The Speech Cartel

Should Big Tech be allowed to cancel the First Amendment?

Yesterday I was interviewed on the Ricochet Podcast to discuss the deplatforming of Trump and Parler and more generally the status of our free speech rights in the era of Big Tech monopolies. What follows is a transcript with minor edits for length and clarity.

Please note:

• I recognize that in a crisis, emergency measures are warranted, especially to prevent violence, which happened when Trump supporters stormed the Capitol in a unique outrage to our democracy.
• I’m talking here about decisions to permanently ban or deplatform individuals and/or businesses with no ability to appeal. Big Tech went beyond emergency measures and took permanent actions.
• What happens when the immediate crisis lifts? Will we continue to allow a consortium of powerful tech companies to control who has the right to assemble and speak online?

You can listen to the pod here or please read below:

Q: How would you characterize what Jack Dorsey said [in his tweetstorm] about his decision [to permanently ban Trump’s Twitter account] and what do you think the implications are going forward?

A: Jack is my favorite oligarch because after taking these actions and there’s a huge outcry, he then gets very contemplative and introspective about them. I think he made a really important point: he said, when we (Twitter) decided to deplatform the president, one of the key arguments that we considered is that he could go to other sites and other platforms and have his speech rights that way. And he basically concedes that this argument was undermined by the fact that every other major tech company then did the same thing.

I think he’s really on to something there, which is that when one company denies you service, maybe that isn’t such a big deal because you can go somewhere else. But when all of Big Tech does the same thing, they are now forming a monopolistic cartel, if you will, to deny you your free speech rights.

And this is the thing that concerns me. It’s not specifically Trump. Frankly, the silence over the last several days has been blissful. I’m not necessarily going to miss his tweets at all, but I’m concerned about the precedent that we are giving a consortium of Big Tech companies the power to deny us our free speech rights. If these companies had gotten together as a consortium to raise prices, everybody would be up in arms, but instead they’ve gotten together to deny us our civil liberties and no one’s saying a word. Isn’t that more important? Aren’t our civil liberties more important than money?

Q: You mentioned something interesting when you talked about cartels and getting together. In the old days, we imagine these plutocrats with their watch chains — after their lunch of oysters — would sit down and carve things up and form these monopolies. Do you think they actually sat down and planned what some people are calling a coordinated, conservative purge or was this a spasmodic reaction?

A: What Jack said is that we didn’t “coordinate” with all of these providers of core internet infrastructure, but we “emboldened” each other. I think if you were to apply an antitrust overlay to the First Amendment, you could call that a kind of signaling. There’s this game of follow-the-leader. Why is everyone from Snap — does the president even have an account on Snap? — why are they suddenly joining this bandwagon? Twitter and Facebook, and Apple and Google especially, are the leaders in our industry. This signaling by companies that normally compete with each other, for the purpose of raising prices, would clearly be a violation of antitrust law. What I’m suggesting is that cartels should not be allowed to get together for the purpose of regulating speech either. I would like to see some enterprising lawyer run with this take and provide this antitrust overlay on free speech law.

I got a lot of pushback on Twitter for this point of view that I’ve been espousing over the last week. What they say to me is that these are all private companies. They should be able to do whatever they want. But here’s the problem with that: our free speech rights have been privatized. Speech became digitized when it moved to the internet, and now that is the place where most political speech occurs. Speech has become digitized, privatized, and centralized in the hands of a small number of players, a handful of oligarchs, if you will.

The framers of the Constitution could never have anticipated that that was going to happen. Back in their day, there were a multiplicity of places you could go for speech. There were town squares all over the country, and those town squares have all now been replaced by a handful of social media sites. So what are your effective free speech rights today if you get deplatformed? You don’t have them.

