What the Post-Debate Polling Landscape Looks Like
September 17, 2024
The first complete post-debate polling is done. And except for an obvious outlier (see discussion below), Harris has gained ground.
I doubt the gunman on the golf course in Florida on Sunday will give Trump any advantage - even a temporary bump. What it might do is blunt some of the absolute craziness swirling around his campaign and Republicans over the weekend related to his strange relationship with a fascist podcast grifter. Whether it is because of politics or something personal, it was a story threatening to spin out of control before some guy appeared 400 yards away from Trump with a gun yesterday. One thing that appears probable: Republicans were starting to see the same down-ballot carnage that Democrats saw when they began to suggest Biden step aside. Does that stop now, or will it get worse? Time will tell.
Here are the first completely post-debate national polls:1
ABC News/Ipsos: Harris +6 (52-46)
Reuters/Ipsos: Harris +5 (47-42)
Yahoo News/YouGov: Harris +4 (50-46)
TIPP Insights: Harris +4 (47-43)
Data for Progress: Harris +4 (50-46)
Angus Reid Global: Harris +4 (49-45)
Big Village: Harris +7 (51-43)
Morning Consult: Harris +6 (51-45)
Redfield & Wilton Strategies: Harris +2 (44-42)
AtlasIntel: Trump +3 (51-48)
The outlier: AtlasIntel has Trump up by three points nationally. A lot of Trump supporters on social media were crowing about how this pollster was the only one that got the 2020 margin correct (not true; TIPP did too). But that tells us nothing about whether this poll is an outlier. Pollsters do not get it right every election; even the best miss and sometimes the lousy ones hit. But does the margin and Trump leading tell us this poll is an outlier? Not exactly, but it is very different that the other pollsters mostly having Harris at +4 or better. That’s a seven point swing or more. No, the real evidence that this is an outlier is a little deeper in the data.
When we look at the other topline results and crosstabs that we see serious problems with this poll. On issue questions, the poll has Harris just +2 on abortion and Trump +5 on healthcare! It has Walz -15 in popularity. It's way off. None of those results make sense. This looks as obvious an outlier as ABC News/Washington Post’s 2020 Wisconsin poll that had Biden up by 17 points. I would not take this poll seriously (the poll, not the pollster). Unless they have captured shocking new movement within the electorate on everything (or unless they are using an electorate model that is way off and they don’t correct for it), this result will not be replicated.
While we see movement to Harris, and that has mostly been true since she joined the race, what is interesting (but not surprising) is that Trump’s support has stayed the same - usually in the 45-47% range. From his polling and his past two national election results, it appears more and more likely that he has a ceiling of support that goes no higher than 47% nationally.2 I think the Trump folks know this. It’s probably why they wanted Kennedy in the race, even though it was a miscalculation and he has been a bigger drain on Trump support than Harris. (This is why they are now trying to get Kennedy removed from the ballot in battleground states.) Trump’s ceiling of support is also why the campaign is going so negative and offensive. Aside from the fact that Trump is a negative and offensive person, the campaign needs to turn voters off this fall. They assume their base will turn out no matter what, but that if they can turn off enough potential Democratic voters or push them to minor candidates like West and Stein they can eke out a victory in the Electoral College. The assumption that their base will turnout no matter what has been tested, as we saw with the reaction to Trump trying to moderate his position on abortion rights.
We are not seeing much state polling yet. Suffolk University has a post-debate poll out for Pennsylvania that has Harris +3 (49-46). Insider Advantage, a lower-rated pollster, has Trump +2 (50-48) in the Keystone State but has Harris up +2 in Wisconsin. Suffolk is the only top-tier quality pollster that has yet released a post-debate battleground state poll.
There is an interesting development in Iowa. Selzer & Co., a highly-regarded polling firm (especially in Iowa elections), released a poll on Sunday that had Harris trailing there by just four points (Trump 47, Harris 43). Iowa was once a strong Democratic state, having supported the nominee as recently as 2012, but has since shifted strongly to the Republicans. Trump beat Clinton by over nine points in 2016 and Biden by just over eight points in 2020. The Congressional delegation likewise has shifted to the Republicans. But there have been some rumblings that Iowa might be swinging back in the Democrats direction this year, even if it falls short of victory for Harris. Ann Selzer seems to really have her fingers on the pulse of Iowans, but this poll could be an outlier or maybe just as good as Harris can do there. It is notable that this is not a post-debate poll. Fieldwork began on September 9th, the day before the debate. According to Politico, Democrats are dumping a lot of money into IA-01 and IA-03, where they think they may have shots at beating Republican incumbents in those districts which were considered unwinnable as recently as a few months ago.
With Florida and Texas both within five points in the polling averages, this makes one more state Trump cannot take for granted. He cannot lose any of these states and expect to win - even losing just Iowa would blow up his path to victory. And even more competitive than the presidential race in Florida is the Senate race. With the Democrats newfound belief that they can win Congressional and Senate races that only a few months ago were considered solid GOP seats - a belief that is undoubtedly supported by some internal polling - we may have a clue into why Republicans were erupting this weekend about Trump’s relationship with the podcaster and his and Vance’s doubling-down on racist lies about Haitian immigrants in Ohio.
Perhaps Republican Congressional leaders are seeing what Democratic leaders were seeing in June after the first debate with Biden. The guy at the top of the ticket might be threatening their incumbents’ chances at reelection (and any chance they might have at majorities). Republicans have made a lot of humiliating compromises for Trump. If he is going to cost them their jobs, then what value does he have for them anymore?
As of this morning, these are all of the national polls that were in the field no earlier than September 11th. There are some polls that were in the field on or before the debate on September 10th and then afterwards, but those are not included here. These polls are the one that were in the field entirely after the debate.
This is another reason to believe the AtlasIntel poll, which pegs Trump at 51%, is an outlier.
I really appreciate your analyses.