Forecasts of the 2026 midterm elections from polls, prediction markets, professional forecasters, academic models and the PLSC 2219 class model, updated daily.
A forecast of every 2026 House and Senate race, built by the class in problem sets 4 through 7. It uses presidential approval, the economy and each seat's partisan lean. As of 2026-09-30.
The model stops updating on November 1, 2026. The forecast as of that date is final.
The model is re-run every day with that day's national inputs and the district maps in force that day. Values through 2026-09-28 were computed afterward, using the approval average available on each date and current economic data.
Latest: 94%
Latest: 41%
Latest: 236 · line at 218
Latest: 50 · line at 51
Latest: 54.6% · line at 50%
Latest: 36.9%
Latest: +16.0%
Latest: +1.1%
The share of 10,000 simulated elections that end with each number of Democratic seats. Blue bars are Democratic majorities.
Senate totals include the Democratic seats not up this year. A 50–50 Senate counts as Republican control.
Races where each side has at least a 5% chance, ordered from least to most likely to be won by the Democratic side.
| State | Democratic side | Republican side | Predicted D share | 80% range | P(D wins) |
|---|---|---|---|---|---|
| MS | Scott Colom | Cindy Hyde-Smith | 41.0% | 33.2–48.8 | 7% |
| NE | Dan Osborn | Pete Ricketts | 41.4% | 33.6–49.2 | 8% |
| KY | Charles Booker | Andy Barr | 41.6% | 33.9–49.4 | 8% |
| AL | Everett Wess | Barry Moore | 41.8% | 34.0–49.5 | 9% |
| SC | Annie Andrews | Lindsey Graham | 43.0% | 35.2–50.8 | 12% |
| KS | Adam Hamilton | Roger Marshall | 43.2% | 35.4–51.0 | 13% |
| AK | Mary Peltola | Dan Sullivan | 44.5% | 36.7–52.3 | 18% |
| LA | Jamie Davis | Julia Letlow | 45.0% | 37.3–52.8 | 21% |
| FL | Angela Nixon | Ashley Moody | 45.4% | 37.6–53.1 | 22% |
| OH | Sherrod Brown | Jon Husted | 45.5% | 37.7–53.2 | 23% |
| MT | Alani Bankhead | Kurt Alme | 45.8% | 38.0–53.6 | 24% |
| IA | Josh Turek | Ashley Hinson | 48.8% | 41.0–56.6 | 42% |
| TX | James Talarico | Ken Paxton | 49.0% | 41.2–56.7 | 43% |
| NC | Roy Cooper | Michael Whatley | 52.6% | 44.8–60.4 | 67% |
| ME | Troy Jackson | Susan Collins | 52.8% | 45.0–60.6 | 68% |
| MI | Abdul El-Sayed | Mike Rogers | 53.6% | 45.8–61.4 | 72% |
| NH | Chris Pappas | John E. Sununu | 55.4% | 47.6–63.2 | 81% |
| MN | Peggy Flanagan | Michele Tafoya | 55.8% | 48.0–63.6 | 83% |
| GA | Jon Ossoff | Mike Collins | 57.1% | 49.4–64.9 | 88% |
| IL | Juliana Stratton | Don Tracy | 58.9% | 51.1–66.7 | 93% |
| District | Democratic side | Republican side | Predicted D share | 80% range | P(D wins) |
|---|---|---|---|---|---|
| NC-8 | Colby Watson | Mark Harris | 42.9% | 37.4–48.5 | 5% |
| OK-5 | Jena Nelson | Stephanie Bice | 43.0% | 37.4–48.6 | 5% |
| TN-6 | Mike Croley | Johnny Garrett | 43.2% | 37.7–48.8 | 6% |
| FL-26 | Nicole Locklin | Mario Diaz-Balart | 43.4% | 37.8–49.0 | 7% |
| IN-5 | J. D. Ford | Victoria Spartz | 43.5% | 37.9–49.1 | 7% |
| MD-1 | Dan Schwartz | Andrew P. Harris | 43.6% | 38.0–49.2 | 7% |
| TX-15 | Bobby Pulido | Monica de la Cruz | 43.6% | 38.0–49.2 | 7% |
| WI-6 | Brad Smith | Glenn Grothman | 43.6% | 38.0–49.2 | 7% |
| NC-5 | Chuck Hubbard | Virginia Foxx | 43.7% | 38.1–49.3 | 7% |
| NC-10 | Ashley Bell | Pat Harrigan | 43.7% | 38.1–49.3 | 7% |
