Welcome to FiveThirtyEight’s weekly politics chat. The transcript below has been lightly edited.
sarah (Sarah Frostenson, politics editor): President Trump’s approval rating is back in the low 40s. He’s gotten poor marks for how he’s handled protests around George Floyd’s death and for how he’s managed the coronavirus pandemic. A number of national polls even show the race for president widening, with former Vice President Joe Biden comfortably in the lead. So are Americans changing their opinion of Trump? What evidence do we have that that’s the case?
We’ll talk about Trump’s relationship with independent voters and whether he might even be losing some support among Republicans, but first, let’s start with overall public opinion on Trump. Is it changing? Trump has had a 41 percent approval rating before — in fact, his approval rating has dropped even lower — so how do we contextualize what’s happening now?
lee.drutman (Lee Drutman, senior fellow at New America and FiveThirtyEight contributor): Certainly, this is not a great moment for Trump. Polls show that three-quarters of Americans believe that there is systemic racial injustice in the U.S., and Trump is clearly on the wrong side of public opinion at a moment in which these issues are super salient. That has to hurt his standing.
geoffrey.skelley (Geoffrey Skelley, elections analyst): It’s worth noting, too, that Trump’s recent approval slide is particularly noticeable as his numbers probably had been artificially high — the highest they’ve been since the start of his presidency — thanks to a “rally-around-the-flag effect” due to the coronavirus crisis. However, public opinion of his response has since soured. Combine that with the largely negative response to his handling of the protests, and you can better understand why his approval rating has been hit so hard.
perry (Perry Bacon Jr., senior writer): At the same time, it does look like Biden is having a surge — with his average lead in the polls now around 8 points, instead of 5 to 6 points. I do think that one CNN poll is getting too much attention and is probably shaping perceptions too much. I don’t think Biden is going to win by 14 points, and I’m not really sure anyone else thinks that either. It is true, though, that Biden has been leading in the polls most of the year, and Trump has been pretty unpopular for a long time.
geoffrey.skelley: I agree that it’s unlikely Biden is going to win by 15 points, given just how polarized our politics are. The last time a president won by a double-digit margin was all the way back in 1984, too, so it’s been a long time since that happened.
lee.drutman: It’s important to understand a lot of Biden’s surge as a “NOT TRUMP” vote, though.
geoffrey.skelley: Referendum elections going to referendum.
sarah: Tell us more, Lee.
lee.drutman: In that latest CNN poll that everybody’s buzzing about, 60 percent of Biden supporters say their vote is more of a vote against Trump than a vote for Biden.
This is negative partisanship, plain and simple. Trump has been wildly unpopular for a while, and while Biden has never generated a ton of enthusiasm, the worse Trump does, the more Biden rises.
sarah: Some of this might boil down to an obsession — is obsession the right word? — with treating 41 percent or 40 percent as a magical dividing line for Trump. That is, anything below that mark means he’s severely underwater; whereas, if he’s in the upper range of his approval rating bound, so between 42 to 45 percent, he’s fine.
But that simplification arguably misses two things: One, 42 to 45 percent isn’t exactly great as you all have said. And two, the fact that he’s experiencing a downturn now has to hobble his reelection chances, right?
What do we know about incumbent presidents’ reelection chances when they’ve had such low approval ratings?
geoffrey.skelley: Gallup’s final polls for Jimmy Carter in 1980 and George H.W. Bush in 1992 put their approval ratings below 40 percent. They both lost. Generally, presidents who win reelection are at or above 50 percent in those polls.
lee.drutman: Two things are different about American politics in this moment, though. One, because of Trump’s coalition (more rural, more rust-belt working class, and therefore over-represented in the Electoral College), he can still plausibly win with just 46 percent of the popular vote. Second, because of the power of partisanship, Trump doesn’t have to be popular. He just has to be less unpopular than his opponent in a handful of key swing states. And his campaign is definitely going to do everything between now and November to make Biden seem like a dangerous extremist radical who wants to unleash chaos and rioting and socialism on America.
