Midterm Elections and Presidential Party Losses in US history follow a strong pattern: the president’s party lost House seats in 18 of the 20 midterms from 1946 to 2022, averaging 25.6 seats, and lost Senate seats in 13 of 20, averaging 3.6. The 2026 midterm took place on November 3, 2026, with President Trump’s approval near 37% in polling averages, and results were still being counted on November 4, though ABC News projected that Democrats had won enough seats to take control of the House.
Midterm Elections and Presidential Party Losses in US 1946–2026 – Introduction
Midterm Elections and Presidential Party Losses in US history is one of the most reliable patterns in American politics. Since 1946 the party holding the White House has lost House seats in 18 of 20 midterm elections, and it has gained Senate seats only six times. The exceptions are rare and have clear causes, such as the national rally after the September 11 attacks in 2002.
This report uses the American Presidency Project’s midterm table at UC Santa Barbara, last updated July 13, 2026, which draws on Gallup approval data and on House and Senate historian records. It covers House losses, Senate changes, presidential approval, exceptions and the 2026 election. It is current as of November 4, 2026. The 2026 results are incomplete, so the 2026 row shows the pre-election setup and early projections, not final seat counts.
Interesting Facts About Midterm Elections and Presidential Party Losses 1946–2026
HOUSE SEAT CHANGE FOR THE PRESIDENT'S PARTY, AVERAGES 1946-2022
All 20 midterms ████████████████████████████░░░░░░░░░░░░ -25.6
Approval below 50% (11) ██████████████████████████████████████░░ -34.7
Approval 50% or higher (9) ████████████████░░░░░░░░░░░░░░░░░░░░░░░░ -14.6
Last four (2010-2022) ██████████████████████████████████░░░░░░ -31.5
| Metric | Value (1946–2022 Unless Noted) |
|---|---|
| Midterms studied | 20 |
| Midterms where the president’s party lost House seats | 18 |
| Average House change | –25.6 seats (median –22) |
| Largest House loss | –63 (2010, Obama) |
| Largest House gain | +8 (2002, George W. Bush) |
| Midterms where the president’s party lost Senate seats | 13 of 20 (gained in 6, no change in 1) |
| Average Senate change | –3.6 seats |
| Average House loss, approval below 50% / 50% or higher | –34.7 / –14.6 |
| Midterms where the president’s party gained seats in both chambers | 1 (2002) |
| House majority changed hands (1946–2022) | 7 times |
| 2026 House / Senate seats the president’s party was defending | 220 / 20 |
Data source: The American Presidency Project, UC Santa Barbara, Seats in Congress Gained/Lost by the President’s Party in Mid-Term Elections (updated July 13, 2026), based on Gallup approval data, the House Historian and the Senate Historian; report calculations.
The table covers all 20 midterms from 1946 to 2022. The president’s party lost House seats 18 times, with an average loss of 25.6 seats and a median of 22. The Senate shows a smaller pattern: 13 losses, 6 gains and one unchanged result, for an average of –3.6. Only 2002 produced gains in both chambers.
Approval is the strongest single indicator. When approval was below 50% in the final Gallup reading, the average House loss was 34.7 seats. When it was 50% or higher, the average was 14.6. The last four midterms (2010 to 2022) averaged 31.5 House seats lost, a little above the 28.4 average for 1946 to 1994. Midterm Elections and Presidential Party Losses in US history is therefore a story about both timing and public mood.
