the sample is carefully combined with the other pre-election data sources to make sure these additional interviews don’t overrepresent or underrepresent important demographic and ideological groups.
Many voters will have already cast their ballot by the time they take the survey, but if they haven’t yet voted, they will be recontacted on Election Day to verify their final vote choice and method.
Then, on Election Day, in-person interviews will be conducted with voters at polling places in 15 states – Alaska, Florida, Georgia, Iowa, Maine, Michigan, Minnesota, Nebraska, Nevada, New Hampshire, North Carolina, Ohio, Pennsylvania, Texas, and Wisconsin – adding approximately 25,000 interviews across those states. Interviews are also being conducted at early-vote centers in Alaska in late October.
The data from the pre-election and in-person surveys are then blended to ensure the combined surveys are representative of the estimated 2026 electorate in each state where an exit poll is conducted.
The result is a comprehensive national survey that draws on interviews from all 50 states to provide a detailed snapshot of the country as a whole, as well as separate state-by-state results in every state where a statewide race for U.S. Senator or governor, is taking place. Interviews conducted in the five states without a statewide race – Indiana, Missouri, North Dakota, Utah and Washington – will be included in the national survey, but will not be reported separately.
Voters are asked a standard set of demographic questions, such as their age, sex, race and ethnicity, education level and ideology. They’re also asked to share their views on what they think about the country’s current leadership and direction, how they are assessing the candidates on the ballot, and what they believe are the key issues facing the country, such as crime, the economy, foreign policy, healthcare and immigration.
The questionnaire asks about voters’ choices in races for U.S. Senate and governor, depending on where those races are happening. This year, California voters will also be asked how they voted on Proposition 40, a proposed one-time tax on billionaires in the state.
In addition, participants in the national survey are asked if they voted for the Democratic or Republican candidate in their local U.S. House race. This national finding – often called the “generic” House ballot – sheds light on the shape of the race for Congress across the country, but does not include results from specific congressional districts.
None of this information is a replacement for the vote count, and the topline findings – how many voters overall say they supported a specific candidate or party – won’t be reported. Instead, the survey is designed to provide valuable insight into how specific groups voted, and why.
The AP Voter Poll will be an integral part of the AP’s coverage of the 2026 election, helping explain voters’ motivations and illuminate how key groups shaped the outcome.
For example, in the 2025 New Jersey governor’s election, the AP Voter Poll results showed that Democrat Mikie Sherrill won after getting support from about half of white voters in the state and having a strong showing among Black, Hispanic and Asian voters.
And in the New York City mayoral race, the 2025 AP Voter Poll found that younger voters overwhelmingly voted for Zohran Mamdani in his successful campaign, while Andrew Cuomo’s coalition was much older.
The Voter Poll is also a valuable tool available to the AP Decision Team as it analyzes races on Election Night. Races are not called based solely on Voter Poll results, but the survey can shed additional light on a contest and help determine whether a winner can be declared. Other factors that the Decision Team will consider when calling a race include election results, vote history, registration and turnout data, demographic information and other data.
In some contests where one political party has an established history of lopsided victories and where other election data, such as overwhelming advantages in voter registration, advance ballot returns and pre-election polling, all point to a victory by that party, results from The Voter Poll may be the final data point that allows the AP to declare a winner as soon as polls close.
