The Antitrust Assault on Ad Tech: A Law & Economics Critique

Executive Summary

For years, regulators and competition watchdogs have expressed concern about competition in the digital advertising business. They note that digital advertising appears to be dominated by a few dominant firms, such as Google, Facebook, and—to a lesser extent—Amazon. Some claim that this dominance allows these firms—and Google, in particular—to engage in anticompetitive conduct to extend their market power and to earn supercompetitive profit at the expense of advertisers, publishers, and consumers. This paper investigates the digital advertising market and assesses some of these claims. We conclude based on the information that is publicly available that many of the most significant claims made against Google’s advertising technology (ad tech) business are based on a misunderstanding of U.S. antitrust law, or of the details of the  ad tech market itself.

Digital advertising provides the economic underpinning for much of the Internet. Targeted digital advertising on independent websites is often facilitated by intermediaries that match advertisers and websites automatically, displaying ads to users for whom they are most relevant. This intermediation has advanced enormously over the past three decades. Some now allege, however, that the digital advertising market is monopolized by its largest participant: Google.

In particular, a lawsuit filed in December 2020 by the State of Texas and nine other U.S. states (later joined by five more states in March 2021) alleges anticompetitive conduct related to Google’s online display-advertising business. It has been reported that the U.S. Justice Department (DOJ) may bring a similar lawsuit before the end of 2022.

Moreover, a bipartisan group of U.S. senators introduced the Competition and Transparency in Digital Advertising Act in May 2022. If passed by Congress and signed into law, the bill would require some of the largest Internet firms—such as Google, Facebook, and Amazon—to break up their digital advertising businesses. A summary of the bill claims that Google is the “leading or dominant” firm in “every part” of the ad tech “stack” and that it uses this dominance to “extract monopoly rents” from advertisers and publishers.

Our paper focuses on the open-display digital advertising business. Display ads are designed to be visually engaging, combining text, images, and a hyperlink to a website. These are distinct from search ads, which are text ads displayed along with organic Internet search results. Most of today’s digital display advertising appears on heavily trafficked owned-and-operated sites such as TikTok, Instagram, YouTube, and Amazon, in which the company providing the advertising space is the same company that operates the platforms that place the ads. In contrast, open-display space is supplied by independent publishers—such as The New York Times (nytimes.com), MLB.com, or The Weather Channel (weather.com)—and is usually facilitated by intermediaries.

This paper begins with an overview of digital advertising. The market’s history is one of dynamic innovation, with many new developments arising to solve problems created by previous innovations, address new innovations, and respond to market developments. The market’s structure has changed dramatically as advertisers, publishers, and consumers have responded to new technologies. These changes and innovations all must balance the competing demands of advertisers, publishers, and consumers to maximize the total value of the advertising platform. Thus, any antitrust evaluation of digital advertising must consider whether the market’s structure and the conduct of its participants may be procompetitive responses to prior market changes, as well as the extent to which the overall market structure may mitigate superficially problematic elements within it.

Of particular importance, digital advertising intermediaries that are vertically integrated into some or all components of the digital advertising “stack” of services use the prices charged to each side of the market to optimize overall use of the platform. As a result, pricing in these markets operates differently than pricing in traditional markets: pricing on one side of the platform is often used to subsidize participation on another side of the market, increasing the value to all sides combined. Consequently, pricing (or other terms of exchange) on one side of the market may appear to diverge from the competitive level when viewed for that side alone. While one side of the market may pay superficially higher fees, this cost can be offset by the benefits from increased participation on the other side of the market. In this way, using subsidies to increase participation on another side of the market creates valuable network effects for the side of the market facing the higher fees.

In the second half of the paper, we address some of the specific allegations of anticompetitive structures and conduct made in the Texas Complaint and by critics of the digital advertising industry. We conclude that a flawed premise underlies many of these allegations. It is a version of the “big is bad” argument, in which conduct by dominant incumbent firms that makes competition more difficult for certain competitors is viewed as inherently anticompetitive—even if the conduct confers benefits on users. Under this approach, the largest firms are seen as acting anticompetitively if they do not share their innovations or reveal their business processes to competing firms. As a result, creating new and innovative products, lowering prices, reducing costs through vertical integration, and enhancing interoperability among existing products is miscast as anticompetitive conduct.

In contrast, we note that U.S. antitrust law is intended to foster innovation that creates benefits for consumers, including innovation by incumbents. The law does not proscribe efficiency-enhancing unilateral conduct on the grounds that it might also inconvenience competitors, or that there is some other arrangement that could be “even more” competitive.

Moreover, U.S. antitrust law does not second guess unilateral conduct simply because it may hinder rivals. Any such conduct would first have to be shown to be anticompetitive—that is, to harm consumers or competition, not merely certain competitors. In multisided markets, this means finding not simply that some firms on one side of the market are harmed, but that the combined net effect of challenged conduct across all sides of the market is harmful. The Texas Complaint, however, is built on the alleged harm of reduced revenue to publishers, without considering the corresponding benefit of lower prices to advertisers (and the consumers of advertised products and services).

Based on the information publicly available, we conclude that many of the most significant claims made against Google’s ad tech business are based on a misunderstanding of U.S. antitrust law, or of the details of the ad tech market itself. Although we cannot be sure how the Texas et al v. Google case will develop once its allegations are fleshed out into full arguments, many of its claims and assumptions appear wrongheaded. If the court rules in favor of these, the result will be to condemn procompetitive conduct and potentially to impose costly, inefficient remedies that function as a drag on innovation.

Legislators, too, who may be concerned about Google’s conduct and tempted to impose regulatory requirements on tech companies should bear in minds the risk of the “Nirvana fallacy,” in which real-life conduct is compared against a hypothetical “competition-maximizing” benchmark and anything that falls short is deemed worthy of intervention. That fanciful approach would pervert businesses’ incentives to innovate and compete and would make an unobtainable “perfect” that exists only in the minds of some economists and lawyers the enemy of a “good” that exists in the market.

Introduction

For years, regulators and competition watchdogs have expressed concern about competition in the digital advertising business. They note that digital advertising appears to be dominated by a few exceptionally large firms, such as Google, Facebook, and—to a lesser extent—Amazon. Some claim that this dominance allows these firms—and Google, in particular—to engage in anticompetitive conduct to extend their market power and to earn supercompetitive profits at the expense of advertisers, publishers, and consumers. This paper investigates the digital advertising market and assesses some of these claims. We conclude, based on the information that is publicly available, that many of the most significant claims made against Google’s advertising technology (ad tech) business are based on a misunderstanding of U.S. antitrust law, or of the details of the ad tech market itself.

Digital advertising provides the economic backbone for much of the Internet. By providing websites and apps a means to monetize their products without having to charge user fees, advertising enables access to entertaining and informative content without payment. Targeted advertising allows companies to inform potential customers of new products, giving new entrants a way to compete with popular incumbents, while effective targeting avoids wasting the time of those who aren’t likely to be interested. Advertising can endow products with new characteristics in customers’ minds and make consumers aware of product features they may not have known about.

Advertising on independent websites is often facilitated by intermediaries that match advertisers and websites automatically, targeting ads at the users to whom they are most relevant. This intermediation has advanced enormously over the past three decades. Some now allege, however, that the digital-advertising market is monopolized by its largest participant: Google. These allegations originate from various sources, including policy discussions, lawsuits, draft legislation, and academic reports.

In particular, a lawsuit filed in December 2020 by the State of Texas and nine other U.S. states (later joined by five more states in March 2021) alleges, among other things, anticompetitive conduct related to Google’s online display-advertising business.[1] This action is one of three currently pending lawsuits brought against Google by government antitrust enforcers in the United States; the other two relate to Google’s distribution agreements and search design.[2] It has been reported that the U.S. Department of Justice (DOJ) may bring a similar lawsuit before the end of 2022.[3] Along similar lines, the European Commission opened an investigation into Google’s display-advertising services in June 2021[4] and the German competition authority has published a report regarding its inquiry into non-search advertising.[5] Among other things, the European Commission is investigating whether Google “has made it harder for rival online advertising services to compete in the so-called ad tech stack.”[6]

These ongoing cases follow regulatory reports and hearings examining the market, including a year-long study by the United Kingdom’s Competition and Markets Authority (CMA). The CMA investigation of digital advertising (including search and social-media advertising) has thus far produced recommendations for a code of conduct and “pro-competitive interventions” into the market, as well as a new regulatory body to oversee these measures.[7] The Australian Competition and Consumer Commission is also conducting its own study of the digital advertising market,[8] and both houses of the U.S. Congress have held hearings on the market in recent years.[9] In October 2020, the Democratic majority staff of the U.S. House Judiciary Committee’s Subcommittee on Antitrust, Commercial, and Administrative Law issued a report that recommended, among other things, regulation for the display advertising market.[10]

The digital  advertising industry has also drawn the attention of legislators. In May 2022, a bipartisan group of U.S. senators introduced a bill that would require some of the largest Internet firms to break up their digital advertising businesses.[11] Dubbed the Competition and Transparency in Digital Advertising Act (CTDAA), the measure was introduced by Sen. Mike Lee (R-Utah) and co-sponsored by Sens. Amy Klobuchar (D-Minn.), Ted Cruz (R-Texas), and Richard Blumenthal (D-Conn.). Sen. Lee’s summary of the CTDAA identifies several allegations against the largest firms in the digital advertising business, with an emphasis on Google.[12] The summary claims that Google is the “leading or dominant” firm in “every part” of the “ad tech stack” and that it “exploits” conflicts of interest to “extract monopoly rents” from advertisers and publishers. Because of these monopoly rents, consumers are harmed in the form of higher prices for advertised goods and services and lower quality of online content, according to Lee’s press release.[13]

A 2020 paper published by the Omidyar Network—itself based on an interim CMA report produced during the authority’s then-pending investigation[14]—alleged anticompetitive practices within Google’s display advertising business and laid out a “roadmap” for a prospective antitrust case.[15] Other legal and economic consultants have also voiced concerns about Google’s role in the digital advertising industry.[16] These critiques were published before the Texas Complaint and provide more detail underlying the allegations and arguments described in the Texas Complaint. For that reason and because it may influence further lawsuits and regulatory interventions in the digital advertising market, including the DOJ’s, we also evaluate several of the Roadmap’s findings and conclusions.

This paper investigates the digital advertising market and assesses the aforementioned claims that it is uncompetitive. It explains some of the complex dynamics that underpin this market, thereby shedding light on the weaknesses and deficiencies in many of the arguments made about it, particularly those behind the Texas Complaint. This analysis is relevant to the entire Internet and to the wider economy, not just to Google and the display-advertising market. Many of the allegations made in the Texas Complaint would, if upheld by a court, have profound implications for antitrust law by establishing precedents that successful platforms effectively have a legal duty to act as essential facilities for their competitors; that efficiency-enhancing innovation by incumbent platforms is anticompetitive (particularly when it is not shared with competitors); and that courts or regulators can impose remedies that put these duties into effect without consideration of the harmful tradeoffs and unforeseen consequences that could themselves constrain competition and innovation. Such an approach would severely affect not only Google and the ad tech industry, but also businesses operating in unrelated markets and industries.

We begin with an overview and history of digital advertising. It is a history of dynamic innovation, with many new developments arising to solve problems created by previous innovations, address new innovations, and respond to market developments. These innovations must balance the competing demands of advertisers and publishers. The market’s structure has changed dramatically as advertisers, publishers, and consumers have responded to recent technologies. Because of this dynamism, we argue that it is a mistake to conclude that market structure and firm conduct at some specific point in time was ideal or better from a competition point of view, or that deviations from that paradigm represent a problem for competition enforcers to correct.

In the second half of the paper, we address some of the specific allegations of anticompetitive structures and conduct made in the Texas Complaint and by critics of the digital advertising industry. We conclude that a flawed premise underlies many of these allegations. Fundamentally, the allegations stem from an assertion that conduct engaged in by dominant, incumbent firms that makes competition more difficult for competitors is anticompetitive—even if the conduct confers benefits on users. This often amounts to a claim that the largest firms are acting anticompetitively by innovating and developing their business processes and products in ways that create benefits for one or more digital advertising constituents and for the advertising ecosystem more generally, including by creating new and innovative products, lowering prices, reducing costs through vertical integration, and enhancing interoperability between existing products.

I. Overview of the Digital Advertising Industry

For many people, digital ads are “just there.” They appear on one’s Facebook timeline, get inserted into one’s Twitter feed, or show up in the middle of an online news article. Unseen to most users is a complex stack of services that match advertisers with advertising space, using real-time auctions and other algorithms to deliver ads targeted to produce a user response, such as buying a product, supporting a cause, or visiting another website.

In this section we explain that digital advertising is just one small part of a much broader advertising and marketing industry. We provide a concise history of how the business evolved from simple banner ads to highly targeted display ads. Through this evolution, digital advertising has become a multisided market where intermediaries must balance the demands of advertisers, publishers, and users to maximize the total value of the advertising platform. Because of this balancing, it is a mistake for policymakers and regulators to focus only on a single set of users or a single segment of the stack of digital advertising services.

A.   Digital advertising is part of a broader advertising and marketing market

The Texas Complaint alleges that Google has market power in six distinct product markets, each of which the states claim to be nationwide in geographic scope. Four of these distinct product markets involve intermediation in the sale of “open” display ads on third-party websites and two involve intermediation in the sale of “in-app” display ads on mobile devices.[17] As we note in our earlier paper, it is likely that the states allege overly narrow product-market definitions.[18] In particular, structure and conduct viewed within a broader digital advertising market, overall advertising market, or marketing market indicates than no single firm has significant market power.

Digital advertising comprises about half of U.S. advertising revenues (Figure 1), while advertising itself accounts for about half of marketing activities. Marketing includes advertising, as well as events, sales promotion, direct marketing, telemarketing, product placement, and merchandising. Within digital advertising, advertisers have a broad set of options about where and how to run ads, including:

• Search ads, in which the ad is displayed as a search-engine result (e.g., Google, Bing, DuckDuckGo);
• Display ads on a site owned and operated by the firm that sells the ad space (e.g., Facebook, YouTube, Amazon Marketplace);
• “Open” display ads on a third party’s site (e.g., The New York Times, Dallas Morning News, Runner’s World); or
• “In-app” display ads served on mobile apps.

While total advertising spending in the United States has increased by about 15% since 2009, as a share of the economy spending has been relatively flat at slightly less than 1% of GDP.[19] As mentioned, about half of total U.S. advertising dollars go to digital channels, up from approximately 10% a decade ago. Approximately 30% of ad spending goes to TV and less than one-quarter goes to radio, newspapers, magazines, billboards, and other “offline” forms of media.[20]

Figure 1: U.S. Advertising Spending Over Time

Source: Benedict Evans, News by the Ton

It is well-understood that television broadcasters and cable networks compete with digital services.[21] And they do so on virtually all dimensions: for user attention, for labor, for content and other inputs—and for advertising. The same is true of competition for advertising among digital publications, newspapers, radio, magazines, video games, music streaming, and podcasts. The fact that offline and online advertising—to say nothing of marketing more broadly—employ distinct processes does not consign them to separate markets. Indeed, it is widely accepted that online advertising has drawn advertisers from offline markets, as previous technological innovations drew advertisers from other channels before them.[22] Moreover, while there is evidence that, in some cases, offline and online advertising may be complements as well as substitutes,[23] the distinction between these cases is becoming less and less meaningful as the revolution in measuring the effectiveness of advertising has changed how marketers approach different levels of what is known as the marketing “funnel.”[24] For example, economist David Evans’ review of the literature concludes that digital advertising is a segment of the broader advertising business in which different forms of advertising compete and complement each other:

Advertisers base decisions about the level and allocation of their budgets on formal or informal analyses of the rate of return on investment. For these ad campaigns, the different advertising methods can be substitutes to the extent they provide alternative ways of delivering messages to an audience, and complements to the extent they can reinforce each other.[25]

Economists Avi Goldfarb and Catherine Tucker demonstrate that display-advertising pricing is sensitive to the availability of offline alternatives.[26] Although advertising technology and both supplier and consumer preferences continue to evolve, the weight of evidence suggests a far more unified and integrated economically relevant market between offline and online advertising than their common semantic separation would suggest:

We believe our studies refute the hypothesis that online and offline advertising markets operate independently and suggest a default position of substitution. Online and offline advertising markets appear to be closely related. That said, it is important not to draw any firm conclusions based on historical behavior.[27]

In summary, there is evidence that open-display and in-app ads compete with search ads, while digital ads compete with offline advertising. Thus, courts and regulators should be skeptical of overly narrow market definitions focused on only small slices of a much larger relevant market for advertising.

B.   A simplified description of digital display advertising

The combination of software and processes that facilitate digital advertising transactions is known as the “ad tech stack.” The stack consists of several software components to match advertisers with publishers.

Advertiser ad servers are used by advertisers and media agencies to store ads, deliver them to publishers, track their activity, and assess their effectiveness (by, for example, tracking conversions). Demand-side platforms (DSPs) automate the purchase of advertising inventory by collecting bids in real-time auctions from multiple advertiser ad servers. DSP bids are based on the advertiser’s objectives, data about the final user, and data on impressions or conversions. Publisher ad servers manage publishers’ inventory and determine whether and where to serve a particular ad on a publisher’s site. Supply-side platforms (SSPs) automate the sale of publishers’ inventory, typically through real-time auctions involving multiple DSPs.

In general, the process of buying and selling digital ads through open-display auctions works as follows (Figure 2):

1. When a user opens a webpage (or uses an app), the publisher’s ad server sends a bid request to SSPs for the advertising impressions available on that page for that user.
2. The SSPs send bid requests to multiple DSPs for the ad impressions.
3. DSPs evaluate the advertising opportunity based on user data and the objectives of their advertisers’ campaigns and send bids to the SSPs.
4. SSPs rank the bids received based on price and other priorities set by the publisher. The SSPs send their winning bids to the publisher.
5. The publisher ad server compares bids received from SSPs, along with any pre-existing direct deals between the publisher and specific advertisers and decides which ad to serve on the page.

Figure 2: A Simplified View of the Digital Advertising Stack

Source: OECD, CMA

While this process applies to most programmatic transactions, there are many variations. For example, there are diverse ways in which SSPs are contacted and asked to submit their bids. To the extent that a publisher and advertiser have a pre-existing, direct agreement, there are differences in how these arrangements are handled and integrated with deals arranged through automated platforms. The specific approach used to match ads with ad inventory will reflect a balance among different sides of a multisided market. One approach might increase the prices received by sellers (publishers) but expose buyers (advertisers) to increased risk of overpayment. Other methods might reduce risks to advertisers, but also reduce the prices received by publishers.

C.   A brief history of digital advertising

This history of the digital advertising market is a history of iterative innovation, with new developments and services arising to solve problems created by previous innovations and to respond to changing market conditions. At the heart of these innovations is an attempt to balance the competing demands of advertisers, publishers, and consumers. Given that this is a dynamic market, it would be mistaken to conclude that the market structure at some specific point was the “correct” one from a competition point of view. Moreover, it would be a mistake to conclude that deviations from some previous “ideal” world present a problem that can be corrected by disruptive regulation.

Digital advertising originally worked similarly to conventional print and broadcast advertising. Online publishers would negotiate with advertisers (or their ad agencies) to sell ad space on their websites, giving relevant advertisers information about their readership gathered through market research. All users would see the same ads. The resulting system was poorly targeted, inefficient, and carried high fixed costs, including the cost of things like market research and the transaction costs of publishers hiring sales teams and advertisers hiring ad agencies to do business with one another. Inevitably, these fixed costs meant that only larger publishers and larger advertisers could engage in the online market profitably.

In 1993, O’Reilly & Associates Inc. introduced its Global Network Navigator (GNN) magazine and other ad-supported online publications, which first rolled out clickable ads. O’Reilly is credited with the first attempt to create an “advertising medium” on the Internet.[28] The price of ads ranged from $500 for a one-page business profile to $5,000 for up to 25 pages about the company placing the ad.[29] A year later, Wired magazine’s digital affiliate HotWired ran what later became known as the web’s first banner ads. The ad—for AT&T’s “You Will” campaign—cost $30,000 for a three-month dedicated placement in a section of HotWired’s site.[30]

The first step toward automating this process came with the introduction of ad-server software on both the publisher and advertiser sides, which allowed each side to automate parts of the ad-placement process. Ad servers allowed publishers to automatically describe the type of content on their pages, which in turn allowed advertisers to place ads tailored to that content. An article about hiking could automatically indicate to a department store to place ads selling walking boots, for instance. It also allowed the publisher to sell to many advertisers without having to transact directly with any of them. Ad servers also allowed advertisers to browse and manage campaigns across a large, aggregated number of publishing sites, instead of having to interact with sales teams one by one. This process was, however, still negotiated directly, and often left publishers with unused “remnant” advertising space that they had not been able to sell.

To solve this problem, ad networks entered the market. These functioned as intermediaries between advertisers’ and publishers’ ad servers, aggregating unsold remnant ad space and allowing advertisers to buy that space en masse without having to deal directly with each publisher. Ad networks did not replace direct sales, but they allowed for residual space to be bought and sold more easily, increasing the amount of ad inventory available and lowering the fixed costs to use it. This, in turn, made the market feasible for smaller publishers, who would otherwise be unable to attract direct deals to sell ad space, and for advertisers to conduct large-scale ad campaigns across many publishers (including small ones).

In 1995, GNN was acquired by AOL.[31] That same year, marketing-communications agency Poppe Tyson spun off its Internet advertising division as DoubleClick, with the objective of “responding to advertisers’ need to be able to buy millions of impressions on the Internet without having to buy from hundreds of different sites.”[32] The company created “subnets” of publishers to target specific categories of consumers.[33]

Also in 1995, ad agency WebConnect, the first ad network, began to collaborate with its clients to identify the websites that their ideal consumers visit. WebConnect then placed ads on the websites where they were more likely to be seen by the audience most relevant to their clients. The company also produced a tool to prevent “ad fatigue,” which occurs when users are repeatedly shown the same ad.[34]

In 1998, GoTo.com, which was renamed Overture in 2001, launched its ad-supported search engine.[35] Search result rankings were based on an open-market bidding process. Advertisers on GoTo were informed of the amounts other advertisers were bidding for a click-through within the results for a given search term, and any advertiser could increase its bid to obtain a higher ranking, a process GoTo described as “pay for performance.”[36] One of GoTo’s key innovations was linking advertising pricing to click-throughs, rather than to page views.[37]

To drive home just how efficient these ads could be, [GoTo.com’s founder Bill Gross] came up with an audacious pricing scheme. Instead of paying for page-views—an old-media model that had come to dominate the Web—advertisers would pay only when people actually clicked on their ads. And their placement on the GoTo.com results page would be determined through an auction, so that more desirable keywords would command higher prices, while less common keywords could be had for as little as a penny per click. As a search engine, GoTo.com had nothing on Google. But as a way of making money on searches, it was ingenious.[38]

During the dot-com boom of the late 1990s, banner ads spread throughout the Internet, though growth was tempered by user complaints that the ads slowed page loading.[39] The dotcom bust wiped out many of the firms that were the biggest buyers of digital banner ads. In response, Wired predicted that digital advertising would undergo a “facelift.”[40]

Old media revenues will wither as mainstream advertisers storm the Net. Instead of stuffing junk mail into the mailbox outside your house, they’ll send it directly to your inbox. And companies get smarter, choosing sites that take better aim at their quarry, er, potential customers.

“It’s very much a targeted medium,” Robin Neilfield, co-founder of NetPlus Marketing. “You have to analyze the content on a site, you can’t just buy based on demographics.”[41]

Because ad networks were not comprehensive—they did not carry the entire inventory of the Internet—publishers began to use yield managers (later known as SSPs) to compare bids for their ad space and to decide which to accept. SSPs effectively allowed publishers to aggregate demand from a larger number of ad networks, which themselves aggregated demand from advertisers. This process allowed ad space to be more easily commoditized, with an SSP helping to identify an ad space’s relevance to potential advertisers.

As indirect sales became possible, ad exchanges emerged to sell ad space using real-time auctions. Ad space could be tagged according to characteristics like audience type, relevance to the advertiser, and/or prominence and quality of the ad, with bids gauged accordingly.

Finally, DSPs on the advertiser side allowed advertisers to engage with many ad networks and ad exchanges at one time. These also allowed advertisers to track campaigns and measure performance of different ads with different publishers, and to adjust their campaigns accordingly. Most ad exchanges now have DSP functionality built in.

In 2000, Google introduced a new self-service advertising product called AdWords (now Google Ads) that allowed businesses to purchase text ads on search-results pages. At the time it was reported that AdWords was designed to attract small-to-midsize advertisers with budgets of $5,000 or less.[42] AdWords differed from GoTo/Overture in a major way. GoTo/Overture placed ads within search results, with results ranked by bid. In contrast, AdWords placed ads separate from search results with pricing based on pageviews.[43] In this way, Google could display ads without compromising the relevance of search results. In 2002, it launched AdWords Select, its pay-per-click, auction-based search-advertising product.[44]

In 2003, Google acquired Applied Semantics, whose AdSense display advertising product allowed it to sell targeted ads on third-party websites.[45] With AdSense, the display-ad server was able to read text on a publisher’s site and serve relevant ads, considering factors like the user’s geographic location, age, demographics, and the search made.[46] AdSense was the forerunner of programmatic display advertising, the process of automating the buying and selling of ad inventory in real time through an automated bidding system. In 2005, Google introduced the Quality Score model, which considers an ad’s click-through rate, as well as the bid price, in placing ads.[47]

YouTube was launched in 2005 and acquired by Google the following year, when the company also introduced video ads. In 2007, Google acquired DoubleClick.[48]

Around 2015, “header bidding” began to roll out, with publishers Meredith Corp. and Townhall Media as two of largest early adopters.[49] Before header bidding, it was difficult for every demand-side partner to submit a bid for every ad request. As a result, publishers relied on approaches such as “ad waterfalls”[50] to try to get the most from each partner. Because of the way ad waterfalls are configured (based on historical, not real-time, data), publishers believed ad waterfalling led to winning bids that were below what some bidders might be willing to pay.[51] Client-side[52] header bidding was adopted as a way to increase real-time price competition among multiple SSPs, leading to higher returns for publishers and a more efficient allocation of ad space to advertisers.

Despite the widespread adoption of header bidding—as of the second quarter of 2021, about two-thirds of publishers were using it[53]—the technology has its own challenges. For example, the addition of extra code on the webpage, which client-side header bidding requires, can slow down the publisher’s website, driving away users.[54]

As an alternative to client-side header bidding, server-side header bidding was introduced. Prebid launched in 2015 as an independent and open-source option. Google released Open Bidding in April 2016 and Amazon introduced Transparent Ad Marketplace (“TAM”) at the end of 2016. In these alternatives, the auction among SSPs takes place in a remote server controlled by a third party (the provider of the server-side header-bidding solution) instead of in the user’s browser. This helps to improve site-load speed. On the other hand, this solution leads to less revenue for publishers and reduces the availability of data to advertisers and publishers.[55]

Over the past decade, the price of digital advertising has fallen steadily, while output has risen. U.S. digital-ad spending grew from $26 billion in 2010 to $189 billion in 2021, an average annual increase of 20%.[56] Over the same period, the Producer Price Index for Internet-advertising sales declined by an annual average of 4%.[57] The rise in spending in the face of falling prices indicates that the number of ads bought and sold increased by approximately 25% annually. The combination of increasing quantity, decreasing cost, and increasing total revenues are consistent with a growing and increasingly competitive market, rather than one of rising concentration and reduced competition.

D.   Digital advertising is a multisided market

The digital advertising market can be thought of as a complex multi-step and multisided market that involves three key parties—advertisers, publishers, and intermediaries—and is aimed at a fourth: consumers. In contrast, critics of the current structure of and conduct in the digital advertising industry have characterized it as a “straightforward and traditional” market in which publishers supply an inventory of ad space and advertisers are buyers of the ad space.[58] In this simplistic account of the market, for a given supply of inventory, publishers would seek to maximize the price received per ad, while advertisers would seek to minimize the price paid per ad. Targeting of ads would be based on the demographics of a publisher’s readership or the content of the publication, rather than the individual characteristics of each reader. In general, this is how the market initially operated before the introduction of clickable ads. But even this simple formulation is quite complex. Advertisers expect to maximize the return on their investment in advertising. Even at a low price, advertising expenditures would be wasted if that investment were not converted to increased sales of the advertiser’s product or service.

The invention of clickable ads with which users could interact changed the objective function of digital advertising. Publisher revenues and advertising costs became linked to individual consumers acting on an ad by, for example, clicking on it. Rather than paying or receiving a price-per-ad based on the size of a publication’s user base, advertising expenditures became a function of a price-per-click (or other action) and the number of clicks. This meant that the rewards for relevance—as well as the complexities of determining relevance—were greater because some viewers might be persuaded to act there and then.

In this multisided market, ad intermediaries must balance the interests of at least three constituencies: (1) advertisers creating ads and placing them; (2) publishers defining inventory and displaying ads; and (3) users consuming published content who view and act on ads. Intermediaries in these markets often benefit from network effects, through which the value of the platform to each user depends in part on the number and quality of other users on the platform.[59]

The quality and relevance of users is assessed by collecting information on the users as they browse the web. This information can include which ads they have viewed and clicked in the past, their geographical location, as well as their demographics, financial situation, and topics of interest. Broadly speaking, a larger network with diverse users provides more information and is better able to target ads to relevant users, benefiting advertisers, publishers, and consumers.

Network effects are not always positive, however, nor are they always captured by the platform that facilitates them.[60] While access to consumer data can help to improve the quality of the ads displayed—and increase the value of those ads to advertisers and publishers—claims that such access provides increasing returns to scale are not borne out by the burgeoning empirical literature on the topic. Summarizing these empirical findings, economist Catherine Tucker concludes that “empirically there is little evidence of economies of scale and scope in digital data in the instances where one would expect to find them.”[61]

Intermediaries in multisided markets often face difficult optimization problems caused by the interrelated demands of participants on different sides of the market, each group of whom benefits from the existence and size of the other, but whose interests conflict across many margins.[62] This highlights the key distinction between “straightforward and traditional” markets and multisided markets.

Ad tech intermediaries that are vertically integrated into some or all components of the ad tech stack use prices charged to each side of the market to optimize overall use of the platform. As a result, pricing in these markets operates differently than pricing in traditional markets. Pricing on one side of the platform is often used to subsidize participation on another side of the market, increasing the value to all sides combined. Consequently, pricing (or other terms of exchange) may appear to one side of the market to diverge from the competitive level when viewed for that side alone. While one side of the market may pay higher fees, this cost can be offset by the benefits from increased participation on the other side of the market. Thus, using subsidies to increase participation on another side of the market creates valuable network benefits for the side of the market facing the higher fees.

For example, among the criticisms of digital advertising business practices is the use of “second-price auctions” rather than “first-price auctions.”[63] First-price auctions are those most familiar to people: multiple bidders offer prices, and the highest bidder wins the auction and pays an amount equal to her winning bid. In a second-price auction, the highest bidder wins the auction but pays an amount based on the next-highest bid. In markets with many bidders possessing the same information, first-price auctions and second-price auctions would be expected to produce the same amount of revenue under the well-known auction-theory concept of revenue equivalence.[64]

The choice of auction approach reflects the tensions between different sides of the market in a multisided market. On the one hand, under certain circumstances, a first-price auction tends to increase the prices received by sellers (here, publishers), but exposes buyers (here, advertisers) to an increased risk of overpayment.[65] On the other hand, under certain conditions, a second-price auction reduces risks to advertisers, but also reduces the prices received by publishers.[66] It would be expected that an ad tech intermediary would balance these competing interests to maximize total revenues flowing through the ad tech stack, to maximize its profitability. In such a multisided market, it would be a mistake to focus only on one side of the market and ignore the effects that decisions such as this have on the other participants.

The extent to which ad tech intermediaries—in particular, vertically integrated services like Google’s—act to optimize the overall value of the platform is critical to understanding how these markets work. It also highlights how misleading it can be to assume that these processes can be analyzed as “straightforward and traditional” markets.

II. Antitrust Primer: Effective Competition Is not an Antitrust Offense

A flawed premise underlies much of the Texas Complaint, the Omidyar Network’s Roadmap report, and the CTDAA legislation. Fundamentally, most of the charges that each of these level against Google and Facebook’s ad tech businesses derive from an assertion that conduct engaged in by dominant incumbent firms that makes competition more difficult for competitors is anticompetitive—even if the conduct confers benefits on users. This often amounts to a claim that the largest firms are acting anticompetitively by innovating and developing their business processes and products in ways that create benefits for one or more digital advertising constituents and for the advertising ecosystem more generally. These include creating new and innovative products, lowering prices, reducing costs through vertical integration, and enhancing interoperability between existing products, among other things.

This approach entails an argument—made explicit in the Texas Complaint and the Omidyar Roadmap—that Google harms competition by creating obstacles for rivals without offsetting “incremental efficiencies.”67F[67] According to the report, this means that, even if Google’s practices produce benefits for such constituents as advertisers, publishers, or consumers, they could possibly be reimagined to create even more competition or achieve the same benefits in ways that better prop up rivals. According to the Roadmap, the practices should therefore be condemned as anticompetitive.68F[68]

But that is not how the law—or the economics—works. Such an approach converts manifestly beneficial aspects of Google’s ad tech business into anticompetitive defects, essentially arguing that successful competition creates barriers to entry that merit correction through antitrust enforcement. The CTDAA takes this argument a step further by imposing “best interests,” “best execution,” and “transparency” obligations on large firms and mandating divesture of parts of the largest firms to facilitate more entry by competitors. This approach turns U.S. antitrust law (and basic economics) on its head. As some of the most famous words of U.S. antitrust jurisprudence have it:

A market may, for example, be so limited that it is impossible to produce at all and meet the cost of production except by a plant large enough to supply the whole demand. Or there may be changes in taste or in cost which drive out all but one purveyor. A single producer may be the survivor out of a group of active competitors, merely by virtue of his superior skill, foresight and industry. In such cases a strong argument can be made that, although, the result may expose the public to the evils of monopoly, the Act does not mean to condemn the resultant of those very forces which it is its prime object to foster: finis opus coronat. The successful competitor, having been urged to compete, must not be turned upon when he wins.69F[69]

U.S. antitrust law is intended to foster innovation that creates benefits for consumers, including innovation by incumbents. The law does not proscribe efficiency-enhancing unilateral conduct on the grounds that it might also inconvenience competitors, or that there is some other arrangement that could be “even more” competitive. Under U.S. antitrust law, firms are “under no duty to help [competitors] survive or expand.”70F[70]

To be sure, the arguments are couched in terms of anticompetitive effect rather than being described merely as commercial disagreements over the distribution of profits. But these effects are simply inferred, based on assumptions that Google and Facebook’s vertically integrated business models entail an inherent ability and incentive to harm rivals. For example, Google is alleged to be able to surreptitiously derive benefits from display advertisers by “leveraging” its search-advertising capabilities,71F[71] or by “withholding YouTube inventory,”72F[72] rather than altruistically opening it up to rival ad networks, or by using its access to data to improve its products without sharing that data with competitors.

All these charges may be true, but none is inherently anticompetitive. Under U.S. law, companies are not obligated to deal with rivals and certainly are not obligated to do so on rivals’ preferred terms.73F[73] In the Texas Complaint, for example, the court, citing Charych v. Siriusware, noted, “[D]efendants were under no obligation to develop an interface that was compatible with plaintiffs’ product.”74F[74] As long ago as 1919, the U.S. Supreme Court held that “[i]n the absence of any purpose to create or maintain a monopoly, the [Sherman Act] does not restrict the long recognized right of [a] trader or manufacturer engaged in an entirely private business, freely to exercise his own independent discretion as to parties with whom he will deal.”75F[75] More recently (in 2004) the Court held:

Firms may acquire monopoly power by establishing an infrastructure that renders them uniquely suited to serve their customers. Compelling such firms to share the source of their advantage is in some tension with the underlying purpose of antitrust law, since it may lessen the incentive for the monopolist, the rival, or both to invest in those economically beneficial facilities.76F[76]

Moreover, U.S. antitrust law does not second guess unilateral conduct simply because it may hinder rivals. Any such conduct would first have to be shown to be anticompetitive—that is, to harm consumers or competition, not merely certain competitors.77F[77] In two-sided markets, this means finding not simply that some firms on one side of the market were harmed, but rather that the combined net effect of challenged conduct across all sides of the market was harmful.78F[78]

In the platform context, understanding whether there is harm to competition at all requires an assessment of the effects of conduct on all sides of the platform. “[N]o economic basis exists for establishing a presumption that ‘harm’ on one side of a two-sided platform is sufficient to demonstrate that market output has been restricted, or that consumer welfare has otherwise been harmed.” In fact, “[s]eparating the two markets allows legitimate competitive activities in the market for general purposes to be penalized no matter how output-enhancing such activities may be.”79F[79]

The Texas Complaint, however, is built on the alleged harm of reduced revenue to publishers, without considering the corresponding benefit of lower prices to advertisers, or the net effect on consumers.80F[80]

Beyond that, antitrust law does not condemn conduct on the basis that an enforcer (or a court) is able to identify or hypothesize alternative conduct that might provide similar benefits at lower cost. In alleging that there are ostensibly “better” ways that Google could have pursued its product design, pricing, and terms of dealing, both the Texas Complaint and Omidyar Roadmap do just that—assert that, had the firm only selected a different path, an alternative could have produced even more benefits or an even more competitive structure. This line of thinking seems to be one motivation for the CTDAA’s remedies.

The reason that the possibility of “better” theoretical arrangements cannot serve as the basis for antitrust intervention is that there are limits to what can be achieved through intervention, not least because of limitations on legislators’ and enforcers’ knowledge about the competitive dynamics of the markets they seek to regulate.81F[81] A practice’s departure from a theoretical competitive benchmark may be inextricably linked to the social benefits it generates. When this is the case, enforcement that requires the practice or product to change in order to adhere to a theoretical standard may end up undermining the benefits of the practice in the first place. That is particularly true in the context of the sort of “vertical foreclosure” arguments leveled against Google in the advertising space, in which it is alleged that the combination of different levels of the ad-supply chain by Google limits the ability of competitors to enter and compete effectively.82F[82] It is surely conceivable that the product improvements conferred by the combination of different functions into a single platform—e.g., greater efficiency, realization of network effects, more effective targeting—could be replicated by a different means that might also facilitate “even more competition.” But such an approach is fraught with the risk of serious and costly error.83F[83]

The alleged cure of tinkering with benefit-producing unilateral conduct by applying an “even more competition” benchmark is worse than the supposed disease. The adjudicator is likely to misapply such a benchmark, deterring the very conduct the law seeks to promote. As then-Judge Stephen Breyer explained in the context of above-cost low pricing (another “defect” that both the Texas Complaint and the Roadmap claim constitutes anticompetitive conduct by Google84F[84]), “the consequence of a mistake” is “to penalize a procompetitive price cut,” conduct that, from an antitrust perspective, is “the most desirable activity.”85F[85] That commentators or enforcers may be able to imagine alternative, theoretically more desirable, conduct is beside the point.

Similarly, subjecting the kinds of product-design decisions at issue in the Google case to refined balancing of benefits and harms would deter innovation. “To weigh the benefits of an improved product design against the resulting injuries to competitors is not just unwise, it is unadministrable. There are no criteria that courts can use to calculate the ‘right’ amount of innovation, which would maximize social gains and minimize competitive injury.”86F[86] Put simply, “no one can determine with any reasonable assurance whether one product is ‘superior’ to another.”87F[87]

For these reasons, a “product improvement by itself does not violate Section 2, even if it is performed by a monopolist and harms competitors as a result.”88F[88] “Any other conclusion would unjustifiably deter the development and introduction of those new technologies so essential to the continued progress of our economy.”89F[89] A benefit-creating product design, even if it hinders rivals, is “necessarily tolerated by the antitrust laws.”90F[90]

Nor does U.S. law condemn a firm’s decision not to share a product improvement with rivals on terms rivals might prefer, even when such sharing might lead to greater competition in the short term. “Compelling” innovators “to share the source of their advantage” with rivals, among other evils, “may lessen the incentive for the monopolist, or rival, or both” to invest in innovation.91F[91] Except in extremely limited circumstances, firms can decide the terms on which they offer their products and services.

Directly rejecting the Roadmap’s suggestion—and the CTDAA’s mandate—of compelling dealings on terms that might produce greater competition, the Supreme Court has decreed that the “Sherman Act . . . does not give judges carte blanche to insist that a monopolist alter its way of doing business whenever some other approach might yield greater competition.”92F[92] Firms are not obliged to go into new lines of business or abandon existing lines to throw lifelines to rivals.93F[93]

The law similarly encourages vertical integration, because it tends to foster innovation-enhancing synergies and lower prices by eliminating double marginalization.94F[94] As the Roadmap elsewhere admits, it is “not in itself uncommon” to see vertical integration result in “fewer and fewer companies,” even “in competitive markets.”95F[95] Thus, vertical integration by internal expansion—even by a monopolist—is presumptively lawful.96F[96] The Roadmap and the CTDAA, however, simply disregard this, instead presumptively condemning vertical integration that hinders rivals by creating efficiencies.97F[97] Again, this is simply not a defensible interpretation of U.S. antitrust law, nor should it be.

III. Allegations Against Google in Digital Display Advertising

Critics of the digital advertising industry—and Google’s role in it—have leveled numerous allegations. These include claims that Google “leverages” its ownership of YouTube to obtain and exert market power in the buying and selling of other digital-display ads. Some claim that Google anticompetitively uses cross-subsidies, charging supercompetitive prices at one end of the ad tech stack to subsidize supra-competitive prices at another end of the stack. It is also alleged that Google has superior information about consumers that it will not provide to competitors, giving Google an anticompetitive advantage. It is claimed that steep entry barriers—some allegedly erected by Google—inhibit entry and allow the company to achieve a supercompetitive “take rate” from its intermediation services. While the lawsuits may provide additional information and data to support these claims, we argue that, with the limited public information available to us, it is not clear that any of them constitute anticompetitive conduct.

A.   ‘Leveraging’ market power in video streaming into the digital open-display market

The Omidyar Roadmap argues that Google, by virtue of its vertical integration throughout the intermediary stack and into the supply side (as the owner of YouTube), has the incentive and ability to derive unwarranted benefits from its display advertising business. It alleges, for example, that, by offering a single interface for placing both search and display ads, “Google leverages its monopoly position in search to coerce advertisers into using Google’s display products.”98F[98] In support it cites the CMA as saying:

Google may also be able to leverage its market power in search into the open display market. Smaller advertisers often choose to single-home to minimize transaction costs. Advertisers that wish to single-home have a strong incentive to use Google Ads as they can use it to access Google search advertising and YouTube inventory as well as the open display market.99F^[99]

An earlier version of the Texas Complaint echoed these claims:

Google’s practice of withholding YouTube video inventory from rival ad buying tools… effectively locks single-homing small advertisers into Google’s ad buying tool. In addition, other providers of ad buying tools indicate that it does not make economic sense to try to compete with Google Ads for small advertisers, because they cannot achieve sufficient scale with smaller advertisers who want to buy display, YouTube, and even search ads, through just one tool.100F[100]

And, similarly, the Roadmap also argues that most sources of demand for Google Ads purchase ad space through AdX because Google “designed its exchange in such a way that it operates more efficiently with requests from Google’s own ad server than it does when requests come from rival ad servers.”101F^[101]

All these assertions describe efficiency-enhancing behavior as anticompetitive. The report does not allege that Google preferences its own ad exchange in ways that harm advertisers; rather, the company’s products simply work better together (which is not unusual when different software products must interact) and it is thus in advertisers’ best interests for Google to act this way.

U.S. law, rightly, does not consider efficiencies obtained from vertical integration in this way to be anticompetitive. Nor do efficiencies that rivals cannot beat qualify as “barriers to entry.” The alternative—requiring Google to refrain from using the cheapest and/or fastest option available, because doing so makes its product better than all competitors—would mean reduced innovation, higher overall costs, and no benefit to either advertisers or publishers.

Later, the Roadmap makes another similar allegation: that Google “leverages” its ownership of YouTube, and the fact that only Google’s DSP can place ads on YouTube, to give itself an anticompetitive advantage in open-display advertising because rival DSPs are inherently limited by being unable to place ads on YouTube. An earlier version of the Texas Complaint echoed this claim.102F[102]

The Roadmap characterizes this conduct as “a contractual way to deny interoperability,”103F[103] but there is no contractual restraint here. How Google distributes YouTube’s ad inventory is a unilateral distribution decision permitted under U.S. law. And Google’s policy is not unusual in any way: many other websites that carry video advertising—including Hulu, Instagram, and Twitter—self-distribute their own inventory and do not make it available for resale by third parties.104F[104] Google does not have a duty to maximize its competitors’ profits by allowing them to resell YouTube inventory.

Access to YouTube is also not essential to a DSP’s success. Before Google stopped third-party platforms from buying YouTube ad inventory, it reported that only a “small amount” of buying was being done through Google’s AdX, which allowed third-party platforms to bid. At the time, AdExchanger reported that “[b]y ’small amount,’ that reportedly means 5%.”105F[105] A competing DSP, TubeMogul, said that this decision was an “unfortunate development” but “immaterial, since less than 5% of total ad spend through our software in Q2 was directed to YouTube.”106F[106]

This is consistent with the fact that there are several successful DSP competitors that compete with Google, despite not having access to YouTube’s ad inventory. The Trade Desk went public for $1 billion in 2016, processed more than $6.2 billion in transactions in 2021, and had a market cap of more than $25 billion in the first week of August 2022. 107F[107] Other DSPs, like Amazon’s and Xandr (formerly AppNexus), both continue to compete with Google vigorously without access to YouTube inventory, as the Omidyar Roadmap admits in the case of AppNexus.108F[108]

The Roadmap further -alleges that Google’s owned-and-operated properties—including Search, YouTube, Shopping, Flights, and News—confer an anticompetitive advantage because “Google pays no ‘traffic acquisition costs’” for the advertising space on its own sites: “When Google places ads on YouTube, just as when it places ads on its own search results pages, Google pays no ‘traffic acquisition costs’ because it needn’t pay any publisher for access to the ‘eyeballs’ that will see or interact with the ads it helps place.”109F^[109]

Google’s parent Alphabet reported that the company’s traffic-acquisition costs were approximately 20% of its revenues in 2021.110F^[110] Over the past few years, 40-50% of Alphabet’s expenditures have been on “cost of revenues,”111F[111] and of these, roughly half have involved traffic-acquisition costs.112F[112] Alphabet defines traffic-acquisition costs as (a) “the amounts paid to our distribution partners who make available our search access points and services” and (b) amounts paid for ads displayed on Google Network Members properties. It identifies “distribution partners” as browser providers, mobile carriers, original equipment manufacturers, and software developers.

Contrary to the Roadmap’s insinuations, there is nothing to suggest that these expenditures become less burdensome as a company increases in scale. Indeed, the opposite may be true, if it is more costly to gain access to marginal users than inframarginal ones, consistent with Google’s traffic-acquisition costs increasing over the years as it has grown.113F[113] While Google does not have to pay itself for the use of its own display inventory, there is clearly an opportunity cost to displaying its own inventory rather that of another firm. The claim that the company faces no traffic-acquisition costs for these properties is inaccurate.

The Roadmap’s focus on traffic-acquisition costs also overlooks content-acquisition costs—the payments to content providers from whom Google licenses video and other content for distribution on ad-driven and subscription services such as YouTube and Google Play. While Google does not pay a publisher for access to “eyeballs” on its owned-and-operated properties, it pays substantial and increasing amounts for content on those properties that attract the “eyeballs.”114F^[114] Alphabet CFO Ruth Porat indicates, for example, that YouTube pays content creators “a majority of our revenue.”115F[115] Leaving this expense out of the calculation is another example of the over-simplification that characterizes many of the claims that Google’s ad tech business is a simple (and simply anticompetitive) business.

B.   Excess pricing

Where Google’s critics diverge most significantly from the spirit of U.S. antitrust law is in their overriding concern for how advertising revenues are distributed among the recipients of advertisers’ payments: intermediaries (Google) and publishers. The Texas Complaint alleges that Google has a higher “take rate” than competing exchanges,116F[116] is able to increase its take rates without losing market share,117F[117] and “manipulates auctions to increase its take rate.”118F[118] This follows the Roadmap’s speculation—based on the CMA Interim Report—that Google may take a larger cut of advertising spending than its competitors.119F^[119] And these allegations echo claims made in another report that Google introduces “hidden” fees that increase the overall cut it takes from ad auctions.120F[120]

First, it should be noted that the basis for these claims in the Roadmap are drawn from the CMA investigation’s interim report, published in December 2019. In the final report, after further investigation, the CMA abandoned this claim. The final report describes the CMA’s analysis of all the billions of Google Ad Manager open auctions related to U.K. web traffic during the period between March 8 and March 14, 2020. This, according to the CMA, allowed it to observe any possible “hidden” fees, as well. The CMA concludes:

Our analysis found that, in transactions where both Google Ads and Ad Manager (AdX) are used, Google’s overall take rate is approximately 30% of advertisers’ spend. This is broadly in line with (or slightly lower than) our aggregate market-wide fee estimate outlined above. We also calculated the margin between the winning bid and the second highest bid in AdX for Google and non-Google DSPs…. We found that Google’s average winning margin was similar to that of non-Google DSPs. Overall, this evidence does not indicate that Google is currently extracting significant hidden fees. As noted below, however, it retains the ability and incentive to do so.121F^[121]

This is a crucial finding that severely undermines the allegations that Google extracts excessive or “hidden” fees. It also undermines the claim that there are “missing funds accruing to Google.” While these conclusions do not eliminate the possibility that the industrywide price could itself be above competitive levels (and it remains to be seen whether the plaintiffs states in the Texas case will produce different evidence), they do mean that the best evidence currently available calls into question the charge that Google exploits a lack of interoperability by prioritizing its own products or that it engages in opaque pricing to conceal hidden charges of which its customers are unaware.

More fundamentally, absent evidence of Google deceiving advertisers and publishers to extract above-competitive margins, claims that its prices are “too high” or its revenue sharing “too low” are at odds with established antitrust law. U.S. antitrust law does not attempt to derive “proper” prices and impose these obligations on companies to ensure a “fair” outcome. Absent anticompetitive defects in the process, even monopolists are free to charge monopoly prices. The alternative would be for some agency—a court or a regulator—to regulate pricing and second guess every business decision made by dominant firms.

C.   Cross-subsidies

At the same time, the Roadmap alleges that Google can “charge low prices at one end of the stack, to drive out competitors, while charging high prices at the other to counterbalance any losses.”122F[122] But even if true, this would not be anticompetitive. It is a widely understood feature of platforms that they can shift prices from one side of a multisided market to another to maximize the platform’s total value. For example, a marketplace may make sellers bear the burden of fraud or mis-selling to give assurance to customers, and grow the consumer side of the platform market, just as a ridesharing app may discount rides to attract customers to build a large enough base to induce drivers onto the app.

This is a normal part of platform economics, which has long recognized that offering one side a low, zero, or negative price can be efficient and procompetitive.123F[123] As the U.S. Supreme Court held in Ohio v. American Express, an integrated competitive-effects analysis should look at the overall effect on output, not the effect on one side of the market; the relevant market must include both sides of the platform or the market would not exist at all.124F[124] There is no reason to think that this kind of behavior would generally be classed as “predatory pricing” in the absence of other behavior, such as raising prices after driving out competitors.125F[125]

But neither the Texas Complaint nor the Roadmap allege that Google’s prices were predatory. On the contrary, their sole claim in this respect is that, after being acquired by Google, DoubleClick lowered its prices (by a factor of ten, according to the Roadmap126F[126]), which it then maintained at these lower levels. This price reduction is facially procompetitive, however. It is unusual, to say the least, to describe a price reduction, with no subsequent price rises, as anticompetitive. If less-efficient competitors were unable to compete with these lower prices, that is competition in action. The law does not preclude nonpredatory low prices, nor even predatory prices without recoupment.127F[127] Sustained price reductions are one of the primary goals of antitrust.

Moreover, the source of the Roadmap’s claim that these price reductions were done “to drive out competitors” was, notably, a company that was not actually driven out of business by these price reductions. The source was an ad server, Smart, which claimed that Google’s price reductions “made the provision of publisher ad server difficult to sustain as a standalone business. This was the main reason why Smart felt the need to expand into the provision of SSP services.”128F[128] A competitor of Google’s responding to price reductions by broadening its own offerings is, again, procompetitive, not anticompetitive.

D.   Data gathering and integration

The Texas Complaint and the Roadmap describe several pro-privacy measures Google has adopted or plans to adopt as being detrimental to its competitors, including the decision to disable third-party cookies (which allow digital advertising companies to track users across the web to serve them relevant targeted ads) on the Chrome browser.129F[129] The Complaint argues that this shift benefits Google to the detriment of other ad tech companies, because (it says) Google, but not its competitors, has other data sources it can use to target ads at users.130F[130] In the same vein, the Roadmap points to Google’s decision not to share with advertisers raw data that it compiles about users.131F[131]

The Complaint ignores regulatory causes of these changes altogether, and the Roadmap dismisses the suggestion that they may be driven by the European General Data Protection Regulation (GDPR), on the basis that “data sharing here in the U.S., where we have no privacy regulatory scheme akin to that which is in place in Europe” has also been curtailed.132F[132] Both forget data-privacy laws in U.S. states, such as the California Consumer Privacy Act (CCPA). And, even if that weren’t the case, the claim that GDPR would have no effect outside Europe ignores that companies may find it easier to comply with such laws by changing their practices globally, rather than on a country-by-country or state-by-state basis.133F[133] Many companies have done this: Microsoft, for example, announced in November 2019 that it would “honor California’s new privacy rights throughout the United States.”134F[134]

Both also ignore the possibility that these provisions may be a response to demand from users of Google Chrome. Google may have good reasons to maintain a reputation for protecting user privacy, particularly because of the wide range of services it provides where user privacy is often of paramount importance to many users: Search, Maps, Gmail, YouTube, and Chrome itself.

Apple and Mozilla, neither of which has a significant online display advertising arm (and thus, have no incentive to block cookies simply to disadvantage display advertisers, as the Complaint alleges Google has done) have taken similar steps to increase user privacy. These are direct competitors of Google Chrome’s, and when Apple made blocking third-party cookies the default in its Safari browser, it was reported by one major outlet as “beating Google by two years to the privacy feature.”135F[135] Indeed, one of the reasons that Google delayed its disabling of third-party cookies was reportedly to implement technologies to “make it easier for advertisers to target certain demographics without laser-sighting down to specific people, ensure that the infrastructure many sites use for logins don’t break, and help provide some level of anonymous tracking so advertisers can know if their ads actually converted into sales.”136F[136]

That Chrome’s competitors, neither of which has an incentive to hurt ad tech companies, have taken the same steps that the Complaint alleges Google is taking for anticompetitive reasons should be compelling evidence that Google, too, is responding to user demand and/or regulation. Under U.S. law, the fact that these are legitimate moves and benefit users interested in privacy—and, indeed, may be a response to competition in the browser market—undermines claims that Google has failed to maximize competition along other dimensions.