Q: You have some legal training. You’re able to think in legal categories. “Congress shall make no law abridging the freedom of speech.” Clearly the First Amendment concerns what government may and may not do. You’ve already stated this problem. Twitter, Facebook, Google, they’re all private enterprises. At the same time… I hear antitrust law, and I’m thinking Theodore Roosevelt and the 1880s. It seems to me that the legal framework for dealing with questions of free speech in 2021 simply doesn’t exist. Even antitrust law is out of date. If that’s correct, who fixes that problem? Is it legislation? How do you grapple with this?

A: I think you’re right, that what’s missing is an online bill of rights. The First Amendment explicitly says that Congress shall make no law. It’s a prohibition on government action. The point I’m making with the antitrust argument is that when a cartel of super powerful monopolies gets together and decides something, whether explicitly through coordination or by emboldening each other to take the exact same position, that is effectively like a government decision. You have no effective ability to contest that. Jack basically admitted as much. If you look at his tweetstorm, he said once everybody started doing the same thing, it felt like a government action, as opposed to the action of one private company.

That’s my point. This is why we have antitrust law — because we realize that although companies in general, as a starting point, should be free to do what they want, their power in the market can reach such a point that they need to be regulated and constrained. There’s a doctrine in antitrust law called “essential facilities.” Let’s say there’s a drawbridge that is the only way to get across the river and a monopoly begins denying use of that drawbridge for its own nefarious purposes. We would declare that to be an essential facility and stop them from doing that because they are acting with the power of a government. I’m trying to provide a way of thinking about what Big Tech is doing that should be broadly unacceptable to Americans, regardless of what side of the political spectrum you’re on.

Q: So the legal analogies in antitrust law exist already?

A: Absolutely. It’s just that we’re not used to thinking about antitrust as applying to civil liberties. We’re only used to thinking of antitrust law as applying to consumer harm. But again, aren’t civil liberties more important than money? If we want to prevent cartels from getting together to raise prices, why in the world would we not prevent them from getting together to deny us our civil liberties? How do we functionally still have a free speech right in this country if it can be taken away by a clique of tech elites?

Q: How much leverage do the employees have? How much power do the employees at Twitter or Facebook have in making, helping or driving these decisions?

A: They have a lot of power. They’ve certainly banded together to create a tremendous amount of pressure on the leadership of the company. These decisions aren’t being made in a vacuum but based on pressure that comes from both below and above. The employees are pressuring from below. There’s also pressure coming from the Senate Judiciary Committee, from above. These tech companies are simultaneously being told by a number of senators who will now control that committee that: A) we don’t think you’re doing enough to censor and B) we’re contemplating breaking you up. It’s sort of the protection end of the extortion racket: “listen, unless you guys start getting tough…”

Q: “Nice little social media site you got here.” Yeah…

A: I find that aspect of this very disturbing. We have the state exercising soft power on Big Tech to engage in censorship that would never be allowed under the First Amendment if the state were to do it itself.

Q: If you’re Jack or the CEO of a large content platform, trying to attract engineers in a tight labor market and now you have a Democrat controlled Senate, wouldn’t you cave if you were on the Board? Wouldn’t you say, Jack, the Board votes to cave?

A: I just think our commitment to freedom of speech needs to be made of sturdier stuff than that. It’s always the case throughout history that the free-speech test cases require the defense of very unpopular people saying very unpopular things. It’s a canard that in order to defend free speech, you must believe in the worst things being said. This was the historical ACLU position.

I’m taking a lot of hits on Twitter from people who are saying: if you’re defending free speech, you must be saying that you believe in incitement or violence or whatever provocation Trump or somebody on the right has said. But that’s not really an argument. That’s an intimidation tactic.

Q: There’s a firehose of data coming through the cloud. How can Facebook or Twitter tell, how can they possibly read it all? How does that funnel work? Like if you use the word “decapitate.” Where does that go? To whose office? How does this work?

A: These are the kinds of details that are really interesting to get into. Because I know something about the operations of these tech companies. I can tell you that when you actually try to operationalize this idea of censorship, of taking down problematic posts, you realize what a gigantic mess it is. After my tweets, I had a social media founder, not Facebook or Twitter, but another site that deals with user-generated content, tell me: “David, you’re more right than you even know.” Because what happens in practice is you start with this policy that seems really sensible: we need to take down content that’s XYZ. But it’s very hard to know exactly how to apply that policy.

