My Take on "Harris +3 in Iowa" and the Battle for Narrative

Topic: Perspective

Yesterday, I planned to minimize my time consuming election information and posting on social media and instead work on a few battleground state predictions, and wound up pushing out two – Nevada and Arizona.  That plan changed for worse when Ann Selzer, the vaunted “gold standard” pollster of Iowa, pushed out a Harris +3 poll for Iowa.  Not the national popular vote, but for Iowa.

That would be the same Iowa that Trump won by 9.4% and 8.2% in 2016 and 2020, respectively.  The same Hawkeye State that neither campaign mentions, visits, or sends any serious resources to…and the same state that correlates very closely with voter sentiment and behavior in the critical decisive state, Wisconsin.  Selzer’s reputation for accuracy is such that I even wrote an article about her eye-popping Trump +18 poll against Joe Biden in June, and how it spelled certain doom for Biden in Wisconsin, and serious danger in Minnesota even with all of the “election fortification” that has been put in place by their Democrat trifecta.  Selzer called Trump by 7% in both of his previous races, which were off to the left by 2.4% and 1.2%, respectively – certainly not a sign of dishonesty, especially given that the state voted for Barack Obama twice before the working-class realignment in favor of the GOP under an America First agenda.

The good news about this poll is that no steel worker in Cambria County, Pennsylvania, is lamenting its release and planning to sit things out on Tuesday, and certainly Trump’s swarms of working-class voters in Marathon County, Wisconsin, don’t have the slightest worry in the world.  I suppose you could spend a week in both counties and not find a soul who has even heard her name, let alone considered her polling record.  I am usually not the first person to jump on a new outrage, although I do believe in conspiracy.  This poll reeks of dirty play, and I’ll tell you why.

First, her initial Trump-Harris poll in September showed Trump +4, with too many undecideds – but that result was still four points to the right of her September 2020 poll showing Trump and Biden tied, which she adjusted right seven points for final release to come close to the final mark.  A Harris +3 poll at go-time means she pulled herself seven points left, and 21 points left from Biden +18 in June, when practically all other battleground states are polling the same or further right than they were right after Biden bombed the debate, with the same light blue states coming into view today, such as Virginia or New Mexico. Yesterday, Morning Consult, probably the most consistently bad 2024 major pollster, showed Trump at +9 in Ohio – so here we have the two Midwestern Obama states that are now rock-solid Trump states moving in opposite directions?  I don’t think so.

Here are the facts on Iowa today, compared with insights about Selzer’s poll:

First and foremost, the Democrats have lost more than 203,000 active registrations, including many to a voter roll purge that also clipped Republicans, but at a much-reduced rate.  The result has been a major shift toward the GOP, with the state moving from R+1.0% in active registrations to R+10.3% today.  You can see the final numbers here and represented in the below chart I made last month that captures the gist of what I’m getting across to you right now:

To believe Iowa is sailing left, let alone as far left as Selzer maintains, one would need to believe the active Democrat registration totals fell by over 203,000 and all 99 counties became more Republican or less Democrat as a share of registration.  Voter registration has been perfectly predictive of all elections in Iowa this century:

One reason Iowa has barely been polled this year is because most people can read those numbers, as well as interpret candidate disinterest in visiting the state, to mean that Iowa is destined to back Trump by a bigger margin than it did in 2020, most likely to the north of 10%.  In fact, Emerson College must have caught wind that Selzer was playing dirty pool, because they cut her off at the pass like badasses yesterday and dumped a Trump +10.5% lead in Iowa right before she beclowned herself on the big stage.  Unlike a few pollsters the media vilify as biased to the right, Emerson doesn’t fit that mold.  For all the virtue signaling the left are doing online about Selzer releasing a non-herded poll, I don’t see a whole lot of consideration for Emerson’s result, which jives with reality. Emerson also seems to consider the state of Early Voting, which shows a GOP turnout edge for the first time that I can find. Notice the GOP leads Early Voting by 7,276 voters, whereas they trailed it in 2020 by 128,375 and in 2016 by 41,881 on the way to Trump blowouts:

The methodology for the Iowa poll is atrocious.  It sounds like she polled the Lincoln Project or a women’s psych ward, because “protecting our democracy” and “abortion” came well ahead of immigration and the economy as priority issues among those polled.  The recall vote of the poll (who did you vote for in 2020?) was Biden +2, over ten points left of the Trump +8.2 result (meaning she pulled her answers from a group that backed Joe Biden by two points, not a group that backed Trump by a margin matching the state’s certified result - this would be like polling Massachusetts Republicans and assembling a poll of their answers suggesting the state would turn red), with D+3 used as party identification and not the current R+10.3.  Ryan Girdusky gave some more details:

This article was supposed to be short and to the point, so read those splits yourself.  So, why would this poll be released?  My mind suggests the words from the immortal “Million Dollar Man” Ted DiBiase – everybody’s got a price.  Can I prove that?  No, at least not yet.  But here is what has happened since the poll came out:

Polymarket’s presidential winner market has pulled close again, with Trump down from the 66.9% favorite three days ago, to up just 10 percent right now. Of course, this presents a buying opportunity for Trump, but narrative matters in modern elections.  It is Trump who has commanded the narrative around this election for the entire year, and finding enough ballots to pull this election off in the eyes of the public will require a reversal of narrative favoring Harris.  One major indicator the public is looking for a Trump win is the disposition of the betting markets.  This Iowa poll plunged Trump’s odds in the national betting markets and flipped Wisconsin hard to the left, and damaged Trump’s odds in the other key Midwestern states.

To me, every aspect of this poll should be brought into question, as it suggests Iowa should vote to the left of Pennsylvania, Michigan, and Wisconsin this year, falling right in line with Minnesota.  No indicator, including polls by others, suggests anything even near this result.  This result is so laughable that it is probably less troublesome than if she had showed Trump +5, which given her reputation for accuracy, would have opened major cause for concern given that most people in election analysis think Trump needs to get Iowa by at least 9 to be in the driver’s seat in adjacent Wisconsin.

This poll, if we can even call it that, stinks to high heaven and raises many questions as to just what in the hell is going on behind the scenes that so much manipulation is occurring at the ground level in quadrennial rituals used to set public expectation for election outcomes.  You can’t sell me a steal if you don’t set the narrative, and this poll has all the makings of a major PSYOP used to influence markets and set expectations.  It is insulting to all who analyze and understand election data.

---

Seth Keshel, MBA, is a former Army Captain of Military Intelligence and Afghanistan veteran. His analytical method of election forecasting and analytics is known worldwide, and he has been commended by President Donald J. Trump for his work in the field.