The 2024 presidential race is close, and well within the polling margins of error (both currently and historically). So, at quick glance let's say each candidate has a 50/50 chance.
But in my mind, the top of ticket polling (especially national polling) shows us little. There are other metrics that matter come election day that seemingly give the advantage to Democrats / Harris:
which candidate or party has more money to close with (+Democrats)
who is winning the down ballot statewide races in tossup states (+Democrats)
which party has been favored most of the season on the generic ballot (+Democrats)
which party won a majority of the battlegrounds in the last elections (+Democrats)
…and specifically looking at the last presidential election: which candidate lost a majority of these battleground states (-Trump).
I’m not saying this because I want Harris to win; I’m saying this because there are campaign funding totals that we can look at, there are down ballot races we can look at, there are past elections we can look at. I go back to an article I wrote for the Independent Center in July called Why 2024 Looks A Lot Like 2012 – it’s going to be close. And even if I were to say all of these little metrics only give Harris a 60/40 advantage, 40% odds of something happening happen all the time (4 out of 10 times to be exact).
Before we get into the data, a little disclaimer: I am not pro-Harris. I am not pro-Trump… I voted this year, and didn’t support either of them (I wrote someone in). I have no dog in the fight other than I’d like either one of the candidates to clean up electorally so that we don’t have months of recounts.
Oh, and I can’t see into the future. That said, let’s get into it.
Anyone reading the polls closely needs to read this article from ABC showing how ‘off’ state polls have been over the years. Or this fantastic op-ed by Ezra Klein aptly titled Ignore the Polls in which he details how “In 2016, an American Association for Public Opinion Research postmortem found that the average error of the national polls was 2.2 points, but the polls of individual states were off by 5.1 points. In 2020, the national polls were off by 4.5 points and the state-level polls missed, again, by 5.1 points.”
(GASP… “Brett, how dare you say this, aren’t you a pollster?”… yep.)
The (inaccurate) argument I hear is that the polls underrepresented Trump in 2016 and 2020 – which is true of national polling, not necessarily true of state polling (see Pennsylvania, Georgia, Arizona, Nevada…). Alternatively, people say likely voter polls inflate the Democrat… Tell that to Obama in 2012 in which he outperformed his polls by three points (and I would argue that Harris’ campaign looks a lot more like his 2012 campaign than Clinton’s ’16 or Biden’s ’20 campaigns).
So, with the table set, and showing that polling can be off, I generally say ‘look at the aggregates.’ But that’s tricky this year because even there we’re not seeing agreement. I bet if you look at RCP, you think Trump is going to win. If you look at NYT, Harris is going to win. That’s because now the aggregates have wildly different projections.
At the time of writing this (Sunday, October 27, 2024), the nationwide race looks like this by aggregate:
RCP: Trump + 0.1
The Hill: Harris + 0.7
NYT: Harris + 1
538: Harris + 1.3
To show the difference a little further, RCP has Trump leading in every tossup state, whereas The Hill, 538 and NYT have Harris up in Michigan. 538 and NYT have Harris up in Wisconsin. NYT has Harris up in PA and Nevada (The Hill and 538 have Nevada and PA as dead even – either tied or at less than half a point).
SO, polling in the battlegrounds this year is well within the 5.1 margin of error for state polling. If I were to look at any of the aggregates, I’d say it’s close.
But I would break the tie by looking at the reverse coattail effect (down ballot candidates bringing up the top-of-ticket). To explain this, look at the presidential race in battlegrounds on 538:
It looks “even.” But now look at the race with the down ballot ticket along side it:
This tells me the Democrat / Harris ticket has the edge because there is either more support among Democrats for their statewide candidates, more Independents are leaning toward the Democratic ticket, and/or more Republicans are willing to split ticket. Whichever it is, this indicates a net positive for Harris and the Democrats.
RCP, which we showed leaned a little more Trump than the others, also favors down ballot Democratics in Presidential battlegrounds, if at a narrower margin (note that MT is not a Presidential battleground):
North Carolina is the only Presidential battleground with a gubernatorial race this election (and no Senate seat up there), and the polls heavily favor the Democratic candidate in that race:
When I show this data, I generally get the same responses. Either “Does reverse coattail voting ever happen in Presidential elections” (yes – see google), or a different, less logical “Well if the top of the ticket is tied, that means Trump will bring up the down ballot candidates.”
But there are two major problems with the latter argument. First, in the last three elections (2018, 2020, and 2022), when Trump was at the helm of the party, his candidates in battleground areas did not fare well. So why would it be different this year? See ‘negative coattail effect.’
Second, when I reverse engineer this argument, the logic seemingly flips in people’s minds. If I say, “If Trump and Harris were tied in Arizona, Michigan, Nevada, North Carolina, Pennsylvania and Wisconsin, BUT the next race on the ballot is statewide and favors the Republican…” the same people suggesting that Trump will bring up the down ballot candidate react with “That means Trump will definitely win!”… But you can’t have it both ways. If down ballot Republicans are ahead and that ‘means’ Trump will win, then down ballot Democrats leading should ‘mean’ Harris will win. Logic doesn’t flip just for your party.
Listen, with the exception of early / absentee voting which I’m not putting much stock into because people can only vote once, and running up a score early means you might not close as strong, and whatever other pseudoscience people are using to predict the voting trend differences from 2020 to 2024 (I find to be bologna) – I believe a lot of the other indicators favor Democrats and the Harris team (money, favorability, down ballot candidates, ground staff, etc.).
I mean, if it was truly a Trump runaway, we wouldn’t be talking about North Carolina (which he won in the last two elections). We wouldn’t have state polls ‘even.’ Heck, we wouldn’t have 3 of the 4 aggregates noted above giving Harris the edge. And we definitely wouldn’t be talking about statewide Republicans in Texas and Nebraska fighting for reelection, or heck, the Montana Senate race being somewhat close. But alas, we are, and that’s what makes me think while it may close, Dems / Harris have a slight edge this year.
At the end of the day though, I’d love to put out a little reminder: whether you’re team Red or Blue, please please please remember we’re all team Red, White and Blue (Team ‘merica!)
Comments