IT’S HIGH TIDE AGAIN IN INTERNET MARKETS

By Josh Palmer¹

"The Internet is a tidal wave. It changes the rules." When Bill Gates wrote this in an internal Memo to his executive staff in May 1995, it was in large part a warning that Microsoft had to focus on getting to the forefront of this wave or else have its dominant position in computing washed away. Microsoft’s ensuing practices would result in the United States Department of Justice (DOJ) filing an antitrust case against the company in 1998. A multitude of private actions in the U.S. followed, as did similar cases globally.

Twenty-five years after Gates’s memo, the Internet tidal wave seems to be cresting again. This time in the form of looming antitrust cases against companies that have enjoyed large success, by developing digital platforms that leverage the Internet (and other) capabilities Bill Gates highlighted in his Tidal Wave Memo.

In this article, I briefly review Microsoft’s conduct pursuant to Gates’s memo and the antitrust cases against Microsoft that followed. I then review the economic and antitrust lessons from the Microsoft cases as espoused by the authors of a widely used industrial organization textbook. I conclude by discussing what these lessons, combined with commentary from economists analyzing digital platforms and competition, suggest about the next wave of antitrust cases in these markets. Given the complexity of digital platforms and range of antitrust concerns that have been raised, this article in no way should be taken as exhaustive. Rather, I attempt to highlight some of the issues that seem most prevalent and interesting in the antitrust litigation context.²

I. MICROSOFT AND THE FIRST INTERNET TIDAL WAVE

Much as an oceanic tidal wave is the consequence of gravitational interactions between the Sun, Moon, and Earth,³ Bill Gates recognized that the Internet tidal wave was the result an interaction of multiple technologies, including TCP/IP protocols allowing computers to operate on distributed networks, HTML and other extensions allowing information to be presented in a more structured manner, modern communications infrastructure that could be purchased through commodity bids, and "most important[ly]", a place to publish public content.⁴

[Page 70]

Gates also predicted that the Internet tidal wave created by these technological interactions created enough energy to fundamentally alter not only how information was processed, but also how businesses functioned and consumers interacted and learned, projecting that: "In the next 20 years the improvement in computer power will be outpaced by the exponential improvements in communications networks. The combination of these elements will have a fundamental impact on work, learning, and play."⁵

Of primary concern to Gates was the possibility that a computer’s operating system would no longer be an essential part of the computing ecosystem. In particular, he warned that if web browsers were able to operate as middleware compatible across operating systems, Microsoft’s monopoly position in operating systems with its Windows operating system would be severely jeopardized. Gates specifically called out Netscape, "[a] new competitor ‘born’ on the Internet . . . pursuing a multi-platform strategy where they move the key API into the client to commoditize the underlying operating system."⁶

Three years after Gates’s Tidal Wave Memo informing his executives and directors of plans to "define an integrated strategy" across the company, with the goal to "protect and grow our Windows asset",⁷ the DOJ filed an antitrust suit claiming that Microsoft engaged in conduct to protect its operating system monopoly. Consistent with Gates’s directive, the allegations included a charge that Microsoft was bundling its own browser, Internet Explorer, with Windows to leverage its monopoly into the emerging browser market. The DOJ further alleged Microsoft used exclusionary agreements that precluded downstream customers from buying, using, distributing, or promoting other companies’ products, as well as restricted the rights of its customers to provide services or resources to Microsoft’s actual and potential software competitors.⁸ A number of private antitrust cases followed, including class actions on behalf of consumers in various states.⁹ Microsoft’s conduct also led to antitrust actions around the world.¹⁰

[Page 71]

After reviewing the evidence in the DOJ and consolidated states’ cases—including Gates’s Tidal Wave Memo—the Court released its findings of fact on November 5, 1999. Among the facts found to have been proven by a preponderance of evidence was that:

To the detriment of consumers, however, Microsoft has done much more than develop innovative browsing software of commendable quality and offer it bundled with Windows at no additional charge. As has been shown, Microsoft also engaged in a concerted series of actions designed to protect the applications barrier to entry, and hence its monopoly power, from a variety of middleware threats, including Netscape’s Web browser and Sun’s implementation of Java. Many of these actions have harmed consumers in ways that are immediate and easily discernible. They have also caused less direct, but nevertheless serious and far-reaching, consumer harm by distorting competition.¹¹

Although the subsequent litigation resulted in a lower court ordering Microsoft broken up and in 2001 the D.C. Circuit unanimously affirming (en banc) the company’s liability for monopolization, Microsoft ultimately entered a consent decree with the DOJ, which included conduct-related instead of structural remedies.¹² Microsoft also agreed to settlements in the billions of dollars in private cases. The California indirect-purchaser plaintiff class, for example, obtained a settlement valued at over $1 billion.¹³ Microsoft also settled related lawsuits with a number of firms, including Novell ($536 million), AOL-Time Warner ($750 million), and Sun Microsystems ($700 million).¹⁴

Pointing to, in part, a precedent-setting structured reasonableness framework for evaluating a dominant firm’s conduct under Sherman Act §2, as well as clarifying that antitrust laws extend to nascent actual and potential competitive harms, even in rapidly changing innovation markets. Jonathan Baker claims that "Microsoft was the most important U.S. antitrust decision of the past three decades."¹⁵

Not surprisingly, given the importance of the technological and competitive issues involved, a variety of sources have proffered lessons from the Microsoft cases. Economists Dennis Carlton and Jeffrey Perloff, authors of the widely used industrial organization textbook, "Modern Industrial Organization", have offered a summary of both the industrial organization and antitrust lessons from the Microsoft case, which is especially relevant for anticipating the potential antitrust litigation in digital markets.