House Seats Lost by the President’s Party 2026 Historical Table: 1946–2022
HOUSE SEAT CHANGE, PRESIDENT'S PARTY (LOSSES SHOWN AS BARS)
2010 Obama ████████████████████████████████████████ -63
1946 Truman ███████████████████████████████████░░░░░ -56
1994 Clinton ██████████████████████████████████░░░░░░ -54
1958 Eisenhower██████████████████████████████░░░░░░░░░░ -48
1974 Ford ██████████████████████████████░░░░░░░░░░ -48
1966 Johnson █████████████████████████████░░░░░░░░░░░ -47
2018 Trump ██████████████████████████░░░░░░░░░░░░░░ -41
2006 G.W. Bush ███████████████████░░░░░░░░░░░░░░░░░░░░░ -30
1950 Truman ██████████████████░░░░░░░░░░░░░░░░░░░░░░ -28
1982 Reagan ████████████████░░░░░░░░░░░░░░░░░░░░░░░░ -26
1954 Eisenhower███████████░░░░░░░░░░░░░░░░░░░░░░░░░░░░░ -18
1978 Carter █████████░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░ -15
2014 Obama ████████░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░ -13
1970 Nixon ███████░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░ -12
2022 Biden █████░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░ -9
1990 G.H.W. Bush█████░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░ -8
1986 Reagan ███░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░ -5
1962 Kennedy ██░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░ -4
1998 Clinton ░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░ +4
2002 G.W. Bush ░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░ +8
| Year | President (Party) | House Seats Defended | House Change | Share of Defended Seats Lost |
|---|---|---|---|---|
| 1946 | Truman (D) | 244 | –56 | 23.0% |
| 1950 | Truman (D) | 263 | –28 | 10.6% |
| 1954 | Eisenhower (R) | 221 | –18 | 8.1% |
| 1958 | Eisenhower (R) | 203 | –48 | 23.6% |
| 1962 | Kennedy (D) | 264 | –4 | 1.5% |
| 1966 | Johnson (D) | 295 | –47 | 15.9% |
| 1970 | Nixon (R) | 192 | –12 | 6.3% |
| 1974 | Ford (R) | 192 | –48 | 25.0% |
| 1978 | Carter (D) | 292 | –15 | 5.1% |
| 1982 | Reagan (R) | 192 | –26 | 13.5% |
| 1986 | Reagan (R) | 181 | –5 | 2.8% |
| 1990 | G. H. W. Bush (R) | 175 | –8 | 4.6% |
| 1994 | Clinton (D) | 258 | –54 | 20.9% |
| 1998 | Clinton (D) | 207 | +4 | Gain |
| 2002 | G. W. Bush (R) | 220 | +8 | Gain |
| 2006 | G. W. Bush (R) | 233 | –30 | 12.9% |
| 2010 | Obama (D) | 257 | –63 | 24.5% |
| 2014 | Obama (D) | 201 | –13 | 6.5% |
| 2018 | Trump (R) | 241 | –41 | 17.0% |
| 2022 | Biden (D) | 222 | –9 | 4.1% |
Data source: The American Presidency Project, UC Santa Barbara (updated July 13, 2026); House Historian party-division records. The share of defended seats lost is a report calculation (seats lost divided by seats the president’s party held before the election). Some outlets report 2018 as –40 or –42; the American Presidency Project lists –41.
The five worst House losses were 63 seats in 2010, 56 in 1946, 54 in 1994, 48 in both 1958 and 1974, and 47 in 1966. Those years all combined low or falling approval with a large exposed majority. Obama’s party lost 24.5% of the seats it defended in 2010, and Ford’s lost 25.0% in 1974 after the Nixon resignation and pardon.
Smaller losses cluster in two situations. Presidents with strong approval, such as Kennedy in 1962 (61%), lost 4 seats and Reagan in 1986 (64%) lost 5. Presidents whose party held fewer seats to lose, such as Biden in 2022 with 4.1% of his seats lost, also fared better. The House changed hands in seven midterms: 1946, 1954, 1994, 2006, 2010, 2018 and 2022. These numbers are the core of Midterm Elections and Presidential Party Losses in US analysis.
