- Trump's approval sits at ~39% (April 2026) — well below the 45% threshold that historically signals major midterm losses
- When a president's approval is below 50%, his party loses an average of 37 House seats; below 45% losses accelerate sharply (Obama 2010: −63, Clinton 1994: −54)
- No single poll is reliable — always use aggregated averages from multiple pollsters to cancel out house effects
- We track 6 independent aggregators (Ballotpedia, CNN, DDHQ, FiftyPlusOne, RCP, Race to the WH) and average them daily
What Is a Presidential Approval Rating?
A presidential approval rating is the percentage of Americans who say they approve of the job the president is doing. It is measured through a simple survey question — typically something like: "Do you approve or disapprove of the way [president's name] is handling his job as president?" Respondents answer approve, disapprove, or no opinion.
The approval rating is a binary measure: it collapses complex feelings about a president into a single yes-or-no answer. This simplicity is a feature, not a bug — it makes approval ratings highly consistent across time, pollsters and methods, allowing meaningful historical comparisons going back to Harry Truman in the 1940s.
The disapproval rating is the mirror image. When added together, approval + disapproval + no opinion = 100%. Historically, as a president's term continues, the "no opinion" share shrinks as people form firmer views.
How Approval Ratings Are Measured
Different pollsters use different methodologies, which can produce approval ratings that differ by several percentage points even when measuring the same underlying sentiment. The main methods are:
Live telephone interviews (CATI) — A human interviewer calls random phone numbers and reads questions to respondents. Historically considered the gold standard, but declining response rates (now below 5%) raise questions about representativeness.
Interactive Voice Response (IVR) — Automated "robopoll" calls where respondents press numbers on their keypad. Cheaper and faster than live interviews; can only reach landline numbers in many states.
Online panels — Respondents recruited to take surveys online. Now the most common method. Quality depends heavily on how the panel was recruited and how responses are weighted to match the population. Pollsters like YouGov, Ipsos and Morning Consult use this method.
Each method has systematic biases — called "house effects" — that cause a given pollster to consistently show slightly higher or lower approval than others. This is why aggregated averages are more reliable than any single poll.
Historical Patterns
Most presidents begin their term with a "honeymoon" effect — elevated approval ratings in the 55-65% range as the public gives a new president the benefit of the doubt. This typically erodes as the president makes decisions that disappoint some supporters and energizes opponents.
George W. Bush started his first term at around 57% and fell to 51% by September 2001 — then surged to 90% after the September 11 attacks in the largest single approval spike ever recorded. By 2008 he had fallen to 25%, among the lowest ever recorded.
Barack Obama started at 67% in January 2009 and gradually declined, finishing his first term around 48-50%. His second term ended around 58% — relatively high for a two-term president.
Donald Trump is historically unusual. He began his first term at 45% — the lowest inauguration-day approval in the polling era — and never exceeded 49% approval during his entire first term (2017-2021). He ended his first term at 34%. His second term began at approximately 47% in January 2025 and has declined to approximately 39-40% as of April 2026.
Joe Biden started at approximately 54% in January 2021, declined steadily, and ended his presidency at approximately 38% — one of the lowest final approval ratings on record.
The 50% Threshold and the "Six Percent Rule"
Political scientists have documented a strong relationship between presidential approval ratings at the time of midterm elections and the number of House seats the president's party wins or loses. The key threshold is 50% approval.
Presidents with approval ratings above 50% at midterm time rarely lose significant ground. The clearest exception is George W. Bush in 2002: with approval ratings still elevated above 60% due to post-9/11 sentiment, Republicans gained 8 House seats — the only time in modern history a president's party gained seats in a first-term midterm.
Presidents with approval ratings below 50% consistently lose seats. When approval falls into the low 40s or below, losses accelerate dramatically:
- 2010 (Obama, ~45%) — Democrats lost 63 House seats, Republicans took the majority
- 1994 (Clinton, ~46%) — Democrats lost 54 seats, ending 40 years of Democratic control
- 1974 (Ford/Nixon, ~40%) — Republicans lost 49 seats
- 2022 (Biden, ~42%) — Democrats lost 9 seats (smaller than models predicted due to abortion issue post-Dobbs)
The historical average loss for a president's party when approval is below 50% is approximately 37 House seats. With Trump at approximately 39-40% in April 2026, models suggest Republicans face significant exposure — though generic ballot polling, fundraising, candidate quality and the specific districts in play all factor into the final result.
How We Track Approval Ratings
USPollingData.com compiles Trump's approval rating from six major aggregators, drawing on Wikipedia's compilation methodology. Rather than tracking individual pollsters (each of which has different house effects), we use aggregators that have already combined dozens of polls:
- Ballotpedia — Nonpartisan political encyclopedia, tracks approval via averaging
- CNN — Media aggregator combining major national pollsters
- DDHQ (Decision Desk HQ) — Data journalism outlet with its own weighting model
- FiftyPlusOne — Polling aggregator focused on electoral data
- RealClearPolitics (RCP) — Simple average of recent polls, widely cited
- Race to the White House — Tracking aggregator focused on electoral trends
We average these six aggregators daily to produce a single tracking number. This meta-aggregation approach minimizes the influence of any individual pollster's house effects and provides the most stable, representative measure of presidential approval available.
| President | Inauguration | Peak | At Midterm | Final | House Midterm Result |
|---|---|---|---|---|---|
| Truman | 82% | 87% | 40% (1950) | 32% | Dems −29 |
| Eisenhower | 69% | 79% | 62% (1954) | 59% | R −18 |
| Johnson | 79% | 79% | 44% (1966) | 49% | Dems −47 |
| Nixon | 59% | 67% | 49% (1970) | 24% | R −12 |
| Carter | 66% | 75% | 37% (1978) | 34% | Dems −15 |
| Reagan | 51% | 68% | 42% (1982) | 63% | R −26 |
| Clinton | 58% | 73% | 46% (1994) | 66% | Dems −54 |
| Bush W. | 57% | 90% (9/11 spike) | 38% (2006) | 25% | R −30 |
| Obama | 67% | 67% | 45% (2010) | 58% | Dems −63 |
| Trump (1st) | 45% | 49% | 42% (2018) | 34% | R −41 |
| Biden | 54% | 57% | 42% (2022) | 38% | Dems −9 (smaller than models predicted) |
| Trump (2nd) | 47% | 47% | ~39% (Apr 2026) | — | TBD — Nov 3, 2026 |
Frequently Asked Questions
What is Trump's current approval rating?
As of April 2026, Trump's approval rating is approximately 39-40% in aggregated averages. Our site updates this daily by averaging data from six major aggregators: Ballotpedia, CNN, DDHQ, FiftyPlusOne, RealClearPolitics and Race to the White House. See the Polls page for the current figure.
How does approval rating predict midterm results?
When a president's approval is below 50% at midterm time, their party loses an average of 37 House seats. The lower the approval, the greater the expected losses. Trump's current approval of approximately 39-40% puts Republicans in a historically dangerous position heading into November 2026, though final results also depend on the generic ballot, economic conditions and candidate quality.
Which approval rating is most reliable?
Aggregated averages combining multiple pollsters are the most reliable measure. Individual polls have "house effects" — systematic biases from specific methodologies. Aggregators like RealClearPolitics and FiveThirtyEight combine many polls to smooth out these effects. We use six aggregators and average them for maximum stability.