[Page 72]

II. ECONOMIC AND ANTITRUST LESSONS FROM MICROSOFT

From an industrial organization perspective, Carlton and Perloff profess that Microsoft teaches that "in a rapidly changing technological environment with scale economies, complementary products, and network effects a dominant firm can use strategic behavior to preserve and increase its dominance even though the product changes dramatically over time."¹⁶ Though this may seem obvious, Microsoft and others had questioned whether the dynamic aspect of technology markets negated the ability of firms to exercise market power over the long run.¹⁷

However, from an antitrust litigation perspective, as Carlton and Perloff point out, Microsoft also teaches (or, at least reaffirms) that strategic behavior to disadvantage rivals—which the authors note is a core component of most business programs—may or may not benefit consumers even when it successfully results in market power, depending on the context. Consequently, they posit as another lesson the need to identify and weigh both pro- and anti-competitive effects—that is, the need to conduct a rule-of-reason analysis of strategic conduct in technology markets.¹⁸

As Internet-related technologies and the use of digital platforms have expanded in the 25 years since Bill Gates’s Tidal Wave Memo, so too has the range of potential strategic behavior available to digital platform firms.¹⁹ This suggests that rule-of-reason analyses in digital platform litigation is likely to be much more involved and costly. Although the full set of strategic conduct is wide-ranging and outside the scope of this article, I review the primary economic features of digital markets and how these features are likely to impact the challenges of conducting rule-of-reason analyses in these markets. First, however, I briefly review the firms that were powered by the Internet tidal wave Gates predicted—Google, Apple, Amazon, and Facebook—and that are the focus of most of the current wave of antitrust scrutiny of big tech.

Lastly, after reviewing the primary economic features of digital platforms and how they may impact rule-of-reason analyses, I point to commentary from leading economists that suggests antitrust litigation involving digital platforms may well require not only more, but also new types of evidence and analyses.

[Page 73]

III. THE RISE OF GOOGLE, APPLE, AMAZON, AND FACEBOOK

As Bill Gates predicted in his 1995 Tidal Wave Memo, the Internet has changed the rules not only on how technology works, but how consumers engage in economic, social, educational, and nearly every other type of activity. While Gates accurately predicted the events, he missed on the firms that would come to dominate. Specifically, of the "Big-Four," only Apple was mentioned as a potential competitor, and that was due to Apple’s early adoption of TCP and its "strength in education". Netscape and Yahoo were the browser and search competitors Gates singled out.²⁰

Of course, it would have been difficult for Gates to foresee the rise of any of these other companies. Amazon still had not sold its first online book when Gates wrote his Tidal Wave Memo, and Google and Facebook were still years away from starting. The Stigler Committee on Digital Platforms—a committee of legal, economic, and political science academics that spent over a year studying digital platforms—recently concluded:²¹

One of the key defining factors of the past decade is the rise of Digital Platforms (DPs), such as Google, Facebook, Amazon, Apple. As more and more of our economy and society moved online, these companies ascended from non-existent or nearly bankrupt in the early 2000s to join Microsoft as global behemoths, exceeding (as of August 2019) more than 4 trillion dollars in market capitalization.
This meteoric rise is not surprising. These companies invented new products and services that revolutionized the way we work, study, travel, communicate, shop, and even date. In the process, they created trillions of dollars in consumer surplus.

In 2019, Apple, Microsoft, Google-parent Alphabet, Amazon, and Facebook were amongst the world’s Top 10 digital companies.²² "Collectively, the Big Five tech companies generate over $800 billion in revenue each year, making them bigger than Saudi Arabia’s entire economy."²³ Apple, Amazon, Microsoft, and Google-parent Alphabet also dominate the stock market, "making up 17% of the S&P 500’s total market value."²⁴

[Page 74]

Given their size and dominance, it is no surprise that these large firms now face their own tidal wave of antitrust challenges.²⁵ In the U.S., for example, federal and state investigations into these companies are ongoing,²⁶ as are various class actions.²⁷ There are also investigations or actions in (at least) Australia, the United Kingdom, Germany, the European Commission, France, Israel, and Japan.²⁸

As the wave of pending antitrust cases builds against these digital firms, the lessons from the Microsoft case—that strategic behavior by these firms can have both beneficial and harmful impact so that a rule-of-reason analysis is required—provide a lens through which likely economic inquiries and issues that will be litigated can be brought into view.

IV. THE CHALLENGES OF RULE-OF-REASON ANALYSES IN DIGITAL MARKETS

Digital platforms encompass an ever-widening range of technologies, industries, and consumers. Nonetheless, a core set of features typically capture the most relevant aspects for understanding the underlying economics. After reviewing these core features, I argue that the complexity of digital platform environments indicates that the variety and extent of economic analyses in rule-of-reason analyses will also need to scale. Hence, the already long and difficult task of identifying net competitive effects is likely to get longer and more difficult for all participants.

