Democrats are favored by nine points on the generic ballot and Donald Trump has a 55 percent approval rating. That’s what Atlas Intel released this week, and it got liberals all giddy. That is until you dig into the figures. Mark Mitchell of Rasmussen Reports was not impressed. First, he said that Democrats aren’t ahead by nine points. Second, he noted the sample: “With the sample sizes involved, the atlas poll is less statistically believable than the Selzer Iowa poll.”
That infamous poll, which some allege was a suppression effort by desperate Democrats, had Kamala Harris beating Donald Trump in the Hawkeye State by three points. It was facially untrue. That would mean pollsters, the media, and campaign operatives from both parties missed a massive 16-point swing. Harris was never in Iowa if that was an indicator of anything.
Bookmark this:
— Mark Mitchell, Rasmussen Reports (@honestpollster) May 31, 2025
The Atlas Dem+9 GCB is going to act like a magnet pulling every "high quality" pollster left.
Then the media is going to hammer how unpopular Republicans are all summer.
Some of the crosstabs made no sense, but Mitchell added to bookmark his prediction:
The Atlas Dem+9 GCB is going to act like a magnet pulling every "high quality" pollster left.
Then the media is going to hammer how unpopular Republicans are all summer.
He explains more about how Atlas, which was one of the most accurate polls in 2024, might have lost its way on this survey:
No, Democrats are not up nine points.
— Mark Mitchell, Rasmussen Reports (@honestpollster) May 30, 2025
Recommended
There is simply no such thing as a D+9 registered voter sample in the U.S., not even LV. https://t.co/SuYEaRbwdN
— Rich Baris The People's Pundit (@Peoples_Pundit) May 30, 2025
EVERY pollster in the industry is far to the right of Atlas' Dem +9.
— Mark Mitchell, Rasmussen Reports (@honestpollster) May 31, 2025
That's all you need to know.
The results had Republicans for DEI. pic.twitter.com/8bfLItmPbp
— Brittany Rae (@legitbrittFLA) May 30, 2025
With the sample sizes involved,
— Mark Mitchell, Rasmussen Reports (@honestpollster) May 30, 2025
the atlas poll is less statistically believable than the selzer Iowa poll
J Ann Selzer's sample was like five times smaller than the atlas sample...
— Mark Mitchell, Rasmussen Reports (@honestpollster) May 30, 2025
Just saying.
Basically Atlas with their latest poll:
— Mark Mitchell, Rasmussen Reports (@honestpollster) May 30, 2025
"I, for one, welcome our socialist overlords..." pic.twitter.com/urjdtAsthG
The atlas poll has Democrats up nine points.
— Mark Mitchell, Rasmussen Reports (@honestpollster) May 30, 2025
Also important to know that in the poll, AOC is nine points more popular than ending DEI in the government.
D+9?
— Mark Mitchell, Rasmussen Reports (@honestpollster) May 30, 2025
But Trump is down almost 30 points with women, all when other pollsters show Trump rebounding?
I normally keep my mouth closed, but this is J Ann Selzer level.
Sorry for the lack of context, I'm talking about the Atlas poll which I don't think is right.
— Mark Mitchell, Rasmussen Reports (@honestpollster) May 30, 2025
Join the conversation as a VIP Member