A guide to post-debate poll-watching

What to look for in the polls — and the model.

Kamala Harris had a strong debate last night. Although you might equally say that Donald Trump had a weak debate: repeatedly veering off-topic, especially when goaded by Harris, to talk about things like a conspiracy theory about migrants eating cats and dogs (?!?) in Ohio.

Yes, readers, this was the former President of the United States speaking. And there’s still a good chance that Trump could become president again — although Harris probably reduced the chances of that last night. In the CNN “instapoll” of debate-watchers, Harris won 63-37, a 26-point margin. That isn’t as large as, say, Trump’s 34-point win over Joe Biden in June — or Mitt Romney’s 35-point margin against Barack Obama following the first presidential debate in 2012. But it’s more than the average winning margin in this poll, which is 18 points.

While the instapoll can be Democratic-leaning, perhaps because the debate-watching audience is also predominantly Democratic¹ — in fact, Romney in 2012 and Trump against Biden are the only cases of Republicans winning in the CNN poll since 2000 — it typically does translate into gains in post-debate polls. So the rest of this newsletter is about when you might expect Harris’s gains, if she has any, show up in the Silver Bulletin polling averages and forecast. I’ll even do something dangerous, speculating on how big the impact might be. Harris entered the evening with a 38 percent chance of winning the Electoral College in our model. What are the chances she’ll be at 50 percent or higher once all the post-debate polling is accounted for?

The 5 phases of post-debate poll movement

After major events like debates and conventions, or otherwise when there’s been a shake-up in the race, there are some prototypical patterns of polling movement. The polls don’t follow this trajectory every single time, but I’ve been doing this for long enough to know it’s a common course. And it proceeds in roughly five phases.