Welcome back to Beyond the Polls. It's holiday season, and you know what that means. Beyond the

Poll's annual Festivist episode. Let's dive in. Well, you know what time of year that is? It's

Festivist season, and like everyone else who's obsessed with Festivist, I'm going to start with

the airing of the grievances.

Yes, I've got a lot of grievances with you, people, especially the polling industry.

Now, come on, polling industry.

You've been subject to a lot of criticism over the last 10 decades, and it's certainly understandable

why you would be under that sort of focus, that sort of criticism.

It's because the error rate has popped up.

The error rate is larger than it has been historically.

And it's certainly understandable why that would be the case.

As listeners to this podcast know, what's happened is the failing response rate that

people just don't pick up their phones when they have phones anymore to talk to pollsters

means that the whole theoretical basis of polling, which is a random sample of a certain amount,

reflects the opinions of a much larger population within a certain

degree of error no longer holds. You can't get a random sample. So all pollsters have been

moving towards modeling or waiting, trying to estimate what an electorate looks like. Sometimes they get it right. Increasingly,

they get it wrong. And that's one of the reasons why in 2024 all these polls had Kamala Harris

up by a point or so, and a leading pollster had Kamala Harris winning every single swing state,

when, of course, it didn't happen. People who were listening

to this podcast understood why that might happen, and people who read my election prediction

noted exactly why I disagreed with the bulk of the pollsters. It was because the systematic

...