The former vice president led President Trump by seven points in both an NBC News-Wall Street Journal poll and an NPR-PBS-Marist poll; 10 points in a Washington Post-ABC News poll; 11 points in a Monmouth poll; and by a stunning 14 points in a CNN poll released early Monday morning. He also led Trump by 12 points in all-important Michigan, according to a Detroit Free Press poll. And an automated poll even showed him neck-and-neck in long-red Texas — again.
Just about all of which have drawn a predictable response: What about Hillary Clinton?
As Biden’s favorite status takes hold, Clinton’s claim to that label in 2016 stalks both him and the Democrats who might otherwise be more optimistic about what lies ahead in 2020. Yes, Biden is up in the polls, but that was true of Clinton, too. And look what happened there.
Trump himself sought to make this point in a response to the CNN poll Monday morning, saying CNN’s polls had shown the "same numbers, and worse, against Crooked Hillary.”
CNN Polls are as Fake as their Reporting. Same numbers, and worse, against Crooked Hillary. The Dems would destroy America!— Donald J. Trump (@realDonaldTrump) June 8, 2020
Except Trump’s numbers against Biden are, in fact, worse than they were against Clinton. And there are a few key ways in which the situations are different. Let’s run through them.
The size of the lead
CNN’s Harry Enten hit on this point on Monday morning. He noted that, while Clinton led for the vast majority of the 2016 campaign, she rarely flirted with the 50 percent that Biden is getting in many of these recent polls.
The RealClearPolitics average of polls during the 2016 general election, in fact, generally had her between 43 and 48 percent. Her biggest lead in the average was around seven points and was often significantly smaller — even at times falling behind Trump.
Biden’s average lead, by contrast, is consistent and often larger than Clinton’s ever was. He is right at 50 percent (49.9 percent, to be exact), and his average lead is eight points.
As Enten notes, here’s what that means, practically speaking:
When Trump closed the gap in the waning days of the 2016 campaign, he had to convince very few Clinton supporters to vote for him. Trump merely had to pick up support from those who were undecided or backing a third-party candidate.
The Michigan poll
Michigan was one of the states in which the limited polling we had in 2016 missed the mark. So when Biden leads by 12 points in that Detroit Free Press-EPIC-MRA poll this weekend, skepticism understandably seeps in.
But the points above apply in Michigan as well as anywhere else. While Clinton led in Michigan, her lead was generally far smaller than 12 points. The final RCP polling average had her lead at 3.6 points; she wound up losing it by less than a point.
Among the polls that seemingly missed the mark was the same EPIC-MRA poll, which on the eve of Election Day had her up 42 to 38 points. But notice how low those numbers are. Fully 13 percent of voters were still undecided despite it being the end of the campaign. Some polls had lower numbers of undecideds, but the very few high-quality, live-caller polls we had generally showed large numbers of them. And those undecided voters broke for Trump, according to exit polls.
Even if that doesn’t account for the entire miss, we’re talking about a miss of a couple points, not anywhere close to double-digits.
Another reason Democrats are concerned, despite Biden leading in the polls, is that Trump’s political standing is largely unchanged. While the new CNN poll showed him with an approval rating of 38 percent, that’s the first high-quality poll in eight months to show him dipping below 40 percent. And that 40 percent number is essentially where he was on Election Day 2016, when 4 in 10 voters had a favorable impression of Trump. If he won with that image then, why not now?
Two points. One is that Biden is more popular than Clinton was in 2016. Her unfavorable rating in the RCP average never dropped below 52 percent once the 2016 general election got rolling, and on Election Day she was just as disliked by voters as Trump was, according to exit polls. Biden, by contrast, has about equal numbers of voters who like and dislike him, with an unfavorable rating average of 45.7 percent.
It’s certainly plausible that Biden will be dragged down to Trump’s level as the campaign progresses, but Clinton basically began as someone voters disliked.
And now for point No. 2: Even if Biden’s image becomes sullied, polls suggest it might not matter as much. One of the most decisive reasons Trump won was that, among the large cross-section of Americans (about 1 in 5) who disliked both him and Clinton (what have been dubbed “double-haters”), he won by a large margin.
As Philip Bump has written, exit polls show they favored Trump by 21 points in Michigan, 25 points in Pennsylvania and 37 points in Wisconsin, easily accounting for his tiny margins in the three most important states.
That’s now reversed. An April Quinnipiac University poll showed voters who dislike both Trump and Biden favored Biden by 32 points. An NBC-WSJ poll the same month showed the margin at 50 points — 60 percent to 10 percent.
Trump in 2016 took 46 percent of the popular vote, despite only 4 in 10 voters actually liking him. Current polling, though, suggests he’s having a difficult time expanding his support among voters who don’t like him, which is why his share of the polls and his approval rating are often in lockstep — struggling to climb out of the low 40s.
That could certainly change if Biden becomes more unpopular, but these double-haters voted for change in 2016, and that appears to be what they’re inclined to do again in 2020, this time to Trump’s detriment.
Given the coronavirus outbreak and the racial unrest in this country right now, getting accurate polling is a challenge. And that’s perhaps especially the case in Michigan, which has also dealt with flooding on top of everything else. Even if you ignore what happened in 2016 in that state, that’s a reason for skepticism.
Some polls are indeed showing unusual differences in their partisan compositions, as the Economist’s G. Elliott Morris notes, which suggests Biden’s lead might not actually be growing:
Live-caller phone polls not adjusting for the partisan composition of their data could be for 2020 what not weighting by education was for 2016— G. Elliott Morris (@gelliottmorris) June 8, 2020
That’s a key point. But even if the partisan makeup of these polls is off, one particular finding is extremely notable: independents.
In that Michigan poll, they favor Biden by an astounding 40 points — 63 to 23. That’s a number that is, frankly, very difficult to believe. But other polls also show Biden winning independents by large, consistent margins. The CNN poll has him ahead by 11 points (52 to 41), Monmouth has him up by nine (50 to 41), and The Post-ABC poll has him up by 17 (56 to 39). By way of comparison, the final Post-ABC poll in 2016 had Trump leading this group — by five points — and Trump went on to win them by four, 46 to 42.
If Biden can hold any advantage among these voters in the 2020 campaign, it becomes virtually impossible for Trump to win — especially given that more Democrats tend to turn out to vote than Republicans. And these numbers reinforce that, however accurate the current composition of these polls, they suggest Biden is indeed ahead.
Plenty can change in the months ahead, of course, but right now Biden’s lead isn’t exactly comparable to Clinton’s in 2016; it’s superior to it.