When you should panic about the polls
Most of the time, you shouldn't — but here are the exceptions.
Ever since I’ve been covering elections in 2008, Democrats have been prone to bouts of panic about the state of the horse race.
Sometimes it’s justified. If anything, there was too little panic from Team Blue about 2016, which was dismissive of polls showing a tightening race down the stretch run and a tenuous Hillary Clinton lead in the Electoral College. And sometimes it isn’t. Although John McCain had periods of strong polling in 2008, especially after the Republican Convention, in the end Barack Obama won in a landslide that didn’t justify Democrats’ constant anxieties.
If anything, this tendency is growing more pronounced in the Trump era. Republicans tend to be boastful of polls showing strong results for Trump, while Democrats tend to be perpetually anxious — to the point that Democratic campaigns can even find it strategically advantageous to portray themselves as underdogs because it tends to boost fundraising.
The media can play a role, too. For various reasons — but mostly the scar tissue left over from 2016 — the incentives are probably toward portraying Republican prospects in a favorable light. In 2022, this got out of hand with predictions of a “red wave” that were never actually justified by polls or models that had Democrats on track for an above-average midterm.
This week, Democratic panic has been elevated. Why? A combination of little things. The recognition that the election is only four weeks away. Scary news in the Middle East — and from hurricanes in the Southeast. Misreading signals from prediction markets. Concern about Kamala Harris’s performance in interviews.1
Of course, if Democrats see another Trump term as an existential threat, they should be panicked. The chances of a Trump win in the Electoral College are about even. If there’s a hurricane bearing down on your backyard, you probably don’t want to parse the finer points of whether the odds of a direct hit are precisely 40 percent or 48 percent or 52 percent so long as you’re in the Cone of Uncertainty.
But generally speaking, the polls aren’t nearly so fast-moving as hurricanes and the outlook doesn’t change that much from day to day.
Take the set of Quinnipiac polls released on Wednesday afternoon, for instance. They were pretty good for Trump: Quinnipiac has him ahead by 3 in Michigan and 2 in Wisconsin, though Harris leads by 3 in Pennsylvania.2 The previous Quinnipiac polls of these states had been quite a bit better for Harris. And they have relatively large sample sizes.
And yet, there are dozens of polls released every week. And Quinnipiac polls tend to be “bouncy”: they’re not one of those firms you can accuse of herding. So how much should they change your view of the race?