Liz Wolfe, in recent weeks, seems to have displaced Elizabeth Nolan Brown as the chief tech platform apologist at Reason. In her most recent article on this theme, “Google CEO Sundar Pichai Is Right: Companies That Dominate Today May Be Gone Tomorrow” (July 26), the general lines of her argument are simple: “Today’s Big Tech critics forget that companies with significant market share come and go….The great antitrust push of the 2020s rests on the strange assumption that this sort of creative destruction won’t happen, despite the fact that the historical record indicates the opposite.” 

It’s hardly limited to Reason, of course. Recently I saw prominent libertarians on social media mocking Warren for tweeting that Facebook was a monopoly (she was using Twitter to complain about Facebook — get it?). Aside from the alleged self-own of disproving her own claim by using one platform in what’s essentially a two-platform market to criticize the other platform for its monopoly power, she also ignored the fact that Facebook supplanted Myspace, and so on, and so on.

What these arguments get wrong is that it’s the structure of the market at any given time that determines monopoly power, not whether one giant firm periodically replaces another in a market that itself retains the same structure over time. It’s not so much that tech critics “forget” that giant firms come and go, but that it’s irrelevant to the existence of monopoly power.

What matters is that, while Facebook or Twitter may eventually be replaced by some other social media giant, the tipping point will be reached only when they have a user community in the hundreds of millions that offers sufficient network effects to spur mass migration from one platform to the other. And it necessarily follows from the size of the user community required for such network effects — at least in the hundred of millions — that the overwhelming majority of the social media market at any given time will be controlled by a small handful of platforms. 

What’s more, this market structure results from a protectionist legal regime — a legal regime which neither Wolfe nor Brown has ever bothered to acknowledge in any of these gee-whizzy pieces about poor old Mr. Zuckerberg vs. Dollar Democracy — that requires achieving such network effects as a condition for becoming a major player in the market. 

Specifically, these protectionist laws function as entry barriers by requiring a very large minimum size even to get on the field with the big boys. They achieve this by making it impossible for small players to expand on a granular or modular basis, and thereby gradually erode the market share of incumbents. To see how attempts at doing so work out, just take a look at Mastodon and other instances in the Fediverse. Despite several waves of people very noisily leaving Twitter for the Fediverse in response to the latest ham-handed mismanagement by Jack Dorsey, and encouraging others to do the same, the total number of users in Fediverse instances peaked at about 2% of Twitter’s total user base. Mastodon and other Fediverse instances are a quirky niche market, and will never be anything else so long as the present legal basis for Twitter’s market power remains in place. I set my own account up at Octodon.social, but rarely use it because 1) most of my friends on Twitter aren’t there, 2) I’m not willing to leave behind my closest Twitter friends by migrating completely to the Fediverse, and 3) I can’t afford to devote the same time and attention to yet another platform while maintaining my commitments on the old ones. Those network effects again, in other words.  

I keep referring to the legal monopoly power of Facebook and Twitter. But what, specifically, does that mean? The most important thing prohibited by the protectionist legal regime is what Cory Doctorow calls adversarial interoperability. That is, it prevents people from setting up their own independent, user-governed instances and piggybacking them on the Facebook or Twitter platform, importing their their Facebook and Twitter contact lists, and posting cross-platform — all without permission from Facebook or Twitter. As Doctorow explains it, 

Facebook alternatives like Diaspora could use their users’ logins and passwords to fetch the Facebook messages the service had queued up for them and allow those users to reply to them from Diaspora, without being spied on by Facebook. Mastodon users could read and post to Twitter without touching Twitter’s servers. Hundreds or thousands of services could spring up that allowed users different options to block harassment and bubble up interesting contributions from other users — both those on the incumbent social media services, and the users of these new upstarts.

Under the present legal regime, where this is prohibited, the entry barrier presented by network effects means you have to either get big or go home. So no matter how many times one of the handful of giant platforms is replaced by another, the market will still be dominated by the same tiny handful of platforms that are big enough to achieve network effects. But adversarial interoperability totally does away with the need for network effects — it lets people go to new instances with governance rules and user interfaces more to their liking, while keeping the networks they already have at Facebook or Twitter. 