sarah: Right, and Biden’s favorables as they currently sit aren’t that stellar despite his polling lead, right?
lee.drutman: Yeah, Biden’s underwater.
geoffrey.skelley: But only a little bit. Right before the 2016 election, Clinton had a -13 net favorability rating (favorable rating minus unfavorable rating). Right now, Biden is at about -1, so Trump’s campaign still has a long way to go in lowering that.. That said, Trump has a better net favorability rating than he did back in 2016, too — he’s at -13 now compared to -21 in 2016, thanks to some Republicans coming into Trump’s fold.
The 2016 election sometimes felt like a race to the bottom because you had the two most unpopular nominees in modern times running. And while the 2020 contest might feel like this sometimes, too, the candidates — at least right now — aren’t as unpopular.
perry: So I think something that’s actually more important than the recent polling shift is people like former Secretary of Defense James Mattis, former President George W. Bush and Sen. Lisa Murkowksi, in addition to other conservatives, making Trump-skeptical moves in the past week or so.
If we end up with a kind of permission structure on the right that signals basically, “Trump is too extreme and not worth saving, vote for Biden or stay home and you are still a good Republican,” that could be big.
I know — conservative elites broke with Trump in 2016 and Republican voters did not. But Trump barely won — and more conservative elites doing this in 2020 could signal, in particular, that a certain type of Republican (college-educated, more cosmopolitan) cannot vote Republican this year.
sarah: Perry raises an interesting point, and one I don’t know the answer to. On the one hand, rebukes from figures like Mattis and Bush seem big, but at the same time I’m not sure how much they erode Trump’s core support.
But again, as Perry says, that’s not who this message is aimed at, and as we’ve long argued as a site, Trump’s base isn’t enough.
lee.drutman: Perry’s certainly right about the permission structure. And I think there may also be some Republican voters who think that Trump losing represents a chance to clean house in the Republican Party, and that Biden is a moderate they can live with.
I’m not sure how widespread this sentiment is outside of political elites, but it could be enough to sink Trump in an electorate in which even just 1 or 2 percent of voters make the difference.
sarah: But how big of a bloc of voters do we think Bush and Mattis are speaking to? My thought is this is either a very small group of voters or they’ve already broken from Trump, so I’m not sure how much of a factor it plays in 2020. I’m thinking, too, of all the Republicans who spoke out in 2016 and the small effect they seemed to have.
perry: The way Biden has a big electoral win (and one that I think is possible) is a bloc of Republicans in the suburbs of Atlanta, Dallas and Houston decide not to vote, at least for Trump. (Maybe they just vote for the GOP Senate candidate). Bush lives in the Dallas suburbs!
That seems entirely possible to me.
Not to mention, in 2018 these types of voters flipped to Democrats.
lee.drutman: I agree that it’s not a big bloc. But there is a bloc of maybe 10 percent of Trump voters who are unsure of who to vote for, and they tend to be more liberal on most issues than the Republican Party as a whole.
But they are also younger, so I’m not sure how much they will care about what a bunch of senior Republicans have to say. Then again, they might care more about issues of racial injustice and the future of the Republican Party beyond Trump.
perry: I am not going to be able to beat back the “But 2016” rejoinder. I get it. I just think this race might be different.
sarah: That’s fair, Perry. It’s a good point. And I’m definitely guilty of overcompensating for 2016 from time to time.
perry: But I think Lee was laying out the more likely scenario earlier — Republicans run a lot of ads attacking Biden, reminding GOP voters that Biden has taken a lot of liberal stances.
