What are the bases of the digital platforms’ power? In Chapter 3 of our Digital Platforms and Concentration ebook in advance of this month’s blockbuster antitrust conference at Chicago Booth, Lina M. Khan points to the gatekeeper power, leveraging power, and information exploitation power of America’s tech giants.
Editors’ note: This piece is adapted from a forthcoming article in the Georgetown Law Technology Review. See also our 2017 conference volume Is There a Concentration Problem in America? Other chapters from this year’s ebook can be read at the following links:
A handful of tech platforms mediate a large and growing share of our commerce and communications. Over the last year, the public has come to realize that the power these firms wield may pose significant hazards. Elected leaders ranging from Senator Elizabeth Warren (D-MA) to Senator Ted Cruz (R-TX) have expressed alarm at the level of control that firms like Amazon, Alphabet, and Facebook enjoy. In a recent poll, a majority of Americans expressed concern that the government wouldn’t do enough to regulate US tech companies. As the editor of BuzzFeed observed, a “major trend in American politics” is “the palpable, and perhaps permanent, turn against the tech industry,” now viewed as “sinister new centers of unaccountable power.”
New revelations continue to unveil the degree of power these firms wield and its consequences. The potential effects range from stifling startups and undermining innovation to manipulating the flow of information and enabling foreign interference in our elections. Despite growing recognition of platform power, public conversation about why this power exists and what to do about it is still in its early stages. This essay seeks to help advance that discussion by identifying forms and sources of platform power, explaining how this power is being or could be exploited, and exploring historical analogies and legal hooks that could help us tackle it.
Forms and Sources of Platform Power and Its Abuses
The markets in which these firms operate and the specific mechanics of their business models somewhat vary. For this reason, more extensive studies of platform power would benefit from being platform-specific. But despite their differences, Amazon, Alphabet, and Facebook share key forms and sources of power.
The first is gatekeeper power. The source of this power is the fact that these companies serve effectively as infrastructure for digital markets. They have captured control over technologies that other firms rely on to do business in the online economy. Fifty-five percent of online shopping searches, for example, now begin on Amazon’s platform; last year the company enjoyed over 40 percent of online revenue in the United States. Alphabet and Facebook together capture 73 percent of all digital advertising in the country and 83 percent of all growth, while Apple and Alphabet jointly account for 99 percent of the world’s smartphone operating systems. For producers, retailers, advertisers, and app developers looking to reach users and consumers, these platforms are vital intermediaries, the railroads of the 21st century.
The degree of market control enjoyed by dominant platforms is protected both by network effects and the self-reinforcing advantages of data, which serve as an entry barrier. Their entrenched positions are reflected partly in their skyrocketing valuations; Wall Street is pricing their stock at multiples that seem to reflect market power. Newcomers that have attempted to compete with a platform in a platform market (like Jet.com) have been acquired by other giants (Walmart).
This means that not only are the platforms vital intermediaries, but—in many instances—they are the only real option. Even when producers, retailers, advertisers, publishers, and app developers manage to find alternate channels, those narrower paths can only really supplement access on the margins. The platforms generate too much business and attract too many eyeballs for firms to bypass them entirely. This renders business users highly dependent on the platforms—a finding confirmed by a recent study undertaken by the European Commission. The EC wrote, “Many of the business users have indicated that they try to avoid any conflict with platforms, fearing a negative impact on their business. This applies especially to conflicts with the largest platforms, as business users indicate that often no viable alternative for these major platforms exists due to their scale, geographic range and the number of (potential) customers active on the platforms.”
Platforms can use their gatekeeper power to extort and extract better terms from the users that depend on their infrastructure. For example, Amazon has disabled the “buy-buttons” for book publishers in order to extract better terms; executives have also described how the company tweaks algorithms during negotiations to remind firms of its power to sink their sales. Recently the company has started offloading costs onto suppliers, subsidizing its shipping costs by raising fees for the companies that sell through its platform. Merchants attempting to negotiate with Amazon risk seeing their accounts suspended, and getting kicked off its platform often means not just seeing lower revenue but having to lay off employees. Google and Facebook’s ad duopoly, meanwhile, gives them ample power to raise prices. Last quarter Facebook hiked the average price per ad by 43 percent.
