The Supreme Court's Reasoning Prohibits the Deplatforming Limits that the Parties Care About

As the Moody v. NetChoice majority noted, the parties focused on Facebook's and YouTube's main feeds. The majority similarly focused on those main feeds in its extensive discussion of First Amendment principles, and, as co-blogger Ilya notes, it left little doubt about the unconstitutionality of the Texas and Florida statutes as applied to them. NetChoice didn't invalidate the laws on their face because of uncertainty about how the statutes apply to other services that were barely discussed (if at all) in the briefing below.

The article that co-blogger Eugene links to in his 12:19pm post today made probably the strongest argument in favor of the Texas and Florida laws at issue in NetChoice (and both states' lawyers cited Eugene's article in their oral arguments): that platforms can be treated like common carriers such that the regulation of the hosting of users does not implicate the First Amendment. The Alito opinion (concurring in the decision to remand but rejecting the majority's application of the First Amendment to the statutes as issue) largely agreed with Eugene's reasoning, but the majority did not.

Eugene says that "the majority did not decide whether the First Amendment extends to platforms' many other functions—such as platforms' decisions whether to 'deplatform' users in a way that keeps readers from seeing the user's posts even when they deliberately seek out those posts," and I think that's correct for platforms other than the curated ones like Facebook's and YouTube's main feeds. I could easily imagine that the Court might find that some platform like Gmail cannot deplatform anyone based on their viewpoint, on the theory that Gmail does not engage in expression for First Amendment purposes. But Texas and Florida did not enact these statutes to quash Gmail's exclusion of users based on their viewpoint.

Is there any argument that under NetChoice regulating the exclusion (i.e., deplatforming) of users from the kinds of main feeds that the majority focused on does not implicate the First Amendment? I don't think so. The passages that Ilya quotes (and many other passages like it in the opinion) are very clear about the application of the First Amendment to these platforms. Beyond that, the logic of applying the First Amendment to the exclusion of messages (which the majority clearly does) also applies to the exclusion of users. A platform can exclude pro-KKK messages as part of its own communication that it abhors those messages. Similarly, excluding the KKK as a group, or the Grand Wizard of the KKK as a person, such that their posts cannot be seen by other users, would communicate that it wants nothing to do with the KKK. As the majority stated,

Suppose, for example, that the newspaper in Tornillo had granted a right of reply to all but one candidate. It would have made no difference; the Florida [right-of-reply] statute still could not have altered the paper's policy.

As the majority noted, the platforms rely on algorithms to implement their exclusion standards. A decision to implement an algorithm excluding all messages relating to the KKK (or the Kentucky Derby) would be covered by the First Amendment. And if one of the ways they chose to implement that exclusion was to exclude all messages from the Grand Wizard of the KKK, or the Kentucky Derby organization, they are still making an editorial decision that constitutes speech for First Amendment purposes. We might object that their decision excludes more speech than we would like (what's wrong with the Kentucky Derby?), just as we might object to the decisions made by the parade organizers in Hurley or the newspaper in Tornillo. But that doesn't change the fact that they are communicating by doing so.

Anyway, I'm guessing Eugene and I will discuss this at the upcoming online symposium on these cases. See below to sign up for what passes for fireworks among law professors.

Comments

[Mitch]

The speech people care about is their own, on their own wall/page, not the "what's trending" stuff that few people read. To put this in PruneYard terms, a mall can create its own daily schedule. But if one of the mall's tenants puts up a sign on her wall (say, a vegan restaurant urging people to avoid meat), that's the tenant's speech, not the mall's. Shurtleff held that although the flagpoles belonged to the City, the flags belonged to the private groups.

[David Nieporent]

the “what’s trending” stuff that few people read.

Then why do the companies offer this?

[MollyGodiva]

At the state level yes, but at the federal level the President can order social media companies to do whatever s/he wants, under penalty of imprisonment, fines, or dissolving the entire company. How so? Because this is an "official act" and the President is immune from all laws.

