In this essay, I pursue two paths. In the first, I revisit the beginnings of U.S. antitrust law to emphasize that that law has long had a policy of permitting firms to grow organically into dominant market positions. The Sherman Act created an anti-trust policy, not a broad anti-monopoly policy. And that remained true even as U.S. antitrust law moved in 1914 to supplement the Sherman Act with The Federal Trade Commission Act and The Clayton Act. I then turn to the second path. The history described in the first section suggests that, by design, there are fundamental limits on how U.S. antitrust law can respond to firms that have achieved great, legitimate success. But firms can stray and that brings them within antitrust. Case settlements move faster than litigation and put on the table many of the tools that we associate with regulation, such as nondiscrimination obligations, access rights, and line-of-business restrictions.
By Randal C. Picker1
Sue them? Regulate them? Both? The rise of the big tech firms has created a moment of possible change in how those firms are regulated. You really do need a scorecard to keep track of the pending antitrust suits and investigations of Amazon, Apple, Facebook and Google. (Microsoft, so far, seems to be mainly below the radar). The same is true of pending bills in the U.S. Congress. It is easier in Europe, where both the Digital Services Act and Digital Markets Act are under consideration, though as soon as we turn to the member states possible responses grow quickly, where Germany has been an early mover. If we switch to Asia, South Korea has moved to regulate app stores. This is platform regulation organized around the gatekeeping positions of the big tech firms.
Antitrust and regulation are different approaches to possible controls over these firms. U.S. antitrust laws are organized around ideas of fault, market definition, and market power. Litigation in the U.S. is a slow path to change and the same is true in Europe. New laws offer the promise of a quick regime change, though the lessons of regulatory statutes like the Telecommunications Act of 1996 make clear that vague statutes can also lead to litigation timelines measured in decades over the language of new statutes.
In this essay, I pursue two paths. In the first, I revisit the beginnings of U.S. antitrust law to emphasize that that law has long had a policy of permitting firms to grow organically into dominant market positions. The Sherman Act created an anti-trust policy, not a broad anti-monopoly policy. And even as U.S. antitrust law moved in 1914 to supplement the Sherman Act with The Federal Trade Commission Act and The Clayton Act, leading voices of that era — soon-to-be-Justice Louis Brandeis and soon-to-be-President Woodrow Wilson — made clear that even though they opposed the trusts, they were not opposed to firms that achieved their market positions though legitimate competition and organic growth.
I then turn to the second path. The history described in the first section suggests that, by design, there are fundamental limits on how U.S. antitrust law can respond to firms that have achieved great, legitimate success. But firms born through innovation do not necessarily always stay on the righteous path and that, appropriately and again by design, brings them within the grasp of antitrust. Case settlements can be powerful tools in part because they can move faster than full-tilt litigation with trials and appeals and in part because settlements put on the table many of the tools that we associate with regulation, such as nondiscrimination obligations, access rights, and line-of-business restrictions.
As that suggests, there is some overlap between how antitrust can operate and how new platform legislation is likely to be framed. The AT&T cases in the U.S. provide a useful touchstone here, as antitrust and legislation moved in parallel even as they pursued similar ends with many of the same instruments. Right now, in Europe, the momentum behind legislation there almost certainly reflects frustration over the inability of the competition policy authorities to achieve more on-the-ground competitive changes even as they have, so far at least, issued a series of fines and rulings against Google and in an earlier era against Microsoft. In the U.S., the public cases against, so far, Amazon, Facebook and Google are still quite young and possible legislation seems to have more momentum as both Republicans and Democrats express often different frustrations with the big tech firms. That said, the big tech firms have achieved their positions by providing products valued by the public and regulation, via antitrust or through new statutes, should ensure that the value of those products is not lost through new clumsy rules.
I. LEGITIMATE MONOPOLIES IN EARLY U.S. ANTITRUST LAW
The best place to start to understand what Senator Sherman hoped to accomplish is with the text of the legislation he introduced. On August 14, 1888, Sherman introduced S.3445 into the 1st Session of the Fiftieth Congress and the bill was referred to the Senate Committee on Finance. The title of the bill — “To declare unlawful trusts and combinations in restraint of trade and production” — gives a good sense of Sherman’s central goal. In the first session of the next Congress, Sherman once again introduced his bill, now as S.1. On March 18, 1890, Sherman introduced a revised version of the bill to navigate possible objections that had been raised regarding the constitutionality of the bill.1
But the core of Sherman’s approach had not changed between 1888 and 1890. Sherman was focused on joint activity, not single firm activity. His bill targeted “arrangements, contracts, agreements, trusts, or combinations between two or more citizens or corporations.” Which of those were declared “to be against public policy, unlawful and void?” Those which were made with a view to, or which tended “to prevent full and free competition” relating to imports or in “articles of growth, production, or manufacture.” In addition to those limits, Sherman wanted to declare unlawful “arrangements, trusts or combinations between such citizens or corporations” “made with a view to or which tend to advance the cost to the consumer of any such articles.” Sherman was looking to ban joint activity that interfered with competition, or which raised prices to consumers.