Facebook and Twitter become Meta-Facebook and Meta-Twitter, just infrastructures or platforms for hosting hundreds or thousands of separate, self-governing instances, and the network effects of Facebook and Twitter become the equivalent of cows skeletonized by piranha. Potentially the same with a Meta-Amazon hooked into by independent sellers, a Meta-Uber used as a platform by ride-sharing co-ops, etc. It’s like converting a Walmart supercenter into the hosting site for a bazaar of a thousand booths, with Walmart retaining control only of the lights and air conditioning and the janitorial service.

But this is illegal. We’ve got a legal regime that requires you to have sufficient capital and resources to get the same network effects as Zuckerberg or Dorsey, before you can replace them. Here are some of the specific laws (from Doctorow again):

CFAA is nominally an anti-computer-intrusion statute, which criminalizes “exceeding your authorization” on a computer that doesn’t belong to you. Even when it passed, more than 40 years ago, technologically clued-in scholars and practicioners [sic] warned that this was way too broadly defined, and that someday we might see this rule used to felonize normal activities involving computers we owned, because the computers would have to talk to a server to accomplish part of their work, and the server’s owner could use onerous “user agreements” and “terms of service” to define our authorization….

40 years later, those fears are vindicated: CFAA is used to threaten, intimidate, sue, and even jail people engaged in otherwise perfectly lawful activity, merely because they have violated some term of service on the way. The metastasis of terms of service into sprawling novellas of impenetrable legalese has created a world where anything you do to frustrate the commercial ambitions of digital monopolists is a potential criminal offense.

Then there’s Section 1201 of the Digital Millennium Copyright Act of 1998, a Bill Clinton bill that creates a felony for “bypassing an effective means of access control” (AKA Digital Rights Management or DRM) for copyrighted works….

Together, the CFAA and DMCA have given digital businesses access to a shadowy legal doctrine that was never written by Congress but is nevertheless routinely enforced by the courts: Felony Contempt of Business-Model.

…Uber and Lyft have lengthy terms-of-service that set out the rules under which you are authorized to communicate with Uber and Lyft’s servers. These terms of service prohibit using their servers to locate drivers for any purpose other than booking a ride. They certainly don’t permit you to locate a driver and then cancel the booking and re-book with a co-op app.

And Uber and Lyft’s apps are encrypted on your phone, so to reverse-engineer them, you’d have to decrypt them (probably by capturing an image of their decrypted code while it was running in a virtual phone simulated on a desktop computer). Decrypting an app without permission is “bypassing an effective means of access control” for a copyrighted work (the app is made up of copyrighted code).

Now let’s stop to take a look at the other argument, that the existence of Twitter somehow demonstrates Facebook’s lack of monopoly power. Well, technically it does; as Lionel Hutz put it, it’s the best kind of correct — technically correct. What Facebook and Twitter have is actually oligopoly power. But it’s a distinction without much of a difference. The power relation between corporations and consumers is pretty much the same either way.

Whether in monopoly or oligopoly markets, firms are price makers rather than price takers. In an oligopoly market this generally takes the form of a “price leader” system, in which the other firms follow the lead of the dominant firm in setting their own prices. And the firms in an oligopoly market have considerable leeway to limit competition in business models, product design, etc., and to carefully spoon out improvements. Remember the gentlemen’s agreement between the Big Three auto companies sixty years ago not to introduce a number of product innovations until all three were ready to do so?

Of course social media’s a bit different because it’s free — your eyeballs are the source of revenue. But the restricted competition in terms of quality and basic business model is the same as in any other oligopoly market. Facebook and Twitter are both pretty arbitrary and authoritarian, and generally introduce changes in their user interface that bear no appreciable relation to what a majority of users want. Facebook’s design and governance are godawful because users have nowhere else to go but Twitter, and vice versa.

So Elizabeth Warren is right. Facebook is a monopoly, with unaccountable power over its users, and it wields that power in close partnership with the state. And it’s not the critics of big tech who miss the point, but Liz Wolfe. 

Of course Warren’s approach to breaking Facebook’s or Twitter’s monopoly power is entirely wrong-headed. Typical of an old-style liberal, she sees the corporate economy through the lenses of the 20th century mass production era, and wants to break up Facebook like it was Standard Oil. But Facebook’s power needs to be broken. And the way to do it is not by passing new laws, but by repealing old ones that keep us from breaking it — or maybe by developing the technical means for circumventing existing ones without getting caught.