The generic congressional ballot asks respondents which party's congressional candidate they would support, generally without naming the actual candidates in their district.
It provides a view of the national partisan environment. It is not a nationwide House election result or a direct count of future seats.
What the question leaves out
A generic question simplifies the choice. Real ballots include candidates with individual records, campaign strengths, and local connections.
A voter may prefer a party nationally while supporting a particular representative from another party. Another respondent may express a party preference but not participate in the election.
For that reason, always read the question and the population definition. A survey of adults and a survey of likely voters need not estimate the same thing.
Why analysts follow it
A consistently measured generic ballot can help show changes in broad congressional preference. The most useful comparison uses similar methods over time.
Pew's historical explanation describes this measure as national preference for a party's congressional candidate. Its examples concern an earlier election and should not be mistaken for current 2026 figures. Source: Pew Research Center.
Why a lead does not equal a seat total
Suppose a hypothetical generic ballot gives Party A 46% and Party B 43%. Translating that difference into a House majority requires assumptions about turnout and where support is located.
District boundaries and the distribution of close races matter. A national swing could change many seats or relatively few, depending on the underlying map.
Do not apply a fixed rule such as “every national point equals a certain number of seats” without identifying and testing the model behind it.
How to use the measure
Track the fieldwork dates, eligible population, sponsor, wording, and uncertainty. Compare multiple credible releases and inspect whether a reported trend comes from actual movement or a change in methodology.
Use the generic ballot alongside district evidence. For Senate control, focus on the states with relevant Senate contests; national House preference cannot directly resolve them.
Related reading: national vs. state polls and polling averages.