Senate Seat Changes for the President’s Party 2026 Historical Table: 1946–2022
SENATE SEAT CHANGE, PRESIDENT'S PARTY
1958 Eisenhower ████████████████████████████████████████ -13
1946 Truman ████████████████████████████████████░░░░ -12
2014 Obama ███████████████████████████░░░░░░░░░░░░░ -9
1986 Reagan ████████████████████████░░░░░░░░░░░░░░░░ -8
1994 Clinton ████████████████████████░░░░░░░░░░░░░░░░ -8
1950 Truman ██████████████████░░░░░░░░░░░░░░░░░░░░░░ -6
2006 G.W. Bush ██████████████████░░░░░░░░░░░░░░░░░░░░░░ -6
2010 Obama ██████████████████░░░░░░░░░░░░░░░░░░░░░░ -6
1974 Ford ███████████████░░░░░░░░░░░░░░░░░░░░░░░░░ -5
1966 Johnson ████████████░░░░░░░░░░░░░░░░░░░░░░░░░░░░ -4
1978 Carter █████████░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░ -3
1954 Eisenhower ███░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░ -1
1990 G.H.W. Bush ███░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░ -1
| Year | President (Party) | Senate Seats Defended | Senate Change |
|---|---|---|---|
| 1946 | Truman (D) | 21 | –12 |
| 1950 | Truman (D) | 21 | –6 |
| 1954 | Eisenhower (R) | 11 | –1 |
| 1958 | Eisenhower (R) | 20 | –13 |
| 1962 | Kennedy (D) | 18 | +3 |
| 1966 | Johnson (D) | 21 | –4 |
| 1970 | Nixon (R) | 7 | +2 |
| 1974 | Ford (R) | 15 | –5 |
| 1978 | Carter (D) | 14 | –3 |
| 1982 | Reagan (R) | 12 | +1 |
| 1986 | Reagan (R) | 22 | –8 |
| 1990 | G. H. W. Bush (R) | 17 | –1 |
| 1994 | Clinton (D) | 17 | –8 |
| 1998 | Clinton (D) | 18 | 0 |
| 2002 | G. W. Bush (R) | 20 | +2 |
| 2006 | G. W. Bush (R) | 15 | –6 |
| 2010 | Obama (D) | 15 | –6 |
| 2014 | Obama (D) | 20 | –9 |
| 2018 | Trump (R) | 9 | +2 |
| 2022 | Biden (D) | 14 | +1 |
Data source: The American Presidency Project, UC Santa Barbara (updated July 13, 2026); Senate Historian party-division records. “Defended” counts members of the relevant electoral class who were up for election.
Senate results depend on which class of seats is up. Eisenhower’s party defended 20 seats in 1958 and lost 13, and Truman’s defended 21 in 1946 and lost 12. Nixon’s party defended only 7 in 1970 and gained 2, and Trump’s defended only 9 in 2018 and gained 2 even as it lost 41 House seats. The average Senate loss was 3.6.
Six presidents gained Senate seats in a midterm: Kennedy (1962), Nixon (1970), Reagan (1982), George W. Bush (2002), Trump (2018) and Biden (2022). Five of those six still lost House seats. In 2026 Republicans defend 20 Senate seats in the regular classes, plus two special elections in Florida and Ohio, so the map is less favorable than 2018. Senate swings show why Midterm Elections and Presidential Party Losses in US results often split by chamber.
Presidential Approval and Midterm Outcomes 2026: Gallup Ratings and Seat Losses
AVERAGE HOUSE LOSS BY FINAL GALLUP APPROVAL BAND, 1946-2022
Below 40% (2 midterms) ████████████████████████████████████████ -43.0
40-44% (6) █████████████████████████░░░░░░░░░░░░░░░ -27.3
45-49% (3) ████████████████████████████████████████ -44.0
50-59% (4) ███████████████████████████░░░░░░░░░░░░░ -29.0
60% or higher (5) ███░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░ -3.0
| Final Gallup Approval | Midterms | Average House Change | Years |
|---|---|---|---|
| Below 40% | 2 | –43.0 | 1946 (27%), 2006 (37%) |
| 40–44% | 6 | –27.3 | 1950, 1966, 1982, 2014, 2018, 2022 |
| 45–49% | 3 | –44.0 | 1978 (45%), 1994 (48%), 2010 (45%) |
| 50–59% | 4 | –29.0 | 1958, 1970, 1974, 1990 |
| 60% or higher | 5 | –3.0 | 1954, 1962, 1986, 1998, 2002 |
| Below 50%, all | 11 | –34.7 | Average Senate change –4.6 |
| 50% or higher, all | 9 | –14.6 | Average Senate change –2.3 |
| Correlation, approval vs House change | 20 | 0.57 | Report calculation |
Data source: The American Presidency Project, UC Santa Barbara (Gallup, late October or latest available reading before each midterm); report calculations. The 1950 reading is 41% and the 1978 reading is 45%.