[Mitch]

A president would not have authority to do that. Immunity simply means he could not be punished for such an order, not that social media platforms would need to comply.

[gVOR08]

And if said prez ordered DOJ and/or the IRS to harass the company, could they not be forced to comply? Would he not still be immune?

[ReaderY]

Disagree. I think the majority opinion limits the platforms’ First Amendment rights to derivative work that they themselves create from their users’ original work – “curated content” like Facebook’s news feed. But it does not cover users’ own original work, which is the users’ speech, not the platforms.

So Facebook can exclude an entire user, not just individual posts, from appearing in its news feed. But it can, consistent with the First Amendment, be prohibited from deleting users and their content entirely. Interested listeners need not be required to pay attention to the platforms’ derivative works, like the news feed. They can instead choose to search for their preferred content on their own, in effect curating their own content.

I see the majority opinion as accepting the idea that social media platforms operate a publisher on top of a common carrier. The common carrier can be prohibited from excluding disfavored viewpoints. And users can choose to use only the common carrier functions of the platform and completely ignore the publisher functions. While platforms can have the right to exclude users from the publisher functions, they do not have such a right with respect to common carrier functions.

While the exact boundary between common carrier and publisher functions was left for another day, the opinion’s very lack of clarity on this point permits a narrow conception of the publisher functions. For example, to pick on Facebook again, if user group administrators’ curate their groups’ own content, I think under the opinion platforms can be prohibited from interfering with the groups. The speech is the groups’ not the platforms. This is so even though the platform can exclude them from its news feed.

In effect, the news feed is just a user ofbthe common carrier, one common carrier user among many. If a platform wants users to pay attention to the platform’s news feed and other published content rather than using its underlying common carrier functionality to create their own, it can be made to compete for users’ attention by having to offer something users want to see, just like everybody else.

If the town printer also runs a newspaper, it can of course exclude whomever it wants from its newspaper. But it doesn’t follow that it can also exclude anyone it wants from its printing press. It can be forced to print, as a printer and not as a publisher, whatever content customers want printed, including competing newspapers. The situation here is similar.

[Brett Bellmore]

"And users can choose to use only the common carrier functions of the platform and completely ignore the publisher functions."

Meanwhile, back in the real world, the platforms move Heaven and Earth to make it impossible for the users to do that. Just TRY excluding promoted content from your FB feed. There are browser extensions that try to help you do that, FB routinely alters their code to break them.

[ReaderY]

How exactly does FB make users read the FB feed? They can just ignore the feed and use the search function to look for what they want. As I read the opinion, it potentially permits FB to exclude users from the feed, but not the search function. Moreover, again as I read it, states can potentially require FB and other large social media platforms to maintain a search function in addition to the feed.

This means that FB could constitutionally be prohibbitted from “deplatforming” users. Although it is a publisher with respect to the content it itself curates and present, under an interpretation of the publisher side of te boundary that is somewhat narrow but doesn’t contradict anything said in the opinion, it is merely a common carrier with respect to the original raw content that users themselves post, and is a publisher only with respect derived works that the platform itself derives from that content.

[ReaderY]

Again using FB as an example, user could simply go into a user group curated by someone whose editing philosophy the user prefers. As I read the opinion, user-curated content is not Facebook’s and Facebook can be prohibited from censoring that.

In a sense, the FB feed can be regarded as simply a user group that all FB members automatically become members of when they join FB and which is curated and therefore published by FB. But if users don’t like it, they can simply join and focus their time on some other group curated by a user. FB the curating publisher is simply another user of FB the common carrier underlying technology platform. I think states could even say that FB the common carrier technology platform can’t unduly discriminate in favor of FB the curating publisher but has to let other would-be publishers use its technology platform.

It think it could still set things up so that its feed automatically and prominently appears on the screen. But users could still ignore it and look at something else if they want.

