Ann Selzer is inarguably the best pollster in the state of Iowa and she released a shocking weekend poll that has raised more questions than answers. In bright red Iowa – a state that went for Trump by eight points in 2020 and ten points in 2016 – Selzer’s poll showed Harris up by three.
Instead of chastizing Selzer as many of the right have done, Selzer deserves credit because she had the integrity to release the poll knowing the backlash that was coming.
Good pollsters will tell you that about 1 in 20 properly conducted surveys produces a result that is considered to be an outlier. There wasn’t anything necessarily done improperly with the modeling, but the sentiments recorded from those who were accurately surveyed don’t reflect the reality of the population as a whole.
What you’ve been seeing in political polling over the past few months is a phenomenon called herding. When the majority of pollsters produce a poll that is a significant deviation away from the averages of the other polls, they discard it for fear of being wrong and jeopardizing their lucrative private sector contracts. The problem is this practice manipulates the totality of polling and creates stagnant averages that don’t pick up late-breaking trends.
While it’s unlikely that Selzer’s poll accurately reflects the facts on the ground in Iowa, the survey’s shift among women and senior citizens away from Trump and towards Harris should be the primary concern among Trump supporters. It appears that the Puerto Rico comments in the New York City rally are having a meaningful impact at the end of this race. But is it enough to overcome the Trump advantage in early voting? Here’s what I think:
A Squirrel, a Raccoon, and an Election