Platforms also use their gatekeeper power to entrench their gatekeeper power, limiting the ability of third-party merchants to reach users independently. Amazon, for example, closely monitors communications between third-party Marketplace merchants and consumers, penalizing merchants who direct consumers to their own independent websites or other sales channels. Gatekeeper power now also risks shaping the content and production of news. Dependence on Facebook and Google for traffic has led publishers to package news according to the dictates of the platforms’ algorithms. As a bill recently introduced by House Representative David Cicilline stated, “An entity with the power to dictate the terms of distribution of news has the power to dictate the content of news.” The head of the Newspaper Association of America noted, “Facebook and Google are our primary regulators.”
A second form of power is leveraging. The source of this power is the fact that the platforms not only serve as critical infrastructure, but are also integrated across markets. This enables a platform to leverage its platform dominance to establish a position in a separate or ancillary market. By placing a platform in direct competition with the firms using its infrastructure, this form of integration also creates a core conflict of interest, incentivizing a platform to privilege its own goods and services over those offered by third parties.
Last year the European Commission announced that this form of discrimination violates European competition laws. It fined Google $2.7 billion for “systematically giv[ing] prominent placement to its own comparison shopping service” and “demot[ing] rival comparison shopping services in its search results,” leading traffic to third-party websites to plummet. The EU competition authority is also conducting investigations into potentially anticompetitive leveraging tactics Google engaged in through its Android operating system and AdSense. Apple, meanwhile, has previously blocked updates to Spotify from the App Store; Spotify alleges this tactic sought to undermine Spotify as a rival to Apple Music. If gatekeeper power gives platforms the ability to extort, leveraging power gives platforms the incentive to discriminate.
|By placing a platform in direct competition with the firms using its infrastructure, this form of integration also creates a core conflict of interest, incentivizing a platform to privilege its own goods and services over those offered by third parties.|
A third form of power is information exploitation. The source of this power is the various forms of data that platforms collect, in multiple markets. Platforms gather enormous amounts of information, ranging from the amount of time you hover your mouse over a particular button and the number of days an item sits in your shopping basket, to every location you’ve visited with your phone and how you psychologically react to different posts and words.
In some cases, platforms also track user activity on third-party websites and applications. Platforms can use this data in a host of ways, altering what information you see based on your profile. Platforms can also harness this data to engage in first-degree price discrimination, charging each consumer a different price for the same good or service. Uber, for example, has admitted that it engages in personalized price discrimination. The degree to which other platforms are engaging in similar practices has not been publicly documented. Separate from the risks of discrimination, the extent of platforms’ data-gathering creates significant privacy threats. Even robust privacy controls would only go so far to protect users, given the security vulnerabilities that inevitably arise when data is concentrated in a single entity.
Platforms also engage in information exploitation against the businesses that use their services to reach markets. Amazon, for example, collects swaths of information on the merchants selling through its Marketplace. It routinely uses this data to inform its own sales and products, exploiting insights generated by third-party retailers and producers to go head-to-head with them, rolling out replica products that it can rank higher in search results or price below-cost. In this way Amazon’s platform functions as a petri dish, where independent firms undertake the initial risks of bringing products to market and Amazon gets to reap from their insights, often at their expense. Facebook has similarly developed systematic ability to exploit information. Through acquiring Onavo, a privacy-enhancing technology, Facebook closely tracks which competing applications are diverting attention from Facebook’s own app. Using this information, Facebook can either make an aggressive acquisition bid, taming the nascent threat by bringing it in-house, or can introduce an identical app, eating into its business.
The issue here is not that the platforms introduce rival goods—thereby increasing competition—but that their strategies are based on a significant information asymmetry that exists between the platforms and everyone else. The ability to intervene at the very earliest stages of a company’s growth means platforms can effectively nip emerging rivals in the bud.
|The issue here is not that the platforms introduce rival goods—thereby increasing competition—but that their strategies are based on a significant information asymmetry that exists between the platforms and everyone else.|
To be sure, platforms exhibit other forms and mechanisms of power. But these three sources—gatekeeper power, leveraging power, and information exploitation power—go far to explain the current dominance these firms enjoy.
Ways to Address Platform Power
Breaking down platform power into its specific forms and sources allows us to distill what about platform power, if anything, is actually new. In other words, we can understand which facets of platform power we have grappled with in the past, and which aspects present new issues that require new thinking and/or new policy action.
