The United States might well be at an anti-monopoly moment. Many Republicans and Democrats believe that existing high-tech platforms have far too much economic power, control and influence, and that we should at least start thinking about imposing sanctions on all monopolies. There’s a good chance, however, that none of the currently proposed solutions will be able to curb most of these firms’ alleged monopoly power in an effective and timely manner. The case by the FTC and the States against Facebook has been dismissed without prejudice, although it will be refiled soon. Like DOJ and State cases against Google, these cases are likely to be quite large, and easily could take years to resolve. The enforcers will face formidable challenges to demonstrate that defendants engaged in anticompetitive conduct. Because of the size of these cases and the enforcers’ many other responsibilities, would either agency also have the necessary resources to mount an effective case against other possible monopolies, such as Amazon or Apple?
If only there were a faster, less resource-intensive way to curb monopolies’ powers! A solution that would not require additional legislation, or the detailed examination of so much of a company’s history necessary to demonstrate that it engaged in anticompetitive conduct. An approach that even a conservative Supreme Court might adopt because it follows from a textualist interpretation of Section 2 of the Sherman Act.
The optimal solution is no-fault monopolization (sometimes called per se monopolization). This article argues that Section 2 of the Sherman Act is a no-fault statute. It was designed to impose sanctions on all monopolies and attempts to monopolize, without inquiring into whether the firm achieved or attempted to achieve its monopoly using anticompetitive conduct. This conclusion emerges from the first-ever textualist interpretation of Section 2, a statute that does not contain any language requiring anticompetitive conduct.
Textualism is a form of statutory interpretation long championed by Justice Scalia, which today always is used by at least three Supreme Court Justices, and sometimes by all nine. Justice Scalia and Bryan Garner wrote a book explaining and analyzing textualism, and a relatively complete textualist analysis of Section 2 required a 92-page law review article. Nevertheless, a basic textualist analysis can be described relatively simply.
Textualism does not attempt to discern what Congress “intended to do” other than by examining the words and phrases in the statutes. A textualist analysis does not add or subtract from the statute’s exact language and does not create exceptions or interpret statutes differently in special circumstances. No requirement should be read into a law unless, of course, it is explicitly contained in the legislation. No exemption should be inferred to achieve some overall policy goal Congress arguably had, or to interpret a statute in a manner a court believes is reasonable.
Rather, textualism begins and ends its analysis by interpreting only the words and phrases used in the relevant statutes, with each word and phrase being given its fair, plain, ordinary, original meaning. To do this, textualism relies mostly upon the definitions contained in reliable dictionaries of the period in which the statute was enacted. This is supplemented by analyzing these terms as they were used in contemporaneous legal treatises and cases. Crucially, textualism ignores statutes’ legislative history (i.e., the associated floor debates and Congressional committee reports). The only two relevant complications are the literalness doctrine and the absurdity doctrine, which will be discussed below.
For Section 2 of the Sherman Act the textualist task produces a clear and straightforward result. Its violation does not require anticompetitive conduct.
Section 2 of the Sherman Act makes it unlawful for any person to “monopolize, or attempt to monopolize . . . any part of the trade or commerce among the several States . . . .” There is nothing, no language in Section 2, requiring anticompetitive conduct. A textualist interpretation of Section 2 therefore needs only to determine what the terms “monopolize” and “attempt to monopolize” meant in 1890. This examination will demonstrate that these terms meant the same things they would mean today if they were “fairly,” “ordinarily,” or “plainly” interpreted, free from the legal baggage that has grown up around them by a multitude of court decisions. These terms did not require anticompetitive conduct in 1890, and they should not today.
When the Sherman Act was passed the word “monopolize” simply meant to acquire a monopoly. The term was not limited to monopolies acquired or preserved by anticompetitive conduct.
As noted, Justice Scalia was especially interested in the definitions of key terms in contemporary dictionaries. Scalia and Garner believe that six dictionaries of the 1851 to 1900 period are “useful and authoritative.” All six were checked for definitions of “monopolize,” as were other dictionaries, including the Oxford English Dictionary, which Scalia and Garner consider useful and authoritative even though the volume containing “monopolize” was published somewhat later (in 1908).
The contemporary dictionaries all show that the principal definition for “monopolize” was simply that a firm had acquired a monopoly. None required anticompetitive conduct for a firm to “monopolize” a market. For example, the 1897 edition of Century Dictionary and Cyclopedia defined “monopolize” as: “1. To obtain a monopoly of; have an exclusive right of trading in: as, to monopolize all the corn in a district . . . .”
Second, and serendipitously, a definition of “monopolize” was given in the Sherman Act’s legislative debates, just before the final vote on the Bill. Although normally a textualist does not care about anything uttered during a congressional debate, Senator Edmund’s remarks should be significant to a textualist because he quotes from a contemporary dictionary that Scalia considered useful and reliable:
[T]he best answer I can make to both my friends is to read from Webster’s Dictionary the definition of the verb ‘to monopolize’: 1. To purchase or obtain possession of the whole of, as a commodity or goods in market, with the view to appropriate or control the exclusive sale of; as, to monopolize sugar or tea . . . .
There was no requirement of anticompetitive conduct.