[Brett Bellmore]

I used to use an extension called FB Purity, that allowed you to control what showed up on your screen. Hardly a week went by that FB didn't change something in the code to kill it.

I'm telling you, FB does NOT let you do that. No matter what settings you use, they insert into your feed content you didn't ask for. It's their business model, selling your eyeballs to other people!

Now, perhaps you can envision future rulings that would force FB to cut it out, and let users curate their own experience, the way they could when it was in its growth phase. But that isn't today's FB.

[ReaderY]

Look. If FedEx sent you literature along with delivering your packages, you couldn’t make it stop either. But you don’t have to read it. And just because FedEx also produces literature of its own doesn’t give it a First Amendment right to not deliver your packages just because you write and want to send something it disagrees with.

Same here. You can’t stop FB from sending you its feed. But you don’t have to read it. And while FB can exclude you from its feed, it can’t deplatform you entirely as the OP claims. Further, while its ability to ban you from its own feed definitely will definitely hurt your chances of getting noticed, it can’t stop you from posting content or delivering it by common carrier as distinct from publishing means and functions. Extrapolating from examples given in the opinion, I think common carrier functions could include the messaging functions (you can PM content), the search functions (users who want to can search for your content themselves), and user groups (user group administrators who curste content are the publishers, and not FB, so they post whatever content they want to their groups.)

That’s a lot of alternative ways to get your message out without using FB’s publishing functions and hence without implicating its rights as a publisher. And there might be others as well.

[Josh R]

After discussing Facebook's News Feed and YouTube's homepage use of prioritization and labeling, Kagan says:

But sometimes, the platforms decide, providing more information is not enough; instead, removing a post is the right course. [...] The platforms thus unabashedly control the content that will appear to users, exercising authority to remove, label or demote messages they disfavor.

[ReaderY]

But not long after that very passage, Kagan says

“Like the editors, cable operators, and parade organizers this Court has previously considered, the major social media platforms are in the business, WHEN CURATING THEIR FEEDS, of combining “multifarious voices” to create a distinctive expressive offering.” [emphasis added.]

So when they aren’t curating their feeds, they are in a different business. Kagan says exactly this! I’m just running with it and taking it to its obvious logical conclusion.

FB in particular has a variety of other ways of communicating besides its feed. It has PM features, private groups that can be accessed outside the news feed, and a search feature that lets users search posts themselves to find them.

When activities outside the feed occur, Facebook is in a different business. It is not “combining ‘multifarious voices’ to create a distinctive expressive offering.” It is doing something else entirely. It is acting as a simple common carrier, nothing more.

The decision, narrowly but fairly interpreted, holds only that the feed, and the feed only, is protected by the First Amendment. Nothing else.

My interpretation is entirely consistent with Kagan’s language.

[Josh R]

The opinion said Facebook has a First Amendment right to remove posts.

[ReaderY]

From its feed.

The opinion also said Facebook performs non-publisher functions not protected by the First Amendment.

What goes in what category was largely left for another day. I’m proposing what I think is the most expansive interpretation of the non-publisher functions that remains consistent with the opinion.

You’re entitled to argue for a different boundary and narrower interpretation of non-publishing, with more functions being considered publishing. The question was left open. The boundary remains to be seen.

The opinion nonetheless recognizes that social media platforms are providing multiple services and functions and operating multiple businesses, some of which involve publishing protected by the First Amendment and some of which do not. There is some sort of boundary between them.

[Josh R]

Are you sure the removal applies only to the feed? I'm honestly not sure.

But even assuming you are right, under this ruling, can YouTube refuse to reveal the video exists when someone searches for it? Can they do likewise when someone clicks on the poster's channel? Can YouTube make sure the only way to view it is to know the URL, and even then can they post a message before the video is played that the video is bad?

[Mitch]

This was from an article I published last week, which accords with the difference between printing and publishing,

WWBFD (What Would Ben Franklin Do?)

