abstract. American competition policy has four big problems: Amazon, Apple, Facebook, and Google. These companies each reign over a sector of the digital marketplace, controlling both the consumer experience and the possibility of competitive entry. This Essay argues that the conventional account of how antitrust law allowed this consolidation of market power—that it failed to evolve to address the market realities of the technology sector—is incomplete. Not only did courts fail to adapt antitrust law from its smoke-stack roots, but they gave big tech special dispensation under traditional antitrust doctrine. Swayed by prevailing utopic views about digital markets in the early 2000s—that they were uniquely dynamic, innovative, and competitive—these courts carved out special exceptions to antitrust rules about tying and the duty to deal with competitors. The tech companies have used this blank check to entrench their market power and keep start-ups from becoming what they themselves once were: the next big thing.

Introduction

American competition policy has a big problem. Actually, it has four big problems: Amazon, Apple, Facebook, and Google. What was once a dynamic pool of smaller start-ups, the high-tech sector has now coalesced around just four companies that together reported over $773 billion of revenue in 2019.1 Each reigns over its own segment of the high-tech marketplace: Amazon controls the retail sector, Apple dominates devices and apps, Facebook owns social media, and Google virtually governs the internet itself. To the extent Silicon Valley still churns out a steady stream of startups, it is more to feed these beasts by acquisition than to produce meaningful rivals to their empires.2

Of course, not everyone agrees that this state of affairs is a problem at all. To some, the size of these firms is merely a symptom of their success. Relentless innovation, a customer-is-king mentality, network effects that benefit consumers, and economies of scale have made these firms ever larger and their products ever better for American consumers. Some even contest the idea that they are large at all by arguing that in a properly defined market, each firm faces significant rivalry and thus lacks market power. Some think that American antitrust law should pat itself on the back for fostering the competitive conditions that let these innovative companies thrive.3

However, this view is increasingly unpopular, and for good reason. Each of these companies, in its own way, holds the keys to competitive entry in many important online markets. To bring an app to market, a developer must deal with Apple; to reach online shoppers, retailers must use Amazon, and so on. Without a meaningful choice between platforms, independent sellers, developers, and websites must pass through a privately maintained bottleneck often on unfavorable terms. These restrictions on competition harm consumers by reducing the output and raising prices for goods that must pass through the bottleneck, and by reducing firms’ incentives to innovate—if they know a large portion of their profits will be appropriated by the platform, they have less incentive to bring new products to market. And by controlling the throttle of technological innovation, each dominant firm can stave off the possibility that one of these nascent companies will build a rival network—a platform that can break the bottleneck itself.4 Long-term, stable platform dominance means consumers likely will not see the kind of Schumpterian innovation associated with great technological leaps forward.5 Rather, consumer welfare depends on these platforms’ internal incentives to innovate, which are weakened in the absence of true rivalry.6 In short, there is a growing recognition that as much as these companies have innovation to thank for their success, their current tactics are making it hard for the next generation of disruptive innovators to take over. If antitrust law continues to stand by, consumers will pay the price.

Resolving this debate about whether high-tech dominance is an antitrust success story or a tragic failure of competition policy is beyond the scope of this Essay. Rather, I take as given the view—increasingly vocalized in the political and academic spheres alike—that these firms have market power, and the accumulation and stability of this market power is a problem for consumers that antitrust law should address.7 This Essay addresses a different question: how did antitrust law let this happen? To those who want more dynamism from high-tech markets, the question is crucial. Those who do not understand antitrust’s past failures are condemned to repeat them.

The conventional wisdom for why antitrust was unable to prevent the growth of these companies is that the Sherman Act and its first century of case law was unprepared to deal with the unique challenges that tech presented.8 It is a statute that was written in 1890 in response to commodity cartels and monopolies;9 the most influential cases interpreting it involve steel,10 oil,11 coal,12 aluminum,13 and tobacco.14 Thus, as American antitrust law entered its second century, it had only low-tech legal tools to confront high-tech market power. It was stuck with doctrines—like market
definition15 or predatory pricing16—that failed to capture the realities of an internet-based economy driven by data, network effects, “free” products, and extreme product integration. The first wave of high-tech case law, starting in 1995, has been a halting, imperfect attempt to modify these old rules to address new anticompetitive arrangements.17 Scholarly criticisms of antitrust’s failure to adapt to the new economy are relatively common.18 Also easy to find are criticisms of the federal government’s underenforcement of the antitrust laws, especially merger enforcement, against the tech sector.19