Further, highly regarded industrial organization economists have argued that the evidence and analyses presented in digital platform litigation will likely need to extend beyond traditional approaches. I consider some ways in which this may manifest.

A. Economic Features of Digital Markets

Discussions related to digital platforms and antitrust typically revolve around the cost structure of digital technology, network effects, and zero-price aspects.

Economies of scale—declining per unit costs as output increases—are characteristic features of digital platforms. On the supply side, hardware, software, and heavy reliance on intellectual property all contribute to economies of scale. For example, as Bill Gates recognized, the Internet allows digital assets to be distributed with low marginal costs.²⁹ Similarly, the software and intellectual property on which digital business is built and conducted also results in economies of scale.³⁰ All else equal, industries with significant economies of scale in costs result in larger firms.

[Page 75]

Digital platforms can also exhibit direct or indirect network effects (or both) on the demand side.³¹ Direct network effects occur when users on one side of the platform gain additional benefits from users on the same side of the platform. Communication or social networks are an example—the more people I can interact with on my network, the more I value it.

Digital platforms also exhibit indirect network effects, which "refer to the situation in which participants on one side of the market value having more participants on the other side with whom they can have a mutually beneficial interaction."³² Indirect network effects occur because the greater density of relevant parties a platform can bring to one side of the platform, the greater the value of the platform to the other side of the platform.³³ Evans and Schmalensee point to OpenTable as an example. The more restaurant patrons use OpenTable, the more valuable it is to restaurants. Similarly, the more restaurants use OpenTable, the more valuable it is to restaurant patrons. Indirect network effects generate positive-feedback loops (virtuous cycles) for successful firms.³⁴ Complementary products, like games that are developed for a particular social media cite, also generate indirect network effects.³⁵

Like economies of scale, network effects tend to promote market structures comprised of large firms. The strength of network effects can be mediated by the ability (and willingness) of users to use multiple platforms (often referred to as "multi-homing"). The easier it is to use multiple platforms for a given service, the weaker the network effects.³⁶

[Page 76]

Indirect network effects also generate externalities because the decision of an economic agent to join or leave a platform impacts the value of that platform for other economic agents. Consequently, the indirect network effects result in interrelated pricing structures. That is, pricing on one side of the platform must take into account how any demand effects will propagate through to the other side(s). In particular, a price increase will tend to decrease demand, and the decreased number of participants will reduce the value of the platform to users on the other side(s) of the platform. As a result, optimal pricing policies over the platform may make price-cost comparisons on a particular side of the platform less informative of market power than in non-platform markets.³⁷ Consequently, the price-cost relationship on any particular side of the platform may not be an accurate measure of market power.

Another key aspect of the digital platforms operating on the Internet, which Bill Gates foresaw, is zero-price services.³⁸ Many digital platforms involve sides on which one party does not pay even a nominal price for the provided service.³⁹ For example, Google does not charge its search engine users a fee. In these cases, firms compete for consumers’ attention or presence, not their dollars. These markets are often referred to as "attention" markets to distinguish them from "transaction" markets. Multi-sided platforms can be comprised of both attention and transaction markets. When consumers do not directly pay a money price, measures of quality and innovation become more important in evaluating consumer welfare.⁴⁰

The lesson from Microsoft that a rule-of-reason analysis is necessary to determine the net competitive effects of strategic behavior by technology firms indicates that the economies of scale, network effects, and zero-price elements inherent in digital platform markets will impact how the looming digital platform antitrust cases are litigated.

[Page 77]

B. Strategic Behavior and Rule-of-Reason Analysis in Digital Markets

The complexity of digital markets suggests that evaluating strategic behavior under a rule-of-reason framework will likely mean a multiplicative, not additive, increase in necessary facts and analyses.⁴¹ That is, even traditional areas of disagreement like the relevant market and appropriate counterfactual will be more intensely contested.⁴²

Market definition is a predictable element of disagreement in most antitrust cases. As Baker points out, even in traditional markets, there need not be a unique set of products that comprise an antitrust market.⁴³ Instead, a variety of product groupings may satisfy, for example, the hypothetical monopolist test that forms the conceptual framework for identifying relevant antitrust markets in the DOJ and FTC’s horizontal merger guidelines.⁴⁴ Plaintiffs and defendants typically offer competing market definitions. In multi-sided markets, this disagreement is likely to spill over to each side of the at-issue platform, multiplying the number of market definition arguments.⁴⁵

A related issue is how to measure and interpret shares and concentration within multisided markets. For example, when a particular side of the platform does not pay a price, what measure will be used?⁴⁶ The complexity of the underlying business practices and technology also adds to the challenge of obtaining accurate shares. Bitton and Lewis (2020), for instance, argue that misunderstanding the underlying technology can (and has) led to inaccurate claims regarding Google’s market shares in the digital advertising space.⁴⁷

In addition to determining the relevant markets, shares, and concentration levels, there is still no consensus amongst economists regarding which of the sides of the multi-sided platforms must be analyzed.⁴⁸ The issues of what are the relevant antitrust markets, how competitive are they, and the range of relevant economic impacts will generally be more complicated for digital platforms.

