The Volokh Conspiracy
Mostly law professors | Sometimes contrarian | Often libertarian | Always independent
Christopher Yoo on Regulating Social Media Platforms as "Common Carriers"
The Univ. of Pennsylvania legal scholar makes the most thorough critique yet of this approach to justifying regulations that bar social media firms from engaging in most types of content moderation.

One of the main arguments advanced to justify the Florida and Texas social media laws challenged in NetChoice v. Paxton and Moody v. NetChoice, (cases currently before the Supreme Court) is the claim that social media firms are "common carriers." Therefore, it is argued, the states can enact laws barring them from using most types of content moderation, even if such restrictions would otherwise violate the First Amendment.
University of Pennsylvania law Prof. Christopher Yoo recently published an article that is the most thorough takedown of the common carrier theory so far. Here is the abstract:
Courts and legislatures have suggested that classifying social media as common carriers would make restrictions on their right to exclude users more constitutionally permissible under the First Amendment. A review of the relevant statutory definitions reveals that the statutes provide no support for classifying social media as common carriers. Moreover, the fact that a legislature may apply a label to a particular actor plays no significant role in the constitutional analysis. A further review of the elements of the common law definition of common carrier demonstrates that four of the purported criteria (whether the industry is affected with a public interest, whether the social media companies possess monopoly power, whether they are involved in the transportation and communication industries, and whether social media companies received compensating benefits) do not apply to social media and do not affect the application of the First Amendment. The only legitimate common law basis (whether an actor holds itself out as serving all members of the public without engaging in individualized bargaining) would again seem inapplicable to social media and have little bearing on the First Amendment. The weakness of these arguments suggests that advocates for limiting social media's freedom to decide which voices to carry are attempting to gain some vague benefit from associating their efforts with common carriage's supposed historical pedigree to avoid having to undertake the case-specific analysis demanded by the First Amendment's established principles.
I agree with almost all of Yoo's analysis. In particular, I think he is right that social media firms don't fit any of the traditional rationales for common carrier status, and that states cannot simply create such status by legislative fiat (or at least, if they do, it cannot override constitutional constraints on their regulatory authority).
I offered some related critiques of the common-carrier rationale for social media regulation here:
The standard rationale for common carrier regulation is that the the firms in question have some kind of monopoly power. A classic example is a situation where there is only one railroad available to move freight from Point A to Point B, in an era where the only alternative modes of transportation (e.g.—horse-drawn wagons) were vastly slower and less efficient. It is often argued that "Big Tech" social media have some sort of monopoly over the distribution of political information, especially online.
The reality is very much otherwise. Recent survey data compiled by the Pew Research Foundation finds that many more Americans get news by means other than social media than use the latter. For example, 68% of Americans indicated they regularly get news from media websites and apps, 68% from television, and only 53% from social media sites. Among the overwhelming majority (about 96% of the total sample) who use more than one type of media to get news, 35% preferred TV, 26% preferred news websites and apps, and only 11% said they preferred social media. The same study also found that, on average, Americans trust news from social media sources less than that from television and news websites.
What is true of news is also true of opinion and commentary about political and social issues in the news. Most TV news channels, media websites, and other similar information sources carry extensive commentary and opinion pieces. And, of course, they routinely print and broadcast statements by politicians, activists, and other public figures.
To the extent we are specifically concerned with access for conservative viewpoints, there are large right of center players in both TV media and online news and opinion. These include such major outlets as Fox News, the Wall Street Journal editorial page, the Washington Times, the New York Post, and others….
Other rationales for imposing common carrier rules on social media firms are even weaker than the monopoly theory. For example, Eugene Volokh and others cite analogies to telephone lines or mail carriers. Most people wouldn't want phone companies to bar calls by those whose ideologies they disapprove of.
But such analogies are misplaced. With rare exceptions, phone calls and letters only reach a small, specifically intended audience….. By contrast, the whole point of most political discourse on social media is the ability to reach a large audience all at once. But an information product that reaches a large audience simultaneously usually works better if it has at least some moderation rules, and other constraints that enable consumers to find the material they want, while avoiding harassment, offense, and other things that make the experience annoying, unpleasant, or simply a waste of time.
For that reason, moderation rules and content restrictions are crucial for social media, in a way that is rarely, if ever, true for phone lines or mail delivery services….
[E]ven if social media platforms sometimes adopt flawed rules, the fact remains that such rules are often a valuable part of the product they provide. And it is far better for the quality… of such rules to be determined by competition in the market than by one-size-fits-all government mandates—or by a common carrier mandate imposing a near-total ban on such rules….