The 1946–2022 record shows a clear break at 60%. Every president at 60% or higher lost an average of only 3.0 House seats. Those same five midterms include gains in 1998 and 2002. Presidents in the 40–44% band lost 27.3 on average, and those below 40% lost 43.0, though the sample is only two (Truman at 27% in 1946 and George W. Bush at 37% in 2006).
The relationship is strong but not exact. The correlation between final approval and House change is 0.57 across the 20 midterms (report calculation). The 45–49% band lost 44.0 on average because it includes 1994 and 2010, two wave years driven by factors beyond approval. Joe Biden’s 40% in 2022 produced a loss of only 9, the sharpest gap between approval and outcome. Approval predicts direction better than size in Midterm Elections and Presidential Party Losses in US history.
Historical Exceptions and Political Trends 2026: When the President’s Party Did Not Lose
EXCEPTIONS AND SPLIT RESULTS (HOUSE / SENATE CHANGE)
1998 Clinton (65% approval) House +4 Senate 0
2002 G.W. Bush (67%) House +8 Senate +2
1962 Kennedy (61%) House -4 Senate +3
2018 Trump (44%) House -41 Senate +2
2022 Biden (40%) House -9 Senate +1
| Midterm | Approval | House | Senate | Context |
|---|---|---|---|---|
| 1998 | 65% | +4 | 0 | Strong approval during the impeachment fight; party defended 207 House seats |
| 2002 | 67% | +8 | +2 | Rally after September 11, 2001; only midterm with gains in both chambers |
| 1962 | 61% | –4 | +3 | Approval above 60%; smallest House loss in a first-term midterm in the table |
| 2018 | 44% | –41 | +2 | Democrats flipped the House; Republicans defended only 9 Senate seats |
| 2022 | 40% | –9 | +1 | Smallest loss for a president under 45%; Dobbs ruling and candidate quality cited by analysts |
| Lame-duck midterms (6) | Mixed | Average –20.0 | Average –7.0 | 1958, 1986, 2006 and 2014 included heavy Senate losses |
| All other midterms (14) | Mixed | Average –28.1 | Average –2.1 | Includes 2002 gain and 1962 loss of 4 |
Data source: The American Presidency Project, UC Santa Barbara (updated July 13, 2026); explanations as summarized by Dave Manuel and The Rational Moderate; Gallup. Group averages are report calculations. Lame-duck midterms follow the American Presidency Project’s flags: 1950, 1958, 1986, 1998, 2006 and 2014.
Only 1998 and 2002 produced House gains for the president’s party. Both came with approval above 65%, and each had a clear driver: the impeachment backlash in 1998 and the national security rally after September 11 in 2002. In 1962, 2018 and 2022 the president’s party still lost House seats but gained in the Senate. Five midterms had that split pattern: 1962, 1970, 1982, 2018 and 2022.
The trend since 2010 is large losses in wave years and smaller ones in others. The last four midterms averaged 31.5 House seats lost, but 2022 saw only 9. Analysts attribute the recent smaller losses in part to a smaller universe of competitive districts after redistricting and partisan sorting. NPR noted this in October 2026. Lame-duck midterms averaged 20.0 House seats lost against 28.1 for the other 14, a gap of 8.1 seats, though they lost more Senate seats (7.0 against 2.1) (report calculations).
