Mark Mitchell, the self-taught pollster for Rasmussen Reports, has a little but revealing secret to tell about his mysterious business.
“It’s not rocket science,” said the Wharton School graduate and retired nuclear Navy officer who just completed his first presidential cycle with the best Election Day record in the industry.
Mitchell is an outlier in the insular industry. He doesn’t live in Washington. He’s the first to criticize the Washington-based polls when he sees bias. And he’s quick to respond to critics of his polls.
“This is my first presidential cycle. So I guess an untrained, non-statistics person can come in and school all these clowns,” he told Secrets.
https://t.co/CNtxskeFt1 pic.twitter.com/wkvdH7CUS0
— Rasmussen Reports (@Rasmussen_Poll) November 7, 2024
Only two polls called the popular vote nearly exactly: Rasmussen Reports and the Wall Street Journal. Both had the final vote as 49% for President-elect Donald Trump and 46% for Vice President Kamala Harris. Trump won 50%-47%.
The difference in the two polls, Mitchell said, is that Rasmussen had Trump leading Harris for 15 straight weeks, while the Wall Street Journal polled before the conventions, after them, and right before the election. What’s more, the Wall Street Journal showed a Harris convention bump that Mitchell said didn’t happen.
As for the other polls, especially those showing Harris winning, Mitchell blamed liberal bias.
“I’ve been calling these people out this entire cycle as shilling for Democrats and being absolutely shameless liars,” he said. Mitchell even said that some polls acted as “psyops” for Democrats.
At the end of the election, the gold standard Iowa survey from pollster Ann Selzer showed Harris winning the state in what turned out to be a 16-point error. Some in the political world raised their eyebrows when it was published, but most in the media went with the wrong results to claim it showed a big anti-Trump wave coming on Election Day.
State Forecasting Accuracy by Organization, Battleground States Only.
Atlas Intel 100%
PollFair 86%
Rasmussen 86%
Trafalgar 71%
Insider Advantage 71%
Real Clear Politics 71%
Fabrizio 67%
Remington 67%
Emerson 57%
Wall Street Journal 50%
Beacon (Fox) 50%
Marist 50%
Morning…— PollFair (@poll_fair) November 7, 2024
Meanwhile, as Mitchell showed Trump consistently ahead nationally, “everybody laughed at us” and called Rasmussen a “right-wing” pollster, he said.
To keep his polls accurate, Mitchell said he was constantly tweaking his sample to make it the closest possible to the voting public. And he said that even with that effort, his poll sample tilted slightly left.
He nailed most of the state battleground surveys. Three came in exactly as he predicted: North Carolina, Pennsylvania, and New Mexico. His only miss was in Texas.
SEE THE LATEST POLITICAL NEWS AND BUZZ FROM WASHINGTON SECRETS
In the end, he said his “absolute” polling error was 1.9 points, less than half of most others. “You can’t do much better than that,” he said, adding that “it’s impossible to get to zero without luck.”
Mitchell said polling will always lean left as long as the “legacy media” fund them. In a YouTube video for his website, he said, “I really don’t think the industry is going to change until legacy corporate mainstream media dies.”