[Page 78]

As a further complication, antitrust analyses are inherently reliant on the presumed counterfactual—is the market less competitive than it would have been but-for the at-issue conduct? Are consumers worse off than they would have been but-for the at-issue conduct? Given the conceptual nature of any but-for world, even once a relevant market is decided upon, there will certainly be differing opinions offered on the counterfactual conduct and its impact on the market dynamics and equilibrium outcomes.⁴⁹

The supply and demand economies of scale, for example, raise interesting issues when analyzing counterfactuals. For instance, since the structure of multisided digital platforms makes these markets inherently prone to tipping (that is, being dominated by one or a few large firms),⁵⁰ what counterfactual aspects would prevent one dominant firm from being replaced by another dominant firm in the but-for world? That is, how much additional market power does a defendant have relative to the amount of market power in the counterfactual?⁵¹ Similarly, of the many available alternatives, what but-for conduct will be proffered on each side of the platform and on what basis?⁵²

Even settling on the but-for conduct, what market dynamics and equilibrium outcomes would have resulted (and when) in each of the relevant markets? For example, network effects indicate that consumer preferences, and hence demand functions, change as the platform adds or loses users. If the but-for conduct would have resulted in less dominance by the defendant, what would the but-for demand curves look like? Efficiency and rationality arguments in the presence of dynamic preferences are more complicated than those based on stable demand curves.

[Page 79]

Research into the economics of digital platforms is relatively new and evolving.⁵³ The lack of established tools and empirical work will likely raise arguments about whether proffered evidence on the assumed counterfactual conduct and consumer effects is sufficiently reliable, rather than merely theoretical.⁵⁴

Antitrust litigation is complex. It can be difficult for judges and juries to understand and evaluate the economic evidence underlying necessary legal elements. Further, there are negative consequences to both false positives and false negatives.⁵⁵ The predictable disputes over market definition and counterfactuals can be expected to inherit the complexity of the of the underlying strategies and technology, resulting in intensified calls for specialized triers of fact.⁵⁶

V. HOW MIGHT DIGITAL PLATFORMS ACTIONS ALTER ANTITRUST LITIGATION?

Beyond just requiring more of the same when it comes to market definition arguments, counterfactual assertions, and calls for specialized triers of fact, numerous highly regarded economists argue that new types of economic evidence and analyses may be needed when evaluating digital platforms.

A. Overwhelming Complexity as a Driver for New Analytical Tools

The complex interdependencies among the various platform sides have led well-established economists to call into question the sufficiency of traditional economic analyses and tools to handle competition analysis in digital platforms. Some illustrative examples follow.

Jean Tirole, winner of the 2014 Nobel Memorial Prize (Economics) and author of a popular industrial organization textbook argues that:⁵⁷

[Page 80]

When it comes to multi-sided platforms, these [traditional principles of competition policy] simply are not applicable in many cases. New guidelines for adapting competition policy to two-sided markets would require that both sides of the market be considered together, rather than analyzed separately. . . . This will require care and a new analytical approach.

David S. Evans and Richard Schmalensee, economics professors at University College London and MIT, respectively, and widely cited by the Supreme Court in the two-sided platform case Ohio v. American Express, assert that:⁵⁸

New tools may well be necessary to apply traditional principles appropriately in markets with multi-sided platforms[.]

Kate Collyer, Hugh Mullan, and Natalie Tilman, economists at the Competition and Markets Authority in the UK, claim that:⁵⁹

Many of our standard tools for assessing market power are more complex to apply in multi-sided markets and may need to be adapted.

Fiona Scott Morton and her subcommittee on digital platforms concluded that:⁶⁰

From an economic perspective, there is no single new characteristic that would make competition in digital platforms different from more traditional markets. Rather, it is the coincidence of several factors at a scale that has not been encountered before that makes the problem unique and requires new analysis of market structure and market power.

Assertions by such well-known industrial organization and competition economists questioning the reliability of standard tools and noting the relatively unsettled body of empirical knowledge are likely to be raised in litigation, especially by defendants.

B. The Importance of Zero- and Non-Price Aspects

Zero- and non-price attributes also threaten to complicate digital platform antitrust litigation. For starters, as Alfred Marshall—a founder of modern economics and the marginal revolution—explained in his classic microeconomics principles text: "Economic laws, or statements of economic tendencies, are those social laws which relate to branches of conduct in which the strength of motives chiefly concerned can be measured by a money price."⁶¹

When consumers obtain a product or service which they do not directly pay a monetary price for, the typical signal of their valuation of that product or service is absent. In its absence, economists have turned to cognitive resources like attention as the object of interest when using zero-price services.⁶² As standard economics was not developed to evaluate consumer welfare effects as measured by non-price elements, behavioral economics and marketing have become important areas of expertise.⁶³ This raises the question of whether behavioral economists or marketing experts, who may have more relevant subject matter expertise in some instances than traditional economists, will be used as expert witnesses in digital platform antitrust litigation. For example, the economists on the Market Structure and Antitrust Subcommittee, make this point about behavioral economics on a number of occasions:

[Page 81]

Economists and lawyers will have to develop tools to explain to courts the role of behavioral biases in the creation of market power and in their effect on the quality of content.⁶⁴
Behavioral economics has had a profound influence in the conduct of economic policy that will become even more prevalent as more knowledge is digested and applied. It is of great relevance for our understanding of internet economics because, as information flows improve and some physical barriers are removed, human factors are more likely to provide the frictions that have increasing effects on market outcomes.⁶⁵
Given the prevalence of behavioral effects in the digital economy, the measurement of consumer welfare must be carried out very carefully. As we have mentioned, behavioral economics is now a well-established discipline that can help sort different online behaviors and business practices. Incorporating this knowledge into the legal practice’s toolbox may help develop better measures of output and quality.⁶⁶