The platforms repeatedly cite the observation that Benjamin Franklin’s newspaper was not a “stagecoach, with seats for everyone.” MCAC v. Halleck, 587 U.S. 802, 813 (2019). Franklin explained that his subscribers paid him to provide desirable content, and it would breach that promise to provide “Libelling and Personal Abuse.” Benjamin Franklin, Apology for Printers, Pa. Gazette, June 10, 1731. But Franklin balanced the competing imperatives in a way that deserves emulation today.

Like Milton, Franklin celebrated speech’s public function. He insisted that “when Men differ in Opinion, both Sides ought equally to [be] heard by the Publick,” because “when Truth and Error have fair play, the former is always an overmatch for the latter.” He squarely rejected the kind of bottleneck imposed by Jack Dorsey and Mark Zuckerberg in recent years. “If Printers ought not to print any Thing but what they approve . . . an End would thereby be put to Free Writing, and the World would afterwards have nothing to read but what happen’d to be the Opinions of Printers.” And such suppression was not needed to protect autonomy, because it is “unreasonable to imagine Printers approve of every thing they print . . . since in the way of their Business they print such great variety of things opposite and contradictory.”

Franklin’s solution was to “separately” publish material unworthy of inclusion in his own paper. This served the public function by enabling consideration of other opinions—and his own private interest in avoiding reputational harm and legal liability.
Today’s platforms can achieve these competing goals by letting subscribers speak, providing a disclaimer, and invoking section 230 when needed.

Publishing is easier today. No longer does one need a printing press; writers can self-publish from their own desktop, which alters the balance between the competing imperatives. The Court favored autonomy in Elenis, so the website designer could not be forced to compose and express a message with which she disagreed. But a different balance should be struck if a customer leased desktop publishing software (or a Twitter page) from a company and used it herself to create speech the company disliked.

[Stephen Lathrop]

If you want to posit a legitimate comparison between you and Twitter, then you have to do what Twitter does. You do not get to posit that Twitter must do for you whatever you want them to, simply because (contrary to fact) if you were like Twitter, you could enjoy Twitter's freedom to publish at pleasure.

To publish at pleasure, you must actually be the publisher, not a mere content provider. Content providers are subject to the pleasure of their publishers. Always have been in the past. Always should be going forward. Any government intervention to change that would prove self-defeating, and mess up publishing for publishers and contributors alike.

[Stephen Lathrop]

Is there any argument that under NetChoice regulating the exclusion (i.e., deplatforming) of users from the kinds of main feeds that the majority focused on does not implicate the First Amendment?

The 1A argument is that government cannot impose content requirements on publishers. Stop demanding that, and all the contradictions go away. Facebook publishes what it pleases to publish. You publish what you please to publish. If Facebook wants to publish on your platform, government cannot force you to carry Facebook's stuff. And vice versa.

[Lee Moore]

1. The headline and the article seem to be pointing in different directions. The headline

The Supreme Court's Reasoning Prohibits the Deplatforming that the Parties Care About

leads me to believe that the platforms can be forced by law to let people chatter on their platforms, when they would prefer to deplatform them. But the article seems to be saying that it is laws forbidding deplatforming that the SC disapproves of. I assume the latter is what is intended, but if so , the headline could do with some further work.

2. A platform can exclude pro-KKK messages as part of its own communication that it abhors those messages. Similarly, excluding the KKK as a group, or the Grand Wizard of the KKK as a person, such that their posts cannot be seen by other users, would communicate that it wants nothing to do with the KKK.

Really ? Does this not also imply then that a platform can exclude not merely pro-gay or pro-black messages, but also gay folk or black folk, as part of its communication that it wants nothing to do with gay folk and black folk ?

And then we wonder if people selling a "platform" for chatterers have enhanced 1A rights as compared to people selling cakes and flowers ? Certainly the chatterers themselves have enhanced 1A rights as against the cake and flower buyers, since chattering is speech, but as to the businesses providing the services it's not obvious why those providing chattering space trump those providing cakes or flowers. Or ironmongery.

