It will be that close

All realistic signs point to a very close election; everything else is vibes

“Maybe the American people [are] quietly cooking something up, something we don’t know about. I think they are and I think it’s this: a Romney win…. All the vibrations are right.”

-Peggy Noonan, 11/5/2012

Josh Marshall has a recent piece arguing “Maybe It Won’t Be That Close?” The piece is appropriately hedged, noting that polls have underestimated Trump support in the last two presidential elections. But he also argues that we’re choosing to ignore some important data points — including a Democratic enthusiasm advantage, Kamala Harris’ increasing favorability ratings, and some curious polls in places like Iowa — suggesting that Democrats may be on the path to a substantial win. I appreciate what he’s saying, but the very smart money is still on this being an extremely close election.

In 2016, there was every reason to expect a substantial win by Hillary Clinton. Clinton had led in public opinion polls all year, sometimes by 7 or 8 points. Trump was caught bragging about committing sexual assault literally a month before election day. Clinton had absolutely mauled him in three successive presidential debates. Democrats weren’t being willfully obtuse about their coming loss — the available evidence really did point to a win. Nate Silver was one of the few poll-watchers to note that undecided voters were behaving strangely in 2016, but it was mainly a way of saying that there was more uncertainty than usual, not that there was a Trump win impending. Any indication that Trump was about to win the Electoral College came from things like crowd enthusiasm, Michael Moore’s gut feelings, angry e-mails from retired uncles in Michigan who used to be in labor unions — you know, vibes ‘n shit. These are typically not very reliable or replicable indicators.

In 2020, there was every reason to expect a substantial win by Joe Biden. Biden had led in public opinion polls all year, by margins greater than Clinton had led by four years earlier. On top of that, we know that presidents are generally held accountable by voters for crises that occur on their watch, and nearly half a million Americans had died from a pandemic and the economy was in utter chaos during Trump’s last year in office. The only reason most Democrats weren’t outwardly crowing about Biden’s coming landslide was that 2016 had burned them so severely.

Trump received 46% of the vote in 2016. He received 47% of the vote in 2020. This translated into a 304-227 Electoral College win for Trump in 2016 and a 232-306 loss for Trump in 2020. Biden won the popular vote by more than Clinton did, but the real difference was determined by a few tens of thousands of voters in places like Pennsylvania, Wisconsin, and Michigan. Polling is pretty good these days, even given the array of challenges it faces, but it can’t really tell us anything about the 50,000 or so voters who will make the difference among the 150 million or so who will cast their votes this year. We just don’t have instruments that accurate.

One could imagine a situation in which Harris well overperforms. There are at least a handful of datapoints, including a few very positive state polls, suggesting some strong support for her. But one could also imagine situations in which Trump overperforms. It’s plausible his supporters have grown even more disgusted by the media and pollsters and are even less likely to respond to surveys than usual, but they’ll still show up and vote. But it’s not clear to me that one of these outcomes is any more likely than the other.

What we do know is that Harris has, on average, about a 3-point lead in public opinion polls nationally, people are already voting, and this 3-point margin is about what Democrats have needed to pull to draw the Electoral College to a tie. What we can be pretty sure of is that this 3-point margin isn’t likely to shift around a whole lot between now and November, that Trump will probably get around his usual 47% of the vote, and that the third party vote will be small but not negligible (especially if RFK is still on a few state ballots).

We can guess at other things, but really that’s not based on anything more than cherry picking some data or just going with one’s gut. What we can gather from the past two elections with Donald Trump on the ballot all points to another very close contest with no one winning by anything considered a landslide. Past performance is no guarantee of future behavior, but it tends to be a better predictor of future behavior than our guts.