Sherman clearly was targeting the rise of the trusts, hence the characterization of this proposed bill as “anti-trust.” The trusts represented an effort by capitalists to bring competing enterprises together in a single great organization. Unsurprisingly, trusts had their defenders and Samuel Dodd, the lawyer for and brains behind the Standard Oil Trust, offered a spirited defense of trusts in the New York Daily Tribune on February 2, 1890. But Dodd was running against the popular tide.3
Sherman lost control of his bill in the Senate and on March 27, 1890, it was referred to the Senate Committee on the Judiciary. A week later, on April 2, 1890, Senator Edmunds came to the full Senate with an amended version of Sherman’s bill. The amendment struck all of Sherman’s language — really all — and the new language would become what we think of as the now-familiar language of the “Sherman” Act when it was enacted on July 2, 1890. And as is probably clear from the discussion so far, Sherman’s original bill did not use the word “monopoly” or “monopolize” and only pursued joint activities and not those by single firms.4
On April 8, 1890, the Senate turned to consider the Judiciary Committee draft. Sherman himself had exited the debate, saying that he would vote for the new draft deeming it “the best under all the circumstances that the Senate is prepared to give in this direction.” But even as the Senate was moving to vote, Senator Kenna raised a question about the meaning of the new language in Section 2. Kenna posed a hypothetical involving someone who “by his own skill and energy, by the propriety of his conduct generally, shall pursue his calling to in such a way as to monopolize a trade.” As Kenna continued to talk, he ran at the hypo again, focusing on someone “who happens by his skill and energy to command an innocent and legitimate monopoly of a business.” The proposed new statute would be a criminal statute and Kenna wanted to know if this behavior was a crime.5
Senator Edmunds who had brought the revised draft from the Judiciary Committee assured Senator Kenna that his hypos were outside the scope of Section 2: “Anybody who knows the meaning of the word ‘monopoly,’ as the courts apply it, would not apply it to such a person at all; and I am sure that my friend must understand that.” Edmunds would quickly take a second run at his answer: “It does not do anything of this kind, because in the case stated the gentlemen has not any monopoly at all. He has not bought off his adversaries. He has not got possession of all the horned cattle of the United States. He has not done anything but compete with his adversaries in trade if he had any, to furnish the commodity for the lowest price. So, I assure my friend that he need not be disturbed upon the subject.”6
The byplay on the floor of the Senate between the two senators gave a sense of what was excluded and included under the new language of Section 2. Legitimately obtained monopoly was outside of Section 2 and of course Section 1 only addressed joint activity and so would not cover single-entity monopoly. Monopolization required more, such as buying off adversaries or some other illegitimate business act.
Jump to the presidential election of 1912. The trusts had continued to be the subject of inquiry and possible legislation. And in its 1911 ruling in Standard Oil, the U.S. Supreme Court had resolved a key interpretative question regarding Section 1 of the Sherman Act in concluding that that section barred only unreasonable restraints of trade. Louis Brandeis, who would join the Supreme Court on June 1, 1916, was then a private lawyer in Boston but he was a central participant in the discussion about what to do about the trusts.7
Brandeis drew a clear line between the trusts — a tool that typically brought competing businesses together to limit competition — and what Brandeis sometimes termed “natural” monopoly. Those were firms that had achieved their position not through combination but rather through natural, organic, internal growth. Firms that had succeeded through legitimate competition in the marketplace. Brandeis was skeptical that the latter really existed, but he was clear about how more efficient firms should be treated: “There is nothing in our industrial history to indicate that there is any need whatever to limit the natural growth of a business to preserve competition. We may emphatically declare: ‘Give fair play to efficiency.’”8
Brandeis’s framing gives rise to an obvious question: if we were to focus on the large industrial enterprises of his era, what was the mix between the competition-limiting trusts and firms that had achieved their position through organic growth? I do not know the answer to that question, but even as Brandeis was speaking and writing, there was one prominent example of Brandeis’s so-called natural monopoly: the Aluminum Company of America. On May 19, 1912, The New York Times offered an update on five recent cases: The Powder Trust; The Standard Oil Trust; The Tobacco Trust; The Electric Lamp Trust; and what it termed The Aluminum Trust. Even as it described — almost certainly mistakenly — Alcoa as a trust, it explained why the government had not moved to dissolve the company: “That company is not a combination of former competitors but has obtained a practical monopoly of the business through its own growth, with valuable patents and almost complete control of known deposits of bauxite, the base of aluminum.” A business built on government patents to be sure but built on its own internal growth and success.9