The Roadmap also presents a hypothetical circumstance that amounts to an allegation that Google “captures” data from ads served to publishers to “expropriate” publishers’ investments in content that attracts a particular audience:

Some publishers have invested in content that attracts and retains a specific type of consumer, for example, readers of the Wall Street Journal or Golf Digest; this in turn allows them to support their business by selling valuable ads to advertisers looking for exposure to those audiences. Google has two ways to expropriate that value. First, rather than serve an ad on the Wall Street Journal at a high price, it can track the user who visited the Wall Street Journal and wait until she visits a site that sells space at low prices, for example, a local recipe blogger. Google can then sell Wall Street Journal users to advertisers in a way that does not benefit the Wall Street Journal at all and costs advertisers much less. A second strategy used by Google is to take the data describing these differentiated audiences and use it to create an imitation portfolio of consumers that mimic the characteristics of the publisher’s audience. For example, Google could create an audience of consumers similar to the people who read Golf Digest. Then Google sells access to this group of consumers when they visit inexpensive websites. Advertisers are happy to buy these ads because the consumers likely belong to the specialized audience of interest but are available at a much lower price. In these ways the unique audience assembled by the publisher is copied and expropriated.137F[137]

It should be noted that the Roadmap does not conclude that Google engages in these practices, but merely describes strategies Google “can” undertake to “expropriate” publishers’ investments. The Roadmap concedes that advertisers would be “happy” under such hypotheticals, because they are buying effective ads at a “much lower price.” In the Roadmap’s example, the hypothetical recipe blogger is “happy” that it earned revenues from selling an impression and the advertiser is “happy” that it paid a lower price than it would have had the impression been sold to the Wall Street Journal. The Wall Street Journal may not be so “happy” that it did not serve that particular ad, but that display space did not sit empty; the hypothetical lost ad was replaced by another impression that was served. And it is neither Google’s nor antitrust law’s job to make specific publishers better off—nor to make publishers better off at the expense of advertisers—but to ensure that the market as a whole is competitive and acting in consumers’ interests.

These hypotheticals again highlight the tensions discussed above between the different sides of a multisided market. Actions that make advertisers “happy” may come at the expense of publishers’ advertising revenues and actions that increase publishers’ revenues may increase costs to advertisers. One of the goals of a multisided market intermediary such as Google is to balance these competing interests to maximize total revenues flowing through the ad tech stack.

The Roadmap concludes that, through its “entire family of products,” Google collects and analyzes substantial amounts of information about its users. It uses this information to maximize the “effectiveness,” “precision,” and “value” of the ads it intermediates.

First, Google offers an entire family of products—everything from Gmail and Google Maps to the Google Calendar, Google Chrome, Android mobile operating system and the search engine—that gather valuable personal data about its users. Second, the products across the ad stack further collect data on consumer activities that the company then integrates to maximize the effectiveness and precision of ad targeting and attribution and thereby the value of the ads.138F[138]

Rather than “expropriating” publishers’ data, it would be reasonable to conclude that Google is adding value to the data provided by publishers, advertisers, and consumers to better target ads. For example, the Wall Street Journal may not know that a consumer recently did a Google Search for “running shoes.” By adding valuable information from Search, the consumer might be served a relevant running shoe ad on the Wall Street Journal’s site. This benefits the publisher who is paid for serving a valuable impression, the advertiser who sells a pair of shoes, and the consumer who obtains useful information and purchases the product she was seeking.

E.   Accelerated Mobile Pages (AMP) and header bidding

The Texas Complaint, like the Roadmap, alleges that Google designed Accelerated Mobile Pages (AMP) “[t]o respond to the threat of header bidding… [by making it] essentially incompatible with JavaScript and header bidding. Google then used its power in the search market to effectively force publishers into using AMP.”139F[139] But this gets several key facts wrong. First, AMP is an open-source industry collaboration project and Google cannot unilaterally impose a design standard on it. 140F[140] Second, a version of header bidding can work with AMP.141F[141] Third, it is mistaken to assert that “non-AMP-formatted results often do not even show up on the first page of results, regardless of their relevance.”142F[142] AMP has been a prerequisite only for inclusion in the top news story carousel, while other listings are ranked by relevance and speed.

Importantly, the argument ignores the main benefit of AMP to publishers and users: faster load times for mobile users who may be on slow connections. One of header bidding’s chief downsides is that it increases page-load latency. It is obvious why an HTML framework built to maximize load times would not be compatible with header bidding. Because AMP confers undisputed benefits on users and publishers, Google and the other companies involved in the AMP project have no obligation to re-engineer AMP to be compatible with header bidding. Any conclusion otherwise would involve a court deciding that users should be forced to use a slower Internet so that websites can use header bidding.

F.   Alleged barriers to entry in the open digital display ad market

Claims about Google’s alleged market power in display advertising rest on assumptions that the company enjoys the benefits of significant barriers to entry throughout the ad tech stack, thus enabling it to extract supercompetitive rents without fear of competition: “With these barriers in place,” it is claimed, “entry seems nearly futile.”143F[143]

A key element in establishing a company’s durable market power—and thus, its ability to impose anticompetitive costs on its users—is the presence of entry barriers. Even a market with only a single company—a true monopoly—cannot act like a monopoly if entry into its market is easy; if it did profitably raise prices, new competitors would enter the market and undercut it.144F[144]

As the Roadmap concedes, “[m]arket power is not permanent, of course. It can be undercut by, among other things, new entrants that offer better quality or lower prices.”145F[145] This notion of “contestability” is a fundamental part of assessing the competitiveness of markets under U.S. antitrust law.146F[146] In the absence of barriers to entry, it is well-established that assumptions of future competitive harm from ongoing conduct cannot be sustained.147F[147] Thus, the Roadmap bases much of its brief against Google on the presence of barriers to entry, “which heighten[] the prospect that Google can engage in conduct that harms competition without restraint from new entrants or potential new entrants.”148F[148] On the strength of these asserted barriers, the Roadmap’s authors interpret ambiguous conduct as anticompetitive.

According to the Roadmap, “the CMA’s findings reveal a number of significant barriers to entry into the digital advertising market.”149F[149] But most of its assertions in this regard are flawed, either because the CMA did not, in fact, make “findings” in the ways it suggests, or else because it reaches incorrect conclusions that certain conduct constitutes a barrier.150F[150] The Texas Complaint’s assertions of similar barriers to entry are likewise problematic.151F[151]

1.     Consumer location information

Although the Texas Complaint does not discuss it as an explicit barrier to entry,152F[152] one of the Omidyar Roadmap’s key assertions about barriers to entry relates to Google’s access to user-location data. It asserts that:

The CMA concluded that Google has nearly insurmountable advantages in access to location data, due to the location information it receives from the Android operating system, Google search, and other applications…. An entrant into the ad tech stack requires information about the consumer to target an ad effectively. Because Google accounts for nearly the entirety of the mobile search sector in the UK—97%—and controls many of the known sources of location data, such an entrant faces a large barrier to entry.153F[153]

But the CMA does not, in fact, “conclude[] that Google has nearly insurmountable advantages in access to location data,” either in the CMA Interim Report to which the Roadmap refers, nor in the CMA Final Report. The CMA never makes any claim of “insurmountable advantage.” Indeed, it does not use the word “insurmountable” at all, except to note that “rival platforms did not suggest that accessing consumer data was an insurmountable barrier to entry.”154F[154] Rather, to support this claim, the Roadmap cites to a portion of the CMA Interim Report recounting a suggestion made by Microsoft regarding the “critical” value of location data in providing relevant advertising.155F[155]

Moreover, that portion of the CMA Interim Report, as well as the suggestion made by Microsoft, is about search advertising, not display advertising. While the CMA does not characterize this data in the way the Roadmap claims, it does allege that “Google has exclusive access to a large amount of user data that can be used for targeted advertising and for measuring advertising outcomes, collected through its consumer-facing services. Data collected on its search platform is particularly valuable for targeting purposes in open display as it reveals users’ purchasing intent.”156F[156]

While location data may also be valuable for display advertising in a comparable way, it is not clear that the GPS-level data that is so valuable in providing mobile-search ad listings is particularly useful for display advertising, which may be just as well-targeted by less granular city- or county-level location data.

Consider the Roadmap’s illustrative example:

A digital ad for a brick-and-mortar running store in Des Moines is of little use to a runner looking to test out new shoes in Omaha, and, if shown to the Omaha runner, is unlikely to prompt a click, much less a purchase, from the Des Moines store.157F[157]

This is certainly correct. But GPS or even cell-tower location data is not necessary to determine in which city a user is located. Publicly available databases of IP address locations can provide this information, and they are readily and often freely available to all competitors. It is difficult to imagine that display advertising uses location data at any greater level of granularity except in unusual circumstances; it simply would not be particularly useful or effective.158F[158]

Furthermore, to the extent that location data (like other consumer data) may be useful for display advertising, the most significant issue affecting its availability to advertisers is not Google’s presence in the ad tech stack; it is privacy regulations that limit the collection, use, and sharing of such data.159F[159] These privacy regulations, such as the GDPR, limit the ability of digital firms to sell or otherwise pass user data to third-party advertisers. What seems like unequal treatment, in this regard, is really a case of privacy regulation in action.

These laws may have the indirect effect of favoring larger digital conglomerates that can collect user data through one service and use it to target ads in another. In this sense, it could be true that Google has informational advantages over rivals, though in a critically different way than that alleged by the Roadmap. But it can hardly be considered anticompetitive if the source of such advantage is legal constraints on information sharing. Indeed, an empirical study by economists Avi Goldfarb and Catherine Tucker found that (pre-GDPR) privacy regulation in the EU “restricted advertisers’ ability to collect data on Web users in order to target ad campaigns. We find that, on average, display advertising became far less effective at changing stated purchase intent after the EU laws were enacted, relative to display advertising in other countries.”160F[160] Along similar lines, Nils Wernerfelt and his co-authors show that access to data from different sources significantly improves ad targeting.161F[161] In turn, this may give a competitive advantage to firms that operate several successful web services and applications.

As Israeli competition law scholars Michal Gal and Oshrit Aviv found with respect to the GDPR, privacy regulations can function as a barrier to entry in several further ways.162F[162] These include creating new economies of scale associated with regulatory compliance, increased litigation risk, and uncertainty around interpretation of the rules. Because they serve to make reputation more central, they also can lead users to become more likely to entrust their data to incumbents but not to unknown, new entrants.163F[163]

2.     Attribution measurement

“Attribution” refers to the method by which advertisers can see which ad led a user to an action, such as visiting a website or making a purchase.164F[164] The Roadmap alleges that Google can design attribution to mislead advertisers by, for example, favoring search ads over display ads.165F[165] This would lead to more of the advertiser’s money going to Google instead of (in part) to a publisher, and (assuming, as the Roadmap implies, that this makes ad campaigns less effective166F[166]) can harm advertisers by misleading them into choosing a less-effective advertising channel.

The Roadmap provides no evidence this is taking place. What it describes is more a complaint about the nature of search advertising in general: that companies will sometimes end up paying for ads in lieu of identical organic search results for their pages. That is an argument to be had elsewhere, but there are clear reasons why it may be in an advertiser’s interest to advertise even in these situations. Paid search ads may give them greater control over how a link is displayed to a user (for example, with text the advertiser has chosen, rather than text that the search engine has retrieved) or guarantee a prominent listing for searches where the advertiser’s URL listing is not always guaranteed to be on top.

Apart from the broader objection to the nature of search advertising, the Roadmap’s authors also object to Google setting an attribution default in its DSP. But advertisers can choose from several different attribution models, not just the default one that the Roadmap objects to, which attributes to search ads the “last click.” Other options include “last non-direct click,” which “ignores direct traffic and attributes 100% of the conversion value to the last channel that the customer clicked through from before buying or converting”; “last Google Ads click”; “First interaction”; and others that give attribution weights according to where and when the customer saw or used the ads during their purchase or conversion “journey.”167F[167] These are precisely the kinds of models that the Roadmap’s authors implicitly believe are more appropriate for campaigns heavy on display advertising. Advertisers can also create their own custom models, and Google has published guides for advertisers to help them choose among the different models.168F[168]

The Roadmap’s objection is thus reduced to being about the choice by Google to use the “last click” attribution model as the default. But some model has to be the default, and “last click” is also the default on, for example, Microsoft Advertising.169F[169] Indeed, according to digital-ad-intermediary company, Outbrain (one of Google’s competitors), it is the most common attribution default across the industry.170F[170] For the Roadmap’s objection to carry any weight, a case based on this claim would need to demonstrate that it was unusual for Google to use “last click” as the default attribution. Even then, given the ease with which advertisers can change the attribution model, the charge would be thin.

3.     ‘Opaque’ pricing

Both the Texas Complaint and the Roadmap allege that Google’s “opaque pricing” constitutes a barrier to entry by impeding “advertisers from switching to a lower-cost for higher-quality” buying tool.171F[171] As the Roadmap puts it, a “new or potential new PAS or DSP cannot credibly claim to be able to undercut the Google products on price if the publishers and advertisers cannot tell how much Google actually is charging.”172F[172] The Texas Complaint further alleges that “Google compounds its exclusionary auction manipulations by purposefully keeping its auction mechanics, terms, and pricing, opaque and ‘nontransparent.’ This makes it nearly impossible for publishers and advertisers to discover Google’s misrepresentations, and even harder for rivals to neutralize or offset.”173F[173] Both the Texas Complaint and the Roadmap also suggest that competition is undermined when publishers and advertisers do not know the fee structure of the intermediaries they are using.

But it is not unusual for businesses’ costs and prices to be private to their competitors, and it is not a barrier to competition. Grocery stores do not need to know how much it cost a farmer to grow an orange or how much their rivals are paying for transportation, unless they are attempting to anticompetitively coordinate their prices; they just need to work to make their own costs as low as possible and to reduce their prices to consumers by as much as possible. Similarly efficient firms are perfectly able to offer competitive prices simply by making the best offer based on their own fundamentals; only less-efficient firms will struggle (as they should).

Along these lines, for competition to work effectively in display advertising, Google’s competitors do not need to know what Google is charging; they need to offer a price and product that is more attractive overall to potential customers than Google’s is. Similarly, a publisher does not need to know how much an advertiser bid to place an ad, nor does the advertiser need to know how much the publisher received to serve the ad. The advertiser’s competition concern is whether an effective ad can be served at a lower price from a different intermediary and the publisher’s competition concern is whether it can earn greater revenues from a different intermediary. One should not be surprised that Google does not reveal information on which competing intermediaries can free ride. Indeed, this is widely considered to be one of the hallmarks of vigorous competition.

Conclusion

As we have argued, many of the most significant claims made against Google’s ad tech products are based on a misunderstanding of U.S. antitrust law, or of the details of the ad tech market itself. Although we cannot be sure how the Texas, et al. v. Google case will develop once the allegations in the Complaint are fleshed out into full arguments, many of its initial claims and assumptions are wrongheaded. Based on the information currently available, if the court rules in favor of these, the result will be to condemn procompetitive conduct and potentially to impose costly, inefficient remedies that function as a drag on innovation.

Legislators, too, who may be concerned about Google’s conduct and tempted to impose regulatory requirements on it and other tech companies should bear in minds the risk of the Nirvana fallacy, in which real-life conduct is compared against a hypothetical “competition maximizing” benchmark, and anything that falls short is deemed problematic and in need of intervention.174F[174] That approach would pervert the incentives of businesses to innovate and compete, and would make an unobtainable “perfect” that exists only in the minds of some economists and lawyers the enemy of a “good” that exists in the market right now.

 

[1] Third Amended Complaint, Texas v. Google, 21-md-3010-PKC (S.D.N.Y. Jan 14, 2022) at 105 (hereinafter, “Texas Complaint”).

[2] See Complaint, United States v. Google LLC, No. 1:20-CV-03010 (D.D.C. Oct. 20, 2020); see also, Complaint, State of Colorado, et al. v. Google LLC, 1:20-CV-03715 (D.D.C. Dec. 17, 2020).

[3] DoJ Expected to File Antitrust Lawsuit Against Google in Weeks—Bloomberg News, U.S. News (Jul. 14, 2022), https://money.usnews.com/investing/news/articles/2022-07-14/doj-expected-to-file-antitrust-lawsuit-against-google-in-weeks-bloomberg-news.

[4] Antitrust: Commission Opens Investigation into Possible Anticompetitive Conduct by Google in the Online Advertising Technology Sector, European Commission (Jun. 22, 2021), https://ec.europa.eu/commission/presscorner/detail/en/ip_21_3143.

[5] Bundeskartellamt Publishes Report on Non-Search Online Advertising for Public Discussion, Bundeskartellamt (Aug. 29, 2022), https://www.bundeskartellamt.de/SharedDocs/Meldung/EN/Pressemitteilungen/2022/29_08_2022_SU_Online_Werbung.html?nn=3599398.

[6] Id.

[7] Online Platforms and Digital Advertising Market Study, U.K. Competition and Markets Authority (Jul. 1, 2020), https://www.gov.uk/cma-cases/online-platforms-and-digital-advertising-market-study (hereinafter “CMA Market Study”); Online Platforms and Digital Advertising, Market Study Final Report, U.K. Competition and Markets Authority (Jul. 1, 2020), https://assets.publishing.service.gov.uk/media/5fa557668fa8f5788db46efc/Final_report_Digital_ALT_TEXT.pdf (hereinafter “CMA Final Report”), at 21 & 37.

[8] See Josh Frydenberg, Competition and Consumer (Price Inquiry—Digital Advertising Services) Direction 2020 (Feb. 10, 2020); Ad Tech Inquiry Issues Paper 5, Australian Competition & Consumer Commission (Mar. 10, 2020); Digital Advertising Services Inquiry, Australian Competition & Consumer Commission (Feb. 26, 2021), https://www.accc.gov.au/focus-areas/inquiries-finalised/digital-advertising-services-inquiry/submissions-to-interim-report.

[9] Investigation of Competition in Digital Markets, Majority Staff Report and Recommendations, Subcommittee on Antitrust, Commercial and Administrative Law of the Committee on the Judiciary (Oct. 4, 2020), available at https://templatelab.com/competition-in-digital-markets/4493-519; Hearing on Stacking the Tech: Has Google Harmed Competition in Online Advertising?, Committee of the Judiciary, Subcommittee on Antitrust, Competition Policy and Consumer Rights (Sep. 15, 2020), https://www.judiciary.senate.gov/meetings/stacking-the-tech-has-google-harmed-competition-in-online-advertising.

[10] Investigation of Competition in Digital Markets, Majority Staff Report and Recommendations, id., at 20.

[11] Competition and Transparency in Digital Advertising Act, S.4258, 117th Congress (2021-2022), https://www.congress.gov/bill/117th-congress/senate-bill/4258/text.

[12] Lee Introduces Digital Advertising Act, Mike Lee US Senator for Utah (May 19, 2022), https://www.lee.senate.gov/2022/5/lee-introduces-digital-advertising-act.

[13] Support Mounts for Lee’s Digital Advertising Act, Mike Lee US Senator for Utah (May 27, 2022), https://www.lee.senate.gov/2022/5/support-mounts-for-lee-s-digital-advertising-act.

[14] Online Platforms and Digital Advertising Market Study Interim Report, U.K. Competition and Markets Authority (Dec. 18, 2019), https://assets.publishing.service.gov.uk/media/5ed0f75bd3bf7f4602e98330/Interim_report_—_web.pdf (hereinafter, “CMA Interim Report”).

[15] Fiona M. Scott Morton & David C. Dinielli, Roadmap for a Digital Advertising Monopolization Case Against Google, Omidyar Network (May 2020), https://omidyar.com/wp-content/uploads/2020/09/Roadmap-for-a-Case-Against-Google.pdf (hereinafter, “Roadmap” or “Omidyar Roadmap”). One of the Roadmap’s authors testified at a Senate hearing on the display-advertising market, and the report has been widely cited. See, e.g., Gilad Edelman, Here’s What an Antitrust Case Against Google Might Look Like, Wired (May 18, 2020), https://www.wired.com/story/antitrust-case-against-google-roadmap-paper.

[16] Damien Geradin & Dimitrios Katsifis, An EU Competition Law Analysis of Online Display Advertising in the Programmatic Age, 15 Eur. Comp. J. 55 (2019); Damien Geradin & Dimitrios Katsifis, “Trust Me, I’m Fair”: Analysing Google’s Latest Practices in Ad Tech from the Perspective of EU Competition Law, 16 Eur. Comp. J. 11 (2020); Damien Geradin & Dimitrios Katsifis, Online Platforms and Digital Advertising Market Study: Observations on CMA’s Interim Report, TILEC Discussion Paper No. DP2020-044 (Feb. 13, 2020), https://ssrn.com/abstract=3537864; Damien Geradin & Dimitrios Katsifis, Competition in Ad Tech: A Response to Google, TILEC Discussion Paper No. DP2020-038 (Jun. 3, 2020), https://ssrn.com/abstract=3617839.

[17] The markets alleged in the Texas Complaint involve (1) publisher ad servers, (2) ad exchanges, (3) ad-buying tools for large advertisers, (4) ad-buying tools for small advertisers, (5) in-app mediation tools, and (6) in-app networks. The complaint does not relate to other forms of advertising on the Internet, such as targeted text-based ads sold by search engines, video ads that run before or during video content, or shareable ads on social media platforms.

[18] This section is distilled from our much longer discussion of the broader market surrounding digital advertising. See Eric Fruits, Geoffrey A. Manne & Lazar Radic, Relevant Market in the Google AdTech Case, ICLE Issue Brief 2022-06-01 (2022), https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4125569.

[19] Benedict Evans, News by the Ton: 75 Years of US Advertising (Jun. 15, 2020), https://www.ben-evans.com/benedictevans/2020/6/14/75-years-of-us-advertising; Benedict Evans, TV, Merchant Media and the Unbundling of Advertising (Mar. 18, 2022), https://www.ben-evans.com/benedictevans/2022/3/18/unbundling-advertising.

[20] See Fruits, Manne & Radic, supra note 18.

[21] Michael Schneider & Kate Aurthur, R.I.P. Cable TV: Why Hollywood Is Slowly Killing Its Biggest Moneymaker, Variety (Jul. 21, 2020), https://variety.com/2020/tv/news/cable-tv-decline-streaming-cord-cutting-1234710007 (“[B]asic cable feasted on a dual revenue stream of subscriber fees and advertising revenue. But that gravy train started going off the rails when the streaming services arrived.”).

[22] At the same time, as Benedict Evans notes, not all digital advertising is drawn from offline sources: “[I]f you talk to people at both Google and Facebook and in the agency world, you’ll hear that a lot of the money spent on Google and Facebook is money that was never spent on traditional advertising—it’s coming from SMEs [small and medium enterprises] and local businesses that might have spent in classified at most but probably wouldn’t have done even that.” Evans, News by the Ton, supra note 19 (emphasis in original).

[23] See Xi He, Rigoberto Lopez & Yizao Liu, Are Online and Offline Advertising Substitutes or Complements? Evidence from U.S. Food Industries, 15 J. Agricultural & Food Indus. Org. 1 (2017).

[24] David Bardey, Jorge Tovar & Nicolas Santos, Characterization of the Relevant Market in the Media Industry: Some New Evidence, Toulouse School of Economics Working Paper 16-719 (2016), https://www.tse-fr.eu/publications/characterization-relevant-market-media-industry-some-new-evidence (“The results show substitution and complementary patterns across certain media outlets. An increase in price for advertising in radio, for instance, leads to higher demand for newspapers and outdoors. Similarly, complementarity relationships between media outlets are observed, suggesting that advertising across the various media platforms is, overall, interwoven.”).

[25] David S. Evans, The Online Advertising Industry: Economics, Evolution, and Privacy, 23 J. Econ. Persp. 37, 49 (2009).

[26] Avi Goldfarb & Catherine Tucker, Search Engine Advertising: Channel Substitution When Pricing Ads to Context, 57 Management Sci. 458 (2011) (The authors find that the price of “ambulance chaser” lawyer ads was significantly more expensive in states prohibiting direct-mail solicitation by attorneys. This leads them to conclude that “online advertising substitutes for online advertising”).

[27] Avi Goldfarb & Catherine Tucker, Substitution Between Offline and Online Advertising Markets, 7 J. Competition L. & Econ. 37, 43 (2011).

[28] Daniel S. Levine, Ad-Supported Cyber-Magazines to Launch on Internet, Adweek (Sep. 10, 1993).

[29] Id.

[30] Brian Morrissey, How the Banner Ad Was Born, Digiday (Apr. 12, 2013), https://digiday.com/marketing/how-the-banner-ad-was-born.

[31] Chris Lapham, AOL and GNN Partner to Build Launch Pad, CMC Magazine (Jul. 1, 1995), https://www.december.com/cmc/mag/1995/jul/cutting.html.

[32] Kim Cleland, Poppe Creates Web Net, Advertising Age (Oct. 30, 1995).

[33] Id.

[34] The History of Online Advertising, OKO Ad Management (Jul. 19, 2019), https://oko.uk/blog/the-history-of-online-advertising.

[35] The company changed its name to Overture, which was acquired by Yahoo! in 2003.

[36] GoTo.com Announces First Round of Financing, Totaling More Than $ 6 Million, Led by Draper Fisher Jurvetson, Business Wire (May 19, 1998), https://www.internetnews.com/marketing/goto-com-raises-6-million-in-first-round-of-financing.

[37] Will Oremus, Google’s Big Break, Slate (Oct. 13, 2013), https://slate.com/business/2013/10/googles-big-break-how-bill-gross-goto-com-inspired-the-adwords-business-model.html.

[38] Id.

[39] Dean Schmid, The History of Display Advertising: Everything You Need to Know, DisruptorDaily.com (Aug. 14, 2017), https://www.disruptordaily.com/the-history-of-display-advertising-everything-you-need-to-know.

[40] Julia Scheeres, Death of Banner Ads Exaggerated, Wired (Jan. 26, 2001), https://www.wired.com/2001/01/death-of-banner-ads-exaggerated.

[41] Id.

[42] Breaking News, AdAge (Oct. 23, 2000).

[43] Mark Evans, Investors Leap off Overture Roller Coaster: Rival Google Elbows In, National Post (Feb. 21, 2002).

[44] Oremus, supra note 37.

[45] Google Grabs Applied Semantics, EuropeMedia (Apr. 25, 2003); Google Expands Advertising Monetization Program for Websites, Google Press Release (Jun. 18, 2003), http://googlepress.blogspot.com/2003/06/google-expands-advertising-monetization.html.

[46] Dean Schmid, The History of Display Advertising: Everything You Need to Know, DisruptorDaily (Aug. 14, 2017), https://www.disruptordaily.com/the-history-of-display-advertising-everything-you-need-to-know.

[47] Kate Walsh, Search Marketing: Understanding the Basics, B2B Marketing Magazine (March 2006).

[48] Louise Story & Miguel Helft, Google Buys DoubleClick for $3.1 Billion, The New York Times (Apr. 14, 2007), https://www.nytimes.com/2007/04/14/technology/14DoubleClick.html.

[49] Sarah Sluis, The Year Header Bidding Went Mainstream, AdExchanger (Dec. 27, 2016); Townhall Media Selects OpenX for Patent-Pending Header Bidding Solution, BusinessWire (Sep. 18, 2015), https://www.businesswire.com/news/home/20150918005110/en/Townhall-Media-Selects-OpenX-for-Patent-Pending-Header-Bidding-Solution.

[50] As the name suggests, ad waterfalls enable publishers to sell their inventory seriatim, beginning with premium, direct sales and flowing through the most historically profitable ad servers in succession to unload unsold inventory before offering its remnant inventory in the open display channel. See, e.g., Maciej Zawadzinski, What Is Waterfalling and How Does it Work?, Clearcode (Aug. 20, 2021), https://clearcode.cc/blog/what-is-waterfalling.

[51] See, e.g., Header Bidding, OKO Ad Management, https://oko.uk/topic/header-bidding (retrieved July 27, 2022).

[52] Client-side header bidding is so-named because it operates via a small piece of java script embedded in the header of a publisher’s website and executed within the user’s browser (i.e., client). See, e.g., Maciej Zawadzinski, What Is Header Bidding and How Does it Work?, Clearcode (Aug. 20, 2021), https://clearcode.cc/blog/what-is-header-bidding.

[53] Header Bidding Facts and Statistics 2021, Automatad (Jun. 27, 2021), https://headerbidding.co/header-bidding-statistics. Today, 70% of the top 10,000 U.S. publishers use header bidding. See Header Bidding (HBIX) Tracker, kevel (retrieved Nov. 1, 2022), https://www.kevel.com/hbix.

[54] See, e.g., CMA Final Report, supra note 7, at Appendix M, ¶ 33.

[55] See, e.g., Vishveshwar Jatain, Header Bidding Integrations: Client Vs. Server-Side, Explained, Blockthrough (Apr. 15, 2021), https://blockthrough.com/blog/header-bidding-integrations-client-vs-server-side-explained.

[56] IAB and PwC, IAB Internet Advertising Revenue Report, 2010 Full Year Results (Apr. 2011), available at https://www.iab.com/wp-content/uploads/2015/05/IAB_Full_year_2010_0413_Final.pdf; Megan Graham, Digital Ad Revenue Jumped 35% in the U.S. Last Year, Biggest Gain Since 2006, Wall Street Journal (Apr. 12, 2022), https://www.wsj.com/articles/digital-ad-revenue-jumped-35-in-the-u-s-last-year-biggest-gain-since-2006-11649759401.

[57] Producer Price Index by Commodity: Advertising Space and Time Sales: Internet Advertising Sales, Excluding Internet Advertising Sold by Print Publishers, U.S. Bureau of Labor Statistics, https://fred.stlouisfed.org/series/WPU365; Producer Price Index, December 2009—February 2021, U.S. Bureau of Labor Statistics, https://fred.stlouisfed.org/graph/?g=vtTd.

[58] Scott Morton & Dinielli, supra note 15, at 9.

[59] Importantly, however, network effects are not monolithic; nor do they increase forever. For different types of networks at different points in their growth, adding more users might not increase the value of the platform and could even reduce the platform’s benefits. See, e.g., D’Arcy Coolican & Li Jin, The Dynamics of Network Effects, Andreesen Horowitz (Dec. 13, 2018), https://a16z.com/2018/12/13/network-effects-dynamics-in-practice.

[60] See Stan J. Liebowitz & Stephen E. Margolis, Network Externality: An Uncommon Tragedy, 8 J. Econ. Persp. 133 (1994).

[61] Catherine Tucker, Digital Data, Platforms and the Usual [Antitrust] Suspects: Network Effects, Switching Costs, Essential Facility, 54 Rev. Indus. Org. 683, 686 (2019).

[62] See, e.g., David S. Evans, Economics of Vertical Restraints for Multi-Sided Platforms, University of Chicago Institute for Law & Economics Olin Research Paper No. 626 (Jan. 2, 2013), https://ssrn.com/abstract=2195778.

[63] See, e.g., Stylianos Despotakis, R. Ravi & Amin Sayedi, First-Price Auctions in Online Display Advertising, 58 J. Marketing Research 888 (2021). See also Display Advertising Switched to First-Price Auctions After Adoption of Header Bidding, New Study Finds, Tepper School of Business (Apr. 22, 2020), https://www.cmu.edu/tepper/news/stories/2020/april/display-advertisting-research-ravi.html.

[64]  Jonathan Levin, Auction Theory (Oct. 2004), available at https://web.stanford.edu/~jdlevin/Econ%20286/Auctions.pdf.

[65] Maciej Zawadziński, How Do First-Price and Second-Price Auctions Work in Online Advertising?, Clearcode (Aug. 12, 2021), https://clearcode.cc/blog/first-price-second-price-auction.

[66] Id.

[67] Texas Complaint, supra note 1, at ¶ 351 (“Overall, the lack of transparency prevents more efficient competition that would drive greater innovation, increase the quality of intermediary services, increase output, and create downward pricing pressure on intermediary fees.”); Scott Morton & Dinielli, supra note 15, at 18 (“Based on the public facts known at the moment, however, it does not seem plausible that the incremental efficiencies created by the conduct described here could outweigh all the harms to competition resulting from this broad pattern of behaviors.”); Scott Morton & Dinielli, supra note 15, at 38 (“It also is true that Google has allowed some rivals to survive (although not necessarily to thrive). It is possible that Google adopted a strategy of incomplete foreclosure specifically so that it can paint an illusion of healthy competition when the reality is quite different. Indeed, to the extent Google has adopted ‘pro-competitive’ concessions, the narrative here demonstrates that they simply have not succeeded in addressing the harms or lowering the barriers to entry.”).

[68] Id. at 3 (“It is clear even to us as lay people that there are less anticompetitive ways of delivering effective digital advertising—and thereby preserving the substantial benefits from this technology—than those employed by Google.”).

[69] United States v. Aluminum Co. of America, 148 F.2d 416, 430 (2^nd Cir. 1945) (Learned Hand, J.) (emphasis added).

[70] Cal. Computer Prods., Inc. v. Int’l Bus. Machine Corp., 613 F.2d 727, 744 (9th Cir. 1979) (“IBM, assuming it was a monopolist, had the right to redesign its products to make them more attractive to buyers whether by reason of lower manufacturing cost and price or improved performance. It was under no duty to help [its competitors] survive or expand.”).

[71] Scott Morton & Dinielli, supra note 15 at 18.

[72] Texas Second Amended Complaint at ¶ 113.

[73] See Verizon Commc’ns Inc. v. Law Offices of Curtis V. Trinko, LLP, 540 U.S. 398, 408 (2004). The exception—”at or near the outer boundary of § 2 liability” (id. at 409)—is the extremely narrow case in which a monopolist (i) sacrifices profits, by (ii) terminating a prior course of dealing, (iii) for no purpose except to harm competition. See Novell v. Microsoft, 731 F.3d 1064, 1074-75 (10th Cir. 2013) (Gorsuch, J.) (holding that a refusal-to-deal claim requires terminating “a preexisting voluntary” course of dealing where the “monopolist decided to forsake short-term profits,” and “the monopolist’s conduct” is “irrational but for its anticompetitive effect”).

[74] Opinion and Order, Texas, et al. v. Google, 21-md-3010-PKC (S.D.N.Y, Sep. 13, 2022) (citing Charych v. Siriusware, Inc., 790 Fed. App’x 299, 302 (2^nd Cir. 2019)).

[75] United States v. Colgate & Co., 250 U.S. 300, 307 (1919).

[76] Trinko, 540 U.S. at 407-08

[77] See Brunswick Corp. v. Pueblo Bowl-O-Mat, Inc., 429 U.S. 477, 488 (1977) (“The antitrust laws, however, were enacted for ‘the protection of competition not competitors.’”) (quoting Brown Shoe Co. v. United States, 370 U.S. 294, 320 (1962)).

[78] See Ohio v. American Express, 138 U.S. 2274, 2285 (2018) (“Due to indirect network effects, two-sided platforms cannot raise prices on one side without risking a feedback loop of declining demand…. Price increases on one side of the platform [] do not suggest anticompetitive effects without some evidence that they have increased the overall cost of the platform’s services.”).

[79] Geoffrey A. Manne, In Defence of the Supreme Court’s ‘Single Market’ Definition in Ohio v American Express, 7 J. Antitrust Enforcement 104, 111 (2019) (quoting Brief for Amici Curiae Antitrust Law & Economics Scholars in Support of Respondents at 19, Ohio v. American Express, 138 U.S. 2274 (2018) (No. 16-1454) and United States, et al. v. American Express, 838 F.3d 179, 198 (2^nd Cir. 2016)).

[80] Among innumerable examples, see Texas Complaint, supra note 1, at ¶ 297 (“Google’s harm to the competitive process has harmed customers in this market, i.e., online publishers.”). Notably, the Texas Complaint does, in places, recognize that identifying the incidence of benefits and harms in multisided markets is complex—it just fails to carry its analysis to its logical conclusion. Thus, in ¶157 the Complaint notes that “[t]he higher advertising revenue publishers make from exchanges permits publishers to offer consumers better quality content and lower-priced or free access to their content.” (Emphasis added). Undoubtedly, this is true. But if it is correct, then it must also be correct that, at the same time, the correspondingly higher prices advertisers pay for advertising through exchanges limits their ability to provide marketing benefits directly to consumers and may increase the price to consumers of the advertised goods. It is an empirical question which effect is larger, but the mere possibility that one set of consumers could benefit from a different arrangement is insufficient on its own to identify harm when another set of consumers would be harmed by it.

[81] See Harold Demsetz, Information and Efficiency: Another Viewpoint, 12 J. L. & Econ. 1, 1-2 (1969) (“In practice, those who adopt the nirvana viewpoint seek to discover discrepancies between the ideal and the real and if discrepancies are found, they deduce that the real is inefficient…. The nirvana approach is… susceptible… to committing three logical fallacies—the grass is always greener fallacy, the fallacy of the free lunch, and the people could be different fallacy.”) (emphasis in original).

[82] See, generally, Thomas Nachbar, Less Restrictive Alternatives and the Ancillary Restraints Doctrine, Virginia Law and Economics Research Paper No. 2020-18 (2021) (forthcoming U. Seattle L. Rev.) at 57-8, available at https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3723807 (“The more general risk to tech markets comes from the intangible nature of the products and services they produce. Although many of the cases cited for less restrictive alternatives are horizontal cases, it is the vertical context (which normally receives more permissive antitrust review) in which less restrictive alternatives present the greatest likelihood of destabilizing current law because of the difficulty of specifying what is and is not less restrictive with regard to the intangible products produced by today’s ‘big tech’ economy. To the extent that less restrictive alternatives present problems of incrementalism, those problems will be exacerbated in the ‘big tech’ markets.”).

[83] See Geoffrey A. Manne, Error Costs in Digital Markets, Global Antitrust Institute Report on the Digital Economy (Joshua D. Wright & Douglas H. Ginsburg, eds., 2020) 33, 76, available at https://gaidigitalreport.com/wp-content/uploads/2020/11/Manne-Error-Costs-in-Digital-Markets.pdf (“The concern with error costs is especially high in dynamic markets in which it is difficult to discern the real competitive effects of a firm’s conduct from observation alone. And for several reasons, antitrust decision-making in the context of innovation tends much more readily toward distrust of novel behavior, thus exacerbating the risk and cost of over-enforcement.”).

[84] Among many other examples, see Texas Second Amended Complaint at ¶138 (“Then, through Dynamic Allocation, Google’s ad server passed inside information to Google’s exchange and permitted Google’s exchange to purchase valuable impressions at artificially depressed prices. Competing exchanges were deprived of the opportunity to compete for inventory and left with the low-value impressions passed over by Google’s exchange.”); Omidyar Roadmap, supra note 15, at 20 (“[A]fter purchasing DoubleClick, which became its publisher ad server, Google apparently lowered its prices to publishers by a factor of ten, at least according to one publisher’s account related to the CMA. Low prices for this service can force rivals to depart, thereby directly reducing competition.”).

[85] Barry Wright Corp. v. ITT Grinnell Corp., 724 F.2d. 227, 235 (1^st Cir. 1983) (Breyer, C.J.).

[86] Allied Orthopedic Appliances, Inc. v. Tyco Health Care Grp. LP, 592 F.3d 991, 1000 (9^th Cir. 2010).

[87] Berkey Photo, Inc. v. Eastman Kodak Co., 603 F.2d 263, 287 (2^nd Cir. 1979). See also, Manne & Wright, Innovation and the Limits of Antitrust, 6 J. Comp. L. & Econ. 153–202 (March 2010), https://doi.org/10.1093/joclec/nhp032.

[88] Allied Orthopedic, 592 F.3d at 999-1000; see also, California Computers Prods. v. IBM, 613 F.2d 727, 744 (9^th Cir. 1979); Foremost Pro Color, Inc. v. Eastman Kodak Co., 703 F.2d 534, 543-45 (9^th Cir. 1983).

[89] Foremost Pro Color, 703 F.2d at 543.

[90] Allied Orthopedic, 592 F.3d at 1000.

[91] Verizon Commc’ns, Inc. v. Law Offices of Curtis V. Trinko, 540 U.S. 398, 400-41 (2004).

[92] Trinko, 540 U.S. at 415-16.

[93] Id.; see also New York Merc. Exch., Inc. v. Intercontinental Exch. Inc., 323 F.Supp.2d 559 (S.D.N.Y. 2004) (dismissing Section 2 claim and reiterating limited exceptions in which forced sharing is appropriate).

[94] See, e.g., Jack Walters & Sons Corp. v. Morton Bldg., Inc., 737 F.2d 698, 710 (7^th Cir. 1984) (“We just said that vertical integration is not an improper objective. But this puts the matter too tepidly; vertical integration usually is procompetitive. If there are cost savings from bringing into the firm a function formerly performed outside it, the firm will be made a more effective competitor.”). There is a robust body of empirical research indicating that vertical integration is generally procompetitive or benign. For a summary of the leading meta-studies by DOJ and FTC economists and others, see Koren W. Wong-Ervin, Antitrust Analysis of Vertical Mergers: Recent Developments and Economic Teachings, The Antitrust Source (February 2019), https://www.americanbar.org/content/dam/aba/publishing/antitrust_source/2018-2019/atsource-february2019/feb19_wong_ervin_2_18f.pdf. See also, Francine Lafontaine & Margaret Slade, Vertical Integration and Firm Boundaries: The Evidence, 45 J. Econ. Lit. 677 (2007) (“In spite of the lack of unified theory, overall a fairly clear empirical picture emerges. The data appear to be telling us that efficiency considerations overwhelm anticompetitive motives in most contexts. Furthermore, even when we limit attention to natural monopolies or tight oligopolies, the evidence of anticompetitive harm is not strong.”). See also, generally, Geoffrey A. Manne, Kristian Stout & Eric Fruits, The Fatal Economic Flaws of the Contemporary Campaign Against Vertical Integration, 68 Kansas L. Rev. 923 (2020).

[95] Scott Morton & Dinielli, supra note 15, at 17.

[96] See, e.g., Port Dock & Stone Corp. v. Oldcastle Ne., Inc., 507 F.3d 117, 123-25 (2^nd Cir. 2007) (affirming dismissal of a Section 2 claim and finding that even a monopolist’s “vertical expansion into another level of the same product market will ordinarily be for the purpose of increasing its efficiency, which is a prototypical valid business purpose”). Moreover, single-firm conduct that supposedly projects power into another market, even through anticompetitive means, does not violate Sherman Act Section 2 unless the practices threaten monopoly power in that distinct second market. Harming competition is not enough. See Trinko, 540 U.S. at 415 n.4 (citing Spectrum Sports, Inc. v. McQuillan, 506 U.S. 447, 459 (1993)).

[97] Thus, the Omidyar Roadmap condemns Google’s supposed integration of data “to maximize the effectiveness and precision of ad targeting and attribution and thereby the value of an ad,” Scott Morton & Dinielli, supra note 15, at 20, even though the conduct makes Google’s offering to advertisers more attractive.

[98] Id. at 18-19.

[99] CMA Interim Report, supra note 14, at ¶ 5.89.

[100] Texas Second Amended Complaint at ¶ 113.

[101] Scott Morton & Dinielli, supra note 15, at 22.

[102] Texas Second Amended Complaint at ¶¶ 284-91 (“Cutting off access to YouTube foreclosed competition in the ad buying tool markets and protected Google’s market power in these markets. Many DSPs stopped growing, many others went out of business, and the market overall has been closed to entry.”).

[103] Scott Morton & Dinielli, supra note 15, at 22.

[104] See, e.g., Ryan Joe, The Big Story: Call of the Peacock, AdExchanger (Jan. 22, 2020) at 31:05-31:26, https://www.adexchanger.com/podcast/the-big-story/the-big-story-call-of-the-peacock (indicating that NBC’s Peacock streaming service will have only direct sales when it launches); Kevin Weiss, What Is the Amazon Demand Side Platform (DSP)?, Amplio (July 2019), https://www.ampliodigital.com/blog/what-is-the-amazon-demand-side-platform-dsp (“Amazon DSP is the only way to access advertising inventory exclusively available on Amazon’s collection of owned online properties and devices like: Kindle; Fire TV; IMDb; Amazon Owned & Operated properties”); Tim Cross, Xandr Launches New Demand-Side Platform ‘Xandr Invest’, VideoAdNews (Jun. 10, 2019) https://videoadnews.com/2019/06/10/xandr-launches-new-demand-side-platform-xandr-invest (“Xandr [AT&T-Time Warner’s ad tech division] has announced it will be the exclusive source of inventory from Community, its recently announced video marketplace which includes content from various WarnerMedia brands as well as Vice, Hearst Magazines, Newsy, Philo, Tubi and XUMO.”).

[105] Neal Mohan, Focusing Investments to Improve Buying on YouTube, Google (Aug. 6, 2015) https://doubleclick-advertisers.googleblog.com/2015 (“To continue improving the YouTube advertising experience for as many of our clients as possible, we’ll be focusing our future development efforts on the formats and channels used by most of our partners. To enable that, as of the end of the year, we’ll no longer support the small amount of YouTube buying happening on the DoubleClick Ad Exchange.”); see also, Kelly Liyakasa, Google to Yank YouTube Inventory out of AdX by Year’s End, AdExchanger (Aug. 6, 2015), https://www.adexchanger.com/ad-exchange-news/google-to-yank-youtube-inventory-out-of-adx-by-years-end.

[106] Liyakasa, id.

[107] See Lara O’Reilly, Ad Tech Company The Trade Desk Goes Public at $28.75 Per Share—A Huge Pop on its $18 Price Target, Business Insider (Sep. 21, 2016), https://www.businessinsider.com/the-trade-desk-ipo-2016-9; Trey Titone, The Bill That Could Break Up Google and Shake Up Ad Tech, Ad Tech Explained (May 23, 2022), https://adtechexplained.com/competition-and-transparency-in-digital-advertising-act-ctda; Trade Desk Market Cap, YCharts, https://ycharts.com/companies/TTD/market_cap.

[108] Scott Morton & Dinielli, supra note 15, at 16 n.70 (identifying AppNexus as a “vigorous competitor to Google”).

[109] Id. at 2, 28-29.

[110] Annual Report (Form 10-K) for Year Ending December 31, 2021, Alphabet Inc. (Feb. 02, 2022), https://www.sec.gov/ix?doc=/Archives/edgar/data/0001652044/000165204422000019/goog-20211231.htm.

[111] Id.

[112] Id.

[113] Id.

[114] Id.

[115] Rachel Kaser, YouTube Claims to Share Billions in Ad Money with Creators, Unlike Instagram, The Next Web (Feb. 5, 2020), https://thenextweb.com/facebook/2020/02/05/youtube-claims-share-billions-ad-money-creators-unlike-instagram.

[116] Texas Complaint, supra note 1, at ¶¶ 61, 156, 253, 288.

[117] Id. ¶ 157.

[118] Id. ¶ 21.

[119] Scott Morton & Dinielli, supra note 15, at 14 (“The CMA estimates Google’s take rate, or price, at 40%, which it deems a supra-competitive price for the services provided by the Google-controlled players in ad tech stack. A recent study by the Incorporated Society of British Advertisers (ISBA) found that publishers received 51% of the price, while the amount they could track going to intermediaries was 34%. The study could not find where the remaining 15% of the price went. As we will describe below, Google has such a dominant position across all elements of the ecosystem, it seems likely that these missing funds are accruing to Google at least in part, which would support the CMA’s findings.”).

[120] Geradin & Katsifis, “Trust Me, I’m Fair,” supra note 16 (“[L]ack of competition across the ad tech chain enables Google to exploit advertisers and publishers by charging hidden fees for ad intermediation on top of its disclosed commission…. Unfortunately, we conclude that Google’s latest switch does nothing to increase auction transparency. Worse, it seems to strengthen Google’s ability to extract hidden margins from its customers, while undermining the competitive pressure exercised by header bidding.”).

[121] CMA Final Report, supra note 7, at 275 (emphasis added).

[122] Scott Morton & Dinielli, supra note 15, at 20.

[123] See, e.g., Jean-Charles Rochet & Jean Tirole, Platform Competition in Two-Sided Markets, 1 J. Eur. Econ. Ass’n 990 (2003); Bruno Jullien, Price Skewness and Competition in Multi-Sided Markets, IDEI Working Paper 504 (March 2008), available at https://core.ac.uk/download/pdf/6375977.pdf.

[124] See Joshua. D. Wright & John. M. Yun, Burdens and Balancing in Multisided Markets: The First Principles Approach of Ohio v. American Express, 54 Rev. Industrial Organization 717 (2019); Manne, In Defence of the Supreme Court’s ‘Single Market’ Definition in Ohio v American Express, supra note 79.

[125] As described here, true pricing is theoretically possible but difficult in practice: “To successfully engage in predatory pricing means taking enormous and rising losses that grow for the ‘predatory’ firm as customers switch to it from its competitor. And once the rival firm has exited the market, if the predatory firm raises prices above average cost (i.e., to recoup its losses), there is no guarantee that a new competitor will not enter the market selling at the previously competitive price. And the competing firm can either shut down temporarily or, in some cases, just buy up the ‘predatory’ firm’s discounted goods to resell later.” Sam Bowman, Buck’s “Third Way”: A Different Road to the Same Destination, Truth on the Market (Oct. 27, 2020), https://truthonthemarket.com/2020/10/27/bucks-third-way-a-different-road-to-the-same-destination.

[126] Scott Morton & Dinielli, supra note 15, at 20.

[127] See, e.g., Barry Wright, 724 F.2d at 234-35.

[128] CMA Interim Report, supra note 14, at Appendix H, ¶ 194.

[129] Texas Complaint, supra note 1, at ¶ 477.

[130] Id. at ¶¶ 473-476.

[131] Scott Morton & Dinielli, supra note 15, at 28.

[132] Id. at 28.

[133] As one study on the effects of GDPR (in this case, on app development) notes, “While 42.1 percent of EU-developed apps exit in the year following GDPR, the analogous figure averages between 37.7 and 50 percent in the other six countries, confirming the difficulty in finding an untreated part of the world.” Rebecca Janßen, Reinhold Kesler, Michael E. Kummer, & Joel Waldfogel, GDPR and the Lost Generation of Innovative Apps, NBER Working Paper 30028 (May 2022) at 19-20, available at https://www.nber.org/papers/w30028.

[134] Julie Brill, Microsoft Will Honor California’s New Privacy Rights Throughout the United States, Microsoft Blog (Nov. 11, 2019), https://blogs.microsoft.com/on-the-issues/2019/11/11/microsoft-california-privacy-rights.

[135] Nick Statt, Apple Updates Safari’s Anti-Tracking with Full Third-Party Cookie Blocking, The Verge (Mar. 24, 2020), https://www.theverge.com/2020/3/24/21192830/apple-safari-intelligent-tracking-privacy-full-third-party-cookie-blocking.

[136] Dieter Bohn, Google to “Phase Out” Third-Party Cookies in Chrome, but not for Two Years, The Verge (Jan. 24, 2020), https://www.theverge.com/2020/1/14/21064698/google-third-party-cookies-chrome-two-years-privacy-safari-firefox.

[137] Scott Morton & Dinielli, supra note 15, at 30.

[138] Id. at 20.

[139] Texas Complaint, supra note 1, at ¶¶ 407-408. See also, Scott Morton & Dinielli, supra note 15, at 26.

[140] See David Besbris, Introducing the Accelerated Mobile Pages Project, for a Faster, Open Mobile Web, Google (Oct. 7, 2015), https://blog.google/products/search/introducing-accelerated-mobile-pages/.

[141] See Automated Team, Header Bidding on AMP—A Complete Guide, Automated (Jan. 10, 2020), https://headerbidding.co/header-bidding-amp.

[142] Scott Morton & Dinielli, supra note 15, at 26.

[143] Id. at 17.

[144] See William J. Baumol, Contestable Markets: An Uprising in the Theory of Industry Structure, 72 Am. Econ. Rev. 1, 14 (1982) (“In the limit, when entry and exit are completely free, efficient incumbent monopolists and oligopolists may in fact be able to prevent entry. But they can do so only by behaving virtuously, that is, by offering to consumers the benefits which competition would otherwise bring. For every deviation from good behavior instantly makes them vulnerable to hit-and-run entry.”).

[145] Scott Morton & Dinielli, supra note 15, at 15.

[146] See, generally, William J. Baumol, John C. Panzar, & Robert D. Willig, Contestable Markets and the Theory of Industry Structure (1982).

[147] United States v. Microsoft Corp., 253 F.3d 34 (D.C. Cir. 2001) (“Because a firm cannot possess monopoly power in a market unless that market is also protected by significant barriers to entry… it follows that a firm cannot threaten to achieve monopoly power in a market unless that market is, or will be, similarly protected.”).

[148] Scott Morton & Dinielli, supra note 15, at 15.

[149] Id.

[150] See CMA Final Report, supra note 7, at 252–55 for a discussion of barriers to entry.

[151] See Texas Complaint, supra note 1, at ¶ 127: In addition to these barriers, Google’s own anticompetitive conduct imposes additional barriers to entry and expansion. As addressed below in Section VII.A, from 2010 to present, Google has tied its ad server to its ad exchange, requiring publishers to use Google’s ad server in order to receive live, competitive bids from Google’s ad exchange. This tie effectively forces almost every large publisher to use Google’s ad server. And because it is difficult-to-impossible for a publisher to use multiple ad servers simultaneously, requiring publishers to use Google’s ad server effectively prohibits them from using a competitor’s ad server. Google’s anticompetitive conduct creates an unnatural and nearly insurmountable barrier to entry.

[152] An earlier version of the Texas Complaint did make assertions regarding Google’s abuse of monopoly power through the “use [of] its data advantages to trade on inside information” (Texas Second Amended Complaint at ¶ 311), by which the state plaintiffs may mean (or have meant) to encompass location data, among other things.

[153] Scott Morton & Dinielli, supra note 15, at 15.

[154] CMA Interim Report, supra note 14, at 189 (emphasis added).

[155] Id., at ¶ 3.71 (“Microsoft suggested that accessing at-scale location data from user devices is a critical input to providing relevant, localized results. It indicated its belief that Google has unique advantages in this area, due to the location data that it receives from the Android operating system and the location data it receives when users access Google Search or other apps like Google Maps/Waze.”).

[156] CMA Final Report, supra note 7, at ¶ 5.268.

[157] Scott Morton & Dinielli, supra note 15, at 15.

[158] To be sure, location data can be helpful in assessing the efficacy of advertising by, for example, enabling an advertiser to better evaluate whether an advertisement led users to go to the advertiser’s physical location. But this function hardly seems necessary to a well-functioning market, and other sources of such information (e.g., questionnaires) are available.

[159] Regulation (EU) 2016/679 of the European Parliament and of the Council of 27 April 2016 on the Protection of Natural Persons with Regard to the Processing of Personal Data and on the Free Movement of Such Data, and Repealing Directive 95/46/EC (General Data Protection Regulation), OJ L 119, 4.5, 2016; See, e.g., Bert Peeters, Processing of Location Data: Navigating the EU Data Protection Framework, CiTiP Blog (Feb. 4, 2021), https://www.law.kuleuven.be/citip/blog/processing-of-location-data-navigating-the-eu-data-protection-framework. (“The general understanding seems to be that, while European law does not qualify location data as a ‘special category’ of data under article 9 of the GDPR, location data should for all intents and purposes be treated with the utmost of care.”).

[160] Avi Goldfarb & Catherine Tucker, Privacy Regulation and Online Advertising, 57 Mgmt. Sci. 57, 57 (2011).