The flaw here seems to be in imagining that the platform performatively expressing its disapproval of someone else's speech by denying service, is a speech act (and so protected by 1A) while an ironmonger denying service to performatively express its disapproval of someone else's non speech action is not a speech act (and so not protected by 1A.)

But this is not logical, Captain. A performative act to express disapproval cannot be speech or not, according to whether what is disapproved of is itself speech.

Further work required.

[M L]

Good points.

[David Nieporent]

Really ? Does this not also imply then that a platform can exclude not merely pro-gay or pro-black messages, but also gay folk or black folk, as part of its communication that it wants nothing to do with gay folk and black folk ?

See BSA v. Dale.

[Lee Moore]

BSA v Dale outlined a right of “expressive association” for organisations with memberships. It did not rely on the much simpler principle that Prof Benjamin stated, which is applicable to anyone.

If it is enough – as Prof Benjamin suggests – that the service provider simply wishes to express their own disapproval of or distancing from X, by refusing service to someone or something X-related; then you should not need to be a “platform” – performative speech in the form of denial of service should be open to anyone.

The Masterpiece cake guy would have won 9-0 in twenty five seconds by simply saying – “I didn’t want to bake that cake because by refusing to do so I communicate that I want nothing to do with same sex weddings.”

But IIRC he got no more than two votes for that.

[David Nieporent]

The flaw here seems to be in imagining that the platform performatively expressing its disapproval of someone else’s speech by denying service, is a speech act (and so protected by 1A) while an ironmonger denying service to performatively express its disapproval of someone else’s non speech action is not a speech act (and so not protected by 1A.)

The "service" here is publishing their speech, not mongering iron.

[Lee Moore]

So ?

The service that is being refused is irrelevant to whether the refusal to provide it is a performative speech act.

[David Nieporent]

So the choice to distribute or not distribute speech is itself protected by the first amendment; the choice to monger iron is not.

[Lee Moore]

Assumes the conclusion.

In order for the first amendment to protect A's choice to distribute or not distribute B's speech, that choice must itself be regarded as A's speech.

To avoid begging the question, your sentence needs to be adjusted thus :

So the choice to distribute or not distribute speech is itself speech ; the choice to monger iron is not.

But stated like that it's illogical. We do not conclude that the choice to monger or not to monger iron is itself ..... iron, just because the item about which the ironmonger makes this choice is iron.

The character of the platform's action in distributing or not distributing speech - if it is intended to communicate some sort of message - may indeed be speech. But the same goes for the ironmonger's action in mongering or not mongering iron.

The general principle is fairly straightforward.

To be regarded as engaging in speech, you do not need to be speaking about speech. Speaking about iron, or fish, or your neighbor's deplorable morals is also engaging in speech.

If refusing to provide services to X is intended to communicate your opinion about X, then if the law accepts that performative acts can constitute speech, that must be so regardless of the particular services that are declined, and regardless of X, or your reasons for wishing to disassociate yourself from X.

[Josh R]

The First Amendment does not protect refusing non-speech servcies, even if the provider wants to send a message by not providing service. But, the First Amendment protects refusing to provide speech services when doing so gets in the way of the provider's message. Hence, limo drivers can't refuse gay weddings, but websites can.

Whether Facebook has a First Amendment right to deplatform depends on whether the user's messages get in the way of Faecbook's. That's not settled, but Kagan strongly implies Facebook has such a right when she concluded they have a right to remove user posts.

[Lee Moore]

I appreciate the distinction that you make in your first paragraph, but, like Nieporent's and Prof Benjamin's, it is a conclusion not a reason.

Specifically :

"the First Amendment protects refusing to provide speech services* - when doing so gets in the way of the provider’s message."

* aka X

Why is the X in italics to be treated differently from any other X - such as supplying cake, flowers or nails ? (The fact that the service in question concerns "speech" is certainly relevant to whether any rule impinges the 1A rights of the speaker who is being provided with services, but that is not what is in question here.)