And say if Biden picks Sens. Kamala Harris or Elizabeth Warren — two of the most liberal members of the Senate — as his running mate, maybe you’re in a situation where Trump is getting 44 percent of the vote and is viable in swing states.
geoffrey.skelley: The problem for the Trump campaign is they already have very good numbers among Republicans. So they really need Biden’s numbers to crater among independents much like Clinton’s did. Biden’s favorables among independents are in the 30s, so they aren’t good, but Clinton’s numbers were toxic: There were polls going into Election Day where her favorability among those voters was in the 20s and her unfavorables were between 60 and 70 percent.
lee.drutman: Also, right now the public is on the side of the protesters. But there’s a (very likely) scenario in which criminal justice reform dies in the Senate, cities walk back the progressive promises they’re making now, and there’s tremendous frustration that nothing will change. Then, protests get more aggressive, and progressive demands to “defund the police” become more polarizing.
An additional complication is that if many ballots wind up being cast remotely because of the coronavirus pandemic, there could be lots of first-time, vote-at-home voters. That means enthusiasm and mobilization could be a big issue. And I’m not sure how this plays. It could go either way, helping or hurting both Democrats and Republicans.
perry: Yeah, I could picture a situation where Biden doesn’t motivate young black people — or young people more broadly — if he sticks to his more centrist impulses, i.e. “most police officers are good.” Another potential problem for Biden could arise if Harris or Warren is on the ticket and his campaign overreaches with its promises for racial justice (those two want to be leaders on racial justice), landing him in Clinton’s “deplorable” zone. That all could hurt Biden with white swing voters.
lee.drutman: Perry is right. And Republican operatives almost certainly have a plan to broadcast those moves to the groups most likely to to not vote.
geoffrey.skelley: The Trump campaign even tried to dampen turnout among black voters in 2016, for example.
sarah: These are good points, but I thought Michael Tessler over at the Washington Post’s Monkey Cage had some good points on why the protests won’t help Trump politically, which I think is especially true, if as you say, Lee, some of those Republicans who are most unsure are more likely to be more liberal on issues of racial injustice.
lee.drutman: Right — I think there is a sense among protesters and supporters that maybe this time really is different and their voices are being heard. But turning protest into political action isn’t so straightforward. And if anything comes out of Washington (unlikely) it will be very minor.
sarah: That’s fair. It’s also, as FiveThirtyEight contributor Shom Mazumder wrote for the site, really hard to know how these protests might play out electorally, as protests can have unintended consequences.
But OK, we’ve talked a little about who is driving this downturn in Trump’s numbers, but one question I have is how do independents fit into this? Is Trump seriously underwater there?
lee.drutman: As always when we’re talking about independents, we have to be careful. Most independents are not swing voters, and swing voters aren’t always independents.
geoffrey.skelley: Lee’s right that “independent” is a complicated label for many of the voters who self-identify that way. For that reason, I wish more pollsters included independents who lean toward a party in their party ID breakdowns. But what we do know is that right now Biden has an edge on Trump among independents in most recent national polls — though the leads vary quite a bit. That CNN poll was very favorable to Biden, for instance, and it had him up 11 points, but a Monmouth University poll also had him up 16 among independents. Some other polls give Biden a smaller edge with this group — he was up 4 points in a survey from NPR/PBS NewsHour/Marist College, for example — but they agree that Biden’s ahead.
lee.drutman: Interestingly, there is some evidence, too, that there are Trump-Biden voters, who tend to be economically liberal, socially conservative — like Obama-Trump voters. My thought is some of these voters may just be perpetually dissatisfied with Washington and in a persistent throw-the-bums-out mood.
perry: And Lee, to his credit, has been writing about economically liberal, socially conservative voters since Trump’s rise.