| WI-8 | Rick Crosson | Tony Wied | 43.7% | 38.1–49.3 | 8% |
| NC-6 | Cyril Jefferson | Addison McDowell | 43.9% | 38.3–49.5 | 8% |
| AZ-8 | Bernadette Greene-Placentia | Abraham Hamadeh | 44.0% | 38.4–49.6 | 8% |
| FL-18 | Curtis Gibson | Scott Franklin | 44.0% | 38.5–49.6 | 9% |
| AR-2 | Chris Jones | French Hill | 44.2% | 38.6–49.8 | 9% |
| GA-11 | Chris Harden | John Cowan | 44.2% | 38.6–49.8 | 9% |
| NC-13 | Paul Barringer | Brad Knott | 44.3% | 38.7–49.8 | 9% |
| FL-21 | James Martin | Brian Mast | 44.4% | 38.8–50.0 | 10% |
| FL-8 | Jennifer Jenkins | Mike Haridopolos | 44.4% | 38.8–50.0 | 10% |
| TX-24 | Kevin Burge | Beth Van Duyne | 44.4% | 38.8–50.0 | 10% |
| NC-9 | Richard Ojeda | Richard Hudson | 44.5% | 38.9–50.1 | 10% |
| OH-8 | Vanessa Enoch | Warren Davidson | 44.6% | 39.0–50.2 | 11% |
| WI-7 | Fred Clark | Michael Alfonso | 44.7% | 39.1–50.2 | 11% |
| FL-12 | Kimberly Overman | Gus Bilirakis | 44.7% | 39.1–50.3 | 11% |
| OK-1 | John Croisant | Mark Tedford | 44.8% | 39.2–50.3 | 11% |
| SC-5 | Mallory Dittmer | Wes Climer | 44.8% | 39.2–50.4 | 12% |
| TN-5 | Chaz Molder | Charlie Hatcher | 44.8% | 39.2–50.4 | 12% |
| MN-8 | Trina Swanson | Pete Stauber | 44.8% | 39.3–50.4 | 12% |
| NC-14 | LaKesha Womack | Tim Moore | 44.9% | 39.3–50.5 | 12% |
| TX-2 | Shaun Finnie | Steve Toth | 44.9% | 39.3–50.5 | 12% |
| TX-10 | Caitlin Rourk | Chris Gober | 45.0% | 39.4–50.6 | 13% |
| AZ-2 | Jonathan Nez | Eli Crane | 45.1% | 39.5–50.6 | 13% |
| SC-2 | Zyon Khalifa | Joe Wilson | 45.1% | 39.5–50.7 | 13% |
| TX-22 | Marguette Greene-Scott | Trever Nehls | 45.1% | 39.5–50.7 | 13% |
| GA-12 | Ceretta Smith | Rick Allen | 45.1% | 39.5–50.7 | 13% |
| FL-27 | Eliott Rodriguez | Maria Elvira Salazar | 45.1% | 39.5–50.7 | 13% |
| GA-10 | Pamela DeLancy | Houston Gaines | 45.2% | 39.6–50.7 | 13% |
| AK-AL | Bill Hill | Nick Begich | 45.2% | 39.7–50.8 | 14% |
| NE-1 | Chris Backemeyer | Mike Flood | 45.3% | 39.7–50.9 | 14% |
| TX-21 | Kristin Hook | Mark Teixeira | 45.4% | 39.8–51.0 | 14% |
| NC-7 | Kim Hardy | David Rouzer | 45.5% | 39.9–51.0 | 15% |
| WA-4 | John Duresky | Amanda McKinney | 45.5% | 39.9–51.1 | 15% |
| NC-3 | Raymond Smith | Greg Murphy | 45.5% | 39.9–51.1 | 15% |
| FL-13 | Leela Gray | Anna Paulina Luna | 45.7% | 40.1–51.3 | 16% |
| TN-9 | Justin Pearson | Brent Taylor | 45.7% | 40.1–51.3 | 16% |
| TX-38 | Melissa McDonough | John Bonck | 45.7% | 40.1–51.3 | 16% |
| NY-21 | Blake Gendebien | Anthony Constantino | 45.8% | 40.2–51.4 | 17% |
| MN-1 | Jake Johnson | Brad Finstad | 45.9% | 40.3–51.4 | 17% |
| VA-5 | Tom Perriello | John McGuire | 45.9% | 40.3–51.5 | 17% |
| AZ-5 | Elizabeth Lee | Mark Lamb | 45.9% | 40.3–51.5 | 18% |
| NY-2 | Patrick Halpin | Andrew Garbarino | 46.0% | 40.4–51.6 | 18% |
| WA-5 | Carmela Conroy | Michael Baumgartner | 46.1% | 40.5–51.7 | 19% |
| NJ-2 | Zack Mullock | Jeff Van Drew | 46.2% | 40.6–51.8 | 19% |
| TX-9 | Leticia Gutierrez | Alex Mealer | 46.3% | 40.7–51.9 | 20% |
| FL-4 | LaShonda Holloway | Aaron Bean | 46.3% | 40.7–51.9 | 20% |
| OH-7 | Brian Poindexter | Max Miller | 46.7% | 41.1–52.3 | 22% |
| CO-5 | Jessica Killin | Jeff Crank | 46.7% | 41.1–52.3 | 22% |
| CO-3 | Dwayne Romero | Jeff Hurd | 46.7% | 41.1–52.3 | 23% |
| FL-2 | Amanda Green | Austin Rogers | 47.1% | 41.5–52.7 | 25% |
