Chart library
Long-history charts that set the present against the record: how states, parties, voters and forecasts have moved, election by election. Each one redraws itself as new results arrive. Link to it, embed its image, or take its numbers.
Also: Explore 2026 in charts
- The States That Swapped Sides
- How Close Washington Came
- The Races With No Poll on Record
- The Vanishing Crossover Senator
- When the Forecast Disagreed With the Polls
- The Diploma Divide
- When Third Parties Could Have Decided
- The Ballot Nobody Finishes
- Is This a Wave Year?
- The 2016–2024 Wind Map
History · President
The States That Swapped Sides
Arkansas and Vermont swapped sides of the country
Each line is one state's presidential margin compared with the nation's, election by election since 1976. Seventeen states crossed the national line and moved at least fifteen points; the three that moved furthest each way are highlighted.
Swipe the chart sideways to see all of it.
How to read it. Left of the centre line: the state voted more Republican than the country that year. Right: more Democratic. A highlighted line's colour is the side the state was on in the latest election shown. Washington, D.C., which voted sixty-four to ninety-one points more Democratic than the nation, is off the scale and left out.
Source. MIT Election Data and Science Lab, U.S. President 1976–2024 (two-party vote; a nominee's votes on every ballot line). The chart adds 2028 when that year's results are in the file. Drawn November 2, 2026, 2:14 AM Pacific.
Link to this chartChart image pageDownload the data (CSV)Free to republish with credit to FutureBallot.
History · President, Senate and House, 2004–2024
How Close Washington Came
The thinnest line to power since 2004: the presidency in 2020, decided in WI by 0.6 points
For each election shown, the seat that decided who held power: the state that cast the 270th electoral vote for the winner, the race that gave the Senate majority its deciding seat, and the 218th House seat for the majority. Each dot is the margin by which that deciding seat was won. The lower the dot, the closer Washington came to going the other way.
Swipe the chart sideways to see all of it.
How to read it. Colour is the side that won control. Margins are points of all votes cast in that state or district, the winner over the runner-up; a dot above fifteen points is drawn at the top, with its true margin written beside it. Seats are ranked from the winner's safest down: the deciding seat is the one that took the winner past the line, so in a landslide it is a safe seat and the dot stands high. The Senate lane starts in 2008: the seats not up each year are counted from the results of the six years before, and FutureBallot holds too few Senate results before 2004 to count them; 2012 and 2014 are left out, because a senator not up that year is recorded without a party or caucus (Alaska's 2010 write-in winner). The vice president's tie-breaking vote counts for the vice president's party.
Source. Electoral votes from each state's House seats plus two (D.C. three), with Maine's and Nebraska's district electors from the National Archives' records; presidential results from the MIT Election Data and Science Lab; Senate and House results as recorded by FutureBallot in each race's deciding round (Georgia's and Louisiana's runoffs included). Unopposed House races, which record no votes, count as the safest seats. Drawn November 2, 2026, 2:14 AM Pacific.
Link to this chartChart image pageDownload the data (CSV)Free to republish with credit to FutureBallot.
History · polls, 2004–2026
The Races With No Poll on Record
In 2024, 81% of contested House races had no poll in FutureBallot's archive; three of thirty-three Senate races had none
Every contested Senate, governor and House race of each election year as a book on a shelf: the more polls FutureBallot holds of a race, the thicker and taller its spine. Races with no poll on record stand as faint outlines. Senate races are shelved first, then governors, then the House.
Swipe the chart sideways to see all of it.
How to read it. Spine width grows with the square root of the number of polls, so one heavily polled race does not fill a shelf. A contested race is one where a second candidate took at least 5% of the vote (for this year, one with two or more candidates on the ballot). The right-hand column is the share of contested House races with no poll on record.
Source. FutureBallot poll archive, general-election questions. The archive is the polls FutureBallot has collected, not every poll ever taken: campaign polls that were never released, and some published House polls in earlier years, are not in it. A ghost means no poll on record, not proof that nobody polled. Drawn November 2, 2026, 2:14 AM Pacific.