But it feels like these voters may not be exactly extra-motivated by civil rights issues, right?
sarah: That’s a good question, Perry. That was definitely part of the narrative around these voters in 2016 — they broke for Trump because of their views on immigration and on racial discrimination. It’s unclear to me if that will motivate them again in the same way in 2020, or if there are other considerations. At this point, though, how sticky do we think this downturn in Trump’s approval rating really is? As we said at the outset of this chat, his approval rating has been at 41 percent before. So do we really think a shift in public opinion is happening now? Or might the economy — think more news like last week’s job report — pose a silver lining for Trump, and his approval rating will maybe bounce back?
lee.drutman: Trump’s had a really bad past few weeks, and the polls reflect that. Though given just how bad a period he’s had, it’s sort of remarkable his approval rating hasn’t fallen further, which is a testament, I suppose, to the resilience of his 40 percent floor.
That said, I don’t think the economy will have that much of an effect, given how little he was rewarded for a good economy, and how little he’s been punished for a bad economy. There’s about 5 percent of the electorate that seems to go back and forth on Trump, mostly based on how Trump behaves, and how bad the coverage of him is. Based on regression to the mean, I’d guess Trump’s approval rating will go back up, unless there is some kind of cascade of elite Republicans turning on him.
geoffrey.skelley: The good news for Trump is that voters aren’t completely short-term when it comes to their attitudes toward economic conditions. After all, Trump continues to poll decently when it comes to his handling of the economy — I suspect he’s still getting some credit for the strong past performance. So economic improvement could help him. Still, Lee’s right that the roaring economy didn’t give Trump a great overall approval rating before, so it’s unlikely to alone save him.
perry: I think Trump will end up back in the 42-to-43 percent approval range, which is bad, but not horrible. The tear gas incident alone drew so much negative coverage that I doubt he will do that particular thing again.
geoffrey.skelley: I don’t know, Perry. On Tuesday, he tweeted out conspiracy theories about a man who was injured by police in Buffalo. I wouldn’t bet against him not doing something!
lee.drutman: Yeah, I agree with Geoffrey. I think Trump’s instinct is to create controversy, chaos and division, because he thinks that makes him look strong and leader-ish, and that’s the lesson that he learned from 2016.
He might not break up a crowd to do another photo op, but he’ll continue to amplify unlikely and provocative fringe views just to set the pinballs in motion.
sarah: So to conclude, do we think there’s evidence that public opinion of Trump is really changing this time Or is that too hard to answer at this point?
geoffrey.skelley: I think if Trump’s approval slides a little below 40 percent, we could point to this being an especially bad moment for him. He’s only really been there once in our tracker since the midterm, and that was during the government shutdown, when his approval fell to 39 percent. So I think the jury is still out on this being a moment where opinions truly shift.
One other way to think about this is how much opinions against Trump have hardened.
For instance, we know that a lot of voters have said in the past they didn’t want to vote for Trump, and most people who disapprove of him do so strongly. But do some of those who said they “somewhat disapprove” now say they “strongly disapprove” of Trump? Or do they move back more to the middle, to even “somewhat approve?” That will help determine whether there’s much leeway for him to recover electorally-speaking.
lee.drutman: Bottom line: It seems clear Trump’s had a bad few weeks, and public opinion reflects that. If you think he’ll continue to have weeks as bad as the last few for the foreseeable future, or he’ll get more criticism from the right, then yes, public opinion is really changing. But my guess is that the news cycle will shift, controversies will change, and Trump will recover.
perry: I think it’s too early to say. Let’s consider: 1) COVID-19; 2) The use of tear gas; 3) The high jobless rate because of COVID-19; and 4) The protests. They are all electorally bad for Trump, but I doubt bad things like that all happen at the same time in October, right before the election. But Trump was an underdog before most of this happened and remains so.
lee.drutman: But to Geoffrey’s point, I think it’s possible that if there’s a deeper slide (say, below 35 percent approval), there might be some genuine panic and potential circular firing squad action among Republicans.
Lots of scenarios are possible, but at this point I’d put my money on Republicans closing ranks around Trump as the election nears and his approval ticking up, but Biden still winning by a narrow margin. Basically, in five months we’ll be more or less where we’ve been for the last three months.
geoffrey.skelley: The more things change, the more they actually stay the same.