| TX-32 | Dan Barrios | Jace Yarbrough | 47.1% | 41.5–52.7 | 26% |
| OH-15 | Don Leonard | Mike Carey | 47.1% | 41.6–52.7 | 26% |
| NC-11 | Jamie Ager | Chuck Edwards | 47.4% | 41.8–53.0 | 27% |
| NY-1 | Chris Gallant | Nick LaLota | 47.4% | 41.8–53.0 | 27% |
| MO-2 | Fred Wellman | Ann Wagner | 47.4% | 41.8–53.0 | 28% |
| GA-1 | Amanda Hollowell | Jim Kingston | 47.7% | 42.1–53.3 | 30% |
| IA-1 | Christina Bohannan | Mariannette Miller-Meeks | 47.8% | 42.2–53.4 | 31% |
| PA-8 | Paige Cognetti | Rob Bresnahan | 47.8% | 42.2–53.4 | 31% |
| WI-3 | Rebecca Cooke | Derrick Van Orden | 47.9% | 42.3–53.5 | 32% |
| FL-11 | James Pericola | Joe Strada | 48.1% | 42.6–53.7 | 34% |
| OH-10 | Kristina Knickerbocker | Mike Turner | 48.1% | 42.6–53.7 | 34% |
| KY-6 | Zach Dembo | Ralph Alvarado | 48.2% | 42.6–53.8 | 34% |
| NV-2 | Teresa Benitez-Thompson | David Flippo | 48.5% | 42.9–54.1 | 37% |
| TX-23 | Katy Stout | Brandon Herrera | 48.5% | 42.9–54.1 | 37% |
| VA-1 | Shannon Taylor | Rob Wittman | 48.6% | 43.0–54.2 | 37% |
| MI-4 | Sean McCann | Bill Huizenga | 48.7% | 43.1–54.2 | 38% |
| PA-10 | Janelle Stelson | Scott Perry | 48.7% | 43.2–54.3 | 39% |
| FL-16 | Kelly Kirschner | Sydney Gruters | 49.0% | 43.4–54.6 | 41% |
| SC-1 | Nancy Lacore | Mark Smith | 49.1% | 43.5–54.7 | 42% |
| WI-1 | Mitchell Berman | Bryan Steil | 49.2% | 43.6–54.8 | 43% |
| IA-3 | Sarah Trone Garriott | Zach Nunn | 49.4% | 43.9–55.0 | 45% |
| FL-7 | Bale Dalton | Ryan Elijah | 49.6% | 44.0–55.2 | 46% |
| MT-1 | Sam Forstag | Aaron Flint | 49.7% | 44.2–55.3 | 48% |
| PA-7 | Bob Brooks | Ryan Mackenzie | 50.0% | 44.4–55.5 | 50% |
| TX-35 | Johnny Garcia | Carlos De La Cruz | 50.5% | 44.9–56.1 | 54% |
| FL-22 | Pia Dandiya | Casey Askar | 50.6% | 45.0–56.2 | 55% |
| MI-7 | William Lawrence | Tom Barrett | 50.6% | 45.0–56.2 | 55% |
| IA-2 | Lindsay James | Joe Mitchell | 50.6% | 45.0–56.2 | 56% |
| AZ-6 | JoAnna Mendoza | Juan Ciscomani | 50.7% | 45.1–56.3 | 57% |
| ME-2 | Matthew Dunlap | Paul LePage | 50.7% | 45.1–56.3 | 57% |
| FL-9 | Darren Soto | Dan Green | 50.7% | 45.2–56.3 | 57% |
| CA-22 | Randy Villegas | David Valadao | 50.8% | 45.2–56.4 | 57% |
| CO-8 | Manny Rutinel | Gabe Evans | 50.9% | 45.3–56.5 | 58% |
| NJ-7 | Rebecca Bennett | Thomas Kean | 51.0% | 45.4–56.6 | 59% |
| VA-2 | Elaine Luria | Jennifer Kiggans | 51.1% | 45.5–56.7 | 60% |
| PA-1 | Bob Harvie | Brian Fitzpatrick | 51.6% | 46.0–57.2 | 64% |
| MI-10 | Christina Hines | Mike Bouchard | 52.2% | 46.6–57.8 | 69% |
| NY-17 | Cait Conley | Michael Lawler | 52.3% | 46.7–57.9 | 70% |
| AL-2 | Shomari Figures | Rhett Marques | 52.3% | 46.7–57.9 | 70% |
| NC-1 | Don Davis | Laurie Buckhout | 53.5% | 47.9–59.1 | 79% |
| AZ-1 | Amish Shah | Jay Feely | 53.6% | 48.0–59.2 | 80% |
| OH-9 | Marcy Kaptur | Derek Merrin | 54.0% | 48.4–59.6 | 82% |
| FL-14 | Kathy Castor | Mike Beltran | 54.0% | 48.4–59.6 | 82% |
| TX-28 | Henry Cuellar | Tano Tijerina | 54.1% | 48.5–59.7 | 82% |
| TX-34 | Vicente Gonzalez | Eric Flores | 54.2% | 48.6–59.8 | 83% |
| FL-25 | Jared Moskowitz | Scott Singer | 54.6% | 49.0–60.2 | 85% |
| CA-6 | Richard Pan | Kevin Kiley | 55.5% | 49.9–61.1 | 90% |
| NH-1 | Maura Sullivan | Anthony DiLorenzo | 55.8% | 50.2–61.4 | 91% |
| WA-3 | Marie Perez | John Braun | 56.4% | 50.8–62.0 | 93% |