Perhaps the problem is not that social media giants monopolize any audience in some economic sense, but that they have too much influence over political discourse relative to some egalitarian baseline. Why should Mark Zuckerberg's views have any more clout than those of the average American? But we can make exactly the same argument for the owners and editors of Fox News, the New York Times, and any other outlet with a large audience. They too have vastly more influence over public discourse than the average American does….
Giving government a free hand to impose common carrier restrictions on any website or media outlet that "monopolizes" a particular audience or otherwise has "too much" influence is a power that can and will be abused. Call it "common carrier creep!"
Editor's Note: We invite comments and request that they be civil and on-topic. We do not moderate or assume any responsibility for comments, which are owned by the readers who post them. Comments do not represent the views of Reason.com or Reason Foundation. We reserve the right to delete any comment for any reason at any time. Comments may only be edited within 5 minutes of posting. Report abuses.
Please
to post comments
Let’s actually make outlets pick one of or the other and make the choice count. Common carrier with (more) robust liability protections then we have now or non-common carrier and you have strict liability for anything thats posted.
Farcebook can become common carrier and still cleans up spam and straightforward stuff but gives up editorial control. In return they can’t be sued anymore for failing to control bad thought. Any viewpoint enforcement the authorities want they can handle themselves. No more suits against farcebook by people who claim farcebook facilitated something bad. Its all on LEO.
Otherwise they can keep any moderation policy they want but are 100% liable. Someone upset they were libeled by a farcebook post? They should be able to sue farcebook.
No more of this quantum state of being a common carrier but not depending on the circumstance.
I too used the "false dichotomy" when dealing with my kids when they were under about 6 years old.
Do you have a serious proposal? You know, one that could pass in Congress and get signed by a President?
You are under the mistaken impression that § 230 makes social media companies into common carriers. It does not. It simply provides immunity from suit for user-generated content. (Remember that § 230 applies to virtually every website and online app in existence; it is not something special for Facebook.)
" It does not. It simply provides immunity from suit for user-generated content."
Which they already had as a matter of existing precedent. No, as written, what it did was allow them to retain that immunity even if they engaged in certain forms of moderation.
As interpreted by the courts, it allowed them to retain that immunity even if they exercised arbitrary editorial control over user content.
But, immunity for user generated content? They'd already had that, if they just refrained from picking and choosing which user content was permitted.
I didn’t say it made them into ones legally but the fact remains that they fill the same roles as the mail and the telephone as an increasingly indispensable way people communicate in the modern era. If they want to reap the benefits the old mail and telly companies did by acting like them... they should shoulder the same responsibilities. Especially as a monopoly.
They do not fill the same roles as the mail or telephone. Nor, of course, are they monopolies.
Role is not dispositive. The starting point for analysis of common carriage is transport of someone or of something. An elevator, a Ferris wheel, and an escalator can all be used in common carriage.
"Role is not dispositive," he says, as he describes a role.
Transport is a function not a role.
I agree that the analogy between mail or phone companies vs social media is a flawed one. However, that’s because the digital age had no precedent before this.
The problem remains that most people aren’t off the grid. We aren’t Amish or living like hunter-gatherers. There are many people who wouldn’t be able to make a living, shop, or find a job if they were barred from FB, YouTube, GoFundMe, OnlyFans, Amazon, and LinkedIn for just a few examples. That is the reality we live in now. Cutting people off from these things for their political views or other reasons would lead to not just social suicide but actual suicide.
Agree. Pretending they don’t serve similar functions to traditional common carriers because they have a different operating model or don’t meet some technical definition is just silly. They serve similar functions in society to traditional common carriers. They combine virtual mail and similar with virtual public squares.
Yes, the social media companies also do some publishing , bit fhat doesn’t their entire busimess, or even most of it, a publishing business.
You can get out of liquor laws by serving a stale sandwich with the drinks and claiming to be a restaurant, not a bar. But you shouldn’t be able to get out of common carrier law that way.
Agreed. Pretending they do serve similar functions to traditional common carriers despite the fact that anyone with an IQ over 5 knows that they don't is just stupid, however.
Traditional common carriers are about private point-to-point transportation or communication. Social media companies are about the opposite.
Using Facebook, you can make a public post which mirrors your understanding of social media. You can also make posts limited to your FB friends, which is a more personal communication. You can also private message someone using FB messaging, which is exactly like text messaging over a phone network.
FB exercises editorial control over all of these functions.
The internet is public infrastructure, but the gateways and entrances are all privately controlled. How do you suggest we gain access?
How do you suggest we gain access?
Start and operate your own private access point. To do it will cost you one-millionth or less than it would cost to start a common carrier. And nobody, including government, will get in your way. Which ought to tell you something.
