For much of 2023 the owners of Substack have explained that their corporate position on content moderation is, well, largely not to moderate it. Their position is that the cure for bad speech is more and better speech. As Substack co-founder Hamish McKenzie put it recently, “we believe that supporting individual rights and civil liberties while subjecting ideas to open discourse is the best way to strip bad ideas of their power.”
McKenzie’s formulation sounds compelling, and there is some logic to it. While Substack has attempted a social media component through its Notes feature, the site functions more like a publishing platform than a interactive social media site where their algorithms determine the content. Ken “Popehat” White is correct when he writes, “On Twitter, Nazis were constantly in my face, I had to painstakingly block them one by one, and the interface recommended that I follow them. Here I generally only encounter them if I look for them or, very occasionally, if one wanders into my comments for me to block. You can publish here and comment here and never encounter Nazis stuff here.”
Furthermore, if the marketplace of ideas is functioning well, then one could argue that the best way for toxic ideas to die is exposure to criticism — i.e., better ideas. No one is censored, the best ideas win out — it’s a perfect equilibrium!
Upon further reflection, however, McKenzie’s statement is a little weird. For one thing, his articulation situates Substack as akin to the federal government, in that even if users articulate blatantly offensive messages, they have no choice but to let them publish. But Substack is a private, for-profit corporation and not a government constrained by the First Amendment. There are plenty of platforms out there if someone is shut out of Substack — some of them pretty dedicated to the hosting of extremist content.
Another problem is that McKenzie’s model of how the marketplace of ideas functions is… not how the marketplace of ideas actually functions. As I explained in The Ideas Industry, sites like Substack have lowered the barriers to entry into the public sphere but they have also raised the barriers to exit. In other words, it is much easier for stupid ideas to enter the marketplace of ideas and much harder for them to die. There are ways in which this dynamic has existed since the days of John Stuart Mill, but 21st century forces have made the problem much worse. McKenzie is trying to persuade people that Substack exists in a world without long-term intellectual tradeoffs and that just ain’t so.
Finally, while one can try to defend a “no content moderation” policy, that is not in point of fact Substack’s actual policy. There clearly are areas — pornography, doxxing, calls for incitements to violence — where Substack moderates content. And as I noted earlier this month:
I don’t think the official Substack opinion makes a ton of sense. The moment one acknowledges that content moderation is necessary, the debate then becomes where to draw the line. Maybe, maybe their position would be more defensible if McKenzie and [Chris] Best were not the same guys who decided it would be a good idea to promote and publicize the likes of Richard Hanania. Their choices keep piling up and they keep leaning in one particular, odious direction.
This is why a few weeks ago about 200 Substackers, myself included, simultaneously posted a query to Substack:
In the past you have defended your decision to platform bigotry by saying you “make decisions based on principles not PR” and “will stick to our hands-off approach to content moderation.” But there’s a difference between a hands-off approach and putting your thumb on the scale. We know you moderate some content, including spam sites and newsletters written by sex workers. Why do you choose to promote and allow the monetization of sites that traffic in white nationalism?
That query attracted enough attention for Hamish McKenzie to post a response, which I quoted from above. But this is the part that jumped out at me:
There also remains a criticism that Substack is promoting these fringe voices. This criticism appears to stem from my decision to host Richard Hanania, who was later outed as having once published extreme and racist views, on my podcast, The Active Voice. I didn’t know of those past writings at the time, and Hanania went on to disavow those views. While it has been uncomfortable and I probably would have done things differently with all the information in front of me, I ultimately don’t regret having him on the podcast. I think it’s important to engage with and understand a range of views even if—especially if—you disagree with them. Hanania is an influential voice for some in U.S. politics—his recent book, for instance, was published by HarperCollins—and there is value in knowing his arguments. The same applies to all other guests I have hosted on The Active Voice, including Hanania’s political opposites.
It should be acknowledged that the archive of McKenzie’s podcast suggests Hanania was an outlier guest. That does not change the fact that McKenzie is peddling a copious amount of bullshit in that paragraph. As Ken White noted:1
McKenzie has a choice of whom to invite on his podcast. McKenzie could have invited David Duke or Nick Fuentes, who are also prolific contributors to the public discourse, but selected Richard Hanania. Presumably McKenzie thinks that Hanania is a plausible, within-the-Overton-window voice and the others aren’t. That’s a value judgment that Hanania’s racism is inside-the-circle, and it’s one that promotes Hanania. Moreover, McKenzie’s choice to accept Hanania’s deeply dubious confessional to “reforming” also reflects a value judgment — a judgment in favor of slack-jawed credulity, in my view….
McKenzie’s apologia deeply annoys me because it treats me like I’m a moron. It’s the equivalent of yelling over the wall of my walled garden “don’t worry, those guys three gardens over really just like Hugo Boss, and also they have some points on tax policy.” There’s a difference between the ethos of “we’re a platform that’s decided not to make value judgments about offensive speech, if that’s okay with you, you’re welcome” and the ethos of “we’re a victim of cultural Marxism and we see that a lot of these guys are not that bad and we’re doing a service to humanity by platforming them and listen to their guest spot on my podcast.”
Radley Balko makes similar points:
Once you ban consensual porn, nudity, or sex work while still allowing Nazism, it isn’t unfair to conclude that you find porn, nudity, and sex work more offensive than Nazism. And it’s perfectly reasonable to conclude that those are some pretty fucked up priorities.
