It often seems that the pieces of writing that one cares about least get the biggest response. My piece from two weeks ago, “Against Polling,” was written quickly; It’s what we call in this business a “take.” I didn’t want to go through each claim and argument and check it for soundness, and I have no editor or fact checker to force me to do it, so I took the license granted to opinion writers and expressed an opinion, which I thought was more or less well-informed. Because I didn’t want to have to defend the piece, I wasn’t entirely glad to see that by the afternoon, it was getting traction online. I was surprised to see it was popular with pollsters themselves, who seemed relieved by its frank diminution of the pretensions of their business. But it was also greeted with consternation, alarm, and perhaps even a little bit of hurt feelings. Critics maintain that it’s an irresponsible and even scurrilous piece of opinion-mongering that pays little attention to the underlying facts of the matter, misunderstands the basic issues, and levels careless and unfair charges at people who are working in good faith. To which I reply, fair enough.

In Vox, Eric Levitz has taken up the pen on behalf of the unfairly maligned data boys. The headline says my ideas are “dangerous,” and later on, he says they are “insidious.” Oh dear! My piece, according to Levitz, is a mix of truisms and outrageous statements that I do not support with evidence and likely cannot be supported because they are untrue. Levitz finds “[both] this piece and its reception…puzzling” and says of the excitement it’s generated, "I’m not certain why this is.” He has a psychological reading. Apparently, I’m licensing progressives to ignore reality and follow their whims, and I “[offer] an elaborate rationalization for dismissing any data one does not like.”

I don’t want to go point by point and argument by argument, and to consider all the counterfactuals and what-could-have-beens, because that sounds boring to me. I’m also largely innumerate and will get easily tripped up talking about percentiles. If you want to get more rigorous, I suggest Matt Bruenig’s takedown of “popularist” reasoning or Jay Caspian Kang’s piece from 2024 on how polling turns politics into fantasy sports.

I think Eric and I are speaking different languages. I’m not trying to optimize for election outcomes. He doesn’t get what I’m doing and admits to being puzzled by it. He dismisses the central claim of the piece as obvious: that when people respond to a poll, we don’t know what they mean, we don’t know what a poll is saying; we receive a highly abstract statement and project a coherent meaning onto it. We are interpreting it and pretending that it’s a direct, unmediated representation of reality. As an empiricist, he believes there’s a straightforward reality out there that can be captured with a poll, and I’m irresponsibly encouraging people to ignore it; I just don’t agree. I think even when done in good faith, there are a ton of unexamined assumptions in this scientistic approach to politics. The way the questions are posed in a poll is never value-neutral; it frames the world in a particular way. I don’t even think everybody is capable of forming an opinion in the way we typically think about it. We have no idea whether or not the people answering the poll understand the words that are being said to them in the way the pollster intends. Public opinion doesn’t exist out there: it is always in the process of being formed, and polls are part of that process. They are themselves manipulative. And I don’t think tinkering around the edges with survey design can solve all this. Levitz finds this frustrating. How dare I deny reality? I just deny the premise that the polls are unproblematically capturing reality; I think they are mutilating it and pretending to be objective; sometimes this is because of the inherent problems in the method, and sometimes I suspect that mutilation is, either consciously or unconsciously, because of the pecuniary and ideological interests of the mutilators.

Levitz says that I’m involved in a process of rationalization, or at least, I’m allowing it to happen. But one could easily charge that the data boys are involved in the same thing. They pretend innocence. “It’s just the numbers! Yes, we know, values are involved, but regrettably, these are just the hard facts! They disguise, as Levitz does, political discipline with the discipline of reality itself. Why is it that their “scientific” method always spits out the same results? Moderate. Moderate. Moderate. To them, that is Moses and the Prophets: Moderate. Move to the center. Nothing in my piece recommends a specific policy agenda or type of rhetorical appeal. Eric suspects I am using this to advance my social democratic agenda, which is fair because politics is always a matter of interests and power, not pure ideas. But when one inquires about the interests and power behind the data-heavy approach, then one is being unfair and irresponsible. Now, you get called to account. Now, you’re breaking party discipline. “Come, come, be reasonable.” Why?