[161] Nils Wernerfelt, Anna Tuchman, Bradley Shapiro, & Robert Moakler, Estimating the Value of Offsite Data to Advertisers on Meta, University of Chicago, Becker Friedman Institute for Economics Working Paper No. 114 (August 22, 2022) at 1, available at https://ssrn.com/abstract=4198438 (“Taken together, our results suggest a substantial benefit of offsite data across a wide range of advertisers, an important input into policy in this space.”).

[162] Michal S. Gal & Oshrit Aviv, The Competitive Effects of the GDPR, 16 J. Competition & Econ. 349 (May 18, 2020). See also James Campbell, Avi Goldfarb, & Catherine Tucker, Privacy Regulation and Market Structure, 24 J. Econ. & Mgmt. Strategy 47, 68 (2015) (“[A] potential risk in privacy regulation is the entrenchment of the existing incumbent firms and a consequent reduction in the incentives to invest in quality. These incentives are stronger when firms have little consumer-facing price flexibility, as is the case in online media.”).

[163] Gal & Aviv, id. at 16.

[164] See, e.g., Cheok Lup, Explaining Marketing Attribution Models [Scenario Example], tinkerEdge (Nov. 12, 2015), https://www.tinkeredge.com/blog/web-analytics/explaining-marketing-attribution-models. (“On Day #1: User wants to purchase a coffee table for his new house, and perform a keyword search on Google. He clicks on one of the organic listings on Google Search Engine Result Page (SERP) to land onto Overstock.com. On Day # 2: He continues his search for his coffee table, and clicks on one of the PPC ads on Google SERP to land onto Overstock.com again. He subscribes to the email newsletter this time. On Day #3: He receives an eDM [electronic direct mail] from Overstock.com with a promotional offer of 30% discount sale, and clicks the “Buy Now” button from the eDM to enter the website. Unable to resist the discount offer, he decides to make a purchase of the furniture from the website.”). Attribution metrics determine which channel gets credit for the ultimate sale.

[165] Scott Morton & Dinielli, supra note 15, at 29

[166] “Furthermore, the default makes the advertiser believe that search ads are very effective relative to display ads, so the advertiser has no reason to change the default.” Id.

[167] See About the Default MCF Attribution Models: Learn How Each MCF Model Assigns Conversion Credit, Google Analytics Help (last visited Nov. 1, 2022), Attribution Models, Google, https://support.google.com/analytics/answer/1665189?hl=en. For the Google Analytics 4 version of these attribution models currently being implemented, see [GA4] About Attribution and Attribution Modeling, Google Analytics Help (last visited Nov. 1, 2022), https://support.google.com/analytics/answer/10596866. Google even created a guide called “Beyond Last Click Attribution” to help advertisers select the most appropriate model. See Beyond Last Click Attribution: Official Guide to Attribution Modeling in Google Ads, Google Ads Help (last visited Nov. 1, 2022), https://support.google.com/google-ads/answer/7003286.

[168] See Joan Arensman & Wilfred Yeung, Move Beyond Last Click Attribution in AdWords, Google Blog (May 10, 2016), https://adwords.googleblog.com/2016/05/move-beyond-last-click-attribution.html.

[169] How Does Conversion Tracking Work?, Microsoft Advertising (n.d.) https://help.ads.microsoft.com/#apex/ads/en/56710/2.

[170] Nir Elharar, How to Choose the Right Marketing Attribution Model for Your Content, Outbrain (Apr. 8, 2019), https://www.outbrain.com/blog/marketing-attribution-model-content.

[171] Texas Complaint, supra note 1, at ¶ 195.

[172] Scott Morton & Dinielli, supra note 15, at 17

[173] Texas Complaint, supra note 1, at ¶ 351.

[174] See Demsetz, supra note 81.

Executive Summary

Section 512 of the Copyright Act, passed as part of the Digital Millennium Copyright Act of 1998, was created to preserve “strong incentives for service providers and copyright owners to cooperate to detect and deal with copyright infringements that take place in the digital networked environment,” while also providing “greater certainty to service providers concerning their legal exposure for infringements that may occur in the course of their activities.” The idea was to provide a safe harbor to online service providers (OSPs) that would also help to fight piracy.

In practice, Section 512 has reduced OSPs’ liability risk and thereby promoted the growth of distribution services. Piracy, however, has grown exponentially. Among the factors driving this growth is that the courts’ interpretation of Section 512 has pushed OSPs toward a reactive “file-containment” approach, rather than encouraging them to seek proactive solutions to piracy on their services. Indeed, citing both technological change in the intervening years and the judicial construction of Section 512, the U.S. Copyright Office has concluded that “Congress’ original intended balance has been tilted askew.” It would therefore be appropriate for Congress to revisit the law, applying lessons learned over the more than two decades since its enactment.

Much of the challenge in combating online piracy stems from the “volume problem.” The amount of content that traverses online services makes it unreasonable to expect OSPs to catch every intentional or inadvertent infringement by their users. But the amount of infringement that slips through and harms copyright holders is nonetheless substantial, accelerated by technological innovations like more comprehensive search engines, faster upload and download speeds, and the emergence of peer-to-peer file-sharing services.

One solution would be for OSPs to license content directly from rightsholders. As intermediaries, OSPs can potentially license content more efficiently and cost-effectively than individual copyright holders and users of Internet services could negotiate licenses among themselves. Licensing by OSPs could also remove litigation risk, enable their service’s users to benefit from the content, and ensure copyright holders’ rights are respected. TikTok, Facebook, Snapchat, Instagram, and YouTube, for example, are increasingly licensing at least some content, so that their users have an authorized way to incorporate that content into posts and streams.

Unfortunately, the current safe-harbor regime gives OSPs little incentive to license content or otherwise proactively deter pirated content on their services, insofar as they can presumptively monetize infringing content until rightsholders issue takedown notices. Under Section 512, to be protected by the safe harbors, OSPs must have neither “actual knowledge” of infringement, nor “red-flag knowledge”—i.e., awareness of facts that make infringement apparent. Judicial interpretations of Section 512, however, have essentially collapsed the red-flag standard into the actual-knowledge standard, while progressively narrowing the scope of the actual-knowledge standard; the bar for legally relevant knowledge of infringing activity is now quite high.

To address this, the standard for when an OSP is considered to have “knowledge” of infringement ought to be changed from what an “ordinary” person might infer from the circumstances to what a reasonable person in the user-generated content business would infer, even absent notification by a rightsholder. This broader knowledge standard would then be used to condition the safe harbors offered by Section 512 on OSPs taking reasonable steps both to prevent infringement and to remove that infringing content that does slip through.

Even where OSPs do not host or display infringing content, they may sometimes facilitate its dissemination by others. To be eligible for the safe harbors, OSPs should be obligated to provide the identity of infringing parties and to prevent further access to the infringing content, even when the OSPs are not at fault for the underlying infringement. Around the world, these sorts of “no-fault injunctions” have been used effectively to combat piracy with no interference with OSPs’ normal operations. Indeed, in some cases, private companies have voluntarily partnered with rightsholders to restrict access to content that a court has declared infringing.

Congress originally expected OSPs to collaborate with rightsholders in the development of standard technical anti-piracy measures, such as filtering. In the nearly quarter century since Section 512’s enactment, however, no standard technical measures have emerged. Recently proposed legislation—the SMART Copyright Act—endeavors to fix this problem. It would empower the Office of the Librarian of Congress to engage in rulemaking proceedings to develop standard technical measures with the relevant multistakeholder community. Despite some ambiguities and shortcomings in the bill’s text, it offers a promising framework to address one of Section 512’s longstanding deficiencies.

Finally, Section 512 requires OSPs to have policies to terminate service to repeat infringers and to reasonably implement those policies. Courts, however, have interpreted these requirements loosely. The purpose of the safe harbors is to provide platforms greater certainty regarding litigation risk when they act responsibly and to assure copyright holders that their rights will be reasonably protected in exchange for the liability limitations the platforms receive. That bargain is not achieved unless the platforms and their users know that costly repeat infringement will not be tolerated. To better address this goal, the Copyright Office should be authorized to provide guidance on the minimum requirements necessary to meet the repeat-infringer policy obligation, including by creating a model repeat-infringer policy that will be presumed to comply.

Introduction

The birth of the commercial Internet was among the most important technological developments of the past century and crucial to its success have been the rules governing the Internet’s use. These include architectural rules to address such issues as management of the Domain Name System (DNS) and the resolution of disputes over domain names. Other important rules concern the relationships between and among commercial entities and individuals operating online. They include such familiar legal rules as torts, copyright, and antitrust.

In certain respects, these rules of general applicability apply differently on the Internet than they do in other contexts, typically to address the scale, scope, and speed with which information can be shared online via intermediaries (i.e., digital platforms). This leads to the general legal presumption that intermediary liability should be specially tailored online to ensure that business models that rely on user-generated content can thrive, while also suitably protecting broader social interests.  Section 230 is one of the most well-known U.S. laws dealing with this subject, and it has received significant attention, aimed at understanding how best to frame policy around intermediary liability online. As we have noted in previous work on that subject, intermediary-liability laws should be focused on balancing the benefits that platforms can provide with the negative externalities they can generate.[1] Any legal policy that requires intermediaries to moderate more than they currently do will remove some harmless content; the relevant question, however, is whether the marginal reduction in harmless speech is justified by the marginal increase in the deterrence of illegal content. In essence, crafting intermediary-liability policy is about conducting a cost-benefit analysis that implicitly assumes that the goal is optimally to minimize both the loss of user-generated content and the harms of illegal activity facilitated by platforms.[2]

Given the uncertainties and complexity in locating that middle ground between costs and benefits, erring on the side of granting full immunity from third-party liability­­ to platforms—as Section 230 largely does—may conceivably be the best possible outcome. Without strong evidence to support this position, however, it is highly improbable that a legal regime that results in complete immunity for platform operators for the harmful activity that can occur on their services is socially optimal. This is particularly true, given the long history of common-law attempts to parse exactly this kind of liability for offline intermediaries, which suggests that courts and lawmakers can indeed shape regimes to allow both liability and room for firms to operate their services.[3]

In other words, simply claiming that costs would rise if intermediaries were held liable, or that liability itself is a harm to the platforms, is insufficient. Liability has some harmful consequences everywhere. The relevant question is whether those harms outweigh the ones avoided by not imposing legal accountability.

It is notable that copyright protection was specifically exempted from the liability shield created by Section 230.[4] It is also notable that the subject of this paper—Section 512 of the Copyright Act (“Section 512”)—adopted, at least in theory, a different approach than Section 230. But as we discuss below, while Section 512 did not create a complete liability shield, its practical effect, largely through judicial interpretation, has come closer to that end than its drafters likely envisioned.

Section 512 altered key elements of how secondary-liability rules for copyright are applied online. Most notably, Section 512 created a safe harbor for online service providers (“OSPs”) for potentially infringing content generated by users of their platforms.[5]

Nearly a quarter-century after its passage, Section 512 is due for reform. When Congress added Section 512 to the Copyright Act, it had two things in mind. First, that copyrighted content merited protection online, just as it did offline. Second, that the then-nascent OSPs would struggle to bear the full weight of direct or secondary liability for all the copyrighted content their users might disseminate without authorization. Thus, Congress intended Section 512 to preserve “strong incentives for service providers and copyright owners to cooperate to detect and deal with copyright infringements that take place in the digital networked environment,” while providing “greater certainty to service providers concerning their legal exposure for infringements that may occur in the course of their activities.”[6]

Given the monumental challenges that OSPs would face in trying to prevent any unauthorized dissemination of copyrighted content by their users, they particularly feared the precedent set by 1993’s Frena decision, which imposed direct copyright liability on a bulletin-board operator for storing infringing images uploaded by users.[7] To address the perceived threat to OSP viability if Frena were broadly followed, Section 512 contains safe harbors that essentially codify the precedent in 1995’s Netcom case, which instead countenanced only secondary liability for OSPs that host infringing content.[8] Broadly speaking, the law grants OSPs conditional immunity for unwittingly disseminating unlicensed copyrighted material without authorization.^[9] Importantly, Section 512’s immunity is conditioned on OSPs acting to curb infringement once they have actual knowledge of its existence, such as when notified by the copyright owner, or when the infringement is apparent. This latter category of apparent infringement is sometimes referred to as “red flag” knowledge.[10]

The regime Section 512 established has produced mixed results in practice. By enabling OSPs to transmit content across the Internet at greatly reduced risk of liability, Section 512 has, without question, facilitated the rapid growth of distribution services that also benefit content producers and consumers. At the same time, however, the proliferation of pirated content has grown exponentially. The law imposes little obligation on OSPs to mitigate the dissemination of infringing content other than to react ex post—at which point, the damage has already been done.

It is therefore not surprising that there have been growing calls for stronger copyright and other forms of content protection, both in the United States and around the world. Australia’s Competition and Consumer Commission (ACCC) has moved to impose “neutrality” requirements on tech platforms,[11] as well as to curtail platforms’ ability to monetize news content without licensing it at what news organizations believe are fair rates.[12] The European Union passed reforms to its Copyright Directive intended to provide greater protection for rightsholders.[13] And following years of public meetings and stakeholder input, the U.S. Copyright Office published its long-anticipated Section 512 report, which concluded that the safe harbors should be adjusted to better address online piracy, as “Congress’ original intended balance has been tilted askew.”[14]

In this paper, we examine whether Section 512 set the right balance between, on the one hand, mitigating unreasonable copyright-litigation risks that may arise from user-generated content, and on the other, holding online platforms accountable when they unreasonably fail to curb foreseeable infringement risks. We ultimately conclude, as the Copyright Office did, that the law’s proper balance has been tilted askew. We therefore recommend adjustments to Section 512.

I. The Development of Section 512

When Congress passed the Digital Millennium Copyright Act (“DMCA”) in 1998, only about 30% of the U.S. population used the Internet in any fashion,^[15] and only 12% of American adults reported daily online use.^[16] Indeed, while the Internet’s potential to dramatically alter the way consumers access information and buy all manner of goods and services had become clear by the mid-1990s, much of the technology that would come to shape the Internet as we know it remained either still in its infancy or did not exist at all. Amazon didn’t start selling books until 1995,[17] MP3.com wasn’t launched until 1997,[18] and Google wasn’t founded until 1998.[19] Napster, one of the earliest drivers of massive-scale digital piracy, didn’t exist until 1999.[20] Facebook wasn’t launched until 2004.[21] YouTube started in 2005 and was bought by Google in 2006.[22] Twitter became a company in 2007.[23]

By contrast, recent estimates find that, today, 81% of Americans have mobile Internet access, while 28% of Americans say they are online “almost constantly.”[24] Global Internet access has jumped from 3.14% in 1998 to more than 50% in 2020.[25] This explosion of online access has benefitted consumers and businesses in ways that weren’t necessarily obvious in 1998. Amazon’s e-commerce revolution, for example, benefitted not just Amazon, but a host of firms that wished to exploit the Internet as a distribution channel. At the same time, service has offered consumers a fast, convenient, and affordable way to shop for nearly every conceivable product. Similar revolutions have been seen in media and the arts, where digital distribution has provided users an outlet for their own creativity, while creating another channel for traditional media entities to reach audiences.

But Internet distribution reduces friction not just for legitimate transactions, but also for unlawful ones.[26] The question becomes how to combat that unlawful activity without hindering the development of legitimate content and the online platforms that distribute it—either through excessive regulation or unending litigation risk stemming from user-generated content.

It was this question that Congress sought to address in the mid-1990s when it attempted to balance two competing interests regarding copyright policy online. On the one hand, it recognized that the incipient Internet platforms of the day would have great difficulty if they were subject to direct or secondary copyright liability for all their users’ posts. On the other hand, rightsholders had a valuable interest in protecting their works. The balance Congress struck is embodied in Section 512, which was added to the Copyright Act through passage of the Digital Millennium Copyright Act of 1998.^[27]

Importantly, Section 512 did not create an absolute shield for OSPs against copyright-infringement claims. Rather, Section 512 created a set of “safe harbors” that would grant various types of OSPs protection from copyright claims arising from user-generated content, provided the service providers promptly took down instances of infringement.[28] The safe harbors cover four specific categories of activity. Section 512(a) covers OSPs that merely serve as conduits for material directed at third parties;[29] Section 512(b) covers OSPs that temporarily cache content as it is being transmitted;[30] Section 512(c) covers OSPs that host material for third-party users;[31] and Section 512(d) covers OSPs that “link” to content—for example, search engines or directories.[32] Each of Section 512’s safe harbors imposes certain obligations on OSPs before they will merit protection from liability. For example, service providers are required to comply with a notice-and-takedown procedure,[33] as well as to act on both “apparent” and “actual” knowledge of infringement.[34]

As noted above, Section 512 was passed at a relatively immature stage in the development of online technologies. At the time, the web existed largely as a collection of static, primarily text-based pages. Usenet also existed and was then a major conduit for pirated content, although its relative importance has since declined.[35] By today’s standards, search technologies were crude^[36] and, while there undoubtedly were private servers dedicated to content-sharing, it would have been difficult to find such servers in the relatively disorganized web of the day. High-speed Internet service was also rare and, outside of universities and large corporations, most users connected to the Internet via analog modems and telephone lines.^[37] Moreover, the peer-to-peer services that would make file-sharing more efficient effectively did not exist.^[38] And yet, even at the time Section 512 was enacted, it was understood that “[d]ue to the ease with which digital works can be copied and distributed worldwide virtually instantaneously, copyright owners will hesitate to make their works readily available on the Internet without reasonable assurance that they will be protected against massive piracy.”^[39]

Thus, even in the Internet’s infancy, when broadband speeds were a tiny fraction of today’s and with only rudimentary file-sharing services available, Congress believed that “[w]ith this… evolution in technology, the law must adapt in order to make digital networks safe places to disseminate and exploit copyrighted materials.”^[40]

A.   The evolving legal backdrop

Section 512 emerged in response to federal case law regarding online copyright infringement. Two cases in particular—Playboy Enterprises Inc. v. Frena and Religious Technology Center v. Netcom On-Line Communication Services Inc.—represented the dominant poles of jurisprudence.

In Frena, Playboy Enterprises sued George Frena, the operator of an online bulletin-board system.[41] The complaint concerned copyrighted photos that were stored on Frena’s servers and had been uploaded by users of the bulletin board without Frena’s knowledge.[42] As soon as he became aware of the infringing material, Frena removed it.[43] The court, however, framed Frena’s activity as one of direct infringement, rather than contributory infringement.[44] After walking through a fairly routine copyright analysis, it found Frena liable.[45]

By contrast, the Netcom decision two years later framed the relevant copyright analysis differently. In Netcom, an affiliate organization of the Church of Scientology sued the operator of a bulletin board, Thomas Klemesrud, and his Internet-service provider (ISP), Netcom, for hosting portions of copyrighted works that it owned.[46] The works were not posted by either Klemesrud or Netcom, but by a user of Klemesrud’s bulletin board named Dennis Erlich.[47] The court dismissed the direct infringement claims, characterizing storage by a bulletin-board operator and transmission by an ISP as “incidentally making temporary copies,” and thus insufficiently tangible to support a direct infringement claim.[48]

The court, however, went on to examine the plaintiffs’ other claims of contributory infringement, as well as vicarious liability.[49] In this regard, it found that the plaintiffs had raised genuine issues of fact: whether Netcom and Klemesrud had sufficient knowledge that an infringement was occurring, and whether they were in positions to stop such infringement.[50] Consequently, the court denied the motion for summary judgment, and allowed the case to proceed on the secondary-liability theories.[51]

Thus, under either dominant approach to examining the infringing acts of users of an online service, there was a distinct possibility that providers could be found liable. At the same time, some rightsholders objected that basing service providers’ liability on their having sufficient knowledge of infringing activities (however that term would come to be defined) would encourage OSPs to choose to be willfully blind.[52]

In response to these cases, Congress drafted and passed Title II of the Digital Millennium Copyright Act, which was subsequently codified as Section 512 of the Copyright Act:

There have been several cases relevant to service provider liability for copyright infringement.  Most have approached the issue from the standpoint of contributory and vicarious liability. Rather than embarking upon a wholesale clarification of these doctrines, the Committee decided to leave current law in its evolving state and, instead, to create a series of “safe harbors,” for certain common activities of service providers. A service provider which qualifies for a safe harbor, receives the benefit of limited liability.[53]

Section 512’s explicit goal was to balance the competing interests of rightsholders and service providers in a way that preserved strong incentives for service providers and copyright owners to cooperate in detecting and resolving copyright infringements in the digital networked environment.[54] At the same time, it provided greater certainty to service providers concerning their legal exposure for infringements that may occur in the course of their activities.[55]

II. The Costs and Benefits of Section 512

It is to be expected that a legal regime constructed before the commercial Internet had truly taken shape would require updates as new harms and user behaviors emerged. With the benefit of hindsight that the drafters of Section 512 lacked, this section offers an overview of the positive and negative effects that the law has produced.

A.   The positive effects of Section 512

Section 512 has delivered on at least some of its promise. Online-distribution services have grown dramatically: transforming entire industries; significantly altering how we consume books, music, and videos; and increasing the availability of creative works, including user-generated works. Without a doubt, OSPs have generated enormous benefits to society,[56] and a substantial proportion of those benefits have come from legitimate dissemination of high-quality content (e.g., Apple Music, Spotify, Netflix, Amazon Prime Video, etc.).

It would be difficult to quantify the full value that online services carrying user-generated content have offered society, but it is surely quite large. Just seven years after Section 512’s passage, YouTube was founded as a platform for individuals to store and stream everything from their own amateur films and home movies to grassroots outreach on important civic and political issues.[57] It now draws 2 billion monthly users.^[58] Facebook has 1.9 billion daily users.^[59] Google processes at least 2 trillion searches annually.^[60] And all of these services are nominally free to users, thanks to the platforms’ ability to monetize their services successfully with ad revenue. In the process of hosting and serving ads, the platforms also generate value for advertisers, who are better able to match their offerings to targeted users, and for users, who receive more relevant offers for products and services.

The rise of closed systems like iTunes (now Apple Music) and Spotify also demonstrates the benefits that can flow, in part, from predictable legal liability around digital content. Numerous third-party services allow content creators to feed their music into Apple’s and Spotify’s ecosystems.[61] It would be virtually impossible to vet rights claims at scale for every piece of content that these services host. Platforms like Bandcamp and Soundcloud provide services analogous to YouTube, but for independent musicians. Those sites allow millions of individual artists to release and market their music to a broad array of consumers. And social media like Reddit, Facebook, Instagram, and SnapChat would similarly be unable to offer multimedia-sharing services to millions or billions of users without something like the liability-limiting provisions of Section 512.

Research published in the Proceedings of the National Academy of Sciences (PNAS), using willingness-to-accept choice experiments, finds that users assign relatively high values to Internet services.[62] For example, survey respondents in 2017 indicated they would require the following payments to give up each of the following services for one year (95% confidence interval):

Search engines:	$14,000-22,000
Email:	$6,900-10,200
Maps:	$2,700-5,100
Video streaming:	$940-1,490
E-commerce:	$700-1,000
Social media:	$240-430
Messaging:	$115-210
Music:	$130-215

 

Authors Erik Brynjolfsson, et al., believe that one explanation for the high valuations that users place on Internet services is that many see them as essential to their jobs and would thus be reluctant to give them up, even for significant compensation. Moreover, the authors argue, because most consumers do not pay for these services directly, nearly all of their willingness-to-accept represents consumer surplus.

There are also intangible benefits that flow from Internet platforms and user-generated content.  For example, experiments comparing Internet versus non-Internet research have found that searchers are more likely to find an answer to a question using Internet search; that an Internet search takes significantly less time to complete; and that searchers will consult significantly more sources on the Internet.[63] When searching factual questions for which there is a clear correct answer, Internet searches are significantly more likely to find the correct answer.

B.   The negative effects of Section 512

Along with the value created by online platforms, however, has also come the widespread, unauthorized dissemination of copyrighted content. This has almost certainly diminished some of the investment-backed expectations of content creators and rightsholders; raised costs for producers, and, thus, consumers; and reduced the quantity, breadth, and quality of content available to audiences. Early on, for example, YouTube quickly became a major venue for users to upload copyrighted content illegally,[64] while other legitimate social-media platforms similarly have hosted large quantities of unauthorized content.[65] The past quarter-century has, of course, also seen the development of myriad other sites specifically dedicated to the mass, unlawful dissemination of copyrighted content.[66]

At the same time, direct piracy is not the only negative effect suffered by rightsholders. The broad dissemination of pirated content as an alternative to legitimate content, coupled with Section 512, has placed downward pressure on the value of licenses for content. The Phoenix Center’s T. Randolph Beard, George S. Ford, and Michael L. Stern conducted an economic analysis examining the distortions that online platforms have had on the market for licensing by examining the relationship between market rates and the rates that YouTube paid for digital music.[67] They concluded that YouTube’s use of Section 512, given the widespread piracy on its service, “reduces revenues to artists and labels in the U.S. by at least hundreds of millions and by perhaps more than one billion dollars each year.”[68] If YouTube were to pay rates closer to the market level, it would generate between $650 million and $1 billion in additional revenue for content creators.[69]

This is not to suggest that OSPs like YouTube do not attempt to control piracy on their services. Soon after acquiring YouTube, Google developed its innovative ContentID “fingerprint” filtering system.[70] Such systems detect attempts to upload unauthorized content, allowing copyright holders to determine whether to permit dissemination and whether to monetize it.[71] While this is the sort of technological innovation that Congress hoped to encourage with Section 512, filtering systems are available on only a few platforms and they extend primarily to the largest copyright holders, often to the exclusion of smaller content creators.[72] Moreover, such systems do nothing to stem access to infringement on dedicated piracy sites, whose unlawful offerings can be found relatively easily, whether through word of mouth, linking sites, or search services.[73]

Compounding matters, there have been significant changes in the legal and technological landscape over the last two decades that undermine some of the assumptions underlying Section 512. As we discuss at length in the remainder of this paper, judicial interpretations have systematically diminished the effectiveness of many of the provisions of Section 512 that were meant to aid rightsholders in combatting piracy. Although red-flag knowledge of suspicious activity that could be infringement was supposed to constitute grounds on which OSPs were expected to act, judicial interpretation of Section 512’s relevant provisions has so narrowed the scope of those grounds as to render them a nullity.[74] Rightsholders have, moreover, largely been unable to seek adequate redress in U.S. courts by seeking the sorts of no-fault injunctions that have been widely successful in other jurisdictions.[75] And the original standard technical measures envisioned by Section 512’s drafters, which would have enabled collaborative approaches between platforms and rights holders to control piracy, have not adequately materialized.

III. The Costs and Extent of Piracy

From some perspectives, the balance embodied by Section 512 has largely worked well.[76] As Jennifer Urban, et al., characterized the current enforcement around Section 512:

Overall, the fundamental compromise in section 512—to manage liability and enforcement costs for OSPs and rightsholders—holds in essence. The basic compromise still underpins negotiations between OSPs and rightsholders over responsibility as Internet services and distribution channels evolve.[77]

For those who take this perspective, where the basic compromise has failed, it has done so in the direction of over-enforcement, suggesting that the protections Section 512 offers to rightsholders should be limited further still.[78] More directly, the primary concern of many who support the status quo is the extent to which Section 512 could be used as a tool to stifle free expression, not to facilitate piracy.[79]

Nonetheless, stakeholders on both sides of the Section 512 bargain generally acknowledge that the sheer scale of online piracy has tended to overwhelm the law’s notice-and-takedown provisions.[80]

Developed in a world that operated at dial-up speeds, Section 512 was targeted at isolating infringing content and preventing its spread. When an infringing file appeared, the relatively slow speeds of dial-up Internet access were a natural barrier that tended to prevent rapid dissemination, giving platforms and rightsholders time to issue a series of notices and counter-notices to stop further infringement. But the underlying assumptions of that bargain have been undermined by the advance of technology. Today, a file-containment approach leads to the well-known game of copyright “Whac-A-Mole” that does little, if anything, to control the spread of massive online piracy.[81]

A core defect of the current Section 512 regime is that it places little onus on platforms to prevent either the initial unlawful dissemination or the repeat posting of files that are known (or easily knowable) to be infringing. While Google faces a gargantuan task in processing millions of takedown notices, rightsholders face an even larger collective challenge in searching across all platforms to discover, investigate, and report on cases of infringement.^[82] Many do not have the resources of major movie studios or record labels and must make tough decisions about how to adequately police infringement of their property.

U.S. consumers logged 725 million visits to pirate sites for movies and television programming in April 2020 alone.[83] Close to 90% of those visits were attributable to illegal streaming services.[84] In the United States, there are more than 9 million subscribers to Internet protocol television (IPTV) services specializing in pirated content, which reap more than $1 billion annually in ill-gotten gains.[85] Globally, there are more than 26.6 billion illicit viewings of U.S.-produced movies and 126.7 billion illicit viewings of U.S.-produced television episodes each year, annually costing U.S. rightsholders between $30 and $70 billion, costing the sector between 230,000 and 560,000 jobs, and costing the overall economy between $45 and $115 billion in GDP.[86]

For larger rightsholders, policing this infringement represents a significant cost (albeit one that addresses only a fraction of the online infringement that affects them). For smaller rightsholders, it can be a prohibitive barrier that prevents them from effectively policing unlicensed use of their property and existentially threatens their livelihood.[87] The Copyright Office took note of the explosion of piracy since the mid-1990s in its Section 512 report:

[B]etween 1998 and 2010, Google received notices for less than three million URLs containing content that allegedly infringed a copyrighted work. The scale of notices grew with time, and in 2013, Google received notices for approximately three million URLs—more than the total received by Google during the previous twelve years. Since then, the volume of infringement notices has rocketed up. In 2017, Google received notices identifying about 882 million URLs, and has processed requests to delist more than 4.6 billion URLs for copyright violations to date.[88]

From January to June 2021, Microsoft reported receiving more than 11 million takedown requests involving 103 million URLs.[89] The company rejected only about 0.33% of these requests.[90]

Because takedowns occur after illicit dissemination, by the time the takedown process is initiated, copyright holders by definition have already suffered significant harm to their exclusive rights to determine whether, how, and under what terms their content may be disseminated. Moreover, the future market for the works likely has been impeded, as at least some portion of the future audience will probably be able to access the content at no cost. Recognizing the smaller addressable market, distributors (whether online or traditional media) will offer copyright holders less to license the content than they would have if the illicit dissemination had never occurred. The volume of takedown notices offers evidence of the massive scope of online piracy, which in turn can have a huge effect on a copyright’s value to the rightsholder.

The status quo is also not without cost to the OSPs. Complying with the volume of requests generated pursuant to Section 512’s safe-harbor requirements entails significant investment of resources.[91] Just as smaller rightsholders face disproportionately large challenges in policing infringement of their copyrights, it is smaller platforms and new market entrants that are least able to bear the costs of safe-harbor compliance.

There is, moreover, the potential problem of takedown notices that were filed fraudulently. Estimating real costs in this regard is difficult, as the costs of compliance depend on several factors, including internal technology and compliance staff. Some sense of the scale is available, however, from public data. For example, Automattic—the makers of a host of popular web-publishing software, including WordPress and WooCommerce—reports that, generally, between 5% and 10% of the takedown requests they receive are “abusive.”[92] In 2021, this accounted for about 530 notices.[93] A larger provider like Google faces substantially more abusive takedown demands. In one study of a single fraudulent effort to force takedowns by misrepresenting ownership of content, a researcher discovered 33,988 illegitimate takedown efforts.[94] Even beyond fraud, erroneous takedowns can be a problem, with another study estimating that up to 30% of the takedown notices in its sample were potentially in error.[95]

IV. Legal Developments in the Section 512 Regime

There are two general trends that can be observed in the evolution of Section 512’s legal standards: toward relatively less participation on the part of platforms to deter illegal content on their services and toward greater burdens on rightsholders to police piracy of their content.[96] While the platforms have enjoyed exponential growth and the accompanying financial rewards, the law has not always kept pace with that growth by ensuring that platforms more properly internalize the social costs of their activity.^[97]

For example, courts have consistently interpreted Section 512’s grant of immunity as being almost completely undisturbed by red-flag knowledge.^[98] Courts have also interpreted a key rightsholder’s ability to seek subpoenas and injunctions against actual infringers in so restrictive a manner as to effectively neuter that section of the law.[99] Further, in practice, the ways that OSPs process takedown notices essentially requires copyright holders to proceed URL-by-URL—a linear process doomed to failure and frustration in the face of logarithmic piracy.

A.   Knowledge, red-flag knowledge, and the duty to monitor

To receive the benefit of a Section 512 safe harbor, OSPs engaged in hosting or search services must act to address copyright infringement by users of their services when they have either: 1) actual knowledge of infringement, or 2) awareness of facts that make it apparent that infringement is occurring—i.e., red-flag knowledge.[100]

Despite some suggestions to the contrary from the U.S. Supreme Court,[101] courts have not generally imposed a legal obligation on OSPs to proactively mitigate infringement by their users to qualify for the safe harbor; rather, courts have instead only required service providers to curtail infringement after the fact. In this regard, courts have relied upon Section 512(m), which explicitly declines to condition application of a safe harbor on an OSP “monitoring its service or affirmatively seeking facts indicating infringing activity, except to the extent consistent with a standard technical measure complying with the provisions of subsection (i).”[102] There is room, however, between actively monitoring to discover evidence of actual or impending infringement and taking preventative measures to avoid infringement where such evidence presents itself—either because it has been affirmatively called to the OSP’s attention or has otherwise become apparent.

Indeed, the legislative history of Section 512 describes actual and red-flag knowledge as two distinct ways through which OSPs may become aware of infringing material that requires action on their part.

[A] service provider need not monitor its service or affirmatively seek facts indicating infringing activity… in order to claim this limitation on liability (or, indeed any other limitation provided by the legislation). However, if the service provider becomes aware of a ‘‘red flag’’ from which infringing activity is apparent, it will lose the limitation of liability if it takes no action. The ‘‘red flag’’ test has both a subjective and an objective element. In determining whether the service provider was aware of a ‘‘red flag,’’ the subjective awareness of the service provider of the facts or circumstances in question must be determined. However, in deciding whether those facts or circumstances constitute a ‘‘red flag’’—in other words, whether infringing activity would have been apparent to a reasonable person operating under the same or similar circumstances—an objective standard should be used.[103]

To be sure, this can be a complicated standard to adjudicate. A court was expected, first, to determine whether an OSP had subjective red-flag knowledge of infringement. Next, if a court found that the OSP did not have subjective knowledge, it would have to determine if the lack of such knowledge was objectively reasonable. If the OSP’s lack of knowledge was objectively unreasonable, the OSP would be required to remove the infringing material or lose the safe harbor. Thus, in the original formulation of Section 512, actual or red-flag knowledge of infringement could theoretically arise from a range of potential situations: rightsholders pointing out a violation, an employee discovering (either through automated functions or plain observation) such material, or from evidence that would lead a reasonable person to recognize that infringement might be occurring.

A series of court decisions, however, have significantly enhanced the requirements to meet these knowledge standards. In the Viacom case, the 2^nd U.S. Circuit Court of Appeals described actual knowledge as whether the OSP “‘subjectively’ knew of specific infringement.”[104] Other circuits and district courts have largely followed this view of actual knowledge.[105]

As for red-flag knowledge, the Viacom court described it as turning “on whether the provider was subjectively aware of facts that would have made the specific infringement ‘objectively’ obvious to a reasonable person.”[106] Along these lines, the 9th U.S. Circuit Court of Appeals ruled that “general knowledge” that an entire category of hosted content was likely to contain copyrighted material was insufficient to create red flag awareness in an OSP.[107] Curiously, the 9th Circuit held that red-flag knowledge didn’t exist even when a suspected infringer went so far as to label its files “stolen” or “illegal,” as such labels merely increased the “salacious appeal” of the content.[108]

In effect, courts have collapsed the distinction between red-flag knowledge and actual knowledge by disallowing “red flags” to arise from general awareness of infringement on a service. The 9th Circuit’s Veoh opinion is emblematic:

Although the parties agree, in retrospect, that at times there was infringing material available on Veoh’s services, the DMCA recognizes that service providers who do not locate and remove infringing materials they do not specifically know of should not suffer the loss of safe harbor protection.[109]

As noted above, the 2nd Circuit’s Viacom decision essentially agreed with the requirement that subjective knowledge of infringement be “objectively reasonable” to avoid constituting red-flag knowledge.[110] Yet it also followed the 9th Circuit’s approach of pinning the knowledge requirement to “specific” acts of infringement. This substantially narrows the circumstances under which an OSP could theoretically be said to form a reasonably subjective view of potential infringement.[111]

It is helpful to unpack the 2nd Circuit’s 2016 Vimeo decision to understand the scope of the problem. Vimeo is a website that allows users to post videos. Several record labels and music publishers sued Vimeo for direct, contributory, and vicarious copyright infringement, documenting at trial a variety of Vimeo employee messages about incorporating copyrighted songs in uploads. Examples included three members of the Vimeo content-moderation team individually:

• telling a user that Vimeo allowed uploading “lip-synch” videos containing copyrighted music, but would take them down if and when asked by a rightsholder;
• telling a user, “don’t ask, don’t tell,” in response to a question about including copyrighted music in original videos the user created;
• saying to a user “[w]e can’t officially tell you that using copyright music is okay. But…” in response to a question about using a song by the band Radiohead;
• telling fellow employees that she was “[i]gnoring, but sharing” internally a user message that included a link to a video and asked what Vimeo did about people using copyrighted music on Vimeo; and
• responding to a user question about including Bobby McFerrin’s “Don’t Worry, Be Happy” in a home video that “[t]he Official answer I must give you is: While we cannot opine specifically on the situation you are referring to, adding a third party’s copyrighted content to a video generally (but not always) constitutes copyright infringement under applicable laws… Off the record answer… Go ahead and post it….”[112]

The company’s vice president of product and development also sent a message to two members of the content-moderation team and every employee in the “[email protected]” email group asking: “Who wants to start the felons group, where we just film shitty covers of these [EMI] songs and write ‘FUCK EMI’ at the end?”[113]

The 2nd Circuit nonetheless ruled that the messages in this case did not constitute red-flag knowledge because they did not relate to the specific infringement claims at issue and because the mere viewing by employees of videos containing recognizable songs would not be sufficient.[114] Citing its prior decision in Viacom, the court said that, to possess red-flag knowledge, the service provider must be subjectively aware of facts that would make the specific infringements at issue objectively obvious to a reasonable person.[115] Moreover, that “reasonable person” is “an ordinary person” without any expertise regarding music or copyright law, and without any obligation to investigate whether the content is copyrighted or the poster is engaged in a licensed or fair use.[116]

According to the court, it would not be sufficient to establish red-flag knowledge that there were facts that would lead a reasonable person to infer that infringement occurred. Rather, the service provider must have actual knowledge of the significance of those facts, and those facts would need to lead an ordinary, reasonable person to infer infringement was occurring.[117]

This appears to be both a poor interpretation of the statute and bad policy. Indeed, the Copyright Office does not appear to believe that Congress intended to erect so high a bar as the Vimeo court and others have suggested:

The Office believes a standard that requires an OSP to have knowledge of a specific infringement in order to be charged with red flag knowledge has created outcomes that Congress likely did not anticipate. The Copyright Office reads the current interpretations of red flag knowledge as effectively removing the standard from the statute in some cases, while carving an exceptionally narrow path in others that almost requires a user to “fess up” before the OSP will have a duty to act. OSPs are correct that Congress likely did not intend to adopt a general awareness standard for red flag knowledge, since such a standard would consume many OSPs Congress otherwise sought to protect. Yet courts have set too high a bar for red flag knowledge, leaving an exceptionally narrow space for facts or circumstances that do not qualify as actual knowledge but will still spur an OSP to act expeditiously to remove infringing content.[118]

B.   Subpoenas and injunctions

Another notable trend in the law, both in the United States and abroad, has been the growing challenges rightsholders’ face in seeking to identify suspected infringers. Under the EU’s General Data Protection Regulation, for example, it has become significantly more difficult to obtain valid WHOIS contact information for the owners of domains that host infringing content.[119] In a similar vein, the European Commission has been wrestling with so-called “structural infringement”—infringement that occurs as a core component of a provider’s business model and is aided by online anonymity.[120] According to a recent report prepared for the Commission, rightsholders in the EU find that such structural infringement is compounded by the paucity of requirements that intermediaries be identifiable to hosting providers.[121] There is little recourse under current EU law to identify anonymous parties, leading to calls to impose “know your customer” requirements on intermediaries.[122]

In the United States, Section 512 theoretically grants rightsholders some ability to unmask anonymous infringers through subpoenas.[123] Although it is not explicit on this point, Section 512(h) has been narrowly interpreted by courts to apply only to hosting providers, and not to Internet service providers (ISPs).[124] Technically, there are other means to potentially identify infringers, including relying on FRCP (26)(d)(1) motions to seek the identity of infringers.[125] This process, however, requires additional pleading and litigation expense as compared to Section 512(h), which allows rightsholders to apply for an unmasking order as of right once a takedown request has been filed. Limiting subpoenas solely to storage providers often ignores “the most relevant OSPs for uncovering the identity of individuals using BitTorrent and similar file-sharing protocols.”[126]

Another area where Section 512 has failed to evolve is in the use of injunctions. On its face, Section 512(j) appears to provide a broad remedy to rightsholders. It allows courts to grant injunctions: 1) to disable access to infringing content, 2) to limit service to the subscriber who is infringing, or 3) to provide other relief the court deems necessary to limit infringement of copyrighted content at a specific online location.[127] The first two forms of relief, however, only address access to a specific unauthorized copy of the copyrighted content or continued service access by a specific subscriber, rather than preventing unauthorized access more broadly to the copyrighted work. This perpetuates the Whac-A-Mole problem.

The third form of injunctive order might prove more useful, but it is rarely issued. Before granting any injunction under Section 512(j), a court must perform a balancing test to determine whether the burden placed on the OSP outweighs the harm to rightsholders.[128] The Copyright Office has observed that courts have generally found that the burden on OSPs from broader orders would outweigh the benefit to rightsholders.[129] Short of a major reconsideration of how injunctive remedies should work, Section 512(j)’s injunctive remedy is unlikely to be of much use to rightsholders in the foreseeable future:

The cost and expense of seeking an injunction in federal court against an OSP, particularly one that has previously demonstrated a willingness to litigate subpoenas and other matters relating to claims of online infringement, likely has some deterrent effect on rightsholders’ willingness to test the outer boundaries of section 512(j). Thus, while there may be some untapped “potential” in section 512(j) for combating online infringement, it is unlikely that changes to section 512(j) would play a significant role in restoring the balance under section 512. Nonetheless, the Office notes that, even in the absence of legislative change, courts have been overly narrow in their consideration of available injunctive relief under section 512(j).[130]

C.   Failure of voluntary, industrywide solutions

One final aspect of the legal development of Section 512 is worth addressing. As noted previously, Section 512 was enacted to create a set of tools for rightsholders and OSPs to work collaboratively to mitigate piracy, while also facilitating the growth of the commercial Internet. Ostensibly, part of this collaboration was to be the voluntary development of standard technical measures (“STMs”) that could effectively prevent infringement.[131] Yet, after more than 20 years, no STMs have been adopted.[132] The most effective preventive measures produced to date have been the filtering solutions adopted by YouTube,[133] Facebook,[134] and Audible Magic,[135] but neither filtering nor other solutions have been adopted industrywide. As the Copyright Office has observed:

While consensus-based fixes would be the ideal approach to improving the U.S. notice- and-takedown system, it has become clear that this is one instance where the perfect should not become the enemy of the good. Throughout the Study, the Office heard from participants that Congress’ intent to have multi-stakeholder consensus drive improvements to the system has not been borne out in practice. By way of example, more than twenty years after passage of the DMCA, although some individual OSPs have deployed DMCA+ systems that are primarily open to larger content owners, not a single technology has been designated a “standard technical measure” under section 512(i). While numerous potential reasons were cited for this failure— from a lack of incentives for ISPs to participate in standards to the inappropriateness of one-size-fits-all technologies—the end result is that few widely-available tools have been created and consistently implemented across the internet ecosystem. Similarly, while various voluntary initiatives have been undertaken by different market participants to address the volume of true piracy within the system, these initiatives, although initially promising, likewise have suffered from various shortcomings, from limited participation to ultimate ineffectiveness.[136]

The Copyright Office sounds a somewhat pessimistic note on this situation, seeing proposed STMs as “likely to encounter opposition from one or several groups of stakeholders.”[137] This concern is well-taken. For example, in their analysis of Section 512, Urban, et al., regard YouTube’s use of Content ID as a negative for the Internet ecosystem, fearing that it may become a standard for effective rights enforcement:

From the perspective of some other [online service providers], Google’s size, its prominence in the politics of notice and takedown, and its role in litigation, combined with its early adoption of DMCA Plus measures like content filtering on YouTube, trusted sender programs, autocomplete restrictions, and search result demotion, make it a dangerous elephant in the room. It is capable of adopting practices that could move collective perceptions of what is required for good practice, or even for safe harbor protection. When Google adopts DMCA Plus measures, these OSPs see their own practices under threat, as they fear the norm-setting potential of these moves.^[138]

The ideal solution to control widespread piracy would likely be a set of standards that evolve naturally and that work for both rightsholders and platform operators. In lieu of that, unfortunately, Section 512 likely needs to be reevaluated to discover where incentives can be better aligned.

In that vein, Sens. Patrick Leahy (D-Vt.) and Thom Tillis (R-N.C.)—the chair and ranking member, respectively, of the U.S. Senate Judiciary Committee’s Subcommittee on Intellectual Property—recently introduced S. 3880, the SMART Copyright Act. The bill would amend Section 512 to require OSPs to comply with a slightly heightened set of obligations to deter copyright piracy on their platforms.[139] Among other changes, the Leahy-Tillis bill would empower the Office of the Librarian of Congress (“LOC”) with broad latitude to recommend STMs for everything from off-the-shelf software to open-source software to general technical strategies that can be applied to a wide variety of systems. This would include the power to initiate public rulemakings in which the LOC could either propose new STMs or revise or rescind existing STMs. The STMs could be as broad or as narrow as the LOC deems appropriate, including being tailored to specific types of content and specific types of providers.

Critically, the SMART Copyright Act would not hold OSPs liable for the infringing content itself, but only for failure to make reasonable efforts to accommodate the STM (or for interference with the STM). Courts finding an OSP to have violated their obligation for good-faith compliance could award an injunction, damages, and costs.

Indeed, this approach comports with general principles of intermediary liability. The common law has deployed these principles in analogous situations, where the incentives of private actors are not aligned with the socially optimal outcome. As Doug Lichtman and Eric Posner have observed:

[R]ules that hold one party liable for wrongs committed by another are the standard legal response in situations where . . . liability will be predictably ineffective if directly applied to a class of bad actors and yet there exists a class of related parties capable of either controlling those bad actors or mitigating the damage they cause. . . . [W]hile indirect liability comes in a wide variety of flavors and forms . . . , it is the norm.^[140]

And as we have detailed in work examining nearly this exact concept in the context of Section 230:

Generally speaking, the law of negligence has evolved a number of theories of liability that apply to situations in which one party obtains a duty of care with respect to the actions of a third party. One legal obligation of every business is to take reasonable steps to curb harm from the use of its goods and services…. If the business has created a situation or environment that puts people at risk, it has an obligation to mitigate the risk it has created.[141]

Services that depend on user-generated content have been a boon to free expression, commerce, and likely much more. With that said, these services are inherently likely to surface illicit content if they are not adequately maintained. It is widely debated today what sort of changes are needed to reform Section 230 in order to prevent some of the harms that have emerged in the last quarter century.[142] Section 512 is due no less for this sort of reform, where OSPs should be obligated to take reasonable steps to ensure that their services are not vulnerable to piracy. Section 512 originally contemplated that voluntary standards would emerge to achieve this end. History has demonstrated that a more positive obligation may be necessary.

V. Potential Solutions

A range of possible reforms to Section 512 could better mitigate piracy and offer incentives for OSPs and rightsholders to engage in licensing negotiations. Properly applied safe harbors should encourage OSPs to help prevent unlawful dissemination of copyrighted content—as the obligation to act in the face of red-flag knowledge would largely do, absent its amelioration by the courts.[143] Ideally, such rules would also encourage OSPs to license content, enabling them and their users to benefit from such content without litigation risk. Indeed, in theory, it should be significantly more efficient for OSPs to negotiate license agreements with rightsholders than for rightsholders to do so with each of the service providers’ many users.[144]

But, as noted above, the current safe-harbor regime offers few incentives for OSPs either to curb piracy or to license content at market rates; indeed, they can obtain de facto unlicensed access to the content at no cost and with effectively no risk of liability.[145] To help address these misaligned incentives, Section 512’s safe harbor should be conditioned on OSPs taking reasonable steps: 1) to prevent infringement proactively, and 2) to stop infringement either a) when they have actual knowledge, such as when notified by a rightsholder, or b) when infringement would be apparent to a reasonable person.[146]

Below, we discuss potential adjustments to the legal standards that lead to application of safe harbors, as well as the practical steps that OSPs would need to take to qualify. We also examine some relatively less dramatic changes that could nonetheless contribute to a healthier online ecosystem that deters piracy and preserves the freedom of OSPs to innovate.

A.   Clarification of the knowledge standards

As noted above, judicial interpretations of Section 512 have essentially collapsed the red-flag standard into the actual-knowledge standard, while progressively narrowing the scope of the actual-knowledge standard; the bar for legally relevant knowledge of infringing activity is now quite high.[147]

To remedy this, the statute should be revised to effectively overturn the subjective element of red-flag knowledge applied in Vimeo. OSPs that host user-generated content should be attributed more knowledge than an “ordinary” person. Thus, red-flag knowledge would be present when information exists that would objectively lead a reasonable person in the business of facilitating dissemination of user-generated content (i.e., running a website that hosts such content) to infer infringement is taking place, even if a rightsholder has not alerted the site to a specific instance of infringement.

OSPs that host user-generated content seek to monetize that content. The DMCA presumes a significant likelihood that much of that content includes copyrighted material. Just as Congress wanted to ensure that platforms had room to grow and operate, so too did it intend to provide opportunity for rightsholders and platforms to work together to meaningfully control piracy. A service in the business of distributing content, where there is an elevated risk of infringing content, should be expected to act according to a higher standard than that to which we would hold an uninformed lay person.

This obligation to behave reasonably should exist even if that requires some degree of investigation and remediation on the part of the OSP once they have information that would objectively lead a reasonable platform to infer infringement is taking place. The obligation should obtain even if that information did not come from the rightsholder and was not related to a specific instance of infringement.

Such a standard would still offer reasonably responsive OSPs a safe harbor, which is appropriate, given that it would be impossible for platforms to catch all instances of infringement. The standard is not and should not be one of perfect content moderation, but one of reasonable content moderation. So long as a platform takes objectively reasonable steps to prevent and remediate infringement, the fact that other infringement slips through should not result in loss of the Section 512 safe harbor.

B.   Authentication, anonymous users, and subpoena authority

The Internet has long facilitated anonymous or pseudonymous communications. Fully anonymous communication systems obviously make it more difficult for rightsholders to pursue the parties responsible for infringement. Further, they can create a sense of safety (real or imagined) for would-be infringers, who may believe they can infringe with impunity.

Thus, another beneficial reform would require OSPs that host content likely to contain infringing material to reasonably ensure that they know their users’ identities. This would both discourage users from engaging in piracy and make it harder for those users to evade enforcement (and to continue infringing) simply by changing account names once caught. It would also help rightsholders to seek redress, including in cases where all they want is to ask users who are infringing unintentionally to cease doing so. Identities could remain confidential, disclosed to third parties only when needed to resolve a case of infringement.

Such disclosure might be provided voluntarily by the service provider, subject to any applicable requirements regarding the user’s privacy. The disclosures might also be provided pursuant to subpoenas issued under Section 512(h), which provides that “[a] copyright owner or a person authorized to act on the owner’s behalf may request the clerk of any United States district court to issue a subpoena to a service provider for identification of an alleged infringer in accordance with this subsection.”[148]

Relatedly, as the Copyright Office has explained, “this provision has proven to be little-used by rightsholders, in part because of how restrictively courts have interpreted it and in part because the information gleaned from such subpoenas is often of little use.”[149] Such subpoenas can be costly to obtain and are frequently ineffective; the data are often inaccurate or useless, and the OSPs may have already deleted it.[150] Moreover, courts have held that only OSPs that host material that can be removed pursuant to section 512(c)(3)(A) may be the subject of a Section 512(h) subpoena. In practice, this means such subpoenas cannot be used to obtain information from “mere conduit” ISPs.[151]

The Copyright Office is not convinced that Congress intended to exclude ISPs from section 512(h). Only an ISP is likely to be able to determine the identity of a user behind an IP address, information essential to filing an infringement claim.[152] The Copyright Office has therefore recommended clarifying that Section 512(h) applies to conduit ISPs.[153] OSPs are supposed to help rightsholders combat infringement in exchange for safe-harbor protection. Section 512 should be amended to require all OSPs seeking safe-harbor protection to provide whatever identifying information their service collects pursuant to a Section 512(h) subpoena.

C.   Filtering, takedown, and staydown

Section 512 does not require OSPs to proactively filter infringing content.[154] It was, however, expected that private industry would collaborate to develop and implement widespread standards for proactive technological controls to deter piracy. Section 512(i), for example, requires OSPs to accommodate STMs that would deter piracy and that have been developed through a voluntary, consensus process. While there have been piecemeal developments toward that end, the imagined innovations have thus far failed to materialize industrywide.

Congress must consider ways to prevent the initial sharing of pirated works, rather than the prevailing outdated file-containment approach. The latter presumes that post-hoc notice-and-takedown will be sufficient to control the spread of infringing material, which has not proven to be the case. New filtering solutions could empower OSPs to contribute significantly toward fighting piracy.[155]

Were OSPs to adopt reasonably effective filtering technologies, infringing files that are flagged and removed could more reliably be prevented from being reposted in the future. Indeed, as far back as 2007, several media and platform companies developed a set of best practices to control the proliferation of piracy. The proposed principles for user-generated content called on “websites to implement filtering technology that can recognize copyrighted works and notify rightsholders of any matches; rightsholders may then determine how the match should be treated.”[156]

Such filtering technologies already exist. Google employs proprietary filtering systems, for example.[157] DropBox has also implemented filtering in the past.[158] Audible Magic makes its filtering technology available for use by others,[159] and even small websites have found ways to employ filtering.[160] Indeed, even more prosaic technologies that have existed for decades—like web crawlers and digital fingerprinting accessible through basic APIs—could be adopted as effective STMs.

As the Copyright Office has noted, however, private firms could do more to advance broader access to these technologies for a wider range of firms.[161] A legal requirement to filter content for copyright infringement would help to foster a market for additional filtering solutions. This, in turn, could drive down costs and help to address the concern that smaller entities lack the resources to either create or obtain filtering solutions.[162]

The Copyright Office could help facilitate this process. Congress could empower the office to work with industry on specifications for STMs and to establish guidelines that ease their implementation. This could include determining minimum levels of functionality that such filters must include, which in turn should take a given platform’s size into account.

Indeed, this idea is central to the proposed SMART Copyright Act, discussed at the end of Part IV.[163] In this regard, the bill is directionally correct legislation, with two important caveats: it all depends on the kinds of STMs the LOC recommends and on how a “violation” is determined for the purposes of awarding damages.