Whether Facebook has a First Amendment right to deplatform depends on whether the user’s messages get in the way of Faecbook’s.

But why is the same not equally true of the cakemaker's denial of service ? If the cakemaker wishes to proclaim that God is Great, and He abhors Devilworshippers (or whatever) - requiring the cakemaker to provide cake for these Devilworshippers gets in the way of his message.

(Note that I agree you are quite correct to say that we need to be focussing of Facebook's message NOT its business - 1A is a speech-protectin' rule, not a business protectin' rule. Thus the fact that the cakemaker is not in business to proclaim the word of God, is irrelevant to the question of whether his message is being interfered with.)

Whether Facebook's message, or the cakemaker's message is being interfered with is of course a question of fact, as you say. But it's a question that has to be answered under common principles, whether you are a social media platform, a cakemaker or indeed a racist landlady.

[Lee Moore]

Let us look at the Volokh Conspiracy by way of example. What do we find ?

1. Various headers and links that can take us off to other articles
2. Advertisements
3. Articles by the learned Conspirators
4. Comments by the learned, and not so learned, commenters

Whose speech is each of these ?

1. is the speech of the VC qua VC (or maybe Reason, but either way it’s “the platform’s” speech)
2. is the speech of the people paying to plant the ads
3. is the speech of the Conspirators themselves (and possibly the VC qua VC too)
4. is the speech of the commenters

If the government required the VC to accept or deny ads (box 2) – would that “abridge” the VC’s own speech ? Accept – yes. The ads clog up space that could be used for the VCs speech. (Likewise if the government tried to control the position of ads on the page in a way that impinged on the VC’s positioning of its own speech.) Deny – no. It might be bad for VC’s business, but it wouldn’t affect the VC’s own speech.

What about the comments ? As the VC comments are hidden in a box that you have to click on to get into, it doesn’t seem to me that either the comments, or any regs the government might make concerning the content of those comments would impinge on the Conspirators or VC’s own speech. (Though it might impinge on individual Conspirators own speech to the extent that they participate in the comments section. And if as I assume they participate there, when they do, in their individual capacity, the VC’s speech is not impinged on.)

Thus government regulation of the VC comment section as currently constituted would not affect VC’s speech and thus would not abridge its 1A rights. (Though it might well abridge the commenters 1A rights.)

But Shirley ! The VC wants to have an orderly, expletive-light comment section, which readers enjoy, and which is not clogged up with trolls and other folk offering get rich schemes. It’s part of the whole “VC experience” that the comments are full of witty, learned, and good humored stuff, with never an angry word said. Sure, But that’s the VC’s business (or hobby if it is not in it for the money.) It’s not its speech.

So if the government chose to require the VC not to moderate the comments section and not to deplatform anyone, even Queen Amalthea ! , it’s hard to see how that could “abridge” the VC’s freedom of speech (or press.)

[Josh R]

I’m guessing even Eugene (who supports government’s ability to keep Facebook from censoring user speech based on viewpoint) would disagree with your analysis of The Conspiracy because he does not likely view this blog as a common carrier. Rather, I suspect he views the comments as more like letters to the editor.

[Lee Moore]

Stipulating that VC comments are in the same category as Letters to the Editor, what is the argument that 1A prohibits the government from requiring the VC to publish them ?

They’re not the VC’s speech, and they don’t impinge on the VC’s speech in the way that Letters to the Editor of a dead tree newspaper occupy space otherwise usable for the newspaper’s own speech.

[Josh R]

Of course it is a conclusion, and one you don't like. But, it is grounded in precedent and logic.

The precedent: when the service does not involve speech, Ollie's Barbeque (Katzenbach) and the Heart of Atlanta Motel do not have a First Amendment right to refuse to serve blacks because doing so would get in the way of the service provider's message that blacks are not worthy of service. In contrast when the service involves speech, there is a First Amendment right to refuse service if the provider's message is at stake (Tornillo, Pacific Gas, Turner, 303 Creative).