Is it so unreasonable to ask, why is it that the data brigade and the positivists are constantly urging a move rightwards, why they happen to be the same faction that wants to mend bridges with the business world and Silicon Valley, and why they have the ear (and wallets) of the donors? Why should I grant their pretensions of embodying reason and factuality itself? That’s the very definition of ideology: a tendency that claims not to be a tendency, to be in fact, the absence of tendency and pure neutrality. There’s no such thing. The polling shit is part of an ideology that hides values in value-neutral language.

What I’m doing, for all its haphazardness, is less caught up with a particular wing of the left politics. It could very well be that a rhetorical approach will appeal to the conservative sentiments of an audience. I’m not denying the need for political pragmatism, far from it: I’m offering another set of tools. What I think the rhetorical approach makes possible, by not dividing the world into little bits of discrete reality out there, is what the empiricists think is impossible: candidates who can be both moderate and radical. Like say, F.D.R., Ronald Reagan, or Abraham Lincoln. It was Lincoln I was thinking of when I was writing the piece, and he constantly frustrated the radical wing of his party. He did not have polls to go on. He did have the language of Shakespeare and the King James Bible. Let me lay my cards on the table: There is, in my piece, also an implicit critique of certain aspects of the left, not just the center, namely the dull and alienating cant they constantly engage in and the silly shibboleths they insist everyone repeat.

I think the reason people enjoyed my piece is precisely because it was highly rhetorical. Levitz is flabbergasted that I could dare be Against Polling or dare say that it’s “garbage” or “bullshit,” and then later qualify my claims. Yes, I was exaggerating for effect, and it worked! People paid attention. Then I explained more thoroughly what I meant, although not to Eric Levitz’s satisfaction. If I wrote, “Some Slight Issues with the Positivist Approach to Politics,” it would put people to sleep. Levitz is very flattering: he writes—in my favorite part of the piece, by the way—that “I’m a brilliant writer” and on Twitter he confessed to being “jealous” of my work. (So, he does understand rhetoric a bit, he knows that I’m less likely to go apeshit if he says some nice things to me.) But this is my whole point: thank you, but it’s not that brilliant, it’s just that I’m willing to employ rhetoric. I read old things, and their language is imaginative, vigorous, and exciting, and then I emulate them. It’s a trick, a trade, anybody can learn it. Well, not anybody, but it’s easier than it looks. The reason that many pollsters were not offended is that they got what I was doing and enjoyed the cathartic or emotional part of just saying, “You know what, I’m sick of this tedious bullshit.”

Political writing and rhetoric should define a political field; it should outline an opponent, a topic, and an issue. It opens a debate and presents a vision. And my frustration with the data-driven approach resonated. They got what I meant and who I was talking about. Political discussions are for the public; they should not require a degree in statistics. Also, don’t take it so literally, Eric. It’s figurative language. Do sinister interests lurk behind the vulgar positivism of the center right? Yes, absolutely, but I think this vulgar positivism itself is the problem, the way it has made people talk about their “priors” instead of their beliefs. It inserts a dead and dulling corporate language that kills any thought or imagination. It turns people away from politics or towards the part of politics that’s exciting: populist demagogery and conspiracy-mongering, you know, fun stuff. It constantly forecloses interesting public debates about values and moral problems by pointing to data, which, by the way, they often aren’t even very transparent about. It’s a type of rhetoric without admitting to itself that it is; it gestures at reason, moderation, empiricism, but it does not wholly capture those things. I’m objecting to the whole style of Vox itself, the plodding explanations and endless logic-chopping. I’m objecting to what Erik Baker called “the aggressively anti-humanistic posture of much liberalism today, downstream from Obama-era STEM mania,” which reduces everything into a question of the correct procedures of bureaucratic administration overseen by wonks, instead of a shared process of public deliberation. I’m objecting to all these tedious tautological mantras: “Popular things are popular; Good things are good, actually.” Enough! It’s a metaphysical and aesthetic critique, almost more than a political one. Mostly, I’m just tired of it. And so, I think, whether they know it or not, is the public: It’s generating nothing but boring debates and uninspiring campaigns.