The law would magnify the incentive for private firms to work together with rightsholders to develop STMs that more reasonably recruit OSPs into the fight against online piracy. In this sense, the LOC would be best situated as a convener, encouraging STMs to emerge from the broad group of OSPs and rightsholders. The fact that the LOC would be able to adopt STMs with or without stakeholders’ participation should provide more incentive for collaboration among the relevant parties.

Short of a voluntary set of STMs, the LOC could nonetheless rely on the technical suggestions and concerns of the multistakeholder community to discern a minimum viable set of practices that constitute best efforts to control piracy. The least desirable outcome—and the one most susceptible to failure—would be for the LOC to examine and select specific technologies.[164]

Among the concerns that surround promulgating new STMs are that they could potentially create cybersecurity vulnerabilities or sources for privacy leaks, or that they could accidentally chill speech.[165] In light of the potential unforeseen harms that can arise from implementation of an STM, the SMART Copyright Act’s requirements should be modified. If a firm does, indeed, discover that a particular STM, in practice, leads to unacceptable security or privacy risks, or is systematically biased against lawful content, there should be a legal mechanism that would allow for good-faith compliance, while also mitigating the STM’s unforeseen flaws. Ideally, this would involve working with the LOC in an iterative process to refine relevant compliance obligations.

While adopting filtering solutions would represent a new cost for many OSPs, complying with the existing notice-and-takedown system has costs, as well.[166] Using filtering solutions to prevent the unauthorized dissemination of copyrighted material would reduce the number of takedown notices that rightsholders would need to send and that OSPs would need to process. All parties would save time, hassle, and money—very possibly reducing the overall cost of Section 512 compliance.

1.     Fears of illegitimate blocking are overstated

Some critics have raised concerns that flaws in filtering technology could lead to mistakenly blocking legitimately disseminated content.[167] Among the faults commonly attributed to filtering solutions are failures to recognize licensed content or content disseminated pursuant to the Fair Use doctrine, or instances in which a rightsholder is mistaken as to the scope of its copyright.[168] The extent to which such faults are common is difficult to assess empirically. The experience of companies that offer content filters is proprietary information, and there are any number of reasons a platform might choose not to contest a takedown request that it or its users believe to be illegitimate. But some inferences may be drawn from available data, which generally suggest that harms stemming from this concern do not outweigh the known costs of piracy.

Actual false positives—instances where a user has an unambiguous right to use a file (e.g., they are the actual creator or have a license)—are likely to be exceedingly rare. A party who is the unambiguous creator of a work understandably has strong incentives to ensure that she can use her work as she sees fit.

What remains are false positives that fall somewhere on the spectrum of fair use—that is to say, they are somewhere between probably authorized under the fair-use factors and probably not authorized. If this reasoning is correct, then looking at the known instances of objections to takedown requests offers some sense of the possible scope of any false-positives problem. Some of the testimony offered during a December 2020 hearing of the U.S. Senate Judiciary Committee is illustrative in this regard.

According to Katherine Oyama, global director of business public policy for Google, uploaders disputed less than 1% of the Content ID claims made from January through June 2020.[169] Thus, in the set of all takedowns, 99% of the uploaders did not feel sufficiently entitled to use as to lodge an objection, or lodging an objection was otherwise not worth their time due to a variety of costs.[170] The remaining 1% captures some significant portion of those who are both definitely entitled (i.e., the original creator or a licensee who was misidentified) and those legitimately entitled to use under the fair-use affirmative defense.[171]

Among that 1% of uploaders, slightly more than half of the disputes were resolved in the uploader’s favor.[172] Even if all the remaining disputed uploads were kept down in error—which is unlikely—that would represent an error rate of less than 0.5% of all content flagged by Content ID. It is likely that most, if not all, of the errors were remedied in the slightly more than 0.5% of material flagged by Content ID that was allowed to proceed onto YouTube. Further, according to Noah Becker—president and co-founder of Adrev, a digital-rights-management company that administers Content ID claims for rightsholders—70% of Content ID disputes are “false claims of fair use, false claims of having procured a license, or false claims that the content is in the public domain. Most of these illegitimate disputes contain perjurious information from the user.”[173]

Using a similar set of numbers as those offered by Katherine Oyama, economist Stan Liebowitz estimated the differential between the number of infringing files uploaded to YouTube annually and the number of files taken down as infringing but later restored.[174] By his analysis of publicly available data, in 2016, roughly 2 million takedown disputes were resolved in favor of the uploader, while 600 million files were taken down and stayed down.[175] That is to say, 0.3% of takedowns were in error, which he characterized as a conservative estimate.[176]

It is, of course, possible that the percentage of false takedowns is greater than 1%, given that we can extrapolate only from known takedowns and putbacks. But if some unaccounted-for number of false takedowns are never challenged, it would appear that those users have deemed it not worth the trouble. Given the relative simplicity of the platforms’ takedown-challenge process,[177] a reluctance to engage would suggest the uploader does not place much value in the upload (i.e., there is not much benefit to be had and, by implication, limited social value). In such cases, the social cost of the false positive (erroneous takedown) is presumably not very significant compared to the value of enforcing IP rights.

By contrast, a false negative (erroneously leaving content up), even on a small site, can create significant harms to a rightsholder. Even a single unauthorized version of copyrighted content can quickly become available to the world, due to the global nature of the Internet and the availability of search, linking sites, and Internet-protocol-enabled piracy devices. Moreover, a single unauthorized version can multiply quickly. Thus, without more evidence, there is currently no reason to assume that content-filtering solutions like Content ID result in such widespread removal of legitimate uses of copyrighted material as to justify failing to enforce rightsholders’ claims rigorously.

While an industry standard for proactive filtering would help, the incredible scale of the takedown problem highlights deeper flaws in Section 512. The law should also be extended to place affirmative obligations on OSPs to employ filtering to prevent the recurrence of known infringing material once it has been discovered—a so-called “staydown” obligation, similar to recent experiments in the EU.[178] Expecting a rightsholder to repeatedly notify an OSP each time an infringing file reappears on the service makes little sense: it places rightsholders—especially small rightsholders—at a constant disadvantage.

Indeed, the Copyright Act does not currently require a copyright holder to scour a site for every instance of infringement of a specific work and to itemize every URL. Section 512(c)(3)(A)(ii) specifies that a takedown notice must identify the “copyrighted work claimed to be infringed, or, if multiple copyrighted works at a single online site are covered by a single notification, a representative list of such works at that site.”[179] Section 512(c)(3)(A)(iii) makes clear that the copyright holder need provide only “information reasonably sufficient to permit the service provider to locate” the infringing material.[180] Even if a copyright holder fails to “substantially comply” with the notice information requirements, the OSP must take “reasonable steps” to work with the copyright holder to gather the missing information.[181] If reasonably sufficient information remains unavailable, the copyright holder’s notice would not be considered the source of actual or red-flag knowledge.[182] The OSP may, however, obtain such knowledge from other sources.

Despite the broad obligations placed on OSPs by the Copyright Act’s statutory language, courts have rendered decisions that suggest rightsholders must provide near-exhaustive information, including URLs, for each and every specific infringement before OSPs have essentially any obligation to act.[183] The legislative history makes clear that such detailed information, while certainly helpful, is not required:

Where multiple works at a single on-line site are covered by a single notification, a representative list of such works at that site is sufficient. Thus, for example, where a party is operating an unauthorized Internet jukebox from a particular site, it is not necessary that the notification list every musical composition or sound recording that has been, may have been, or could be infringed at that site. Instead, it is sufficient for the copyright owner to provide the service provider with a representative list of those compositions or recordings in order that the service provider can understand the nature and scope of the infringement being claimed.

New subsection (c)(3)(A)(iii) requires that the copyright owner or its authorized agent provide the service provider with information reasonably sufficient to permit the service provider to identify and locate the allegedly infringing material. An example of such sufficient information would be a copy or description of the allegedly infringing material and the so-called ‘‘uniform resource locator’’ (URL) (i.e., web site address) which allegedly contains the infringing material. The goal of this provision is to provide the service provider with adequate information to find and examine the allegedly infringing material expeditiously.[184]

Some OSPs impose additional technical requirements for notices, such as creating an account or using an online form, instead of emailing the designated agent registered with the Copyright Office.[185] To mitigate this problem, the Copyright Office should be authorized to create model forms deemed to provide adequate notice, as well as to specify what kind of information is necessary and sufficient to require takedown. This information could be revised in a periodic process to give the office and stakeholders opportunities to properly shape the contours of this set of requirements.

D.   Repeat-infringer policies

Section 512 already requires that OSPs have policies to terminate service to repeat infringers, and to reasonably implement those policies. Indeed, Section 512(i) requires that, to be eligible for safe harbor, an OSP must have “adopted and reasonably implemented, and inform[ed] subscribers and account holders of the service provider’s system or network of, a policy that provides for the termination in appropriate circumstances of subscribers and account holders of the service provider’s system or network who are repeat infringers.”[186] As the Copyright Office has observed:

Both the House Commerce and Senate Judiciary Committee Reports explained that “those who repeatedly or flagrantly abuse their access to the Internet through disrespect for the intellectual property rights of others should know that there is a realistic threat of losing that access.”[187]

Courts, however, have historically interpreted the repeat-infringer policy requirement rather loosely. In Ventura Content v. Motherless, Inc.,[188] a site that allowed users to upload pornographic images and videos was found not to be in violation of the repeat-infringer policy requirement, even though the operator did not have a formal, detailed policy or keep a list of the number of times a user infringed.[189] The operator testified that he instead relied on his memory and terminated some repeat infringers but not others based on his own “gut” judgment, after considering a variety of unwritten factors that were not publicly available.[190] The court nonetheless concluded that this met the obligation to have a policy and to reasonably implement it.[191]

Where violations have been found, such cases have typically involved such egregious fact patterns as to provide little generalizable guidance. In Capitol Records v. Escape Media Group, an online-music service was found in violation where it failed to keep adequate records of infringement, prevented copyright owners from collecting information necessary to issue takedown notices, and did not terminate repeat infringers.[192] In UMG Recordings v. Grande Communications Networks, an ISP was found in violation where it had “utter[ly] fail[ed] to terminate any customers at all over a six-and-a-half-year period despite receiving over a million infringement notices and tracking thousands of customers as repeat infringers.”[193] And in BMG Rights Management v. Cox Communications, an ISP was found in violation where it capped the total number of notices a copyright holder could provide in a day; only counted one copyright-holder notice per subscriber per day; only considered terminating users after 13 strikes; and, if it did terminate them, reinstated the users after a break and restarted the strike counter so that, as an employee email indicated, the company could “collect a few extra weeks of payments for their account.”[194]

The point of the DMCA safe harbor is to provide platforms greater certainty regarding litigation risk when they act responsibly and to assure copyright holders that their rights will be reasonably protected in exchange for the liability limitations the platforms receive. That bargain is not achieved unless the platforms (and their users) know that costly repeat infringement will not be tolerated. To better address this goal, the Copyright Office should be authorized to provide guidance on the minimum requirements necessary to meet the repeat-infringer policy obligation, including by creating a model repeat-infringer policy that will be presumed to comply. This would offer platforms, their users, rightsholders, and courts more clarity on what behavior will be sanctioned.

E.   No-fault injunctions

Even where U.S. courts have ruled that websites have willfully engaged in infringement, stopping the infringement can be difficult, especially when the parties and their facilities are located outside the United States. One solution would be for a court to direct a non-party to the case, such as a U.S. -based ISP, to cut off access to a website held to be infringing.

Although Section 512 does allow courts to issue injunctions, there is ambiguity as to whether it allows courts to issue injunctions that obligate OSPs not directly party to a case to remove infringing material. Section 512(j) provides for the issuance of injunctions “against a service provider that is not subject to monetary remedies under this section.”[195] The “not subject to monetary remedies under this section” language could be construed to mean that such injunctions may be obtained even against OSPs that have not been found at fault for the underlying infringement.[196] But, “[i]n more than twenty years … these provisions of the DMCA have never been deployed, presumably because of uncertainty about whether it is necessary to find fault against the service provider before an injunction could issue, unlike the clear no-fault injunctive remedies available in other countries.”[197]

Indeed, more than 40 countries—including Denmark, Finland, France, India, England, and Wales—have enacted or are under some obligation to enact no-fault-injunction provisions directing ISPs to disable access to websites that predominantly promote copyright infringement.[198]

[L]egally and factually, these remedies turn on the infringing conduct of the pirate site at issue; they do not entail any finding of fault on the part of the intermediaries. No-fault injunctive remedies have been applied against a wide range of intermediaries, after first giving the intermediary notice of the infringing conduct taking place through its platform, all without entailing any inquiry into whether the intermediary may or may not have behaved in a manner that would incur primary or secondary liability. That question simply is not relevant to this form of relief, which turns on the simple finding that “such intermediaries are best placed to bring such infringing activities to an end.”[199]

Relatedly, Google has been working with rightsholders to delist pirate sites in response to “‘no fault’ orders directed at ISPs.”[200] To date, they have delisted nearly 10,000 sites in this manner.[201] The Motion Picture Association claims that its partnership with Google to delist pirate sites results in a “1.5 times larger traffic decline,” when compared with no-fault injunctions applied strictly at the ISP level.[202]

Thus, Section 512 should be amended to similarly grant U.S. courts authority to issue no-fault injunctions that require OSPs to block access to sites that courts have ruled are willfully engaged in mass infringement. Comparable authority was included in the Stop Online Piracy Act that was defeated in 2012, amid hyperbolic claims that allowing such orders would “break the Internet.” Notably, however, such sky-is-falling predictions have not materialized in the other nations that have authorized no-fault injunctions.[203] In fact, there is evidence that such orders can be quite useful in curbing piracy,[204] with one study demonstrating that no fault injunctions led to a more than 90% decrease in piracy.[205]

F.   Preservation of rights-management information

Digital copyrighted files often have embedded rights-management information, which indicates who holds the copyright and how the content may be used. OSPs and others who use the work sometimes strip out such information,[206] which is unlawful under Section 1202 of the Copyright Act. But to obtain redress, a rightsholder must demonstrate both that the OSP or other entity: 1) intentionally removed the rights-management information or disseminated the copyrighted work with knowledge that the rights-management information had been removed, and 2) did so with knowledge that its actions would facilitate infringement.[207] In practice, this section has been difficult to enforce, because the second requirement has proven a very high bar.

For example, in Stevens v. CoreLogic, Inc.,[208] CoreLogic’s software removed rights-management information from the photographs of two professional real-estate photographers when it compressed them for uploading to the Multiple Listing Service database.[209] The photographers, alleging that this might have led to unauthorized use of their photographs, sued under Section 1202.[210] The court upheld summary judgment for CoreLogic, on grounds that there was no evidence CoreLogic knew or had reason to know its actions would “induce, enable, facilitate, or conceal” infringement.[211] Similarly, in Philpot v. AlterNet Media Inc.,[212] photographer Larry Philpot alleged that AlterNet posted on its Facebook page a copyrighted photograph he took of Willie Nelson, and that it did not include the associated rights-management information.[213] The court granted AlterNet’s motion to dismiss on grounds that Philpot failed to plead facts showing that AlterNet knew or had reason to know that removal of the rights-management information would induce, enable, facilitate, or conceal an infringement.[214]

The lack of accurate rights-management information makes it harder for copyright holders to enforce their rights, as well as for individuals willing to license content to determine whom to approach to do so. Consequently, anyone disseminating the copyrighted work, including OSPs that may monetize the content through advertising or other means, should have an obligation to ensure that rights-management information included by a copyright holder remains intact and accurate. The concern here is not that the entity may be infringing the copyright or that there has been some intent to cause harm (although both may be true in some cases), but that the entity’s carelessness increases the likelihood that someone else will use the work without the copyright holder’s authorization. Consequently, Congress should consider amending Section 1202 to make it unlawful to negligently, recklessly, or knowingly remove rights-management information, or to negligently, recklessly, or knowingly disseminate a copyrighted work without that rights-management information, regardless of whether there was an intent to facilitate infringement.

Conclusion

Revising the Copyright Act as described above would encourage OSPs both to prevent initial infringement and to more effectively curtail ongoing or repeat infringement. OSPs could decline to implement these content-protection requirements, but the consequence would be losing the safe harbors and becoming subject to the ordinary standards of copyright liability. OSPs also might more widely choose to license copyrighted works that are likely to appear on their platforms. That would benefit copyright holders and Internet consumers alike. The providers themselves might even find it leads to increased use of their service—as well as increased profits.

Ultimately, however, it is important to advance copyright reforms that take seriously both the constraints on OSPs as well as the real harms that the failures of Section 512 have caused rightsholders for more than two decades. Real reform can be accomplished in a way that preserves both the benefits of a free and open Internet, as well as healthy legal protection for intellectual-property rights.

At the same time, it is important to be cognizant that effective reforms must move through a political process that can be challenging. What is set forth in this paper is a vision of what comprehensive reform would look like. But, short of a full reform, there are select measures proposed above that we believe could, even standing on their own, provide significant benefit. Foremost among these would be the expanded use of no-fault injunctions in the United States. As we note above in Section V(e), no-fault injunctions have been successfully employed around the world in a way that protects both the interests of copyright holders, as well as the interests of OSPs and private citizens. Moreover, we believe that a reform that facilitates a no-fault injunctive regime would go very far in controlling the most egregious forms of organized piracy.

[1] Geoffrey A. Manne, Kristian Stout, & Ben Sperry, Who Moderates the Moderators?: A Law & Economics Approach to Holding Online Platforms Accountable Without Destroying the Internet, at 38-39, International Center for Law & Economics, ICLE (2021), available at https://laweconcenter.org/resource/who-moderates-the-moderators-a-law-economics-approach-to-holding-online-platforms-accountable-without-destroying-the-internet.

[2] Id. at 27 (“The relevant questions [when considering intermediary liability rules] are: To what degree would shifting the legal rules governing platform liability increase litigation costs, increase moderation costs, constrain the provision of products and services, increase ‘collateral censorship,’ and impede startup formation and competition, all relative to the status quo, not to some imaginary ideal state? Assessing the marginal changes in all these aspects entails, first, determining how they are affected by the current regime. It then requires identifying both the direction and magnitude of change that would result from reform. Next, it requires evaluating the corresponding benefits that legal change would bring in increasing accountability for tortious or criminal conduct online. And finally, it necessitates hazarding a best guess of the net effect.”).

[3] Id. at 139-95 and accompanying text.

[4] 47 U.S.C. § 230(c)(1)-(2).

[5] Congress added Section 512 to the Copyright Act through amendments adopted in Section 202 of the Digital Millennium Copyright Act of 1998. See Pub. L. No. 105-304, sec. 202, 112 Stat. 2860, 2877.

[6] Digital Millennium Copyright Act, H.R. Rep. No. 105-796, at 72 (1998) (Conf. Rep.).

[7] Playboy Enterprises, Inc. v. Frena, 839 F. Supp. 1552, 1554 (M.D. Fla. 1993). Frena is discussed, infra, at notes 37-41 and accompanying text.

[8] Religious Tech. Ctr. v. Netcom On-Line Commc’n Servs., Inc., 907 F. Supp. 1361, 1365–66 (N.D. Cal. 1995). Netcom is discussed, infra, at notes 42- 47, and accompanying text.

[9] 17 U.S.C. § 512.

[10] See, e.g., 17 U.S.C. § 512(c)(1) (providing safe harbor on the condition that the online service provider “(A)(i) does not have actual knowledge that the material or an activity using the material on the system or network is infringing; (ii) in the absence of such actual knowledge, is not aware of facts or circumstances from which infringing activity is apparent; or (iii) upon obtaining such knowledge or awareness, acts expeditiously to remove, or disable access to, the material; (B) does not receive a financial benefit directly attributable to the infringing activity, in a case in which the service provider has the right and ability to control such activity; and (C) upon notification of claimed infringement as described in paragraph (3), responds expeditiously to remove, or disable access to, the material that is claimed to be infringing or to be the subject of infringing activity.”).

[11] See, e.g., Dirk Auer et al., Submission on the Final Report of the Australian Competition and Consumer Commission’s Digital Platform Inquiry, International Center for Law & Economics (Sep. 12, 2019), https://laweconcenter.org/resource/submission-on-the-final-report-of-the-australian-competition-and-consumer-commissions-digital-platforms-inquiry.

[12] Australian News Media to Negotiate Payment with Major Digital Platforms, Australian Competition and Consumer Commission (Jul. 31, 2020), https://www.accc.gov.au/media-release/australian-news-media-to-negotiate-payment-with-major-digital-platforms; see also Journalism Competition and Preservation Act, S. 673, 117^th Congress (2021); Dean Miller, France and Australia to Google and Facebook: Pay for News, The Seattle Times (Apr. 24, 2020), https://www.seattletimes.com/opinion/france-and-australia-to-google-and-facebook-pay-for-news, (France suing Google under the new EU copyright directive for harm to news producers); Inti Landauro & Emma Pinedo, Alphabet to Reopen Google News in Spain After Govt Amends Rules, Reuters (Nov. 3, 2021), https://www.reuters.com/technology/alphabet-reopen-google-news-spain-soon-after-govt-changed-regulation-2021-11-03, (Google required to negotiate with Spanish news producers under EU copyright updates).

[13] Copyright Reform Clears Final Hurdle: Commission Welcomes Approval of Modernised Rules Fit for Digital Age, European Commission (Apr. 15, 2019), https://ec.europa.eu/digital-single-market/en/news/copyright-reform-clears-final-hurdle-commission-welcomes-approval-modernised-rules-fit-digital.

[14] U.S. Copyright Office, Section 512 of Title 17: A Report of the Register of Copyrights 1 (May 2020), available at https://www.copyright.gov/policy/section512/section-512-full-report.pdf [hereinafter “Section 512 Report”].

[15] Data Bank: World Development Indicators, The World Bank,  http://databank.worldbank.org/data/reports.aspx?source=world-development-indicators (last visited Oct. 11, 2022).

[16] See, e.g., The Internet News Audience Goes Ordinary, Pew Research Center for The People & The Press (January 1999), http://www.people-press.org/1999/01/14/the-internet-news-audience-goes-ordinary.

[17] Amazon Opens for Business, History.com: This Day in History (Jul. 27, 2019), https://www.history.com/this-day-in-history/amazon-opens-for-business.

[18] Hope Hamashige, MP3.com Founder Michael Robertson Discusses His Revolutionary Company, CNNMoney (Feb. 20, 2000), https://money.cnn.com/2000/02/28/electronic/q_mp3.

[19] From the Garage to the Googleplex, Google (last visited Oct. 11, 2022), https://about.google/our-story.

[20] Tom Lamont, Napster: The Day the Music Was Set Free, The Guardian (Feb. 23, 2013), https://www.theguardian.com/music/2013/feb/24/napster-music-free-file-sharing.

[21] Anne Sraders, History of Facebook: Facts and What’s Happening, TheStreet (Feb. 18, 2020), https://www.thestreet.com/technology/history-of-facebook-14740346.

[22] Paige Leskin, Youtube Is 15 Years Old, BusinessInsider (Oct. 11, 2022), https://www.businessinsider.com/history-of-youtube-in-photos-2015-10.

[23] Jack Meyer, History of Twitter, TheStreet (Jan. 2, 2020), https://www.thestreet.com/technology/history-of-twitter-facts-what-s-happening-in-2019-14995056.

[24] Monica Anderson, Mobile Technology and Home Broadband 2019, Pew Research Ctr. (2019), https://www.pewresearch.org/internet/2019/06/13/mobile-technology-andhome-broadband-2019, (reporting that 81% of American adults owned a smartphone in 2016, up from 35% in 2011); Andrew Perrin & Madhu Kumar, About Three-in-Ten U.S. Adults Say They Are ‘Almost Constantly’ Online, Pew Research Ctr.: FactTank (Jul. 25, 2019), https://www.pewresearch.org/fact-tank/2019/07/25/americans-going-onlinealmost-constantly.

[25] See ITU Telecomm. Dev. Bureau, Measuring Digital Development Facts and Figures 2019, 1 (2019), available at https://www.itu.int/en/ITU-D/Statistics/Documents/facts/FactsFigures2019.pdf; Internet Usage Statistics: World Internet Users and 2020 Population Stats, Internet World Stats (last visited Oct. 11, 2022), http://www.internetworldstats.com/stats.htm; Internet Growth Statistics, Internet World Stats (last visited Oct. 11, 2022), https://www.internetworldstats.com/emarketing.htm.

[26] See Geoffrey A. Manne & Julian Morris, Dangerous Exceptions: The Detrimental Effects of Including ‘Fair Use’ Copyright Exceptions in Free trade Agreements, International Center for Law & Economics, ICLE White Paper 2015-1 (2015) (“Technologies such as DVRs and MP3 players that predominantly enable users to shift the time, location and/or format of consumption may increase the value of the creative work to the consumer and hence the creator. But there is also a risk that they will be used for illegal distribution of works, reducing income to the creator—since it is almost impossible for the creator to appropriate that value. To the extent that the value added by the technologies may be appropriated by creators, it is in the interests of the creator to permit their use—and to develop technologies that minimize infringing uses. But if the infringing use dominates—as was clearly the case with Napster, for example—then it is in the creators and society’s interest for the use to be prohibited.”); See also Benjamin Klein et al., The Economics of Copyright “Fair Use” in a Networked World, 92 Am. Econ. Ass’n Papers & Proc. 205, 208 (2002).

^[27] See H.R. Rep. No. 105-796, at 72 (1998) (Conf. Rep.) (noting that the purpose of Section 512 was to “preserve[] strong incentives for service providers and copyright owners to cooperate to detect and deal with copyright infringements that take place in the digital networked environment” while “provid[ing] greater certainty to service providers concerning their legal exposure for infringements that may occur in the course of their activities.”); See also Irina Y. Dmitrieva, I Know It When I See It: Should Internet Providers Recognize Copyright Violation When They See it?, 16 Santa Clara Computer & High Tech. L.J. 233, 235-39 (2000) (discussing debates that emerged in academic circles over the problems faced by both rightsholders and online service providers under the pre-DMCA case law).

[28] 17 U.S.C. § 512.

[29] Id. § 512(a).

[30] Id. § 512(b).

[31] Id. § 512(c).

[32] Id. § 512(d).

[33] Intermediaries that qualify on the basis of caching, hosting, or linking must comply with the notice-and-takedown procedures prescribed in Section 512(c)(3). See also Io Grp., Inc. v. Veoh Networks, Inc., 586 F. Supp. 2d 1132, 1148 (N.D. Cal. 2008).

[34] See, e.g., 17 U.S.C. § 512(c)(1)(A) (providing safe harbor on the condition that the online service provider “(i) does not have actual knowledge that the material or an activity using the material on the system or network is infringing [or] (ii) in the absence of such actual knowledge, is not aware of facts or circumstances from which infringing activity is apparent”). Courts have progressively made it more difficult for rightsholders to assert claims against OSPs on the basis that they had such knowledge. For example, in Capitol Records v. Vimeo, the 2^nd U.S. Circuit Court of Appeals construed fair use as a barrier to employees of Vimeo being able to detect potentially illegitimate uses of copyrighted material, even where that employee knew the work was copyrighted. Capitol Records, LLC v. Vimeo, LLC, 826 F.3d 78, 96–97 (2d Cir. 2016), cert. denied, 137 S. Ct. 1374 (2017) (“Even assuming awareness that a user posting contains copyrighted music, the service provider’s employee cannot be expected to know how to distinguish, for example, between infringements and parodies that may qualify as fair use.”). This is particularly striking, because fair use is ostensibly an affirmative defense. See Kristian Stout, A Takedown of Common Sense: The Ninth Circuit Overturns the Supreme Court in a Transparent Effort to Gut the DMCA, Truth on the Market (Sep. 23, 2015), https://truthonthemarket.com/2015/09/23/a-takedown-of-common-sense-the-9th-circuit-overturns-the-supreme-court-in-a-transparent-effort-to-gut-the-dmca.

[35] Ernie Smith, How File Sharing Broke the Internet’s First Forum, Motherboard (Feb. 6, 2018), https://motherboard.vice.com/en_us/article/a34z85/how-file-sharing-broke-the-internets-first-forum-usenet.

^[36] As noted above, Google did not exist until 1998 and the then-available search engines were far inferior. See Caleb Donaldson, Beyond the DMCA: How Google Leverages Notice and Takedown at Scale, 10 Landslide 2 (2017), available at https://www.americanbar.org/content/dam/aba/publications/landslide/2017-nov-dec/beyond-dmca.authcheckdam.pdf.

^[37] See Howard A. Shelanski, The Speed Gap: Broadband Infrastructure and Electronic Commerce, 14 (2) Berkley Tech L. J. 721 (1999), available at https://lawcat.berkeley.edu/record/1116740/files/fulltext.pdf.

^[38] See Tom Lamont, Napster: The Day the Music Was Set Free, The Guardian (Feb. 23, 2013), https://www.theguardian.com/music/2013/feb/24/napster-music-free-file-sharing.

^[39] The Digital Millennium Copyright Act of 1998, S. Rep. No. 105-190, at 8 (1998).

^[40] Id. at 2-3.

[41] Playboy Enterprises, Inc. v. Frena, 839 F. Supp. at 1554.

[42] Id. at 1554.

[43] Id.

[44] Id. at 1559.

[45] Id. at 1555-58. It is noteworthy that the court and litigants did not appear to bring up even the possibility of treating Frena as a secondarily liable party.

[46] Religious Tech. Ctr. v. Netcom On-Line Commc’n Servs., Inc., 907 F. Supp. at 1365–66.

[47] Id. at 1365-66.

[48] Id. at 1368–70.

[49] Id. at 1373.

[50] Id. at 1376-81.

[51] Id. at 1383.

[52] Irina Y. Dmitrieva, I Know It When I See It: Should Internet Providers Recognize Copyright Violation When They See It?, 16 Santa Clara Computer & High Tech. L.J. 233, 237 (2000).

[53] The Digital Millennium Copyright Act of 1998, S. Rep. 105-190, at 19.

[54] Id. at 20.

[55] Id. See also Section 512 Report, supra note 14, at 21 (“The legislative history of section 512 thus acknowledges two key components of the balance that Congress sought to achieve: the assurance that good faith actions to address internet piracy by OSPs would qualify for safe harbors, providing ‘greater certainty’ regarding their liability, and the preservation of ‘strong incentives for service providers and copyright owners to cooperate to detect and deal with copyright infringements that take place in the digital networked environment,’ providing creators with viable remedies against online infringement.”) available at https://www.copyright.gov/policy/section512/section-512-full-report.pdf; The Digital Millennium Copyright Act at 22: What Is It, Why Was it Enacted, and Where Are We Now: Hearing Before the Subcomm. on Intellectual Property of the S. Comm. on the Judiciary, 116^th Cong. (2020) (statement of Senior Judge Edward J. Damich, at 2), available at https://www.judiciary.senate.gov/imo/media/doc/Damich%20Testimony.pdf.

[56] See, e.g., Erik Brynjolfsson, Avinash Collis, & Felix Eggers, Using Massive Online Choice Experiments to Measure Changes in Well-Being, 116 Proc. Nat’l. Acad. Sci. 7250 (April 2019) (finding that digital goods and services have “created large gains in well-being that are not reflected in conventional measures of GDP and productivity.”).

[57] See YouTube, Britannica.com (last visited Oct. 11, 2022), https://www.britannica.com/topic/YouTube.

[58] See YouTube for Press (last visited Oct. 11, 2022), https://blog.youtube/press.

[59] See Number of Daily Active Facebook Users Worldwide as of 4^th Quarter 2021, Statista (last visited Oct. 11, 2022), https://www.statista.com/statistics/346167/facebook-global-dau.

[60] See Danny Sullivan, Google Now Handles at Least 2 Trillion Searches Per Year, Search Engine Land (May 24, 2016), https://searchengineland.com/google-now-handles-2-999-trillion-searches-per-year-250247.

[61] See, e.g., TuneCore (last visited Oct. 11, 2022), https://www.tunecore.com; see also CD Baby (last visited Oct. 11, 2022), https://cdbaby.com.

[62] Erik Brynjolfsson, et al., supra note 53.

[63] Yan Chen, Grace YoungJoo Jeon & Yong-Mi Kim, A Day Without a Search Engine: An Experimental Study of Online and Offline Searches, 17 Exp. Econ. 512 (2014).

[64] Kenneth Li, YouTube Anti-Piracy Software Policy Draws Fire, Reuters (Feb. 16, 2007),  https://www.reuters.com/article/us-youtube-media/youtube-anti-piracy-software-policy-draws-fire-idUSN1321663620070217, (noting that, in 2007, one copyright holder alone was demanding the removal of 100,000 copyright-protected videos from YouTube).

[65] See, e.g., Transparency Report 2021, Reddit.com, https://www.redditinc.com/policies/transparency-report-2021-2, (Reddit reports a 104% increase in takedown requests for 2021, amounting to nearly 1 million pieces of content); see also Notice and Takedown, Facebook Transparency Report, Meta, https://transparency.fb.com/data/intellectual-property/notice-and-takedown/facebook, (In the first half of 2021, Facebook reports taking down more than 3 million pieces of content for copyright violations in response to 738,000 notices).

[66] See, e.g., Brett Danaher, et al., The Effect of Piracy Website Blocking on Consumer Behavior (2019), https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2612063, (discussing the effects of site-blocking on the well-known piracy site The Pirate Bay); see also Brett Danaher & Michael D. Smith, Gone in 60 Seconds: The Impact of the Megaupload Shutdown on Movie Sales (2013), https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2229349.

[67] T. Randolph Beard, George S. Ford & Michael L. Stern, Safe Harbors and the Evolution of Music Retailing, Phoenix Center, Phoenix Center Policy Bulletin No. 41 (2017).

[68] Id. at 3.

[69] Id. at 20.

[70] See Paige Leskin, YouTube Is 15 Years Old. Here’s a Timeline of How YouTube Was Founded, Its Rise to Video Behemoth, and Its Biggest Controversies Along Way, Business Insider (May 30, 2020), https://www.businessinsider.com/history-of-youtube-in-photos-2015-10.

[71] See, e.g., Overview of Copyright Management Tools, YouTube Help (last visited Oct. 11, 2022), https://support.google.com/youtube/answer/9245819. “Monetization” is the process, typically, of permitting YouTube to run ads alongside the content in question and to share the proceeds of that ad sale in different ways (e.g., between YouTube, the video creator, and the copyright holder).

[72] House Section 512 Hearing, 113th Cong. 54 (statement of Maria Schneider, Grammy Award Winning Composer/Conductor/Producer, Member of the Board of Governors, New York Chapter of the Recording Academy); see also Directors Guild of America (“DGA”), Comments Submitted in Response to U.S. Copyright Office’s Dec. 31, 2015, Notice of Inquiry at 8 (Apr. 1, 2016) (“DGA Initial Comments”) (“[I]ndividual creators usually do . . . not have any access to, or in many cases awareness of . . . [content-filtering technologies].  That . . . needs to be rectified.”); FMC, Additional Comments Submitted in Response to U.S. Copyright Office’s Nov. 8, 2016, Notice of Inquiry at 6 (Feb. 21, 2017).

[73] Note, however, that the full picture of enforcement online remains relatively complicated. For example, acknowledging that search results can be a major vector for individuals locating pirated content, Google has begun working with the Motion Picture Association to delist sites in voluntary compliance with no-fault injunctions. See, e.g., Ernesto Van der Sar, MPA: Google’s Delisting of Thousands of Pirate Sites Works, TorrentFreak (Mar. 22, 2022), https://torrentfreak.com/mpa-googles-delisting-of-thousands-of-pirate-sites-works-220322. No-fault injunctions are discussed further, infra, at notes 175-185 and accompanying text.

[74] See section notes 100-118, infra, and accompanying text.

[75] See section notes 119-130, infra, and accompanying text.

[76] Jennifer M. Urban, et al., Notice and Takedown in Everyday Practice at 115, Berkley Public Law Research Paper No. 2755628 (Mar. 17, 2017), https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2755628. See also, e.g., Amazon.com, Comments Submitted in Response to U.S. Copyright Office’s Dec. 31, 2015, Notice of Inquiry at 3 (Apr. 1, 2016) (“Amazon Initial Comments”) (“A key principle of both federal Internet policy and the DMCA is that online service providers should not be required to police the activities of their users or make difficult legal determinations about the nature of any particular content on the service provider’s system.  Lawful services like Amazon and other U.S. Internet companies could not have flourished without such a policy.  This principle is crucial to the growth of the Internet where today, a single service can facilitate real-time discourse among over three billion worldwide users.”); See The DMCA’s Notice-and-Takedown System Working in the 21st Century: Hearing Before the S. Comm. on the Judiciary Subcomm. on Intell. Prop., 116th Cong. (2020) (statement of Abigail A. Rives, Intell. Prop. Couns.).

[77] Urban, et al., id.

[78] Id. at 116-21.

[79] Id. at 116-18; See also Marc J. Randazza, Lenz v. Universal: A Call to Reform Section 512(f) of the DMCA and to Strengthen Fair Use, 18 JETLaw 743 (2020), available at https://scholarship.law.vanderbilt.edu/jetlaw/vol18/iss4/3; Rashmi Rangnath, ​U.S. Chamber of Commerce Uses the DMCA to Silence Critic​, ​Public Knowledge (Oct. 27, 2019), https://www.publicknowledge.org/blog/u-s-chamber-of-commerce-uses-the-dmca-to-silence-critic.

[80] See, e.g., Computer & Communications Industry Association (“CCIA”), Comments Submitted in Response to U.S. Copyright Office’s Dec. 31, 2015, Notice of Inquiry at 11 (Apr. 1, 2016) (“CCIA Initial Comments”); Microsoft Corporation, Comments Submitted in Response to U.S. Copyright Office’s Dec. 31, 2015, Notice of Inquiry at 9 (Mar. 31, 2016) (“Microsoft Initial Comments”) (“In 2012, Microsoft received notices targeting under 1.8 million links to alleged infringing works appearing in Bing’s search results.  In 2015, that number grew to over 82 million alleged links to infringing works appearing in Bing’s search results, with more than 99% of such notices sent using Microsoft’s online forms.  Processing this volume of notices without the benefit of automated tools and processes, using human review, would not be viable.”); Motion Picture Association of America, Inc. (“MPAA”), Comments Submitted in Response to U.S. Copyright Office’s Dec. 31, 2015, Notice of Inquiry at 18 (Apr. 1, 2016) (“MPAA Initial Comments”) (“For smaller owners, the phenomenon may well make the notice-and-takedown exercise cost prohibitive.  One independent film maker, for example, had to send 56,000 takedown notices regarding her film, and that volume of notices did not result in the film’s permanent removal.”)

[81] Section 512 Report, supra note 12, at 81.

^[82] For just one album (“1989” by Taylor Swift), UMG had to hire full-time staff to issue more than 180,000 takedown requests between October 2014 and March 2016. Nonetheless, that album alone was illegally downloaded more than 1.4 million times. See Karen Gwee, How Artists Are Struggling for Control in an Age of Safe Harbors, Consequence (Jul. 8, 2016), https://consequenceofsound.net/2016/07/how-artists-are-struggling-for-control-in-an-age-of-safe-harbors.

[83] Now More than Ever, Creative Future (Jun. 10, 2020), https://creativefuture.org/now-more-than-ever (providing data from research firm MUSO).

[84] Id.

[85] Pirate Subscription Services Now a Billion-Dollar U.S. Industry, Joint Digital Citizens Alliance-NAGRA Report Finds, Digital Citizens Alliance, (Aug. 6, 2020), https://www.digitalcitizensalliance.org/news/press-releases-2020/pirate-subscription-services-now-a-billion-dollar-u.s.-industry-joint-digital-citizens-alliance-nagra-report-finds.

[86] David Blackburn, Jeffrey Eisenach, & David Harrison Jr., Impacts of Digital Video Piracy on the U.S. Economy, Nera Consulting (June 2019), available at https://www.theglobalipcenter.com/wp-content/uploads/2019/06/Digital-Video-Piracy.pdf.

[87] See, e.g., The Internet Association, Comments Submitted in Response to U.S. Copyright Office’s Dec. 31, 2015 Notice of Inquiry at 15 (Apr. 1, 2016) (“Internet Association Initial Comments”) (“[T]he problems of scale are true for Internet platform creators: startups and small businesses lack the sophisticated resources of larger, more established businesses in responding to takedown requests.”)

[88] Section 512 Report, supra note 12, at 32.

[89] Copyright Content Removal Requests Report, Microsoft (last visited Oct. 11, 2022), https://www.microsoft.com/en-us/corporate-responsibility/copyright-removal-requests-report.

[90] Id.

[91] Id.

[92] Intellectual Property 2021: Jul 1-Dec 31, Automattic (last visited Oct. 11, 2022), https://transparency.automattic.com/wordpress-dot-com/intellectual-property/intellectual-property-2021-jul-1-dec-31.

[93] Id.

[94] Over Thirty Thousand DMCA Notices Reveal an Organized Attempt to Abuse Copyright Law, Lumen (Apr. 22, 2022), https://www.lumendatabase.org/blog_entries/over-thirty-thousand-dmca-notices-reveal-an-organized-attempt-to-abuse-copyright-law.

[95] Urban, et al., supra note 76, at 2.

[96] Section 512 Report, supra note 14, at 84 (“Over the decades, the shift in the balance of the benefits and obligations for copyright owners and OSPs under section 512 has resulted in an increasing burden on rightsholders to adequately monitor and enforce their rights online, while providing enhanced protections for OSPs in circumstances beyond those originally anticipated by Congress”).

^[97] Some rightsholders have claimed that the way DMCA safe harbors have been construed has led to a “culture of free” — expectations by users that content has near-zero cost. See U.S. Copyright Office, Docket No. 2015-7, Section 512 Study: Notice and Request for Public Comment (2015), available at https://downloads.regulations.gov/COLC-2015-0013-89806/attachment_1.pdf. According to these commentators, the end result of safe harbors in the presence of a “culture of free” is that the perceived value of licenses themselves goes down, resulting in a vicious circle of devaluation that affects their subsequent bargaining position with the platforms. This is, however, another way of restating the argument here. If the licensing value is diminished because existing safe harbors are improperly biased toward platform owners, then the devalued license reveals that the costs of piracy are born by rightsholders, instead of being allocated more equitably between platforms and rightsholders.

[98] See, e.g., Capitol Records, Ltd. Liab. Co. v. Vimeo, Ltd. Liab. Co., 826 F.3d 78 (2d Cir. 2016); Viacom Int’l, Inc. v. YouTube, Inc., 676 F.3d 19 (2d Cir. 2012); Mavrix Photographs, Ltd. Liab. Co. v. LiveJournal, Inc., 873 F.3d 1045 (9th Cir. 2017).

[99] See infra notes at 87-94 and accompanying text.

[100] See 17 U.S.C. § 512(c)–(d).

[101] In Grokster, the Supreme Court noted that secondary liability for “vicarious” infringement could attach when an OSP directly profits from an infringement, while also declining to exercise its right to stop or limit that infringement.  Metro-Goldwyn-Mayer Studios Inc. v. Grokster, Ltd., 545 U.S. 913, 930 (2005). While not an explicit command to proactively monitor infringement, the Supreme Court recognized some circumstances in which a service provider’s obligations are heightened with respect to deterring the presence of infringing material on their services.

[102] 17 U.S.C. § 512(m).

[103] Digital Millennium Copyright Act, H.R. Rep. No. 105-551, at 53 (1998) (Conf. Rep.).

[104] Viacom, 676 F.3d at 31.

[105] See, e.g., UMG Recordings, Inc. v. Shelter Capital Partners LLC, 718 F.3d 1006, 1021 (9th Cir. 2013) (noting that “actual knowledge” as being “specific” knowledge of “particular infringing activity”); BWP Media USA, Inc. v. Clarity Digital Grp., LLC, 820 F.3d 1175, 1182 (10th Cir. 2016) (“general knowledge of potential infringement could not count as ‘actual’ knowledge”); Sony Music Entm’t v. Cox Commc’ns, Inc., 426 F. Supp. 3d 217, 230–31 (E.D. Va. 2019) (adopting the 9th Circuit’s interpretation of actual knowledge of infringement).

[106] Viacom, 676 F.3d at 31.

[107] Shelter Capital, 718 F.3d at 1023.

[108] Perfect 10, Inc. v. CCBill LLC, 488 F.3d 1102, 1114 (9th Cir. 2007). This case may be a classic instance of bad facts making bad law, with the subject content—pornography—perhaps being viewed negatively by the reviewing court in such a way that users’ increasing the “salacious appeal” of the content was seen as expected behavior.

[109] Shelter Capital, 718 F.3d at 1023 (emphasis added).

[110] Viacom, 676 F.3d at 31.

[111] Id. The court felt that the two provisions did not collapse into each other because of the addition of the “objective” standard for red-flag knowledge, even though the set of circumstances in which the knowledge standard could apply would become virtually identical (“The red flag provision, because it incorporates an objective standard, is not swallowed up by the actual knowledge provision under our construction of the § 512(c) safe harbor. Both provisions do independent work, and both apply only to specific instances of infringement.”).

[112] Vimeo, 826 F.3d at 85-86.

[113] Id. at 86.

[114] Id. at 94.

[115] Id. at 93-94.

[116] Id. at 94.

[117] Id.

[118] Section 512 Report, supra note 14, at 123 (emphasis added).

[119] See WHOIS Database Under GDPR: Temporary Measures in Place, Eurodns (Jul. 17, 2018), https://www.eurodns.com/blog/whois-database-gdpr-compliance. The WHOIS information is now considered protected. A third party with a valid interest can still obtain this information, but the process has become more cumbersome and may require appeals through the ICANN organization.

[120] Jan Bernd Nordemann, The Functioning of the Internal Market for Digital Services: Responsibilities and Duties of Care of Providers of Digital Services, European Parliament IMCO Committee at 51 (2020), available at https://www.europarl.europa.eu/RegData/etudes/STUD/2020/648802/IPOL_STU(2020)648802_EN.pdf.

[121] Id. at 51-54.

[122] Id. at 54.

[123] 17 U.S.C. § 512 (h).

[124] See, e.g., In re Subpoena to Univ. of N.C. at Chapel Hill, 367 F. Supp. 2d 945, 955 (M.D.N.C. 2005); see also Recording Industry Association of America, Inc. v. Verizon Internet Services, Inc., 351 F.3d 1229, 1233 (D.C. Cir. 2003).

[125] See, e.g., Strike 3 Holdings, LLC v. Doe, 964 F.3d 1203, 1213 (D.C. Cir. 2020).

[126] Id.

[127] 17 U.S.C. 512(j)(1).

[128] 17 U.S.C. 512(j)(2).

[129] See, e.g., Wolk v. Kodak Imaging Network, Inc., No. 10 CIV. 4135 RWS, 2011 WL 940056, at *8 (S.D.N.Y. Mar. 17, 2011).

[130] Section 512 Report, supra note 14, at 171.

[131] Id. at 67.

[132] See e.g., Authors Guild, Inc., Comments Submitted in Response to U.S. Copyright Office’s Dec. 31, 2015, Notice of Inquiry at 27 (Apr. 1, 2016) (“As a result, there has been no impetus to conduct the sort of standards creation process to develop STMs that was contemplated by Congress . . . .”); Comput. & Commc’ns Indus. Ass’n (“CCIA”), Comments Submitted in Response to U.S. Copyright Office’s Dec. 31, 2015, Notice of Inquiry at 24 (Mar. 31, 2016) (“CCIA Initial Comments”) (“CCIA is unaware of any successful or emerging inter-industry technological effort that satisfies the requirements of Section 512(i)(2).”); see also SMART Copyright Act of 2022, S. 3880, 117th Congress (2022) [hereinafter “SMART Copyright Act”].

[133] How ContentID Works, YouTube Help (last visited Oct. 11, 2022), https://support.google.com/youtube/answer/2797370?hl=en.

[134] About Rights Manager, Meta for Business (last visited Oct. 11, 2022), https://www.facebook.com/business/help/2015218438745640?id=237023724106807.

[135] Technology, Audible Magic (last visited Oct. 11, 2022), https://www.audiblemagic.com/technology.

[136] Section 512 Report, supra note 14, at 67-68.

[137] Id. at 68

^[138] Urban, et al. supra note 68, at 71.

[139] SMART Copyright Act, supra note 108.

[140] Doug Lichtman & Eric Posner, Holding Internet Service Providers Accountable, 14 Sup. Ct. Econ. Rev. 221, 223 (2006).

[141] Geoffrey A. Manne, Kristian Stout, & Ben Sperry, Who Moderates the Moderators?: A Law & Economics Approach to Holding Online Platforms Accountable Without Destroying the Internet at 38-39, International Center for Law & Economics, ICLE (2021), available at https://laweconcenter.org/resource/who-moderates-the-moderators-a-law-economics-approach-to-holding-online-platforms-accountable-without-destroying-the-internet.

[142] See id. at notes 25-100 and accompanying text.

[143] See T. Randolph Beard, et al., Fixing Safe Harbor: An Economic Analysis, Phoenix Center Policy Paper No. 52 (2017) at 23, available at https://www.phoenix-center.org/pcpp/PCPP52Final.pdf (stating that “the vetting of upload material prior to its availability for consumption on the UUC platform should be encouraged” and that “the protection of the safe harbors could be limited to UUC platforms with formal vetting policies and systems.”).

[144] See, e.g., Reiko Aoki & Aaron Schiff, Intellectual Property Clearinghouses: The Effects of Reduced Transaction Costs in Licensing, 22 Info. Econ. & Pol’y 218 (2010) (noting that third-party clearing houses are “two-sided platforms” that can improve intellectual-property licensing by centralizing information, reducing search friction, solving coordination and externality problems, simplifying contracting, and generally creating economic value by bringing upstream IP owners and downstream IP users together more efficiently); Bruce I. Carlin, Intermediaries and Trade Efficiency, at 6-7 (2005), https://ssrn.com/abstract=779485 (stating that intermediaries add value by allowing suppliers and consumers to trade objects of all quality levels, by alleviating the cost of obtaining a counterparty and decreasing search costs, and by decreasing the transaction cost); John E. Dubiansky, The Licensing Function of Patent Intermediaries, 15 Duke Law & Tech. Rev. 269, 269-70 (2017) (arguing that licensing to intermediaries can provide advantages over unilateral licensing because intermediaries can overcome search and valuation costs, avoid litigation costs, drive licensee demand by reducing uncertainty, and create network effects by increasing the number of prospective licensees accessed through the intermediary).

[145] See George R. Barker, The Value Gap in Music Markets in Canada and the Role of Copyright Law, at 8 (2018), https://ssrn.com/abstract=3320026 (stating that poor copyright law creates a “value gap” by enabling OSPs to commercially exploit copyrighted works at less than market-based rates, if they pay copyright holders anything at all); Daniel Lawrence, Addressing the Value Gap in the Age of Digital Music Streaming, 52 Vand. J. Transnat’l L. 511, 518-522 (2019), available at https://cdn.vanderbilt.edu/vu-wp0/wp-content/uploads/sites/78/2019/05/25124350/9.20Lawrence.pdf (explaining how flaws in the U.S. Copyright Act have caused a value gap); T. Randolph Beard et al., Safe Harbors and the Evolution of Music Retailing, Phoenix Center Policy Bulletin No. 41, at 20 (2017), available at https://www.phoenix-center.org/PolicyBulletin/PCPB41Final.pdf (estimating that the flawed safe harbors in the United States create a value gap of between $650 million and more than $1 billion per year for music, alone).

[146] See Beard et al., supra note 116 at 4-5, 8-10, 20, 21-22, 25-26 (2017) (stating that the United States’ flawed copyright safe harbors promote infringing platforms to the detriment of responsible ones, and recommending that the safe harbors be conditioned on platforms doing more to prevent piracy in the first place).

[147] See, e.g., Capitol Records, LLC v. Vimeo, LLC, 826 F.3d at 78; Viacom v. YouTube; Ventura Content, Ltd. v. Motherless, Inc., 885 F.3d 597, 610–11 (9th Cir. 2018); BWP Media USA, Inc. v. Clarity Digital Grp., LLC, 820 F.3d 1175, 1182 (10th Cir. 2016).

[148] 17 U.S.C. § 512(h)(1).

[149] Section 512 Report, supra note 14, at 6.

[150] Id. at 164.

[151] Verizon, 351 F.3d at 1234–36.

[152] See discussion, supra, at note 125 and accompanying text.

[153] Section 512 Report at 6, 166-67.

[154] But note that the Supreme Court has held that, in some cases, a lack of filtering is a relevant element of an infringement analysis. See Metro-Goldwyn-Mayer Studios Inc. v. Grokster, Ltd., 545 U.S. 913, 913 (2005) (noting that evidence suggested that Grokster and StreamCast were not “merely passive recipients of information about infringing use” but engaged in inducement, and observing that there was “no evidence that either company made an effort to filter copyrighted material from users’ downloads or otherwise impede the sharing of copyrighted files.”).

[155] See Beard et al., supra note 116, at 10 (observing that platforms are now capable of relatively effective filtering and suggesting that such filtering should be a “predicate for safe harbor,” with notice and takedown used as a “backstop.”).

[156] Principles For User Generated Content Services, UGCPrinciples.com (last visited Oct. 11, 2022), available at https://ugcprinciples.com.

[157] See supra notes 109 & 110.

[158] Greg Kumparak, How Dropbox Knows When You’re Sharing Copyrighted Stuff (Without Actually Looking At Your Stuff), TechCrunch (Mar. 30, 2014), https://techcrunch.com/2014/03/30/how-dropbox-knows-when-youre-sharing-copyrighted-stuff-without-actually-looking-at-your-stuff.

[159] Audible Magic (last visited Oct. 11, 2022), https://www.audiblemagic.com.

[160] See, e.g., Ventura Content v. Motherless, Inc., 885 F.3d 597, 616 (9^th Cir. 2018) (indicating a site in the case employed hashing software to police the presence of illicit clips on its service).

[161] Section 512 Report, supra note 14, at n. 501.

[162] Urban, et al. supra note 68, at 124.

[163] See SMART Copyright Act, supra note 108 and accompanying text.

[164] All else equal, firms are more intimately familiar with how their own technology works and how their users interact with that technology than are regulators. As such, regulators are usually best positioned to propose general standards and leave technical implementation details up to actual market participants (so long as such implementation reasonably comports with the requirements of the standard).

[165] Re:Create Statement on Dangerous Technical Mandate and Filtering Bill, S. 3880, re:create (Mar. 18, 2022), https://www.recreatecoalition.org/press_release/recreate-statement-on-dangerous-technical-mandate-and-filtering-bill-s-3880.

[166] See Section 512 Report, supra note 14, at n. 237 (quoting comments of the Business Software Alliance that “BSA members invest significant resources into developing state of the art systems for processing high volumes of takedown notices.”); In re Section 512 Study: Notice and Request for Public Comment, Copyright Office Docket No. 2015-7, Comments of the Information Technology and Innovation Foundation, at 3, (Mar. 21, 2016), available at http://www2.itif.org/2016-section-512-comments.pdf (stating that “[t]he best way to minimize the cost of sending and responding to so many notices of infringement is to use automated techniques. In particular, online service providers can use automated filtering systems that check content as it is uploaded to stop a user from reposting infringing content.”).

