To read a presidential approval rating, look at three numbers together: the share who approve, the share who disapprove, and the difference between them (net approval). Then check who was surveyed and when, and compare against an average of many polls rather than a single release. The trend over several weeks tells you far more than any one day’s number.
The live figure changes every week, so this guide covers how to read it. For the current average, open the Approval tab in Election Tracker, which is described below.
What does a presidential approval rating measure? #
Pollsters ask a version of one question: “Do you approve or disapprove of the way [the president] is handling his job as president?” Gallup has asked it since the 1930s, which is why approval is the longest-running political measure in American polling.
Answers fall into approve, disapprove, and no opinion. Some pollsters also offer “strongly” and “somewhat,” which tells you how hard each side’s opinion is. The question is about job performance, not personal likability; that’s what favorability ratings measure.
How do you read approve, disapprove and net approval? #
Take two hypothetical readings:
| Approve | Disapprove | No opinion | Net | |
|---|---|---|---|---|
| President A | 43% | 47% | 10% | −4 |
| President B | 43% | 54% | 3% | −11 |
Both have 43% approval, but they’re in very different positions. President A has a pool of undecided people who could still come around. President B’s disapproval is higher and firmer, with almost no one left undecided. That is why many analysts watch net approval and the disapprove number at least as closely as the approve number.
A rule of thumb: when disapproval passes 50% and stays there, opinion has usually hardened, and moving it takes a major event.
Why do approval polls disagree with each other? #
On any given week, reputable polls can differ by 5 points or more. The usual reasons:
- Population. Polls of all adults, registered voters and likely voters produce different numbers. Approval polls most often survey adults, which is appropriate for a job-performance question but less predictive of elections.
- Mode. Live phone, online panels and text-to-web samples reach different people.
- Question format. Offering a “neither” option or pushing undecided people to choose changes the split.
- House effects. Some firms consistently run a little higher or lower than the consensus. That’s a methods choice, not necessarily bias. See pollster house effects.
- Sampling error. A 1,000-person poll has about a ±3 point margin on each number, so week-to-week wiggles inside that range usually mean nothing.
Where can you track presidential approval ratings? #
Two kinds of sources are worth knowing.
Individual pollsters that ask the question on a regular schedule, such as Gallup, YouGov/The Economist, Reuters/Ipsos, Quinnipiac, AP-NORC and Marquette Law School. Their trend within their own series is often more telling than their level, because method stays constant.
Averages that combine many pollsters, such as Silver Bulletin, The New York Times, RealClearPolling and Decision Desk HQ. FiveThirtyEight’s average, once the standard reference, ended when ABC News shut the site down in March 2025. How presidential approval ratings are calculated walks through how these averages differ.
On a phone, Election Tracker puts the approval average on one screen: approve and disapprove side by side, the net number, and a label saying whether it’s rising, falling or stable. Beneath that is a 90-day chart of both series (drawn as a 7-day average), then the individual recent polls with sample size, population and field date. The data comes from VoteHub, with the New York Times average as a fallback source, and the screen shows when it was last refreshed. It’s free, needs no account, and runs on iPhone.
How to read an approval trend line #
- Ignore single-day moves under 2 points. That’s sampling noise, especially in a daily tracking poll.
- Look for moves that hold for two or three weeks. Those usually reflect something real: a major policy fight, an economic shift, a foreign crisis.
- Watch both lines. If approval is flat but disapproval is rising, the undecided group is breaking against the president.
- Check for a “rally” spike. Sudden national events can lift approval for a few weeks. They usually fade. George W. Bush’s jump after September 11, 2001 is the extreme example; it was well above anything seen since.
- Compare like with like. Don’t compare an adults-only poll this month to a likely-voter poll last month.
What do approval ratings mean for the 2026 midterms? #
Approval is one of the best single indicators of how the president’s party does in a midterm. The pattern since World War II is consistent: the president’s party almost always loses House seats, and it tends to lose more when the president is unpopular.
| Midterm | President | Gallup approval (approx.) | House seats for president’s party |
|---|---|---|---|
| 1994 | Clinton | 42% | −54 |
| 1998 | Clinton | 64% | +4 |
| 2002 | G.W. Bush | 67% | +8 |
| 2006 | G.W. Bush | 41% | −30 |
| 2010 | Obama | 45% | −63 |
| 2018 | Trump | 42% | −41 |
| 2022 | Biden | 42% | −9 |
Source: The American Presidency Project, UC Santa Barbara.
The two midterms where the president’s party gained House seats since 1934 (1998 and 2002) both came with approval in the 60s. But 2022 shows the relationship isn’t mechanical: Biden’s approval was in the low 40s and Democrats lost only 9 House seats and gained a Senate seat. Candidate quality, redistricting, and issues that cut against the usual pattern all matter. Our explainer on why the president’s party loses seats in midterms covers the theories.
For November 3, 2026, the practical move is to read approval next to the generic congressional ballot, which asks people which party they’d back for Congress. When the two tell the same story, it’s usually the right one. When they diverge, something candidate-specific is going on.
Common mistakes when reading approval ratings #
- Treating one poll as the news. A single poll at a record low is often an outlier. Check the average first.
- Reading approval as a vote share. A president at 42% approval isn’t going to get 42% of the vote. Some people who disapprove still vote for the president’s party, and vice versa.
- Ignoring “no opinion.” Early in a term that group is large; by the second year it usually shrinks.
- Comparing across decades without context. Partisan polarization has compressed approval ranges. Modern presidents rarely see the highs or lows of the 1950s through 1990s.
Frequently asked questions #
What is considered a good presidential approval rating? #
Above 50% is strong by modern standards, and few recent presidents have held it for long. The mid-40s is typical in today’s polarized environment. Readings below 40% have usually signaled serious trouble for the president’s party in the next election.
How often are presidential approval ratings updated? #
Daily or weekly tracking polls (from firms such as YouGov and Morning Consult) publish constantly, while Gallup and several university polls release monthly or less often. Averages update whenever a new poll is added. Election Tracker re-checks its approval source hourly and whenever you pull to refresh.
What is net approval? #
Net approval is the approve percentage minus the disapprove percentage. A president at 44% approve and 52% disapprove has a net approval of −8. It’s useful because it captures both sides in one number and makes trends easier to compare across polls with different “no opinion” shares.
Why do approval polls of adults and of voters differ? #
Adults include people who aren’t registered or rarely vote, and those groups can lean differently from habitual voters. The gap is usually a few points and can run in either direction depending on the president and the moment. For forecasting elections, voter samples are more relevant; for measuring national mood, adult samples are standard.
Does a low approval rating mean the president’s party will lose the midterms? #
Usually, but not always. Low approval has been followed by big House losses in most modern midterms, yet 2022 showed that other factors can soften the blow. Treat approval as a strong indicator, not a guarantee.