If you’ve been following our coverage of the Democratic primary, you’ll know that I don’t think much of media really understands Joe Biden’s popularity among Democrats. That doesn’t mean that Biden is destined to win the primary. In fact, I’d regard him as an underdog relative to the field — that is, I think he has a less than 50 percent chance of getting the nomination — partly for reasons I’ll outline later on in this column. But there have already been several occasions when despite widespread predictions of Biden’s demise, the former vice president rebounded or held steady in the polls.
Furthermore — not totally unlike Donald Trump four years ago — Biden’s support comes mostly from the type of Democrats who are sometimes relatively invisible in media coverage of the campaigns, such as black Democrats and older Democrats without college degrees. That’s another reason to be skeptical about claims that Biden isn’t as popular as polls seem to imply. They sometimes reflect narratives that are filtered through journalists’ college-educated social environments — or conditioned by conversations on social media — with all the implicit biases those can introduce.
So this article about Biden in The New York Times, which alleged a disconnect between the polls and conditions on the ground In Iowa, was a little dismaying for me. Here’s a representative snippet:
But less than two weeks before Labor Day, when presidential campaigns traditionally kick into high gear, there are signs of a disconnect between his relatively rosy poll numbers and excitement for his campaign on the ground here [in Iowa], in the state that begins the presidential nominating process.
The thing is, I actually think the Times is onto something here! But it’s something you can see in the polls. And it’s something that probably has a lot to do with Iowa, where the article was datelined from — and where there are relatively few voters from among the groups that are most enthusiastic about Biden.
One straightforward way for a poll to detect voter enthusiasm is simply for it to ask voters how strongly they feel about a particular candidate. Many polls, for example, ask voters whether they have a very favorable, somewhat favorable, somewhat unfavorable or very unfavorable view of the candidates. If the Times’s hypothesis is correct about Biden, we’d expect to see a lot of lukewarm support for him — plenty of voters in the “somewhat favorable” category, but relatively few in the “very favorable” bucket.
That’s not really what national polls show, though. I averaged data from the four polls I can find that were conducted after the Detroit debate — from Morning Consult, YouGov, HarrisX and Fox News1 — that asked Democratic voters the four-pronged favorability question I described above. By this measure, Biden had just as much enthusiastic support as any other Democrat. In the polling average, 36 percent of Democrats had a very or strongly favorable view of Biden, essentially the same as for Bernie Sanders (37 percent) and Elizabeth Warren (36 percent) and well ahead of the next Democrat, Kamala Harris, at 26 percent.2
There’s plenty of enthusiasm for Biden in national polls …
Average favorability of Democratic primary candidates in national polls
Now, because more voters list Biden as their first choice than list Sanders or Warren, he probably does have his fair share of begrudging supporters. But it’s a fallacy to say that Biden doesn’t have enthusiastic supporters also. According to national polls, he has plenty — about as many as Sanders or Warren. On top of that, he also gets support from other Democrats who are less enthusiastic but support him for other reasons, e.g. electability. Those two paths to support are why Biden’s polling at almost 30 percent while Sanders and Warren are closer to 15 percent.
… And among attentive voters
Average of three Quinnipiac national polls conducted in July and August
|Candidate||All Dems.||Attentive Dems.*|
Furthermore, Biden does plenty well with the most attentive voters, which isn’t quite the same thing as enthusiasm but is certainly related to it. (And in any event, it’s a bullish indicator since the most attentive voters are more likely to eventually vote.) Quinnipiac has routinely been asking Democrats how much attention they’re paying to the campaign and breaking out results on that basis. In the table to the left, you can see how that data looks in an average of their last three national polls, dating back to July.
Biden gets 30 percent of the vote among Democrats who say they’re paying “a lot” of attention to the campaign, essentially the same as the 29 percent support he has from Democrats overall in Quinnipiac’s polling. Warren and Kamala Harris do gain some ground among the most attentive voters — while Sanders loses some ground. But still, Biden tops the field among voters in that category.
If you look at polls of Iowa, though, you see more evidence of an actual enthusiasm deficit for Biden. Unfortunately, there are no especially recent polls of Iowa that asked the four-pronged favorability question of Democratic voters. But there are some from June and July that did, namely a Selzer & Co. poll for CNN and the Des Moines Register, a Change Research poll and a David Binder Research poll.3
… but there’s less enthusiasm for Biden in Iowa
Average favorability of Democratic primary candidates in Iowa polls
Biden’s numbers are OK in Iowa, but not great. Some 31 percent of Democrats say they have a very or strongly favorable view of him, which is essentially tied with Sanders for fourth place in the field. By comparison, 46 percent of Iowa Democrats have a strongly favorable view of Warren. (And remember, these polls are a couple of months old — she’s gained ground in most national and early-state polls since then.) Meanwhile, 40 percent also have a strongly favorable view of Harris and Pete Buttigieg.
Whenever we see a divergence between Iowa polls and national polls, the question is whether it has something to do with Iowa’s demographics or, rather, the fact that the campaign is more active in Iowa than it is nationally. If the differences are demographic in nature, then Biden might not have as much to worry about — the Democrats who turn out to caucus in Iowa are white and liberal, whereas he overperforms among nonwhite and moderate Democrats, who make up bigger parts of the electorate in states such as South Carolina and most of the Super Tuesday states. Alternatively, if Biden’s numbers are middling in Iowa because voters don’t like him as much upon prolonged exposure to him, that could mean it’s more of a canary in the coal mine, and that voters in other states will tire of Biden once they begin paying more attention to the campaign.
These explanations aren’t necessarily mutually exclusive — but on balance, the demographic hypothesis is more persuasive in Biden’s case. In 2016, 91 percent of Iowa’s Democratic caucus electorate was white, as compared to around 60 percent nationally in the Democratic primaries. And almost 70 percent of voters there identified as liberal, as compared to about 50 percent nationally. Those are big differences, and not ones that are favorable for Biden’s coalition, so it’s reasonable to expect he might be doing quite a bit worse there than he is nationally. Furthermore, as I mentioned, national polls show Biden doing just as well with highly attentive voters as with voters overall, which should also lead us to discount the canary-in-the-coal-mine hypothesis.
Actually, the candidate who should be really worried about his Iowa numbers is Sanders. Iowa’s demographics should boost his coalition, but his favorability numbers are just as lukewarm as Biden’s there.
But getting back to Biden: Even if his issues in Iowa are mostly linked to demographics, the state could still be a major concern for him. That’s because Iowa can affect the result of every state that votes after it. These effects seem to have dampened in recent election cycles. (Just ask Ted Cruz or Mike Huckabee — or, for that matter, Barack Obama — how much winning Iowa helped them in New Hampshire.) Nonetheless, to the extent that some of Biden’s support is predicated on electability, some of it could erode if voters see him take a big loss.
Biden may already be trailing in Iowa, in fact. Among the last seven polls there — everything since the first debate in Miami — three have shown Biden leads, but two have had Warren ahead, along with one Biden/Warren tie and one Buttigieg lead. If you combine those top-line polls with the enthusiasm numbers, plus Warren’s strong ground game in Iowa, you can argue that the state is Warren’s to lose more than Biden’s.
Still, the story is a lot more complicated than a cliched narrative about voters choosing heart (Warren/Sanders) over head (Biden), or voting for Biden only through gritted teeth. Biden has plenty of enthusiastic support. But he has less of it in Iowa, and that could be a big problem for him.