How does that help someone who has been barred from LinkedIn but needs it to find a job?
Or barred from YouTube yet the training required by said job is on YouTube?
Completely agree. Social media companies do everything traditionally done by common carriers, plus some additional functions. But the fact that they also do some additional functions should not let them get out of common carrier obligations for the common carrier functions they in fact perform.
1. If the whole has to be categorized one way or the other, it is how society uses them, as their users see them, that controls, not the business they’d like to be in or see themselves to be in. How they see themselves not only doesn’t control, it’s irrelevant. It just doesn’t matter if they make most of their money from publishing. It doesn’t matter if they think they’re in the publishing business. It doesn’t matter if they regard that as their business model. Society uses them primarily for common carrier functions, not the publishing added on top.
2. I think Congress could require them to split their businesses into two, with technical infrastructure and common carrier functions required to be in one business and publishing functions in another, with the publishing business merely a customer of the common carrier business. Given the very different legal regimes applicable to the two, I think Congress can rationally forbid a single business from operating both a common carrier and a publisher, or can take the lesser step of forbidding it from operationally comingling the two.
No, they do almost all stuff that has nothing to do with "traditional" common carrier functions, and may also do a small amount of stuff that is similar to that performed by some things that supposedly are common carriers. But that does not mean there are any such "obligations" anyway.
Wrong. How they "see themselves" — or rather, how they hold themselves out — is determinative.
In exactly the same way that Congress can tell the New York Times that it can't sell ads and subscriptions, but has to pick between the two, which is to say: no, not even close.
You gain access to the Internet through (typically) a cable or phone company. Facebook is not the Internet. Twitter is not the Internet. Google is not the Internet. No website or app is the Internet.
David Nieporent appears never to have encountered a mass mailing.
The traditional test for common carriage includes no requirement of privacy.
There is no "traditional test for common carriage."
Every Massachusetts Judge determines on the basis of the facts and as a matter of common law whether a service is common carriage. I have not researched to the same level of detail, but state judges in California, New York, Pennsylvania, Michigan, and Ohio follow the same procedure -- probably other states.
Primrose v. Western Union Telegraph, 154 U.S. 1, 14 S. Ct. 1098 (1894) discusses the application of state common carriage law in interstate commerce.
I like Christopher Yoo, but not for the reasons given in this link:
https://www.youtube.com/watch?v=Rqnw5IfbZOU
This is just nonsense. Saying it’s so doesn’t make it so.
Social media are subject to network effects, which makes a monopoly the economically most efficient form of organization. Particularly locally. After all, railroads are hardly a monopoly, but they are a clear common carrier because they can be a local monopoly. And it was regulation that made them interoperable. If we look at what historically has been subject to common carrier legislation, potential for monopoly is sufficient, not actual monopoly. The standard being applied here is lifted out of thin air. A lot of widely accepted common carriers would fail it.
I do have misgivings about laws that purport to control the moderation policies of social media platform.
Such a law could easily and effectively require Holocaust education discussion forums to keep content denying the Holocaust or defaming the Judenvolk.
Nevertheless, there is a long tradition of regulating telephone companies as common carriers. These laws kept telephone companies from- for example- cutting off service to stop the spread of ideas of racial equality.
This applies with equal force to social media platforms.
"Such a law could easily and effectively require Holocaust education discussion forums to keep content denying the Holocaust or defaming the Judenvolk."
And what do you think the response would be to such content?
I’ll leave it to the legal minds to work out the details, but to the layman, the obvious policy is that media companies may either choose to act as mere hosts of user content, or as publishers of that content. If they are merely hosts, then they cannot be sued, because it is not their content. If they are publishers, then they can be sued, because it is their content. In neither case is it allowed to post illegal content.
As a consumer, I would be fine with either choice, and in fact I would probably subscribe to content from both, depending. For some sources, I would want a publisher to have first vetted things. For others, I would not.
What I most emphatically don’t want is some hybrid combination of the two, which it seems we have at present.
How do you feel about bookstores? Do you think that Barnes & Noble should be required to choose between selling all books or being liable for what's in each book?
A bookstore is potentially liable to a complaint for distributor libel, but UPS bears no potentially liability from transporting a book from B&N to a purchaser. What is the point of the question?
The two scenarios don't really align, because a book store isn't displaying the contents of books as the product, as a publisher or media company would. Do bookstores get sued for which books they sell? Can't think of an example off-hand, but maybe?
In general, a bookstore is pretty much like any store for any product. I don't expect every store to carry every possible item, obviously, I presume that stores specialize to serve a target market. In this case it would be more just a matter of market segmentation rather than censorship.
