Betting markets think Biden should quit
Democrats — other than the White House — are finally acting like the party that cares more about winning.
Yesterday’s critique of the 538 model was originally supposed to be our Model Talk column for the week, but I wound up running it free (check it out if you haven’t yet). But today’s newsletter includes plenty of frank discussion on probabilities and strategies as well — so we’ll slap the Model Talk label on it, which means that the second half will run as a paid post.
As a consolation prize for free subscribers though — I know there are a lot of you who are new to this newsletter and I’m really glad you signed up — here’s the latest edition of the Risky Business podcast, which was unavoidably another politics-heavy episode. Timing note: we taped this on Wednesday afternoon right before the pressure really began to ramp up again on Joe Biden to drop out, but I think most of it should hold up reasonably well.
Meanwhile, we’re starting to see a big change in our polling average, with Trump now ahead by almost 4 points. This is a matter of Trump’s numbers increasing as undecided and third-party voters come off the fence — rather than Biden’s numbers decreasing — presumably as a result of the assassination attempt against him and the start of the Republican National Convention.
The forecast/forward-looking component of our model is designed to hedge against changes in the polls surrounding the party conventions, so Biden is “only” down to a 26 percent chance of winning the Electoral College. The model will want a couple weeks to see whether Trump can sustain these new, higher numbers. Still, Biden’s win probability is at a new low point for the year.
Markets think Biden should drop out
At Polymarket, however, Donald Trump’s chances of winning have dropped to 62 percent. That’s down quite a bit from a peak of 72 percent on Tuesday morning.
While I have complicated feelings about the relative merits of models and prediction markets1, this is one of those times when prediction markets are probably the more useful tool. That’s because political events have been unfolding faster than polls or models have time to catch up with them. Just in the past week, Trump was shot at and nearly killed, named J.D. Vance as his running mate, and accepted the Republican nomination with a very long speech in Milwaukee. Meanwhile, of course, Democrats are trying to figure out what to do about Biden, and our model isn’t accounting for the possibility that another candidate could replace him.
Are traders reacting negatively to Trump’s meandering acceptance speech yesterday? No, not really. If you zoom into the Polymarket data, you’ll find that the market was essentially unchanged before and after the speech. And that seems reasonable enough. Speeches don’t usually move the needle, and the first 30 minutes or so — where Trump spoke movingly if subduedly about the assassination attempt against him — may offset the self-indulgent hour that followed.
But the speech did serve as a reminder that we have two men with huge egos running for president — and they aren’t necessarily calibrating everything to maximize their chances of winning.
Trump has run the better campaign so far, I’d say — for instance, by downplaying the GOP’s unpopular position on abortion. But the past week looks like one of missed opportunities. Vance gave a better speech than I expected on Wednesday, but he isn’t from a swing state and — on paper, anyway — doesn’t offer much of a bridge to moderate voters. What if Trump had picked, say, Virginia Gov. Glenn Youngkin instead, who was elected in a blue-ish state and could expand Trump’s map by moving Virginia closer to the tipping-point line? Instead, though, Trump was heavily influenced in his VP choice by his son, Donald Trump, Jr.
And I’m not one who can really complain about going long — having just written a book that weighs in at more than 500 pages — but Trump’s speech could really have used an editor.2 I don’t want to overstate the case — there are worse things than being boring, especially if you’re ahead in the race and want to run out the clock — and Trump looks like he’s getting a boost in the polls anyway. But the instinct to take full advantage of the vibe shift and show voters a “New Trump” dissipated after the first half-hour.
The reason the betting markets are moving against Trump, though, is because Democrats are getting their act together in trying to usher Biden off the stage. It’s somewhat useless to link to reporting because it’s changing on a minute-by-minute basis, but party bigwigs — Nancy Pelosi, Barack Obama, Chuck Schumer and Hakeem Jeffries — are being increasingly unsubtle in their efforts to get Biden to quit. Officially, Biden is insisting that he’ll stay in the race, and there’s a lot of contradictory reporting about what his inner circle thinks, but his chances of dropping out have spiked up to 70 percent on Polymarket and were as high as 90 percent at one point yesterday.