Two of these forms of power—gatekeeper power and leveraging power—we have tackled in the past. Gatekeeper power can arise any time there is a network monopoly. Indeed, the gatekeeper power of the railroads—and the railroads’ abuse of this power—gave rise to the anti-monopoly movement in the late 1800s, ultimately leading to the creation of the Interstate Commerce Commission in 1887 and the passage of the Sherman Antitrust Act in 1890. Determining that breaking up the railroads would hamper our national transportation system, Congress designed a regime to prevent railroads from abusing their power. Most notably, railroads had to abide by common carriage rules, providing equal access on equal terms, and had to publicly list their prices. This helped scale back their power to arbitrarily hike prices and extort the farmers and suppliers reliant on the railroads to get to market.
Indeed, common carriage has been a traditional tool for maintaining the benefits of network monopoly while preventing the private firms who manage this monopoly from exploiting their power. Mandating nondiscriminatory access in the form of common carriage has also been applied to inns, ports, stockyards, and grain elevators, to name a few. Most recently, the Federal Communications Commission under the Obama administration adopted common carriage rules in the form of “network neutrality,” prohibiting discrimination by Internet service providers. Introducing common carriage for platforms would be one way to tackle their gatekeeper power. A platform neutrality regime could require a platform to treat all commerce flowing through its infrastructure equally, preventing a platform from using the threat of discrimination to extract and extort.
A set of tools also exists to tackle leveraging power. Structural separations and prophylactic bans could limit the ability of dominant platforms to enter certain distinct lines of business. This, in turn, would limit the ability of dominant platforms to leverage their platform advantage into other areas. Structural separations preventing platforms from engaging in business activity that places them in direct competition with the firms using their platforms would also help eliminate the conflict of interest that platforms face when they own both the pipes and the products flowing through them. As with common carriage, structural separations have been a mainstay tool for tackling the power of network monopolies and other firms that play an infrastructure-like role in the economy. Structural bans have been applied to railroads, telecommunications carriers, TV networks, and banks. Introducing a separations regime for platforms would help prevent leveraging and eliminate a core conflict of interest currently embedded in the business model of dominant platforms.
Information exploitation power presents more of a challenge. To some extent, we have addressed information exploitation in the past, through disclosure regimes and laws requiring public auditing of privately collected information. But two aspects of platforms’ information exploitation power seem new. One is the sheer volume of information that these firms collect, and the security vulnerabilities created when a handful of platforms capture swaths of data. Partly the issue is structural: concentrated data is more vulnerable to security breaches than is that same data dispersed. Partly it comes down to business model: as digital advertising firms, Google and Facebook make money through collecting information. So long as their business models are surveillance-based, they will continue to collect as much information as possible. The other challenge that information exploitation poses is not to privacy but to competition. Gathering data on business activity that relies on the platform gives the platform an information advantage it can use to extort value from those businesses by harvesting their insights, or to thwart nascent rivals in ancillary lines of business.
Tackling information exploitation power is not straightforward. One idea is to regulate their conduct, limiting what information platforms collect and how they use it. This would include introducing privacy regulations like those adopted by Europe in its General Data Protection Regulation (GDPR) and prohibiting platforms from using information collected on their platforms to advantage distinct lines of business. But these forms of regulation risk proving ineffective unless we also address the underlying structure of platforms. Structural reforms would include: structuring competition in platform markets by undoing, for example, Facebook’s acquisition of Instagram and WhatsApp, prohibiting future acquisitions, and granting users ownership rights over their data; requiring social networks and search engines to spin off their ad networks, ending their surveillance-based business models; and prohibiting platforms from entering lines of business that depend on their platform (i.e., the kind of separations regime advocated above). By targeting the underlying structure and business model, these measures target the incentive and ability of platforms to collect and harness information.
The discussion around how to tackle platform power is just beginning. As the debate develops, it’s worth recalling that certain facets of platform power are not new, and that existing levers and concepts can be retooled to ensure that the platforms are structured to align with—and not undermine—open markets, fair competition, and the free flow of information.
Disclaimer: The ProMarket blog is dedicated to discussing how competition tends to be subverted by special interests. The posts represent the opinions of their writers, not necessarily those of the University of Chicago, the Booth School of Business, or its faculty. For more information, please visit ProMarket Blog Policy.