So they create guidelines. And then, because there’s so much content to review — millions and millions of posts — the work gets farmed out. They’re often outsourced to BPOs, which are business process outsourcing organizations in far-flung places like the Philippines, where you have low-paid workers reviewing all this content and quite frankly English might not even be their first language. They’re trying to decipher our political and cultural hot buttons and apply these Byzantine guidelines to millions of social media posts. It doesn’t work at all. This is why you end up with people like Ron Paul getting caught in the dragnet.

I don’t think the leadership of Facebook intended for Ron Paul to be caught in the roundup of the usual suspects. They can see that he’s not a violent insurrectionary, but when these censorship principles are actually operationalized, it’s just a giant mess. Ron Paul got his account reinstated because he’s a prominent public figure and his son is a US Senator. What about the rest of us? What about the rest of us who can’t put public pressure on Facebook to embarrass them into giving us our account back, with them saying, “Oh, sorry, it was just a mistake.” How many mistakes are going to be allowed to happen before we realize that this process is broken?

Q: So you hear these people on Twitter saying they’ve received death threats. I kind of always roll my eyes and say, “Oh, shut up, come on. You didn’t, this is not real.” Am I a bad person?

A: Well, I think we all have our own threshold for what we’re willing to tolerate. I certainly see tons of hate and vitriol on social media sites, usually directed at denouncing the other side’s hate and vitriol. There seems to be something about social media that brings out the worst in people. I see people I know in the real world who are perfectly reasonable people come across like total lunatics on Twitter. They come across as very intolerant people; I don’t think they are in the real world. We all have to fight that instinct when we participate online. But I think there is a serious question here about what is allowable and whose standard should be the standard — because we all have different standards.

This is where I go back to the First Amendment. We have a standard. It’s an external standard that has been defined over the last 230 years by the Supreme Court. It carves out a lot of the speech that’s most problematic to social networks. There’s a lot of speech the Supreme Court has said is dangerous and should not be protected. That speech includes things like incitements to violence. You can’t yell fire in a crowded theater. You cannot defame people. Free speech is not a defense to fraud. And so on the list goes. There’s about nine major categories of problematic speech that are not protected by the First Amendment.

I would like to see that at least with respect to these monopolistic players, who truly control free speech in the modern era, they say, “our standard is going to be the First Amendment. That is what we’re trying to operationalize. We’re going to make mistakes, but at least that is what we’re shooting for. We can all agree on those principles.” That is the standard I would like to see them adopt. [For more on this idea, please see my previous blog on Section 230.]

Q: Should we be on Signal? Should we be more encrypted?

A: It’s almost a separate question because that’s about private communication, and I think what we’re talking about here is the right to speech, which is about public communication. But the reason why it’s come up is because of something I and some other people have predicted: if you try to censor in a heavy-handed way, then you drive a lot of speech underground. To the extent you are worried about people coordinating various kinds of things you want to stop, in some ways it might be better for it to be on social media where you can track it and see it and eventually prosecute it than to drive it completely underground. There is an interesting tradeoff there. To be clear, I think speech that’s inciting a crime should not be allowed. I want to be very clear about that, but I do worry that a heavy hand is simply going to push a lot of the speech underground where we can’t track it at all.

Q: AWS (Amazon Web Services) kicked Parler off. Is AWS a big place with lots of servers, air conditioned, and somebody walks around like that scene in the show, Silicon Valley? What is it?

A: There’s the hardware aspect and the software aspect. The hardware aspect is these giant server farms that create the physical infrastructure that the internet runs on. There’s many of them now, they’re gigantic. It’s a scaled solution. Then there’s the software aspect, which is how AWS makes it really easy for you to plug into that hardware. It’s kind of like an operating system.