| CA-48 | Marni von Wilpert | Jim Desmond | 56.7% | 51.1–62.3 | 94% |
| NE-2 | Denise Powell | Brinker Harding | 56.8% | 51.2–62.4 | 94% |
Shares are of the two-party vote. Where an independent runs in place of a party's nominee (Nebraska, for example), they are counted as that side's candidate.
A regression of the change in the president's party's share of the House vote on 35 elections, 1956 to 2024:
| Term | Estimate | Std. error |
|---|---|---|
| Intercept | -3.763 | 1.675 |
| midterm | -5.298 | 0.673 |
| income_1yr | -0.183 | 0.186 |
| gas_1yr | -0.060 | 0.028 |
| approval | 0.117 | 0.032 |
midterm = 1 in 2026. income_1yr and gas_1yr are percent changes over one year in real disposable income per capita and the average gas price. approval is presidential approval, in percent.
Today's inputs and result:
| Presidential approval | 36.9% | Average of the published approval aggregators, as of 2026-09-30 |
| Gas price | +16.0% | Average of the last 52 weeks ($3.61) against the 52 weeks before; EIA weekly price |
| Real disposable income | +1.05% | Average of the last 12 months released, through 2026-08, against the 12 months before; per capita |
| Predicted change for Republicans | -5.90 | points, from 51.3% in 2024 |
| National Democratic share | 54.58% |
The Democratic share of the two-party vote in each race, fit separately for the House and the Senate on contested races from 2012 to 2024:
| Term | House | Senate |
|---|---|---|
| Intercept | 8.941 (1.823) | 17.597 (8.991) |
| PVI | 0.887 (0.008) | 0.811 (0.045) |
| nat_d | 0.835 (0.036) | 0.663 (0.180) |
| inc_d | 3.534 (0.126) | 4.032 (0.529) |
| Race-level error (points) | 4.04 | 5.93 |
| Races used | 2717 | 240 |
Standard errors in parentheses. PVI is the seat's partisan lean in points: 0.75 × its 2024 presidential lean plus 0.25 × its 2020 lean, on the 2026 lines (2024 alone in states that redrew after 2024). nat_d is the national Democratic share. inc_d is +1 for a Democratic incumbent, −1 for a Republican incumbent, 0 for an open seat.
Each race has two sources of error: how far the national vote could miss (1.99 points nationally, 1.66 in a House race and 1.32 in a Senate race), and race-level error. Combined, the error in a single race is 4.36 points in the House and 6.07 in the Senate. A race's win probability is the chance a normal distribution with that spread, centered on the prediction, lands above 50%.
Seat totals come from 10,000 simulated elections. Each simulation draws one national error shared by every race in both chambers, plus a separate error for each race. The random seed is fixed, so the same inputs give the same numbers.
17 House races are already settled (no Republican on the ballot, or two candidates from the same party) and are fixed at 0 or 1.
Each update uses the latest national inputs: approval (the site's average of the published approval aggregators), the change in the gas price over the last 52 weeks (EIA weekly, via FRED) and the change in real disposable income per capita over the last 12 months released (via FRED). The same one-year rule is used for every date. Each date also uses the district maps in force on that date, so the forecast shifts when a state's new map takes effect. The coefficients stay the same. The model stops updating on November 1, 2026.