Less common are scholarly accounts of another reason why the first wave of high-tech antitrust jurisprudence failed to prevent the kind of competitive problems present in high-tech today. At a time when there was great optimism about the internet and its promise of fostering innovation and lowering barriers to competitive entry, the high-tech sector was given special treatment as a matter of substantive antitrust law. The problem, thus, was not only a failure to modify antitrust law to meet the competitive challenges of tech, but also giving tech somewhat of a free pass under some key antitrust doctrines—even as those doctrines stood at the time. This Essay aims to help fill that gap by identifying two important areas of antitrust law where courts gave tech companies special treatment, citing fears of chilling innovation. Yet today, the big-tech companies that have most benefitted from that dispensation are themselves the biggest threat to innovation in digital markets.

This Essay begins by documenting the enthusiasm many scholars and commentators exhibited for the idea that the internet was a competitive panacea—that it would lead to a more competitive, dynamic, and innovative economy. It then traces how this idea resulted in special treatment for tech companies under the antitrust laws. I focus on the D.C. Circuit’s 2001 decision in United States v. Microsoft Corp.,20 which declined to apply Supreme Court precedent to high-tech product integration even as it dealt the tech sector some significant blows with its other holdings. Then, I trace the death of the “duty-to-deal” doctrine to the Supreme Court’s exceptional treatment of telecommunication networks in its 2004 opinion in Verizon Communication’s Inc. v. Law Offices of Curtis V. Trinko, LLP.21 I close by sketching the ways that big tech has cashed these blank checks, pointing out the central irony of this history: tech was given leeway because it was seen as especially innovative; today, the “big four” systematically suppress technological innovation that might threaten their dominance.

I. techtopia

As President Bill Clinton started his last year in office, he began to build his legacy as high-tech’s biggest champion. “In the new century,” he said in a speech in January 2000, “innovations in science and technology will be key . . . to miraculous improvements in the quality of our lives and advances in the economy.”22 In June of that year, the official White House website highlighted more than forty of his initiatives said to increase access to the internet and streamline America’s transition to online commerce.23 Similarly, Vice President Al Gore campaigned to succeed Clinton as President by emphasizing his tech record. “During my service in the United States Congress,” he told a reporter in an interview, “I took the initiative in creating the internet.”24

As a politician at the turn of the twenty-first century, highlighting one’s record encouraging the high-tech revolution made sense. Companies like Microsoft, Apple, Yahoo, and eBay were fueling a tremendous spike in venture-capital investments, pushing the stock market to record highs.25 Household computer ownership more than doubled between 1990 and 199726 and a gold rush to build the infrastructure that would connect these computers was on. In the late 1990s, telecommunications companies—from legacies like AT&T to startups like Qualcomm—invested a total of $500 billion in fiber-optic cable, wireless networks, and 3G spectrum to build an “information superhighway” for the next millennium.27

The connectivity that came with the internet age promised more than convenience; to some it offered “nothing less than the end of scarcity.”28 And that, in turn, could upend two centuries of microeconomic theory that used scarcity to explain the operations of markets.29 As the marginal cost of digital products fell to zero, prices would follow, “making way for an entirely new way of organizing economic life in an age characterized by abundance.”30 Indeed, the rise of “free” goods offered by a second wave of tech startups in the 2000s—companies like Facebook and Google—seemed to vindicate this prediction of an information utopia.

Theories about how high-tech firms would compete displayed some of this utopic thinking. Some economists encouraged sanguinity about highly concentrated tech markets, citing network effects.31 Markets where the value of a product is measured by how well it connects you to others (Facebook is a good example) tend to coalesce around a single provider.32 Further, some economists argued, dominance in these markets was fragile, meaning they lacked the power of a typical monopolist.33 What appeared to be low barriers of entry to the tech market34—at a time when it seemed like everyone was starting a “dot-com” in their garage or dorm room35—ensured competition and all the benefits associated with it.36

It was in this political and economic context that our antitrust institutions—primarily the antitrust enforcement agencies and the federal courts—first confronted allegations that high-tech firms were violating the Sherman Act. The firms were accused of designing their products and selectively refusing to deal with rivals in ways that reinforced consumers’ dependence on dominant networks, pushing network effects beyond the point that provides benefits to the consumer, and making it hard for a better network to take over. In other words, antitrust had to reckon with the reality that technology facilitated not only communication and commerce, but also monopoly power.

