Are these guys trying to muddle through? Or playing to win?

A debate preview — plus a quick model update.

This is mostly a newsletter about tonight’s debate, but first I wanted to thank you so much for the great reaction to the release of the presidential forecast model yesterday. I have to admit it was kind of amusing seeing the reaction from partisans on Twitter and comparing it to the relatively civil and sophisticated discussion in our comments section (along with subscription numbers that were way toward the high end of my expectations). I think we’re building a nice community here, in other words, and I’d encourage you to sustain the friendly vibes.

Eli and I haven’t quite figured out a rhythm for when we’re going to update the model. The original plan was just once a week until Labor Day, but we’re clearly going to be able to do better than that. In fact, we’ll try to get you an update most of the time once there’s interesting polling or some other data (such as a new economic report) that materially affects the model. In practice, that probably works out to a few updates a week. Though we hope you’ll give us a few mulligans on days where we’re both out of pocket or my dream comes true and I make another World Series of Poker final table or something.

When we run one of these interim updates, we’ll usually just update the landing page without a whole lot of fanfare, other than maybe noting the change on Twitter and Substack Notes. You probably know the drill by now: polling averages are available for free, while the probabilities and forecast components require a paid subscription.

Besides, we don’t think it’s a good use of your time to react to every new poll that comes out anyway, which rarely have all that large an impact on the topline takeaway. Today was a good example of polls with bad headline numbers for Biden that didn’t really move the forecast much, for instance.

The bad headlines came from national polls from Quinnipiac and The New York Times/Siena College that put Biden several points behind, and he now trails Donald Trump by 1.5 points in our national average headed into the debate:

But the impact of national polls on the model can be a bit funky. I’m skipping some nuance here, but our forecasts are mostly determined by state polls rather than national polls, and national polls mostly influence the model only indirectly through how they affect the calculations it makes about state polls. Sometimes national polls can have a big effect, especially in states that haven’t been polled much recently — but sometimes the model just shrugs its shoulders at them.

And sometimes there’s subtle stuff going on beneath the surface. For instance, although Biden got poor numbers in the NYT poll, it did show him doing meaningfully better among likely voters than among registered voters — and that affects the adjustment the model applies to all registered voter polls. These polls also showed a large undecided vote, which increases the model’s uncertainty parameter (helpful for whichever candidate is currently behind). And the new state polls in the past 24 hours were something of a mixed bag: there were a bad set of numbers for Biden from a Republican group, but the model adjusts for the fact that they’re partisan polls.

After yesterday’s debut post, I don’t think I’m at any risk of being accused of shilling for Biden, so I’ll just say this particular set of numbers probably shouldn’t change your view of the race much — unless you were under the impression that Biden had pulled into the lead in national polls, which he hadn’t in our average to begin with. There’s more detail over at the landing page. And if you just can’t get enough election model stuff, Maria and I also discussed it on the latest episode of the Risky Business podcast.

How to watch the debate-watchers

The first of two scheduled presidential debates will be held tonight. (No, you’re not hallucinating: there’s a debate in June.) Debates are something I’ve always found tough to cover. Until you see polling showing how the candidates did, media coverage — including mine — is unavoidably subjective theater criticism.

Only, it’s subjective theater criticism that doesn’t break the fourth wall and admit to being subjective theater criticism. “Straight news” reporters are pressed by their editors to come up with interesting things to say that won’t sound subjective but still seem differentiated and smart. Former deputy assistant press secretaries from some failed campaign 8 years ago join cable news panels to spin for their side — but the game is to avoid making it look like they’re spinning. The degree of partisan ref-working is never higher than during debates — nor is the degree of groupthink. It’s less the wisdom of crowds than the madness of crowds, in other words — and it can sometimes take on an unpredictable life of its own.