[167] See, e.g., Ben Depoorter & Robert Kirk Walker, Copyright False Positives, 89 Notre Dame Law Rev. 319, 322 (2013)  (“Second, and even more problematic, are instances where transaction  costs and risk aversion inhibit wrongly accused infringers from opposing copyright infringement actions.”).

[168] Id.

[169] The Role of Private Agreements and Existing Technology in Curbing Online Piracy: Hearing Before the Subcomm. on Intell. Prop. of the S. Comm. on the Judiciary, 116^th Cong. (2020) (written testimony of Katherine Oyama, at 9), available at https://www.judiciary.senate.gov/imo/media/doc/Oyama%20Testimony.pdf.

[170] Some users undoubtedly lack knowledge of their rights. Others will find the costs in time to outweigh the benefit of pressing a fair use claim. The relevant point here is that the reasons for failing to object are likely diverse. Thus, concerns rooted in a concept that all users who fail to object to a takedown are having their rights improperly ignored is potentially quite misleading.

[171] Admittedly, this is an extrapolation. But it is hard to imagine that a true author would fail to object to a takedown of her content. Similarly, a person who incorporates part of a work into his or her own work and feels entitled under fair use (e.g., a parody or a news commentary) has a relatively strong incentive to object. Whereas a user that incidentally includes a song in the background of a video has a much more ambiguous claim and is much less likely to object.

[172] Id.

[173] Id.; The Role of Private Agreements and Existing Technology in Curbing Online Piracy: Hearing Before the Subcomm. on Intell. Prop. of the S. Comm. on the Judiciary, 116^th Cong. (2020) (written testimony of Noah Becker, at 6), available at https://www.judiciary.senate.gov/imo/media/doc/Becker%20Testimony.pdf. And to the extent the illegitimate disputes were inadvertent, the education requirement discussed above would play an ameliorating role.

[174] Stan J. Liebowitz, Economic Analysis of Safe Harbor Provisions, CISAC, at 10 (2018), https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3143811.

[175] Id. at 11.

[176] Id. (Noting that he is selecting numbers on the low side of the potential range of takedowns and putbacks). In another analysis of a study purporting to demonstrate “high” error rates from false takedown notices, George Ford demonstrates that the actual incidence is less than 0.2% of requests. George S. Ford, Notice and Takedown in Everyday Practice: A Review, 3 (2017), available at https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2963230.

[177] See, e.g., Submit a Copyright Counter Notification, YouTube (last viewed Apr. 7, 2022), https://support.google.com/youtube/answer/2807684.

[178] See DSM Directive, art. 17(4). To avoid liability for an instance of infringement on its service, an OSP must have “acted expeditiously, upon receiving a sufficiently substantiated notice from the rightholders, to disable access to, or to remove from their websites, the notified works or other subject matter, and made best efforts to prevent their future uploads.” (emphasis added)

[179] 17 U.S.C. § 512I(3)(A)(ii).

[180] 17 U.S.C. § 512(c)(3)(A)(iii).

[181] 17 U.S.C. § 512(c)(3)(B)(ii).

[182] 17 U.S.C. § 512(c)(3)(B)(i).

[183] See Perfect 10 v. CCBill, 488 F.3d 1102 (9^th Cir. 2007); Viacom Int’l v. YouTube, Inc., 940 F.Supp. 2d 110, 115 (S.D.N.Y. 2013).

[184] H.R. Report No. 105–551, at 55 (emphasis added).

[185] See Copyright and the Internet in 2020—Reactions to the Copyright Office’s Report on the Efficacy of 17 U.S.C. § 512 After Two Decades: Hearing Before the H. Comm. on the Judiciary, 116^th Cong. (Dec. 15, 2020) (statement of Terrica Carrington, at 8), available at https://copyrightalliance.org/wp-content/uploads/2020/11/Copyright-Alliance-512-DMCA-HJC-Testimony-for-September-30-FINAL.pdf (stating that “[i]ndividual creators face numerous other significant barriers to the effective use of the notice and takedown process, including the lack of uniformity and consistency from one OSP’s web form to the next, and the practice by some OSPs of imposing requirements beyond those prescribed under the law”).

[186] 17 U.S.C. § 512(i)(A).

[187] Section 512 Report, supra note 14, at 102 (citing H.R. Rep. No. 105-551, pt. 2, at 61 (1998); S. Rep. No. 105-190, at 52 (1998)).

[188] 885 F.3d 597 (9th Cir. 2018).

[189] Id. at 607-8.

[190] Id. at 616.

[191] Id. at 619.

[192] Capitol Records v. Escape Media Group, 12-CV-6646 (AJN) (S.D.N.Y. Mar. 25, 2015).

[193] UMG Recordings v. Grande Communications Networks, 384 F. Supp. 3d 743 (W.D. Tex. 2019).

[194] BMG Rights Management v. Cox Communications, No. 16-1972 (4th Cir. Feb. 1, 2018).

[195] 17 U.S.C. § 512(j).

[196] See Copyright Law in Foreign Jurisdictions—How Are Other Countries Handling Digital Piracy?: Hearing Before the Subcomm. on Intellectual Property of the S. Comm. on the Judiciary, 116^th Cong. (Mar. 10, 2020) (statement of Justin Hughes, at 11 & n.65), available at https://www.judiciary.senate.gov/imo/media/doc/Hughes%20Testimony.pdf (“§512(j)(1)(B) makes transmission ISPs eligible for injunctive orders to deny access to subscribers engaged in infringing activity ‘by terminating the accounts of the subscriber or account holder that are specified in the order’ as well as eligible for orders ‘restraining the service provider from providing access, by taking reasonable steps specified in the order to block access, to a specific, identified, online location outside the United States.’ I believe that §512(j)(3) makes it clear that these orders were intended to be “innocent” third party injunctions available without suing the ISP.”).

[197] See Copyright Law in Foreign Jurisdictions—How Are Other Countries Handling Digital Piracy?: Hearing Before the Subcomm. on Intellectual Property of the S. Comm. on the Judiciary, 116^th Cong. (Mar. 10, 2020) (testimony of Stanford K. McCoy, at 6), available at https://www.judiciary.senate.gov/imo/media/doc/McCoy%20Testimony.pdf.

[198] See Directive 2001/29 of the European Parliament and of the Council of 22 May 2001 on the Harmonisation of Certain Aspects of Copyright and Related Rights in the Information Society, art. 8(3), 2001 O.J. (L 167) (EC); Ellen Marja Wesselingh, Website Blocking: Evolution or Revolution? 10 Years of Copyright Enforcement by Private Third Parties, Rev. Internet Derecho Politica 38–39 (October 2014), https://idp.uoc.edu/articles/10.7238/idp.v0i19.2422/ galley/2482/download; Neil Turkewitz, Why the Canadian Supreme Court’s Equustek Decision Is a Good Thing for Freedom — Even on the Internet, Truth on the Market (Jul. 8, 2017), https://truthonthemarket.com/2017/07/08/why-the-canadian-supreme-courts-equustek-decision-is-a-good-thing-for-freedom-even-on-the-internet.

[199] See McCoy, supra note 175, at 3 (quoting Directive 2001/29/EC of the European Parliament and of the Council of 22 May 2001 on the harmonization of certain aspects of copyright and related rights in the information society, recital 59).

[200] Charles H. Rivkin, Working Toward a Safer, Stronger Internet, Motion Picture Association (Mar. 21, 2022), https://www.motionpictures.org/press/working-toward-a-safer-stronger-internet.

[201] Id.

[202] Id.

[203] See Brett Danaher et al., Website Blocking Revisited:  The Effect of the UK November 2014 Blocks on Consumer Behavior at 17 (Apr. 18, 2016) (unpublished article), https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2766795; see also Rettighedsalliancen, Annual Report 2017, 5 (March 2018), available at https://rettighedsalliancen.dk/wpcontent/uploads/ 2018/08/ENGB_RettighedsAlliancen2018.pdf (noting the average 75% decrease in Danish IP traffic to piracy sites in the wake of DNS blocking orders).

[204] See McCoy, supra note 198, at 4 (“Our internal data shows us that site blocking is very effective at cutting traffic to pirate domains – meaning that an order applicable to the main access providers in a given country reduces traffic to a targeted domain by 70% on average and can be as high as 80-90% in some countries. That domain-specific impact is very clear and sustained over time. It becomes even more durable if the remedy specifies the underlying site, rather than just one or a few of the many domain names the site may use at any given time (this is the case in the UK, for example).”); Beard et al., supra note 116 (observing that site blocking has helped to curb digital piracy) (citing Brett Danaher, Michael D. Smith, and Raul Telang, Website Blocking Revisited: The Effect of the UK November 2014 Blocks on Consumer Behavior, Working Paper (Apr. 18, 2016), http://ssrn.com/abstract=2766795; Site Blocking Efficacy in Portugal: September 2015 to October 2016, Incopro (May 2017), available at http://www.incoproip.com/wp- content/uploads/2017/07/Site-Blocking-and-Piracy-Landscape-in-Portugal-FINAL.pdf; Site Blocking Efficacy Study: United Kingdom, Incopro (Nov. 13, 2014), available at http://auscreenassociation.film/uploads/reports/Incopro_Site_Blocking_Efficacy_Study-UK.pdf; Site Blocking in the World, Motion Picture Association (October 2015), available at http://www.mpa-i.org/wp-content/uploads/2016/02/Site-Blocking-October-2015.pdf; Nigel Cory, How Website Blocking is Curbing Digital Piracy Without “Breaking the Internet,” Information Technology & Innovation Foundation (August 2016), available at https://www.researchgate.net/publication/333292640_How_Website_Blocking_Is_Curbing_Digital_Piracy_Without_Breaking_the_Internet; T. Randolph Beard, et al., supra note 117.

[205] Brett Danaher, Michael D. Smith & Rahul Telang, Copyright Enforcement in Digital Age: Emprical Evidence and Policy Implications, 60 Comm’cns ACM 2, 11 (2017).

[206] A frequent reason for violating this provision appears to be simple error. For example, in Stevens v. CoreLogic, No. 16-56089 (9th Cir., Aug. 6, 2018) (Order and Amended Opinion), the defendants employed standard software libraries that did not adequately read and retain embedded metadata in photos, giving rise to a claim under Section 1202.

[207] See 17 U.S.C. § 1202(b) (making it unlawful for someone to “intentionally remove or alter any copyright management information” or to distribute or perform a copyrighted work “knowing that copyright management information has been removed or altered,” if that person also knows or has reasonable grounds to know that doing so “will induce, enable, facilitate, or conceal an infringement of any right under this title.”).

[208] Stevens v. CoreLogic, No. 16-56089 (9th Cir. 2018).

[209] Id. at 7.

[210] Id. at 9.

[211] Id. at 16-17.

[212] No. 18-cv-04479-TSH (N.D. Cal., Nov. 30, 2018).

[213] Id.

[214] Id.

Abstract

Whether self-preferencing is inherently anticompetitive has emerged as perhaps the core question in competition policy for digital markets. Large online platforms who act as gatekeepers of their ecosystems and engage in dual-mode intermediation have been accused of taking advantage of these hybrid business models to grant preferential treatment to their own products and services. In Europe, courts and competition authorities have advanced new antitrust theories of harm that target such practices, as have various legislative initiatives around the world. In the aftermath of the European General Court’s decision in Google Shopping, however, it is important to weigh the risk that labeling self-preferencing as per se anticompetitive may merely allow antitrust enforcers to bypass the legal standards and evidentiary burdens typically required to prove anticompetitive behavior. This paper investigates whether and to what extent self-preferencing should be considered a new standalone offense under European competition law.

I. Introduction

In recent years, widespread concern has emerged that large digital platforms may misuse their market positions by giving preferential treatment to their own products and services. One fear is that, by engaging in self-preferencing, so-called “Big Tech” firms may be able not only to entrench their power in core markets, but also to extend it into associated markets.1F[1] Notably, by controlling ecosystems of integrated complementary products and services—which usually represent important gateways for business users to reach end users—dominant platforms may enjoy a strategic market status that allows them to exercise bottleneck power. As the argument goes, by acting as gatekeepers and regulators within their ecosystems, these platforms represent unavoidable trading partners and may pick winners and losers in the marketplace.

Moreover, digital platforms often serve a dual role, acting as both an intermediary and a trader operating on the platform. Hence, they may be tempted to influence results in their own favor (so-called “biased intermediation”). Indeed, once an intermediation platform is also active in complementors’ markets, it loses its status of neutrality and risks of discrimination against rivals may arise because of potential conflict of interests. Therefore, quoting a slogan delivered by U.S. Sen. Elizabeth Warren (D-Mass.) during the 2020 Democratic Party presidential primary campaign: “you get to be the umpire or you get to have a team in the game—but you don’t get to do both at the same time.”2F[2] European Commissioner for Competition Margrethe Vestager has used a similar sporting analogy—arguing that a platform cannot be both a player who competes against rivals in the downstream market and, at the same time, the upstream referee who determines the conditions of that competition.3F[3]

In short, self-preferencing may allow large digital platforms to adopt a leveraging strategy to pursue a twofold anticompetitive effect—that is, excluding or impeding rivals from competing with the platform (defensive leveraging) and extending their market power into associated markets (offensive leveraging). The latter scenario may take the form of envelopment, in which a platform attempts to both exclude rivals and facilitate its own entry into a target market by tying the core functionalities of its platform to the services offered in that market.4F[4]

The legislative initiatives that have been undertaken around the world posit that, to ensure a level playing field, digital gatekeepers must be prevented from engaging in various forms of self-preferencing. The European Union’s Digital Markets Act (DMA), for example, prohibits gatekeepers from: engaging in any form of self-preferencing in ranking services and products offered by the platform itself; using any non-publicly available data generated through activities by business users to compete with those users on the platform; preventing the removal of preinstalled applications; giving preferential access to hardware, operating-system, or software features to their own ancillary services; and refusing to grant “fair, reasonable, and non-discriminatory” (FRAND) access to app stores, search engines, and social-networking services.5F[5] The United Kingdom’s proposed regulatory regime for digital markets, which imagines the adoption of firm-specific codes of conduct for online platforms with “strategic market status,” includes self-preferencing as an example of exclusionary behavior that large digital platforms sometimes engage in when they exert control over an ecosystem.6F[6] The German Digitalization Act likewise includes a ban on platforms favoring their own offers when they mediate access to supply and sales markets, particularly in cases where they present their own offers in a more favorable manner, exclusively pre-install them on devices, or integrate them in any other way.7F[7]

The American Innovation and Choice Online Act (AICOA) would go even further. The bill would declare it unlawful to engage in conduct that would “unfairly preference the covered platform operator’s own products, services, or lines of business over those of another business user on the covered platform in a manner that would materially harm competition on the covered platform.”8F[8] Accordingly, for example, Google would be prevented from launching only Google Maps in response to a query for restaurants, or from placing Google services at the top of a search-results page unless it is accompanied by all possible rival services. Similarly, Amazon would be constrained from showcasing its branded products or favoring third-party products that use its fulfillment service, while Apple would be banned from supplying prominent app-search results for its own apps or even from preinstalling its own apps.9F[9]

These provisions and others like them would essentially treat digital platforms as common carriers, and therefore subject them to a neutrality regime and utilities-style regulation. In some markets, lawmakers have proposed even more stringent reforms designed to reduce digital platforms’ potential bottleneck and intermediation power, and to prevent conflicts of interest, such as requirements that intermediation and operating units be structurally separated, restrictions on lines of business, and imposed duties to deal.10F[10]

In addition to these legislative initiatives, self-preferencing has also emerged as a theory of harm before European courts and antitrust authorities. After all, much of the behavior prohibited explicitly in the DMA initially attracted attention as part of antitrust investigations. In particular, the ban against self-preferencing appears to have been informed by the European Commission’s decision in the Google Shopping case, in which Google was fined for having systematically demoted the results of competing comparison-shopping products on its search results pages, while having granted prominent placement to its own comparison-shopping service.11F[11] The fact that the decision came following a protracted seven-year investigation has been cited as evidence of the need for an ex ante prohibition of such practices, thus removing the annoying hurdles and burdens posed by standard antitrust analysis.

The European General Court recently upheld the Commission’s decision,12F[12] although it narrowed the original decision’s scope by focusing on the context in which the practice occurred. Rather than articulating a legal test for a new antitrust offense, the Court applied fact-specific criteria to examine the potential for discrimination by a search engine. This approach notably differs from defining self-preferencing as a standalone abuse, as has been supported by the European Commission and some national competition authorities (NCAs).13F[13]

The DMA, it should be noted, will not displace Europe antitrust rules;14F[14] rather, the law will be implemented alongside them. This heightens the potential for interpretative uncertainty regarding the degree to which self-preferencing will or ought to be treated, in practice, as an infringement of competition law. This paper therefore sets out to investigate whether, in the aftermath of the Google Shopping ruling, self-preferencing by digital platforms has peculiar features that justify its consideration as a new theory of harm.

Indeed, one of the primary challenges posed by treating self-preferencing as a competitive harm, from a competition-law perspective, is the lack of an obvious limiting principle.15F[15] Notably, recent European case law suggests that, rather than a standalone theory of harm, self-preferencing is a catch-all category that includes various practices already addressed by antitrust rules. The risk is that labeling self-preferencing as per se anticompetitive would merely provide antitrust authorities with the opportunity to elide the application of legal standards and evidentiary burdens traditionally required to prove anticompetitive behavior.

This paper calls for appreciation of the continuing wisdom of antitrust orthodoxy against the prevailing zeitgeist, arguing that many of the perceived limits of antitrust actually represent its virtues.16F[16] Indeed, the goal of competition law ought not be to satisfy urgent policy objectives. Rather, antitrust is about limiting principles, even where that means it is unpopular.17F[17]

The remainder of the paper is structured as follows. Section II provides an overview of the relevant traditional antitrust theories of harm and emerging case law to analyze whether and to what extent self-preferencing could be considered a new standalone offense in EU competition law. Section III investigates whether platform neutrality more generally belongs to the scope of competition law, according to its legal foundations and settled principles. Section IV concludes.

II. Self-Preferencing as a Standalone Offense

The debate over self-preferencing revolves around its novelty. Antitrust concerns are raised regarding the preferential treatment granted by a vertically integrated dominant firm to its own products and services because of the firm’s dual role as both host and competitor. This is of particular interest when such potential conflicts of interest may result in the leveraging of market power in adjacent lines of business in ways capable of producing exclusionary effects.

From this perspective, competitive risks associated with self-preferencing do not appear significantly different from those that emerge in any scenario of vertical integration. Vertical integration is, indeed, often procompetitive, specifically because it can be used to improve efficiency and reduce transaction costs. Furthermore, while there is some dispute as to whether a dominant firm is required to ensure a level playing field by treating rivals in the same way as it does its own businesses, competition law is already equipped with tools to forbid practices that pursue discriminatory leveraging strategies. The emergence of digital platforms does not, in and of itself, challenge antitrust enforcement. To investigate whether self-preferencing should be considered a new standalone offense, it is necessary to first analyze the scope of relevant antitrust prohibitions and to evaluate the peculiar features of self-preferencing, as illustrated by courts and antitrust authorities that have recently sanctioned this behavior.

A.   Traditional European Theories of Antitrust Harm

Although predatory pricing and loyalty rebates may sometimes lead a firm to favor its own downstream services, our attention will be devoted to those practices that appear closer to self-preferencing: namely, refusal to deal, tying, bundling and mixed bundling, margin squeezes, and discrimination. In particular, the last of these represents the most obviously relevant comparison, as the favorable treatment a platform grants to its own products and services entails discriminatory treatment of rivals.

Under European competition law’s non-discrimination provisions, preferential treatment may be investigated when a vertically integrated firm applies to rivals (primary line injury) or other partners (secondary line injury) more onerous conditions than it applies to its own downstream businesses.18F[18] The second-degree price discrimination is mainly addressed by Article 102(c) TFEU, which establishes the abusive character of applying dissimilar conditions to equivalent transactions with trading parties, thereby placing them at a competitive disadvantage. It has been noted that the provision may be considered a straightforward legal basis for a theory of self-preferencing, as shown by the case law that has predominantly applied the provision in settings where a vertically integrated dominant firm sought to advantage its downstream operations at the expense of rivals.19F[19]

In the aftermath of the MEO ruling and following the effects-based approach affirmed in Intel,20F[20] discrimination is not, in itself, problematic from the point of view of competition law.21F[21] As a consequence, not every disadvantage that affects some customers of a dominant firm will amount to an anticompetitive effect; competitive disadvantages cannot be presumed. Antitrust enforcers are instead required to consider all the circumstances of the relevant case, assessing whether there is a strategy to exclude from the downstream market a trading partner that is at least as efficient as its competitors.

Self-preferencing may also take the form of tying, bundling, or mixed bundling. In the first of these, a dominant player leverages its market position in the tying product, making the purchase of the latter subject to the acceptance of another (tied) product. Bundling refers to the way products are offered and priced. In the case of pure bundling, the products are only sold jointly in fixed proportions. In mixed bundling, the products are also made available separately, but the sum of prices when sold separately is higher than the bundled price.22F[22]

Any of these practices may lead to anticompetitive foreclosure in the tied market and, indirectly, in the tying market. The exclusion of as-efficient-competitors is key to triggering antitrust liability for competition foreclosure. Mixed bundling may be anticompetitive if the discount is so large that equally efficient competitors offering only some of the components cannot compete against the discounted bundle. With bundling, the greater the number of products on which the undertaking exerts market power, the stronger the likelihood of anticompetitive foreclosure. In the case of tying, if an insufficient number of customers would buy the tied product on its own to sustain competitors of the dominant undertaking in the tied market, the tying could lead to those customers facing higher prices. Finally, the risk of foreclosure in tying and bundling strategies is expected to be greater where the dominant player makes it last—e.g., through technical tying (i.e., designing a product in such a way that it only works properly with the tied product and not with alternatives offered by competitors).

As tying strategies can be implemented either through contractual terms or by technical means, antitrust authorities are increasingly prone to challenge platforms’ product-design decisions that favor their own products or services by limiting interoperability, thereby impeding compatibility with rival products or services.23F[23] In Microsoft, the European General Court argued that the ubiquity of a dominant player in the tying market is likely to foreclose competition in the tied market. The Court noted that the practice of bundling a specific piece of software to an operating system through pre-installation allows the tied product “to benefit from the ubiquity of that operating system … which cannot be counterbalanced by other methods of distributing media players.”24F[24]

Foreclosure also may arise when consumers obtain the tied product free of charge and are not prevented from obtaining rival services. In Google Android, the Commission fined Google for having engaged in a leveraging practice to preserve and strengthen its position in the search-engine market by requiring device manufacturers to preinstall Google Search and the Chrome browser as preconditions to license the Google Play app store. By locking down Android in the Google-controlled ecosystem, manufacturers wishing to pre-install Google apps were prevented from selling smart-mobile devices that run on versions of Android not approved by Google (so-called Android “forks”).25F[25] According to the Commission, pre-installation can create a status quo bias, which reduces the incentives for manufacturers to pre-install competing search and browser apps, as well as the incentives for users to download such apps. The therefore affects rivals’ ability to compete effectively with Google. Despite the fact that Android is mostly distributed as open-source software, the Commission rejected both of the justifications Google put forward: that leveraging practices reflected a legitimate appropriation strategy to preserve incentives to innovate in a regime of weak appropriability26F[26] and that fork restrictions fell under governance rules needed to protect multi-sided platforms from negative externalities (in this case, preventing software fragmentation and the potential diffusion of incompatible versions of the software).27F[27]

Taken to its extreme, self-preferencing can result in refusals to deal,28F[28] which explains why European policymakers have invoked the essential facilities doctrine to address such cases. The aim of the doctrine, which imposes on dominant firms a duty to deal with all who request access, is to prevent a firm with control over an essential asset from excluding rivals or from extending its monopoly into another stage of production. Because it requires sacrificing the dominant firm’s freedom of contract and right to property, however, it may weaken incentives to invest, innovate, and compete.

These refusal-to-deal infringements are, under European competition law, generally limited to “exceptional circumstances.” According to Magill, a refusal to deal may trigger an antitrust violation when: (i) access to the product or service is indispensable to a firm’s ability to do business in a market; (ii) the refusal is unjustified; (iii) the refusal excludes competition on a secondary market; and (iv), if intellectual property rights are involved, it prevents the emergence of a new product for which there is potential consumer demand.29F[29] The IMS30F[30] and Microsoft31F[31] judgments substantially dismantled the third and fourth requirements, respectively, by considering the secondary-market requirement met even if that market is just potential or hypothetical, and the new product requirement satisfied even when access to the facility is necessary for rivals to develop follow-on innovation (i.e., improved products with added value).

Nonetheless, pursuant to the interpretation provided in Bronner, the requirement that a requested facility be indispensable remains in place and represents the last bulwark against the dangers of uncontrolled application of the doctrine.32F[32] Indeed, access to an input is considered indispensable if there are no technical, legal, or even economic obstacles that would render it impossible (or even unreasonably difficult) to duplicate. To demonstrate the lack of realistic potential alternatives, a requesting firm must establish that it is not economically viable to create the resource on a scale comparable to that of the firm controlling the existing product or service.

Against this background, the recent Slovak Telekom judgment introduced a relevant novel claim that the conditions laid down in Bronner do not apply where the dominant undertaking does give access to its infrastructure but makes that access subject to unfair conditions.33F[33] In addition, the Court of Justice (CJEU) implied that enforcers do not have to prove indispensability when access to a facility has been granted as a result of a regulatory obligation.34F[34] The implications are particularly relevant to digital markets, as the regulatory framework established by the DMA requiring access to platforms designated as gatekeepers would exempt antitrust authorities from having to demonstrate the indispensability of those facilities.

Finally, self-preferencing may be construed as a “margin squeeze,” which EU competition law defines as a standalone abuse that undermines the condition of equality of opportunity between economic operators. The European Commission initially equated this practice to a constructive refusal to deal, noting that, instead of refusing to supply, a dominant undertaking charges a price for a product on the upstream market that would not allow even an equally efficient competitor to trade profitably in the downstream market on a lasting basis.35F[35] The Commission therefore introduced the so-called Telefonica exceptions to categorize a specific class of cases where Bronner’s requirements would not apply. These exceptions hold that an obligation to supply cannot have negative effects on the input owner’s and/or other operators’ incentives to invest and innovate upstream.36F[36]  The CJEU has, however, gradually moved toward rejecting the concept of an implicit refusal to grant access, holding that margin squeezes should be treated as a separate theory of harm, thereby introducing an even broader exception to Bronner.

Notably, while an essential facility was involved in Deutsche Telekom I, the owner of the facility had a regulatory obligation to share and rivals’ margins were negative.37F[37] Teliasonera found a margin squeeze in a situation where the input of the dominant undertaking was not indispensable, there was no regulatory obligation to supply, and rival firms’ margins were positive, but insufficient, as the rivals were forced to operate at artificially reduced levels of profitability.38F[38] Telefonica39F[39] and Slovak Telekom40F[40] upheld the approach of considering margin squeezes as an independent form of abuse to which the criteria established in Bronner are not applicable.

B.   European Case Law on Self-Preferencing

Against this background, doubts about the potential to identify self-preferencing as a standalone abuse under EU law emerge from the court analysis and antitrust decisions that have been issued to date sanctioning dominant platforms for preferential treatment granted of own products and services. Indeed, recent European case law would appear to question whether self-preferencing is sufficiently novel to constitute a standalone theory of harm, given that it has been readily addressed under existing theories of harm. With the exceptions of the Amsterdam Court of Appeal41F[41] and the Italian Competition Authority,42F[42] courts generally do not even use the term “self-preferencing,” opting instead to label the conduct “favoring.”

1.  UK Streetmap decision and European Commission Google Shopping decision

The birth of self-preferencing as a standalone theory of harm is usually associated with the European Commission’s investigation of Google for having positioned and displayed, in its general search-results pages, its own comparison-shopping service more favorably than rival comparison-shopping services.43F[43]

However, a similar issue was addressed a few months earlier by the High Court of England and Wales in the dispute between Streetmap and Google, which involved the interaction of competition between online search engines and online maps.44F[44] Indeed, Streetmap contended that Google abused its dominant position by prominently and preferentially displaying its own related online-map product. Streetmap contended that, by visually displaying a clickable image from Google Maps (and no other map) in response to certain geographic queries (Maps OneBox) at or near the very top of its search-engine results page (SERP), and consequently relegating Streetmap to a blue link lower down the page, Google abused its dominant position in the market for online search and online search advertising.

Given the evident similarity with the Google Shopping case, Justice Roth’s analysis is worthy of examination. While Streetmap framed Google’s practice in terms of bundling or tying, and referred extensively to the European Microsoft decision, the U.K. Court held that the complaint should have been appropriately characterized as an allegation of discrimination.45F[45] The user who sees the Maps OneBox is, indeed, under no obligation to click on it or to use Google Maps; she remains free to use any other online-mapping provider without penalty. In contrast to Microsoft, where obtaining a competing streaming-media player by downloading from the Internet was regarded by a significant number of users as more complicated than using the pre-installed Microsoft product, the Google SERP includes clickable links to other relevant online maps and users experience no difficulty in clicking on those blue links.

To establish whether Google’s conduct constituted anticompetitive foreclosure, the Court concluded that it was necessary to prove that the effects of that conduct appreciably affected competition, which cannot simply be assumed. Indeed, the Google’s introduction a Maps OneBox containing a thumbnail map was intended to improve the quality of the SERP, and hence must be evaluated as a technical efficiency46F[46]: “The unusual and challenging feature of this case is that conduct which was pro-competitive in the market in which the undertaking is dominant is alleged to be abusive on the grounds of an alleged anti-competitive effect in a distinct market in which it is not dominant.”47F[47] For this reason, evaluating alternative ways that Google might have made this procompetitive improvement without allegedly distorting competition in online maps played a significant role in the Court’s analysis.

If anticompetitive effects are proven, then the issue of objective justification must be considered. This requires a proportionality assessment, which is a matter of fact and degree. Hence, the question of alternatives cannot be considered only with respect to competitive impact: “Where the efficiency is a technical improvement, proportionality does not require adoption of an alternative that is much less efficient in terms of greatly increased cost, or which imposes an unreasonable burden.”48F[48]

Following this line of reasoning, Justice Roth found that the introduction of the Maps OneBox with no shortcut hyperlinks to Streetmap (and other online maps) did not, in itself, have an appreciable effect in steering customers away from Streetmap; it therefore was not reasonably likely to give rise to anticompetitive foreclosure.49F[49] Moreover, even if it was likely to have such an effect, Google’s conduct was objectively justified because the way that it implemented the technical efficiency—i.e., presenting a thumbnail map on the SERP—was not disproportionate.50F[50]

The European Commission reached a different conclusion in Google Shopping. There, the Commission found that, by promoting its own comparison-shopping service in its search results and demoting those of competitors, Google engaged in a strategy of leveraging the dominance of its flagship product (i.e., the search engine) in the adjacent market for comparison-shopping services.

According to the Commission’s findings, Google’s strategy rested on two related practices: ensuring prominent placement for its own comparison-shopping service and demoting rival comparison-shopping services in its search results. Notably, while competing comparison-shopping services could appear only as generic search results—potentially subject to demotion in search listings by Google’s algorithms—Google’s own comparison-shopping service was prominently positioned, displayed in rich format, and free from the risk of demotion to the second page of search results.51F[51]

The Commission concluded that the conduct fell outside the scope of competition on the merits, could extend Google’s dominant position in the national markets for general search services to the national markets for comparison-shopping services (offensive leveraging), would tend to protect Google’s dominance in the former (defensive leveraging).

Rather than recognizing that it was deploying a novel theory of harm, the Commission argued that Google’s conduct belonged to the well-known category of leveraging. Accordingly, there was no need to look for a new legal test, since “it is not novel to find that conduct consisting in the use of a dominant position on one market to extend that dominant position to one or more adjacent markets can constitute an abuse.”52F[52] The Commission therefore found that self-preferencing constitutes a “well-established, independent, form of abuse.”53F[53]

To support its line of reasoning, the Commission invoked disparate case law, including judgments involving either specific theories of harm (e.g., Tetra Pack,54F[54] Irish Sugar,55F[55] and Microsoft,56F[56] with regards to tying and predatory pricing, loyalty rebates, and refusal to deal, respectively) or that are outdated (e.g., Telemarketing57F[57]).58F[58] The Commission’s rationale in offering this selection of decisions is unclear. References to one case in which the essential facilities doctrine was applied (i.e., Microsoft) and to another ruling that has since been replaced by the elaboration of the essential facilities doctrine (i.e., Telemarketing) are even more surprising.

The Commission ultimately dismissed Google’s claim that its conduct could be considered abusive only if Bronner’s criteria were fulfilled: namely, if access to Google’s general search results pages were indispensable to being able to compete.59F[59] According to the Commission, the decision merely required Google to cease the conduct. Hence, the Bronner criteria were “irrelevant in a situation, such as that of the present case, where bringing to an end the infringement does not involve imposing a duty on the dominant undertaking to transfer an asset or enter into agreements with persons with whom it has not chosen to contract.”60F[60]

The case spurred debate over the legal test applied to require Google to grant equal treatment to rival comparison-shopping services and its own service. Among the questions raised by the case are whether the conduct fell more within exclusionary or discriminatory abuses and, if it was the former, whether tying or the essential facilities doctrine was the proper framework to assess such self-preferencing abuse.61F[61] For instance, the experts appointed by the Commission to provide suggestions for the design of a competition policy for digital markets considered self-preferencing a specific technique of leveraging, which is not abusive per se, but subject to an effects test.62F[62] Furthermore, quoting Microsoft, they argued that, according to well-established case law, the owner of an essential facility must not engage in self-preferencing. Nonetheless, they believed that self-preferencing by a vertically integrated dominant digital platform can be abusive, not only under the preconditions set out by the essential facilities doctrine, but also wherever it is likely to result in leveraging market power and is not justified by a pro-competitive rationale.

2.   Amsterdam Court of Appeal ruling in Funda

In May 2020, the Amsterdam Court of Appeal handed down a decision in litigation between VBO Makelaars and NVM, two associations of real-estate agents, brokers, and appraisers.63F[63] NVM owns Funda, the largest online real-estate platform in the Netherlands and which, according to VBO, granted NVM agents more prominent positions in the ranking of properties. Funda also applied higher tariffs to and granted only limited website functionality to VBO agents, who also did not have access to the underlying Funda database. VBO’s complaint charged NVM with anticompetitive discrimination.

Upholding the decision of the district court, the Court of Appeal found that Funda did not abuse its dominant position in favoring the listings of NVM members over those of rival agents. Assessing self-preferencing as discriminatory conduct under Article 102(c) TFEU, the Court cited MEO, arguing that VBO’s complaint failed to demonstrate that the discrimination distorted the company’s competitive position in ways that led to a competitive disadvantage on the downstream market for real-estate services.

In particular, the Court noted that several factors play relevant roles in consumers’ home-purchase decisions and that it is implausible that a buyer would automatically assume that the listing placed highest on a website would be the one that best meets their demands. To the contrary, the Court concluded that the market for homes differs significantly from markets for other consumer products. For example, buyers generally conduct an intensive search over a long period of time to consider all relevant offers. Therefore, the Court found, a lower website ranking would be of minor importance and would not necessarily lead to a competitive disadvantage.

Accordingly, the Court considered comparisons with Google Shopping to be unhelpful.64F[64] In assessing NVM’s preferential treatment in accordance with the principles the CJEU elaborated in MEO, however, the Dutch court did frame self-preferencing as a discriminatory abuse, thus anticipating the approach that the European General Court would ultimately endorse in Google Shopping.

3.   French Competition Authority Apple ATT and Google AdTech decisions

In June 2021, the French Competition Authority (AdlC) followed the European Commission’s lead in investigating practices implemented by Google in the online-advertising sector.65F[65]

Responding to referrals from news publishers who monetize their websites and mobile apps through advertising, the AdlC found that Google engaged in abusive practices to favor its own advertising intermediation technologies, granting preferential treatment to its proprietary technologies offered under the Google Ad Manager brand. Notably, in the Authority’s view, Google used its dominant publisher ad server (DoubleClick for Publishers, or DFP) both to favor its own programmatic advertising sales platform (AdX) and, separately, used AdX to favor DFP in the market for supply-side ad-intermediation platforms (SSPs). Regarding the first practice, the preferential treatment consisted of informing AdX of the prices offered by competing SSPs, thus allowing it to optimize the bidding process by varying the commissions received on impressions sold according to the intensity of competition. Regarding the second practice, Google imposed technical and contractual limitations on the use of the AdX platform through a third-party ad server. As a result, the modalities of interaction offered to third-party ad-server clients were inferior to the modalities of interaction between DFP and AdX, which penalized both third-party SSPs and publisher clients.

Similar concerns about the impact of Google’s conduct in ad-tech services have also been raised by the Australian Competition and Consumer Commission (ACCC). The ACCC concluded that Google’s vertical integration and dominance across the ad-tech supply chain and related services have allowed the company to engage in leveraging and self-preferencing conduct and that this, in turn, has likely interfered with the competitive process.66F[66]

According to the AdlC, the evidence showed that DFP’s favorable treatment of AdX had a foreclosure effect on competition among platforms selling ad space, significantly reducing the attractiveness of rival SSPs. In addition, according to the Authority, DFP’s preferential treatment strengthened Google’s dominant position, impairing the competitiveness of rival ad-server providers, and limiting their ability to compete on the merits. Therefore, as regards the latter, limitations on interoperability were deemed a practice that cannot be considered competition on the merits, as it would tend to impose on rivals a competitive disadvantage by applying to them less favorable technical conditions.67F[67]

By and large, the French decision did not provide insights on the theory of harm or type of abuse that this form of discrimination would constitute. Like the European Commission, the AdlC did not refer to self-preferencing explicitly, instead describing Google’s conduct as favoring. With regards to Google’s leveraging strategy, the AdlC cited Google Shopping and quoted the very same case law the Commission mentioned in that decision.68F[68] The only significant addition made by the Authority was a reference to Slovak Telekom, a margin-squeeze case that, as already mentioned, brought about a remarkable change in confining the application of Bronner’s indispensability condition to “pure” refusals to deal.69F[69]

However, the AdlC is also currently investigating Apple’s privacy policy, where self-preferencing is mentioned explicitly and appears to be framed differently.70F[70] Notably, in March 2021, the French Authority rejected the request for interim measures against Apple’s adoption of the App Tracking Transparency (ATT) framework for applications on iOS 14, which create new consent and notification requirements for app publishers. The ATT solicitation was considered part of Apple’s longstanding strategy to protect iOS users’ privacy and its implementation was not expected to impose unfair trading conditions, such as excessive or disproportionate restrictions on the activities of app developers. The AdlC nonetheless advised that, as part of its investigation into the merits of the case, it would examine how the user-consent collection processes differ between Apple’s own advertising services and third-party advertising services. Differences in those processes might result in a “form of discrimination (or self-preferencing)” if Apple applied, without justification, more binding rules on third-party operators than those it applies to itself for similar operations.71F[71]

The growing suspicion of self-preferencing has likewise prompted the German Competition Authority to initiate its own proceeding on Apple’s ATT policy,72F[72] while the U.K. Competition and Markets Authority (CMA) raised similar concerns in its market study on mobile ecosystems.73F[73]

4.   General Court ruling in Google Shopping

Given this background, the European General Court’s judgment in Google Shopping was much awaited.74F[74] For those who were looking for legal certainty from the judgment, however, those expectations have been not completely met.

What was new in the ruling was its broad interpretation of the general principle of equal treatment, which the Court affirmed obligates vertically integrated platforms to refrain from favoring their own services over rivals.75F[75] While this approach was in line with the Commission’s expansive reading of the special responsibility of dominant firms, however, the ruling framed self-preferencing as a discriminatory abuse.76F[76] Notably, the Court highlighted that the various judgments the Commission cited in its original ruling do not support the conclusion that any use of a dominant position on one market to extend that position to one or more adjacent markets constitutes a “well-established” form of abuse.77F[77] After all, “leveraging” is a generic term covering several practices that are potentially abusive, such as tied sales, margin squeezes, and loyalty rebates.78F[78]

The three rulings the Court cited involve, instead, practices found to be discriminatory abuses specifically because they place third parties at a competitive disadvantage. Two of three involve discriminatory conditions applied by public undertakings operating a commercial port79F[79] and an airport.80F[80] This may support a link with recent legislative initiatives categorizing digital platforms as common carriers and thus subject to the neutrality regime of public utilities-style regulation. Nonetheless, the Court clarified that prohibiting self-preferencing to enforce the policy goal of neutrality is appropriate only when a competitive harm is demonstrated. Indeed, rather than deeming self-preferencing to be per se abusive, the Court moved to its potential anticompetitive effects. This is in line with the effects-based approach affirmed in MEO,81F[81] as well as in other judgments that, although they involve different abusive conduct, entail similar discriminatory elements.82F[82] The Court, however, surprisingly did not even mention MEO.

The Court’s ruling focused on potential exclusionary effects associated with specific leveraging strategies, reflected in the Commission’s original finding of abuse on the basis of certain relevant criteria.83F[83] The Commission had noted that, due to network effects, the traffic that Google’s search engine generates represents a critical asset; that users are significantly influenced by favoring, as they typically concentrate on the first few search results and tend to assume that the most visible results are the most relevant; and that traffic directed from Google’s search-results pages accounts for a large portion of traffic to competing comparison-shopping services, which cannot be effectively replaced by other sources.84F[84]

The Court outlined four criteria that differentiated Google’s self-preferencing from competition on the merits, therefore warranting a finding of antitrust liability.

First, the Court highlighted the “universal vocation” and openness of a search engine as features of its core mission.85F[85] These features distinguish a search engine, which designed to index results that might contain any possible content, from other services referenced in the case law, which consist of tangible or intangible assets (press-distribution systems or intellectual property rights, respectively) whose value depends on a proprietor’s ability to retain their exclusive use.86F[86]

While not explicitly mentioned, the reference is clearly to the essential facilities doctrine case law. Unlike these services, “the rationale and value of a general search engine lie in its capacity to be open” to results from external sources and to display multiple and diverse sources on its general results pages.87F[87] Moreover, the legal obligation of equal treatment that ensues from net-neutrality regulations88F[88] for Internet access providers on the upstream market cannot be disregarded when analyzing the practices of an operator like Google on the downstream market.89F[89]

Second, because Google holds a “superdominant” (or “ultra-dominant”) position on the market for general search services and acts as a “gateway” to the Internet, it is under a stronger obligation not to allow its behavior to impair genuine, undistorted competition on the related market for specialized comparison-shopping search services.90F[90]

Third, the market for general search services is characterized by very high barriers to entry.91F[91]

Fourth, in light of prior considerations (i.e., the mission of a search engine, Google’s dominance, and the presence of very high barriers to entry), the Court found that Google’s conduct is “abnormal.”92F[92] Indeed, for a search engine to limit the scope of its results to its own services entails an element of risk and is “not necessarily rational.” This is especially the case in a situation where, because of barriers to entry and the search engine’s own dominance, it is significantly unlikely that there would be market entry within a sufficiently short period of time in response to the limitations placed on Internet users’ choices.93F[93]

In this scenario, in the Court’s view, Google’s promotion of its own specialized results over third-party results contradicts the basic economic model of a search engine and hence involves a certain form of abnormality.94F[94] The suspicion is strengthened by Google’s “change of conduct.”95F[95] While it initially provided general search services and displayed all the results of specialized search services in the same way and according to the same criteria, once the firm had entered the market for specialized comparison-shopping search services—and after having experienced the failure of its dedicated comparison-shopping website (Froogle)—Google changed its practices and comparison-shopping services were no longer all treated equally.96F[96]

These four criteria suggest that the Court saw Google’s search engine as an essential facility. The Court, indeed, noted that, by envisaging equal treatment for any comparison-shopping services on Google’s general results pages, the Commission’s decision was seeking to provide competitors with access to Google’s general results pages. This was presented as particularly important to competing comparison-shopping services and something that was not effectively replaceable, as it accounted for such a large proportion of traffic to their websites.97F[97] Moreover, the Court acknowledged that the Commission considered Google’s traffic to be indispensable to competing comparison-shopping services.98F[98] As a consequence, the analysis would have required an assessment of preferential treatment pursuant to the conditions set out in Bronne, as Google itself had requested, rather than relying on the case law applicable to abusive leveraging, as the Commission did in its decision.

But despite characterized the features of Google’s general results page as “akin to those of an essential facility,”99F[99] the Court upheld the Commission’s decision not to apply Bronner’s indispensability requirement. In doing so, it drew a line between express refusals to supply and exclusionary practices that do not lie “principally” in a refusal, as such.100F[100] Indeed, “the present case is not concerned merely with a unilateral refusal by Google to supply a service to competing undertakings that is necessary in order to compete on a neighboring market, which would be contrary to Article 102 TFEU and would therefore justify the application of the ‘essential facilities’ doctrine.”101F[101]

Therefore, the Court shared the Commission’s viewpoint that Google’s self-preferencing was a standalone form of leveraging abuse, involving positive acts of discrimination in the treatment of the search results for comparison-shopping services.102F[102]

The Court’s ruling has been generally welcomed for two reasons. By affirming self-preferencing as an independent abuse, the judgment provides legal support to the policy goal of imposing a neutrality regime over large digital platforms, which has informed all the regulatory interventions promoted in different jurisdictions. At the same time, the Court advances a clearer legal qualification of the conduct in question. Indeed, while the Commission’s approach appeared unprecedented—because it revolved around the notion of favoring as a specific form of leveraging—the Court opted for the more defined legal framework of discrimination. The outcome should help to restrain the scope of application for self-preferencing prohibitions in comparison to other traditional practices that, although belonging to the general category of leveraging and including elements of discrimination, reflect specific theories of harm and are assessed according to their respective legal tests.103F[103] By and large, the Court confirms that there is no well-established case law that would forbid any extension of a dominant position in adjacent markets, in contrast with the Commission’s stance.

Nonetheless, the ruling raises new doubts. Notably, the definition of the conduct that would be covered remains unclear. While adopting the general principle of equal treatment as a legal basis to prohibit self-preferencing may allow intrusions into platforms’ design choices,104F[104] the listed criteria appear to define a narrow framework, ultimately calling into question the broad application of self-preferencing as a standalone abuse.

The Court underscored the relevance of the “particular context” in which favoring occurred.105F[105] Namely, the emphasis was on the role played by search engines on the Internet, including their “universal vocation” and “open” business models. This is strengthened by analogies to net neutrality, the characteristics “akin to those of an essential facility,” and the “superdominant position” that made Google a “gateway” to the Internet. Furthermore, the peculiar features of search engines (notably, their openness) are also deemed relevant in assessing the “abnormality” of Google’s behavior—which, in the Court’s evaluation, is indeed at odds with the basic economic model of its search engine. The legal framework is completed with the detection of opportunistic behavior by Google, which changed its strategy once it entered the adjacent market of comparison-shopping services.

It is left far from clear whether Google Shopping is even sanctioning the favoring practice as such. Indeed, the Court describes the anticompetitive strategy in question as formed by a combination of two practices—namely the promotion of Google’s own services and the demotion of its rivals’ services. Therefore, the conduct is not necessarily abusive even if it consists “solely in the special display and positioning” of the platform’s own products and services.106F[106] The practice has instead been judged illegal because it included the relegation of competing services in Google’s general results pages by means of adjustment algorithms. That, in conjunction with Google’s promotion of its own results, there was a simultaneous demotion of results from competing comparison services is considered a “constituent element” of the conduct and moreover plays a “major role” in the exclusionary effect identified.107F[107]

In summary, rather than articulating a legal test for a new antitrust offense, the criteria pointed out in the judgment for considering the preferential treatment abusive appear extremely fact-sensitive: both Google-specific and search engine-specific. Therefore, it is difficult to see how, according to these criteria, a self-preferencing prohibition may be applied to different forms of preferential treatment, digital services, and business models.108F[108]

5.   Italian Competition Authority Amazon Logistics decision

A few weeks after the General Court’s ruling, the Italian Competition Authority (AGCM) handed down a decision that significantly departed from the legal framework elaborated in Google Shopping, thus confirming that the precise contours of self-preferencing abuses under Article 102 TFEU remained anything but clear.109F[109]

In late November 2021, the AGCM issued a mammoth fine against Amazon for granting preferential treatment to third-party sellers who use the company’s own logistics and delivery services (i.e., Fulfilment by Amazon, or FBA). Amazon was accused of having leveraged its dominance in the market for intermediation services on marketplaces to favor the adoption of its own FBA by sellers active on Amazon.it, as well as to strengthen its own dominant position. Under AGCM’s view, this strategy ultimately harmed both competing logistics operators, by putting them at a competitive disadvantage, and competing marketplaces, by creating incentives for sellers to single-home.

Indeed, although third-party sellers are free to manage the logistics associated with their operations on the platform themselves or outsource them to an independent operator (Merchant Fulfilment Network, or MFN), Amazon was deemed to be artificially pushing them to use its own logistics service, thus deterring them from multi-homing.110F[110] Notably, the Authority found that Amazon “tied” the use of FBA to access to a set of exclusive benefits essential for gaining visibility and increasing sales on the marketplace.111F[111]

Among those benefits, the most relevant is the Prime label, which allows sellers to participate in special events promoted by Amazon (e.g., Black Friday, Cyber Monday, Prime Day) and benefit from fast and free shipping. Furthermore, Prime increases the likelihood of sellers’ offers being selected as featured offers displayed in the Buy Box. This is of the utmost importance to sellers, as the Buy Box prominently displays just a single seller’s offer for a given product on Amazon’s marketplaces and generates the vast majority of all sales for that product.

Quoting from several of Amazon CEO Jeff Bezos’ letters to shareholders, the AGCM noted the company believes FBA is the “glue” that links Marketplace and Prime112F[112]: “Thanks to FBA, Marketplace and Prime are no longer two things … Their economics and customer experience are now happily and deeply intertwined.”113F[113] Furthermore, FBA is a “game changer” for sellers because it makes their items eligible for Prime benefits, which drives their sales.114F[114] Pursuant to its leveraging strategy, Amazon prevented third-party sellers from associating the Prime label with offers not managed by FBA. In addition, the AGCM noted that third-party sellers using FBA are not subject to the performance-evaluation metrics that Amazon applies in monitoring non-FBA sellers’ performance. Such metrics can ultimately lead to the suspension of non-compliant sellers’ accounts on Amazon.it. All these benefits derived from the use of FBA were considered, to various extents, to be “crucial” to success on the marketplace.115F[115]

It is worth noting that the European Commission has also launched an investigation of Amazon for facts identical to those already addressed in the Italian inquiry, with the relevant market defined as the European Economic Area, except for Italy.116F[116]

The alleged unequal treatment of non-FBA sellers has also been investigated by the Austrian Federal Competition Authority (BWB).117F[117] Although concerned about potential discrimination against sellers who organize their deliveries independently, the Authority conceded that  a better ranking could have also resulted from the better service offered under FBA, compared with the independent organization of deliveries. Hence, the BWB remained open to the possibility that the appearance of preferential treatment for FBA Marketplace sellers was objectively justified.118F[118] The Austrian Authority concluded that a comprehensive and transparent legal framework was the best way to counter problematic business practices and accepted Amazon’s modifications to its terms and conditions.119F[119]

The link between Amazon Marketplace and FBA was also scrutinized as part of the investigation conducted by the U.S. House Judiciary Committee’s Antitrust Subcommittee into the state of competition in digital markets. The subcommittee’s final report found that many third-party sellers have no choice but to purchase fulfillment services from Amazon to maintain a favorable search-result position.120F[120] The report characterized Amazon’s strategy as tying.121F[121]

For the sake of our analysis, the Italian Amazon decision is especially remarkable because of how it contrasts with Google Shopping. As already mentioned, with the exception of the Amsterdam Court of Appeal, it represents the only decision in which the term self-preferencing is used.122F[122] Self-preferencing is here defined as an unequal and unjustified preferential treatment granted by a dominant player to its own services in pursuing a leveraging strategy, hence falling outside the scope of competition on the merits.123F[123] Therefore, rather than reflecting the criteria set out by the General Court, the Italian decision is clearly inspired by the Commission’s approach in Google Shopping. Indeed, in line with the idea of describing self-preferencing as a new form of leveraging abuse, Amazon’s practice is characterized as a form of tying.124F[124]

This definition of self-preferencing is convenient for enforcers, in that it would allow them to bypass the legal standards otherwise required to prove unlawful tying. Indeed, tying requires a form of coercion, such that customers do not have the choice to obtain the tying product without the tied product. In the Amazon case, by contrast, there is neither a contractual obligation nor technical integration between marketplace services and logistics services. Business users are free to run the logistics by themselves or to outsource it to an independent operator without losing the ability to operate on the Amazon e-commerce platform.

Apart from the legal qualification of the conduct in question (which may be more properly characterized as bundling), finding an abuse in a tying case also requires proof of potential foreclosure effects against equally efficient rivals. Looking at the effects on logistics operators, according to the AGCM’s view, vertical integration between the marketplace and logistics constitutes Amazon’s main competitive advantage, which is unmatchable even by equally efficient rivals.125F[125] Indeed, FBA is an integrated logistics service designed to represent “a one-stop shop solution” for storage, shipping, and customer service within a “closed and complete ecosystem” in which Amazon plays multiple roles.126F[126] While Amazon recently started offering a multi-channel logistics service, few retailers have adopted it due to its high operating costs.127F[127] Moreover, part of the AGCM’s decision concerning complaints raised by Amazon’s major e-commerce rival eBay—which reported that a large portion of its market share had been absorbed by Amazon—ultimately recognized that Amazon’s superiority stemmed from its popularity with users and retailers, especially in the critical areas of trust, shipping, and returns.128F[128]

In short, the thing that has made the playing field uneven has been Amazon’s creation of a successful ecosystem, which provides the company with competitive advantages that cannot be replicated either by pure online marketplaces or pure logistics providers.129F[129] A prohibition on self-preferencing may therefore functionally reflect a bias against ecosystems, which require massive and uncertain investment to create, and which provide significant benefits to both business users and final customers.

In summary, the AGCM endorsed an expansive view of the scope of anticompetitive self-preferencing that was at odds with the legal qualifications and narrow criteria set out by the General Court in Google Shopping, and that lacked the context the General Court had laid out to assess the circumstances in which preferential treatment constitutes discriminatory abuse. Notably, an online marketplace does not share many relevant features with an Internet search engine. Indeed, Amazon’s business model is “a closed and complete ecosystem.”130F[130] Moreover, unlike in the Google Shopping case, Amazon is not accused of changing its conduct in response to its market position. The only elements of the criteria defined in Google Shopping that Amazon could be argued to meet are operating in a market with high barriers to entry and being a vertically integrated firm with a super/hyper dominance in the upstream market.131F[131]

Although the General Court’s ruling is mentioned a few times,132F[132] these appear to be last-minute references included merely to note that the Commission’s decision had been upheld by the General Court.133F[133] Since the Italian Amazon decision was delivered just a few weeks after the Google Shopping ruling, it is possible that the AGCM simply did not have time to adjust its line of reasoning to comport with the Court’s qualifications and criteria.