II. tech exceptionalism in the “dot-com” era

The late 1990s and early 2000s saw an explosion of scholarship that addressed how competition rules should be interpreted in light of the “new economy.”37 Scholars dusted off Schumpeter’s 1942 theory of “creative destruction,” summarized in a 1998 law review article as the idea that “monopoly or highly concentrated markets were more conducive to innovation than smaller firms operating in fully competitive markets.”38 And scholars also argued that “innovation efficiencies are the principal form of economic efficiency which ought to be protected and promoted by laws designed to maintain a competitive process.”39 All this added up to the need for a light touch for antitrust enforcement against high-tech companies.40 Richard Posner closed his article entitled “Antitrust in the New Economy” with this admonition: “[T]he byword of a prudent enforcement agency and a sensible court will be: caution.”41

Fear of chilling innovation through the threat of antitrust liability was also on the minds of the agencies responsible for enforcing the antitrust laws. The Federal Trade Commission (FTC) issued a policy report in 1996 stating that antitrust policy must take care not to dampen incentives sufficient to generate new networks and standards and not to impose remedies that may increase, rather than decrease, competitive problems. Although not novel, this set of issues is assuming great importance in an increasingly technology-driven economy.42 Indeed, the FTC and the Department of Justice (DOJ) identified innovation effects as a concern in forty-seven merger cases during the second half of the 1990s; in the five years prior, the agencies discussed innovation concerns in only four cases.43

The resurrection of Schumpeter,44 and the assertion that “Innovation is King”45 in antitrust, suggested that existing laws focused on static models of competition were inadequate. Theoretically, adjusting antitrust law to defend Schumpeterian innovation could mean increasing liability in some areas and decreasing it in others.46 And indeed, some antitrust scholars have called for aggressive enforcement against exclusionary conduct by high-tech incumbents, in the name of Schumpeterian innovation.47 But in practice—that is, in the case law—the ratchet tended to only work in one direction: to loosen liability for high tech, out of fear that rigorous application of the antitrust laws as they stood would stifle innovation.

A. Microsoft and Technological Tying

The D.C. Circuit’s 2001 opinion in United States v. Microsoft48is often cited as being ahead of its time in recognizing the threat that internet-based platforms might pose to competition.49 Indeed, the opinion—which found Microsoft liable for monopolizing the operating-system market50—is the best legal blueprint we have for a modern-day antitrust case against big tech because it cogently explains how and why a platform will exclude rivals that threaten to undermine its dominance. But the opinion also contains a holding—and some troubling language about the relationship between innovation and antitrust law—that gets overlooked when it is held up as being tough on tech.

In a series of lawsuits that spanned a decade and produced over one hundred judicial opinions, the DOJ sued Microsoft for using its browser, Internet Explorer, to protect its formidable market power in PC operating systems in the 1990s. The DOJ alleged that Microsoft made it exceedingly difficult for Netscape, a competing browser, to reach consumers running Microsoft’s operating system, Windows.51 Widespread adoption of Netscape posed a competitive threat to Windows’s dominant market share by eroding its considerable “applications barrier to entry.”52 The suit alleged that Microsoft engaged in a pattern of behavior to make using Netscape on the Windows operating system inconvenient and costly.53 Netscape’s share of the browser market languished, and its competitive threat to Microsoft was effectively neutralized.54

The government made two claims. First, the government said that Microsoft’s exclusion of Netscape from the browser market was an act of monopolization that violated section 2 of the Sherman Act.55 On this count, the court agreed with the government and found that Microsoft’s design features and some of its contractual terms with third parties were designed more to exclude Netscape than to benefit Windows users.56 Microsoft was liable for monopolization.57

Second, the government claimed that Microsoft’s technological tie between Internet Explorer and Windows was subject to per secondemnation under the Sherman Act.58 The Supreme Court had held in Jefferson Parish Hospital District No. 2 v. Hyde that ties meeting several economic criteria were per se unlawful, that is, could not be defended as better for consumers.59 But the Microsoft court refused to consider the applicability of a per serule, citing the importance of innovation in the high-tech sector. The court explained, “[t]here may also be a number of efficiencies that, although very real, have been ignored in the calculations underlying the adoption of a per se rule for tying.”60 For the court, it was the high-tech nature of the case that justified an exception: “We fear that these efficiencies are common in technologically dynamic markets where product development is especially unlikely to follow an easily foreseen linear
pattern.”61