Link to this chartChart image pageDownload the data (CSV)Free to republish with credit to FutureBallot.
History · Senate, 2004–2024
The Vanishing Crossover Senator
In 2004–2008, twenty-four of 102 Senate races were won by the party that lost the state for president; in 2020–2024, six of 104 Senate races were won by that party
Each row is one year's Senate elections, each block one race, ordered from the state that voted most Republican for president on the left to the most Democratic on the right. Winners from the party that lost their state for president are lit and named. Watch the lit blocks thin out.
Swipe the chart sideways to see all of it.
How to read it. A state's presidential vote is its result in the latest presidential election on or before the Senate race. Colour is the winner's party; a winner recorded without a major party (purple) is not counted as a crossover, and that includes Alaska's 2010 write-in winner, whom the record holds as an independent. Rows hold the races of each even year, not the whole chamber: a senator elected in an earlier year shows in that year's row. Alabama's December 2017 special is left out.
Source. Senate results as recorded by FutureBallot, each race's deciding round (runoffs included), from 2004, the first year the record holds every race; presidential results from the MIT Election Data and Science Lab. Drawn November 2, 2026, 2:14 AM Pacific.
Link to this chartChart image pageDownload the data (CSV)Free to republish with credit to FutureBallot.
Backtest · Senate and governors, 2016–2024
When the Forecast Disagreed With the Polls
Where FutureBallot's rebuilt forecasts and the polls differed by more than three points, the rebuilt forecast came closer more often: the forecast in thirty-three races, the polls in seventeen
Every Senate and governor race from 2016 to 2024 with at least two polls in its final three weeks, placed by the polls' average margin (across) and FutureBallot's forecast (up). Where the two differed by more than three points (fifty races), the mark shows which ended closer to the result.
Swipe the chart sideways to see all of it.
How to read it. On the diagonal, the forecast and the polls agree; the grey band is within three points. These are not forecasts anyone published at the time. They are rebuilt by today's model (version 0.8.7) from only what was known on each election eve, the final snapshot of each race. A model built after these elections has been tested against them, so this is a kinder test than a forecast made in real time. Calibration across all races is on the Accuracy page.
Source. FutureBallot backtest runs of the current race model, each race's last run before its election; polls from FutureBallot's archive that finished in the last twenty-one days, each poll's margin between the forecast's two candidates averaged equally; results as recorded, November round. Margins in points; the Democrat or independent is on the right. Drawn November 2, 2026, 2:14 AM Pacific.
Link to this chartChart image pageDownload the data (CSV)Free to republish with credit to FutureBallot.
History · districts, 2020 and 2024
The Diploma Divide
The most-educated tenth of districts had an average 2024 presidential margin of D+30.5; the least-educated tenth, R+15.1
All 435 House districts as drawn for 2026, sorted from the fewest adults with a bachelor's degree (top) to the most (bottom). Each dot is the district's 2024 presidential margin, and the tail on that dot reaches back to 2020. Holding the lines fixed means every tail is a change in how the same places voted.
Swipe the chart sideways to see all of it.
How to read it. On average the most-educated tenth is 57% graduates and the least-educated 20%. The dot is 2024 and the line runs back to the 2020 result: a dot at the left end of its line moved toward Republicans, one at the right end toward Democrats. Nineteen districts changed sides. Margins beyond eighty points are drawn at the edge.
Source. Census Bureau American Community Survey 2019–2023 (adults twenty-five and over with a bachelor's degree), on each district's 2026 lines; presidential two-party margins as FutureBallot recomputes them on 2026 lines, from precinct, block-level and official district returns, some of them estimates where precincts cross district lines. 2016 is not shown: FutureBallot holds it only by state. Drawn November 2, 2026, 2:14 AM Pacific.
Link to this chartChart image pageDownload the data (CSV)Free to republish with credit to FutureBallot.