Personally, I found the Substack brass’s decision to not just host, but actively promote and recommend a mainstream bigot like Richard Hanania more disappointing than merely hosting the racist stuff. And Substack’s explanation for that decision — the laughable, easily disprovable claim that Hanania’s bigotry was a youthful indiscretion that he has since denounced — wasn’t just hopelessly naive, it was contemptuous of anyone they thought would buy it.
The notion that Hanania had abandoned his racist worldview is a risible claim, and it is difficult not to conclude that McKenzie would not have done anything differently.
So this leaves the hard-working staff here at Drezner’s World at something of a crossroads. As Margaret Atwood points out, Substack provides an “agreeable, well-designed platform,” which is why leaving it would be extremely annoying. I’m also fully aware that content moderation is tricky; as White points out, it can’t just be a Potter Stewart know-it-when-you-see-it kind of thing. Questions about content moderation create difficult tradeoffs for McKenzie and his co-owners.
Substack’s position of pretending those tradeoffs do not exist, however, is insulting and infuriating. As Atwood put it:
You can’t have both the terms of service you have spelled out and a bunch of individual publishers who violate those terms of service. One or the other has got to go, and hiding under the sofa and pretending it isn’t happening will not make your dilemma go away. Nor will some laudable rhetoric about free speech – not when you yourselves have clearly stated that not everything is allowable, including threats of “violence” and “physical harm” to “protected classes.”
So the question is whether exit or voice is the best option given Substack’s official position. This depends on whether Substack’s owners demonstrate more intellectual sophistication than they have to date. It also depends on whether there is another platform that is: a) akin to Substack; and b) less friendly to fascists.
This is a question I need to ponder going forward. I am curious to hear what my paid subscribers think, however, which is why I’m limiting the comments in this column to them. What do you think?
White’s post is worth reading in its entirety. Almost everything I’m saying here he said more cogently last week.
Like you, I’m incredibly annoyed and insulted. For one, I’m annoyed at how the public discussion about this has treated to Substack or not like it was a cost less, frictionless choice; that’s not the case. For another, I’m now facing reputational damages because of the Substack team’s cack-handed policies and explanations—thanks, guys! And as Balko (and I in my own post) noted, this isn’t a Substack or not question—all platforms will face the same questions unless you handle your own tech backend, which is what I came here to avoid! So I am in a wait-and-see mode but keeping enough public discontent that I hope it will pressure the team to do minimal anti Nazi stuff.
I disagree fairly strongly with both DD’s post and most of the comments. A few reasons why:
First off, I think the hypocrisy argument raised by the prohibition against porn is a canard. Even setting aside the fact that Substack's payments processor, Stripe, won't process payments related to porn (a policy that is almost certainly about economics rather than morals; the porn industry has huge numbers of fraud and chargeback issues), hosting porn is also a huge burden because of the cost of developing systems and hiring personnel necessary to guard against hosting child porn, porn that violates copyright, or porn that has been posted without the approval of the participants. Long story short, Substack isn't making an editorial judgment that porn is worse than Nazis. They just don't want to be in the porn hosting business, which is really a completely different business than hosting a blogging platform. The prohibition against spam newsletters is similar.
I suppose that their ban against posting content that “incites violence” or that includes “credible threats of physical harm” comes closer to having a “content moderation policy” in the way that DD then thinks makes it hypocritical to allow Nazi sympathizers to use the platform, but I disagree. Rather than making a policy that prohibits classes of people and ideas as beyond the pale, they instead have rules against specific types of content that apply to everyone.
I think that is the right approach for two reasons:
First, it says clearly right from the beginning that they are not going to be drawn into the game of deciding what ideas are legitimate and what ideas are not but rather will let readers make their own judgements. The former is a path that leads to endless arguments. One can already see the signs of it in these comments with the objections to “transphobes” and “Hamas apologists.” I think it’s wise of Substack not to get drawn into that game, and there is no clearer way of making that point then not to ban Nazis (who everyone, doubtless including the folks at Substack, would agree are the worst of the worst).
Second, I believe the whole deplatforming approach is counterproductive. I think in the long run forcing people who believe in bad ideas to other platforms doesn’t actually change their minds or stop their ideas from spreading, but rather makes people more likely to hold these ideas even more strongly, both because of the inevitable human reaction when people under siege and because they are less likely to be exposed to ideas that could influence their thinking in more positive ways. I think the fact that support for Trump has only become stronger in the three years since he was kicked off Facebook and Twitter is the clearest evidence of that.
And while I could perhaps be convinced that this second belief is wrong, that the marketplace of ideas is broken to use DD’s formulation in ways that make deplatforming the right strategy, he doesn’t ever (at least in this piece) even try to make the case that this is true; he simply asserts it. And especially given how lazy this piece is (with spurious accusations of “spouting BS” and “treating me like a moron” where the case is never really made), I found that it strengthened my belief that the Substack folks have it right.
Also, the fact that this is their policy should be a surprise to no one. They described their philosophy toward content moderation three years ago (https://on.substack.com/p/substacks-view-of-content-moderation). I thought they had it right then, and I think events since have only strengthened that case.
Also worth reading, if folks hadn’t, is Ben Dreyfuss’ take (https://www.calmdownben.com/p/substack-doesnt-have-a-nazi-problem), as well the counter letter that has been circulating (https://www.elysian.press/p/substack-writers-for-community-moderation). In particular, that latter piece speaks for me.