OPINION: This article may contain commentary which reflects the author's opinion.
The polls that were ignored by many in the media in the election between President-elect Donald Trump and Vice President Kamala Harris were the most accurate.
The Real Clear Politics average of polls was the most accurate forecast for the 2024 election, and it continued to show a victory for the president-elect.
The company’s co-founder, Tom Bevan, spoke with Washington bureau chief Carl Cannon and podcast host Andrew Walworth on the Real Clear Politics radio show, where they examined how their polls did.
“In terms of the swing states, we predicted Trump winning five of the seven and Harris leading in the other two by less than half a percent. On average, our averages were within 1.8% of the final outcome in each of these states. Nationally, we had Harris up by 0.1%, and Trump is going to win the popular vote by about 1.5%,” the co-founder said. “So the polls weren’t terrible this year.”
“And all these polls that the folks at the New York Times like to denigrate — Rasmussen, Trafalgar, Insider Advantage — did very well,” he said. “Atlas Intel, the most accurate pollster four years ago, was unbelievably accurate this time. And, by the way, the New York Times/Siena poll was atrocious. They predicted Harris winning four of the seven swing states, had two tied, and Trump winning Arizona. They predicted Harris winning 274 electoral votes without Michigan or Pennsylvania, which is preposterous.”
“And yet the Times criticized our average on the eve of the election,” the company’s Washington bureau chief said. “And said we shouldn’t count these other polls (the polls that were right).”
“They ran this hit piece on us five days before the election,” Bevan explained. “Saying these right-wing polls are distorting the averages… painting this distorted picture that Trump is winning and is going to win the swing states. Look, that’s exactly what happened.”
“Our averages did better than any other average out there, and that’s for two reasons. One, they’re only as good as the polls that go into it. And two, we don’t ‘massage’ the polls. We don’t weight them or futz with them at all. The numbers are just the numbers, despite the fact that some of our competitors have all these bells and whistles, these black box calculations that they’re doing. That ends up making the numbers worse. That’s why the RCP Average continues to be the most accurate and the gold standard in the industry,” he said.
Cannon said, “We have this No Toss-Up Map, and that is whoever is leading in the average gets the state. And until a few days, maybe a week, before the election, our No Toss-Up Map had Trump winning all seven of the swing states that he carried. And then CNN came in late with polls in Michigan and Wisconsin with Harris winning by six points in each, which tipped the averages over [on October 29 and 30]. Had it not been for those CNN polls, we would have had all the states right!”
“I’m not trying to pick on CNN here,” he said. “It’s the New York Times I’m really trying to pick on… I’m waiting for my apology from the New York Times! Do you think I’m waiting in vain?”
Walworth explained the reason when he said, “We do not weight our polls. We take a straight average of the polls that we include. So there is a judgment involved in which polls go into the average, but once they’re in, it is not like we say that this one was more accurate last time, so we’ll put a little magic sauce on that and give it a little bit more emphasis. It’s just an average.”