Debate Reaction and Polling Impacts
September 11, 2024
Brooklyn. Wednesday morning. September 11th. I got up this morning and went to the corner bodega to get a bagel and coffee. As I waited, I looked down and saw the New York Post next to a few other papers. I didn’t process the front page at first because I was thinking, “there are still physical newspapers being sold in bodegas?” When I snapped out of that, I noticed the headlines accusing ABC News and Harris of ganging up on Trump last night. The message was the typical right-wing whine that it is unfair when the lose and no one has a right to call them out on their lying and BS. I knew at that moment that Harris must have destroyed Trump last night. Even his supporters knew it. They were freaking out - and the Post was telling them how to complain about it. Abby Phillip had a fair response.
Something people who don’t know me well are surprised about1 is that there are two things about politics I do not like to watch: conventions and debates. I think people are surprised because these are often the only two political things people volunteer to pay attention to in campaigns. Ads and news coverage is usually thrust at them, and they can look away if they want. But a lot of people choose to watch conventions and debates. Not me.
It’s not true that I never watch any of it, but what I do watch is often after the fact. I prefer to see how everyone else reacts to them and how that impacts the polling afterwards. I am not interested in analyzing the convention speeches or debate performances in any other context except how they impact public opinion and campaign decisions. So, I have not seen any of last night’s debate yet. But I have read the coverage of it.
It’s pretty clear to be from both Democratic and Republican reactions to the debate that Harris absolutely crushed Trump. That’s not a surprising result to me. Harris is really good at this, and has been skilled at this type of discourse for decades stretching back to her time as an Alameda County assistant district attorney. Trump is a terrible debater.2 What he does is bully his opponents, shamelessly lie, and engage in the Gish Gallup tactic. Something most of his opponents in his own party have been unable to handle. But Trump lost his three debates with Clinton and his debates with Biden. But even those opponents had a difficult time figuring out how to handle him (Clinton actually did a better job of it). What is interesting, though, is Trump always behaves the same way (the degree of craziness might waver, but not the craziness itself) and everyone seems concerned that he is going to show up and behave himself and wow the people with expectations so low you’d think that would already disqualify him as a credible candidate. But like the fable of the turtle and scorpion, the scorpion will always sabotage himself in the end because it’s in his nature to act as he does.
So, what does this mean for the horserace? Does a Harris debate win give her a clear lead in the race? Well, first of all there is likely to be some kind of bump in her support for doing this well. Even Romney got a bump in his support after his first debate with Obama in 2012, and Romney was never going to win that race. The real questions are when will see a bump and will it last?3
There is likely to be some polling in the field right now, but that won’t be too helpful unless pollsters report how responses changed from before the debate to afterwards. Even then, it’s not that helpful. We will probably see the first entirely-post-debate polling on Thursday or Friday. It’s good to have a poll in the field for three days, so it may not be until this weekend. That polling should give us an idea of what kind of immediate reaction voters have to the debate. We should wait another week before making any preliminary conclusions that an increase in support for Harris post-debate may be permanent.
The polling has been so close - and there have not been a lot of undecideds, although there are leaners and they might be persuadable - so the change that this debate might create might be almost unnoticeable in the polling but create enough new voters for Harris that she wins close states. Another thing to consider is how can a candidate like Trump still be getting this much support in the polls? The electorate certainly looks like it’s baked in at about 48-48. With at least three minor party candidates on the ballot - and Kennedy still on some ballots - it’s not clear there are many more votes to be won. There is a good chance there will be 2-4% of the vote looking for a none-of-the-above option come November and those votes will go to one of these minor candidates. But the polls might be deceiving us a little right now.
Good pollsters are fairly transparent in the methods. They will tell us who they reached out to, how they did it, and in general how they weighted the results. But their models of electorate turnout are not always clear. Many pollsters consider this kind of modeling to be something akin to their intellectual property or trade secrets. So, we often don’t really know what kind of turnout they assumed in their polling until after the election when they made a mistake and it was apparent in their polling misses. For the really transparent pollsters, it is possible to uncover their turnout models with some expertise and diligence. Digging deep into this area is not something I have the time to do, so I listen to others who make the attempts. And there seems to be some concern out there that a lot of pollsters have created a turnout model that looks like 2020 or 2016. This is likely because turnout both exceeded and came from areas in the population that most pollsters did not take into account in those years. But by using those models, they are favoring a Trump turnout advantage. This is likely why the polling is so tight.
Now, it is possible that turnout will look like 2020 or 2016. We will not really know for sure until the election is over. However, there are indicators that will give us a good idea before Election Day. Voter registration among young people and particularly young women of color has skyrocketed in battleground states according to Tom Bonior, who accurately predicted the Democratic advantage in the 2022 midterms when others were talking about a red wave. Soon, early and mail voting will start. When it does, we can get a picture of the partisan composition of those who have voted early and those who have asked for mail ballots or returned them. There is already some reason to believe that turnout will be more like the 2018 and 2022 midterms than the past two presidential elections. And that kind of turnout would be an advantage for Harris. If that happens, then Harris may actually be up by 4-5 points in the national polling than 2-3 points. It is unlikely at this point that pollsters are missing much of Trump support. There are just not a lot more voters in Trump demographic categories to turnout than there were in 2020. If pollsters are using that as a model, then the polling right now is probably in the ballpark of where the results will be should that turnout assumption be correct.
There is another phenomenon occurring now that also occurred in 2022: polls by conservative PR firms that show Trump doing a lot better than the results from quality polling firms with track records. One reason they do this is to frame public opinion rather than report it. The purpose is similar to push-polling where the survey interviewer is trying to convince the respondent to give a certain answer favorable to the survey sponsor. But there is another, even more insidious reason possibly at play here. To keep expectations that the race is close and Trump has a chance to win, so that if Harris wins a convincing victory they will use their polling as “evidence” of voter fraud in trying to prevent certification of the election. This “evidence” will be used by their allies who have been place in election board positions across some battleground states (like Nevada and Georgia) to refuse certification.
The reaction I am seeing this morning from Trumpland is one of panic. They probably think the election is now going to get out of their control. I don’t know if we can say that with any certainty this year. In a normal year, with a normal GOP nominee, the race would be over. But it should have been over a long time ago. This is, after all, a guy who tried to overthrow the Republic, has threatened violence against opponents, says he will be a dictator on day one, is a convicted felon awaiting sentencing, and he is still polling in the upper 40s.
Actually, even people who do know me well seem continually surprised by this.
People forget how bad Trump was in his debate against Biden. It is only because Biden was so terrible during the first 30 minutes - and it eventually ended his reelection campaign and political career - that Trump’s performance gets so little mention. Against another candidate, he would have looked terrible. That hypothesis was confirmed last night against Harris.
Some commentators call a temporary bump a “bounce” and only use “bump” for permanent bounces. I just use “bump.”