More and more economists acknowledge the role of non-price attributes in consumer welfare, and especially to the extent cognitive and psychological aspects are cited as important features for understanding the effect of firm conduct on consumers.⁶⁷ Consequently, questions of whether and how these aspects are included in both the cost and benefit sides of consumer welfare analyses seem imminent.⁶⁸

[Page 82]

Of course, widening the scope of inquiry not only muddies the waters in terms of what evidence, analyses, and areas of expertise are most relevant and reliable, it also runs into the practical limits of the litigation system.⁶⁹ Courts already have to balance good versus good-enough-for-litigation science.⁷⁰

VI. CONCLUDING REMARKS

Bill Gates foresaw many of the valuable benefits that successful internet companies would bring by integrating previously unavailable technologies. As the Stigler Committee concluded:

Over the past 25 years, that power has exploded with head-spinning velocity: Today, there is no area of human life that has not been affected by the technological innovations made possible by the internet. We now buy goods and services, do banking, pay bills, find information, and talk with multiple groups of friends and acquaintances on the web. The speed, scale, and scope of the internet, and of the ever-more powerful technologies it has spawned, have been of unprecedented value to human society."⁷¹

Thus, as Carlton and Perloff professed after Microsoft, even strategic behavior that harms rivals can increase consumer welfare, necessitating a rule-of reason analysis. However, the complexity and scale of digital platforms promise to carry over to any rule-of-reason analysis, multiplying the areas of disagreement and resources need to adjudicate them. It remains an open question whether non-specialized triers of fact will be able to fulfill this difficult task.

[Page 83]

Moreover, price, the typical signal economists rely on to gauge value, is absent from many digital markets. This presents a challenge to traditional economic analyses. As a result, analyses of consumer welfare effects are more likely to benefit from—if not require—new types of evidence and analyses. This opens the door for behavioral economists, marketing experts, and perhaps others to serve as expert witnesses in antitrust litigation.⁷²

As the wave of antitrust cases against digital platforms crests, it only seems fitting to let Bill Gates deliver the key takeaway from Microsoft: "The Internet is a tidal wave. It changes the rules."

[Page 84]

——–

Notes:

^1. Josh Palmer is a partner at applEcon, LLC. The views expressed here are solely those of Dr. Palmer and not applEcon, LLC. Dr. Palmer would like to thank Breanna-Bode Szopo for her excellent research assistance.

^2. For a more comprehensive view from a policy perspective see Jonathan B. Baker, The Antitrust Paradigm: Restoring a Competitive (Harvard Univ. Press 2019). For a more comprehensive view of the technology and economics see Stigler Comm. on Dig. Platforms, Final Report (Sept. 2019), https://research.chicagobooth.edu/stigler/media/news/committee-on-digital-platforms-final-report [hereinafter Stigler Comm. Report].

^3. See What is a Tidal Wave?, Nat’l Oceanic & Atmospheric Admin., https://oceanservice.noaa.gov/facts/tidalwave.html (last updated June 25, 2018).

^4. See Bill Gates, The Internet Tidal Wave, DOJ, at 1-2 (May 26, 1995), https://www.justice.gov/sites/default/files/atr/legacy/2006/03/03/20.pdf.

^5. Id. at 1.

^6. Id. at 4.

^7. Id.

^8. Complaint, ¶5, United States v. Microsoft Corp., 97 F. Supp. 2d 59 (D.D.C. 2000).

^9. See Consumer Class Action Settlement Information, Microsoft, https://www.microsoft.com/en-us/legal/class/default.aspx (showing California, New York, and Iowa were among the state class actions in which plaintiffs have alleged that Microsoft unlawfully used anticompetitive means to maintain a monopoly in markets for certain software).

^10. See Turning the Page on the Past by Investing in the Future of Canada, Microsoft (Aug. 8, 2018), https://news.microsoft.com/en-ca/2018/08/08/turning-the-page-on-the-past-by-investing-in-the-future-of-Canada/ ("Microsoft and the plaintiffs have reached a settlement and have filed papers seeking approval in the courts of British Columbia, Ontario and Quebec. . . . [L]ike the antitrust allegations made against Microsoft in the United States in the 1990s, the cases we are resolving today are about conduct from the 1980s and 90s."); see also Microsoft Hit by Record EU Fine, CNN (Mar. 25, 2004), https://web.archive.org/web/20060413082435/http:/www.cnn.com/2004/BUSINESS/03/24/microsoft.eu/ ("[T]he European Union has found Microsoft guilty of abusing the ‘near-monopoly’ of its Windows PC operating system and fined it a record 497 million euros ($613 million). . . . [T]he Complaint against Microsoft centered on the Microsoft Media Player, which plays music and video clips. It is a free add-on to Windows.").

^11. See Finding of Facts, ¶409, United States v. Microsoft Corp., 97 F. Supp. 2d 59 (D.D.C. 2000).

^12. See generally Stipulation, United States v. Microsoft Corp., 84 F. Supp. 2d 9 (D.D.C. 1999).

