Member Newsletter
Over the last three months many Democrats have gotten accustomed to seeing absolutely abysmal presidential poll numbers and reacting in one of two ways. Either they slip into despair or they wall off the information on the reasoning that we’re almost a year out from the election and the numbers have limited meaning so far in advance. There was a period in the middle months of the year when Biden was a bit ahead of Trump or the two were roughly tied and then starting in the fall Trump appeared to move meaningfully, though still only marginally, ahead. But something odd happened this week. It illustrated a few points about the ways in which polls operate, how they’re interpreted and what meaning we can draw from them.
Let’s start with the latest NYT/Sienna poll released this week: the lead story ran “Poll Finds Wide Disapproval of Biden on Gaza, and Little Room to Shift Gears.” Not good! Indeed, not simply the article but the poll itself includes lots of ominous or at least troubling news for President Biden. But there’s one detail that didn’t make it into the story: Among likely voters, Biden actually led Trump by two percentage points. Indeed, Democratic data and polling analyst, Tom Bonier — one of the few to have predicted Democrats’ generally positive showing in the 2022 midterms — pointed out something quite notable. An earlier Times/Sienna poll which pushed Dem poll watchers into full doom mode was headlined “Voters in 5 Battlegrounds Favor Trump over Biden.” The new one which shows essentially a tie, albeit with a nominal Biden lead, was headlined “Most Disapprove of Biden on Gaza, Survey Indicates.” Indeed, unless I’m missing it, the original write up of the more recent poll didn’t even note the fact that the poll showed Biden slightly ahead. It’s always bad news for Biden almost regardless of what the numbers are.
Another poll from Quinnipiac on Wednesday told a similar story: chock full of awful numbers for Biden, but the head-to-head with Trump was essentially tied, with Biden taking the support of 47% of registered voters to Trump’s 46%. Yet another poll from Monmouth painted a similar picture but that poll didn’t have a precisely apples-to-apples head-to-head number. So it’s not as clear cut.
What does this all mean, if anything?
One: The first is that these are clearly not great numbers for Joe Biden. He’s running essentially neck and neck with a disastrous former president who launched an insurrection to remain in power after losing reelection and now faces numerous deadly serious felony charges in multiple jurisdictions across the country.
Two: We should be open to the possibility, really ready to accept increasingly abundant evidence, that what would normally be considered abysmal and perhaps fatal numbers for Biden are not at all inconsistent with his winning reelection. Monmouth had him at 34% approval; Quinnipiac 35%; Times-Sienna 37%. Each also had him either tied or marginally ahead of Trump.
But it’s beyond the approval numbers. These polls and numerous ones for months are simply chock full of numbers that are simply eye-popping. Biden wildly unpopular among young people; Biden way behind Trump on an amazing array of issues. And with all this he was just a few points back and now perhaps tied again. What’s going on? My hunch is that a lot of this is sending messages, signals of dissatisfaction, which are not clearly tied to vote selection. Many people have a simpler answer: People can’t stand Biden but they think Trump is worse. I’m not sure this is as persuasive or as on the mark as people think. That’s a topic for another post. For present purposes, the relevant point is that “cross tabs,” the second-order data points in many polls, don’t seem as tied to voting behavior as we’re accustomed to thinking. That’s not terribly surprising. It was one of the big takeaways from last year’s midterm election.
Three: This one’s pretty simple. A month ago Biden seemed like he was running three or four points behind Trump. Now he seems to be running even perhaps slightly ahead. Polls go back and forth. The difference is probably mostly noise. This is going to be another really close election, like it was in 2020. We shouldn’t get too hung up on the poll of the moment.
Four: One thing has become more and more clear to me in recent weeks: a big proportion of the voting population believes it is still unclear who the nominees of each major party will be. That puts a big, big question mark over all of these polls. The absolute reality is that the two nominees are Joe Biden and Donald Trump. Any opinion you have before you realize that central salient fact is liable to change and not especially meaningful.
Five: At least at first, the narrative guides the interpretation of polls. The story is that everything is going wrong for Biden. And it will continue that way even if the polls shift in a contrary direction. Some of the “bias” in the reporting of polls is due to decades of Republicans playing the refs and reporters being afraid of their shadows endlessly overcompensating to avoid any accusations of GOP bias. But it’s not entirely that. It’s also because most reporters at elite national news outlets view Trump as a malevolent clown who threatens the constitutional order. Because of this they find it shocking at some level that Trump is doing so well and view everything in that context as Biden’s failure. In other words, their perhaps unstated assumption that Trump is a bad and dangerous person leads to paradoxically negative coverage for Biden.