Bookstores can be liable for the books they sell, if they are on notice that a book is defamatory but continue to sell it.
Reminder we wouldn't be here if Congress hadn't threatened section 230 unless they censor harrassment, leading to censoring Trump as harrassment, leading to responding threats by his supporters if the companies continued to bend the knee (including counter-threats against 230!)
Move, counter move, much facetious concern for corporate speech rights...when they're bending the knee for us, of course.
At this point, you're all arguing how hard bad prison guards should be allowed to kick prisoners in the nuts, as if it's deep, meaningful stuff.
Congress threatened no such thing, and social media companies did no knee bending, since they told the government no all the time.
You keep posting this over and over as though that makes it true. It is demonstrably untrue.
On the contrary, thanks to the Twitter files, we know there was constant back and forth between the social media and the FBI on content moderation. It was a government assisted enterprise. And certainly still is today.
Thanks to the Twitter files, we know that Krayt's narrative didn't happen. Yes, some government employees requested/suggested that Twitter moderate stuff. And as Twitter itself told a court, those were merely requests or suggestions, and Twitter always made its own decisions about whether it wanted to do that, sometimes doing so and often not.
RJK Jr. would like a word.
Would that word be "fullofshit"? That Twitter decided it didn't want to carry his conspiratorial junk science does not mean anything other than that Twitter decided it didn't want to carry his conspiratorial junk science.
No, not twitter, but President Biden. RFK Jr is saying the Biden admin started censoring him within "37 hours of taking the oath of office".
The tweet:
"HankAaron’s tragic death is part of a wave of suspicious deaths among elderly closely following administration of #COVID #vaccines"
Flagging a tweet as disinformation is not censorship.
The FBI admitted they were actually being paid for helping monitor Twitter. This was a hand in hand enterprise.
I don't understand what that means. I think everyone knows that the FBI is not a volunteer fire department; of course the people who work there are paid.
They shouldn't have been involved at all! - Except in regard to their mission of investigating actual crimes, not thought crimes. In which case they are paid by the taxpayers, they aren't private investigators.
I think some here are unfamiliar with the term, Censorship-Industrial Complex. Michael Shellenberger coined this term with a hat tip to Eisenhower to describe the highly organized and government directed network of censoring agencies, universities, NGOs, and big tech firms that came into existence by order of Obama to make sure there wasn't a "repeat of 2016". This not speculation, this is documented fact. A good one page read is https://environmentalprogress.org/big-news/2023/11/29/censorship-industrial-complex-part-2-michael-shellenberger-testifies-before-congress
the highly organized and government directed network of censoring agencies, universities, NGOs, and big tech firms
Yes, everyone is in on it. But it's all kept secret from you, because the media is in on it as well!
The monopoly power of a carrier is not relevant to determine whether a carrier is a common carrier in many states. In Massachusetts and many other states, a Court determines as a matter of law on the basis of facts whether a carrier is a common carrier. The common law test for common carriage does not include monopoly power. A taxicab company and a bicycle message carrier company are both common carriers in Massachusetts even though neither has monopoly power. Massachusetts caselaw on this issue goes back to the colonial period, and the right to non-discriminatory common carriage is protected by the Ninth Amendment just as the right to sue for libel is protected by the Ninth Amendment despite the Doctrine of Incorporation.
Here is Blackstone on the origin of common carriage law.
A DC Circuit obiter dictum in conflict with a Fifth Circuit obiter dictum is hardly dispositive to determine whether a social medium platform is a common carrier.
I filed a complaint in the District Court of Massachusetts specifically to seek an answer to the question of common carriage. I posed the question to the District Court by invoking supplemental jurisdiction over the relevant Massachusetts statutes.
Neither Somin, Yoo, nor Kavanaugh seem to understand the Title 47 legal technological intersection because no litigation seems to have developed a record that is sufficient for review.
For example, none of the three seem to be aware that the Internet is not a privately owned network like a cable network. The Internet is mostly owned by the government and by the public.
Massachusetts routinely uses LinkedIn to substitute for a Massachusetts website to the point of symbiosis in apparent violation of state-action doctrine. Is such discriminatory inextricable symbiotic entwinement really permissible?
All of that is not just false, but delusional. Which is one of the many reasons your suit was laughed out of court.
Interesting and informative. Thanks.
"plays no significant role in the constitutional analysis...do not affect the application of the First Amendment...have little bearing on the First Amendment."
If social media is evolving to the point that the content on it is the protected free speech of Meta Platforms, Inc. (as opposed to more of a platform for communications between friends and family engaging in their own speech to each other as it resembled initially) then the social media companies will probably have to be held more responsible as the speaker of their own speech.