Taken as a whole, Microsoft does not stand for the high-tech industry getting a free pass under the antitrust laws; the finding of liability on the monopolization claim was properly perceived as an important loss for the company and for the high-tech sector in general.62 But Microsoft’s hawkish section 2 holding has eclipsed the strikingly dovish holding on the tying claim. The Microsoft court side-stepped Supreme Court precedent63 because it believed “the nature of the platform software market affirmatively suggests that per se rules might stunt valuable innovation.”64 In the end, the decision to avoid the Supreme Court test for per se applicability probably did not affect the outcome, because the test itself asks economic questions that the court might have answered in the negative, resulting in rule-of-reason analysis anyway. But the decision to break with binding precedent, invoking the innovative qualities of the defendant, enshrined in case law a powerful idea: as a dynamic sector of the economy, the high-tech industry can claim special dispensation.

B. Trinko and the Duty to Deal

Three years after Microsoft, the Supreme Court dealt the tech sector an even more valuable exemption. That came in Verizon Communications Inc. v. Law Offices of Curtis V. Trinko, LLP,65 where the court severely curtailed antitrust’s “duty-to-deal” doctrine. The suit alleged that Verizon, partial inheritors of the AT&T monopoly, were under no antitrust duty to deal with competing telecom providers, despite a 1996 federal statute requiring the company to interconnect with those very providers.66

Under the Court’s “duty-to-deal” cases, monopolists can be held liable for refusing to deal with their competitors without a legitimate justification. Leading “duty-to-deal” cases found liability where the refusal seemed more motivated by dreams of monopoly than by short-term profit motives. For example, in Aspen Skiing Co. v. Aspen Highlands Skiing Corp., the Court held that Aspen Ski Co. violated section 2 by refusing to sell tickets to a smaller neighboring ski area so that skiers could enjoy both mountains with a convenient “all-mountain” pass.67 The pass was in high demand and would have been profitable for Ski Co.68 But without the “all-mountain” pass the smaller ski area was sure to fail.69 The Court believed causing the smaller mountain to fail was the real motive behind Ski Co.’s refusal and held it liable for monopolization.70

Before Trinko, the key to the Court’s duty-to-deal cases was disambiguating the defendant’s motives for refusing to deal. To find liability in a duty-to-deal case, a plaintiff had to show that the defendant had a strong legitimate motive to deal with a rival, and it nevertheless refused. That behavior supported an inference that the monopolist had an even stronger illegitimate motive in refusing to deal with a rival—namely, creating a monopoly.71 In Aspen, the defendant would have sold more lift tickets at a profitable price if it sold the all-mountain pass; the fact that it refused to do so supported the inference that the move was really about augmenting its market power.72

Under the prevailing Aspen rule, the plaintiffs in Trinko had a convincing case. The defendant, an incumbent telecom provider with monopoly power, had a strong motive to deal with rival local service providers—they were obligated to do so under a federal statute.73 Despite this motive, Verizon refused to comply with its statutory
obligations,74 which should have supported an inference that Verizon’s refusal to deal was more about monopoly than anything else.

Instead, Justice Scalia, writing for the majority, found that Verizon had no antitrust obligation to deal with its competitors, a holding he said safeguarded the “incentive to innovate.”75 In so holding, Justice Scalia ignored the fact that the balance between incentivizing innovation by limiting a competitor’s duty to share its success with rivals and incentivizing innovation by condemning a dominant firm’s exclusionary conduct had already been struck under the Court’s duty-to-deal cases like Aspen. Justice Scalia instead prioritized monopoly rents as the most important spur to innovate,76 and echoed the prevailing sentiment that the high-tech sector must be maximally incentivized to invest in building networks to support the internet superhighway:

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 . . . may lessen the incentive for the monopolist, the rival, or both to invest in those economically beneficial facilities.77

The Trinko opinion went on to limit Aspen to its facts, paving the way for high-tech firms to refuse to deal with competitors even where the best reason to do so was to build or protect a monopoly.