History · President, 1976–2024
When Third Parties Could Have Decided
In 1992, votes for candidates other than the two major-party nominees outnumbered the winning margin in forty-nine of the fifty states; in 2024, in two of fifty
Each column is a presidential election; each chip a state where the votes for every candidate other than the two major-party nominees added up to more than the gap between those two. In those states the others' votes exceeded the margin: enough, had they gone the other way together, to have changed the result. Below, the same count for Senate and governor races.
Swipe the chart sideways to see all of it.
How to read it. More than the margin is a measure of size, not a claim that those voters would otherwise have voted for the loser: they might have split, or stayed home. Washington, D.C., is left out. In the lower rows, a race counts only when both a Democrat and a Republican ran and one of them won; the margin there is between the two of them.
Source. MIT Election Data and Science Lab, U.S. President 1976–2024 (each nominee's votes on every ballot line); Senate and governor results as recorded by FutureBallot, November round, from 2004. Drawn November 2, 2026, 2:14 AM Pacific.
Link to this chartChart image pageDownload the data (CSV)Free to republish with credit to FutureBallot.
History · President, Senate and House, 2004–2024
The Ballot Nobody Finishes
In 2024, about 4.1 million fewer votes were cast for the House than for president, in the forty states where every House race was contested
A ballot whose lines fade down the page: for every hundred votes cast for president, how many were cast for the Senate and the House, each line covering its own set of states. On the right, every presidential year since 2004.
Swipe the chart sideways to see all of it.
How to read it. Only like is compared with like. The House line uses only states where every House race had a second candidate with at least 5% of the vote, since some states record no votes at all for an unopposed race; the Senate line, only states with a contested Senate race that year (where a state held two, the larger total). The two lines cover different states, so compare each with 100, not with each other. These are totals, not people: a ballot left blank and a ballot spoiled look the same in them.
Source. Presidential totals from the MIT Election Data and Science Lab (all votes cast for president); Senate and House totals as recorded by FutureBallot, November round. Drawn November 2, 2026, 2:14 AM Pacific.
Link to this chartChart image pageDownload the data (CSV)Free to republish with credit to FutureBallot.
History · midterms, 1946–2026
Is This a Wave Year?
The president's party trailed in the national House vote in fifteen of the twenty midterms from 1946 to 2022; the only earlier midterm with approval as low as this year's 36% was 1946, when the president's party trailed in the national House vote by 8.6 points
Every midterm since 1946, placed by the president's approval going into the autumn and how the president's party then did in the national House vote. The ringed dot is 2026: this year's approval and the national model's estimate of the House vote, which moves until the votes are counted. The worst result was 1974, when the president's party lost by 17.1 points.
Swipe the chart sideways to see all of it.
How to read it. Up is a better result for the president's party, right a more popular president. Most dots sit below the zero line: the president's party usually loses ground in a midterm. The estimate is not a result; once the votes are counted it becomes an ordinary dot.
Source. Approval: Gallup, August to September of each midterm year (for this year, the mean of published averages moved onto Gallup's scale); House vote: the national popular vote margin; the estimate: FutureBallot's national House model, as on the Environment page. Drawn November 2, 2026, 2:14 AM Pacific.
Link to this chartChart image pageDownload the data (CSV)Free to republish with credit to FutureBallot.
History · President
The 2016–2024 Wind Map
Between 2016 and 2024, twenty-four of the fifty states moved toward Republicans; the biggest shift toward Republicans was in Florida and the biggest shift toward Democrats was in Nebraska
Each state as a tile, with an arrow for how its presidential margin moved from 2016 to 2024: right toward Democrats, left toward Republicans, longer for a bigger shift. The tile's shade is the state's 2024 result.
Swipe the chart sideways to see all of it.
How to read it. Read the arrows for direction and size of the shift; read the tile colour for where the state ended up. Shifts are in two-party margin points.
Source. MIT Election Data and Science Lab, U.S. President 1976–2024, two-party vote. Drawn November 2, 2026, 2:14 AM Pacific.
Link to this chartChart image pageDownload the data (CSV)Free to republish with credit to FutureBallot.