What Approval Ratings Measure — and What They Do Not
A presidential approval rating is the answer to one simple question asked in a poll: "Do you approve or disapprove of the way [the president] is handling his job as president?" The percentage who say "approve" is the approval rating. The percentage who say "disapprove" is the disapproval rating. Some polls also allow "no opinion" or "haven't heard enough to say."
Approval ratings do not measure personal likability, ideology alignment, or policy preferences. A voter can approve of a president's job performance while disagreeing with specific decisions. A voter can disapprove while liking the president personally. The question is purely evaluative: is the overall performance satisfactory?
Approval ratings are also a lagging indicator. They respond to events — economic shocks, foreign policy crises, scandals — but typically with a delay of weeks or months. A single bad news cycle rarely moves approval more than a point or two; sustained negative conditions over months produce meaningful shifts.
Major Pollsters and Methodology Differences
Dozens of organizations poll presidential approval, but methodology differences produce meaningfully different numbers. Three major approaches dominate:
Gallup has tracked presidential approval since Harry Truman in 1945, making it the gold standard for historical comparisons. Gallup uses random-digit-dialing of landlines and cells, interviews roughly 1,500 adults, and publishes results weekly or monthly. Gallup polls all adults, not just registered or likely voters, which tends to produce slightly lower approval numbers than registered-voter polls (because non-voters often have less favorable views of political figures).
AP-NORC conducts less frequent but methodologically rigorous quarterly polls using a probability-based online panel. Its polling is considered among the most accurate due to its sampling approach, though the less frequent cadence limits real-time tracking.
Morning Consult, YouGov and Ipsos use online opt-in panels with large sample sizes (often 10,000+), which allows more frequent tracking but introduces potential selection bias. These pollsters often show slightly different numbers from traditional phone polls for the same time period.
The most reliable single-source comparison uses one consistent pollster over time. Comparing a Gallup number from 1998 to a Morning Consult number from 2026 is methodologically problematic — the differences may reflect survey mode rather than actual approval changes.
| Pollster | Method | Sample | Frequency | Population |
|---|---|---|---|---|
| Gallup | Phone (RDD) | ~1,500 | Weekly/monthly | Adults |
| AP-NORC | Probability panel (online) | ~1,000–1,500 | Quarterly | Adults |
| Reuters/Ipsos | Online opt-in panel | ~1,000+ | Weekly | Adults |
| Morning Consult | Online opt-in panel | ~10,000+ | Daily | Registered voters |
| YouGov/Economist | Online opt-in (matched) | ~1,500 | Weekly | Adults / reg. voters |
| Quinnipiac | Phone (RDD) | ~1,500 | Monthly | Registered voters |
Historical Approval Averages: Every President Since Truman
| President | Term(s) | Avg Approval | High | Low | Key Context |
|---|---|---|---|---|---|
| Harry Truman | 1945–1953 | 45% | 87% | 22% | Post-WWII high; Korea cratered numbers |
| Dwight Eisenhower | 1953–1961 | 65% | 79% | 48% | Sustained popularity; Cold War stability |
| John F. Kennedy | 1961–1963 | 70% | 83% | 56% | Assassinated; high throughout presidency |
| Lyndon B. Johnson | 1963–1969 | 55% | 79% | 35% | Civil Rights high; Vietnam cratered |
| Richard Nixon | 1969–1974 | 49% | 67% | 24% | Watergate drove historic collapse |
| Gerald Ford | 1974–1977 | 47% | 71% | 37% | Nixon pardon tanked initial numbers |
| Jimmy Carter | 1977–1981 | 45% | 75% | 28% | Iran hostage crisis, stagflation |
| Ronald Reagan | 1981–1989 | 53% | 68% | 35% | Recession low in 1983; recovered strongly |
| George H.W. Bush | 1989–1993 | 61% | 89% | 29% | Gulf War spike; recession collapse |
| Bill Clinton | 1993–2001 | 55% | 73% | 37% | Impeachment paradox: approval rose during trial |
| George W. Bush | 2001–2009 | 49% | 90% | 25% | 9/11 spike; Iraq and Katrina collapse |
| Barack Obama | 2009–2017 | 48% | 67% | 38% | Early high; partisan sorting tightened range |
| Donald Trump (1st term) | 2017–2021 | 41% | 49% | 34% | Never topped 50%; tightest range ever |
| Joe Biden | 2021–2025 | 42% | 57% | 36% | Fell sharply after Afghanistan and inflation |
| Donald Trump (2nd term) | 2025–present | ~44% | ~49% | ~41% | Early 2026 range; tariff policy impacted |
Source: Gallup historical polling. Trump 2nd term figures are averages from major pollster composites as of early 2026.
The Partisan Sorting Problem: Why Modern Approvals Have a Tighter Range
A striking feature of modern presidential approval is how narrow the range has become. Eisenhower sustained 65% average approval. Kennedy was at 70%. Even Reagan averaged 53%. Trump's first term averaged 41%, Biden's averaged 42%, and neither president came close to 55% for more than a brief window.
The explanation is partisan sorting. In the 1950s and 1960s, party affiliation was a weaker predictor of approval than it is today. Conservative Democrats in the South and liberal Republicans in the Northeast crossed party lines frequently. Eisenhower won 57% of the popular vote in 1956, with enormous cross-party support.