C.   Preferential access to non-public data

A broad interpretation of self-preferencing could find that it covers the preferential provision of data and information, which could similarly be prohibited as abusive. Following this line of reasoning, the European Commission sent a statement of objections to Amazon in November 2020 informing the company of the Commission’s preliminary view that its practice of systematically using non-public business data from independent retailers who sell on its online marketplace infringes antitrust rules, on grounds that Amazon uses that data to benefit its own retail business that directly competes with those retailers.134F[134]

More recently, the UK CMA has also opened an investigation into how Amazon collects and uses third-party seller data, including whether Amazon gains an unfair advantage in business decisions made by its retail arm.135F[135] Similar concerns were raised by staff to the U.S. House Judiciary Antitrust Subcommittee, whose final report argued that Amazon’s asymmetric access to and use of third-party seller data constitutes unfair treatment of those third-party sellers.136F[136] The ACCC likewise warned that Apple and Google both have the ability and incentive to use their positions as app-marketplace operators to monitor downstream competitors.137F[137] For instance, the ACCC found that Apple’s Developer Agreement allows the company to develop, acquire, license, market, promote, and distribute products and software that perform functions the same or similar to any of the products, software or technologies provided by app developers that use the App Store. By contrast, Apple requires that app developers follow obligations to avoid being copycats.

Similar allegations of unfair use of user data have been levied against Facebook by the European Commission and the U.K. CMA, which charge that the company uses data gathered from advertisers in order to compete with them in markets in which Facebook is active, such as classified ads.138F[138] Finally, one focus of the European Commission’s investigation of Apple’s App Store rule requiring developers to use Apple’s in-app purchase mechanism for the distribution of paid apps and/or paid digital content is the potential that competing developers may be disintermediated from important customer data, while Apple can obtain valuable data about the activities and offers of its competitors.139F[139]

These investigations have inspired the bans on so-called “sherlocking” (i.e., the use of business users’ data to compete against them) included both in the DMA140F[140] and the proposed American Innovation and Competition Online Act,141F[141] as well as calls for structural separation and line-of-business restrictions.142F[142]

Amazon has faced accusations that it takes advantage of its dual role and hybrid business model in serving both as a marketplace-service provider and a retailer on the same marketplace, in competition with independent sellers. The charge is that the company can leverage its access to non-public third-party sellers’ data—such as the number of units ordered and shipped, the revenues sellers earn on the marketplace, the number of visits to sellers’ offers, data related to shipping and to sellers’ past performance, and consumer product claims—to identify and replicate popular and profitable products from among the hundreds of millions of listings on its marketplace.

Notably, according to the European Commission’s preliminary findings, such granular and real-time business data feed into the algorithms of Amazon’s retail business, allowing them to calibrate retail offers and strategic business decisions to the detriment of the other marketplace sellers. Thus, it is argued that the appropriation and the use of third-party sellers’ data enables Amazon to avoid the normal risks of retail competition, such as those associated with investing in a new product or choosing a specific price level, and to leverage its dominance in the market for the provision of marketplace services in France and Germany (i.e., the biggest markets for Amazon in the EU).143F[143]

The U.S. House Antitrust Subcommittee similarly charged that Amazon is able to use marketplace data from third-party merchants to create competing private-label products or to source products directly from manufacturers in order to free ride off sellers’ efforts.144F[144] The impact assessment study supporting the DMA confirmed this suspicion, noting that the launch of Amazon Basics (i.e., the most successful private label brand on Amazon’s marketplace) has negatively impacted the sales on Amazon of third-party products in identified attractive product segments.145F[145]

Leveraging this information exclusively, without sharing it with third-party sellers, is considered a form of self-preferencing because Amazon is in position to use data from its marketplace to gain a competitive advantage in market research and to identify new business opportunities without incurring any financial risk.146F[146] Furthermore, by using information from its Amazon fulfilment program, Amazon can also determine where products offered by third-party merchants are being manufactured and by whom. Since Amazon Basics products are sold in large volumes, Amazon can approach the manufacturers of goods for third-party merchants, buy these items in larger quantities, and sell them for a lower price than the competition on its own platform.147F[147]

This line of reasoning aligns with the core concerns about self-preferencing, such as conflicts of interest and the competitive advantages that a platform’s dual role may yield. But to invoke antitrust in cases where a platform performing a dual role gains a competitive advantage would require demonstrating proof of competitive harm, which isn’t apparent in this case. Indeed, while the impact on innovation appears uncertain,148F[148] Amazon’s practice likely benefits consumers by permitting close price comparisons, increasing output, and forcing sellers to reduce their prices.149F[149] Such effects are even more relevant when it comes to sellers with market power, as the introduction of products and services in competition with third parties would reduce double marginalization.

Moreover, the relevance of non-public third-party merchants’ data in facilitating copying by Amazon is questionable. Indeed, as noted, Amazon’s public product reviews supply a great deal of information and any competitor can obtain an item for the purposes of reverse engineering.150F[150] Conversely, if the products in question are protected by intellectual-property rights, Amazon could be found guilty of infringement. Finally, it is unclear whether and how this form of self-preferencing would meet the legal qualification and criteria set out by the General Court in Google Shopping.

Nonetheless, the European Commission is currently evaluating the commitments offered by Amazon, which has proposed to refrain from using non-public data relating to, or derived from, the activities of independent sellers on its marketplace for its retail businesses that compete with those sellers.151F[151] The relevant data would cover both individual and aggregate data (e.g., sales terms, revenues, shipments, inventory-related information, consumer-visit data, or seller performance on the platform). Amazon commits not to use such data for the purposes of selling branded goods, as well as in its private-label products.

III. Imposing platform neutrality under antitrust law

Because preferential treatment may result from a wide range of practices, self-preferencing potentially covers different types of behavior that are subject to different legal standards and that may include exploitative elements.152F[152] Prohibitions on self-preferencing as per se anticompetitive would therefore grant antitrust enforcers significant leeway to bypass the legal standards ordinarily required to prove traditional anticompetitive harms. As a result, such prohibitions would provide antitrust authorities with a powerful tool to intervene in digital markets. This issue is particularly sensitive in Europe where the DMA entrusts the European Commission with the sole power to apply the new regulation but does not displace national competition law. Hence, national competition authorities will remain in charge of the enforcement of national and European antitrust rules.

Besides its potential as an enforcement shortcut, self-preferencing prohibitions may function to impose a neutrality regime on online gatekeepers. The aim would be to ensure a level playing field that currently appears uneven because of the bottleneck and intermediation power exerted by large online platforms. Such rules also could neutralize conflicts of interests raised by platforms’ dual-mode intermediation. The dual roles that some platforms perform fuel the never-ending debate over vertical integration and the related concern that, by giving preferential treatment to its own products and services, an integrated provider may leverage its dominance from one market to related markets. Indeed, the circumstances that may give rise to conflicts of interests and the circumstances that can give rise to leveraging strategies can be similar.153F[153] From this perspective, self-preferencing is a byproduct of the emergence of ecosystems. By integrating complementary products and services, a platform that controls and operates at all levels of the value chain may have an incentive to favor its own offers.154F[154]

But as antitrust authorities generally recognize, self-preferencing conduct is “often benign.”155F[155] Furthermore, since the value that the ecosystem generates depends on the activities of independent complementors, that value is not completely under the control of the platform owner.156F[156] Firms operating on the platform and competing with the platform owner may be disadvantaged by a variety of legal, technological, and informational measures implicated by self-preferencing, but there also may be legitimate justifications for such conduct that would need to be carefully considered in each instance.157F[157] Platforms implement different business models and are driven by different incentives, which in turn affects their strategies.

Against this backdrop, an outright ban on self-preferencing could undermine the very existence of ecosystems by challenging their design and monetization strategies.158F[158] Given that preferential treatment can take many different forms and have very different effects, the different business models adopted by platforms should be subject to case-by-case and effects-based assessment.159F[159] This is also consistent with the industrial-organization literature, which has found mixed evidence on the impact of duality on welfare, thereby supporting the insight that absolute neutrality is not desirable and interventions should be product- and platform-specific.160F[160]

Finally, and more importantly, antitrust law does not impose a general duty to ensure a level playing field by sharing competitive advantages with rivals. Indeed, a competitive advantage cannot be automatically equated with anticompetitive effects.161F[161] Within this framework, the relevance of the general principle of equal treatment that has been invoked by the General Court in Google Shopping to frame self-preferencing as a discriminatory abuse should be regarded with significant skepticism.

This is even more evident in the aftermath of the recent CJEU ruling in Servizio Elettrico Nazionale, which confirmed that the effects-based approach to the assessment of abusive practices remains core to European competition law.162F[162] Notably, the CJEU definitively stated that competition law is not intended to protect the competitive structure of the market, but rather to protect consumer welfare, which represents the goal of antitrust intervention.163F[163]

Accordingly, as illustrated in Intel, not every exclusionary effect is necessarily detrimental to competition.164F[164] Competition on the merits may, by definition, lead some competitors— those that are less efficient and thus less attractive to consumers from the standpoint of price, choice, quality or innovation—to become marginalized or to depart from the market.165F[165] In particular, given that exclusionary effects do not necessarily undermine competition, a distinction must be drawn between a risk of foreclosure and a risk of anticompetitive foreclosure, since only the latter may be penalized under Article 102 TFEU.166F[166] If any conduct having an exclusionary effect were automatically classified as anticompetitive, such rules would become a means to protect less capable, less efficient undertakings and would in no way protect the more meritorious undertakings that stimulate a market’s competitiveness.167F[167]

By and large, these well-settled principles do not support the claim that antitrust rules are designed to ensure platform neutrality. As acknowledged by the General Court in Google Shopping, self-preferencing cannot be considered prima facie unlawful and therefore outside the scope of competition on the merits. Its assessment instead requires the demonstration of anticompetitive effects, taking account of the circumstances of the case and the relevant legal and economic context.168F[168] Toward this aim, a dominant platform remains free to demonstrate that its practice, albeit producing an exclusionary effect, is objectively justified on the basis of all the circumstances of the case, or that the effects are counterbalanced or outweighed by efficiency advantages that also benefit consumers, such as through lower prices, better quality, or a wider choice of new or improved goods and services.169F[169]

In order to assess the anticompetitive nature of a practice, it is necessary to examine whether the means used come within the scope of normal competition.170F[170] Anticompetitive effects do not amount to a mere competitive disadvantage, but require an impact on efficient firms’ ability and incentive to compete.171F[171] Servizio Elettrico Nazionale also clarified the meaning of competition on the merits, considering outside its scope conduct that is not based on obvious economic reasons or objective reasons.172F[172] It is therefore necessary to assess the ability of equally efficient competitors to imitate the conduct of the dominant undertaking. Exclusionary conduct by a dominant firm that can be replicated by equally efficient competitors does not represent the sort of conduct that would lead to anticompetitive foreclosure; it therefore comes within the scope of competition on the merits.173F[173] In order to assess whether a given practice comes within the scope of competition on the merits, the test of whether it would be impossible for an equally efficient rival to imitate the dominant firm’s conduct arises from the case law on both price-related (e.g., TeliaSonera and Post Danmark II) and non-price-related conduct (e.g., Bronner).174F[174]

Moving away from the goal of ensuring a level playing field, recent European case law on self-preferencing centers instead on the competitive advantages that platforms enjoy due to their dual role. A competitive advantage, however, need not amount to anticompetitive foreclosure. Foreclosure not only needs to be proved, but also assessed against potential advantages for consumers, in terms of price, quality, and choice of new goods and services. It is even less clear how NCAs’ expansive approach toward self-preferencing as a standalone abuse fit within this legal framework. Both the AdlC’s decision in Google AdTech and the AGCM’s decision in Amazon Logistics appear inconsistent both with the legal qualification and criteria defined by the General Court, and with the CJEU principles recalled in Servizio Elettrico Nazionale. Similar doubts are raised by the investigations into the preferential access to and use of non-public business data. Moreover, in these cases, the benefits for consumers appear particularly significant as, for instance in Amazon Marketplace, the conduct under investigation led to an immediate output increase and price reduction: in short, more competition.

IV. Conclusion

In her opinion Post Danmark II opinion, Advocate General Juliane Kokott warned that, in enforcing antitrust rules, the CJEU “should not allow itself to be influenced so much by current thinking (‘Zeitgeist’) or ephemeral trends, but should have regard rather to the legal foundations on which the prohibition of abuse of a dominant position rests in EU law.”175F[175] Accordingly, this paper has addressed the prevailing zeitgeist in digital markets, analyzing the markets’ proclaimed peculiar features and the potential scope of application to evaluate whether it should be considered a novel standalone antitrust prohibition.

Indeed, common-carrier antitrust is on the rise. Following the 2017 decision in Google Shopping, the European Commission and some NCAs have advanced a new theory of harm pointing to the competitive disadvantage suffered by rivals. This therefore constitutes a de facto ban on any preferential treatment granted by dominant platforms to their own products and services. Such a strong stance in antitrust enforcement relies on the premise that the special responsibility that an incumbent dominant player bears implies that they must ensure a level playing field.

It remains the case, however, that European case law questions both the goal of relying on antitrust rules to impose a neutrality regime on dominant platforms and the very existence of self-preferencing as an autonomous abuse. Competition law does not impose a general duty to share competition advantages with rivals and does not protect the structure of the market; hence, not every exclusionary effect automatically undermines competition. Self-preferencing is not, in itself, unlawful and platform neutrality as such is outside the scope of competition law.

In contrast with the European Commission and some NCAs, the European General Court in Google Shopping not only framed self-preferencing as a discriminatory abuse but also highlighted some criteria to assess its potential exclusionary effects and considered it outside the scope of competition on the merits. Such criteria are particularly fact-sensitive, and therefore at odds with its wide application as a standalone abuse.

In summary, against the sirens of a fascinating, popular, and convenient new label, the limiting principles of competition law remind us that it cannot be weaponized to ensure a specific market outcome. Therefore, in the aftermath of the Google Shopping ruling, doubts about the characteristics and boundaries of self-preferencing remain on the table, and we still do not have a legal test that distinguishes such purported new antitrust offenses from other practices aimed at pursuing leveraging strategies and already addressed by antitrust rules.

[1] Jacques Crémer, Yves-Alexandre de Montjoye, and Heike Schweitzer, Competition Policy for the Digital Era, European Commission (2019) 7, available at https://ec.europa.eu/competition/publications/reports/kd0419345enn.pdf; Unlocking Digital Competition, UK Digital Competition Expert Panel (2019) 58, available at https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/785547/unlocking_digital_competition_furman_review_web.pdf.

[2] Elizabeth Warren, Here’s How We Can Break up Big Tech, Medium (2019) available at https://medium.com/@teamwarren/heres-how-we-can-break-up-big-tech-9ad9e0da324c.

[3] Margrethe Vestager, Statement Before the U.S. House of Representatives, Subcommittee on Antitrust, Commercial, and Administrative Law (2020) 2, available at https://www.euractiv.com/wp-content/uploads/sites/2/2020/07/Statement-EVP-Vestager-House-SubCommittee-30-July.pdf. See also id., Technology with Purpose (2020), https://ec.europa.eu/commission/commissioners/2019-2024/vestager/announcements/technology-purpose_en.

[4] Thomas Eisenmann, Geoffrey Parker, and Marshall Van Alstyne, Platform Envelopment, 32 Strateg. Manag. J. 1270 (2011).

[5] Regulation (EU) on Contestable and Fair Markets in the Digital Sector (Digital Markets Act), Article 6(1), (3), (5), (7), and (12), https://www.europarl.europa.eu/doceo/document/TA-9-2022-0270_EN.html.

[6] Impact Assessment – A New Pro-Competition Regime for Digital Markets, U.K. Government (2021) para. 21, https://www.gov.uk/government/consultations/a-new-pro-competition-regime-for-digital-markets.

[7] GWB Digitalization Act (Jan. 18, 2021), Section 19a, https://www.bundesrat.de/SharedDocs/beratungsvorgaenge/2021/0001-0100/0038-21.html.

[8] S.2992 – American Innovation and Choice Online Act, 117th Congress (2021-2022) Section 3(a)(1), available at https://www.congress.gov/117/bills/s2992/BILLS-117s2992rs.pdf.

[9] Richard J. Gilbert, The American Innovation and Choice Online Act: Lessons from the 1950 Celler-Kevaufer Amendment, Concurrentialiste (2022), https://leconcurrentialiste.com/gilbert-innovation-choice-act/?mc_cid=8bdf17d95a&mc_eid=34922555f0; Randal Picker, The House’s Recent Spate of Antitrust Bills Would Change Big Tech as We Know It, Promarket (2021), https://promarket.org/2021/06/29/house-antitrust-bills-big-tech-apple-preinstallation.

[10] Investigation of Competition in Digital Markets’, Majority Staff Reports and Recommendations, U.S. House Judiciary Subcommittee on Antitrust, Commercial, and Administrative Law (2020), 380, https://judiciary.house.gov/uploadedfiles/competition_in_digital_markets.pdf?utm_campaign=4493-519. See also Elettra Bietti, Self-Regulating Platforms and Antitrust Justice, Tex. Law Rev. (forthcoming); Nikolas Guggenberg, Essential Platforms, 24 STLR 237 (2021); Rory Van Loo, In Defense of Breakups: Administering a “Radical” Remedy, 105 Cornell L. Rev. 1955 (2020); Lina M. Khan, The Separation of Platforms and Commerce, 119 Columbia Law Rev. 973 (2019); K. Sabeel Rahman, Regulating Informational Infrastructure Internet Platforms as the New Public Utilities, 2 GLTR 234 (2018).

[11] Case AT.39740, Google Search (Shopping), European Commission (Jun. 27, 2017).

[12] Case T-612/17, Google LLC and Alphabet Inc. v. European Commission, European General Court (Nov. 10, 2021), EU:T:2021:763.

[13] See, e.g., Decision No. 29925, FBA Amazon, Autorità Garante della Concorrenza e del Mercato (Nov. 30, 2021), https://www.agcm.it/dettaglio?db=41256297003874BD&uid=801201274D8FDD40C12587AA0056B614&view=&title=A528-FBA%20AMAZON&fs=Abuso%20di%20posizione%20dominante. Previously, see Decision 21-D-11, Google, Autorité de la Concurrence (Jun. 7 2021) , https://www.autoritedelaconcurrence.fr/en/decision/regarding-practices-implemented-online-advertising-sector.

[14] Giuseppe Colangelo, The Digital Markets Act and EU Antitrust Enforcement: Double & Triple jeopardy, ICLE White Paper (2022), https://laweconcenter.org/resource/the-digital-markets-act-and-eu-antitrust-enforcement-double-triple-jeopardy.

[15] Pablo Ibáñez Colomo, Self-Preferencing: Yet Another Epithet in Need of Limiting Principles, 43 World Competition 417 (2020).

[16] Nicolas Petit, A Theory of Antitrust Limits, 28 Geo. Mason L. Rev. 1399 (2021).

[17] Herbert Hovenkamp, Selling Antitrust, Hastings L.J. (forthcoming).

[18] See Opinion of Advocate General Wahl, 20 December 2017, Case C‑525/16, MEO — Serviços de Comunicações e Multimédia SA v. Autoridade da Concorrência, EU:C:2017:1020, paras. 76-77, arguing that a distinction must be drawn between undertakings that are vertically integrated (and have an interest in displacing competitors on the downstream market) and those that have no such interest. In the case of vertically integrated undertakings, the application by a dominant undertaking of discriminatory prices on the downstream or upstream market is, in reality, similar to first-degree price discrimination, which indirectly affects the undertaking’s competitors. See also Inge Graef, Differentiated Treatment in Platform-to-Business Relations: EU Competition Law and Economic Dependence, 38 YEL 448, 452-453 (2019), distinguishing among pure self-preferencing (whereby a vertically integrated platform treats its affiliated services more favorably than non-affiliated services), pure secondary line differentiation (whereby a non-vertically integrated platform engages in differentiated treatment among unaffiliated services in a market in which it is not active itself), and a hybrid category in which either a vertically integrated or a non-vertically integrated platform engages in differentiated treatment among unaffiliated services in an effort to favor its own business.

[19] Nicolas Petit, Theories of Self-Preferencing Under Article 102 TFEU: A Reply to Bo Vesterdorf, 1(3) CLPD 4 (2015).

[20] Case C-413/14 P, Intel v. Commission, Court of Justice of the European Union (Sept. 6, 2017), EU:C:2017:632.

[21] Id., C‑525/16, MEO v. Autoridade da Concorrência (Apr. 19, 2018),  EU:C:2018:270. See also Wahl, supra note 18, para. 61.

[22] Guidance on the Commission’s Enforcement Priorities in Applying Article 82 of the EC Treaty to Abusive Exclusionary Conduct by Dominant Undertakings, European Commission (2009) OJ C 45/7, paras. 47-62.

[23] See Autorité de la Concurrence, supra note 13, para. 410, binding limits on interoperability with third-party services servers cannot be considered competition on the merits. See also Herbert Hovenkamp, Antitrust and the Design of Production, 103 Cornell L. Rev. 1155 (2018); Pablo Ibáñez Colomo, Product Design and Business Models in EU Antitrust Law, SSRN (2021), https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3925396.

[24] Case T-201/04, Microsoft v. Commission, European General Court (Sept. 17, 2007), para. 1036, EU:T:2007:289.

[25] Case AT.40099, Google Android, European Commission, (Jul. 18, 2018), confirmed by Case T-604/18, Google v. Commission, European General Court (Sept. 14, 2022) EU:T:2022:541.

[26] Dirk Auer, Appropriability and the European Commission’s Android Investigation, 23 CJEL 647 (2017).

[27] Christopher S. Yoo, Open Source, Modular Platforms, and the Challenge of Fragmentation, 1 Criterion Journal on Innovation 619 (2016).

[28] UK Government, supra note 6, para. 21.

[29] Joined Cases C-241/91 P and 242/91 P, RTE and ITP v. Commission, Court of Justice of the European Union (Apr. 6, 1995), EU:C:1995:98.

[30] Id., 29, Case C-418/01, IMS Health GmbH & Co. OHG v. NDC Health GmbH & Co. GH (Apr. 29, 2004), EU:C:2004:257.

[31] European General Court, supra note 24.

[32] Case C-7/97, Oscar Bronner GmbH & Co. KG v. Mediaprint Zeitungs- und Zeitschriftenverlag GmbH & Co. KG, Mediaprint Zeitungsvertriebsgesellschaft mbH & Co. KG and Mediaprint Anzeigengesellschaft mbH & Co. KG, Court of Justice of the European Union (Nov. 26, 1998), EU:C:1998:569.

[33] Id., Case C-165/19 P, Slovak Telekom a.s. v. Commission (Mar. 25, 2021) para. 50, EU:C:2021:239.

[34] Ibid., para. 57. In a similar vein, see Case T‑814/17, Lietuvos geležinkeliai AB v. Commission, European General Court (Nov. 18, 2020), para. 92, EU:T:2020:545.

[35] European Commission, supra note 22, para. 80.

[36] Ibid., para. 82; and Case COMP/38.784, Wanadoo España v. Telefónica, European Commission (Jul 4, 2007). This is likely to occur in two cases: where regulation compatible with EU law already imposes an obligation to supply on the dominant undertaking and it is clear, from the considerations underlying such regulation, that the necessary balancing of incentives has already been made by the public authority when imposing such obligation; or where the upstream market position of the dominant firm has been developed under the protection of special or exclusive rights or has been financed by state resources.

[37] Case C-280/08 P, Deutsche Telekom AG v. European Commission (Deutsche Telekom I), Court of Justice of the European Union (Oct. 14, 2010), EU:C:2010:603.

[38] Id., Case C-52/09, Konkurrensverket v. TeliaSonera Sverige AB (Feb. 17, 2011), EU:C:2011:83.

[39] Id., Case C‑295/12 P, Telefónica SA and Telefónica de España SAU v. European Commission (Jul. 10, 2014), EU:C:2014:2062.

[40] Id., supra note 33.

[41] Case C/13/528337, VBO Makelaar v. Funda, Gerechtshof Amsterdam (May 16, 2020), para. 3.12.3.

[42] Autorità Garante della Concorrenza e del Mercato, supra note 13, paras. 236, 393, 436, 504, 708, 710, 716, and 901.

[43] European Commission, supra note 11.

[44] Streetmap.EU Ltd. v. Google and Others, [2016] EWHC 253 (ch).

[45] Ibid., paras. 51-54.

[46] Ibid., paras. 84 and 147.

[47] Ibid., para. 84.

[48] Ibid., para. 149.

[49] Ibid., para. 139.

[50] Ibid., para. 161.

[51] European Commission, supra note 11, para. 344.

[52] Ibid., para. 649.

[53] Ibid.

[54] Case C-333/94, Tetra Pak International SA v. Commission, Court of Justice of the European Union (Nov. 14, 1996), EU:C:1996:436.

[55] Case T-288/97, Irish Sugar plc v Commission, European General Court (Oct. 7, 1999), EU:T:1999:246.

[56] Id., supra note 24.

[57] Case C-311/84, Centre Belge D’Etudes de Marché – Télémarketing (CBEM) v. SA Compagnie Luxembourgeoise de Télédiffusion (CLT) and Information Publicité Benelux (IPB), Court of Justice of the European Union (Oct. 3, 1985), EU:C:1985:394.

[58] European Commission, supra note 11, para. 334.

[59] Ibid., para. 645.

[60] Ibid., para. 651.

[61] See, e.g., Jay Pil Choi and Doh-Shin Jeon, A Leverage Theory of Tying in Two-Sided Markets with Nonnegative Price Constraints, 13 Am Econ J Microecon 283 (2021); Edward Iacobucci and Francesco Ducci, The Google Search Case in Europe: Tying and the Single Monopoly Profit Theorem in Two‐Sided Markets, 47 Eur. J. Law Econ. 15 (2019); Eduardo Aguilera Valdivia, The Scope of the ‘Special Responsibility’ upon Vertically Integrated Dominant Firms after the Google Shopping Case: Is There a Duty to Treat Rivals Equally and Refrain from Favouring Own Related Business?, 41 World Competition 43 (2018);  Pinar Akman, The Theory of Abuse in Google Search: A Positive and Normative Assessment under EU Competition Law, 2 J. Tech. L. & Pol’y 301 (2017); Ioannis Kokkoris, The Google Case in the EU: Is There a Case?, 62 Antitrust Bull. 313 (2017); John Temple Lang, Comparing Microsoft and Google: The Concept of Exclusionary Abuse, 39 World Competition 5 (2016); Renato Nazzini, Unequal Treatment by Online Platforms: A Structured Approach to the Abuse Test in Google, SSRN (2016), https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2815081; Bo Vesterdorf, Theories of Self-Preferencing and Duty to Deal – Two Sides of the Same Coin?, 1(1) CLPD 4 (2015); Petit, supra note 19.

[62] Crémer, de Montjoye, and Schweitzer, supra note 1.

[63] Gerechtshof Amsterdam, supra note 41.

[64] Ibid., para. 3.12.1.

[65] Autorité de la Concurrence, supra note 13.

[66] Digital Advertising Services Inquiry Final Report, Australian Competition and Consumer Commission (2021), available at https://www.accc.gov.au/system/files/Digital%20advertising%20services%20inquiry%20-%20final%20report.pdf.

[67] Autorité de la Concurrence, supra note 13, paras. 369 and 410.

[68] Ibid., paras. 366 and 369. Since Google did not hold a dominant position in the market for SSPs, the reference to Tetra Pak has also been used to argue that, under specific circumstances, behavior implemented in a non-dominated market that has effects on the dominated market may be considered abusive (see para. 367).

[69] CJEU, supra note 33.

[70] Decision 21-D-07, Apple, Autorité de la Concurrence, (Mar. 17, 2021) https://www.autoritedelaconcurrence.fr/en/decision/regarding-request-interim-measures-submitted-associations-interactive-advertising-bureau.

[71] Ibid., para. 163.

[72] Bundeskartellamt Reviews Apple’s Tracking Rules for Third-Party Apps (press release), Bundeskartellamt (2022) https://www.bundeskartellamt.de/SharedDocs/Meldung/EN/Pressemitteilungen/2022/14_06_2022_Apple.html.

[73] Mobile Ecosystems: Market Study Final Report, U.K. Competition and Markets Authority (2022), Chapter 6 and Appendix J, https://www.gov.uk/cma-cases/mobile-ecosystems-market-study.

[74] European General Court, supra note 12.

[75] Ibid., para. 155.

[76] More recently, see also Case 50972, Google Privacy Sandbox, U.K. Competition and Markets Authority (Feb. 11, 2022), https://www.gov.uk/cma-cases/investigation-into-googles-privacy-sandbox-browser-changes, which considered self-preferencing as a traditional discrimination abuse. In particular, the competitive risks the CMA highlighted involve Google’s self-preferencing its own ad inventory and ad-tech services by transferring key functionalities to Chrome. The Privacy Sandbox Project would offer Google the ability to affect digital-advertising-market outcomes through Chrome in a way that cannot be scrutinized by third parties. It could lead to conflicts of interests because Google operates as publisher and ad-tech provider simultaneously.

[77] General Court, supra note 12, para. 160.

[78] Ibid., para. 163.

[79] Case C-242/95, GT-Link A/S v. De Danske Statsbaner (DSB), Court of Justice of the European Union (Jul. 17, 1997), EU:C:1997:376.

[80] Id., Case C-82/01 P, Aéroports de Paris v. Commission (Oct. 24, 2002), EU:C:2002:617.

[81] Id., supra note 21.

[82] See Lena Hornkohl, Article 102 TFEU, Equal Treatment and Discrimination after Google Shopping, 13 J. Eur. Compet. Law Pract. 99 (2022), mentioning Post Danmark I as an example of primary-line exclusionary discrimination in the predatory-pricing context and TeliaSonera (supra note 38) and Slovak Telekom (supra note 33) as examples of secondary-line exclusionary discrimination in the margin-squeeze context.

[83] General Court, supra note 12, paras. 166 and 175.

[84] Ibid., paras. 169-174.

[85] Ibid., paras. 176-177.

[86] Ibid.

[87] Ibid., para. 178.

[88] Regulation (EU) 2015/2120 laying down measures concerning open Internet access and amending Directive 2002/22/EC on universal service and users’ rights relating to electronic communications networks and services, and Regulation (EU) No 531/2012 on roaming on public mobile-communications networks within the European Union, (2015) OJ L 310/1.

[89] General Court, supra note 12, para. 180. Comparisons to net neutrality have also been made by the U.S. House Judiciary Subcommittee on Antitrust, Commercial, and Administrative Law, supra note 10, 382-383, recommending that Congress consider establishing nondiscrimination rules.

[90] General Court, supra note 12, paras. 180, 182 and 183.

[91] Ibid., paras. 178, 182, 183, and 237.

[92] Ibid., paras. 176 and 179.

[93] Ibid..

[94] Ibid., paras. 176 and 179.

[95] Ibid., para. 181.

[96] Ibid., paras. 182-184.

[97] Ibid., paras. 219-222.

[98] Ibid., paras. 227.

[99] Ibid., para. 224.

[100] Ibid., paras. 232 and 233.

[101] Ibid., para. 238.

[102] Ibid., para. 240.

[103] See Christian Ahlborn, Gerwin Van Gerven, and William Leslie, Bronner Revisited: Google Shopping and the Resurrection of Discrimination Under Article 102 TFEU, 13 J. Eur. Compet. Law Pract. 87 (2022); Friso Bostoen, The General Court’s Google Shopping Judgement Finetuning the Legal Qualification and Tests for Platform Abuse, 13 J. Eur. Compet. Law Pract. 75 (2022), and Hornkohl, supra note 82, arguing that the ruling has resurrected discriminatory abuses as potentially one of the most important tools for regulating the platform economy.

[104] See Elias Deutscher, Google Shopping and the Quest for a Legal Test for Self-preferencing Under Article 102 TFEU, 6 European Papers 1345, 1348 (2021), arguing that the judgment did not address the fundamental question of how far Article 102 TFEU can interfere with the design choices of dominant firms or prohibit them from granting favorable treatment to their own products or services.

[105] General Court, supra note 12, para. 196.

[106] Ibid., para. 187.

[107] Ibid., para. 245.

[108] See Bostoen, supra note 103, arguing that, at best, this form of favoritism may be applicable in other cases of prominent display and positioning in searches (e.g., the conduct of Amazon favoring its own retail offers); and Ahlborn, Van Gerven, and Leslie, supra note 103, noting that, in contrast to price discrimination, discriminatory access to an input can encompass a range of factors that are difficult to disentangle (e.g., greater interoperability or preferential access to core services).

[109] Autorità Garante della Concorrenza e del Mercato, supra note 13.

[110] Ibid., para. 702, describing Amazon’s conduct as an “abusive pressure.”

[111] Id., Italian Competition Authority: Amazon Fined over € 1,128 Billion for Abusing Its Dominant Position (2021), https://en.agcm.it/en/media/press-releases/2021/12/A528.

[112] Autorità Garante della Concorrenza e del Mercato, supra note 13, para. 254.

[113] Ibid..

[114] Ibid., paras. 253 and 737.

[115] Ibid., para. 698.

[116] See C(2020) 7692 Final, European Commission (Nov. 10, 2020). See also Margrethe Vestager, Statement by Executive Vice-President Vestager on Statement of Objections to Amazon for the Use of Non-Public Independent Seller Data and Second Investigation into Its E-Commerce Business Practices (2020), https://ec.europa.eu/commission/presscorner/detail/en/STATEMENT_20_2082. See also CMA Investigates Amazon over Suspected Anti-Competitive Practices, U.K. Competition and Markets Authority (2022), https://www.gov.uk/government/news/cma-investigates-amazon-over-suspected-anti-competitive-practices, opening an investigation into how Amazon sets criteria for allocation of suppliers to be the preferred in the Buy Box and how Amazon sets the eligibility criteria for selling under the Prime label. The European Commission is currently evaluating the commitments offered by Amazon (Commission Seeks Feedback on Commitments Offered by Amazon Concerning Marketplace Seller Data and Access to Buy Box and Prime, European Commission (2022) https://ec.europa.eu/commission/presscorner/detail/en/ip_22_4522). Amazon has committed to apply equal treatment to all sellers when selecting the winner of the Buy Box and to display a second competing offer to the Buy Box winner if there is a second offer that is sufficiently differentiated on price and/or delivery. Both offers will display the same descriptive information and provide for the same purchasing experience. Moreover, regarding Prime, Amazon has committed to set non-discriminatory conditions and qualifying criteria for marketplace sellers and offers, to allow Prime sellers to freely choose any carrier for their logistics and delivery services, and not to use any information obtained through Prime about the terms and performance of third-party carriers for its own logistics services.

[117] Bundeswettbewerbsbehörde (Jul. 17, 2019), available at https://www.bwb.gv.at/fileadmin/user_upload/Fallbericht_20190911_en.pdf.

[118] Ibid., para. 81.

[119] Ibid., para. 87.

[120] U.S. House Judiciary Subcommittee on Antitrust, Commercial, and Administrative Law, supra note 10, 287-291.

[121] Ibid., 287-288.

[122] Autorità Garante della Concorrenza e del Mercato, supra note 13, paras. 236, 504, 710, 716, and 901.

[123] Ibid., paras. 236, 504, 506, 716, 723, and 810.

[124] Ibid., paras. 505, 713, 726, 760, 826, 852, 857, and 874. See also Autorità Garante della Concorrenza e del Mercato, supra note 111.

[125] Id., supra note 13, para. 807.

[126] Ibid., paras. 127, 136, 188, 614, and 804.

[127] Ibid., paras. 834-835.

[128] Ibid., paras. 658-666, 679, and 682.

[129] Ibid., paras. 805-806.

[130] Ibid., para. 136.

[131] Ibid., paras. 506, 609, 610, 680, and 716.

[132] Ibid., paras. 610, 710, 716, and 723.

[133] Ibid., para. 710.

[134] Case AT.40462, Amazon Marketplace, European Commission (Nov. 10, 2020). See also Commission Sends Statement of Objections to Amazon for the Use of Non-Public Independent Seller Data and Opens Second Investigation into Its E-Commerce Business Practices, European Commission, (2020) https://ec.europa.eu/commission/presscorner/detail/en/ip_20_2077.

[135] U.K. Competition and Markets Authority, supra note 116.

[136] U.S. House Judiciary Subcommittee on Antitrust, Commercial, and Administrative Law, supra note 10, 275.

[137] App Marketplaces: Interim Report, Australian Competition and Consumer Commission, (2021), 134, available at https://www.accc.gov.au/system/files/Digital%20platform%20services%20inquiry%20-%20March%202021%20interim%20report.pdf.

[138] Case AT.40684, Facebook leveraging, European Commission (Jun. 4, 2021), https://ec.europa.eu/competition/elojade/isef/case_details.cfm?proc_code=1_AT_40684; CMA Investigates Facebook’s Use of Ad Data, U.K. Competition and Markets Authority (2021), https://www.gov.uk/government/news/cma-investigates-facebook-s-use-of-ad-data.

[139] Case AT.40716, Apple – App Store Practices, European Commission (Jun. 16, 2020).

[140] DMA, supra note 5, Article 6(1).

[141] AICOA, supra note 8, Section 3.

[142] See, e.g., U.S. House Judiciary Subcommittee on Antitrust, Commercial, and Administrative Law, supra note 10, 380. See also Simon P. Anderson and Özlem Bedre-Defolie, Hybrid Platform Model, CEPR Discussion Paper No. 16243 (2021), available at https://cepr.org/active/publications/discussion_papers/dp.php?dpno=16243, arguing that the hybrid business model leads to higher platform fees for third-party sellers, less variety on the platform, and lower consumer welfare, compared to when the platform is a pure marketplace. On a different note, see Herbert Hovenkamp, Antitrust and Platform Monopoly, 130 Yale Law J. 1952 (2021), considering structural separation as the worst solution for the problems raised by Amazon’s strategy. See also Andrei Hagiu, Tat-How Teh, and Julian Wright, Should Platforms Be Allowed to Sell on Their Own Marketplaces?, 53 Rand J Econ 297 (2022), arguing that a structural remedy, such as an outright ban on the dual mode, would be detrimental to consumer surplus or total welfare, since the presence of the intermediary’s products constrains the pricing of the third-party sellers on its marketplace. The authors consider preferable policy interventions that target specific behaviors by the platform, such as a ban on product imitation and on self-preferencing.

[143] European Commission, supra note 134.

[144] U.S. House Judiciary Subcommittee on Antitrust, Commercial, and Administrative Law, supra note 10, 275.

[145] Digital Markets Act: Impact Assessment Support Study (Annexes), European Commission (2020), 301-308, https://op.europa.eu/en/publication-detail/-/publication/2a69fd2a-3e8a-11eb-b27b-01aa75ed71a1/language-en/format-PDF/source-search.

[146] Ibid., 304.

[147] Ibid..

[148] See Hagiu, Teh, and Wright, supra note 142, arguing that a ban on product imitation by a platform restores sellers’ incentive to innovate; Erik Madsen and Nikhil Vellodi, Insider Imitation, SSRN (2022) available at https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3832712, finding that a ban on a platform’s use of marketplace data can either stifle or stimulate innovation, depending on the nature of innovation. Namely, it stimulates innovation only for experimental product categories with significant upside demand potential. On the impact of data usage on consumer welfare, because of the tradeoff between static benefits from lower prices and dynamic costs from lower sellers’ incentives to investment, see Federico Etro, Product Selection in Online Marketplaces, 30 J Econ Manag Strategy 614 (2021).

[149] See Herbert Hovenkamp, The Looming Crisis in Antitrust Economics, 101 B.U. L. 489, 543 (2021), finding no evidence to suggest that the practice is so prone to abuse or so likely to harm consumers in other ways that it should be categorically condemned: “Rather, it is an act of partial vertical integration similar to other practices that the antitrust laws have confronted and allowed in the past. One close analogy is dual distribution, which occurs when a firm sells through both independent franchisees and its wholly owned stores. Such practices nearly always increase output, benefitting consumers and typically even independent competing firms.” On the different impact of Amazon’s practice on consumer welfare and third-party sellers’ welfare, see also Feng Zhu and Qihong Liu, Competing with Complementors: An Empirical Look at Amazon.com, 39 Strateg. Manag. J. 2618 (2018), finding that Amazon tends to enter into high-quality, popular products sold by third-party merchants and that such entry tends to lower prices and lead to the exit of third-party sellers; and Nan Chen and Hsin-Tien Tsai, Steering via Algorithmic Recommendations, (2021) https://www.tse-fr.eu/sites/default/files/TSE/documents/sem2021/tsai.pdf, arguing that Amazon tends to recommend products sold by Amazon Retail to consumers over products sold by third-party retailers, and that this steering is inconsistent with Amazon promoting consumer welfare.

[150] Hovenkamp, supra note 142, 2015-2016.

[151] European Commission, supra note 116.

[152] See Graef, supra note 18, distinguishing Google Shopping and Amazon Marketplace as pure self-preferencing cases (i.e., primary-line injuries whose key objective is to exclude rivals from the market) from the Italian Amazon Logistics case, which belongs instead to a hybrid category that includes a mix of exploitative and exclusionary elements.

[153] Australian Competition and Consumer Commission, supra note 66, 92.

[154] Colomo, supra note 15, 418.

[155] Updating Competition and Consumer Law for Digital Platform Services, Australian Competition and Consumer Commission (2022), 85, https://www.accc.gov.au/media-release/feedback-sought-on-potential-new-rules-for-large-digital-platforms.

[156] Tobias Kretschmer, Aija Leiponen, Melissa Schilling, and Gurneeta Vasudeva, Platform Ecosystems as Meta‐Organizations: Implications for Platform Strategies, 43 Strateg. Manag. J. 405 (2022).

[157] Kevin J. Boudreau and Andrei Hagiu, Platforms Rules: Multi-Sided Platforms as Regulators, in (Annabelle Gawer, ed.) Platforms, Markets and Innovation, Cheltenham, Edward Elgar Publishing (2009), 163; David Evans, Governing Bad Behavior by Users of Multi-sided Platforms, 27 BTLJ 1201 (2012).

[158] See Hovenkamp, supra note 17, considering the pending U.S. self-preferencing legislation “an affront to both antitrust policy and intelligent regulatory policy.” See also Geoffrey A. Manne, Against the Vertical Discrimination Presumption, 2 Concurrences 1, 2 (2020), arguing that forcing platforms to allow complementors to compete on their own terms would affect platform incentives for innovation. Indeed, platforms have an incentive to optimize openness and mandating maximum openness is not necessarily optimal because it would disregard the trade-off faced by platforms. Consequently, any presumption of harm from vertical discrimination is not based on sound economics. In a similar vein, Jonathan M. Barnett, The Host’s Dilemma: Strategic Forfeiture in Platform Markets for Informational Goods, 124 Harv. L. Rev. 1861 (2011).

[159] See Australian Competition and Consumer Commission, supra note 155, 87, arguing that rules might need to be “specifically tailored to each digital platform service with a high level of precision, to target the specific conduct that causes anti-competitive harm.”

[160] See Patrice Bougette, Axel Gautier, and Frédéric Marty, Business Models and Incentives: For an Effects-Based Approach of Self-Preferencing?, 13 J. Eur. Compet. Law Pract.136, 140 (2022). On the welfare effects of Amazon’s dual role and the welfare implications of proposed regulations, see Germán Gutiérrez, The Welfare Consequences of Regulating Amazon, (2021) available at https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3965566, showing that interventions that eliminate either the Prime program or product variety are likely to decrease welfare. See also Guy Aridor and Duarte Gonçalves, Recommenders’ Originals: The Welfare Effects of the Dual Role of Platforms as Producers and Recommender Systems, 83 Int. J. Ind. Organ.102845 (2022), highlighting the importance of targeted restrictions on self-preferencing because of the ambiguity of the welfare implications of a policy remedy separating recommendation and production or imposing unbiased recommendations. However, with regards to app stores and device-funded gatekeepers, Jorge Padilla, Joe Perkins, and Salvatore Piccolo, Self-Preferencing in Markets with Vertically Integrated Gatekeeper Platforms, J. Ind. Econ. (forthcoming) find that consumer welfare would be increased by preventing the device seller from selling its own apps and associated services in competition with third-party apps. See also Morgane Cure, Matthias Hunold, Reinhold Kesler, Ulrich Laitenberger, and Thomas Larrieu, Vertical Integration of Platforms and Product Prominence, Quant. Mark. Econ. (forthcoming), studying the potential effects of self-preferencing in the online hotel-booking industry because of the integration between one of the major online travel agencies (Booking Holdings) and a meta-search platform (Kayak). According to their empirical findings, the horizontal ranking of sales channels for a given hotel indicate that sales channels of online travel agents by Booking Holdings are more often position leaders than price leaders and online travel agents affiliated to Booking Holdings have a higher probability than other online travel agents of being among the visible providers and of being the highlighted sales channel. Moreover, for the vertical ranking of hotels for a search request, hotels are ranked worse in the Kayak search results when an online travel agent of the Expedia Group is the cheapest sales channel.

[161] Colomo, supra note 15, 421; Id., Anticompetitive Effects in EU Competition Law, 17 J. Competition Law Econ. 309, 356 (2021).

[162] Case C-377/20, Servizio Elettrico Nazionale SpA v. Autorità Garante della Concorrenza e del Mercato, Court of Justice of the European Union (May 12, 2022), EU:C:2022:379.

[163] Ibid., para. 46.

[164] CJEU, supra note 20, paras. 133-134.

[165] CJEU, supra note 162, para. 73.

[166] Case C‑377/20, Servizio Elettrico Nazionale SpA v. Autorità Garante della Concorrenza e del Mercato, Advocate General Rantos (Dec. 9, 2021), EU:C:2021:998, para. 43.

[167] Ibid., para. 45.

[168] See CJEU, TeliaSonera, supra note 38; Post Danmark I, supra note 82; Case C-23/14, Post Danmark A/S v. Konkurrencerådet (Post Danmark II), (Oct. 6, 2015), EU:C:2015:651; Intel, supra note 20; Case C-307/18, Generics (UK) and Others v. Competition and Markets Authority (Jan. 30, 2020), EU:C:2020:52; Case C-152/19 P, Deutsche Telekom v. Commission (Mar. 25, 2021) EU:C:2021:238.

[169] CJEU, supra note 162, paras. 84-85.

[170] Ibid., para. 75. See also Advocate General Rantos, supra note 166, paras. 48-50, arguing that demonstrating that a dominant undertaking used means other than those that come within the scope of normal competition is not a requirement that needs to be assessed separately from the restrictive effect of the conduct.

[171] See, e.g., CJEU, MEO, supra note 21, and Post Danmark II, supra note 168. See also Colomo, supra note 161.

[172] Advocate General Rantos, supra note 166, para. 62.

[173] Ibid., para. 68 and CJEU, supra note 162, paras. 77-79.

[174] CJEU, supra note 162, para. 79.

[175] Case C‑23/14, Post Danmark A/S v. Konkurrencerådet, Advocate General Kokott (May 21, 2015), EU:C:2015:343, para. 4.

Introduction

Imagine you are at the grocery-store checkout line and it is to pay. You enter your credit card in the terminal, assuming that your payment will be routed over the network operated by the brand on your card (typically Visa or Mastercard). But you learn after the fact that the grocery store has chosen instead to route it over China Union Pay.

Most of us would be uncomfortable ceding to the merchant the authority to route transactions over the cheapest network, without considering our concerns about security, reliability, and other card features (including rewards). Yet that is already the case for many point-of-sale transactions made with debit cards—the result of a 2011 regulation implemented by the Federal Reserve. Consumers can, however, often still force the transaction to run over their preferred network by pushing the “credit” button.

But new rules under consideration by the Federal Reserve would extend merchants’ ability to determine how debit transactions are routed to online transactions, while also making it more difficult for consumers to control who gets to handle their personal data and process their transactions.[1] Perhaps more worryingly, a new bill (the “Credit Card Competition Act”) introduced by Sen. Richard Durbin (D-Ill.) would, in the name of “competition,” impose similar routing requirements on credit cards, while ignoring important differences in the competitive framework of debit and credit cards.[2]

Since they emerged more than 50 years ago, payment-card networks have come to play an increasingly important role in our lives, both directly and indirectly. Directly, they facilitate hundreds of billions of transactions every year, representing tens of trillions of dollars in value.[3] Indirectly, they have contributed to a near-complete shift from paper-based to electronic value exchange and accounting in the United States and many other countries. This has, in turn, resulted in enormous efficiency improvements and wider social benefits, such as the development of online commerce, greater ease of travel, and reduced tax avoidance.[4]

The shift from paper to electronic value exchange has been driven almost entirely by voluntary decisions made by businesses and consumers. Despite such clear evidence of market success, over the past three decades, governments have increasingly sought to correct alleged “market failures” in payment-card markets. The main tool governments have used is price controls on interchange-fee rates. More recently, however, several governments—including the United States, the European Union, and Australia—have sought to reduce rates further still by regulating the manner in which payments are “routed” (i.e., the way that messages pertaining to a transaction are sent between the merchant and the issuing bank). This has important implications for consumer protection, fraud prevention, and financial inclusion.

In previous studies, we have shown that regulation of interchange fees typically has slowed the shift to more innovative, quicker, more convenient payment systems, while also reducing other benefits and particularly harming poorer consumers and smaller merchants.^[5]

Prohibitions on exclusivity in routing have similar effects as direct price controls. But imposed routing requirements will have additional effects that go beyond those of price controls and would result in various harms to consumers and the economy. This study seeks to delve deeper into the problem, focusing primarily on the justifications for and effects of regulations that affect the way in which transactions are routed. While “routing” may seem arcane, it is fundamental to the effectiveness of payment networks. Understanding the likely consequences of such regulation is thus important. That is the purpose of this paper.

We begin, in Section II, by describing the technological and economic elements of payment-card routing. Supporters of forced routing requirements contend that they will promote more efficient competition in consumers’ payment-card usage. But we show that this superficial argument ignores the basic economic realities of payment-card networks, as well as the fundamentally different nature of consumer competitive choice, both in debit-card markets (where routing requirements currently exist) and in credit-card markets (the intended target of Sen. Durbin’s proposed law). Section III reviews the evidence regarding the effects of regulating payment networks. We summarize the pernicious effects of price controls and then explain how the routing mandate created by the 2011 Federal Reserve regulation, known as Regulation II, has had similar effects. Section IV considers the proposed changes to Regulation II and the new Durbin proposal to regulate credit-card routing, with a particular focus on the likely harmful effects of the changes on the incidence of fraud and the knock-on effects on issuers, cardholders, and merchants. Section V concludes.

[1] Debit Card Interchange Fees and Routing, FR 26189 (2021), available at: https://www.govinfo.gov/content/pkg/FR-2021-05-13/pdf/2021-10013.pdf.

[2] Credit Card Competition Act of 2022, S. 4674, 117th Cong. § 2 (2022), available at: https://www.congress.gov/117/bills/s4674/BILLS-117s4674is.pdf.

[3]  Global Network Card Results in 2021, Nilson Report Issue 1224, https://nilsonreport.com/mention/1672/1link.

[4] See the appendix to this paper and references therein.

[5] See Todd J. Zywicki, The Economics of Payment Card Interchange Fees and the Limits of Regulation, ICLE Financial Regulatory Program White Paper Series (Jun. 2, 2010), available at http://laweconcenter.org/images/articles/zywicki_interchange.pdf; Todd J. Zywicki, Geoffrey A. Manne, and Julian Morris, Unreasonable and Disproportionate: How the Durbin Amendment Harms Poorer Americans and Small Businesses, International Center for Law and Economics (Apr. 25, 2017); Todd J. Zywicki, Geoffrey A. Manne, and Julian Morris, Price Controls on Payment Card Interchange Fees: The U.S. Experience, George Mason Law & Economics Research Paper No. 14-18, (Jun. 6, 2014).

Executive Summary

In contrast to its stated aims to promote a Digital Single Market across the European Union, the proposed Digital Markets Act (DMA) could serve to fragment Europe’s legal framework even further, largely due to overlaps with competition law. This paper provides an analytical overview of areas where conflicts would inevitably arise from dual application of the DMA and European and national-level antitrust rules. It counsels full centralization of the DMA’s enforcement at the EU level to avoid further fragmentation, as well as constraining the law’s scope by limiting its application to a few large platform ecosystems.

Introduction

The Digital Markets Act (DMA) has entered the last and decisive stage of its approval process. With the Council of Europe having reached consensus on its general approach[1] and the European Parliament having adopted amendments,[2] the DMA proposal has moved into the inter-institutional negotiations known as the so-called “trilogue.”

The DMA has spurred a lively debate since it initially was proposed by the European Commission in December 2020.[3] This deliberative process has touched on all the proposal’s features, including its aims and scope, the regulations and rule-based approach it would adopt, and the measure’s institutional design. However, given the positions expressed by the Council and the Parliament, the rationale for DMA intervention and the proposal’s relationship with antitrust law remain relevant topics for exploration.

The DMA is grounded explicitly on the notion that competition law alone is insufficient to effectively address the challenges and systemic problems posed by the digital platform economy. Indeed, the scope of antitrust is limited to certain instances of market power (e.g., dominance on specific markets) and of anti-competitive behavior.[4] Further, its enforcement occurs ex post and requires extensive investigation on a case-by-case basis of what are often very complex sets of facts.[5] Moreover, it may not effectively address the challenges to well-functioning markets posed by the conduct of gatekeepers, who are not necessarily dominant in competition-law terms.[6] As a result, proposals such as the DMA invoke regulatory intervention to complement traditional antitrust rules by introducing a set of ex ante obligations for online platforms designated as gatekeepers. This also allows enforcers to dispense with the laborious process of defining relevant markets, proving dominance, and measuring market effects.

The DMA’s framers declare that the law aims to protect different legal interests than antitrust rules do. That is, rather than seeking to protect undistorted competition on any given market, the DMA look to ensure that markets where gatekeepers are present remain contestable and fair, independent from the actual, likely, or presumed effects of the conduct of a given gatekeeper.[7] Accordingly, the relevant legal basis for the DMA is found not in Article 103 of the Treaty on the Functioning of the European Union (TFEU), which is intended to implement antitrust rules pursuant to Articles 101 and 102 TFEU, but rather in Article 114 TFEU, covering “Common Rules on Competition, Taxation and Approximation of Laws.” Further, from an institutional-design perspective, the DMA opts for centralized implementation and enforcement at the EU level, rather than the traditional decentralized or parallel antitrust enforcement at the national level.

