A district poll estimates a named race. The national generic ballot asks which party voters prefer for Congress. Its benchmark is the national House vote, not the number of seats a party wins.
Select a cycle to inspect its error and the number of eligible estimates. The changing sample is part of the story.
House · individual districts
Original archive calculation: final 1–14 days by median field date; latest estimate per firm/race, same-date ties averaged. Includes retained general, special and runoff contests within source cycles; not AAPOR’s methodology. Outcomes and exclusions are documented below.
| Cycle | Mean absolute error | Mean signed error | Estimates | Contests |
|---|---|---|---|---|
| 1998 | 7.81 | −4.0 | 26 | 24 |
| 2000 | 5.28 | −0.4 | 18 | 15 |
| 2002 | 5.05 | +0.1 | 40 | 28 |
| 2004 | 5.02 | +1.1 | 34 | 27 |
| 2006 | 5.34 | +1.1 | 124 | 75 |
| 2007 | Unavailable | — | 0 | 0 |
| 2008 | 5.93 | +0.7 | 48 | 42 |
| 2009 | Unavailable | — | 0 | 0 |
| 2010 | 7.02 | +0.4 | 102 | 77 |
| 2011 | Unavailable | — | 0 | 0 |
| 2012 | 5.73 | −3.8 | 59 | 43 |
| 2013 | Unavailable | — | 0 | 0 |
| 2014 | 6.57 | +3.9 | 50 | 40 |
| 2015 | Unavailable | — | 0 | 0 |
| 2016 | 6.57 | +5.7 | 19 | 17 |
| 2017 | Unavailable | — | 0 | 0 |
| 2018 | 5.42 | −0.6 | 107 | 79 |
| 2019 | Unavailable | — | 0 | 0 |
| 2020 | 6.52 | +5.7 | 37 | 30 |
| 2021 | Unavailable | — | 0 | 0 |
| 2022 | 5.50 | −2.8 | 29 | 24 |
Mean absolute margin error was 7.81 points in 1998 and 5.50 in 2022, compared with 6.52 in 2020. Across the covered cycles, the smallest observed miss was 5.02 in 2004; the largest was 7.81 in 1998.
The latest cycle contains 29 selected firm/race estimates across 24 contests. Its signed error, −2.8 points, overstated Republican margins. Changing races and pollsters limit comparisons.
Mean absolute margin error was 6.13 points in 1998 and 2.63 in 2022, compared with 4.00 in 2020. Across the covered cycles, the smallest observed miss was 1.31 in 2012; the largest was 6.13 in 1998.
The latest cycle contains 26 selected firm/race estimates across 1 contest. Its signed error, +1.7 points, overstated Democratic margins. Changing races and pollsters limit comparisons.
AAPOR’s weighted analysis · October 23–November 5 · raw D–R margin error
House district polls cover only 20 contests; this is not an estimate for all 435 districts. Presidential congressional-district polls concern electoral votes and are distinct from House races.
Not by itself. Seat allocation depends on district boundaries, where votes are distributed, uncontested races, and individual candidates. A national vote margin is not a seat-count forecast.
No. Polling concentrates in selected, often competitive districts. In AAPOR’s 2024 final-window evaluation there were only 28 published matchups across 20 House contests. A large miss in that sample is not a measured error for every district.
Different windows can contain different pollsters and races. These are descriptive comparisons, not a causal experiment.
The interactive timing chart loads with JavaScript. Every value is also in the downloadable dataset.
Latest per firm/race within each non-overlapping midpoint window.
