When I decided it was time to end the blog, I made myself a quick promise-for the next 50 days or so, that I would be allowed to write as many blog posts as I wanted, but only until the morning of November 4th (i.e. tomorrow). There were ones (like the Halloween blogathon or the Lost articles we've discussed in the last week), that I had actually promised to you, the readers, that I would complete, and I took those more seriously, but there are definitely articles (like this morning's about To Do lists) that I kind of just added because I thought they would be fun. This was sort of my way of saying goodbye to a lot of reoccurring topics we've talked about on this blog, but also a bit of a "patch" for me to get all of my thoughts out before I moved on. I don't have a lot of friends in real life to talk about these things with (it's partially why I have a blog), so I knew that impulse would be hard to control, and while I'm quitting the blog for legitimate reasons in my personal life, I also wanted to give myself a little bit of grace to talk about what I needed to, especially in the heat of the election.
I did not think that I would still be having these urges the day before my self-imposed deadline (I wanted to end the blog with the final article of the Election Night Guide, which I always publish the morning before Election Day, and that will still happen tomorrow), but, well, as I'm still technically in that window I might as well go for it. Yesterday we had the release of the infamous Selzer poll, an historically (and scarily) accurate poll that frequently has caught things that other pollsters have not, and this year's was no exception in terms of upending conversation. So we're going to do an unexpected goodbye to yet another long-running feature of this blog (the "5 Thoughts On..." political dissection articles) and just have four articles for our penultimate day of publishing (cause yes, I have one other article I'll publish later this evening before tomorrow's finale).
 |
| Ann Selzer |
If you're not a political nerd, this is a fair question. The Selzer poll (officially referred to as the Iowa poll) is the longest continually-running statewide newspaper poll in the country. Run by the Des Moines Register, it started in 1943. While political polling has been around in the United States since the 1820's, it became popular in the 1920's when The Literary Digest correctly predicted the winners of every presidential election from 1916-1932. They famously missed the 1936 results (which was a landslide in favor of President Roosevelt, while they had guessed Gov. Alf Landon would win), and the newspaper soon ended publication. George Gallup figured out the mistakes of The Literary Digest, and would instead conduct more scientific, smaller-based surveys that were more accurate, and the Register used that method to create the Iowa Poll, which they still conduct sporadically for surveys of the state.
Ann Selzer came to the Register in 1984, and has conducted the surveys in the state since 1987 (hence why this is frequently shorthanded by many, including me as "the Selzer poll"). Selzer is known for being extremely diligent, having an extensive knowledge of the state that goes back over her 30+ years of surveying it, and for being eerily accurate. She was the only major pollster to predict that Barack Obama would win the Iowa caucuses in 2008 over Hillary Clinton & John Edwards, and she correctly guessed in 2014, 2016, & 2020 that Iowa was going to swing right of the nation. In 2016, in particular, she proved what other polls were glossing over-that Donald Trump was headed to victory. Even when she's wrong, she tends to not be wrong by much-while she guessed the wrong winner in 2018 for the governor's race, she was still only off by 5-points, hardly something to scoff, and damn impressive when you think about a 5-point miss being the biggest gap in her major race record.
 |
| Former President Donald Trump (R-FL) |
The reason the poll was so shocking is that Selzer found that Kamala Harris was leading Donald Trump 47-44% in the Hawkeye State. This is a huge, huge deal if it's correct, and if you've been following along in our Election Night Guide (which, as I pointed out in the Iowa section, I wrote before the Selzer poll so I could stay on deadline), you'll know why: Iowa isn't a swing state anymore. While during the presidential elections of George W. Bush & Barack Obama the state was continually contested, it moved right in 2016 and has stayed that way since, with Joe Biden & Theresa Greenfield both losing there in 2020 while the rest of the nation moved left. If Iowa is actually competitive, what does that say about the rest of the presidential race, which up until now has been billed as an incredibly tight endeavor? Is this a case (like in 2008, 2014, 2016, & 2020) where Selzer is able to find something that other pollsters are simply ignoring in the results?
It has to be noted that backlash to the poll was severe from the GOP, but notably without a lot of Republicans countering with their own polling to counter it (which makes me think it hurt a nerve that backs up some of their own worries). Donald Trump took to Truth Social this morning to point out Emerson (a Republican-leaning pollster) had him up by 10-points and pleaded for farmers to support him. Iowa Republican Gov. Kim Reynolds spent last night attacking the Des Moines Register, but also in the process begging Republicans to vote for Donald Trump to "prove the Register wrong." Some leaks happened that said that Harris being up in Iowa was nonsense, but that Republicans had Trump up by about 5-points, which given Biden lost the state by 8-points, would still be a solid result for Harris and point to an overall win for her on Tuesday (in the nation, if not in Iowa). If this is even remotely true, it could show that Trump has bigger issues than the bulk of polling have suggested.