III. cashing the check

At the turn of the new millennium, big tech was given special treatment under our antitrust laws out of a fear that full enforcement would chill innovation. A great irony of the modern history of antitrust law is that big tech has used that grace to create powerful companies that systematically suppress innovation from rivals in the markets they control, even if they themselves continue to innovate internally.

A. Apple

Apple, a dominant player in devices and apps, has greatly benefited from the deference given to high-tech product design. Apple says the iPhone, which it launched in 2007, “redefined the mobile phone through its seamless integration of hardware and software, its effortless user experience, [and] its simplicity of design.”78 Essential to this “seamless integration” is the closed nature of the iPhone and other Apple devices; only apps available on Apple’s own App Store may be downloaded and used on its devices.

Apple may face competition from other smart phones, such as those running the Android operating system, but their large market share—about half of U.S. smartphones79—means that they control a key access channel for app developers. Indeed, app developers see access to the App Store platform as essential to market their products; since 2008, third-party developers have generated more than $120 billion of sales revenue through Apple.80 But Apple provides access to their platform only on discriminatory and sometimes arbitrary terms.81 Revenue from apps that compete with an Apple product are subject to a thirty percent “commission,”82 while Apple allows firms like Uber—who do not compete with Apple—to keep all revenue generated from iPhone transactions.83 And Apple will sometimes cut off app developers altogether, without explanation or recourse, killing their business.84

Some plaintiffs have challenged Apple’s integration of its phone and the App Store on antitrust grounds.85 Most notably and most recently Epic, maker of the popular video game Fortnite, pulled off an orchestrated attack against Apple by first provoking the platform by refusing to pay the thirty-percent commission on Fortnite purchases. On the same day that Apple kicked Fortnite off the App Store, Epic filed an antitrust lawsuit accusing Apple of monopolization and illegal tying for their refusal to deal.86 Also on that day, Epic released a commercial parodying Apple’s famous 1984 Superbowl ad that depicted the release of the Macintosh as a paradigm-shattering competitive move against IBM, which dominated personal computing in the early 1980s.87 Epic’s parody ad cast Apple, once the upstart David, in the power-hungry Goliath role once held by IBM, and a Fortnite gladiator as the paradigm-smashing innovator.88

Epic’s suit and ad make for high drama, but legally, Epic faces an uphill battle. Microsoft’sspecial treatment for high-tech product integration has taken hold in the case law. In 2010, the Ninth Circuit rejected an antitrust claim against Allied Orthopedic, a pulse-oximeter manufacturer, alleging an unlawful technological tie. The opinion quoted the Microsoft opinions at length:

“As a general rule, courts are properly very skeptical about claims that competition has been harmed by a dominant firm’s product design changes.” . . . “Antitrust scholars have long recognized the undesirability of having courts oversee product design, and any dampening of technological innovation would be at cross-purposes with antitrust law.”89

The Allied Orthopedic court used this language from Microsoft to justify its holding: that if a tie can be said to improve a high-tech product in any way, the competitive effects of its design are beyond antitrust reproach.90

Apple defends its limited access to the App Store as a product-design decision, aimed at creating an integrated tech ecosystem worthy of the Apple brand.91 It remains to be seen whether growing disapproval of Apple’s exclusionary conduct can overcome the language in Microsoft and Allied Orthopedic suggesting that claimed product-design improvements vitiate any competition concerns.92

B. Amazon

Amazon Marketplace is an open online-sales platform that accounts for 52.4% of American online-retail spending; its closest competitor has just 6.6%.93 Amazon allows independent merchants to sell their goods on the Marketplace for a fee and a percentage of sales revenue. It also sells its own products, with 137 private-label brands, including Amazon Basics.94

Not only does Amazon privilege its own products over those of independent sellers,95 but there is evidence that Amazon uses the vast quantity of data it gathers on independent sellers to mimic their businesses. Lina Khan describes the array of information Amazon gathers about rivals’ products—from searches to clicks to sales to inventory—as “microdetails that add up to a formidable—and constantly evolving—arsenal of market intelligence.”96 Despite Amazon’s insistence that it does not use the data to its competitive advantage, many see the evidence in Amazon’s behavior.97 The effect of Amazon’s mimicry is often exit by the independent seller.98

Independent sellers may prefer to sell on Amazon without opening themselves up to the risk of mimic-and-destroy tactics. But under the law of Trinko, Amazon has little duty to deal with Marketplace rivals at all, much less on favorable terms. Consumers lose on two fronts, one actual and one potential. With the threat of Amazon appropriating any profits from a new retail-product idea, independent producers have little incentive to bring it to market, resulting in fewer choices for consumers. And by eliminating smaller competitors by undercutting their prices, Amazon is building market power that it could someday exercise in the form of higher prices to consumers.

