Trump Approval Update
And what recent events tell us about his future popularity.
We finally have enough new surveys to have a good idea of the effect of two major events – the bombing of Iran on June 22, and the passage of the megabill on on July 3.
And the verdict is…nothing, pretty much. According to the G. Elliott Morris polling average, Trump was at 44% approval on June 22; continued to fall to 43%, and then sat right there up to today. The Cook Political Report finds even less change. Nate Silver? He has more of a continuous slump and just a bit more overall loss. But that appears to be that the other two registered some slumping before June 22 that Silver’s methods don’t have, perhaps in response to previous major news story – confrontations over deportation in Los Angeles during the two weeks before the Iran strike.1
The big picture? After all that, depending on which flavor one prefers, Trump is currently at or near his low point in approval. Sticking to Morris’s estimate, the president is about 10 percentage points underwater, which means he remains the second-worst of any polling-era president at this point, almost six months in. Only Trump himself in 2017 was doing worse, which has been the case throughout the year. Just one other polling-era, Bill Clinton, had net-negative approval numbers at this point; only Clinton, Gerald Ford, and Joe Biden reached negative territory at any point during their first year. Only Trump 2017 and Trump 2025 got as low as -10 in that first year.
As far as the effects of recent events, it’s a lot harder to know what’s going on.
The numbers are consistent, for example, with the theory that Trump basically has a normal default approval level in the low 40s, and after reverting to that following a brief honeymoon, very little can dislodge him. It’s quite possible, indeed, that all the apparent fluctuations in the last few months are really just meaningless noise.
What we can probably rule out is that the combined effects of Los Angeles, Iran, and success on Capitol Hill produced a major surge for him…and, with somewhat less confidence, that they produced a significant fall-off in support.2 But beyond that? It’s possible that one or two of the events really helped him, one or two hurt him, and the overall effects canceled each other out. If you squint really hard, it seems more likely that either Los Angeles or Iran hurt him a bit, but we’re just not seeing enough movement, or enough quality polls, to have much confidence in that interpretation.
In other words, when it comes to interpreting these kinds of surveys, squinting hard is precisely the wrong way to go about it. There are just too many events, too close together (including in the month we’re looking at the Texas floods and several major judicial opinions, arrests of various public officials, and more) to be able to separate them out.
Even if none of these events created much — or any — immediate movement, that doesn’t prove that Trump is stuck in that low 40s approval zone no matter what. There are, again, a lot of unknowns here. The increase in partisanship and the decentralized place of the news media may make short-term events a lot less likely to produce short-term approval movement than they did in the days of weaker partisanship and a concentrated news media. The old cues that “Something big has happened! Pay attention!” may no longer work, or work a lot less than they used to.
But then again it’s also possible that these particular events just didn’t pack much of a punch, either because people didn’t care much or because all three of these events sent mixed signals. For example: Passage of the megabill made Trump look like a winner…but most of the provisions of the bill appear to be unpopular. Similarly, military action without US casualties is often popular, but involvement in Iran polls badly.
Nor do a lack of short-term changes — if that’s what happened — necessarily mean there can’t be longer-run effects. The “win” from the megabill will soon be forgotten; cuts to programs, however, could turn out to hurt Trump and the Republicans more as they are implemented even if they didn’t matter much now.
And it’s worth mentioning that positive effects from legislation usually fades rapidly and doesn’t matter much in future electoral cycles; negative effects, however, can stick around. That’s not a prediction, but it does seem to be a likely pattern.
What we can say with some confidence is that Trump is unpopular, and that this appears to be either personal — people just don’t like him — or the effects of his various scandals and unpopular policy positions. It is true that this time around, unlike in 2017, people are reporting unhappiness with the economy, but it’s just as likely that’s an effect of Trump’s unpopularity than it is from inflation, which remains tame, or unemployment, which remains low.
So the one thing that I do think people get very wrong is to say that Trump is immune from scandal. It’s true, of course, that his strongest supporters will overlook practically anything (and yeah, that probably includes the Epstein stuff). But most people aren’t Trump’s strongest supporters.
I still don’t think that Trump is immune from other damage, but until events that normally produce those kinds of changes happen (such as a normal recession), we won’t know. Meanwhile, we’re still where we’ve been. He’s unpopular. Enough so that Democrats have good reason for optimism in 2026, but not enough so that we would expect the president’s party to run away from him or to produce a major landslide.
Why the differences? They use different methods of estimating averages, and they include different polls (Silver takes almost everything, while Cook is just using a small group of theoretically high-quality surveys). I like all of them, although my habit now is to use Morris for most things.
A quick stats point: We don’t know what Trump’s real approval might be; we only can know what we observe in various surveys, all of which are subject to a bunch of different ways of being off from underlying reality. Since what we’re seeing here — the observed effect — is about a two percentage point drop in approval, it’s somewhat more likely that the underlying real effect is negative than positive, just as if we flipped a coin ten times and got seven “heads” we could conclude that it’s more likely that we have a weighted coin that’s apt to come up heads than one that is weighted towards tails. But the coin could easily be a fair one, just as it’s not at all clear that Trump’s approval has changed at all.