Moby Dick is a useless guide to polling, vol. 4
Moby Dick is a useless guide to polling, vol. 4
to the last i grapple with thee; from hell’s heart i stab at thee; for hate’s sake i spit my last breath at thee, polling coverage
The sun sinks, the moon rises on this election, and we conclude this series on why Herman Melville was right that we should not stare into polls for too long, lest we lose ourselves to phantasms and grim fancies, but totally failed to tell us what in fact we should do with polling. My previous missives in this series are here, here, and here - if you’re new to the series, you’ll want to read at least the last one of them before this.
And so we come at last to the question that I think defines this election and thoroughly illustrates how polling coverage has gone askew: what voters does Donald Trump need in order to win next week?
In this respect, much of the media coverage of the election has gotten the headline right: Trump needs men. Specifically, Independent men and young men. Men can be young and Indy, obviously; for this discussion, when I talk about Indies, I’m mostly talking about Millennials and older, and when I talk about young men, I include GenZ independents in that group.
To state the obvious: Indy men is not a big universe of voters, because, as previously noted in this series, most voters knew what they thought of Trump and whether they would vote for him or against him before this campaign cycle even began. The first presidential debate broke and reordered that status quo a bit, as a segment of Democratic voters and a huge chunk of the Independent electorate bailed on the Democratic ticket; virtually all of those Dems and a good portion, perhaps a majority, of the disaffected Independents returned to it after Kamala Harris’s accession (this is where I remind readers that my firm’s work is in qualitative research, including some qual-at-scale, but we do not deal in statistically significant samples).
But, as noted previously, not all of those Independents came back to the Democrats after the first debate. Harris and Trump are at least notionally scrapping over the ones that didn’t. Trump has clearly focused on the men, and not unreasonably so; based on our work, I would say that the percentage of female voters in that segment who did not support Trump initially but could be persuaded to do so was so small that it might not exist at all. If they were voting for him, they had already said so. If not, well, the message we heard from them is that they might not vote for Harris but they damn sure weren’t voting for Trump’s sorry ass. You can see why Harris’s campaign has devoted such time and resource to winning over this segment.
The men in this group seem to present a better opportunity for Trump; less directly threatened by his agenda, personality, political movement, and entire way of being than women, they might be persuaded to come around on him. They care a lot about the economy and tend toward conservative attitudes on it; many of them believe Trump was a good steward of the economy before COVID. And they were more overtly resistant to Harris. At the risk of repeating myself: some of that was perhaps initially understandable, since they didn’t know much about Harris to start with; some of it was back to our old friends race and gender, a raucous double-act still playing to sold-out theaters across the nation.
Yes, these men - we’ve been assured in media coverage for weeks - are the key to Trump’s potential victory. The dudes see something admirable, even aspirational, in him, and will vote for him, another researcher who I will not name informs us, because their bros are doing so.
My problem with this entire line of thinking is that there’s basically no evidence of it in the work we’ve done this cycle, including a deep piece of research we just wrapped with men, especially Indy and conservative men, in Pennsylvania.
Oh, parts of it are true - Independent men are preoccupied with the economy and they do have conservative tendencies on the subject. And a lot of them really do believe that Trump might be the better choice on this issue they really care about. They also despise the man, personally, turned off by his narcissism and impulsivity, both of which affront some of their core beliefs about how a man should behave, what a man should be (more on that in a future missive, perhaps).
It would be hard to overstate how much Trump’s personal style repels this segment. Even ads aimed directly at them - on the economy, on social issues - leave them cold or actively anger them because those pieces of creative, mirroring their candidate, are bombastic, exaggerated, and cruel. Also, not for nothing, these guys do not especially care for women being denied the right to abortion; we found a surprising well of resistance and resentment toward the GOP on this, including from Republican men.
This is not to say that Trump has no prayer with Independent men; some will come over to him, for the expected reasons - perhaps even a small majority. But fighting Harris tooth-and-claw to eek out a technical victory with this segment isn’t going to be good enough to overcome Harris’s staggering advantage with women; he needs to run up the score with these guys and instead we’re finding hardened dislike of him and a surprisingly robust openness to her. The short explanation for the latter of those, incidentally, is that some of these Indy guys appear to have a built predictive system that allows them to vote for her on the basis of toughness and no-nonsense sound judgment - the prosecutor argument.
