If you're a political junkie -- and if you're reading this, odds are that you qualify -- you know the pollster and the poll to which the headline refers. Just days before the election, Ann Selzer released her final 2024 Des Moines Register survey of Iowans, which purported to show Kamala Harris leading Donald Trump in a state he'd twice carried comfortably. The result detonated like a bomb in political circles. As one line of thinking went, if Selzer is right, as she always is, this means that Iowa is at least in play, which would also mean that the 'blue wall' states in the upper Midwest are probably gone for Trump. She hadn't always been right, by the way, but she had developed a strong reputation for good reason. I wrote an analysis of the poll, sharing deep skepticism from sources at the highest levels of Iowa politics. One of the points I made was that if Iowa was seriously on the board for Democrats in 2024, why wasn't that same effect showing up in Ohio polling?
Iowa and Ohio have tracked rather similarly in recent years, migrating from battlegrounds to red states. Nothing is permanent, but that's what they've become lately. Ohio and Iowa voted for Trump at similar rates in 2016 and 2020. They each bucked the national trend and did experience red waves in 2022. There has been a lot of polling of Ohio this year because of an important and competitive Senate race. Have we seen any indications of big blue movement out of Buckeyeland? The numbers have repeatedly shown Trump ahead by a strong margin, with the real question being either the incumbent Democratic Senator can overcome it and hang on for re-election against his Republican challenger in a very tight race. Is it possible that Ohio will surprise us all and be much closer than it was in either of the last few elections? Or is it possible that Ohio will stay quite red, while Iowa veers off in a completely different direction? Again, I suppose anything is possible.
As you can likely gather from that passage (here's the entire piece), I was quite dubious that she was in the ballpark. When the actual election results came in, they represented a catastrophic failure for Selzer. Trump won Ohio by more than 11 points, an expansion of his previous two margins there -- and Iowa followed suit. Trump carried the Hawkeye State by more than 13 points, substantially more dominant than his prior showings. He swept the three key 'blue wall' states, too, en route to comfortably winning the presidency. Selzer's buzzworhty poll turned out to be a breathtaking embarrassment; a 16-point miss. The only real remaining discussion was whether the debacle would prove to be credibility damaging, or credibility killing for her. The answer to that question may have boiled down to the outcome of this investigation:
Gannett, the publishing chain that owns the Des Moines Register, has launched an investigation into the apparent leak of Ann Selzer’s bombshell Iowa poll, two people familiar with the matter confirmed to Semafor. Selzer’s Iowa poll was publicly released by The Des Moines Register/Mediacom on the evening of Nov. 2, making nationwide news and giving Democrats what would turn out to be false hope by showing Kamala Harris leading Donald Trump by three points in the state. (Trump ended up winning Iowa by 13 points.) But roughly 45 minutes prior to the poll’s public release, a stray tweet predicted the poll’s findings. Its author said that Illinois Governor JB Pritzker, a Duke University alumnus, had mentioned the not-yet-released poll during a Duke Democrats meeting that day. (A spokesperson for Pritzker did not respond to an inquiry about the apparent leak.)...Gannett is investigating how Pritzker and possibly other political actors could have learned of the poll early, and is reviewing employees’ emails, one of the sources said.
I'd heard unconfirmed rumors that Selzer's previous poll in the series (post-Biden dropout), which showed Harris within striking distance of Trump, had ended up in Democrats' hands before it was released. The October data set most certainly leaked to Democrats early, and not just by a few minutes. There were whispers swirling a day or two before its publication. Some senior Democrats had it early. Ethical concerns galore. If Selzer's operation, or Selzer herself, had passed exciting-seeming results along to fellow partisans in advance, that is a scandal unto itself. Given the dramatic nature of those results, and how they were framed as reshaping expectations of the whole presidential race by many in the media, the scandal could be wider. Was the poll a humiliating error, or could it have been designed (or at least not preemptively discarded as problematic, which she had done before) to be a narrative shaper, rather than as a reflection of reality? Regardless of how this poll cratered so badly, or how the inaccurate crater was circulating among Democrats before its release, it seems as though Selzer won't be around to bear the brunt of whatever the Gannett probe may uncover. She's out:
Pollster Ann Selzer ending election polling, moving 'to other ventures and opportunities' https://t.co/Z7c7isLFa8 via @DMRegister
— Mike Emanuel 🇺🇸 (@MikeEmanuelFox) November 17, 2024
I knew a few years ago that the election polling part of my career was headed to a close. Over a year ago I advised the Register I would not renew when my 2024 contract expired with the latest election poll as I transition to other ventures and opportunities. Would I have liked to make this announcement after a final poll aligned with Election Day results? Of course. It’s ironic that it’s just the opposite...Over 30 years of polling led to an A+ rating in Nate Silver’s analysis of pollsters’ track records of accuracy. I earned that rating in Silver’s first list, and my grade never dropped. Maybe that history of accuracy made the outlier position too comfortable. Polling is a science of estimation, and science has a way of periodically humbling the scientist. So, I’m humbled, yet always willing to learn from unexpected findings...My integrity means a lot to me. To those who have questioned it, there are likely no words to dissuade. For those who know me best, I appreciate the supportive notes and calls reminding me that what drew me to them as friends, colleagues and clients was commitment to truth and accuracy...
She's claiming she decided to step away from electoral polling more than a year ago and informed the Register that this would be the final cycle of their partnership. The paper printed that assertion from her, which I doubt they would have done if she's making it up to save face. Regardless, she's going out on a decidedly low note, having bungled her final poll in the most high-profile way possible. She should be proud of her track record leading up to this disaster, but the 'A+' grade she treasures will undoubtedly take a major hit, with this prominent 'F' averaged in. I'll leave you with this nugget, which illustrates how massively this last poll failed, and not just at the presidential level:
SHE DID IT@millermeeks destroyed Ann Selzer's bogus "Iowa Poll" in #IA01, and posed for her own "Dewey Defeats Truman" photo
— Matthew Foldi (@MatthewFoldi) November 7, 2024
Selzer's poll predicted that Miller-Meeks would lose by SIXTEEN PERCENT; she won pic.twitter.com/RVQVj923o8
Republican U.S. Rep. Mariannette Miller-Meeks faces a recount for the second time in three campaign cycles after Democratic challenger Christina Bohannan requested a recount in southeastern Iowa's 1st Congressional District race. It's one of four Iowa races headed to a recount from the 2024 general election. Miller-Meeks earned the nickname "Landslide" for her victory by just six votes in 2020 after surviving a recount against Democrat Rita Hart. But the GOP incumbent is confident with a much larger lead this time of about 800 votes that the results will hold.
Miller-Meeks won by literally six votes in 2020. This time, her lead is 800 or so, which will very likely return her to Congress once this recount is completed. Another whopper of a miss. Selzer has coped with this fiasco by publicly suggesting it's possible that the release of her numbers could have motivated Republicans to turn out in greater numbers. This could maybe account for a shift of a point or two, but not 16 points, for goodness sake. That's grandiose, delusional thinking, in my view. Perhaps Selzer has made the correct choice in moving on.