Because the intent of the DMA is to serve as a complementary regulatory scheme, traditional antitrust rules will remain applicable. However, those rules would not alleviate the obligations imposed on gatekeepers under the forthcoming DMA regulations and, particularly, efforts to make the DMA’s application uniform and effective.[8]

Despite claims that the DMA is not an instrument of competition law[9] and thus would not affect how antitrust rules apply in digital markets, the forthcoming regime appears to blur the line between regulation and antitrust by mixing their respective features and goals. Indeed, the DMA shares the same aims and protects the same legal interests as competition law.[10] Further, its list of prohibitions is effectively a synopsis of past and ongoing antitrust cases.[11] Therefore, the proposal can be described as a sector-specific competition law,[12] or a shift toward a more regulatory approach to competition law—one that is designed to allow assessments to be made more quickly and through a more simplified process.[13]

Acknowledging the continuum between competition law and the DMA, the European Competition Network (ECN) and some EU member states (self-anointed “friends of an effective DMA”) have proposed empowering national competition authorities (NCAs) to enforce DMA obligations.[14] Under this approach, while the European Commission would remain primarily responsible for enforcing the DMA and would have sole jurisdiction for designating gatekeepers or granting exemptions, NCAs would be permitted to enforce the DMA’s obligations and to use investigative and monitoring powers at their own initiative. According to supporters of this approach, the concurrent competence of the Commission and NCAs is needed to avoid the risks of conflicting decisions or remedies that would undermine the effectiveness and coherence of both the DMA and antitrust law (and, ultimately, the integrity of the internal market.)[15]

These risks have been heightened by the fact that Germany (one of the “friends of an effective DMA”) subsequently empowered its NCA, the Bundeskartellamt, to intervene at an early stage in cases where it finds that competition is threatened by large digital companies—in essence, granting the agency a regulatory tool that is functionally equivalent to the DMA.[16] Further, several member states are preparing to apply national rules on relative market power and economic dependence to large digital platforms, with the goal of correcting perceived imbalances of bargaining power between online platforms and business users.[17] As a result of these intersections among the DMA, national and European antitrust rules, and national laws on superior bargaining power, a digital platform may be subject to cumulative proceedings for the very same conduct, facing risks of double (or even triple and quadruple) jeopardy.[18]

The aim of this paper is to guide the reader through the jungle of potentially overlapping rules that will affect European digital markets in the post-DMA world. It attempts to demonstrate that, despite significant concerns about both the DMA’s content and its rationale, full centralization of its enforcement at EU level will likely be needed to reduce fragmentation and ensure harmonized implementation of the rules. Frictions with competition law would be further confined by narrowing the DMA’s scope to ecosystem-related issues, thereby limiting its application to the few large platforms that are able to orchestrate an ecosystem.

The paper is structured as follows. Section II analyzes the intersection between the DMA and competition law. Section III examines the DMA’s enforcement structure and the solutions advanced to safeguard cooperation and coordination with member states. Section IV illustrates the arguments supporting full centralization of DMA enforcement and the need to narrow its scope. Section V concludes.

Read the full white paper here.

[1] Proposal for a Regulation of the European Parliament and of the Council on Contestable and Fair Markets on the Digital Sector (Digital Markets Act) – General Approach, Council of the European Union (Nov. 16, 2021), available at https://data.consilium.europa.eu/doc/document/ST-13801-2021-INIT/en/pdf.

[2] Amendments Adopted on the Proposal for a Regulation of the European Parliament and of the Council on Contestable and Fair Markets in the Digital Sector (Digital Markets Act), European Parliament (Dec. 15, 2021), https://www.europarl.europa.eu/doceo/document/TA-9-2021-12-15_EN.html.

[3] Proposal for a Regulation on Contestable and Fair Markets in the Digital Sector (Digital Markets Act), European Commission (Dec. 15, 2020), available at https://eur-lex.europa.eu/legal-content/EN/TXT/PDF/?uri=CELEX:52020PC0842&from=en.

[4] Ibid., Recital 5.

[5] Ibid.

[6] Ibid.

[7] Ibid., Recital 10.

[8] Ibid., Recital 9 and Article 1(5).

[9] Margrethe Vestager, Competition in a Digital Age, speech to the European Internet Forum (Mar. 17, 2021), https://ec.europa.eu/commission/commissioners/2019-2024/vestager/announcements/competition-digital-age_en.

[10] Heike Schweitzer, The Art to Make Gatekeeper Positions Contestable and the Challenge to Know What Is Fair: A Discussion of the Digital Markets Act Proposal, 3 ZEuP 503 (Jun. 11, 2021).

[11] Cristina Caffarra and Fiona Scott Morton, The European Commission Digital Markets Act: A Translation, Vox EU (Jan. 5, 2021), https://voxeu.org/article/european-commission-digital-markets-act-translation.

[12] Nicolas Petit, The Proposed Digital Markets Act (DMA): A Legal and Policy Review, 12 J. Eur. Compet. Law Pract 529 (May 11, 2021).

[13] Marco Cappai and Giuseppe Colangelo, Taming Digital Gatekeepers: The More Regulatory Approach to Antitrust Law, 41 Comput. Law Secur. Rev. 1 (Apr. 9, 2021).

[14] How National Competition Agencies Can Strengthen the DMA, European Competition Network (Jun. 22, 2021), available at https://ec.europa.eu/competition/ecn/DMA_joint_EU_NCAs_paper_21.06.2021.pdf; Strengthening the Digital Markets Act and Its Enforcement, German Federal Ministry for Economic Affairs and Energy, French Ministére de l’Économie, les Finance et de la Relance, Dutch Ministry of Economic Affairs and Climate Policy, (May 27, 2021), available at https://www.bmwi.de/Redaktion/DE/Downloads/XYZ/zweites-gemeinsames-positionspapier-der-friends-of-an-effective-digital-markets-act.pdf?__blob=publicationFile&v=4.

[15] European Competition Network, supra note 14, 6-7.

[16] See Section 19a of the GWB Digitalization Act (Jan. 18, 2021), https://www.bundesrat.de/SharedDocs/beratungsvorgaenge/2021/0001-0100/0038-21.html.

[17] See, e.g., German GWB Digitalization Act, supra note 16; See, also, Belgian Royal Decree of 31 July 2020 Amending Books I and IV of the Code of Economic Law as Concerns the Abuse of Economic Dependence, Belgian Official Gazette (Jul. 19, 2020), http://www.ejustice.just.fgov.be/cgi_loi/change_lg.pl?language=fr&la=F&cn=2019040453&table_name=loi.

[18] Marco Cappai and Giuseppe Colangelo, A Unified Test for the European Ne Bis in Idem Principle: The Case Study of Digital Markets Regulation, SSRN working paper (Oct. 27, 2021), https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3951088.

Executive Summary

Worried that competition between firms is lessening, many economists and policymakers have called for a return to the more aggressive competition policies of the 1960s and 1970s and for the breakup or nationalization of large business, such as tech platforms. We argue, on the contrary, that changes in inter-firm competition are significantly driven by the increasing importance of intangible capital: assets like R&D, brands, software, and organizational development. This has several implications:

• It implies that simply dialing up the intensity of competition policy is the wrong response to the growing gap between leaders and laggards; and
• It raises the importance of competition for consumers’ attention.

Finally, we argue that there is a different aspect of the word “competition” that is affected by the shift to intangible capital: competition between individuals. An intangible-rich economy will see an increase in wasteful signaling. Mitigating this rat race should be a policy priority for educators and governments.

Read the full white paper here.

INTRODUCTION

Over the past 20 years, the ways we pay for goods and services have undergone a revolution. There has been a dramatic shift away from the use of cash and checks and toward the use of payment cards. More recently, this shift has also migrated toward the use of online payments, mobile payments and, to a smaller degree, cryptocurrency. These shifts, demonstrated in Figure 1, have been driven by innovations in payments technologies that have made them quicker, more convenient, more secure, and less costly for both consumers and merchants. And it has continued over the past two years, despite the global COVID-19 pandemic; indeed, the pandemic accelerated the shift to contactless and online payments.1F[1]

At the same time, governments have intervened in the operation of payment systems in various ways. As we have documented previously, these regulations have typically slowed the shift toward more innovative, quicker, more convenient payment systems, while also reducing other benefits and harming, in particular, poorer consumers and smaller merchants.2F[2]

Figure 1: Shares of Non-Cash Payments in the US by Transaction Volume, 2000 – 2020

Source: Authors’ calculations based on data from the Federal Reserve Payment Studies

This literature review revisits and builds upon our previous assessments, incorporating data, analyses, and insights from more recent research. The review is organized as follows. Section I provides a brief overview of the theory of two-sided markets as it pertains to payment-card networks, the role of the interchange fee as a balancing mechanism, and the theory of the optimal interchange fee. Section II describes some of the major ways that jurisdictions have capped interchange fees and posits some hypotheses regarding the likely effects of such caps. Sections III to VIII consider evidence for and against those hypotheses. Section IX concludes.

I.  PAYMENT NETWORKS, TWO-SIDED MARKETS, AND INTERCHANGE FEES

Early card payments involved only two parties: the merchant and the consumer. The “card” (a metal plate) enabled merchants to maintain a record of credit provided to regular customers, who would then settle-up at the end of the month.3F[3] In the 1950s, first Diners Club and then American Express established “three-party” systems, which enabled consumers to use the same card at multiple different merchants.4F[4] In a three-party system, the card issuer pays merchants directly and bills and collects from cardholders directly.5F[5]

During the 1960s, several organizations developed “four-party” systems. As their designation suggests, these systems, typified by those operated by Mastercard and Visa, have four main parties: issuer, consumer, merchant, and acquirer. The issuer contracts with the consumer, providing the card, issuing bills, etc. The acquirer contracts with the merchant, making payment. The rules of the system are set by the network operator, which also facilitates settlement between issuer and acquirer, and monitors for fraud and other abuse.6F[6]

A.   Two-Sided Markets

One of the main challenges faced by any payment system is to persuade both merchants and consumers of its value, regardless of whether it is a three-party or four-party payment card system or some other payment mechanism such as cash, check, or cryptocurrency. If too few merchants accept a particular form of payment, consumers will have little reason to hold it and issuers will have little incentive to issue it. Likewise, if too few consumers hold a card, merchants will have little reason to accept it. Conceptually, economists describe such situations as “two-sided markets”: consumers are on one side, merchants on the other, and the payment system acts as the platform facilitating interactions between them.7F[7] While payment cards are a prominent example of a two-sided market, there are many others, including: newspapers, shopping malls, social-networking sites, and search engines. Indeed, the rise of the Internet has made two-sided markets practically ubiquitous.

All platform operators that facilitate two-sided markets face essentially the same challenge: how to incentivize participation on each side to maximize the joint net benefits of the platform to all participants—and to allocate costs accordingly.8F[8] Thus, the platform operator can be expected to set the respective prices charged to participants on each side of the market to achieve this maximand.9F[9] If the operator sets the price too high for some consumers, they will be unwilling to use the platform; similarly, if the operator sets the price too high for some merchants, they will not be willing to use the platform.

This balancing is not static, but evolves over time in light of changes in costs, benefits, and preferences. Thus, whereas checks were once ubiquitous, now they are rarely used in retail payments, in large part because of their inconvenience and the risk to the merchant that a check will not clear. Similarly, while grocery stores and fast-food chains traditionally eschewed payment cards, their declining costs and increased speed and convenience facilitated faster throughput of the retail experience. Change may come from exogenous forces as well—for example, the COVID-19 pandemic led to a dramatic surge in the use of electronic payments and payment cards, as consumers and merchants sought to avoid physical cash and face-to-face interactions.

Nonetheless, platforms must cover the system operating costs, which arise from the joint interaction of the two sets of participants, mediated by the platform. In the case of payment networks, marginal costs are incurred only if boththe consumer and the merchant choose to interact through the platform.10F[10] Because the costs of operating the network arise from the joint interaction of participants on both sides, there is no “natural” way to allocate them.11F[11] Ultimately, the costs of operating the platform must be covered either by merchants or by consumers, or by some combination of the two.12F[12]

In many cases, this involves one side paying more of the costs than the other side. Consider newspapers, for example, where advertisers are on one side of the market and consumers on the other. Newspapers publish stories to attract readers. While some of the costs of operating a newspaper may come from reader’s payments, often much and sometimes all of the costs are covered by advertisers. Participants on one side of the market, the advertisers, thus effectively subsidize the production of content that is of interest to participants on the other side of the market, the readers. Likewise, search engines return searches that are of interest to consumers while also displaying advertisements (that may be related to the search or to some other activity the searcher has been undertaking), with the revenue from the advertisements funding the (non-advertising related) search returns.

In addition to these cross-subsidies across the two sides of the market (from merchants to consumers), there are also often cross-subsidies among the participants on one side of the market, such as between business and consumer users of a platform service. The Portable Document Format standard (PDF) is a good example: Adobe supplies the PDF-reader software for free, while charging businesses that want to produce PDFs.

In general, the costs of operating the platform tend to fall on the party who is least sensitive to such costs—in the language of economics, the party with the lower price elasticity. In the case of newspapers and search engines, that is advertisers. In the case of payment cards, it is merchants.13F[13]

Thus, it is predictable that the costs of payment-card transactions would gravitate to merchants to cover much of the cost of each transaction. Through this transaction fee, merchants often pay not only all the costs of operating the network but also effectively subsidize participation by consumers, e.g., through rewards programs.  Consumers may also pay an annual fee, although this is rare today, except for cards that offer rewards such as airline miles for which the annual fee defrays the costs of operating the rewards program.

B.   Interchange Fees

In three-party networks, the merchants’ transaction fee is charged directly by the network operator. In four-party systems, merchants pay acquirers a “merchant discount” (known as the “merchant service charge” in some jurisdictions), which includes the acquirer’s processing costs and the “interchange fee.” The interchange fee is a charge made by the network operator, the bulk of which is paid to the issuer (in the form of a deduction from the amount sent by the issuer to the acquirer when settling the transaction).

The various three- and four-party payment networks have been engaged in a decades-long process of dynamic competition in which each has sought—and continues to seek—to discover how to maximize value to their networks of merchants and consumers. This has involved considerable investment in innovative products, including more effective ways to encourage participation, as well as the identification and prevention of fraud and theft. As one of us previously noted:

[Compared to in-house providers of credit,] card issuers have developed the capacity to assess and price risk more accurately, giving them an increased ability both to take on more risk and to allocate the cost of risk within the system. For example, because they draw from a wider array of retailers, card-holders and locations, general credit card issuers can develop more-sophisticated (and less-costly) systems for anticipating and preventing fraudulent practices, reducing the risk of default by particular consumers, or protecting consumers against identity theft. In light of the massive volume of transactions processed and the number of consumers in the system, it has become feasible for card issuers to take increasingly-sophisticated measures to minimize (or appropriately charge for) the risk of non-payment by cardholders. Likewise, this huge database of information has enabled issuers to learn how to attack fraud through effective devices like password authorization, additional digits for card number verification, and special protections for on-line sales. Card issuers deploy extraordinarily-complicated neural networks and intelligent computer systems to detect changing patterns of fraud in real-time. Very few of these protections would be cost-feasible for department store chains (much less supermarkets, small appliance, hardware, or convenience stores), and large-scale card issuers are able to extend affordable credit to a much wider population and to do so much more efficiently. Indeed, these protections represent a traditional economy of scale, the benefits of which redound to both consumers and merchants.14F[14]

It has also involved experimentation with different levels of transaction fees. The early three-party schemes charged a transaction fee of as much as 7%.15F[15] Competition and innovation (including, especially, innovation in measures to reduce delinquency, fraud, and theft) drove those rates down. For U.S. credit cards, the merchant discount rate currently ranges from about 1.5% to 3%.16F[16] Rates vary by industry (for example, lower-margin industries, such as fast food and grocery stores, often have lower rates) and card-not-present transactions typically have higher interchange-fee rates because of higher incidences of fraud associated with them. (The specific rate depends on numerous factors, including the type of merchant and the type of arrangement the merchant has with the acquirer.)17F[17]

The majority of the merchant discount is the interchange fee remitted to the issuing bank, which in the United States ranges from about 1.5% to 3.5% (except for debit cards, subject to the caps discussed below) and averages approximately 2.2%.18F[18] Interchange-fee revenue covers many of the costs of operating the system, such as attracting new customers, card issuance, customer service, and fraud prevention and resolution, as well as such benefits as rewards, fraud protection, and car-rental insurance. Moreover, these services are offered for free today (no annual fee) or even at a negative price, such as when rewards are provided. Finance charges on revolving balances also generate substantial revenue, much of which covers the costs of underwriting, servicing, and charge-offs on credit balances. Revenue provided by interchange fees has risen substantially over time as a percentage of revenues from operations, increasing from 10% in 1990 to more than 20% in 2010, primarily mirroring a decline in the percentage of revenues generated by finance charges and the fact that transaction use (where the card balance is paid in full every month) grew more rapidly than revolving use.19F[19]

C.   What Does Economic Theory Say About Optimal Interchange Fees?

Beginning with William Baxter’s seminal 1983 paper, a rich theoretical literature has sought to understand various aspects of how payment networks operate. This literature has uncovered numerous valuable insights, including:

• The interchange fee exists as a default in open-network schemes, at least in part, because of the high costs of negotiating and enforcing many bilateral agreements between banks.20F[20]
• The interchange fee is set by payment-network operators, thereby internalizing the external costs that would arise if individual issuing banks set their own fees, as individual banks would have an incentive to set their fees excessively high to try to maximize their own revenue without regard to the impact on the value of the system as a whole.21F[21]
• As noted, the optimal allocation of costs between merchants and consumers depends on the price elasticity of the participants on each side of the market.22F[22] And since merchants tend to be less price sensitive, they generally bear more of the cost (via the merchant discount rate, which includes the interchange fee and merchant-acquirer fees), though consumers often also bear some of the cost (for example, in the form of annual fees, which tend to be applied to cards targeted at less price-sensitive consumers).

These factors, in combination with others, mean that the “optimal” interchange fee is indeterminate a priori and varies from place to place and from time to time. The European Commission has noted similarly:

The Commission does not dispute in general that payment systems are characterised by indirect network externalities and that in theory a revenue transfer between issuing and acquiring banks may help optimise the utility of the network to its users. However, whether a collectively fixed interchange fee should flow from acquirers to issuers or vice versa, and at which level it should be set cannot be determined in a general manner by economic theory alone, as theories always rely on assumptions that may not sufficiently reflect market reality.23F[23]

The U.S. Supreme Court put it more succinctly:

To optimize sales, the network must find the balance of pricing that encourages the greatest number of matches between cardholders and merchants.24F[24]

Given the theoretical indeterminacy of the optimal interchange fee, economists have cautioned against intervention unless there is evidence of a significant market failure.25F[25]

II. INTERCHANGE-FEE PRICE CONTROLS AND THEIR EFFECTS

In the three decades since Denmark introduced the first caps on interchange fees in 1990, dozens of jurisdictions have followed suit.26F[26] This section summarizes, in A through D, some of the most significant interchange-fee caps that have been introduced in various jurisdictions, namely Spain, Australia, the United States, and the European Union. It then offers some hypotheses regarding the likely effects of such caps.

A. Spain

Spain’s government twice imposed caps on interchange fees through “agreements” between merchant associations and payment-card schemes.27F[27] The first of these was initiated by the competition authority in April 1999, with a cap of 3.5% coming into effect in July of that year and falling in increments of 0.25% per year until it reached 2.75% in July 2002. This first agreement expired in July 2003.

The second cap was introduced in December 2005 by the Ministry of Finance, Industry, Tourism and Trade, again though an “agreement” between merchant associations and card schemes. The caps came into effect in January 2006 and lasted for five years. These caps were much tighter than those in the previous agreement, as can be seen in the schedule in Table I.

Table 1: Maximum Interchange Fees in Spain Under the ‘Agreement’ Signed in 2005

	2006		2007		2008		2009-2010
Euros (€)	Credit (%)	Debit (€)	Credit (%)	Debit (€)	Credit (%)	Debit (€)	Credit (%)	Debit (€)
0-100 mil.	1.40	0.53	1.30	0.47	1.10	0.40	0.79	0.35
100-500 mil.	1.05	0.36	0.84	0.29	0.63	0.25	0.53	0.21
> 500 mil.	0.66	0.27	0.66	0.25	0.54	0.21	0.45	0.18

Source: Iranzo, et al., 2012, at 14

B.   Australia

In 2002, Australia’s bank regulator, the Reserve Bank of Australia (RBA), introduced a series of regulations affecting the processing of credit-card transactions, including limits on interchange fees charged by four-party card networks. The RBA also prohibited card networks from enforcing restrictions on surcharges.

Under the rules, which came into force in 2003, interchange fees were not subject to an absolute cap but rather were subject to an effective cap set using a “cost-based framework.” Under the framework, a “cost-based measure” was calculated as the total of the eligible costs for each scheme in the preceding year, divided by the value of transactions processed by the scheme during the period. Meanwhile, eligible costs were limited to:

(i) issuers’ costs incurred principally in processing credit card transactions, including the costs of receiving, verifying, reconciling and settling such transactions;

(ii) issuers’ costs incurred principally in respect of fraud and fraud prevention in connection with credit card transactions;

(iii) issuers’ costs incurred principally in providing authorization of credit card transactions; and

(iv) issuers’ costs incurred in funding the interest-free period on credit card transactions, calculated using the average of the cash rate published by the Reserve Bank of Australia over the three financial years prior to the date by which the cost-based benchmark must be calculated.28F[28]

Notably, these eligible costs did not include any cross-subsidies, such as for rewards programs.

In a media release issued Aug. 27, 2002, the RBA noted that “average interchange fees in Australia are expected to fall from around 0.95 per cent of the value of each credit card transaction at present to around 0.55–0.6 per cent in the second half of 2003 – a reduction of around 40 per cent.”29F[29]

The cost-based framework meant that different card schemes were subject to somewhat different effective caps. In February 2005, the RBA launched an inquiry to assess the potential to introduce uniform caps. Following the inquiry, the RBA introduced a common standard in November 2005, under which all cards would be subject to the same caps.30F[30]Under the standard, the caps were calculated based on the weighted average costs across the covered card networks, using the same eligible-cost framework as above.31F[31]

Three-party cards were not originally subject to the regulation. In 2005, the RBA considered but rejected the possibility of regulating three-party cards that had partnered with banks.32F[32]

In February 2006, the RBA changed the rules regarding interchange fees for the domestic debit system—EFTPOS, for “electronic funds transfer at point of sale”—introducing both a floor and a ceiling on EFTPOS transactions.33F[33] At the same time, it prohibited Visa and Mastercard from enforcing their “honor all cards” rules with respect to debit transactions; i.e., if a merchant accepted Visa (Mastercard) credit, then Visa (Mastercard) could no longer require the merchant to accept Visa (Mastercard) debit.

On Sept. 29, 2006, the RBA set a common benchmark (average interchange fee) for Visa and Mastercard of A$0.5 per transaction, to apply for the three years following Nov. 1, 2006.34F[34] In December 2008, the RBA decided to waive the three-year review, so the A$0.5 per-transaction benchmark continued to apply.35F[35]

In 2016, the RBA changed “Standard No. 1,” under which interchange and other credit-card fees were determined.36F[36] Among other things, it set a hard cap of 0.8% per transaction and a maximum average rate of 0.5% over the course of a “reference period.” It also specified that issuers were prohibited from receiving “net compensation” in relation to credit-card transactions.

C.   The United States

In 2010, the United States passed legislation to impose caps on interchange fees on debit-card transactions made using cards issued by banks with more than $10 billion in assets. The caps, imposed under Section 1075 of the Dodd-Frank Wall Street Reform and Consumer Protection Act—a section titled “Reasonable Fees and Rules for Payment Card Transactions,” but better known as the “Durbin amendment” after its sponsor, Sen. Richard Durbin (D-Ill.)—required the Federal Reserve Board to set maximum interchange fees for covered banks that are “reasonable and proportional to the cost incurred by the issuer with respect to the transaction.”37F[37] On July 20, 2011, the Fed Board issued its ruling, known as Regulation II, which capped fees for covered transactions at $0.21 plus 0.05% of the transaction value.38F[38] In a separate rulemaking, the Fed Board determined that an additional $0.01 per transaction may also be charged to defray the cost of investments related to identifying and preventing fraud and suspicious transactions.39F[39]

Regulation II came into effect in October 2011 and the consequences were both immediate and dramatic: average interchange fees for covered debit transactions fell from $0.51 to $0.24, while those for exempt transactions remained constant, at $0.44.40F[40] As can be seen in Figure 2, average interchange fees have remained at roughly the same level since October 2011, averaging $0.23 for covered transactions and $0.43 for exempt transactions.

Figure 2: Average Debit Card Interchange Fees, US, $

Source: Federal Reserve Board

D.   The European Union

Regulations implemented by individual member states notwithstanding, the EU’s actions on interchange fees had, up to 2015, primarily concerned cross-border transactions. For example, the European Commission issued a 2007 decision forcing Mastercard to repeal its fallback interchange fees for cross-border transactions within the European Economic Area.41F[41]

In April 2015, the European Union introduced the Interchange Fee Regulation (IFR), which capped interchange fees for consumer debit cards at 0.2% and for consumer credit cards at 0.3% of the transaction value.42F[42] As a regulation, the caps applied throughout the EU with direct effect; i.e., there was no need for member states to implement enabling legislation and no room for interpretation (although member states were responsible for establishing enforcement authorities). The caps came into force Dec. 9, 2015.43F[43]

The IFR contained numerous other provisions, including:

• mandatory unbundling of card schemes and processors;44F[44]
• a requirement that issuers offer consumers “co-badging” options, where these exist;45F[45]
• a prohibition on the “honor all cards rule” for cards that are not covered by the fee caps;46F[46] and
• prohibitions on restrictions on steering by merchants.47F[47]

Since Dec. 9, 2015, the default interchange fees on covered cards issued by Visa and Mastercard have been 0.2% for debit and 0.3% for credit. Some national card schemes have continued to offer lower interchange fees.48F[48] By contrast, interchange fees for (unregulated) corporate cards currently range from 0.3% to 1.9%.49F[49]

E.   Some Hypotheses Regarding the Effects of Interchange-Fee Caps

Based on the analysis in Section I, we may posit several hypotheses regarding the likely effects of imposing price controls on interchange fees. These hypotheses are spelled out here and assessed in subsequent sections.

First, since interchange-fee revenue is remitted to issuers, a cap on interchange fees would be expected to reduce revenue to issuers. This is our first hypothesis:

H1: Interchange-fee caps reduce (interchange-derived) revenue to issuers.

Second, interchange-fee price controls will reduce the revenues available to fund card operations but issuers will still need to cover a range of costs. As a result, to the extent possible, issuers will respond to the revenue losses caused by price controls by reducing costs directly associated with cards, such as rewards, and also seek alternative sources of revenue, such as by raising fees on other elements of the card contract (such as annual fees) or by increasing the finance charge on revolving balances:

H2: Issuers subject to the interchange-fee cap will reduce benefits associated with cards subject to the cap and recover lost revenue by raising other fees.

But issuers will not reduce the benefits associated with unaffected cards. Indeed, to the extent that an issuer also offers unaffected cards, it may seek to steer consumers toward those unaffected cards. This gives us our third hypothesis:

H3: Issuers that offer both cards that are subject to a price cap and those that are not will seek to steer consumers toward the uncapped cards.

The increase in costs and reduction in benefits associated with cards subject to interchange-fee caps will lead consumers to shift to other payment methods, including to cards that are not subject to the caps, as well as to cash:

H4: Interchange-fee caps will cause consumers to switch to payment methods not subject to the caps.

Evidence from a range of industries suggests that, when input costs change, businesses rarely pass through all of the change to their consumers in the form of higher or lower prices.50F[50] Moreover, the proportion of changes to input prices that are passed through tends to be asymmetric, with reductions in input prices being passed through at a lower rate than increases in such prices.51F[51] Applying this to changes associated with caps on interchange fees gives us our fifth hypothesis:

H5: Merchants will, on average, pass through reductions in costs associated with reduced interchange fees at a rate considerably less than 100%, but issuers will pass through associated cost increases at a higher rate.

By reducing per-transaction card-related revenue to issuers, the interchange-fee cap reduces issuers’ incentive to invest in innovations that might otherwise lead to increased transactions:

H6: Interchange-fee caps will result in reduced investments in innovative technology.

Through a combination of issuers charging higher fees and offering lower rewards, and limited passthrough by merchants, interchange-fee caps have a particularly adverse effect on lower-income consumers:

H7: Through this combinations of factors, interchange-fee caps will adversely affect financial inclusion.

III. HAVE INTERCHANGE-FEE CAPS REDUCED ISSUER REVENUE?

Our first hypothesis is that, since interchange-fee revenue is remitted to issuers, a cap on interchange fees would be expected to reduce revenue to issuers. Studies looking at the effect of caps in Spain, the United States, and the EU generally confirm this effect.

An econometric analysis of Spain’s series of interchange-fee regulations by Valverde, et al., suggested that the mandatory reduction in interchange fees in Spain increased merchant acceptance and contributed to an increase in the volume of transactions.52F[52] However, the analysis by Valverde, et al., only covered the period 1997-2007, so missed most of the more aggressive period of interchange-fee caps (2008-2010). Iranzo, et al., found that, over the five-year period 2006-2010, total bank revenues from interchange fees fell by €3.3 billion.53F[53]

Estimates suggest that the Durbin amendment initially reduced annual interchange-fee revenue for covered banks by between $4.1 and $8 billion.54F[54] Since then, the volume and value of debit-card transactions has continued to grow and, but for the Durbin amendment, would have grown more rapidly (see below). The result is that, relative to the revenue that covered banks otherwise would have realized, the annual lost interchange-fee revenue may have been as much as $8.9 billion in 2012, growing to $14 billion in 2019.55F[55] Extrapolating to 2021, this implies that over the 10 years since it was first implemented, the Durbin amendment may have caused issuing banks to lose $120 billion or more.

In a 2020 study, Edgar Dunn & Co. found that interchange-fee revenue for debit and credit combined fell by 43% following the introduction of the IFR in the EU.56F[56]

IV. HAVE ISSUERS SUBJECT TO INTERCHANGE-FEE CAPS REDUCED REWARDS AND INCREASED OTHER FEES?

Our second hypothesis is that issuers faced with declining revenue from interchange fees will reduce rewards and increase other fees. While responses vary from jurisdiction to jurisdiction, in general, the evidence supports this hypothesis.

Survey evidence reported by Iranzo, et al., suggests that issuers in Spain reduced the rewards available from cards following the reduction in interchange fees.57F[57] From 2008 to 2010, issuing banks increased interest rates from an average of 3% above the European Central Bank (ECB) base rate in 2005 to 4.6% above base.58F[58] As a result, income from interest payments was nearly 80% higher from 2006 to 2010 than in 2005, representing a total incremental increase in income from interest over the period of about €2.6 billion (although this could be an overstatement, since we are only comparing to revenue in 2005). At the same time, average annual fees on credit cards rose by 50% from €22.94 to €34.39, generating incremental revenue over the period of €1.7 billion. And average annual fees for debit cards rose by 56%, from €11.12 to €17.30 per card, increasing revenue by another €0.6 billion. In other words, incremental interest payments and annual fees made up at least a large fraction of what was lost from the interchange-fee caps.

The Reserve Bank of Australia sought to make credit cards less attractive in order to encourage consumers to use the allegedly more “socially efficient” EFTPOS debit system. The data show that it was successful in this endeavor. Between 2003 and 2011, the average spend required to obtain a A$100 shopping voucher through use of a four-party credit card issued by the four largest banks in Australia went from A$12,400 to A$18,400.59F[59] In addition, issuers introduced caps on the total number of rewards that could be earned in a given period.60F[60] This turns the conventional rewards card model on its head: instead of creating incentives to use the rewards card more to achieve specific additional benefits, Australian credit-card issuers now incentivize rewards-card holders to switch cards when they reach the cap.

Australian issuers also increased the fees they charged on four-party credit cards. Between 2002 (the year before the regulation came into effect) and 2004, the annual fee on a “standard” rewards credit card went from A$61 to A$85 (an increase of about 40%). Over the same period, the fee on a “gold” rewards card rose from A$98 to A$128 (a 30% increase). By 2017, although the structure of rewards-card programs had changed somewhat, the average fee on rewards cards (of various kinds) had risen to A$199—significantly higher, even taking inflation into account, than the fee on a gold card in 2002.61F[61]

In sum, consumers in Australia now pay vastly more for their rewards credit cards and receive considerably fewer rewards for each dollar they spend. As a result, consumers have switched from using credit cards to using debit cards. The volume of debit transactions has increased at a much faster rate than the volume of credit-card transactions. Meanwhile, the value of debit-card transactions has gradually been catching up to the value of credit-card transactions.

U.S. banks covered by the Durbin amendment eliminated or reduced card-rewards programs on debit cards.62F[62]Data from the Federal Reserve Board suggest that rewards on debit cards issued by covered banks averaged approximately $0.05 per transaction in 2009 but fell to about $0.02 per transaction afterward.63F[63]

At the same time, covered banks increased the minimum balance required to qualify for free checking accounts and narrowed the types of account that qualified. In 2008, the average minimum deposit required to maintain a free checking account was $109. By 2012, that figure had soared to $723 and has remained around that level since.64F[64]Manuszak and Wozniak estimated that, following the Durbin amendment, the average minimum balance necessary to quality for a basic (non-interest-bearing) free checking account increased by more than $400, while for interest-bearing checking accounts it had increased by nearly $1,700.65F[65]

Meanwhile, the monthly cost to maintain a checking account had already doubled during the second half of 2010, as banks prepared for the revenue effects of the Durbin amendment.66F[66]

According to data from Bankrate, between 2011 and 2013, the proportion of free checking accounts offered by covered banks fell by half, from 76% to 38%, and remained at about the same level until last year, when they rose to 48% (still considerably lower than the level in 2010).67F[67] Meanwhile, using a proprietary data set, Mukharlyamov and Sarin found that the share of free basic checking accounts held by customers at covered banks fell by about half, from 61% to 28%, following the introduction of the Durbin amendment.68F[68]

Studies show that increased account fees have replaced between 40% and 90% of the reduction in revenue from debit-interchange fees.69F[69]

Turning to the EU, Table II shows that issuing banks made up for the losses on interchange-fee revenue following the introduction of the IFR through a combination of increased interest, late-payment fees, and overdraft fees, which rose by nearly 38%, and from increased international-transaction fees, which rose by 22%.70F[70] However, against expectations(i.e., contrary to our hypothesis) average annual fees for both credit and debit fell by an average of 5.6% between 2014 and 2018. Nonetheless, because of the increases in other fees, overall bank revenue rose by 3.7% following the introduction of the IFR.

Table II: Fees charged by issuing banks in the EU, 2014 and 2018

Fees	2014			2018
	Credit	Debit	Combined	Credit	Debit	Combined	Change
Annual Fees	9501	14620	24121	8741	14036	22777	-5.57%
Interest, late & overdraft	13469	8728	22197	15250	15352	30602	37.87%
International transaction	292	985	1277	382	1180	1562	22.32%
Interchange	4343	7527	11870	2258	4474	6732	-43.29%
Total	27605	31860	59465	26631	35042	61673	3.71%

Source: Authors based on Edgar Dunn & Co.71F[71]

V. HAVE ISSUERS ENCOURAGED CONSUMERS TO USE CARDS THAT ARE NOT COVERED BY INTERCHANGE-FEE PRICE CAPS?

The presence of price controls on interchange fees means that consumers will pay higher prices but obtain lower benefits than when fees are set by market forces. Payment devices for which the interchange fee is set by market forces rather than government fiat, by contrast, will offer more net value for consumers. As a result, it would be predicted that consumers will migrate away from those cards that are subject to interchange-fee price controls to those that are unregulated.

This observation leads to two predictions. First (our third hypothesis), where possible, issuers will encourage consumers to switch to cards (and other forms of payment) that are not subject to the caps. Second (our fourth hypothesis), consumers will switch to payment methods not subject to the caps. Again, the evidence from several jurisdictions supports these hypotheses.

Shortly after Australia’s interchange-fee caps for four-party cards came into force in 2003, two banks introduced three-party credit cards with annual fees and rewards similar to those that previously existed on their four-party cards.72F[72]In addition, several issuers introduced packages of two similar premium rewards cards, one that operates on a four-party network and one that operates on a three-party network.73F[73] The reason these “companion cards” existed is that far fewer merchants accept three-party cards than four-party cards; with both cards, consumers could use the higher-earning three-party card where it is accepted and the lower-earning four-party card elsewhere.

Unsurprisingly, the market share of three-party cards, while still relatively small, increased considerably following the 2003 regulations. By volume of transactions, three-party cards increased from about 10% in 2002 to about 16% in 2013 (a 60% increase). By value of transactions, they increased their market share from about 15% in 2002 to more than 20% in 2013 (a 33% increase).

In October 2015, the RBA designated American Express Companion Cards as a “payment system.”74F[74] In its May 2016 Review of Card Payments Regulation, the RBA announced that these cards would—effective July 1, 2017—be subject to the same interchange-fee caps as other cards so designated.75F[75] Following introduction of the caps, companion cards were discontinued and the market share by volume of three-party cards fell back to 7%. But, as seen in Figure 3, it has since risen slightly to about 8%.76F[76] By value, three-party cards’ market share of transactions also fell steeply after mid-2017, but it has risen again since then. In addition, the number of three-party cards held by consumers increased by 36,000 in the year to October 2021, while the number of four-party cards fell by 750,000.77F[77]

These trends suggest that consumers who qualify for three-party cards increased adoption, presumably in no small part because the cards are not subject to interchange-fee caps and therefore are able to offer attractive rewards. Likewise, consumers with both three-party and four-party cards tend to use the former where possible, in order to obtain the larger rewards.

Figure 3: Market Share of Four-Party Cards (Mastercard and Visa, left axis) and Three-Party Cards (American Express and Diners Club, right axis) by Volume of Transactions in Australia, 2002-2021

Source: Reserve Bank of Australia78F[78]

Figure 4: Market Share of Four-Party Cards (Mastercard and Visa, left axis) and Three-Party Cards (American Express and Diners Club, right axis) by Value of Transactions in Australia, 2002-2021

 Source: Reserve Bank of Australia79F[79]

Because the Durbin amendment did not apply to credit cards, U.S. issuers had no reason to reduce rewards on those cards. In response, many consumers switched from using debit to using credit for transactional purposes, paying their balance each month in full rather than revolving.80F[80] As a discussion paper by the Federal Reserve Bank of Philadelphia Consumer Finance Institute put it, this switch from debit to credit was driven by a combination of “regulatory changes in the debit space that limited interchange, making debit rewards less financially viable for depository institutions and a change in preferences by both card issuers and consumers for more and richer rewards as incentives for using a particular form of payment.”81F[81]

Consumers without credit cards—predominantly those with poor credit records—were not positioned to switch from debit to credit. They thus replaced their debit usage with increased use of cash and checks.82F[82] To the extent that debit-only consumers tend to be lower income, the cuts in debit-card rewards and the rising cost of checking accounts, both brought on by the Durbin amendment, have had regressive effects, while higher-income consumers switched to credit cards whose rewards rates remain unaffected.

The EU’s dramatic caps on interchange fees appear to have led issuers to limit card issuance in general and to shift consumers away from credit and toward debit. In a study funded by the European Commission, Ernst & Young and Copenhagen Economics (EY/CE) found that the number of credit cards issued in the EU fell following the introduction of the IFR, while the number of debit cards grew only modestly.83F[83] They observe:

We find that the total number of issued consumer cards in the EU as reported by the schemes changed moderately between 2015 and 2017. The total number of consumer debit cards in the EU grew by 3% and 4% respectively in 2016 and in 2017 … The number of credit cards declined by 1%. The larger decline in the number of credit cards could be related to the large reduction in interchange fees for credit card transactions to meet the interchange fee cap.84F[84]

The EY/CE study also found that the proportion of corporate cards, as a share of all cards, increased in several member states. The shift was particularly dramatic in Ireland, where the proportion of corporate cards went from about 6% in 2015 (before the IFR) to about 14% in 2017.85F[85] Moreover, EY/CE found that the number of commercial-card transactions increased by a weighted average of 12.4 million per member state.86F[86] Since corporate cards are not subject to the IFR, it would seem that issuers sought, in part, to recoup losses by switching eligible consumers to the use of corporate cards.

VI. DO MERCHANTS PASS ON REDUCTIONS IN COSTS ASSOCIATED WITH REDUCED INTERCHANGE FEES?

One of the primary justifications made by legislators and regulators for introducing caps on interchange fees is that these fees drive up the merchant discount rate, which in turn leads merchants to raise the prices of goods and services offered to consumers. For example, an oft-cited 2010 study estimated that low-income U.S. households pay on average an additional $21 per year on goods and services due to credit-card interchange fees and card rewards.87F[87] To arrive at this number, the study assumed that interchange fees are fully passed through to consumers in the form of higher prices.

However, the only reason the overwhelming majority of merchants would pass on all cost savings is that they believe their consumers would otherwise buy their goods and/or services elsewhere—in other words, it assumes the merchants have zero market power. While that might be true for retailers of gasoline and diesel, it is unlikely to be true for retailers of less homogenous goods and services.88F[88] Most retailers have some degree of pricing power because of the idiosyncratic nature of the products they sell and the way in which they are sold (which can include factors such as location, expertise of the staff, the way products are displayed, among others).89F[89] Indeed, the evidence for cost passthrough in general indicates that, while cost increases are often passed through at rates of around 90%, cost decreases are typically passed through at much lower rates.90F[90]

In addition, a 100% pass-through rate would imply that merchants would reap none of the benefit from a reduction in interchange fees. As such, they would have no incentive to lobby for such a change. Since some merchants do lobby for reduced interchange fees, it seems highly likely that they enjoy some degree of pricing power in their given markets and are thus able to pass through less than 100% of the reduction in interchange fees.91F[91]

Hence our fifth hypothesis, that merchants will pass through relatively little of the cost savings from reduced interchange fees, whereas card issuers will pass on much of the losses they incur. As noted above, most of card issuers’ revenue losses are passed on to consumers in the form of higher prices and lower benefits. Merchants, by contrast, demonstrate much lower pass-through rates of cost savings. Moreover, cost savings tend to be passed through at a lower rate to consumers than cost increases.

Iranzo, et al., found that, while acquirers in Spain reduced the merchant service charge (MSC)—the European equivalent of the merchant discount rate—they did so by less than the reduction in the interchange fee.92F[92] Over five years, the difference totaled more than €440 million. However, this may partially reflect unavoidable per-transaction costs, so the net amount received by acquiring banks may have been less.

While acquirers passed through to merchants much (but not all) of the reduction in interchange fees, merchants appear to have passed though little, if any, of their savings to consumers. The survey by Iranzo, et al., of merchant groups asked directly whether their members had reduced prices of goods and services or improved the quality of offerings in response to the interchange-fee caps. Perhaps surprisingly, the response was uniformly negative: merchants had not passed on their savings in any way.93F[93]

Since the additional costs of using payment cards were directly paid by consumers in the form of higher annual fees and interest payments, and since consumers seem not to have obtained any benefit in the form of lower prices or higher-quality goods and services, it seems clear that the main effect of the interchange-fee cap in Spain was to transfer wealth from consumers to big-box merchants.

The Reserve Bank of Australia’s hope that its interchange-fee regulations would save consumers money does not appear to have materialized. In the 19 years since the regulations went into effect, there has been no substantive evidence that merchants have passed savings on to consumers. Even assuming that some savings have been passed on, it is very unlikely that the savings for the average consumer have been anywhere close to the costs imposed on them by the regulation in the form of increased fees on, and reduced benefit from, rewards cards.

The average merchant service charge on four-party cards fell from about 1.4% prior to the interchange-fee cap to around 0.7% in 2020—a reduction of 0.7 percentage points, or 50% in relative terms (see Figure 5).94F[94] Since Mastercard and Visa credit-card transactions make up about a quarter of retail transactions, the overall effect on a typical merchant would have been a reduction in per-transaction costs of about 0.17%. Thus, even if these savings had been fully passed through, the average consumer would have seen prices fall by less than 0.2%. But because the consumer price index in Australia has risen, on average, by 2.3% annually since 2002,95F[95] it would be difficult to discern such an effect.

Figure 5: Total Merchant Service Fees for Various Payment Networks in Australia, 2003—2020

Source: Reserve Bank of Australia

Evidence from Surcharging In Australia

Following the 2002 regulations, merchants have also been able to impose surcharges on payments made with credit cards in Australia. However, most merchants have not introduced surcharges.96F[96] A 2010 survey commissioned by the RBA found that less than 10% of large merchants and less than 5% of small merchants imposed surcharges in 2005. By 2010, however, around a quarter of small merchants and nearly half of all large merchants were surcharging.97F[97]  A subsequent survey conducted by the RBA in 2013 found that just under 7% of Mastercard and Visa credit-card payments were subject to surcharges.98F[98]

Merchants who introduced surcharging did so as a form of price discrimination in instances where consumers have an inelastic demand for using cards (such as online purchases, airline tickets, or hotel rooms) or where merchants are not constrained by repeat purchasers (such as on travel and restaurants).99F[99] So, while surcharging is not at all uniform, it was highly prevalent in areas where price discrimination and rent-seeking are profitable.

Moreover, until caps came into force (see below), merchants almost ubiquitously imposed surcharges at rates that were considerably higher than the cost of acceptance: the RBA survey found that average surcharges in 2013 were 1.5% of payment value. Because average merchant-service fees were 0.8%, this represents a nearly 90% markup over the MSC.100F[100]

Clearly, merchants had been using surcharges as a means of price discrimination against consumers who used credit cards. As a result, many consumers making purchases with credit cards at merchants who imposed surcharges were hit with a triple whammy: higher annual fees, fewer rewards, and higher prices.

In 2016, the Australian Parliament passed legislation prohibiting merchants from applying “excessive” surcharges and capping any surcharge at an amount that reflects “the cost of using the payment methods for which they are charged.” 101F[101] The rules, laid out in more detail in a standard issued by the RBA, came into force in September 2016 for large merchants and September 2017 for smaller merchants.102F[102] The surcharge rules are enforced by the Australian Competition and Consumer Protection Commission, which has issued several fines to companies found to be in violation.103F[103] In response, merchants have reduced their surcharges, as the RBA notes:

“Merchants have generally responded appropriately to the new framework coming into effect. Most notably, the airline industry moved from fixed-fee surcharges (see above) to a percentage-based surcharge with a fee cap. Prior to the reform, a $100 domestic flight would have attracted a surcharge of up to $8.50 for debit and credit cards alike. Following the reform, the same flight would attract a maximum surcharge of $1.30 for credit cards and $0.60 for debit cards (calculated as a percentage of the cost of the airfare).”104F[104]

More precise estimates of the extent of passthrough have been possible for the United States. Using proprietary data from banks and one of the card networks, Mukharlyamov and Sarin estimated that merchants passed through “at most” 28% of their debit-card savings to consumers. Meanwhile, banks passed through 42% of their interchange-fee revenue losses to consumers (with most of those losses passed on to lower-income consumers who pay higher bank fees). They estimate that the net result of this was a $4 billion transfer to merchants, of which $3.2 billion came directly from banks and $0.8 billion from consumers, who paid $2.3 billion in higher checking fees but received only $1.5 billion in lower retail prices.

Wang, Schwartz, and Mitchell, found that, while some merchants received reductions in the merchant discount rate they paid, others actually saw their debit-card acceptance costs increase.105F[105] They found an asymmetric response: merchants who saw their prices increase usually passed those increased costs onto their customers, while very few of those who saw their debit costs decrease passed those costs onto customers. This suggests that there was very little passthrough of savings by merchants (certainly far less than 100%) and that, if there was any substantial passthrough at all, it was greatly delayed.

The story is similar for the EU, where EY/CE estimated that, on average, acquirers passed through about 45% of the reduction in interchange fees (see Table III). Meanwhile, EY/CE also estimated that the average EU-wide pass-through rate by merchants of the lower MSC is 66%.106F[106] Taking these two pass-through rates together, on average, EU consumers would have received about 30% of the reduction in interchange fees in the form of lower costs and/or improved quality.107F[107]

To some extent, these averages, mask differences among member states. For example, EY/CE found that a typical Polish household would save only €1.53 per year, while a household in Italy would save as much as €12.42. However, as Edgar Dunn & Co. noted, banks more than made up for the lost interchange-fee revenue in the form of higher charges. Since those charges were mainly in the form of increases in interest and late-payment fees, it is likely that they would have mainly affected lower-income households that are more reliant on the use of overdrafts and credit cards for short-term credit.108F[108]

Table III: Change in Acquirer’s Margin following the IFR

Fees	Change (EUR million)	Effect on acquirers
Change MSC	-1,200	Revenue loss
– Change interchange fee	-2,680	Cost saving
– Change acquirer scheme fee	280	Cost increase
= Change acquiring margin	1,200	Margin increase

Source: EY/CE study

VII. HAVE INTERCHANGE-FEE CAPS RESULTED IN REDUCED INVESTMENTS IN INNOVATIVE TECHNOLOGY?

Our sixth hypothesis is that interchange-fee caps will reduce issuing banks’ incentive to invest in innovative technology. The evidence seems to support this hypothesis.

Consider, for example, investments in innovation by EFTPOS, Australia’s domestic debit-card network. In its 2015 Review of Card Payments Regulation, the Reserve Bank of Australia noted that “the greater functionality of the international scheme cards (eftpos is still working to develop online and contactless functionality) has also contributed to the shift in market shares.”109F[109] As of September 2021, EFTPOS still had not developed online functionality.110F[110]

The consequence of this lack of innovation for EFTPOS has been little short of catastrophic. Far from becoming the dominant low-cost transaction network for Australia’s consumers, as the RBA had hoped, the proportion of debit transactions undertaken via EFTPOS declined from 82% in 2009 to 40% in 2020, while the value of debit transactions declined by an even larger amount, as can be seen in figure 6.

Figure 6: Value and Volume of Debit Transactions in Australia, 2009- 2020

Source: Blockley, 2021

As Lance Blockley noted in a submission to the Australian Competition and Consumer Commission (ACCC) regarding the proposed merger of EFTPOS, NPP Australia, and BPay, made on behalf of the merging parties:

[T]he debit card market was originally the sole domain of eftpos (from its launch in 1983), until Visa Debit was launched in the market in the 1990’s. Scheme Debit (the term covering Visa Debit and Mastercard Debit) was initially focused on the smaller financial institutions, but eventually became adopted by the four major banks as, unlike eftpos, they were able to be used both online and overseas (making them attractive to consumers); in addition, they generated positive interchange revenue for issuers (at a time when eftpos caused a negative interchange outflow for issuers), making them attractive to the banks.

As Scheme Debit became more issued and used in the market, and consequently eftpos’s share of the total debit card transaction volume fell, the number of eftpos transactions still grew in total due to the strong increases in total debit card volumes. The state of declining share but increasing absolute volume slowed severely for eftpos, however, when contactless debit card transactions became widely adopted, as shown in the charts below.111F[111]

In 2018, the Australian government’s Productivity Commission issued a report about competition in the banking sector that noted with concern the decline in EFTPOS’ market share, asserting:

The decline in market share of eftpos is concerning because eftpos is considered highly price-competitive…. This suggests that competition is stymied by other forces, such as distortions in who pays the costs of card payments.112F[112]

Yet, as noted above, EFTPOS’ own consultant acknowledges that EFTPOS previously had a monopoly in the debit market and that its declining market share is largely a result of consumer preferences for Visa and Mastercard debit that, unlike EFTPOS, enable users both to pay online and to use contactless.

Clearly, EFTPOS’ failure to develop online and contactless functionality was a major drawback and is almost certainly related to its relative lack of investment in innovation, which in turn is a function of its bargain-basement cost model. Logically, the decline in EFTPOS’ market share was driven by competition, not stymied by it.

It should also be noted that Australia has one of the highest penetrations of contactless cards in the world (something that likely helped facilitate continued in-person shopping in the country during the COVID-19 pandemic). This has been thanks almost entirely to Visa and Mastercard, which, unlike EFTPOS, have been able to recover the costs of investment in innovations in other markets.

As with Australia, there is some tentative evidence that domestic networks in the EU may be underinvesting in innovation due to the very low interchange rates mandated by the IFR. As reported by EY/CE, the market share of domestic payment networks declined slightly between 2014 and 2016, with the market share of domestic debit-card schemes falling from 81% to 78% on average and the market share of domestic credit-card schemes falling from 72% to 68% on average (in those member states that have such schemes).113F[113]

In the United States, the Durbin amendment has substantially diminished covered banks’ ability to recover investments in payment innovations that would benefit both merchants and consumers. As a result, some potentially valuable new technologies likely have not been developed and implemented as quickly as they otherwise might have been.114F[114]

Jonathan Reinisch argues that, by dampening investment in innovation, the Durbin amendment has delayed the otherwise “inevitable” shift to mobile payments.115F[115] These reduced investments are driven by two factors: first, the redistribution of costs, as discussed above, which constrains card issuers’ ability to recover investments in innovation, and second, the fee cap itself, which has the effect of reducing the potential savings that merchants might realize from switching to a payment system with lower fees.

Meanwhile, Zywicki predicted that “the arbitrary definition that the Durbin amendment provides for debit cards will promote regulatory arbitrage, as competitors seek to gerrymander products out of the amendment’s net.”116F[116] And that is precisely what we have seen, with various financial-technology (or “fintech”) companies issuing debit cards by partnering with banks whose assets fall below $10 billion.117F[117] The higher interchange-fee revenue from each customer means that, relative to covered banks, fintechs are able to:

• invest more in R&D, thereby improving their offerings and exacerbating their differential from covered banks;
• spend more acquiring new customers in other ways, including by offering reward programs; and
• serve customer niches not currently served by covered banks.

While this arbitrage has allowed various new payment technologies to be rolled out, the pace of roll-out is arguably slower than would be the case if larger banks could also partner with fintechs.

VIII. HAVE INTERCHANGE-FEE CAPS ADVERSELY AFFECTED FINANCIAL INCLUSION?

Our final hypothesis is that caps on interchange fees would adversely affect financial inclusion. Support for this hypothesis comes in part from the observation that more of the costs of regulation tend to be borne by lower-income consumers, which has been noted above for Australia and the EU.

More direct evidence, however, comes from the United States, where caps on debit-card interchange fees have adversely affected access to banking. To see this, it helps to go back a few years before the introduction of the Durbin amendment. As can be seen in Figure 7, during the 2000s, debit-card transaction volume exploded: in 2000, debit was responsible for about one-fifth as many transactions as checks, but by 2007, it had overtaken checks in popularity.

Figure 7: Volume of Payments by Type, U.S., 2000-2018 (billions)

Source: Federal Reserve Payments Studies

This rise in debit-card volume coincided with a significant reduction in the average cost of checking accounts, an increase in the number of “free” checking accounts, and a reduction in the minimum balance required to qualify for free checking. According to a U.S. Government Accountability Office report, between 2000 and 2006, the proportion of banks offering free checking accounts doubled from 30% to 60%, while monthly maintenance fees fell from $6.81 to $5.41.118F[118]

This is not surprising, as the two things are related in a virtuous circle: rising debit-card use increased revenue from interchange fees on debit, which in turn enabled banks to reduce checking-account charges, which increased the number of account holders, which increased debit-card use. At the same time, debit-card purchases mainly replaced purchases made using checks, which are expensive to process. By dramatically expanding access to free checking and eliminating monthly maintenance fees, the introduction and rapid adoption of debit cards dramatically expanded financial inclusion for many consumers who traditionally could not afford a checking account.

