Asking voters which party they want for Congress has long been a critical indicator of the national mood. Here's what the polls have said for elections going back to 1994.
Polls this far out are definitely not super informative, but there is a pattern where the president's party mostly loses ground or slightly gains in the best case scenario.
The polls themselves were collected from a bunch of sources (listed at the bottom of this page). Roper Center, PollingReport.com, RCP, the old FiveThirtyEight pages (via internet archive), and Gallup.
They're then aggregated using a similar system to FiveThirtyEight's poll averages: weighting and house effects adjustments for each poll before averaging. This is my detailed methodology explainer:
Is it a bit early to project 2026 with 2025 polling data?
Polls this far out are definitely not super informative, but there is a pattern where the president's party mostly loses ground or slightly gains in the best case scenario.
The polls themselves were collected from a bunch of sources (listed at the bottom of this page). Roper Center, PollingReport.com, RCP, the old FiveThirtyEight pages (via internet archive), and Gallup.
They're then aggregated using a similar system to FiveThirtyEight's poll averages: weighting and house effects adjustments for each poll before averaging. This is my detailed methodology explainer:
https://plusminus4.substack.com/p/about-ballot-beacon-polling-averages