One of the best pollsters in America recently came out with its latest survey, and it’s good news for Kamala Harris.
The Des Moines Register/Mediacom Iowa poll, conducted by Selzer & Co., found Donald Trump at 47% to the vice president’s 43% – within the margin of error – in a state the former president has twice won comfortably. (The poll was conducted last week, before Sunday’s apparent assassination attempt against Trump.)
While the Hawkeye State is unlikely to be pivotal in November, the fact that Selzer found a close race in a state Trump has dominated could signal good things for Harris, both in terms of the accuracy of polling and for her chances in next-door Wisconsin …CONTINUE READING
What makes the Selzer survey so important is that it’s been accurate in an era when other pollsters have struggled. Four years ago, it had Trump up in Iowa by 7 points days before the election, when other polls had Democrat Joe Biden in a much better position. I noted at the time that “this one poll is giving Trump backers hope and Democrats anxiety.”
Democrats had good reason to worry. Not only did Trump end up winning Iowa by 8 points, he also vastly outperformed his polling in Wisconsin – nearly winning a state where he had trailed by high single-digits in preelection polling. Trump would do significantly better in many other battleground states as well.
A similar scenario played out in 2016. Selzer’s final poll had Trump ahead by 7 points in Iowa. He went on to win the state (by 9 points) and the election over Hillary Clinton, doing better than most swing-state surveys said he would.
This year, a lot of people, myself included, have wondered whether there could be another polling misfire. Even as many pollsters have tried to change how they conduct and/or weight their polls to prevent what happened in 2016 and 2020, they could still get it wrong.
Selzer’s latest Iowa poll, however, suggests that other pollsters who show a race that is way too close to call are not underestimating Trump at this point.
The result also makes sense when you look at the numbers coming out of neighboring Wisconsin, which has comparable demographics. Wisconsin is one of the seven battleground states where the polling remains tight. Harris, though, has received some of her best polling from the Badger State.
The most recent poll from the reputed Marquette University Law School put Harris at 52% to Trump’s 48% among likely voters. An average of recent polls from Marquette, CBS News/YouGov and CNN/SSRS has Harris up by 4 points.
A 4-point advantage is not wide and can easily be overcome by Trump, but it stands out among the other key swing states. Harris and Trump tend to be closer in the other six battleground states (Arizona, Georgia, Michigan, Nevada, North Carolina and Pennsylvania).
Moreover, the 4-point Harris edge is better than Biden’s winning margin in Wisconsin in 2020 – 0.6 points. That result fell well short of the lead he had in preelection state polling. The Badger State polls also overestimated Clinton four years earlier to such an extent that she became the first Democrat to lose Wisconsin since 1984.
The fact that the Selzer poll is indicating a similar shift in Harris’ favor in Iowa makes me think the Wisconsin polls may be on to something.
Related articleThe one Harris comment Trump and his allies have spent $38 million for voters to see
Now, a win by Harris in Wisconsin would not guarantee her victory by any stretch. Even if you give her the state’s 10 electoral votes (along with the states already leaning her way), she’d need to carry at least two of the other six battleground states.
Carrying Wisconsin, however, would improve Harris’ chances of winning. Instead of the race being roughly 50/50, most modelers would give her closer to a 75% chance of winning if she won Wisconsin. That’s a big shift in Harris’ direction.
Still, a 75% chance of victory is not a guarantee of anything. Trump’s chance of winning would basically be the equivalent chance of a coin landing on heads in two consecutive tosses.
This Selzer poll could also be an outlier, and much could change between now and Election Day. Selzer’s September 2020 poll was a lot more favorable to Biden (showing a tie) than its final survey that year.
But the current poll is a far better result than the previous Selzer survey from June – when Biden was still in the race. It had Trump up by 18 points.
The bottom line: You’d much rather have the Selzer poll in your corner than not. And right now, the survey shows a better result for Harris than it did in the end for either Clinton or Biden.
CORRECTION: This story has been updated to reflect that polling overestimated Hillary Clinton’s margin in Wisconsin during the 2016 election.