Woodrow Wilson would win the 1912 presidential election. One of the issues that he campaigned on was the trust issue and he would subsequently publish a book of those speeches, framed in the overarching vision of what he called The New Freedom. In Chapter VIII of that book, Wilson addressed Monopoly, or Opportunity?. Even as Wilson highlighted the risks posed to society by the trusts, especially in his view the money trust or what he said really was the credit trust, Wilson repeatedly distinguished his views on the trusts from those of big businesses that grew organically through legitimate competition. Wilson announced that “I am for big business, and I am against the trusts.” Wilson did not fear big businesses: “I admit that any large corporation built up by the legitimate processes of business, by economy, by efficiency is natural; and I am not afraid of it, no matter how big it grows.” And Wilson closed the chapter by returning to those ideas: “I know, and every man in his heart knows, that the only way to enrich America is to make it possible for any man who has the brains to get into the game. I am not jealous of any business that has grown to that size. I am not jealous of any process of growth, no matter how huge the result, provided the result was indeed obtained by the processes of wholesome development, which are the processes of efficiency, of economy, of intelligence, and of invention.”10
II. THE INTERSECTION OF ANTITRUST AND PLATFORM REGULATION
There are a few takeaways from that brief look at the origins of U.S. antitrust law. It was, first and foremost, anti-trust law. The trusts were artificial combinations that limited competition and restoring that competition was the goal. And government-created monopolies bestowed by kings and queens were condemned. But there was no general condemnation of monopoly as such as is clear in the through line from the 1890 Senate floor debates over the draft bill to Brandeis and Wilson during the 1912 election. A firm could compete legitimately and grow to dominate its market. Size as such was not the issue, but how the firm behaved mattered for whether liability would be found.
That suggests that there was a core limit on what antitrust can do against firms that grew to dominate their markets. But, at the same time, successful firms could overstep and face antitrust suits, and while winning those suits would usually require a showing of fault, settlements can be powerful levers for changing market competition. Take one prominent example: the U.S. government’s settlements with AT&T in 1956 and 1982. Both of those were consensual settlements cut between AT&T and the U.S. government, and, given that, neither required a full-blown finding of fault. It was enough that a finding of fault was possible and that both sides saw a deal as the way forward to resolve the situation.
In both situations, it is worth paying attention to the timelines of the settlements. On January 15, 1949, the U.S. government brought an antitrust suit against AT&T seeking to split up AT&T and to force AT&T to license certain of its patents on reasonable and nondiscriminatory terms. Seven years later, on January 24, 1956, the government and AT&T settled the case. Seven years for a settled case. The 1956 AT&T final judgment had two core provisions both of which are of the sort that we often associate with platform regulation. The first was an access provision that opened AT&T’s patents on extremely favorable terms, while the second was a line-of-business restriction that boxed in AT&T in communications markets and excluded it from other markets, particularly the young computing market. Both of those provisions almost certainly had important consequences. The patents licensing regime boosted innovation, while barring AT&T from computing almost certainly made it easier for IBM to build up its market position.11
On November 21, 1974, the U.S. government once again brought an antitrust lawsuit seeking to break up the firm. As that suit continued to run, Congress considered new laws to address telecommunications competition. On October 7, 1981, by a 90-4 vote, the Senate approved a draft bill that would have resolved many of the pending competition issues and if the House of Representatives had move forward, it was likely the government would have dropped the pending lawsuit.12
But on January 8, 1982, the U.S. government and AT&T announced a settlement of the pending antitrust suit, this time one that would breakup AT&T. AT&T would eventually be split into eight large firms: AT&T proper as a long-distance company freed of most of the restrictions of the 1956 final judgment and seven regional Bell operating companies (“RBOCs”) providing local telephone service. The settlement created nondiscrimination obligations to boost long-distance competition while also imposing line-of-business restrictions on the RBOCs and much more limited ones on AT&T.13
The AT&T case frames the current situation with the big tech firms nicely. The U.S. government and the various states have antitrust suits pending against Amazon, Facebook and Google. There are private antitrust suits against Apple and rumors of a possible government suit against Apple. The European Commission has investigations pending against Amazon and Apple and ongoing appeals of decisions regarding Google. Full litigation of these cases will likely take years, though as the AT&T timeline makes clear, even settlements of cases may come slowly.
Just as was the case with AT&T, legislation is a natural alternative. In many ways, as was the case with AT&T, negotiated settlements and the likely legislative outcomes might run along similar lines. The pending bills in the U.S. Congress and their counterparts in Europe suggest the likely tools are the traditional tools that we use to regulate network industries, including a mix of nondiscrimination rules, access rules and business line restrictions. As I have argued elsewhere, getting this right is tricky and depending on how it is framed of course, legislation could be quite disruptive to the world that consumers currently live in.14
I have not gone through the pending antitrust cases with enough care to assess the merits of the claims against the big tech firms. But the positions of these firms — Amazon, Apple, Facebook, and Google — reflect in many ways the type of innovation and efficiency that the framers of the Sherman Act and then subsequently Brandeis and Wilson believed should be free of antitrust liability. That of course is not to say that those firms have not violated the antitrust laws — again, I have not expressed a view on that here — but rather that there is core there that we should be sensitive to, whether changes are pursued through antitrust remedies, imposed or agreed, or through legislation.