The logic: if you are right, then all anti-discrimination law is de facto unconstitutional. That's absurd.

[Lee Moore]

I think we are seeing the same game. The distinction rests on precedent and the precedent rest on the logic of …… not liking the consequences. ie we get there by working back from the consequences rather than forward from the text and facts. Which is not a good direction.

I agree that “all anti-discrimination law is de facto unconstitutional” is potentially a problem. Especially if you think anti-discrimination laws are a good thing. But there is another way out, other than working back from the required conclusion.

And that is by concluding that “the freedom of speech” in 1A is limited to actual speech, which is a perfectly tenable reading of the text. We can distinguish between “speech” and “communication” – presenting the middle finger, turning your back, blocking the road, refusing service to a potential customer etc may be (depending on your intent) a communication. But they are only “speech” if we metaphorically extend that concept to performative acts that are intended as communication, but do not include speech. We don’t need to conclude that declining service is speech.

[Josh R]

Then, we agree?

Facebook posts are speech. If Facebook is a publisher, it can refuse service because the government cannot force facebook to publish speech it does not want to.

The same does not apply to other services where the refusal to provide service is at most expressive conduct (assuming only for the sake of argument, it is expressive conduct).

[Lee Moore]

No I don’t think we agree.

Facebook’s decision to display your speech, or not, is not actual speech, it is only colourable as speech if performative acts are treated as metaphorical speech for 1A purposes.

Thus 1A protects Facebook’s choice not to display your speech on the metaphorical speech interpretation of 1A. Not on the actual speech interpretation.

And likewise for the cake maker. His refusal to service Devilworshippers has 1A (speech) protection on the metaphorical speech interpretation. But not on the actual speech interpretation.

So you can make the cake maker bake the cake for Devilworshippers if you adopt the actual speech interpretation. But if you do you can make Facebook display the message it doesn’t want to display.

But if you adopt the metaphorical speech interpretation, Facebook is protected by 1A from being forced to provide service. But then so is the cake maker.

You can’t chop and change your interpretative basis according to the identity of the litigant.

[Lee Moore]

I didn’t want to clog up my reply answering “Facebook posts are speech” at the trivial level. So I will do so here.

Facebook posts are indeed speech, but they are not Facebook’s own speech. So 1A does not prevent Facebook from being required by the government to display them unless Facebook’s own speech is impinged on by the requirement.

Which it could be - see above - if Facebook’s refusal to display can be seen as a performative act of metaphorical speech. Or if the display requirement impinges on Facebook’s own speech by clogging up space that Facebook would like to use for its own speech, as with a dead tree newspaper. Or if the required display could be mistaken by the reader for Facebook’s own speech rather than the user’s.

[Josh R]

But if Facebook’s message is affected by hostng the user’s speech (takes up space, mistaken as being endorsed by Facebook, or perhaps other reasons), then Facebook has a First Amendment right not to be compelled to host that speech.

That’s the crux of this debate and as I read the opinions, Kagan is likely to agree with Facebook while Alito (and Thomas) are likely to agree with the government. But as many others have noted, the issue is not settled.

Note that the above analysis does not depend on a “metaphorical interpretation” of speech that opens the door to invalidating anti-discrimination law. The service has to itself be speech and the provider's message must be affected by hosting that speech (as it was in Tornillo, Pacific Gas, Turner and 303 Creative).

[Lee Moore]

But if Facebook’s message is affected by hosting the user’s speech (takes up space, mistaken as being endorsed by Facebook, or perhaps other reasons), then Facebook has a First Amendment right not to be compelled to host that speech.

I agree. And likewise for the cakemaker. If the denial of service is not per se speech (ie if we reject the metaphorical interpretation), 1A can still apply to protect Facebook's actual speech from interference in the ways that you describe. And so with the cakemaker's actual speech.

