2026 General Election · Nov 3
Each step ≈ ~5 points of national popular-vote margin. The slider auto-starts at the blended generic-ballot + approval prediction below; drag to apply your own call. The House feels the full wave; the Senate feels about half (it's map- and candidate-driven). "Cook forecast" never moves.
| Year | President | Approval | House vote | Seats | |
|---|---|---|---|---|---|
| 1994 | Clinton (D) | 46% | R+7.0 | R +54 | 🌊 |
| 1998 | Clinton (D) | 66% | D+1.0 | R +-5 | |
| 2002 | G.W. Bush (R) | 63% | R+5.0 | D +-8 | |
| 2006 | G.W. Bush (R) | 38% | D+8.0 | D +30 | 🌊 |
| 2010 | Obama (D) | 45% | R+6.8 | R +63 | 🌊 |
| 2014 | Obama (D) | 42% | R+5.7 | R +13 | |
| 2018 | Trump (R) | 42% | D+8.6 | D +40 | 🌊 |
| 2022 | Biden (D) | 41% | R+2.8 | R +9 |
The out-party almost always wins the national House vote; a "wave" (out-party clears ~+5, i.e. wave ~+1.0) hit in 1994, 2006, 2010, 2018. The generic ballot measures that national House vote directly; approval only predicts it (correlation ≈ 0.66–0.71), which is why both are blended above. Either way the Senate is map- and candidate-driven and barely responds — in 2018 a D+8.6 House wave still lost Senate seats.
Cook ratings as of 2026-09-11.
Temporary — resets on reload. Updates this forecast only.
Temporary — resets on reload. Updates this forecast only.
Kalshi prediction-market prices · races with no market are toss-ups.
Temporary — resets on reload. Updates this forecast only.
Cook ratings as of 2026-09-11.
Temporary — resets on reload. Updates this forecast only.
Temporary — resets on reload. Updates this forecast only.
Kalshi prediction-market prices · races with no market are toss-ups.
Temporary — resets on reload. Updates this forecast only.