1 James Parker Hall Distinguished Service Professor of Law, The University of Chicago Law School.
2 A New Anti-Trust Bill, The New York Times, Mar. 19, 1890, p6.
3 Samuel Dodd, A Defence of Trusts, New York Daily Tribune, Feb. 2, 1890, p18.
4 For a deeper dive into this background, see William Kolasky, Senator John Sherman And the Origin of Antitrust, Antitrust, 24:1, Fall 2009, p85-89. Before Sherman introduced S.3445 in 1888, on July 10, 1888, he introduced a resolution that called for the Senate Committee on Finance to investigate the issues at stake in the legislation that he would subsequently propose. The resolution was framed around tariff policy, but that resolution, which paralleled in many ways S.3445, did use the word “monopoly” but that language was dropped in S.3445 and S.1. See 19 Cong. Rec. 6041 (1888).
5 21 Cong. Rec. 3145 (1890); 21 Cong. Rec. 3151 (1890).
6 21 Cong. Rec. 3151-3152 (1890). The back and forth between Senators Kenna and Edmunds has received a substantial amount of attention before. It was important in Bork’s 1966 examination of the legislative history of the Sherman Act. See Robert H. Bork, Legislative Intent and the Policy of the Sherman Act, 9 J.L. & Econ. 7, 29-31 (1966). For commentary and disagreement with Bork, see Herbert Hovenkamp, Distributive Justice and the Antitrust Laws, 51 Geo. Wash. L. Rev. 1 (1982); Christopher Grandy, Original Intent and the Sherman Antitrust Act: A Re-examination of the Consumer-Welfare Hypothesis, 53 J. Econ Hist. 359 (1993). See also Nicola Giocoli, Free from what? Competition, regulation and antitrust in American economics, 1870-1914 (published in Luciano Fanti, Oligopoly, Institutions and Firms’ Performance (Pisa Univ. Press 2017)).
7 The Standard Oil Co. of New Jersey v. The United States, 221 U.S. 1 (1911).
8 See Louis D. Brandeis, Shall We Abandon the Policy of Competition, 18 Case & Comm. 494 (Feb 1912) and Louis D. Brandeis, Competition, 44 Amer. Leg. News 5 (Jan 1913), both of which are reprinted in Osmond K. Frankel, ed., The Curse of Bigness: Miscellaneous Papers of Louis D. Brandeis (The Viking Press 1934).
9 Five Trust Decrees, The New York Times, May 19, 1912, p13.
10 Wilson Pillories Steel Trust in Gary, The New York Times, Oct. 5, 1912, p6; Woodrow Wilson, The New Freedom, pp. 166, 180-81, 191 (Doubleday, Page & Co. 1918).
11 Charles Zerner, U.S. Sues to Force A.T.&T. To Drop Western Electric Co., The New York Times, Jan. 15, 1949, p.1; Anthony Lewis, A.T.&T. Settles Antitrust Case; Shares Patents, The New York Times, Jan. 25, 1956, p1. On the innovation consequences of the patent access rules, see Martin Watzinger, Thomas A. Fackler, Markus Nagler & Monika Schnitzer, How Antitrust Enforcement Can Spur Innovation: Bell Labs and the 1956 Consent Decree, 12 Amer. Econ. J.: Econ. Policy 328 (2020).
12 Merrill Brown, Senate Oks Bill Allowing AT&T To Enter Unregulated Industries, The Washington Post, Oct. 8, 1981, pD11.
13 Eileen Shanahan, U.S. Sues to Divest A.T.&T. Of Western Electric Co., Charges Wide Conspiracy, The New York Times, Nov. 21, 1974, p1; Ernest Holsendolph, U.S. Settles Phone Suit, Drops I.B.M. Case; A.T.&T. to Split Up, Transforming Industry, The New York Times, Jan. 9, 1982, p1.
14 Randy Picker, Forcing Interoperability on Tech Platforms Would Be Difficult to Do, ProMarket.org, Mar. 11, 2021; Randy Picker, Europe Lacks a Vision for How Apple’s App Store Fees Should Work, ProMarket.org, May 5, 2021; Randy Picker, The House’s Recent Spate of Antitrust Bills Would Change Big Tech as We Know It, ProMarket.org, June 29, 2021; Randal C. Picker, Security Competition and App Stores, Concurrentialiste Antitrust Law, Aug. 23, 2021.