Member Newsletter
I want to commend to your attention this article from Nate Cohn at the Times. It looks at the weak point for the NYT-Siena poll, and, indeed, many other polls this election cycle. In short, Donald Trump’s current lead is heavily focused on people who didn’t vote in 2020 and tend not to vote in general. They tend not to follow politics closely or pay much attention to traditional news sources. This isn’t new to our discussion. It’s sort of the internal anatomy of the gap between polls of registered voters and likely voters.
This is a fairly big deal. It has always struck me as inherently unlikely that what many suspect will be a relatively low turnout election (relative to recent cycles) will be determined by voters who tend not to vote and didn’t vote in 2020 — voters who, in this election, are supporting Trump in greater numbers. It’s not impossible. But it’s hard to figure. And this is what Trump’s current lead in most polls is based on. To be clear, this isn’t some hidden defect in the NYT-Siena poll or Cohn’s earlier logic. He’s discussed this issue throughout. And this article today focuses on it.
Before you get too excited about this pattern, there’s a lot of churn among the “non-voter” category. People won’t vote in one election and then they will in the next. Cohn’s piece has good statistics on this. Some people who didn’t vote in 2020 will vote in 2024. The question is how many and which ones. And that’s complicated and to a significant degree, currently, unknowable. The best interpretation from the Trump perspective is that these voters not only tend to be electorally marginal but also economically marginal. They’re young voters, voters of color, poorer voters, etc. So if the driver this election cycle is persistent unhappiness about higher prices for economically stressed voters, especially after early-post-pandemic wage gains at the bottom of the economic scale have flattened out, well … maybe these are precisely the voters who are going to turn out after not doing so in 2020.
I don’t know the answer to this. And I don’t think anyone else does either. But this is the main reason why there are real questions about the current poll numbers. It’s not “denial,” as a lot of screechy insider sheets put it. There are significant questions, which make it very plausible (though by no means certain) that Biden could be in a substantially better position than the polls suggest. The current polls tend to be polls of registered voters. So they mostly don’t look at these turnout questions at all.
Cohn prepared this chart which visually captures this pattern.
This is a pretty stunning differential. And like I said, it’s not new. It’s really a different way of looking at the “Likely Voter” screen issue.
Another aspect of this part of the polls is that this is not just a turnout question. Biden’s big problem is not only with people who tend not to vote but people who are not following the election. In other words, they’re people who will tend to be more up for grabs than usual when you move into the high-media-saturation phase of the campaign. That doesn’t mean they’re going to switch to Biden. If they’re really pissed at Biden maybe they’ll stay that way. The point is that you’re talking about the voting intentions of people who aren’t paying any attention to the election yet and will probably only start at the end of the summer.
What I take from all this is not that Joe Biden is secretly ahead. But I think there’s a very good chance the race is significantly more favorable to Biden then it looks. What I take with some confidence is that the race is close and that it’s quite possible, without any wishful thinking, that Biden is better positioned than he looks. So any hand-wringing and throwing up of hands like it’s hopeless is not only a bad idea for all the standard reasons. It’s just not a realistic or accurate take on what the available data, including the polls, is telling us.