Ending Discrimination by Twitter

On March 25th, Elon Musk wrote “Given that Twitter serves as the de facto public town square, failing to adhere to free speech principles fundamentally undermines democracy. What should be done?” He was quote-tweeting himself from a couple of days earlier, where he had polled followers, stating “Free speech is essential to a functioning democracy” and asking “Do you believe Twitter rigorously adheres to this principle?” Of over two million voters, 70.4 percent answered “No.” In answer to his later “What should be done,” a lively debate proceeded both in the replies and in opinion pieces across the popular media.

In one such opinion piece, appearing recently here in Quillette, Angel Eduardo disagreed with Musk’s characterization of Twitter. In “Twitter is Not the Town Square,” Eduardo declared that “No matter what we think, want, or feel, Twitter is not, has never been, and likely can never be a public square.” His argument was that because Twitter is a private company, it “can have whatever rules it likes—and it can apply them with whatever level of consistency it likes.” A social media company is a business, he says, and businesses don’t come to have obligations simply because of how we use them or what we’ve decided they’re for.

Twitter is Not the Town Square

Refashioning the private company as a legitimate town square would require more change than we realize.

QuilletteAngel Eduardo

Eduardo makes an analogy to Starbucks to illustrate this point: many people use Starbucks as a co-working space, and we can imagine someone declaring that in virtue of this, Starbucks fulfills an important public purpose and so has certain obligations in that regard. Eduardo rejects this claim, saying “Starbucks can turn every one of its locations into a walk-up window without any seating or wifi, and there’d be nothing wrong with that in principle—and there’d be nothing we can say about it with any moral or political authority. You’d just have to write your screenplay or hold your startup meetings somewhere else.”

Sam Harris made a similar argument a few months ago on Triggernometry, asking the audience to imagine that his own podcast had become so big and successful that suddenly the government forced him to have Alex Jones or Donald Trump on as guests. He asks “shouldn’t I be able to have anyone I want on my podcast?” He goes on to argue that social media platforms should be able to exclude any group they like, and if people don’t like what they’re doing, they should boycott. He gives the examples of a social media network just for beautiful people, or just for “guys who are over 6 feet 2, and blonde hair and blue eyes.” He qualifies “I wouldn’t have said this in 1964 when we’re … when we have to pass the Civil Rights Act.” But “at this point in history,” Harris thinks, this is how things should go.

Eduardo notes that if Twitter were to become a “legitimate public square,” then Musk “wouldn’t be the one making these decisions.” Private companies have “terms of service … at the whim of whomever owns [them],” while a legitimate public square would not have.

But is it true that private companies can have terms of service at the whim of whoever owns them; that because Twitter is a private company it can have whatever rules it likes? Legally and morally, this is false, and the reason why is hinted at in Harris’s reference to the United States in 1964—although I will talk instead in terms of Australia and the UK.

This question is personal for me, for the reason that Twitter’s having “whatever rules it likes” has profoundly affected me. I am a gender-critical feminist. A gender-critical feminist believes that women are a sex caste: that it is femaleness that women have in common, and historical oppression on the basis of femaleness that explains much of women’s continuing inequality with men. Feminism, according to a gender-critical feminist, is by, for, and about females. This puts gender-critical feminists at odds with mainstream feminists and other progressives, many of whom think of feminism in terms of an expansive concept of gender, and where it is important to include in gender categories people who want to be included (transwomen) and exclude people who want to be excluded (transmen). You might imagine that this is simply a debate to be hashed out on the pages of left-wing blogs and in academic journals and in lengthy social media threads; not something that a social media platform with nearly 400 million users would pick a side on and use its policy to enforce. And yet that appears to be exactly what has happened.

I have been permanently banned from Twitter since June 2019, most likely for “misgendering” under their Hateful Conduct policy, which includes “targeted misgendering or deadnaming of transgender individuals” under the subheading “Repeated and/or non-consensual slurs, epithets, racist and sexist tropes, or other content that degrades someone.” I once referred to a rather unpleasant trans activist who had been harassing my employer about my political views using the pronouns “him” and “he,” and once asked a trans activist about their sex and asserted that transwomen are not female. Because I am banned from Twitter, I cannot participate in discussion with the (remaining) gender-critical Twitter community, and that matters because gender-critical feminism has become my primary area of research. I cannot share my work or the work of other gender-critical people who I support. Perhaps most frustratingly, I cannot correct the record when I am vilified and misrepresented. In 2021, while a campaign to have me disciplined by my employer was being waged from Twitter, I was unable to defend myself.

If Eduardo and Harris are right, then there is nothing wrong with my being banned from Twitter (along with many other gender-critical people, including British feminist campaigner Kellie-Jay Keen Minshull, Irish comedy writer Graham Linehan, and until very recently, Canadian feminist writer Meghan Murphy). If Eduardo and Harris are right, there would also presumably be nothing wrong with the fact that I am permanently banned—for similar reasons—from the essay-writing platform Medium, and from the allegedly lesbian-friendly dating apps HER and Hinge.