Note that the above analysis does not depend on a “metaphorical interpretation” of speech that opens the door to invalidating anti-discrimination law.

I agree. The above analysis applies on the actual speech interpretation, but in each case you need to find some actual speech by Facebook (not its users) that is interfered with by the government obligation to display the user's message. And likewise with the cakemaker.

The service has to itself be speech and the provider’s message must be affected by hosting that speech (as it was in Tornillo, Pacific Gas, Turner and 303 Creative).

And this (the first seven words) is where we differ. The service for an online platform like Facebook or VC includes speech of the various kinds that I described above for the VC. So Facebook has 1A protection for its own speech - both its own speech that appears on its platform, and its own speech conducted otherwise (eg spokesperson appearing on TV.) Exactly the same applies to the cakemaker.

Whether a government regulation in fact interferes with Facebook's own actual speech, or the cakemaker's, are questions of fact. But the same approach must apply to each person who claims that his 1A rights are infringed by an obligation to provide service.

Including those who are not in business at all. So if you as a student society for example wish to hold a "transmen only" parade, with appropriate banners, to proclaim the message that transmen are men, and fun, and wholesome, the question of whether the government can require you to include transphobes with hateful banners in your parade, has to be considered under this same 1A analysis - ie that their presence and their banners tread on your speech - even though you are not providing any services, commercial or otherwise.

That you have a business consisting in displaying a mixture of your own speech and that of other people doesn't get you any extra 1A rights. The question remains (if you avoid the metaphorical speech approach) - did this government regulation impinge on my actual speech ? What your business happens to be, and whether your business is impinged on, is irrelevant.

[Josh R]

.

Me: The service has to itself be speech

You: And this (the first seven words) is where we differ.

OK. But, that just means you either don’t accept or understand the compelled speech doctrine as applied to hosting someone else’s speech. It’s a two-step analysis that applies in the same manner for Facebook, the cake maker and the transmen parade.

Step 1: Is the government requiring you to to host speech you don’t want to. If no, the government wins. If yes, go to Step 2.

Step 2: Does hosting the speech affect your message. If no, the government wins. If yes, you win.

Facebook and the transmen parade pass Step 1, the cake maker doesn’t. We don’t yet know if Facebook passes Step 2. The transmen parade (per Hurley) passes Step 2.

[ReaderY]

But not long after that very passage, Kagan says

“Like the editors, cable operators, and parade organizers this Court has previously considered, the major social media platforms are in the business, WHEN CURATING THEIR FEEDS, of combining “multifarious voices” to create a distinctive expressive offering.” [emphasis added.]

So when they aren’t curating their feeds, they are in a different business. kagsn says exactly this! I’m just running with it and taking it to its obvious logical conclusion.

FB in particular has a variety of other ways of communicating besides its feed. It has PM features, private groups that can be accessed outside the news feed, and a search feature that lets users search posts themselves to find them.

When activities outside the feed occur, Facebook is in a different business. It is not “combining ‘multifarious voices’ to create a distinctive expressive offering.” It is doing something else entirely. It is acting as a simple common carrier, nothing more.

The decision holds only that the feed, and the feed only, is protected by the First Amendment. Nothing else. My interpretation is entirely consistent with Kagan’s language.

[David Nieporent]

It is acting as a simple common carrier, nothing more.

It is, of course, doing no such thing, since it isn't offering services for hire and isn't holding itself out as serving all comers.

[ReaderY]

I recognize I should have spoken more carefully. The relevant dichotomy is whether the activities are protected by the First Amendment or not. I recognize that there are a variety of activities that aren’t traditional common carriers in nature, yet don’t have First Amendment protection. Presenting First Amendment protected vs. common carrier as a dichotomy results in a false dichotomy, which you legitimately pointed out. Whether or not non-feed activities are traditional common carriers, a question for another day, they aren’t protected by the First Amendment under Kagan’s opinion, as discussed above.