Today, voters are much more consistently sorted by party. Republicans almost universally disapprove of Democratic presidents; Democrats almost universally disapprove of Republican presidents. Since each party represents roughly 30-35% of the electorate, the floor for any president is around 30-35% (their own base) and the ceiling is around 55-60% (their base plus most independents). Genuine cross-party approval is rare.
This means a modern president at 44% approval is not doing dramatically worse than one at 48% — both are losing independents and a sliver of their own party. The politically meaningful threshold is whether a president can sustain above 50%, which is now extremely difficult for any president after their first year.
How Approval Predicts Midterm Outcomes
The relationship between presidential approval and midterm House results is one of the most robust patterns in American electoral history. The mechanism is straightforward: voters who disapprove of the president's performance are more motivated to vote against his party's candidates, while supporters of the opposition need less encouragement than usual to turn out.
The rule of thumb is simple. At above 50% approval, the president's party typically loses 5 to 15 House seats — modest losses, sometimes small enough to maintain a majority. At 45-50% approval, losses of 15 to 30 seats are typical. At below 45% approval, the losses tend to be severe: 30 to 60+ seats. The 2010 and 1994 wave elections both occurred when the sitting president was below 45%.
The relationship is not perfect. 2022 was an outlier: Biden's approval was around 40-42%, models predicted 40-60 Democratic losses, but the actual result was only 9 seats — a historically small loss for an unpopular president. Explanations include the Dobbs abortion decision (June 2022), candidate quality in Republican primaries, and strong Democratic fundraising. No model is perfect.
For 2026, Trump's approval in the low-to-mid 40s as of early 2026 puts him in the range where historical patterns strongly favor Democratic gains. Whether those gains reach 5+ seats (the majority threshold) is uncertain — but the directional signal is clear.
| President Approval (Oct of midterm year) | Average House Seat Loss | Historical Examples |
|---|---|---|
| 55%+ | ~5 seats | 1998 Clinton (66%), 2002 Bush (63%) |
| 50–54% | ~10–18 seats | 1990 Bush (52%), 1986 Reagan (53%) |
| 45–49% | ~20–35 seats | 2006 Bush (38%), 2014 Obama (43%) |
| Below 45% | ~35–65 seats | 2010 Obama (45%), 1994 Clinton (47%) |
Note: 2022 was an outlier (Biden ~41%, only -9 seats). Historical averages exclude 1998 and 2002, which were unusual national security/impeachment cycles.
Trump 2025-2026: Where His Approval Stands and What It Implies
Donald Trump began his second term with approval ratings slightly higher than his first-term average, hovering around 47-49% in inauguration-week polling. The initial bump faded within months as tariff announcements, DOGE spending cuts and various executive actions polarized public opinion further.
By early 2026, Trump's approval in major pollster composites averaged roughly 44-45%, with disapproval at approximately 52-53%. This is historically consistent with his first term range — he remains deeply popular with Republicans (85%+ approval) but broadly disapproved among Democrats (5-10%) and split among independents (roughly 40-45% approval).
At that approval level, historical models suggest the Republican Party faces meaningful House losses in November 2026. The Democratic majority threshold of roughly 5 net gains is well within the range that correlates with a 44-45% presidential approval — though as 2022 demonstrated, the translation from approval to seats is not mechanical.
The approval number to watch between now and November 2026 is Trump's rating among independents. When independents break strongly against an incumbent president, the wave effect is amplified significantly. When they are split, losses are more modest even with low overall approval.
Frequently Asked Questions
What is a good presidential approval rating?
In the modern partisan era, above 50% is considered strong and above 55% is exceptional. Historically the average across all presidents is about 53%, but modern polarization has compressed the range. No president since George W. Bush (briefly after 9/11) has sustained above 60% for any extended period. A president at 48% is doing reasonably well by contemporary standards; at 42% they are in politically dangerous territory for their party.
Why do approval ratings rarely change much?
Partisan sorting means most people's view of the president is largely determined by party affiliation, which is very stable. The persuadable middle — true independents and weak partisans — is a smaller share of the electorate than it used to be. Dramatic events (9/11, major recessions, domestic crises) can move approval by 10-20 points, but most news cycles move it by fractions of a point if at all.
Is there a difference between approval of the president and approval of the job?
Yes, though they tend to be highly correlated. Personal favorability (do you have a favorable or unfavorable opinion of X?) is distinct from job approval (do you approve of the job X is doing?). A president can have higher personal favorability than job approval if people like them personally but think the economy or foreign policy is going wrong. Bill Clinton had high personal favorability even during his lowest job approval periods during the Lewinsky scandal.
How is a presidential approval average calculated for this site?
We use a simple average of available polls from major pollsters (Gallup, Reuters/Ipsos, AP-NORC, YouGov/Economist, Quinnipiac) taken within the last 30 days, weighted equally. We do not apply house effect adjustments. For the most sophisticated model, FiveThirtyEight's approval tracker (now at ABC News) applies pollster quality weights and recency weighting. Our figure is intended as a readable overview, not a model.