^13. Declan McCullagh, Judge OKs $1.1 Billion Microsoft Deal, CNET (Aug. 29, 2003), https://www.cnet.com/news/judge-oks-1-1-billion-microsoft-deal/. Two of applEcon’s founding partners, Professor Jeffrey MacKie-Mason and Dr. Janet S. Netz provided expert witness testimony on behalf of the California class (and other states). At that time, the author worked as a staff analyst supporting these experts.

^14. Steve Lohr & Paul Meller, Microsoft to Pay $536 Million to Novell in Antitrust Case, N.Y. Times (Nov. 9, 2004), https://www.nytimes.com/2004/11/09/technology/microsoft-to-pay-536-million-to-novell-in-antitrust-case.html.

^15. Baker, supra note 2, at 197.

^16. See Dennis Carlton & Jeffrey M. Perloff, Modern Industrial Organization, Ch. 11§ 4 (4th Global ed. Pearson 2000).

^17. See generally Erik Brynjolfsson & Michael D. Smith, Frictionless Commerce? A Comparison of Internet and Conventional Retailers, 46 Mgmt. Sci. n.4, 563, 563-85 (2000).

^18. See Carlton and Perloff, supra note 16, at Ch. 11 § 4.

^19. See David S. Evans & Richard Schmalensee, Antitrust Analysis of Platform Markets: Why the Supreme Court Got It Right in American Express 15 (Competition Policy International 2019) (hereinafter Evans & Schmalense). In this article, the term digital platform indicates an Internet-based product or service that facilitates interactions between distinct customers of the platform. The distinct customers comprise different sides of the platform’s business. Digital platforms, therefore have two or more sides. Id. at 15. For example, Google’s search engine facilitates transactions between advertisers, search users, and web content creators, indicating a 3-sided platform. This is consistent with academic definitions and the Supreme Court’s definition in Ohio v. Am. Express Co., 138 S. Ct. 2274, 2280 (2018).

^20. Gates, supra note 4, at 2.

^21. Stigler Comm. Report, supra note 2, at 6.

^22. Top 100 Digital Companies List, Forbes, https://www.forbes.com/top-digital-companies/list/.

^23. See Zia Muhammad, Alphabet, Amazon, Apple, Facebook, Microsoft: How Big Tech Companies Earn Revenue, Digital Info. World (May 12, 2019), https://www.digitalinformationworld.com/2019/05/how-tech-giants-make-billions-infographic.html.

^24. Michael Sheetz, Apple, Amazon, Microsoft and Alphabet and the Road to $1 trillion, CNBC (Jan. 31 2020, 1:42 PM), https://www.cnbc.com/2020/01/31/apple-amazon-microsoft-and-alphabet-and-the-road-to-1-trillion.html.

^25. See, e.g., Herbert Hovenkamp, The Antitrust Enterprise, Principle and Execution 109 (Harvard Univ. Press 2005); see also, e.g., Jean Tirole, Regulating the Disrupters, LiveMint (Jan. 1, 2019), https://www.livemint.com/Technology/XsgWUgy9tR4uaoME7xtITI/Regulating-the-disrupters-Jean-Tirole.html ("Given the scale and scope of these firms’ impact on our societies, it is no surprise that they inspire both hope and fear in the public consciousness").

^26. Russell Brandom, The Regulatory Fights Facing Every Major Tech Company, The Verge (Mar. 3, 2020), https://www.theverge.com/2020/3/3/21152774/big-tech-regulation-antitrust-ftc-facebook-google-amazon-apple-youtube.

^27. See, e.g., Adam Liptak & Jack Nicas, Supreme Court Allows Antitrust Lawsuit Against Apple to Proceed, N.Y. Times (May 13, 2019), https://www.nytimes.com/2019/05/13/us/politics/supreme-court-antitrust-apple.html; see also, e.g., Amazon.com Antitrust, Hagens Berman, https://www.hbsslaw.com/cases/amazon-antitrust ("An independent investigation by Hagens Berman’s legal team and expert antitrust attorneys has revealed that Amazon.com has violated federal antitrust price-fixing laws, causing consumers everywhere to pay artificially increased prices for products purchased via online retailers across the internet. Consumers have now filed a class action lawsuit against Amazon for driving up prices for online purchases made from other retailers.").

^28. Stigler Comm. Report, supra note 2, at 28.

^29. Gates, supra note 4, at 2 ("[A]nother unique aspect of the Internet is that because it buys communications lines on commodity bid basis and because it is growing so fast, it is the only ‘public’ network whose economics reflect the latest advances in communications technology. The price paid for corporations to connect to the Internet is determined by the size of your ‘onramp’ to the Internet and not by how much you actually use your connection. Usage isn’t even metered. It doesn’t matter if you connect nearby or half way around the globe. This makes the marginal cost of extra usage essentially zero encouraging heavy usage.").

^30. See, e.g., Richard A Posner, Antitrust Law 245-46 (The Univ. of Chicago Press 2nd ed. 2001); see also Tirole, supra note 25.

^31. Baker, supra note 2, at 123.

^32. Evans & Schmalensee, supra note 19, at 10.

^33. See id. at 10-11; see also Kate Collyer et al., Measuring Market Power in Multi-Sided Markets, Competition Pol’y Int’l Antitrust Chron., Sept. 2017, at 2 [hereinafter Collyer et al.].

