538 + 1968 does not add up to meta-panic

 From Andrew Gelman:

"How would the election turn out if Biden or Trump were replaced by a different candidate?"


The piece is definitely worth reading, informative and well argued, but the problem is Gelman never actually discusses replacing either Biden or Trump. Admittedly, the passive voice does buy a little wiggle room, but not enough. Even if we assume the replacement has already occurred, the event itself would radically change the landscape. You can't simply plug another candidate in for either of these two men and make it anything more than an alternate history thought experiment.

More to the point, that's not what this conversation has been about up to this point. We have had prominent political commentators such as Nate Silver and Ezra Klein seriously proposing that an incumbent president having already easily cinched renomination, should withdraw from the race taking his vice president and heir apparent with him. What's more, these commentators have presented this as obviously the Democrats' best chance at holding on to the White House and preventing a disastrous second term for Trump.

In any comparison there are two quantities of interest, in this case, the probability of Joe Biden winning if he, more or less, continues doing what he's doing, compared to the chances of the Democrats winning if they make this desperate and unprecedented move. Let's look at that second number.

Josh Marshall and Scott Lemieux among others have already pointed out numerous holes in the arguments for what Marshall calls ThunderDome primaries, but while we are here let's do a quick run through. Though Klein and the rest have framed their arguments as data-based, they never really get past the level of underwear gnome plans and headless clown analogies. There's really no way they could. What's being suggested here is unprecedented, and to the extent there are partial precedents of shaking up the ticket, contesting primaries for incumbents, ThunderDome primaries, or trying big Hail Marys when running badly behind, they pretty much all undercut the argument.

Starting in 1952 (which is about as far back as we can safely go when looking for historical precedents) we have an interesting non-example where there was a great deal of pressure to dump Nixon. Eisenhower chose the calm and steady path and though we can debate as to whether or not the end result was good for the country, it does not appear to have hurt the party either then or in 1956.

In 1972, we have a twofer, a ThunderDome primary and a shakeup of the ticket. Though the Democrats' fate was probably sealed going into the election, I don't think anyone would argue these things help their chances.

In 1980, we have Ted Kennedy challenging the incumbent in the primaries. Once again, while it probably wouldn't have affected the outcome, it certainly didn't help

In 2008, John McCain tried to catch up with Obama with a big Hail Mary pass. He shocked pretty much everyone by choosing an unconventional and largely unknown but genuinely charismatic running mate who also shored up his support in the evangelical base. In many ways, Sarah Palin looked wonderful on paper, but few would now claim this was the senator's smartest move.

I skipped 1968 to save the best for last. This is the only time in living memory when an incumbent chose to step down, and this may be the only election on this list where we probably can draw some lessons, particularly if we remove RFK from the scenario. (While there is an RFK in this race, there is no one analogous to RFK, no big name obvious front runner waiting in the wings.) Ironically but unsurprisingly, based on polling I've seen of Democrats about the 2028 election, the candidate with the most support and name recognition is Harris, whom the meta-panic crowd also wants to dump.

Another really troubling parallel between then and now is a highly divisive antiwar movement, and as much damage as the Democrats saw in 1968, 2024 would have the potential to play out even worse. As angry as the pro-Palestinian faction may be, when you take out the Jill Stein voters and the heightened the contradictions leftists (sizable overlap there) who were unlikely to actually show up for Biden in November anyway, the sane faction who remain has a strong incentive to vote a straight Democratic ticket on election day. For all their threats, they know that a Trump presidency would make conditions far worse in Gaza.

In a ThunderDome convention, this anti-war contingent would have tremendous leverage and every incentive to demand costly concessions from the eventual candidate. According to the polling I've seen, most people are somewhere in the middle on this issue, appalled by both the terrorist attacks and the humanitarian crisis in Gaza. For all its controversy, Biden had staked out the popular position. That wouldn't be an option without an incumbent as the nominee. Expect similar battles over issues like abortion. At the moment, the Democrats have an incredibly popular message in large part because Biden doesn't have to go into specifics beyond restoring Roe V Wade. We are sitting on the global maximum. Any move is a move down.

Other than being younger, the candidate who would emerge from this process would still have all of Biden's other weaknesses. Just as Hubert Humphrey inherited Vietnam, any Democrat running in 2024 will inherit the perceptions of a bad economy and the turmoil in the Middle East. Given the names currently being floated, they will be largely unknown and unvetted (or worse), trying to unite a party consisting of people who recently chose someone else in a bitter primary.

The chances of this producing a winning candidate are small at best, but what matters here is relative size. If Joe Biden's situation was looking like Mondale in September, it still probably wouldn't be worth the risk but at least you could make an argument.

What do the models tell us about Biden's chances at the moment?



Even if we go with the far less optimistic economist forecast, Biden is still at one out of four, hardly where we would like to be, but nowhere near the 1% or 2% level needed to justify the ThunderDome.

Of course, this whole conversation is beyond silly. None of this ever had the slightest chance of actually happening. What's important here is not the proposals themselves, but the fact that serious people (or at least people we treat as serious) made these suggestions and that they got so much traction.