Structural Separation: antitrust's tried-and-true weapon for monopolists who bottleneck markets

Back in 2017, a law student named Lena Khan made waves in policy circles with the publication of her massive, brilliant, game-changing 24,000-word article in the Yale Law Journal, Amazon's Antitrust Paradox, which revisited the entirety of post-Ronald-Reagan antitrust orthodoxy to show how it had allowed Amazon to become a brutal, harmful monopoly without any consequences from the regulators charged with ensuring competition in our markets.


Now, Khan (who is now a Columbia Law fellow) is back with The Separation of Platforms and Commerce — clocking in at 61,000 words with footnotes! — that describes the one-two punch of contemporary monopolism, in which Reagan-era deregulation enthusiasts took the brakes off of corporate conduct but said it would be OK because antitrust law would keep things from getting out of control, while Reagan-era antitrust "reformers" (led by Robert Bork and the Chicago School) dismantled antitrust).


As Khan points out, it's hard to find a theory of antitrust that monopolist-friendly courts and theoreticians will accept. Bork and his pals said that the only time that antitrust should kick in was when there was "consumer harm" in the form of higher prices in the short term (which effectively makes it impossible to enforce antitrust against free-to-use internet services, where there are no prices); but in last year's Ohio v. American Express Co, the Supremes ruled that in "two-sided markets" (like a credit card company, but also like, say, Ebay, Itunes, Amazon, etc), you had to show harm to both the buyers and the sellers, "effectively creating an insurmountable hurdle." And despite innovation's centrality to "dynamic efficiency and long-term welfare," "innovation harms seem to go unaddressed under the consumer welfare framework."

In the meantime, regulators have continued to allow megamergers between giant, market-dominating companies (NBC-Universal, Ticketmaster-Live Nation) on the condition that the new conglomerates accept "conduct remedies" — rules barring them from doing nasty stuff with their new position, like using Ticketmaster's dominance to get acts to book Live Nation venues, or using Live Nation's dominance to force acts to sell their tickets through Ticketmaster, but the agencies that are supposed to monitor this activity are not structured to do so, and find it nearly impossible to keep up with abuses, or mete out punishments when they do manage to catch an abuse ("a highly enfeebled and impoverished set of tools for confronting dominant intermediaries in network industries").


And when you go to court to sue over a firm's monopolistic abuses, the court is apt to punt it to the toothless, impoverished regulator.


Khan has an answer to this conundrum: "Structural separation," another one of those "elegant weapons from a more civilized age," where companies that dominated in key sectors would simply be prohibited from doing certain things: railroads were not allowed to own the freight they shipped; TV networks were not allowed to produce or syndicate programming, banks were not allowed to own businesses, and the phone company wasn't allowed to get into the computing industry.


These "intermediaries" all owned "bottlenecks" that could be used to spy on their competitors, and they all had sources of enormous profit they could use to subsidize their entry into new fields, and by issuing blanket prohibitions on certain activities, regulators could strip down complex questions about whether harms had occurred into simple binary questions: "Did this TV network make a TV show?" or "Did this bank buy a business?" (Regulators also eventually banned affected companies from owning stock in other businesses or having interlocking directorships, closing off other loopholes). The Supreme Court was OK with this: "between 1870 and 1910, the Supreme Court four times upheld rules enjoining banks from owning commercial businesses."

The threat of structural separation enforcement from a variety of agencies (including the FDIC)headed off many potential vertical mergers, stopping Walmart, Home Depot and Target from merging with banks.


This problem of "companies in infrastructure-like sectors that compete with the businesses using their services [with] an incentive to favor their own goods or services over those owned by rivals" may sound a little familiar to people who've been paying attention to late-stage internet capitalism, and Khan finishes with a set of notes explaining how structural separation could be applied to the online world, preferably through legislation (the FTC has some authority to act here, but Khan estimates that they'd have hard, expensive, time-consuming sledding in doing so).


There's an unparalleled appetite to apply some kind of antitrust framework to the Big Tech companies, but vogueish ideas like "nondiscrimination" rules forcing Amazon not to give its house brand products over competitors' but enforcing those rules is notoriously hard — you have to prove that Amazon's ranking algorithm isn't being gimmicked by Amazon to give it an advantage, which is expensive, time-consuming and confusing. As Khan points out, simply forcing Amazon to either be in the "store" business or the "merchandise" business makes the whole matter much cleaner and easier to enforce.


A handful of digital platforms enjoy increasing control over key arteries of online commerce and communications. How lawmakers and regulators should respond to this concentration of market power is now the subject of a global debate. Public authorities around the world are studying digital platforms to understand how antitrust and competition tools can be applied to markets mediated by digital technologies. These studies vary slightly in their methods and conclusions, but they generally demonstrate that digital platform markets today are governed neither by real competition nor regulation—giving dominant platforms astounding power to shape market outcomes.

In the United States, the process of exploring how to respond to dominant platforms has been stunted by the fact that we are living through a major regulatory gap. The abandonment of traditional regulatory tools in favor of antitrust—followed by the partial collapse of antitrust—has left us with a diminished sense of the policy levers available to address dominant network intermediaries. This Article joins an emerging field of scholarship that is responding to this sense of impoverishment by exploring how traditional principles of economic regulation may apply in the digital age.

The process of identifying how to confront the challenges posed by dominant platforms requires, first, an understanding of the relevant problems and, second, an understanding of the relevant set of legal tools and principles available to confront them. Recovering our understanding of structural separations—traditionally a mainstay regulatory principle for confronting dominant intermediaries—is one part of this process. Reviewing the tradition of separations, moreover, underscores the broader set of values and concerns that traditionally informed how we assessed and arrived at the proper form of intervention when confronted with dominant intermediaries.

Recent events, meanwhile, seem to be driving the public discussion toward separations. Earlier this year, India began enforcing a structural separation on foreign online retailers—requiring Amazon to separate its private-label business from its marketplace. In March, Senator Elizabeth Warren rolled out, through her presidential campaign, a proposed separations regime for dominant tech platforms, even drawing support from some tech workers.

Getting the policy right will require careful case-by-case analysis and further study to assess the relevant trade offs. Arriving at the proper set of interventions, however, requires first knowing the full set of available tools. Recognizing the tradition of structural separations helps recover not just a mainstay regulatory principle, but also a broader framework for diagnosing and addressing the set of problems that stem from integration by critical gatekeepers.


The Separation of Platforms and Commerce [Lina Khan/Columbia Law Review]