2026 Midterm Election: Forecasts, Approval and Early Results
2026 PRE-ELECTION INDICATORS
Trump approval, polling average (Oct) ██████████████████████████████░░░░░░░░░░ 37%
Trump approval, Gallup (late Aug) ███████████████████████████░░░░░░░░░░░░░ 33%
Generic ballot, Democratic lead (RCP) ███████░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░ 8.9 pts
Cook projected Democratic House seats ████████████████████████████████████████ 230-240
House seats needed for majority ████████████████████████████████████████ 218
| Indicator (2026) | Value |
|---|---|
| Election Day | Tuesday, November 3, 2026 |
| Seats on the ballot | 435 House, 35 Senate, 36 governorships |
| House before election | 218 Republicans, 214 Democrats, 1 independent, 2 vacant |
| Senate before election | 53 Republicans, 47 Democrats |
| Democratic net gain needed | 3 House seats; 4 Senate seats |
| Trump approval, polling average | 37% (tied for lowest pre-midterm in half a century, per NPR) |
| Trump approval, Gallup | 35% (early August), 33% (late August) |
| Generic ballot (RealClearPolitics) | Democrats +8.9 |
| Cook Political Report final House projection | Democrats 230–240 seats |
| Sabato’s Crystal Ball House projection | Democrats 239, Republicans 196 |
| Decision Desk HQ chance of Democratic House win | More than 80% |
| Result status, November 4, 2026 | ABC News projects Democrats win House control; counting continues |
Data source: American Presidency Project (Gallup approval, 2026 seats to defend); NPR (October 6, 2026); Newsweek, ABC Australia, AOL and Forbes (final forecasts, November 2–3, 2026); ABC News projection (November 3–4, 2026); Associated Press results page for the House (as of 4:02 a.m. UTC, November 4, 2026), which showed counting still underway.
Going into November 3, every major forecaster expected Republicans to lose the House. Cook projected a Democratic gain of 12 to 22 seats, which would bring Democrats to 230–240, and Sabato’s Crystal Ball predicted 239 Democratic seats to 196 Republican. The generic ballot favored Democrats by 8.9 points. Trump’s approval of 37% matched George W. Bush’s 37% in 2006, when Republicans lost 30 House seats. Gallup’s 33% reading in late August is lower than any final pre-midterm reading in the table except Truman’s 27% in 1946.
Results were still being counted early on November 4. ABC News projected that Democrats would win enough seats to take control of the House, and the Associated Press count was still changing. The Senate outcome was not final: Democrats needed a net gain of 4 seats, while 2026’s Republican-held Senate class had 20 seats up, plus two special elections. Early counts have historically been more Republican than final results, so seat totals can move for days. Final totals will go in the 2026 row of Midterm Elections and Presidential Party Losses in US tables once certified.
Midterm Elections and Presidential Party Losses in US 2026 FAQ
How Many Seats Does the President’s Party Usually Lose in the House in 2026 Midterms?
The average since 1946 is 25.6 seats, with a median of 22. Presidents under 50% approval average 34.7.
How Many Midterms Has the President’s Party Lost House Seats Before 2026?
18 of the 20 midterms from 1946 to 2022. The only gains were in 1998 (+4) and 2002 (+8).
How Many Senate Seats Does the President’s Party Usually Lose in 2026 Midterms?
An average of 3.6 since 1946. The party lost Senate seats in 13 of 20 midterms and gained in 6.
What Was the Biggest Midterm Loss in US History Before 2026?
63 House seats in 2010 under Obama. Truman lost 56 in 1946 and Clinton 54 in 1994.
Which Presidents Gained House Seats in a Midterm Before 2026?
Bill Clinton in 1998 (+4) and George W. Bush in 2002 (+8). Both had approval above 65%.
Does Presidential Approval Predict Midterm Losses in 2026?
Yes, loosely. Presidents at 60% or higher lost an average of 3.0 House seats, and presidents below 50% lost 34.7. The correlation across 20 midterms is 0.57.
What Was the Approval Rating Before the 2026 Midterm?
About 37% in polling averages in October 2026, with Gallup at 33% in late August.
Did Democrats Win the House in the 2026 Midterms?
ABC News projected on election night that Democrats had won enough seats for House control. Final seat totals were still being counted on November 4, 2026.
How Many Seats Do Democrats Need to Win the House in 2026?
A net gain of 3 seats from the pre-election 218–214 Republican edge, since 218 seats make a majority.
Why Do Presidents’ Parties Lose Midterm Seats in 2026 and Before?
Common explanations include voter reaction to the party in power, high exposure of the president’s party after a presidential win and falling approval. The record shows the pattern but not the cause.
How Often Does the House Change Hands in a Midterm Through 2026?
Seven times from 1946 to 2022: 1946, 1954, 1994, 2006, 2010, 2018 and 2022.
Disclaimer: The data research report we present here is based on information found from various sources. We are not liable for any financial loss, errors, or damages of any kind that may result from the use of the information herein. We acknowledge that though we try to report accurately, we cannot verify the absolute facts of everything that has been represented.