Whether voters react to what they see through their own eyes or to the spectacle as filtered through the media is a whole different question. If you’d asked me ten years ago, I’d have told you it’s the media coverage that matters more. But I’ve become less sure of that: the media is so distrusted that people barely believe it about relatively objective facts, let alone its sense for the vibes. Biden’s State of the Union address got pretty good media coverage, for instance. But its impact on the polls was ambiguous: probably a little bit helpful for Biden¹ but certainly nothing that fundamentally reset the race.

I’d expect the media to be particularly weird about how it covers Biden’s age and any accusations of diminished capabilities when it comes to extemporaneous speaking or thinking on his feet. This is an area where the partisan ref-working perhaps has been effective: after a period of intense coverage of Biden’s age in the spring, there’s been less lately, and the media has mostly bought the White House’s line about recent videos edited to portray Biden in an unflattering light as Republican “cheap fakes”.

If you haven’t been following these controversies, don’t worry about it, because the downstream effect on what you really ought to care about — voter preferences — is hard to say. The trickle-down effect between how the New York Times frames a certain headline and what some politically disengaged swing voter in Latrobe, Pennsylvania thinks about Biden is unclear at best. And if voters see headlines that seem like they’re spinning for Biden, they may not only be unpersuaded by them but may come to have less trust in the news outlet that publishes them.

My original plan for combatting all this weirdness was to watch the debate in an “isolation chamber”, meaning shutting myself off from any social media reaction. You were still going to get my subjective theater criticism — but at least it would be subjective theater criticism unburdened by groupthink. Unfortunately, that’s not going to work out, since it turns out that I’m going to need to be somewhere tonight where I’ll be watching the debate tonight surrounded by other people who very much will be tuned into the vibes. Instead, we’ll go with the more traditional approach of this preview today, probably followed by a follow-up reaction late tonight or early tomorrow — although if I don’t have anything interesting to say, I might pass.

What would a higher-variance strategy for Biden look like?

One of my consistent points of criticism of the White House is that it isn’t acting as though it’s behind — perhaps because it’s engaged in some poll unskewing and doesn’t believe it’s behind. For instance, the Biden campaign’s tactic of moving this debate to June was arguably a risk-averse move, since there’s now just one debate after Labor Day when it would likely have a larger effect on the Nov. 5 results.

How would Biden’s strategy in the debate differ if he accepts that he’s behind? One way might be by being more combative with Trump by — I don’t know, pointing his finger at him and saying he’s beneath the dignity of the office of president or something. That could make for some good video clips if pulled off effectively but could also fall flat, especially since the debate (at the White House’s request) won’t be before a live audience. Another route might aggressively pivoting to the center on an issue like immigration — Biden sorta tried that in the State of the Union, but then hedged by later apologizing in an interview for referring to an “illegal” rather than “undocumented” immigrant.

Biden was once an aggressive debater — like against Paul Ryan in 2012 — but an uncomfortable question is whether his age affects his ability to deliver such attacks crisply. Without getting into too many age-related stereotypes, older people tend to be risk-averse and they don’t tend to be good at pivoting and changing course. And both Biden and Trump — who recently turned 78 — figure to be rusty; neither have debated at all since 2020, since Trump refused to attend the Republican primary debates and the Democrats didn’t even bother to hold any. 

If the White House thinks of its primary objective tonight as muddling through and averting disaster — and then spinning around what are surely low expectations for Biden — then maybe that’s the right tactical play given the cards in their hand. But it also suggests they have a poor hand to play — and for the moment, it’s a losing hand, at least if you believe the polls.

For Trump, there might be merit to a more risk-averse approach, and perhaps even to a more sober performance where he doesn’t risk seeming unhinged or like that he’s bullying Biden. Do I expect Trump to deliver that performance? LOL. Trump has lost every general election debate he’s ever had, according to post-debate snap polls, and hasn’t shown much ability to shift gears. So you may wind up with the candidate who’s ahead in the polls acting like he’s behind and going for a knockout blow, and vice versa. I’ll be back — probably — with a reaction tonight or tomorrow.

1. It’s hard to say because the timing coincided with both candidates wrapping up their nominations, which often does have an effect on the polls.