^34. See, e.g., Evans & Schmalensee, supra note 19, at 11 ("Indirect network effects result in a positive feedback loop between the two sides."); see also Gates, supra note 4, at 2 ("[M]ost important is that the Internet has bootstrapped itself as a place to publish content. It has enough users that it is benefiting from the positive feedback loop of the more users it gets, the more content it gets, and the more content it gets, the more users it gets.").

^35. Stigler Comm. Report, supra note 2, at 38.

^36. Id. ("Multi-homing lessens network effects because a consumer can enjoy the size of both networks, rather than having to choose one.").

^37. See Collyer et al., supra note 33, at 2 ("[I]n a multi-sided market, the price structure reflects the interlinked demands of the two groups of consumers and the need to get both sides on board. This often results in complex pricing where the price to each group of consumers does not reflect the marginal cost of supplying them."); see also Evans & Schmalensee, supra note 19, at 12 ("An important consequence of this [price] interrelationship is that the prices charged to either group are not informative, by themselves, about the degree of competition for the platform.").

^38. Gates, supra note 4, at 3 ("[F]or users who connect to the Internet some way other than paying us for the connection we will have to make MSN very, very inexpensive—perhaps free. The amount of free information available today on the Internet is quite amazing. Although there is room to use brand names and quality to differentiate from free content, this will not be easy and it puts a lot of pressure to figure out how to get advertiser funding.").

^39. See Stigler Comm. Report, supra note 2, at 30 ("Digital platforms are characterized by free services. ‘Free’ is not a special zone where economics or antitrust do not apply. Rather, a free good is one where the seller has chosen to set a monetary price of zero and may set other, non-monetary, conditions or duties."); see also Fiona M. Scott & David C. Dinielle, Roadmap for a Digital Advertising Monopolization Case Against Google, Omidyar Network, 39 (May 2020), https://www.omidyar.com/sites/default/files/Roadmap%20for%20a%20Case%20Against%20Google.pdf ("[S]ervices provided by Google and other technology platforms provide extraordinary benefits to consumers and to society, free of charge.").

^40. See Stigler Comm. Report, supra note 2, at 31 ("The existence of zero money prices means that measurement of quality will be critical.").

^41. See Collyer et al., supra note 33, at 3 ("Many of our standard tools for assessing market power are more complex to apply in multi-sided markets and may need to be adapted."); see also Evans & Schmalensee, supra note 19, at 191 ("The economic literature analyzing two-sided platforms is new, complex, and evolving."); see also Stigler Comm. Report, supra note 2, at 63-64 ("On the one hand, targeted advertising to wise and well-informed consumers is welfare improving insofar as it allows advertisers to send the right information to the right people, improving their choices and fostering competition among suppliers. On the other hand, in the modern economy this simple model becomes more complex because of the cost to the consumer, namely loss of privacy. There is an open empirical question as to whether the tradeoff is worthwhile to consumers").

^42. A flavor of the likely arguments can be seen in Evans and Schmalensee, chapter 5 of which includes arguments put forth by various economists that the authors consider to be no more than "red herrings".

^43. Baker, supra note 2, at 184.

^44. See DOJ & FTC, Horizontal Merger Guidelines 8-9 (2010).

^45. Stigler Comm. Report, supra note 2, at 31-32 ("Market definition will vary according to what consumers are substituting between, whether there is competition on the platform between complements, or competition between platforms, or competition between a platform and potential or nascent competitors regarding possible future markets.").

^46. For a detailed discussion of the challenges of measuring market share and concentration in multi-sided markets, see Collyer et al.

^47. See generally Daniel S. Bitton & Stephen Lewis, Clearing up Misconceptions About Google’s Ad Tech Business, Australian Competition & Consumer Commission (2020), https://www.accc.gov.au/system/files/Google%20%20Report%20from%20Daniel%20Bitton%20and%20Stephen%20 Lewis%20%285%20May%202020%29.pdf.

^48. Compare Baker, supra note 2, at 185 with Evans & Schmalensee, supra note 19, at 27, 51-54.

^49. See Reference Manual on Scientific Evidence (Third) 439 (2011) ("[O]ne party’s damages analysis may hypothesize the absence of any act of the defendant that influenced the plaintiff, whereas the other’s damages analysis may hypothesize an alternative, legal act. This type of disagreement is particularly common in antitrust and intellectual property disputes."); see also Stigler Comm. Report, supra note 2, at 32 ("The need to identify the specific anticompetitive exclusionary conduct and analyze it may raise enforcement costs given all the possible variants of exclusionary conduct possible in digital markets.").

^50. See Stigler Comm. Report, supra note 2, at 29 ("[Digital] markets often have extremely strong economies of scale and scope due to low marginal costs and the returns to data. Moreover, they often are two-sided and have strong network externalities and are therefore prone to tipping. If so, the competitive process shifts from competition in the market to competition for the market. This combination of features means many digital markets feature large barriers to entry. The winner in these settings often has a large cost advantage from its scale of operations and a large benefit advantage from the scale of its data. An entrant cannot generally overcome these without either a similar installed base (network effects) or a similar scale (scale economies), both of which are difficult to obtain quickly and cost-effectively."); see also Tirole, supra note 25 ("That today’s information-technology markets are highly concentrated is beyond dispute. In most cases, a single company dominates a given market. There is nothing abnormal about this, as users are prone to flocking to just one or two platforms, depending on the service.").