As noted in Section IV, the introduction of caps on debit-card interchange fees reversed this trend: debit-card interchange-fee revenue fell, so banks were less able to subsidize free checking accounts. Mukaharlyamov and Sarin estimated that, in the absence of the Durbin amendment, the proportion of free checking accounts would have risen to 66%. Since the actual number was 29%, they infer that Durbin caused a reduction of 37 percentage points.119F[119]

Unsurprisingly, the loss of access to free checking accounts and the increase in monthly fees both disproportionately affect lower-income consumers. Mukharlyamov and Sarin found that:

[O]ver 70 percent of consumers in the lowest income quintile (annual household income of $22,500 or less) bear higher account fees, since they fall below the average post-Durbin account minimum required to avoid a monthly maintenance fee ($1,400). In contrast, only 5 percent of consumers in the highest income quintile (household income of $157,000 or more) keep balances falling below this threshold.120F[120]

By reducing the availability of free checking and increasing the cost of bank accounts, the Durbin amendment likely resulted in many lower-income people becoming “underbanked” and some people to become unbanked (exiting the banking system altogether). Federal Deposit Insurance Corp. (FDIC) data show that, between 2009 and 2011, the number of unbanked households rose by one million and the number of underbanked households rose by three million. As a result, many of those unbanked and underbanked individuals will have relied on more costly alternatives, such as money orders, prepaid cards, and check-cashing services.121F[121] Plausibly, some of that increase was due to the lower availability of free checking and the increased cost of bank accounts, as banks readied for the implementation of the Durbin amendment.

The proportion of unbanked households has gradually fallen since peaking in 2011; in 2019, it was lower than in 2009. But it is noteworthy that among those consumers who remain unbanked, whereas 12.7% of respondents gave “minimum balance requirements too high” as a reason for not having a bank account in 2009, by 2019, 29% of households gave that as the main reason for not having a bank account and 48.9% identified this obstacle as a factor.122F[122]In addition, 34% identified “bank account fees too high” as a reason, with 7.3% citing it as the “main” reason.123F[123] This suggests that, as the economy has grown stronger, the cost increases brought about in response to the Durbin amendment have become a more significant cause of households being unbanked.

Mukharlyamov and Sarin looked specifically at the difference in the proportion of unbanked individuals between 2011 and 2013 and found an 81% increase in the percentage of unbanked consumers who said that high account fees was their main reason for not having a bank account.124F[124] Further support for the hypothesis that the Durbin amendment has caused households to become unbanked comes from the finding that residents of states with the highest proportion of deposits at banks with assets in excess of $10 billion were most likely to attribute their unbanked status to high fees.125F[125]The growth in the recently unbanked was also highest in states with more affected banks, where the increase in account fees is most pronounced.

These findings contrast dramatically with the hypothetical claims made by Schuh, et al., who claimed that lower-income consumers could save as much as $21 per year in lower retail costs from interchange-fee price controls.126F[126] To make that claim, they assumed full passthrough by merchants of the reduction in acceptance costs, while implicitly assuming no passthrough of revenue losses by banks (such as through reduced access to free checking, higher required minimum balances, and higher bank fees). Even if Shuh, et al., were correct about the nominal savings by lower-income consumers, these are clearly massively outweighed by higher bank fees, which are estimated in the range of $115 to $172 per year.127F[127]

IX. Conclusion

Overall, it is clear that imposing price caps on interchange fees has had many pernicious effects and that, in contrast to the claims of those who support them, they have done far more harm than good.

First and foremost, interchange-fee caps have harmed the very people they were supposed to help. Wherever they have been implemented, they have resulted in lower revenue for issuing banks, which have responded by increasing fees for consumers, either on bank accounts, on credit cards, or both. These fee increases have in general been highly regressive, hurting those with lower incomes the most.

Second, in some cases, such as with the Durbin amendment in the United States, the higher fees have resulted in many people becoming unbanked.

Third, in nearly all cases, issuers have reduced the rewards on payment cards. But those with higher incomes and/or better credit records have often been able to switch to alternative payment cards that are not subject to the caps. So, the reductions in rewards have mainly harmed the poor and those with poor credit records.

Fourth, the rate at which merchants have passed through reductions in costs associated with lower interchange fees (in the form of lower-priced goods) has been less than the rate at which banks have passed through losses in fee revenue—in the form of higher-priced accounts, cards and services, and reductions in rewards. As such, consumers have lost out on net.

Fifth, interchange-fee caps have had somewhat predictable effects on modes of payment and hence on investments in those modes of payment. Thus, the Durbin amendment, which exclusively affected debit, led to a shift in payments toward credit and impeded investment in debit-related payment technologies (until fintech companies realized they could partner with exempt financial institutions). By contrast, the caps in Australia and the EU were effectively tighter for credit than for debit (since the difference between the interchange-fee rates before the caps and after the caps was greater for credit than debit), which led to shifts away from credit and toward debit. However, the caps were so low that investment in domestic debit systems seems to have been impaired, especially in Australia.

* Todd Zywicki is the George Mason University Foundation Professor of Law at the Antonin Scalia Law School, senior

fellow at the Cato Institute, and an academic affiliate of the International Center for Law & Economics (ICLE). Julian Morris is a senior scholar with ICLE. Geoffrey A. Manne is ICLE’s president and founder.

[1] See Developments in Noncash Payments for 2019 and 2020: Findings From the Federal Reserve Payments Study, Federal Reserve Board,(December 2021), available at https://www.federalreserve.gov/publications/files/developments-in-noncash-payments-for-2019-and-2020-20211222.pdf, and the various previous studies and associated data, https://www.federalreserve.gov/paymentsystems/frps_previous.htm.

[2] See Todd J. Zywicki, The Economics of Payment Card Interchange Fees and the Limits of Regulation, International Center for Law and Economics, ICLE Financial Regulatory Program White Paper Series, (Jun. 2, 2010), available at http://laweconcenter.org/images/articles/zywicki_interchange.pdf; Todd J. Zywicki, Geoffrey A. Manne, & Julian Morris, Unreasonable and Disproportionate: How the Durbin Amendment Harms Poorer Americans and Small Businesses, International Center for Law and Economics, (Apr.25, 2017); Todd J. Zywicki, Geoffrey A. Manne, & Julian Morris, Price Controls on Payment Card Interchange Fees: The U.S. Experience, George Mason Law & Economics Research Paper No. 14-18, (Jun. 6, 2014).

[3] Claire Tsosie, The History of the Credit Card, NerdWallet.com, (Mar 15, 2021), https://www.nerdwallet.com/article/credit-cards/history-credit-card; see also Jeremy Norman, The Charga-Plate, Precursor of the Credit Card, Circa 1935 to 1950, HistoryofInformation.com,https://www.historyofinformation.com/detail.php?id=1710.

[4] See Zywicki, supra note 2. Several banks also attempted to establish three-party cards during the 1950s. Most of these were unsuccessful. The exception was Bank Americard, which subsequently became a four-party system and eventually rebranded as Visa.

[5] The issuer may arrange separate underwriting. More recently, the processing of three-party card transactions are sub-contracted to other payment processors. But the fundamental three-party legal arrangements remain the same.

[6] For a more detailed explanation of the operation of payment card systems, see Zywicki, supra note 2, at 27-30.

[7] See Zywicki, supra note 2; see also Jean-Charles Rochet & Jean Tirole, Two-Sided Markets: A Progress Report, 37 Rand J. Econ. 645 (2006); As the U.S. Supreme Court wrote in Ohio v. American Express Co. (585 U.S. Slip Op, 2018, at 2):

By providing these services to cardholders and merchants, credit-card companies bring these parties together, and therefore operate what economists call a “two-sided platform.” As the name implies, a two-sided platform offers different products or services to two different groups who both depend on the platform to intermediate between them.”… For credit cards, that interaction is a transaction…. The key feature of transaction platforms is that they cannot make a sale to one side of the platform without simultaneously making a sale to the other.

[8] Bruno Jullien, Alessandro Pavan, & Marc Rysman, Two-Sided Markets, Pricing, and Network Effects, in Handbook of Industrial Organization (Vol. 4), 485-592, (2021).

[9] Thomas Eisenmann, Geoffrey Parker, & Marshall W. Van Alstyne, Strategies for Two-Sided Markets, Harv. Bus. Rev. (October 2006).

[10] See William F. Baxter, Bank Interchange of Transactional Paper: Legal and Economic Perspectives, 26 J. L. & Econ. 541 (1983).

[11] See Ronald H. Coase, The Problem of Social Cost, 3 J. L. & Econ. 1 (1960).

[12] See Zywicki, supra note 2, at 31-32.

[13] See Zywicki, supra note 2, at 33.

[14] Zywicki, supra note 2, at 10.

[15] Tsosie, supra note 3.

[16] See Frank Kehl, The Complete Guide to Merchant Account & Credit Card Processing Fees, Merchant Maverick, (Nov. 12, 2021). https://www.merchantmaverick.com/the-complete-guide-to-credit-card-processing-rates-and-fees. For a description of CNP transactions, see Ben Dwyer, Card Present vs. Card Not Present Transactions, CardFellow, (Apr. 6, 2020), https://www.cardfellow.com/blog/card-present-card-not-present-definition.

[17] Visa USA Interchange Reimbursement Fees, Visa Public, (Oct. 15, 2021), available at https://usa.visa.com/content/dam/VCOM/download/merchants/visa-usa-interchange-reimbursement-fees.pdf; Mastercard USA Interchange Rates, Helcim, https://www.helcim.com/mastercard-usa-interchange-rates.

[18] Id.

[19] See Thomas A. Durkin, Gregory Elliehausen, Michael E. Staten, & Todd J. Zywicki, Consumer Credit and the American Economy, (2014), at 347, Fig. 71.

[20] Eliana Garcés & Brent Lutes, Regulatory Intervention in Card Payment Systems: An Analysis of Regulatory Goals and Impact, working paper, (Sep. 21, 2018), https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3346472, at 8. As Garces and Lutes note: “Practically all open network schemes have set some default interchange fees that apply automatically when no bilateral agreement exists between banks. No widely adopted international scheme relies solely on bilateral negotiations for the interchange fee. This may be due to the excessive level of information complexity that a system of bilaterally negotiated fees would imply for merchants. To assess the cost of a card payment, the merchant would have to know not only the brand and type of the card used, but also the identity of the issuer. Additionally, given that most card systems impose an “honor all cards” rule on merchants, the absence of a common interchange fee may lead some issuing banks to impose high interchange fees for the cards that they issue and that the merchant is forced to accept. Although there are open network schemes that have operated without interchange fees, these are very rare and with limited regional scope.”

[21] Baxter, supra note 10, at 572-582.

[22] Zywicki, supra note 2, at 33; Jean-Charles Rochet and Jean Tirole, Platform Competition in Two-Sided Markets, 1 J. Europ. Econ. Assoc. 990-1029, (Jun 1, 2003), at 997; Julian Wright, Pricing in Debit and Credit Card Schemes, 80 Econ. Lett 305-309, (2003).

[23] Summary of Commission Decision of 19 December 2007 Relating to a Proceeding Under Article 81 of the EC Treaty and Article 53 of the EEA Agreement (Case COMP/34.579 — MasterCard, Case COMP/36.518 — EuroCommerce, Case COMP/38.580 —Commercial Cards), Official Journal of the European Union, (Nov. 6, 2009), available at https://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=OJ:C:2009:264:0008:0011:EN:PDF, at 19.

[24] Ohio v Amex, supra note 7, at 13.

[25] Jean-Charles Rochet & Jean Tirole, An Economic Analysis of the Determination of Interchange Fees in Payment Card Systems, 2 Rev. Netw. Econ. 69-79, (January 2003).

[26] For a fairly comprehensive overview of such caps, see Fumiko Hayashi & Jesse Leigh Maniff, Public Authority Involvement in Payment Card Markets: Various Countries, August 2020 Update, Federal Reserve Bank of Kansas City, (August 2020), available at https://www.kansascityfed.org/documents/6660/PublicAuthorityInvolvementPaymentCardMarkets_VariousCountries_August2020Update.pdf.

[27] Juan Iranzo, Pascual Fernández, Gustavo Matías, & Manuel Delgado, The Effects of the Mandatory Decrease of Interchange Fees in Spain, Munich Personal RePEc Archive, MPRA Paper No. 43097, (October 2012), available at https://mpra.ub.uni-muenchen.de/43097/1/MPRA_%20paper_43097.pdf.

[28] Standard No. 1 The Setting of (Wholesale) Interchange Fees, Reserve Bank of Australia, (2002). The same standard was issued to each four-party card issuer.

[29] Media Release: Reform of Credit Card Schemes in Australia, 2002-15, Reserve Bank of Australia, (Aug. 27, 2002), https://www.rba.gov.au/media-releases/2002/mr-02-15.html.

[30] Media Release: Payments System Board – November 2005-13, Reserve Bank of Australia, (Nov. 25, 2005), https://www.rba.gov.au/media-releases/2005/mr-05-13.html.

[31] Common Benchmark for the Setting of Credit Card Interchange Fees, Reserve Bank of Australia, (November 2005), https://www.rba.gov.au/payments-and-infrastructure/credit-cards/cc-fees-benchmark.

[32] Media Release: Payments System Reform, 2005-02, Reserve Bank of Australia, (Feb. 24, 2005), available at https://www.rba.gov.au/payments-and-infrastructure/credit-cards/cc-fees-benchmark/pdf/cc-fees-benchmark.pdf.

[33] Media Release: Payments System Reforms Number 2006-02, Reserve Bank of Australia, (Apr. 27, 2006), https://www.rba.gov.au/media-releases/2006/mr-06-02.html.

[34] Credit Card Benchmark Calculation, Reserve Bank of Australia, (2006), available at https://www.rba.gov.au/media-releases/2006/pdf/mr-06-08-benchmark-calc-credit-card.pdf.

[35] Payment Systems (Regulation) Act 1998 Variation of Interchange Fee Standards, Reserve Bank of Australia, (2008), available athttps://www.rba.gov.au/media-releases/2008/pdf/mr-08-28-vifs.pdf.

[36] Payment Systems (Regulation) Act 1998, Standard No. 1 of 2016, The Setting of Interchange Fees in the Designated Credit Card Schemes and Net Payments to Issuers, Reserve Bank of Australia, (May 26, 2016), available at https://www.rba.gov.au/payments-and-infrastructure/review-of-card-payments-regulation/pdf/standard-no-1-of-2016-credit-card-interchange-2016-05-26.pdf.

[37] H.R.4173 – Dodd-Frank Wall Street Reform and Consumer Protection Act, s.1075(a)(3).

[38] Debit Card Interchange Fees and Routing; Final Rule, 76 Fed. Reg. 43,393-43,475, (Jul. 20, 2011).

[39] The fraud-prevention measures were initially spelled out in an interim final rule published on the same day: Debit Card Interchange Fees and Routing; Interim Final Rule, 76 Fed. Reg. 43,477-43,488, (Jul. 20, 2011). They were subsequently incorporated into an amended final rule: 77 Fed. Reg. 46,258-46,282, (Aug. 3, 2012).

[40] See Zywicki, et al., Unreasonable and Disproportionate, supra note 2. See also Regulation II (Debit Card Interchange Fees And Routing), Board of Governors of the Federal Reserve System, (Jul. 18, 2019), https://www.federalreserve.gov/paymentsystems/regii-average-interchange-fee.htm.

[41] The decision related to a case originally brought in response to a 1992 complaint by the British Retail Consortium related to the interchange fees and network rules of Europay (now Mastercard) and Visa-branded cards.  See Non-Confidential Version of Commission Decision of 19 December 2007, COMP/34.579 MasterCard, COMP/36.518 EuroCommerce and COMP/38.580 Commercial Cards, The Commission of the European Communities, (Dec. 19, 2007), available at https://ec.europa.eu/competition/antitrust/cases/dec_docs/34579/34579_1889_2.pdf. For a summary, see Summary of Commission Decisions, supra note 23.

[42] Regulation (EU) 2015/751 of the European Parliament and of the Council of Apr. 29, 2015, on Interchange Fees for Card-Based Payment Transactions, Official Journal of the European Union, (Apr. 29, 2015), at 10-11, Articles 3 and 4. Note that these caps did not apply to corporate cards, as discussed below.

[43] Id., at 14, Article 18.

[44] Id. Article 7.

[45] Id. Article 8. Co-badging is defined as: “The inclusion of two or more payment brands or payment applications of the same brand on the same card-based payment instrument.” (Per Regulation (EU) 2015/751, Art. 2).

[46] Id. Article 10.

[47] Id. Article 11.

[48] Study on the Application of the Interchange Fee Regulation: Final Report, prepared by Ernst & Young and Copenhagen Economics (“EY/CE study”) for the European Commission, (2020), available at https://ec.europa.eu/competition/publications/reports/kd0120161enn.pdf, at 89.

[49] Intra-EEA – Intercountry Interchange Fees, Mastercard, (Jan. 22, 2021); Intra Europe EEA Multi-lateral Interchange Fees, Visa, (Oct 16, 2021).

[50] EY/CE study, supra note 48, at 169-170; Cost Pass-Through: Theory, Measurement, and Potential Policy Implications, prepared by RBB Economics (“RBB Economics study”) for the U.K. Office of Fair Trading, (February 2014).

[51] Sam Peltzman, Prices Rise Faster Than They Fall, 108 J Polit. Econ. 466-502, (June 2000).

[52] Santiago Carbó Valverde, Sujit Chakravorti, & Francisco Rodríguez Fernández, The Role of Interchange Fees in Two-Sided Markets: An Empirical Investigation on Payment Cards, 98 Rev. Econ. Stat. 367–381, (Apr. 21, 2016), https://doi.org/10.1162/REST_a_00502.

[53] Iranzo, et al., supra note 27.

[54] See Benjamin S. Kay, Mark D. Manuszak, & Cindy M. Vojtech, Competition and Complementarities in Retail Banking: Evidence From Debit Card Interchange Regulation, 34 J. Fin. Intermediation 91, 92 (2018); see also Vladimir Mukharlyamov & Natasha Sarin, Price Regulation in Two-Sided Markets: Empirical Evidence From Debit Cards, working paper, (December 2019) (estimating $5.5 billion annual revenue loss to banks from interchange-fee reductions); Bradley G. Hubbard, The Durbin Amendment, Two-Sided Markets, and Wealth Transfers: An Examination of Unintended Consequences Three Years Later, working paper, (May 20, 2013), http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2285105, at 20 (estimating annual revenue loss of $6.6 billion to $8 billion from the Durbin amendment).

[55] See Out of Balance: How the Durbin Amendment Has Failed to Meet Its Promise, Electronic Payments Coalition, (December 2018), available at https://www.electronicpaymentscoalition.org/wp-content/uploads/2018/12/EPC.DurbinStudiesPaper.pdf.

[56] Interchange Fee Regulation Impact Study, Edgar Dunn & Co., (April 2020), available at https://edgardunn.com/wp-content/uploads/2020/04/Interchange-Fee-Regulation-Impact-Study-EDC-v1.1.pdf.

[57] Iranzo, et al., supra note 27, at 34-37.

[58] Id., at 27. See also marginal lending facility rates from European Central Bank, https://sdw.ecb.europa.eu/browse.do?node=9691107.

[59] Iris Chan, Sophia Chong, & Stephen Mitchell, The Personal Credit Card Market in Australia: Pricing Over the Past Decade, Reserve Bank of Australia Bulletin, (March 2012), available at https://www.rba.gov.au/publications/bulletin/2012/mar/pdf/bu-0312-7.pdf.

[60] Robert Stillman, William Bishop, Kyla Malcolm, & Nicole Hildebrandt, Regulatory Intervention in the Payment Card Industry by the Reserve Bank of Australia: Analysis of the Evidence, CRA International, (2008), at 16.

[61] Emily Perry & Christian Maruthiah, Banking Fees in Australia, Reserve Bank of Australia Bulletin, (June 2018), available athttps://www.rba.gov.au/publications/bulletin/2018/jun/pdf/banking-fees-in-australia.pdf, at 5.

[62] See Darryl E. Getter, Regulation of Debit Interchange Fees, Congressional Research Service, (May 16, 2017), at 8; see also Out of Balance: How the Durbin Amendment Has Failed to Meet Its Promises, Electronic Payments Coalition, (Dec. 2018), at 7. Eliminating rewards, such as cash-back on purchases, is functionally equivalent to a price increase.

[63] Kay, et al., supra note 54, at 99, citing 2009 Interchange Revenue, Covered Issuer Cost, and Covered Issuer and Merchant Fraud Loss Related to Debit Card Transactions, Federal Reserve Board, (2011), available at: www.federalreserve.gov/paymentsystems/files/debitfees_costs.pdf. 2011 Interchange Fee Revenue, Covered Issuer Costs, and Covered Issuer and Merchant Fraud Losses Related to Debit Card Transactions, Federal Reserve Board, (2013), available atwww.federalreserve.gov/paymentsystems/files/debitfees_costs_2011.pdf121913_2013_payment_study_release.pdf.

[64] Mukharlyamov & Sarin estimated a 21% increase in the monthly minimum needed to be eligible for free checking and to avoid having to pay monthly fees.

[65] Mark D. Manuszak & Krzysztof Wozniak, The Impact of Price Controls in Two-Sided Markets: Evidence From US Debit Card Interchange Fee Regulation, Board of Governors of the Federal Reserve System Finance and Economics Discussion Series, FEDS Working Paper No. 2017-074, (Jul. 10, 2017), at 5. Two recent surveys by consumer finance websites of the largest banks’ account terms also illustrate the high cost to those who no longer qualify for free checking accounts. A survey by Moneyrates.com reported an average monthly maintenance fee of $14.39 ($172.68 per year) on the accounts it surveyed. See Richard Barrington, How Much Are Bank Fees—The Latest MoneyRates Update, Moneyrates.com, (Aug. 20, 2020), https://www.money-rates.com/research-center/bank-fees. A study by another website reported an average monthly fee on the accounts reviewed of $9.60 ($115.20 per year) and the minimum balance necessary to waive the fee was $1,010. See Theresa Kim, Checking Account Fee Comparison at Top U.S. Banks, Mybanktracker.com, (Jul. 23, 2020), https://www.mybanktracker.com/news/checking-account-fee-comparison-top-10-us-banks. This survey includes one Internet bank that reports no monthly fee, with no minimum balance required. If that bank is excluded, the average monthly fee would be $10.61.

[66] Mukharlyamov & Sarin, supra note 54, at 2-4. They concluded that average monthly checking account fees increased from $3.07 to $5.92. Id., at 4.

[67] Matthew Goldberg, Survey: Free Checking Accounts on the Rise as Total ATM Fees Fall, Bankrate.com, (Oct. 20, 2021), https://www.bankrate.com/banking/checking/checking-account-survey.

[68] Mukharlyamov & Sarin, supra note 54, at 2-4.

[69] Mukharlyamov & Sarin, supra note 54, at 4, estimated that issuers lost approximately $5.5 billion in reduced interchange-fee income annually, of which they recouped about $2.3 billion in higher bank fees, or approximately 42% of lost revenue. Meanwhile, Kay, et al., supra note 54, at 99, estimated that banks increased other fees by about $4 billion in total, offsetting about 90% of the revenue reduction of approximately $4.1 billion in debit-interchange fees. Note that these studies did not seek to extrapolate the losses, so the proportions are based on non-extrapolated losses.

[70] Edgar Dunn & Co, supra note 56, at 23.

[71] Id.

[72] Chan, et al., supra note 59.

[73] Companion Cards Increase Credit Card Rewards, Mozo, (Dec. 8, 2009), https://mozo.com.au/credit-cards/articles/companion-cards-increase-credit-card-rewards/19502078.

[74] Designatton Under the Payment Systems (Regulation) Act 1998, Designation No 1 of 2015, Reserve Bank of Australia, (Oct. 18, 2015), available at https://www.rba.gov.au/media-releases/2015/pdf/mr-15-19-designation-2015-01-american-express-companion-card.pdf.

[75] Standard No. 1 of 2016, The Setting of Interchange Fees in the Designated Credit Card Schemes and Net Payments to Issuers, Reserve Bank of Australia, (May 26, 2016), amended version available at https://www.rba.gov.au/payments-and-infrastructure/review-of-card-payments-regulation/pdf/standard-no-1-of-2016-credit-card-interchange-2018-05-31.pdf.

[76] C2: Market Shares of Credit and Charge Card Schemes, Reserve Bank of Australia, (March 2019), https://www.rba.gov.au/statistics/tables/xls-disc/c02hist.xls.

[77] George Lekakis, Amex and Diners Claw Back Market Share, Banking Day, (Dec. 8, 2021), https://www.bankingday.com/amex-diners-claw-back-share.

[78] Statistics, Credit and Charge Cards – Market Shares of Card Schemes – C1.3, Reserve Bank of Australia, available at https://www.rba.gov.au/statistics/tables/xls/c01-3-hist.xlsx?v=2022-03-01-12-57-31.

[79] Id.

[80] See Zywicki, et al., Price Controls, supra note 2, at 18-22; see also Vladimir Mukharylyamov and Natasha Sarin, The Impact of the Durbin Amendment on Banks, Merchants, and Consumers, Faculty Scholarship at Penn Law, (January 2019), at 34-35; Tom Ankana, Consumer Payment Preferences and the Impact of Technology and Regulation: Insights From the Via Payment Panel Study, Federal Reserve Bank of Philadelphia Consumer Finance Institute, Discussion Paper DP 19-01, (February 2019).

[81] Ankana, supra note 80, at 11.

[82] See Sergei Koulayev, Marc Rysman, Scott Schuh, & Joanna Stavins, Explaining Adoption and Use of Payment Instruments by US Consumers, 47 Rand J. of Econ. 293 (2016). Kloulayev, et al., also note the same trends associated with higher and lower education levels.

[83] EY/CE study, supra note 48 at 135.

[84] Id.

[85] Id. at 139-140.

[86] Id. at 140-141.

[87] Scott Schuh, Oz Shy, & Joanna Stavins, Who Gains and Who Loses From Credit Card Payments? Theory and Calibrations, Federal Reserve Bank of Boston, Research Department Public Policy Discussion Paper No. 10–3, (August 2010).

[88] Among the rare exceptions are globally traded commodities like oil; changes in the price of crude oil are fully reflected in the price of diesel within four weeks. See Erwan Gautier, Magali Marx, & Paul Vertier, How do oil prices pass through to fuel prices?, Eco Notepad, (Oct. 14, 2021),https://blocnotesdeleco.banque-france.fr/en/blog-entry/how-do-oil-prices-pass-through-fuel-prices.

[89] Thomas Bittmann, Jens-Peter Loy, & Sven Anders, Product Differentiation and Cost Pass-Through: Industry-Wide Versus Firm-Specific Cost Shocks, 64 Aust J Agric Resour Econ. 1184–1209, (Sep. 6, 2020), https://doi.org/10.1111/1467-8489.12399.

[90] EY/CE study supra note 48, at 169-170; RBB Economics study, supra note 50.

[91] Ian Lee, Geoffrey A. Manne, Julian Morris, & Todd J. Zywicki, Credit Where It’s Due: How Payment Cards Benefit Canadian Merchants and Consumers, and How Regulation Can Harm Them, Macdonald-Laurier Institute, (2013), at 27.

[92] Iranzo, et al., supra note 27.

[93] Iranzo, et al., supra note 27 at 34-37.

[94] Kateryna Occhiutto, The Cost of Card Payments for Merchants, Reserve Bank of Australia Bulletin, (March 2020).

[95] Inflation Calculator, Reserve Bank of Australia, https://www.rba.gov.au/calculator/annualDecimal.html.

[96] This is consistent with theory. See Wright, supra note 22.

[97]Australian Merchant Acquiring and Cards Markets: Special Question Placement Report, prepared by East & Partners for the Reserve Bank of Australia, (December 2010).

[98] Crystal Ossolinski, Tai Lam, & David Emery, Changing Way We Pay: Trends in Consumer Payments, Reserve Bank of Australia, (June 2014), available at http://www.rba.gov.au/publications/rdp/2014/pdf/rdp2014-05.pdf.

[99] Todd J. Zywicki, Geoffrey A. Manne, & Kristian Stout, Behavioral Economics Goes to Court: The Fundamental Flaws in the Behavioral Law & Economics Arguments Against No-Surcharge Law, 82 Mo. L. Rev. 822, (2017).

[100] See also Stillman, et al., supra note 60, at 25–26, citing survey evidence from 2004–07 indicating that surcharges ranged from 15 to 81 basis points higher than the merchant-service charge.

[101] Competition and Consumer Amendment (Payment Surcharges) Act 2016 (Austl.).

[102] Payment Systems (Regulation) Act 1998 Standard No. 3 Of 2016, Scheme Rules Relating to Merchant Pricing for Credit Debit and Prepaid Card Transactions. Reserve Bank of Australia, (May 26, 2016), available at https://www.rba.gov.au/payments-and-infrastructure/review-of-card-payments-regulation/pdf/standard-no-3-of-2016-scheme-rules-relating-to-merchant-pricing-2016-05-26.pdf.

[103] ACCC Enforcement of Credit Card Surcharging Laws, Australian Federation of Travel Agents, available at https://www.afta.com.au/news-hub/accc-enforcement-of-credit-card-surcharging-laws.

[104] Cameron Dark, Chay Fisher, Kim McBey, & Ed Tellez, Payment Surcharges: Economics, Regulation and Enforcement, Reserve Bank of Australia Bulletin, (December 2018), available at https://www.rba.gov.au/publications/bulletin/2018/dec/pdf/payment-surcharges-economics-regulation-and-enforcement.pdf.

[105] Zhu Wang, Scarlett Schwartz, & Neil Mitchell, The Impact of the Durbin Amendment on Merchants: A Survey Study, 100(3) Econ. Q. 183, (2014) (3^rd Quarter). Some merchants saw their acceptance costs increase because, prior to Dodd-Frank’s price controls, some merchants, especially smaller merchants, had received discounts on acceptance costs. But the imposition of price ceilings also effectively created a price floor, leading some merchants to pay higher fees than previously.

[106] EY/CE study, supra note 48, at 170.

[107] 66% of 45% is 30%.

[108] The Household Finance and Consumption Survey: Results From the 2017 wave, European Central Bank, No. 36, (March 2020).

[109] Review of Card Payments Regulation Issues Paper, Reserve Bank of Australia, (March 2015), https://www.rba.gov.au/payments-and-infrastructure/review-of-card-payments-regulation/developments-card-payments-mkt.html.

[110] Olivia Gee, Debit Cards vs EFTPOS, Mozo, (Sep. 9, 2021), https://mozo.com.au/debit-cards/guides/debit-cards-vs-eftpos; EFTPOS’ 2021 annual report asserts that the system will be ready for ecommerce by May 2022. See EFTPOS Annual Financial Report, EFTPOS, (2021), available at https://www.eftposaustralia.com.au/sites/default/files/2021-11/ANNUAL%20REPORT%202021%20FINAL%20HR.pdf, at 8.

[111] Lance Sinclair Blockley, Expert Industry Opinion in Relation to the Application to the Australian Competition and Consumer Commission for Authorisation of the Proposed Amalgamation of BPay Group Pty Limited and BPay Pty Ltd, EFTPOS Payments Australia Limited and NPP Australia Limited, Australian Competition and Consumer Commission, (Mar. 18, 2021).

[112] Competition in the Australian Financial System, Inquiry Report, Productivity Commission, (2018), at 471.

[113] EY/CE study, supra note 48, at 61.

[114] Todd J. Zywicki, Durbin’s Innovation Killer: The Durbin Amendment Would Raise Costs for Consumers, Increase Fraud, and Kill Innovation, The American (Jun. 11, 2011), https://www.aei.org/articles/durbins-innovation-killer.

[115] Jonathon Reinisch, Swipe Freeze: How the Durbin Amendment Is Preventing Your Mobile Phone From Replacing Your Wallet, 63 DePaul L. Rev. 123, (2013), at 137.

[116] Zywicki, supra note 114.

[117] Tanay Jaipuria, Fintech and Regulatory Arbitrage: Understanding the The Durbin Amendment, Tanay’s Newsletter, (May 24, 2021), https://tanay.substack.com/p/fintech-and-regulatory-arbitrage.

[118] Bank Fees: Federal Banking Regulators Could Better Ensure That Consumers Have Required Disclosure Documents Prior to Opening Checking or Savings Accounts, U.S. Government Accountability Office, (January 2008), available at https://www.gao.gov/assets/gao-08-281.pdf, at 15.

[119] Mukharlyamov & Sarin, supra note 54, at 37.

[120] Id., at 30.

[121] See Zywicki, et al., Price Controls, supra note 2.

[122] 2009 FDIC National Survey of Unbanked and Underbanked Households, Federal Deposit Insurance Corporation, (2009), https://www.fdic.gov/analysis/household-survey/2009/index.html, at 25; How America Banks: Household Use of Banking and Financial Services, Federal Deposit Insurance Corporation, (2019), https://www.fdic.gov/analysis/household-survey, at 2.

[123] On the eve of Dodd-Frank’s enactment, the 2009 FDIC report found only 6.3% of consumers said “service charges too high” was a reason why they did not have a bank account (FDIC 2009, supra note 122). There were some changes in how the survey questions were asked over time that makes it difficult to compare precise figures, although the trend lines appear clear.

[124] Mukharlyamov & Sarin, supra note 54, at 30.

[125] Id.

[126] Schuh, et al., supra note 87.

[127] The authors’ analysis applies to credit cards, not debit cards. But oddly, they assume that the annual fee on credit cards would remain the same after imposing interchange-fee price controls. Experience in Australia and elsewhere reveals this assumption to be completely unfounded. See Chan, et al., supra note 59.

Executive Summary

A quarter-century since its enactment as part of the Communications Decency Act of 1996, a growing number of lawmakers have been seeking reforms to Section 230. In the 116^th Congress alone, 26 bills were introduced to modify the law’s scope or to repeal it altogether. Indeed, we have learned much in the last 25 years about where Section 230 has worked well and where it has not.

Although the current Section 230 reform debate popularly—and politically—revolves around when platforms should be forced to host certain content politically favored by one faction (i.e., conservative speech) or when they should be forced to remove certain content disfavored by another (i.e., alleged “misinformation” or hate speech), this paper does not discuss, nor even entertain, such reform proposals. Rather, such proposals are (and should be) legal non-starters under the First Amendment.

Indeed, such reforms are virtually certain to harm, not improve, social welfare: As frustrating as imperfect content moderation may be, state-directed speech codes are much worse. Moreover, the politicized focus on curbing legal and non-tortious speech undermines the promise of making any progress on legitimate issues: The real gains to social welfare will materialize from reforms that better align the incentives of online platforms with the social goal of deterring or mitigating illegal or tortious conduct.

Section 230 contains two major provisions: (1) that an online service provider will not be treated as the speaker or publisher of the content of a third party, and (2) that actions taken by an online service provider to moderate the content hosted by its services will not trigger liability. In essence, Section 230 has come to be seen as a broad immunity provision insulating online platforms from liability for virtually all harms caused by user-generated content hosted by their services, including when platforms might otherwise be deemed to be implicated because of the exercise of their editorial control over that content.

To the extent that the current legal regime permits social harms online that exceed concomitant benefits, it should be reformed to deter those harms if such reform can be accomplished at sufficiently low cost. The salient objection to Section 230 reform is not one of principle, but of practicality: are there effective reforms that would address the identified harms without destroying (or excessively damaging) the vibrant Internet ecosystem by imposing punishing, open-ended legal liability? We believe there are.

First and foremost, we believe that Section 230(c)(1)’s intermediary-liability protections for illegal or tortious conduct by third parties can and should be conditioned on taking reasonable steps to curb such conduct, subject to procedural constraints that will prevent a tide of unmeritorious litigation.

This basic principle is not without its strenuous and thoughtful detractors, of course. A common set of objections to Section 230 reform has grown out of legitimate concerns that the economic and speech gains that have accompanied the rise of the Internet over the last three decades would be undermined or reversed if Section 230’s liability shield were weakened. Our paper thus establishes a proper framework for evaluating online intermediary liability and evaluates the implications of the common objections to Section 230 reform within that context. Indeed, it is important to take those criticisms seriously, as they highlight many of the pitfalls that could attend imprudent reforms. We examine these criticisms both to find ways to incorporate them into an effective reform agenda, and to highlight where the criticisms themselves are flawed.

Our approach is rooted in the well-established law & economics analysis of liability rules and civil procedure, which we use to introduce a framework for understanding the tradeoffs faced by online platforms under differing legal standards with differing degrees of liability for the behavior and speech of third-party users. This analysis is bolstered by a discussion of common law and statutory antecedents that allow us to understand how courts and legislatures have been able to develop appropriate liability regimes for the behavior of third parties in different, but analogous, contexts. Ultimately, and drawing on this analysis, we describe the contours of our recommended duty-of-care standard, along with a set of necessary procedural reforms that would help to ensure that we retain as much of the value of user-generated content as possible, while encouraging platforms to better police illicit and tortious content on their services.

The Law & Economics of Online Intermediary Liability

An important goal of civil tort law is to align individual incentives with social welfare such that costly behavior is deterred and individuals are encouraged to take optimal levels of precaution against risks of injury. Not uncommonly, the law even holds intermediaries—persons or businesses that have a special relationship with offenders or victims—accountable when they are the least-cost avoider of harms, even when those harms result from the actions of third parties.

Against this background, the near-complete immunity granted to online platforms by Section 230 for harms caused by platform users is a departure from normal rules governing intermediary behavior. This immunity has certainly yielded benefits in the form of more user-generated online content and the ability of platforms to moderate without fear of liability. But it has also imposed costs to the extent that broad immunity fails to ensure that illegal and tortious conduct are optimally deterred online.

The crucial question for any proposed reform of Section 230 is whether it could pass a cost-benefit test—that is, whether it is likely to meaningfully reduce the incidence of unlawful or tortious online content while sufficiently addressing the objections to the modification of Section 230 immunity, such that its net benefits outweigh its net costs. In the context of both criminal and tort law generally, this balancing is sought through a mix of direct and collateral enforcement actions that, ideally, minimizes the total costs of misconduct and enforcement. Section 230, as it is currently construed, however, eschews entirely the possibility of collateral liability, foreclosing an important mechanism for properly adjusting the overall liability scheme.

But there is no sound reason to think this must be so. While many objections to Section 230 reform—that is, to the imposition of any amount of intermediary liability—are well-founded, they also frequently suffer from overstatement or unsupported suppositions about the magnitude of harm. At the same time, some of the expressed concerns are either simply misplaced or serve instead as arguments for broader civil-procedure reform (or decriminalization), rather than as defenses of the particularized immunity afforded by Section 230 itself.

Unfortunately, the usual course of discussion typically fails to acknowledge the tradeoffs that Section 230—and its reform—requires. These tradeoffs embody value judgments about the quantity and type of speech that should exist online, how individuals threatened by tortious and illegal conduct online should be protected, how injured parties should be made whole, and what role online platforms should have in helping to negotiate these tradeoffs. This paper’s overarching goal, even more important than any particular recommendation, is to make explicit what these tradeoffs entail.

Of central importance to the approach taken in this paper, our proposals presuppose a condition frequently elided by defenders of the Section 230 status quo, although we believe nearly all of them would agree with the assertion: that there is actual harm—violations of civil law and civil rights, violations of criminal law, and tortious conduct—that occurs on online platforms and that imposes real costs on individuals and society at-large. Our proposal proceeds on the assumption, in other words, that there are very real, concrete benefits that would result from demanding greater accountability from online intermediaries, even if that also leads to “collateral censorship” of some lawful speech.

It is necessary to understand that the baseline standard for speech and conduct—both online and offline—is not “anything goes,” but rather self-restraint enforced primarily by incentives for deterrence. Just as the law may deter some amount of speech, so too is speech deterred by fear of reprisal, threat of social sanction, and people’s baseline sense of morality. Some of this “lost” speech will be over-deterred, but one hopes that most deterred speech will be of the harmful or, at least, low-value sort (or else, the underlying laws and norms should be changed). Moreover, not even the most valuable speech is of infinite value, such that any change in a legal regime that results in relatively less speech can be deemed per se negative.

A proper evaluation of the merits of an intermediary-liability regime must therefore consider whether user liability alone is insufficient to deter bad actors, either because it is too costly to pursue remedies against users directly, or because the actions of platforms serve to make it less likely that harmful speech or conduct is deterred. The latter concern, in other words, is that intermediaries may—intentionally or not—facilitate harmful speech that would otherwise be deterred (self-censored) were it not for the operation of the platform.

Arguably, the incentives offered by each of the forces for self-restraint are weakened in the context of online platforms. Certainly everyone is familiar with the significantly weaker operation of social norms in the more attenuated and/or pseudonymous environment of online social interaction. While this environment facilitates more legal speech and conduct than in the offline world, it also facilitates more illegal and tortious speech and conduct. Similarly, fear of reprisal (i.e., self-help) is often attenuated online, not least because online harms are often a function of the multiplier effect of online speech: it is frequently not the actions of the original malfeasant actor, but those of neutral actors amplifying that speech or conduct, that cause harm. In such an environment, the culpability of the original actor is surely mitigated and may be lost entirely. Likewise, in the normal course, victims of tortious or illegal conduct and law enforcers acting on their behalf are the primary line of defense against bad actors. But the relative anonymity/pseudonymity of online interactions may substantially weaken this defense.

Many argue, nonetheless, that holding online intermediaries responsible for failing to remove offensive content would lead to a flood of lawsuits that would ultimately overwhelm service providers, and sub-optimally diminish the value these firms provide to society—a so-called “death by ten thousand duck-bites.” Relatedly, firms that face potentially greater liability would be forced to internalize some increased—possibly exorbitant—degree of compliance costs even if litigation never materialized.

There is certainly some validity to these concerns. Given the sheer volume of content online and the complexity, imprecision, and uncertainty of moderation processes, even very effective content-moderation algorithms will fail to prevent all actionable conduct, which could result in many potential claims. At the same time, it can be difficult to weed out unlawful conduct without inadvertently over-limiting lawful activity.

But many of the unique features of online platforms also cut against the relaxation of legal standards online. Among other things—and in addition to the attenuated incentives for self-restraint mentioned above—where traditional (offline) media primarily host expressive content, online platforms facilitate a significant volume of behavior and commerce that isn’t purely expressive. Tortious and illegal content tends to be less susceptible to normal deterrence online than in other contexts, as individuals can hide behind varying degrees of anonymity. Even users who are neither anonymous nor pseudonymous can sometimes prove challenging to reach with legal process. And, perhaps most importantly, online content is disseminated both faster and more broadly than offline media.

At the same time, an increase in liability risk for online platforms may lead not to insurmountable increases in litigation costs, but to other changes that may be less privately costly to a platform than litigation, and which may be socially desirable. Among these changes may be an increase in preemptive moderation; smaller, more specialized platforms and/or tighter screening of platform participants on the front end (both of which are likely to entail stronger reputational and normative constraints); the establishment of more effective user-reporting and harm-mitigation mechanisms; the development and adoption of specialized insurance offerings; or any number of other possible changes.

Thus the proper framework for evaluating potential reforms to Section 230 must include the following considerations: To what degree would shifting the legal rules governing platform liability increase litigation costs, increase moderation costs, constrain the provision of products and services, increase “collateral censorship,” and impede startup formation and competition, all relative to the status quo, not to some imaginary ideal state? Assessing the marginal changes in all these aspects entails, first, determining how they are affected by the current regime. It then requires identifying both the direction and magnitude of change that would result from reform. Next, it requires evaluating the corresponding benefits that legal change would bring in increasing accountability for tortious or criminal conduct online. And, finally, it necessitates hazarding a best guess of the net effect. Virtually never is this requisite analysis undertaken with any real degree of rigor. Our paper aims to correct that.

A Proposal for Reform

What is called for is a properly scoped reform that applies the same political, legal, economic, and other social preferences offline as online, aimed at ensuring that we optimally deter illegal content without losing the benefits of widespread user-generated content. Properly considered, there is no novel conflict between promoting the flow of information and protecting against tortious or illegal conduct online. While the specific mechanisms employed to mediate between these two principles online and offline may differ—and, indeed, while technological differences can alter the distribution of costs and benefits in ways that must be accounted for—the fundamental principles that determine the dividing line between actionable and illegal or tortious content offline can and should be respected online, as well. Indeed, even Google has argued for exactly this sort of parity, recently calling on the Canadian government to “take care to ensure that their proposal does not risk creating different legal standards for online and offline environments.”

Keeping in mind the tradeoffs embedded in Section 230, we believe that, in order to more optimally mitigate truly harmful conduct on Internet platforms, intermediary-liability law should develop a “duty-of-care” standard that obliges service providers to reasonably protect their users and others from the foreseeable illegal or tortious acts of third parties. As a guiding principle, we should not hold online platforms vicariously liable for the speech of third parties, both because of the sheer volume of user-generated content online and the generally attenuated relationship between online platforms and users, as well as because of the potentially large costs to overly chilling free expression online. But we should place at least the same burden to curb unlawful behavior on online platforms that we do on traditional media operating offline.

Nevertheless, we hasten to add that this alone would likely be deficient: adding an open-ended duty of care to the current legal system could generate a volume of litigation that few, if any, platform providers could survive. Instead, any new duty of care should be tempered by procedural reforms designed to ensure that only meritorious litigation survives beyond a pre-discovery motion to dismiss.

Procedurally, Section 230 immunity protects service providers not just from liability for harm caused by third-party content, but also from having to incur substantial litigation costs. Concern for judicial economy and operational efficiency are laudable, of course, but such concerns are properly addressed toward minimizing the costs of litigation in ways that do not undermine the deterrent and compensatory effects of meritorious causes of action. While litigation costs that exceed the minimum required to properly assign liability are deadweight losses to be avoided, the cost of liability itself—when properly found—ought to be borne by the party best positioned to prevent harm. Thus, a functional regime will attempt to accurately balance excessive litigation costs against legitimate and necessary liability costs.

In order to achieve this balance, we recommend that, while online platforms should be responsible for adopting reasonable practices to mitigate illegal or tortious conduct by their users, they should not face liability for communication torts (e.g., defamation) arising out of user-generated content unless they fail to remove content they knew or should have known was defamatory.  Further, we propose that Section 230(c)(2)’s safe harbor should remain in force and that, unlike for traditional media operating offline, the act of reasonable content moderation by online platforms should not, by itself, create liability exposure.

In sum, we propose that Section 230 should be reformed to incorporate the following high-level elements, encompassing two major components: first, a proposal to alter the underlying intermediary-liability rules to establish a “duty of care” requiring adherence to certain standards of conduct with respect to user-generated content; and second, a set of procedural reforms that are meant to phase in the introduction of the duty of care and its refinement by courts and establish guardrails governing litigation of the duty.

Proposed Basic Liability Rules

Online intermediaries should operate under a duty of care to take appropriate measures to prevent or mitigate foreseeable harms caused by their users’ conduct.

Section 230(c)(1) should not preclude intermediary liability when an online service provider fails to take reasonable care to prevent non-speech-related tortious or illegal conduct by its users

As an exception to the general reasonableness rule above, Section 230(c)(1) should preclude intermediary liability for communication torts arising out of user-generated content unless an online service provider fails to remove content it knew or should have known was defamatory.

Section 230(c)(2) should provide a safe harbor from liability when an online service provider does take reasonable steps to moderate unlawful conduct. In this way, an online service provider would not be held liable simply for having let harmful content slip through, despite its reasonable efforts.

The act of moderation should not give rise to a presumption of knowledge. Taking down content may indicate an online service provider knows it is unlawful, but it does not establish that the online service provider should necessarily be liable for a failure to remove it anywhere the same or similar content arises.

But Section 230 should contemplate “red-flag” knowledge, such that a failure to remove content will not be deemed reasonable if an online service provider knows or should have known that it is illegal or tortious. Because the Internet creates exceptional opportunities for the rapid spread of harmful content, a reasonableness obligation that applies only ex ante may be insufficient. Rather, it may be necessary to impose certain ex post requirements for harmful content that was reasonably permitted in the first instance, but that should nevertheless be removed given sufficient notice.

Proposed Procedural Reforms

In order to effect the safe harbor for reasonable moderation practices that nevertheless result in harmful content, we propose the establishment of “certified” moderation standards under the aegis of a multi-stakeholder body convened by an overseeing government agency. Compliance with these standards would operate to foreclose litigation at an early stage against online service providers in most circumstances. If followed, a defendant could provide its certified moderation practices as a “certified answer” to any complaint alleging a cause of action arising out of user-generated content. Compliant practices will merit dismissal of the case, effecting a safe harbor for such practices.

In litigation, after a defendant answers a complaint with its certified moderation practices, the burden would shift to the plaintiff to adduce sufficient evidence to show that the certified standards were not actually adhered to. Such evidence should be more than mere res ipsa loquitur; it must be sufficient to demonstrate that the online service provider should have been aware of a harm or potential harm, that it had the opportunity to cure or prevent it, and that it failed to do so. Such a claim would need to meet a heightened pleading requirement, as for fraud, requiring particularity.

Finally, we believe any executive or legislative oversight of this process should be explicitly scheduled to sunset. Once the basic system of intermediary liability has had some time to mature, it should be left to courts to further manage and develop the relevant common law.

Our proposal does not demand perfection from online service providers in their content-moderation decisions—only that they make reasonable efforts. What is appropriate for YouTube, Facebook, or Twitter will not be the same as what’s appropriate for a startup social-media site, a web-infrastructure provider, or an e-commerce platform. A properly designed duty-of-care standard should be flexible and account for the scale of a platform, the nature and size of its user base, and the costs of compliance, among other considerations. Indeed, this sort of flexibility is a benefit of adopting a “reasonableness” standard, such as is found in common law negligence. Allowing courts to apply the flexible common law duty of reasonable care would also enable the jurisprudence to evolve with the changing nature of online intermediaries, the problems they pose, and the moderating technologies that become available.

Introduction

Economist Ronald Coase devoted an article in the 1974 edition of the American Economic Review to an idea he had observed to be common among his academic colleagues:

(I)n the market for goods, government regulation is desirable whereas, in the market for ideas, government regulation is undesirable and should be strictly limited.

He found the idea strange because, as he argued in the paper, the two markets are not relevantly different. The case for regulation is no weaker in the market for ideas than in the market for goods. After all, it is usually easier for a consumer to know when ordinary goods are faulty than when ideas are bogus. Anyone can tell when a television doesn’t work. It takes unusual dedication to figure out, for example, that Hegel was wrong when he said that “absolute form and absolute content [are] identical — substance is in itself identical with knowledge.”

Coase hoped that devotion to consistency would inspire his peers to adopt a more skeptical attitude toward regulation of the market for goods. He got half of what he hoped for. Academics arguably have become more consistent, but rather than favor laissez-faire in the market for goods, they favor regulation in the market for ideas. This goes to show that consistency is not always something you should seek in your opponents.

Many professors are now keen to restrict the ideas their students hear; or, at least, they are willing to go along quietly with the enthusiasts for such restrictions. They do not seek to protect their students from the incoherent abstractions of 19^th century German philosophers or from any other kind of intellectual error. Rather, they seek to protect them from encountering ideas that will offend them or otherwise make them feel uncomfortable, especially when the topics concern race, sex, sexuality, or some other aspect of “identity.”

Universities are not national or state governments, of course. Their regulatory powers stop at the campus gates. But that doesn’t change the point, which is that many academics appear no longer to believe that the benefits of a free market in ideas are worth the harms that accompany it.

Some outside of universities take the same view, not always drawing the line at private organizations being able to constrain the speech of those with whom they have voluntarily entered contracts. Rather, they want governments to protect consumers of ideas by restricting what can be said. Just as government regulation ensures that only cars meeting certain safety standards are offered for sale, so too should government regulation ensure that only ideas meeting certain safety standards are expressed.

Of course, the market for ideas is already constrained by some safety regulations. For example, an American may not advocate violence or other illegal activity when directed at “producing imminent lawless action.” But beyond this and a few other constraints established by legislation and the courts—such as those entailed by defamation law—the First Amendment to the U.S. Constitution guarantees Americans the freedom to say all manner of harmful things. Some see this as a problem. For example, Richard Stengel, a former managing editor of Time magazine, argued in a 2019 Washington Post op-ed that the United States should follow the lead of other developed nations and develop a hate-speech law. Harvard University law professor Cass Sunstein proposed in his 2021 book Liars that speech deemed by the government to be false and harmful should lose its constitutional protection.

Section 230 of the Communications Decency Act of 1996, which protects “interactive computer services” from being treated as publishers or speakers of the content they host, is also becoming unpopular among those who worry about excessive freedom in the market for ideas. Some of its critics, usually from the political right, think it gives social media firms such as Facebook and Twitter too much freedom to indulge their political biases when moderating content. Other critics, usually from the political left, think it gives such firms too much freedom to host harmful content. Both President Joe Biden and former President Donald Trump have been critical of Section 230, if for very different reasons.

The fashion for private-sector speech prohibitions and proposals for more restrictive legal regimes agitate those who prize freedom of speech. It’s a hot topic in newspaper columns and on talk radio shows. Organizations have even been established to defend free speech, such as the Free Speech Project at Georgetown University and the U.K.’s Free Speech Union.

But defenders of free speech are generally doing their job poorly. Too many merely assert that “you should not have a right not to be offended,” when this is precisely what is at issue. Others follow the 19^th century English philosopher John Stuart Mill and claim that being offended, or suffering hurt feelings more generally, does not count as harm. Again, most seem to simply take this for granted, offering no reason why the offended are unharmed.

The right way to understand harm is economic. Something harms someone if he would pay to avoid it. Since offense and other hurt feelings can pass this test, they can be genuine harm (Section 1). And since speech can cause this harm—and most people believe that legal restrictions on causing harm are generally justified—we have a prima facie case for the regulation of speech.

Indeed, standard economics seems to provide more reason to regulate speech than ordinary goods. If a new car is defective and harms its drivers, people will be reluctant to buy it and its producer will suffer losses. Because the same goes for most goods, regulations that impose product standards are arguably unnecessary (at least, for this reason). Suppliers already have good reason to make their products safe. Speakers, by contrast, often do not bear the cost of the hurt feelings they cause. In other words, hurt feelings are an “external cost” of offensive speech. When someone doesn’t bear all the costs of an action, he tends to do it too much. That is to say, he does it even when the total social cost exceeds the total social benefit.

In his famous 1960 paper “The Problem of Social Cost,” Coase showed that one party holding a legal right not to suffer the external cost of some activity—such as being disturbed by noisy neighbors—needn’t stop it from happening. Nor would giving the neighbors the right to make noise guarantee that the noise continued. This is because, when certain conditions are met, the legally disfavored party will pay the favored party not to enforce his right (Section 2). When this happens, the outcome is efficient: in other words, it maximizes social welfare. Alas, the conditions for such rights trading are rarely met. When they are not, the initial allocation of rights determines the outcome. Which party’s interests should be protected by law therefore depends on who can avoid the harm at the lower cost. The efficient outcome will be produced by giving legal protection to the party facing the higher cost.

Coase’s conditions for trading rights aren’t met in the case of offensive speech (Section 2). We must therefore consider the costs faced by the offenders and by the offended when trying to avoid the offense. This appears to favor speech restrictions. After all, being offended is expensive, keeping your mouth shut is cheap, and each offensive speaker usually offends many hearers. For these reasons, Coasean analysis would seem on first impression to favor revisions to Section 230 that oblige social media platforms to be more assiduous in their moderation of offensive content. A post that would offend millions of the platform’s users can be removed at a low cost to the platform.

But that is merely a first impression. In this paper, I argue that the Coasean case for legal restrictions on offensive speech collapses when confronted with three facts: that being offended is often a masochistic pleasure; that most of the offensive speech that concerns would-be censors occurs on privately owned platforms; and that the proposed restrictions would impose large costs on society. Neither the First Amendment nor Section 230 of the Communications Decency Act should be weakened to remove protection for offensive speech.

Before answering the prima facie Coasean case for restrictions on offensive speech, however, we need to appreciate its force, which begins with recognizing that offense can be a real harm.

Read the full white paper here.