 |
| State Rep. Christina Bohannan (D-IA) |
Before you start having visions of "Harris Wins in a Landslide" dance in your head, I want to give a mild reality check. First off, it's possible Selzer is an outlier. She's better than most, but she still can be wrong (though, to be fair, she's never been this wrong if Emerson is accurate). Secondly, there's also a chance this is mostly an Iowa thing. Selzer's poll showed that Harris's lead was largely predicated on an over-performance with women voters and with senior citizens, both of which are things we've seen in other polls, but they also could be more pronounced in Iowa due to more local issues and local population (Iowa has a greater percentage of the state being white women and senior citizens than most other states). Selzer said on MSNBC this morning that the abortion ban going into effect in the summer of 2024 was a big driver in getting women interested in the election. While Dobbs has had a pronounced effect on voting (more on that in a second), it would be more recent in Iowa, and could be a way for women in the state to punish Republicans to try to overturn the ban. Additionally, Trump's push for tariffs being a key part of his economic plan is going to go over like a lead balloon in a state like Iowa, where farmers and agricultural businesses are far leerier of tariffs than the average voter. Both of these are things that are very local to Iowa, and are worth remembering before you start applying this to races in Wisconsin, Indiana, Nebraska, & Michigan (which a lot of people, including me, were doing on social media last night).
And third, Iowa isn't that important to larger presidential math. Iowa is only worth six electoral votes. As we've seen with Nevada all year, that's not enough in a chess match to counter Harris's need for Pennsylvania or Trump's need for North Carolina. While Trump losing it would give Harris more paths (pretty much any path where Nevada was important suddenly becomes clearer), it still puts an undue amount of pressure for her to win big in places like Pennsylvania, Michigan, & Wisconsin. Where the math is more important is in the House. Though Selzer didn't poll congressional candidates by name (a crucial thing to note, as that wasn't the case with Harris/Trump), she did show Generic Democrats doing MUCH better in IA-1 and IA-3, and nearly winning in IA-2. If Democrats Christina Bohannan & Lanon Baccam were to win, Mike Johnson would have to make up those races somewhere else, in a map where he's already bleeding heavily in California & New York. Put bluntly, I don't see how he loses both of these seats and remains Speaker.
 |
| Vice President Kamala Harris (D-CA) |
The bigger question here is one that I've been asking for months-is Harris being underestimated? We have spent most of the past few weeks seeing virtually identical polls out of all of the seven swing states, which feels absurd that we think all of the seven swing states will be decided by 1-2 points (more on that in a second), but it has to be noted that Selzer isn't totally standing on her own here. Other prominent pollsters have shown similar results, albeit without the headlines that the Selzer poll demands. The Docking institute (polling Kansas) showed Trump up by only 5-points there, while Miami University showed Trump up by 3-points in Ohio, and SurveyUSA showed Trump up by 5-points in Maine's second congressional district. These are races that Trump won by 14, 7, and 7-points respectively. These races would show Harris gaining ground, not losing it as most of the swing states (and especially the national popular vote surveys) have shown, compared to Biden.
One key question that needs to be asked, and I don't know how to delicately ask it so I'm just going to say it straight out-are pollsters ignoring women? In 2016 & 2020, pollsters (save for Selzer) frequently underestimated the "Quiet Trump" voters, and in my opinion have been shifting the environment in 2024 (assuming a situation where Republicans outvote Democrats by 2-3 points, which there's not a lot of indication at this point should be the case given Harris voters appear more enthusiastic than Trump voters as we head into Election Day) to counter that, assuming it'll happen a third time. Given that Democrats have predominantly more support among women, and that (especially white women) have been feeding a lot of the surveys that show Trump losing (specifically the Selzer poll), I do wonder if the polling industry is trying to counter the Quiet Trump voter while at the same time missing the biggest headline of 2024-women are turning out for Harris in a record way. It is not lost on me that, if this is the case, it will have been caught by Selzer, one of the few prominent women in an industry that is largely populated by men. Not a great look for the men of the polling industry if they spent the year ignoring women, and only one of their female peers caught what may well appear obvious on Wednesday morning.
 |
| Harry Enten |
5. Will Herding Hurt the Polling Industry?
The industry has made a point in recent weeks of basically saying the exact same thing-that this is a tied race. But in an actual tied race, you would see more variety in the polls. If Trump is actually neck-and-neck in Michigan, you would normally see results where Trump is up by 3-points and down by 4-points, because polling is not a consistent enough industry for this to look so uniform. When all of the polls are showing the same thing, it implies that the pollsters, not wanting to stray too far from the narrative (or to be wrong) are doing what is known as "herding." Herding essentially means all pollsters are giving the same results to ensure everyone is wrong in the same way, and everyone has covered their butt.
This is a bad look for polling, and has been called out by political analysts like Nate Silver as "cheating." If all of the polls are being shifted and shaped so they say the same thing, they largely become worthless. Pollsters aren't gong to own to that (it looks bad, and makes their data worthless, and puts them out of a job), but the complete lack of outliers from major publications like The New York Times, Fox News, & CNN, particularly countered with Selzer's results, puts up a dilemma. After so badly screwing up in 2016 & 2020, if Selzer once again calls them out and is the only major pollster to catch a Harris wave, what's the point in the polling industry? A lot of egg will be on the face of people like Nate Cohn & Harry Enten, who have spent months insisting on a close race in their polls & analysis...if they don't deliver, it's hard to imagine what credibility they'll still have at the end of this cycle.
No comments:
Post a Comment