^51. Nicholas Economides et al., Handbook of Antitrust Economics 485 (Pablo Buccirossi eds., MIT Press 2008) ("Because inequality is natural in the market structure of network industries, there should be no presumption that anticompetitive actions are responsible for the creation of market share inequality or very high profitability of a top firm. Thus, no anticompetitive acts are necessary to create this inequality. The ‘but for’ benchmark against which anticompetitive actions in network industries are to be judged should not be ‘perfect competition’ but an environment of significant inequality and profits.").

^52. Stigler Comm. Report, supra note 2, at 32 ("The need to identify the specific anticompetitive exclusionary conduct and analyze it may raise enforcement costs given all the possible variants of exclusionary conduct possible in digital markets.").

^53. Evans & Schmalensee, supra note 19, at 191 ("The economic literature analyzing two-sided platforms is new, complex, and evolving.").

^54. Jeffrey M. Perloff et al., Estimating Market Power and Strategies, at xi (Cambridge Univ. Press 2007) ("Economic theory alone cannot tell us how much market power firms exercise or which strategies they use. Thus, empirical work is critical if we are to understand how markets function.").

^55. Carlton and Perloff, supra note 16, at Ch. 11 § 4 ("Figuring out whether a product design is efficient or anticompetitive is likely to be a very hard task in usual cases. Especially if products are changing rapidly, the court could err greatly and could impose large costs on consumers — at least in the usual circumstances with the usual type of evidence.").

^56. See, e.g., Stigler Comm. Report, supra note 2, at 32 (recommending "the establishment of a specialist competition court to hear all private and public antitrust cases which would allow judges to develop some expertise.").

^57. See Tirole, supra note 25.

^58. See Evans & Schmalensee, supra note 19, at 36.

^59. See Collyer et al., supra note 33, at 3.

^60. See Stigler Comm. Report, supra note 2, at 34.

^61. Alfred Marshall, Principles of Economics: Abridged Edition 33 (8th ed. Cosimo Classics 2006) (1890) (emphasis added).

^62. See, e.g., Evans & Schmalensee, supra note 19, at 14-15; see also Stigler Comm. Report, supra note 2, at 40, 42.

^63. Conjoint analysis, a method for determining consumer valuations often used in the marketing literature, is one example of a technique from outside of economics that has been used more recently in litigation. For an informative introduction (centered around its use in patent damages litigation), see Gregory J. Sidak & Jeremy O. Skog, Using Conjoint Analysis to Apportion Patent Damages, Criterion (June 2016), https://www.criterioneconomics.com/docs/using-conjoint-analysis-to-apportion-patent-damages.pdf.

^64. See Stigler Comm. Report, supra note 2, at 31.

^65. Id. at 42.

^66. Id. at 67.

^67. Search algorithms that provide more relevant content, for example, benefit consumers over the full range of topics they care (i.e., value) enough to search for and "[w]hen [platforms] can identify individual tastes at fine levels and personalize their services to this taste, they often improve people’s lives. Search engines can better answer queries or find a nearby destination, cultural and news websites are able to suggest well-suited content, and ecommerce websites can improve matching between buyers and sellers. These are all part of the consumer benefit described previously." Stigler Comm. Report, supra note 2, at 48.

^68. Stigler Comm. Report, supra note 2, at 67 ("Another reason to be pessimistic about measuring traditional surplus concepts is related to the barter nature of the exchange: Users barter attention and personal data for services. With a ‘free’ service, consumers are paying for any expansion of activity with their attention to content. When facing a zero-money price, and when quality is difficult to observe, consumers are not receiving salient signals about the social value of their consumption because the price they believe they face does not reflect the economics of the transaction, and they are ignorant of those numbers."). Similarly, as Marshall pointed out, economists are in the same position—they don’t observe the price that their already-narrowly focused analyses require.

^69. Id. at 67 ("We caution, however, that the legal structure of US antitrust law is not well set up to accommodate this complexity as it opens the door for judges to weigh all manner of social concerns as well as traditional economic effects.").

^70. See Hovenkamp, supra note 25, at 109 ("Moreover, we are not particularly good at locating the line between anticompetitive and innovative practices. An overly aggressive antitrust rule would chill innovative conduct, so we frequently give the firm acting unilaterally the benefit of the doubt."); see also Reference Manual on Scientific Evidence, supra note 49, at xiv ("In the final analysis, a judge does not have the option of suspending judgment until more information is available, but must decide after considering the best available science. Finally, given the enormous amount of evidence to be interpreted, expert scientists from different (or even the same) disciplines may not agree on which data are the most relevant, which are the most reliable, and what conclusions about causation a re appropriate to be derived.").

^71. Stigler Comm. Report, supra note 2, at 27.

^72. Alfred Marshall and Jean Tirole recently have argued, with the right amount of imagination and humbleness, industrial organization and other (non-behavioral) economists can also evolve their tools kits to address the complexity of digital platform markets and antitrust issues. Marshall, supra note 61, at 43 ("The economist needs the three great intellectual faculties, perception, imagination and reason: and most of all he needs imagination, to put him on track of those causes of visible events which are remote or lie below the surface, and of those effects of visible causes which are remote or lie below the surface."); see also Tirole, supra note 25 ("Regulators and economists must be humble; they will learn by doing, and their policies should not be cast in stone.").