So why has so much media coverage focused on the idea that Indy men in swing states will deliver the election for Trump? Two errors, as I see it. The first is simply that they voted Trump in 2016 in numbers that polling did not predict, shocking the country. Believing past to be prologue is a defensible model (in fact, it’s basically a summation of how humans learn anything), but I believe it to be a mistake in this election, which is unlike any in American memory. Even leaving aside the specific peculiarities of Donald Trump, the personality, candidate, and convicted felon: the situation is that a sitting Vice President has replaced a one-term President on her party’s ticket with three months left in a race against a defeated former President now on his third straight presidential campaign. Who is the change candidate here? Whichever one you said: are you sure about that?
Using past as prologue is not just a problem with coverage of Independent men in this election; it is, I believe, the fundamental category error at work in many public polls and their coverage. We are beset by indicators suggesting that this is Harris’s race (the redoubtable Jason Stanford is very good on this subject here, although he cites me more than is perhaps wise); Congressional and even Senate races where Dems are poised to overperform expectations, and registration and early vote numbers suggesting enthusiasm is on the Dem side. And yet we are also left to puzzle over national and battleground polls from major media outlets with the race too close to call anywhere as we are assured that everything from a Trump popular vote victory (this is where CNN’s Harry Enten really lost the plot) to a landslide for Harris is on the table. This is ass-covering substituted for analysis, and it is happening because various outlets have determined that they will not, will not undercount Trump voters again, come what may.
The second error is an error in understanding how voters form their attitudes. There is an undeniable internal logic to “Indy men care about the economy and are conservative on the subject→ Trump is conservative on the economy→ Indy men will vote for Trump because he is conservative on the economy”, it just misses the fact that a ton of Indy men who care about the economy and are conservative will not vote for Trump because they absolutely loathe him. To be fair, most research instruments aren’t designed to pick up the motivation behind an attitude - the why they believe a thing behind the what they believe - very well. Polling isn’t designed to at all, and shouldn’t be asked or expected to for the same reason my firm shouldn’t be expected to produce qualitative insights at statistically significant scale - neither machine is designed for that.
Focus groups do not serve particularly well here, either. I will spare the Several Viewers of this missive a lengthy treatise on focus groups (this time), but the short and relevant element here is that focus groups, properly considered, are really exercises in group dynamics, rather than a deep exploration of a voter or a voting segment’s motivations. A feature of focus groups that I have had repeatedly relayed to me this cycle is men saying they plan to vote for Trump, in spite of his limitations, because of the economy. The natural conclusion that came from this is that men who care about the economy are planning to vote for Trump. You can see how that might happen.
Except that the order is wrong. It’s not “men who care about the economy are voting for Trump”. It’s “men who are voting for Trump say it’s because of the economy.” In fact, we do not know why the man who is voting for Trump is doing so, and we certainly don’t know enough to say what men who care about the economy will do. We have a social permission structure in which people are allowed to vote for virtually anyone “because of the economy” but a less clear one for saying you’ll vote for a candidate because you believe they’ll inflict grievous harm on your perceived enemies or because there’s just something you don’t like about the woman in the race. That doesn’t mean the Trump voter was outright lying, exactly, just that he was choosing to emphasize a socially acceptable reason for his choice because he was asked in a semi-public setting.
If your inclinations and instruments led you to believe that Men Who Care About The Economy Will Vote Trump then you’d look at the vast field of men who clearly state that they care about the economy (a supermajority of men we’ve worked with this cycle, to be clear, and we’re not alone in that experience) and think “wow, he has so many potential voters.” But our work this cycle suggests a loose - at best - association between prioritizing the economy and being inclined to vote for Trump, and that inclination is often shattered by sheer personal dislike of the man. Trump might win Independent men, but he’s not running away with them.
**
There were two groups of men Trump needs. The first were the Independents. The second are young men - GenZ, 18-29 years old. These are literally new voters to Trump, because most of them aged into the franchise since he was last elected. That is also true of GenZ women, but, ah, well, they really do not like him, for the most part. So if he’s going to get GenZ, it’ll have to be the men.
As with the Indy men, there are reasons to think Trump could attract this group. They often have conservative views on the economy and admire entrepreneurship and self-optimization. That Trump is not a successful entrepreneur and is not optimized for anything but glad-handing degenerate dentists before dismissing them to the carving station is, notionally, not a meaningful obstacle; he plays a successful businessman on TV and is certainly richer than they are, which is what counts, maybe? They are more socially conservative than their GenZ women peers - perhaps they furiously resent Woke Democrats? And, so the thinking goes, they are online and angry. Disaffected. They listen to podcasts by people who scorn both political parties, but especially the Democrats. They, we are to believe, hang out on message boards that reinforce views of women that range from dismissive to predatory. Trump is gunning for these bros in a big way, we are told.