Q: If I’m a young entrepreneur and I come into your office to pitch my business to your venture firm. And I say: We’re going to build better server farms with the business principle that we are never going to kick you off. Is that a business?

A: No, it’s not. Because there’s no way you could ever get to scale to compete with Amazon. So what you’re going to provide is something for a tiny sliver of the market that gets kicked off Amazon. It’s not a viable business.

Q: Of all the things we have talked about, the networks, the collection of data streams, the one that has the best chance of being defined as a public utility is AWS. Is that fair?

A: I think AWS has very, very strong market power and there is a monopoly effect in terms of economies of scale. There’s no question that there’s monopoly power there, but an even stronger monopoly than AWS is the Apple/Google mobile operating system duopoly. If they kick you out of their app stores, it’s almost impossible to effectively get distribution on mobile, which is where most of the engagement occurs.

But also, there’s a combined effect, which is why I go back to this idea of a cartel: when Twitter and Facebook banned Trump and other accounts, the argument was, “well, that’s not censorship. Just go create an account somewhere else, go to a different app.” Then Apple and Google banned Parler, which was the different app to go to. And then the argument was, “well, that’s not censorship, just go create a website.” And then AWS bans. And so at some point, when are you going to say that this is an undue imposition on free speech? What am I supposed to do, go create my own internet? All I wanted to do was post a tweet. Come on, let’s not be deliberately or selectively oblivious to the power of these monopolies.

Q: David, one last question. A few years ago, when we all learned about China’s social credit score, we were duly and properly horrified by the idea that the government would assemble such things and keep you from getting an airplane ticket or a good apartment, and it struck us as a totalitarian nightmare and not surprising for China. Now, some people believe there is a sort of de facto social credit score. Your past posts on Parler can be scraped. I read today on Reddit that somebody got a two terabyte archive that they’re bit torrenting of all the videos, you’ve got your post history, you’ve got all these things that people can assemble and point to and correlate and say, this person is slightly sorta kind of smelling like a deplorable and we best not give our Airbnb or Lyft or Uber to them. Are we moving towards not an official social credit score, but a de facto one, or do we already have it? Is this going to get worse until somebody sues the bejesus pants off somebody for it?

A: I think that the people who are decrying Trump’s authoritarianism should be careful not to fall into a similar trap here — of wanting to form mobs, tweet mobs, to not only deny people their free speech, but also to cancel them economically, to cancel their economic livelihoods. It is a disturbing trend. You have this cancel culture that says: I’m not just going to disagree with you, I have to find a way to remove you, not just from social media, but from the entire economy. It’s something to be very concerned about.

That is why we need an online bill of rights. The first question in law is jurisdiction: “who decides?” Right now we are giving Big Tech the power to decide everything, the power to decide our free speech, what we think, what we write, what we get to see and listen to, the power to create a company.

If you get denied access to this core infrastructure on the internet, to what extent can you even create a business in the modern world? Right now they can deny you service for any reason whatsoever. I see these people on social media trying to act as if the terms of service is somehow a protection. The terms of service has almost been fetishized as a guarantee of due process. The terms of service just means whatever the company says it means. They can modify it at any time. It doesn’t give you any rights.

At some point we move beyond Trump, and hopefully that happens very soon, but this question will remain. Once the issue of Trump goes away, I think people are going to be able to see that we have allowed a tech oligarchy to take control of so many aspects of our lives without any rights whatsoever.

People are a bit blinded to that fact right now, because they hate Trump so much. Just because they love the decision to ban and deplatform Trump, they’re willing to go along with this incredible power grab by Big Tech. I am not as concerned specifically about not being able to read Trump’s tweets. Like I said, frankly it’s been blissful not to get these push notifications pop up on my phone every day, alerting me to some new provocation. It’s like, ah thank god. I’m not missing that at all. But I’m worried about the precedents we’re creating here and this transfer of power that we’re just allowing to happen.

Q: Thank you.