Parler’s antitrust case against Amazon is doomed, and there is no basis for Congress to impose any special duties on Amazon to deal with Parler outside of antitrust. 


Parler’s sudden vanishing act has caused progressives to feel a sort of split-personality disorder: One identity takes comfort in knowing that the social media app can no longer be used to plot insurrections; a second identity takes umbrage at the notion that large digital platforms have the power to disappear disfavored edge providers. 

Long a favorite of right-wing conspiracy theorists, Parler can no longer be downloaded from either of the two dominant mobile app platforms, Google’s Play Store and Apple’s App Store. And even more threatening to its survival, Amazon will no longer host Parler on its cloud services. 

On the surface, these appear to be dominant platforms pushing their weight around the internet. But the Parler takedown is less about competition than meets the eye. 

With roughly one-third of global cloud service revenues, Amazon Web Services (AWS) has a commanding lead, yet competes against other internet behemoths. As The Economist reported this week, “It is unlikely that the two other big cloud services, run by Microsoft and Google, will want Parler anywhere near their servers.” With nowhere to turn domestically, Parler was reportedly considering a Russian web host. The dearth of US suitors implies that Amazon’s refusal to deal with Parler is not driven by Amazon’s dominance, but instead by a legitimate business justification.

Invoking the language of a libertarian, Parler’s CEO John Matze told New York Times columnist Kara Swisher that any attempt to police communication on his platform—even plans for civil war—would be tantamount to “authoritarianism.” Because Parler refuses to clean up its act, and because tech platforms are exposed to federal criminal liability even under Section 230, no platform, large or small, would rationally elect to serve as a conduit to Parler’s coup plotting. Moreover, the mere association of any platform with Parler induces howls for a full repeal of Section 230 and other political costs. And platforms must worry about employee morale.

To restore its hosting arrangement, Parler filed an antitrust complaint against Amazon, arguing that AWS’s treatment of Parler was “inconsistent with its treatment of Twitter, indicating a desire to harm Parler.” Setting aside its unorthodox appeal, drafted by an IP attorney, Parler effectively is asking the court to impose a duty to deal on Amazon. This won’t work, as antitrust’s duty to deal is “at or near the outer boundary of § 2 liability,” and even then only applies in limited circumstances, such as when the dominant firm’s present dealings discriminate between rivals and non-rivals. Parler is no more of a rival to Amazon than is Twitter. 

As noted by antitrust attorney Jim Martin, the conspiracy claim could be colorable if Parler could credibly allege an agreement between Twitter and AWS to exclude a growing competitor to Twitter. But there was no exclusionary agreement alleged in the complaint, and no reason to infer one.

Perhaps under older refusal-to-deal law, a case could be made against Amazon under a “lock-in” theory, in which certain dominant companies are endowed with a set of routines and competences that bound their behavior. Yet the lock-in argument is arguably stronger for Parler’s dealings with the dominant app stores, where coordination with its customers would be required to make the switch to a new platform. Nothing stops Parler from switching hosting service to Google or Microsoft, except for the liability and political costs that Parler brings to the table.

Outside of antitrust law, Congress has imposed a duty to deal on dominant platforms on a handful of occasions, even when those platforms were engaged in commercial speech. For example, the 1992 Cable Act imposed must-carry obligations on cable operators in their dealings with broadcasters, and imposed program-access obligations on vertically integrated cable operators in their dealings with distribution rivals. Not only did cable operators have to deal with satellite providers, but the rate at which rivals could access cable-affiliated programming was regulated at fair market value! The 1996 Telecom Act imposed fresh duties, such as open access and interconnection on incumbent local exchange carriers in their dealings with competitive local exchange carriers and long-distance providers, respectively. 

“Although competition problems are endemic across the internet, Parler doesn’t seem to be especially vulnerable to the exercise of market power by a dominant platform.”

Importantly, the motivation to impose these duties on dominant platforms outside of antitrust was meant to solve a competition problem that couldn’t be solved via antitrust. Left unchecked, market forces would allow cable operators to overcharge distribution rivals for cable-affiliated content, leading to weakened competition in the distribution market. The growing power imbalance between cable and broadcast threatened an important source of local news and local employment. Market forces would also ensure that incumbent telephone companies would leverage their power into long-distance or internet services, and cement their dominance in local service. And the slow pace of antitrust, as well as the limits to antitrust’s duty to deal, were seen as obstacles to Congress’s competition concerns.

What is the equivalent competition concern in the case of Parler? No one has seriously suggested that AWS wields selling power over its customers: If Amazon were to raise its hosting rates above competitive levels, content providers would likely turn to Google Cloud or Microsoft Azure, or IBM Cloud, assuming the switching costs weren’t prohibitive. (In contrast, Amazon’s buying power over independent merchants on Amazon’s e-commerce platform is more plausible.) 

Although Apple and Google wield significant buying power over app developers such as Parler, to the extent the platforms’ take rates (typically around 30 percent) are supported by exclusionary restraints, those restraints can be challenged under antitrust laws. Because Parler’s app is free of charge, the question of excessive take rates charged by the app stores is rendered moot. 

And no platform is underpaying Parler for some material contribution Parler makes to the platform’s revenue. Although competition problems are endemic across the internet, Parler doesn’t seem to be especially vulnerable to the exercise of market power by a dominant platform. 

Even nondiscrimination rules, another obligation in the common-carrier panoply motivated by a competition concern (edge innovation), would not assist Parler unless and until Amazon, Google, or Apple vertically integrated into a (similarly situated) social media app. And even then, the nondiscrimination obligation would merely oblige the platform to treat Parler the same way it treats its own social media app. Unlike a duty to deal, a hypothetical vertically integrated Apple could elect not to distribute its own social media app and Parler in its app store, or it could elect to place both apps near the bottom of its search results.

In sum, the case for a duty to deal with Parler, either within or outside antitrust, is exceedingly weak. This is not to say that platforms should be immune from such obligations when dealing with other types of content providers. (Stay tuned for some proposals here!) But where there is no legitimate competition concern, there can’t be any fresh regulatory obligation. 

Let’s not make Parler the poster child for confronting the serious competitive threats posed by dominant platforms.

Disclosure: The author’s firm is involved in litigation against both Apple and Google.