Again, the headline is accurate. The Trump campaign has made a concerted push to get into this segments’ communication channels and win them over. Are we about to see Donald Trump swept into office by a wave of fury from America’s angriest young men?
No.
There are two reasons why this idea is the dampest of squibs. The first is that there aren’t enough of this particular stripe of disaffected young dude to meaningfully sway even the narrowest of margins. As the redoubtable pollster John Ray of YouGov told me:
“Most young guys are not 4chan-dwelling incels. A big faction of the Democratic party activist class is treating them that way and that's the problem. They think we need to be treating them like a sleeping rabid dog to tiptoe around in persuasion stuff but the fact is they're mostly winnable. They legit support Democratic policies pretty much across the board.”
Our work this cycle supports this. We did find a strong streak of small-c conservatism in many of the GenZ men we worked with, and a little less hardened resistance to Trump than we found in the older Independent men. But we also found enormous suspicion of the GOP’s social agenda and, again, a widespread personal dislike of Donald Trump within this segment. They cared a lot more about buying their first house than they did about being revenged against Woke-ism or whatever it is they’re meant to care about. This segment certainly contains its share of angry reactionary types, but not nearly as many as we’re led to believe.
The other reason the idea of the disaffected young bros delivering for Trump is a non-starter is that, well, they are disaffected young bros. It’s not that they haven’t thought about politics - they have. They in fact have a built predictive system for understanding politics and candidates and making a decision. And that predictive system always leads to the same conclusion: none of these politicians are good enough to warrant my attention, much less my vote. Their modal way of engaging with politics is to performatively withdraw from it; their very political identities are built on lofty contempt and disdain. Can you imagine a greater thrill for someone with this modality than having a billionaire* former president, who presents himself and is presented by others to you as the outsider’s outsider, the Chaos Agent supreme, your political deliverance and validation made flesh, come out to dance for you and getting to watch him sweat and flop his way through his routine and go “nah”?
The guy has gone fishing in a pool with a tiny number of fish whose entire identity is based on the premise that We Don’t Bite, Ever. From this, we are asked to believe he will feed his political multitude.
**
There is a strain of magical thinking in much of the coverage of Trump, a tendency to assert that he is surging with or has potentially-decisive latent support among one segment or another, without evidence or explanation. I can only attribute this to the shock of 2016, which was reinforced beyond all due measure in 2020 when it briefly appeared that the same thing might happen (it didn’t; polls underestimated Trump but were closer to the mark, and he lost, as predicted). The idea appears to have taken root within much of the political media and the Democratic political class that the man can summon voters out of nowhere, or is the Pied Piper of Latinos or gamers or whoever it is this time.
But in fact Trump is, or was, frantically hunting for votes amongst a small universe of voters who either don’t like him personally or don’t like politics at all, a fraught endeavor at best. That’s all that’s been happening since Biden stepped down. The Trump campaign was optimized to beat him; in his absence, it has sputtered.
I say “was” because Trump is appearing in my home state of New Mexico today, a state where he will be beaten like a drum (dude also owes Albuquerque almost a quarter of a million dollars from his last visit and I say we send guys to collect). He was, infamously, in New York this weekend, which he has never and will never come close to winning; he has reportedly canceled an appearance at Ohio State v. Penn State in order to go to Virginia, where he was soundly beaten in 2016 and drubbed in 2020. This is not strategy. This is a perverse Red Army Doctrine, in which the battlefield must always expand because to fight to hold threatened territory would be to admit the possibility that the revolution might fail.
This is not because Americans are preparing to reject conservatism, or the Republican Party in general, or to put an end to the gruesome politics that Trump has brought into the mainstream. But they are preparing to reject the man himself.
There have always been two disaster scenarios for the Democratic Party in this election. The first is a loss to Trump, the human consequences of which would be unspeakable. The second is a victory on a scale sufficient to make the party professional class and its leadership believe that the threat has passed, that order has been restored, that the inevitable march of progress is back on after a slight technical hiccup, and no root-and-branch analysis of the party’s relationship with the American people is required, thank you very much.
I now believe the second of these scenarios is the real danger.
The politics of cruelty that Trump awoke will be with us for a while yet; it will not be denied, defeated, and driven from American politics next week.
But he will be.
See you on the other side.