2026 midterms · pending: no count yet
Which Pollster Was Most Accurate in 2026?
The national House popular vote is still being counted, so no 2026 generic-ballot pollster can be scored yet. 23 pollsters published polls in the final three weeks; this page scores each of them as soon as a count is available.
National House popular vote
Pending
The count is under way. We add the national House popular vote from a published count, with its source, the share of the vote counted and the date, and label every figure preliminary until states certify.
Final ElectionsUSA generic ballot average: D+8.0 (119 polls).
Pollsters, final three weeks
CSV →| # | Pollster | Polls | Avg margin | Error | Bias | Last poll |
|---|---|---|---|---|---|---|
| Ipsos | 5 | D+8.6 | pending | pending | Nov 2 | |
| YouGov | 4 | D+8.2 | pending | pending | Oct 30 | |
| ActiVote | 3 | D+6.3 | pending | pending | Oct 30 | |
| Focaldata | 2 | D+8.0 | pending | pending | Nov 1 | |
| Hart Research Associates | 2 | D+7.0 | pending | pending | Oct 30 | |
| Big Data Poll | 1 | D+12.0 | pending | pending | Oct 30 | |
| Clarity Campaign Labs | 1 | D+8.0 | pending | pending | Oct 21 | |
| Cygnal | 1 | D+9.0 | pending | pending | Oct 24 | |
| Echelon Insights | 1 | D+5.0 | pending | pending | Oct 23 | |
| Emerson College | 1 | D+10.0 | pending | pending | Oct 26 | |
| HarrisX | 1 | D+4.0 | pending | pending | Oct 26 | |
| John Zogby Strategies | 1 | D+6.0 | pending | pending | Nov 2 | |
| Marist University | 1 | D+8.0 | pending | pending | Oct 26 | |
| Morning Consult | 1 | D+6.0 | pending | pending | Oct 20 | |
| Pennsylvania State University | 1 | D+4.0 | pending | pending | Oct 20 | |
| Prolific | 1 | D+11.0 | pending | pending | Oct 22 | |
| RMG Research | 1 | D+8.0 | pending | pending | Oct 14 | |
| SSRS | 1 | D+10.0 | pending | pending | Oct 18 | |
| TIPP Insights | 1 | D+10.0 | pending | pending | Oct 30 | |
| Targoz Market Research | 1 | D+3.0 | pending | pending | Oct 22 | |
| The New York Times/Siena College | 1 | D+11.0 | pending | pending | Oct 19 | |
| The Rainey Center | 1 | D+8.0 | pending | pending | Oct 20 | |
| Tufts University | 1 | D+12.0 | pending | pending | Oct 27 |
Margin is Democrats minus Republicans. Error is the average distance between each poll's margin and the result's; bias is the average signed miss (D+ means the polls were more favourable to Democrats than the result). One population per poll, likely voters first; polls are credited to the firm that ran them. Window: polls ending October 13 to November 3, 2026.
How this is scored
This is the same method as the historical pollster scorecard, applied to the national generic ballot: every poll that finished fieldwork in the final 21 days, compared with the national House popular vote. Some districts have no contested race, which pulls the national House vote away from pure party preference, so even a perfect generic-ballot poll can miss by a point or two. We publish a preliminary score from the unofficial count and update it, with the date, when the count changes and again when states certify.
Read the analysis: ElectionsUSA analysis. The data behind this page is in the CSV, one row per poll.