Eduardo asked us to imagine Starbucks being forced to serve the purpose we have given it as a co-working space, Harris asked us to imagine the government forcing him to interview on his podcast guests who he despises. Both were so concerned with keeping the public or the government out of private companies that they overlooked the other side of the issue, which is that private companies cannot legally (and should not morally) exclude customers, clients, or users from protected categories. Equalities law ensures non-discrimination in the provision of public goods and services. In Victoria, Australia, a woman who is breastfeeding in the public library cannot be asked to leave because another user finds it inappropriate, and that’s because breastfeeding is a protected attribute under the Victorian Equal Opportunity Act 2010. A tradesperson cannot refuse your job because he disapproves of your religion, or despises your race, because religious belief or activity, and race, are protected attributes under the same law.

Eduardo and Harris might at this point respond that when they described companies having broad discretion over who they exclude, they didn’t mean involuntary features of persons like race and sex. Harris cannot very well make this claim, given that beauty, height, eye colour, and hair colour all fall into that category; but Eduardo can. However, it’s not just involuntary features of persons that make it to the list of protected attributes. In Victoria there is also employment activity, industrial activity, political belief or activity, religious belief or activity, and personal association with anyone who has one of the other protected attributes.

All of the following would then violate equalities law: someone wearing a “Restore Democracy Sack Dan Andrews” minor party t-shirt being denied service at a cafe by a barista who supports the Labor Party; a public servant working at Centrelink (Australia’s social services bureaucracy) refusing to help someone access welfare payments because when making small talk she discovers that they’re a Scientologist; a tradesperson walking out on a half-completed job because he finds out the client is a defence lawyer. In the interests of a mutually tolerant society, we do not permit people to discriminate in the provision of goods and services, or in employment decisions, on these grounds.

Similarly, in the UK, “philosophical belief” is a protected attribute under the Equality Act 2010. A company that terminated an employment relationship with a woman because of her gender-critical beliefs was found to have discriminated against her.

The question, then, is not whether Starbucks must maintain—and perhaps even strengthen—its role as a co-working space. It is rather whether Starbucks may introduce any rules it likes about who is permitted to co-work in its space, about who it will serve and who may be thrown out upon discovery of some disliked belief or attribute. Starbucks may not throw you out because you are Catholic, or because you are a Republican. No Melbourne cafe can throw you out because you don’t think there should be a Voice to Parliament (constitutional recognition of Indigenous Australians).

Once we think about things this way, it becomes clear that Twitter cannot do anything it likes, and yet it has been doing exactly that. Twitter has been discriminating on the basis of political and philosophical beliefs, obscuring its censorship of those who do not share its far-left views by declaring those views “hateful conduct.” (Musk described this conduct as a “thumb on the scale” and said there would be no more of it).

Whether Twitter is or isn’t a digital town square is less important than whether Twitter, in virtue of being a private company, may do whatever it likes. Eduardo and Harris say it may. They are wrong.

In my view, Musk is correct when he describes Twitter as the “de facto public town square,” and says that “failing to adhere to free speech principles fundamentally undermines democracy.” The “de facto” part is important. Twitter has become the place where people from all around the world go to discuss important issues, both domestic and international. It’s not that because people use it for this purpose, the platform suddenly has obligations to maintain, and perhaps even strengthen, that purpose. If Musk decides to shut Twitter down, of course he may do so. What is important is that Twitter not discriminate in the provision of its services. (And the same can be said for Medium, HER, and Hinge—and these are only the platforms that I have been personally censored by, there are other platforms that have censored other people). Twitter’s service is to offer people a profile from which to tweet and interact with other users. It must offer this service in a way that upholds equalities law.

The challenge Musk faces, going forward, is how to correct for the political partisanship of the previous regime without either overcorrecting (and simply becoming partisan in the other direction) or under-correcting (and leaving in place the imprecise rules that allowed moderators so much discretion over downranking, required deletion of tweets, ineligibility for amplification, temporary suspensions, permanent bans, and other penalties permitted by Twitter policy). To avoid overcorrection it must ensure that it is sanctioning far-left accounts only for conduct that a person of any politics would be sanctioned for, for example when the account is inciting violence or breaking the law in another way. To avoid under-correction it must not talk in general terms about “hateful conduct,” and “hateful imagery and display names,” without giving an explicit and reasonable definition, and examples, of what it means by that. In a country as politically polarised as the United States, it may simply be too much to hope that moderators separate their personal views from their professional role. But the more explicit the rules are, the less discretion politically-motivated moderators can have to abuse their roles to advance their own politics.