One Data Point Told Us Biden was in Big Trouble
July 23, 2024
A weekend can be a long time in politics. On Friday, Republicans thought they had the election already won. Trump had just the night before given a long, boring, and rambling offensive nomination acceptance speech, but none of them seemed to care – even though some delegates left the convention before it was over. Picking Sen. JD Vance (R-OH) as his running mate was both a sign of thinking Republicans could run up the score in November and a glaringly obvious choice for Trump.1 Polling continued to show Biden losing ground and there was evidence this was threatening Democratic candidates down-ballot. Nonetheless, the president remained steadfast that he would not withdraw.
By the end of the weekend everything had changed.
It appears to be a bit of thing on social media to ask where you were when you heard the news on Sunday that Biden had indeed withdrawn. I was walking home from getting coffee listening to a House of the Dragon podcast when several friends started shooting me texts about some big news that they were so sure I had heard about they didn’t actually mention it. It did not shock me like it seemed to shock so many other people. There was plenty of reason to believe that he was going to withdraw, and it would be on Sunday – although some of what I thought was going on behind the scenes wasn’t actually happening.
The battleground polling was brutal and from what we heard internal district-level polling showed big problems for Democrats. Former Speaker Nancy Pelosi (D-CA) and other Democratic leaders had been trying in private to convince Biden for a couple of weeks that he could not win and he was threatening the entire ticket. Monday was going to be a really bad day for Biden if he didn’t withdraw. Congress was going to be back in session, which meant there was going to be plenty of opportunities for Democratic Senators and Congressmen to plot. It was clear that Pelosi was serious about Biden’s chances and his need to withdraw. She was certainly about to go public and call on Biden to do so. I expected an avalanche of public defections on Monday, followed by Pelosi and former President Barack Obama sometime afterwards. It had to be Sunday. Biden had to withdraw on Sunday to prevent a full-scale public repudiation of him by the party.
Well, he did withdraw on Sunday but not necessarily for the reasons I suspected. Biden was isolating in Rehoboth Beach as he recovered from COVID, seeing only a handful of close advisors and family. His longtime close advisor Mike Donilon came to him on Saturday with the news: he had the data now and there was no path to victory. It’s curious why it took them so long to come to this conclusion, but reportedly for the past two months they had not been doing polling in some states (we don’t know which ones – but it sounds like ones they should have been polling). Until that time, Biden was not going to withdraw. But after hearing the news from Donilon, they drafted the withdrawal letter that he posted on the platform-formerly-known-as-Twitter the next day. Biden only informed a few high-level campaign and White House aides of his decision a minute or so before he posted the letter Sunday morning. And that is probably the source of so much shock.
I didn’t expect to hear it this way, either. I assumed he would address the nation Sunday evening.2 He would not have been able to do that without the news leaking out beforehand. For whatever reason, Biden wanted to control how it got out. The only way was to do it the way he did it. His campaign had confirmed just before the letter was released that he was not withdrawing. No one on the campaign apparently knew of his decision until just before the rest of us did. That this was done in so much secrecy makes the coalescing behind Vice President Kamala Harris even more impressive. Thirty-six hours after Biden’s announcement, Harris had secured enough delegates for the nomination and raised about $100 million.
What did Biden and Donilon see in the data that made them realize they could not win? I don’t know for sure, but I know what I saw that convinced me it was nearly an impossible task: the share of Democrats supporting him in battleground states. There were two problems with this data point. First, it was too low. He was getting less than 90% of Democrats supporting him. Biden won 94% of Democrats in 2020. Clinton won 89% in 2016.
Second, Trump’s partisan support was over 90%. This means that Biden would have to work on getting Democrats to vote for him whereas Trump did not have to do the same with Republicans. Trump got fewer Republicans supporting him 2020 than Biden’s partisan share. Trump and Clinton got equal shares in 2016.
There were several public battleground polls released over the past couple of weeks, but the data I am using here to illustrate the problem is from the YouGov poll released last week. Here are the topline results from this poll:
As you can see, in the five battleground states where there are Senate campaigns Democratic candidate is doing significantly better than Biden. And it’s not just that they are doing better, they are all winning and Biden is losing. How does that make sense in an era where there are fewer and fewer voters splitting their tickets? One thing that explains it is the relative partisan support for Biden and the Senate candidates. See Tables 1 and 2, below.
Source: YouGov
When we look at the results for Senate candidates, we can see that there is not a consistent problem of Democrats failing to support Democratic candidates, but of Democrats failing to support Biden.
Source: YouGov
Democratic Senate candidates in battleground states are consistently beating their opponents in partisan share of the vote and are all in the 90s. While Biden just hit 90% in Nevada, he comes up short in every other state. And in all states, including Nevada, Trump beats him.
Biden cannot win if he cannot get more Democrats to support him, as the Senate candidates have done. But this only partially explains why Biden was doing worse than the Senate candidate. The obvious follow-up question to this is: why was Biden doing worse? This question is not directly answered in the data, but another poll found that 21% of Democrats thought Biden’s age was a more important concern than Trump’s felony convictions. It’s hard to believe that any significant number of Democrats do not believe that Trump’s criminal history is a problem, so for one in five of them to say the president’s age is a bigger concern tells us that Biden – and not just any potential Democratic nominee – is the issue. Voters seem to support much of the Democratic program and the party’s Senate candidates, so seeing Biden failing to win Democratic support – in some states, he is down ten points relative to the partisan support the Democratic Senate candidates are drawing – makes no sense unless voters are unwilling to support Biden himself.
We’ll see some polling on Harris out soon.3 I would not place too much stock in previous polling because it was all hypothetical. It no longer is. Quinnipiac has already polled the new race, finding Harris down by two points (49-47). That’s within the margin of error, and it’s a lot better than Biden was doing in the very last poll I saw before his withdrawal (Trump +5). Aside from the topline result, there is good reason to be optimistic about Harris’ chances. She is getting 97% support from Democrats, which is not just significantly more than Biden was getting in the battleground states, but more than the Senate candidates were getting. If this is accurate, it means that Harris doesn’t have to work to get Democratic votes. She has them. She can focus on winning over independent, minor party, and undecided voters – and maybe even a few Republican voters.
If you really pay attention to Trump, you would know he has three deal-breaking criteria for a vice president. That person has to be (1) white, (2) male, and (3) subservient to him. The idea that Tim Scott, Bryan Donald, or Nikki Haley ever had a shot at the role was purely public theater.
He said that he plans to make an address later this week.
Thanks, Robb.