C. Google

Google is a dominant internet search engine that enjoys eighty-eight percent of the U.S. market.99 It provides search for free, but its dominance in search is valuable for its revenue center: advertising, where it makes about $161 billion of revenue annually.100 Google’s monopoly is buttressed by the data it scrapes from the companies, like Trip Advisor, Yelp, Hulu and Netflix, that rival its own properties like Google Maps and YouTube.101 It uses that data to enhance its own ecosystem, at the center of which is Google Search. The effect may be a better search engine, but this nearly unfettered access to data also creates a formidable barrier to entry for anyone trying to topple Google’s empire.102

Google obtains rivals’ data by giving them a Hobson’s choice: either allow us to scrape your data or face delisting on Google search.103 Trinko’s deference to high-tech refusals to deal presents a significant problem to plaintiffs challenging delisting as anticompetitive. For example, on its rise to dominance, Google used “blockage”—the practice of “delisting, de-indexing and censoring”104 a website on its search engine—to suppress competition from a parent-oriented search product called “Kinderstart.com.” When Kinderstart sued, claiming that Google violated its duty-to-deal under section 2, the court cited Trinko in dismissing the claim.105

D. Facebook

Although Facebook is most well-known for its anticompetitive acquisitions,106 it also has maintained its dominance using tactics that might have run afoul of the antitrust laws before the dot-com era. Like Apple, Facebook selectively cuts off access to apps it believes pose a competitive threat, as it did for Twitter’s video feature, Vine.107 Like Amazon, it uses its access to data to mimic rivals’ products and eliminate their competitive threat, as it did for Snapchat.108 And like Google, Facebook conditions its dealings with rivals on receiving their data, even when that data contain the seeds of their own destruction.109

Facebook may defeat an antitrust suit by citing Microsoft and pointing out that creating an integrated, user-friendly Facebook experience requires excluding some rivals from the platform. And it has already used the logic of Trinko to defeat a claim that it violated the Sherman Act when it denied competitor websites direct login access to Facebook.110

E. Whither Innovation?

As a theoretical matter, big tech’s refusals to deal and predatory copying suppress innovation. A retailer with a new idea for a household product will be less inclined to invest in producing it if he knows Amazon can appropriate the returns. A developer with a better “app for that” will be less likely to bring it to market if she believes Apple or Facebook might someday remove it from their platforms. And if a rival search company cannot hope to keep its data private from Google, it will not invest in building a better search engine to try to take on the giant.

Whether big tech stifles innovation as an empirical matter is less clear, but there is anecdotal evidence that it does. During a recent hearing following the House Judiciary Committee’s investigation into competition abuses among high-tech firms, Representative Cicilline read a quote that he said was typical of the entrepreneurs he interviewed: “If someone came to me with an idea for a website or a web service today, I’d tell them to run. Run as far away from the web as possible.”111 Venture capital, while booming overall,112 is shy about funding projects that might compete with Big Tech. The best-case scenario for a start-up is acquisition by one of the big four—a lucrative payday, for sure, but nothing compared to what could come from actually toppling a dominant firm. This puts a ceiling on the upside, and with the ever-present risk of failure, it likely leads to under-investment in new ideas. As one funder put it, “[w]e don’t touch anything that comes too close to Facebook, Google or Amazon.”113

Conclusion: “antitrust Is Greedy”

The promise that we saw in high tech during its first boom—that it would change the way we work, communicate, shop, and play—has largely been realized. Few can argue with the efficiencies that digital communication and commerce have brought to our lives and markets. But, as Professor Herbert Hovenkamp has said, “antitrust is
greedy.”114 It wants not only efficiency in end products, but efficiency in the competitive process that brings them about. During the dot-com era, American antitrust institutions became enthralled with the idea that encouraging the development of dynamic, innovative products required compromising our commitment to dynamic, innovative markets. That compromise contributed—in a way that is often overlooked—to the current competition crisis in big tech.

Rebecca Haw Allensworth is Professor of Law at Vanderbilt Law School.