These definitions are essentially the same as those in the 1898 and 1913 editions of Webster’s Dictionary. The four other dictionaries of the period considered reliable by Scalia and Garner contained essentially identical definitions. In addition, the first edition of the Oxford English Dictionary contained a similar definition of “monopolize:” “1 . . . . To get into one’s hands the whole stock of (a particular commodity); to gain or hold exclusive possession of (a trade); . . . . To have a monopoly. . . . 2 . . . . To obtain exclusive possession or control of; to get or keep entirely to oneself.” Not only does the Oxford English Dictionary equate “monopolize” with “monopoly,” but nowhere does it require that a monopolist have engaged in anticompetitive conduct.
In sum, all of the surveyed roughly contemporaneous dictionaries define “monopolize” as simply to gain a monopoly. A few definitions did give specific examples of how a firm might become a monopoly by anticompetitive means, or by engrossing (which is similar to merging), and some used a phrase like “controlling the market” which is ambiguous as to whether anticompetitive conduct must be involved. None, however, require conduct we would today characterize as anticompetitive, or say that a firm gaining a monopoly by efficient conduct had not “monopolized.” A textualist analysis of contemporary legal treatises and cases yields the same result.
Would a textualist interpretation of Section 2 only mean that firms with a 100 percent market share could illegally “monopolize”? After all, dictionaries of the period defined “monopolize” using such words such as “exclusive,” “the whole of,” and “all.” If so, this would dramatically limit monopolization law.
However, Justice Scalia warned that “the good textualist is not a literalist.” Scalia and Garner explain that the “notion that words should be strictly construed” is untrue. They quote Justice Frankfurter: “Literalness may strangle meaning.” As Justice Kavanaugh observed in Bostock, “Courts must follow ordinary meaning, not literal meaning.” Thus, a textualist would ask how an ordinary person in 1890 would have used the word “monopolize.”
It is noteworthy that the 1895 Supreme Court’s E.C. Knight decision referred to a firm with a 98 percent market share as a “monopoly.” Moreover, if a 100 percent market share were required, a potential defendant often could render the monopolization offense a nullity by deliberately leaving 2 percent of a market to others. As Justice Scalia noted, “Some outcome-pertinent consequences . . . are relevant to a sound textual decision––specifically, those that: (1) cause a private instrument or government prescription to be ineffective.”
The Supreme Court in 1895 was correct to characterize a firm with a 98 percent market share as a “monopoly.” But it is a difficult judgment call as to when a fair, plain, ordinary reading of “monopolize” would consider a firm’s market share too low to qualify. Could a firm “monopolize” with only the 88 percent share of the oil market held by Standard Oil in 1890? If courts insisted on an extremely high minimum market share this would dramatically limit the reach of the monopolization offense. Textualism would call for a minimum or usual minimum market share line or range, but where? It would be as uncertain as the market share requirement for monopolization today, although it should be higher than the current 50–70 percent minimum.
” A textualist interpretation of Section 2 should analyze the word “attempt” as it was used in the phrase “attempt to monopolize” circa 1890. However, no unexpected or counterintuitive result comes from this examination. Circa 1890 “attempt” had its colloquial 21st century meaning, and there is no requirement in the statute that an “attempt to monopolize” requires anticompetitive conduct.
The “useful and authoritative” 1897 Century Dictionary and Cyclopedia defines “attempt” as: “1. To make an effort to effect or do; endeavor to perform; undertake; essay: as, to attempt a bold flight . . . . 2. To venture upon: as, to attempt the sea.— 3. To make trial of; prove; test . . . . .” The 1898 Webster’s Dictionary gives a similar definition: “Attempt . . . 1. To make trial or experiment of; to try. 2. To try to move, subdue, or overcome, as by entreaty.” The 1913 edition of Webster’s Dictionary contains a similar definition. The Oxford English Dictionary, which defined “attempt” in a volume published in 1888, similarly reads: “1. A putting forth of effort to accomplish what is uncertain or difficult . . . .”
However, the word “attempt” in a statute did have a specific meaning under the common law circa 1890. It meant “an intent to do a particular criminal thing, with an act toward it falling short of the thing intended.” One definition stated that the act needed to be “sufficient both in magnitude and in proximity to the fact intended, to be taken cognizance of by the law that does not concern itself with things trivial and small.” No source of the period, however, defined the magnitude or nature of the necessary acts with great specificity (indeed, a precise definition might be impossible).
It is noteworthy that in 1881 Oliver Wendell Holmes wrote about the attempt doctrine in his celebrated treatise, The Common Law:
Eminent judges have been puzzled where to draw the line . . . the considerations being, in this case, the nearness of the danger, the greatness of the harm, and the degree of apprehension felt. When a man buys matches to fire a haystack . . . there is still a considerable chance that he will change his mind before he comes to the point. But when he has struck the match . . . there is very little chance that he will not persist to the end . . . .
Congress’s choice of the phrase “attempt to monopolize” surely built upon the existing common law definitions of an “attempt” to commit robbery and other crimes. Although the meaning of a criminal “attempt” to violate a law has evolved since 1890, a textualist approach towards an “attempt to monopolize” should be a “fair” or “ordinary” interpretation of these words as they were used in 1890, ignoring the case law that has arisen since then. It is clear that acts constituting mere preparation or planning should be insufficient. Attempted monopolization should also require the intent to take over a market and at least one serious act in furtherance of this plan. But “attempted monopolization” under Section 2 should not require anticompetitive conduct.
The Sherman Act has always contained criminal penalties. Do these possible sanctions mean that a no-fault interpretation of Section 2 should be nullified by the textualist doctrine that no statute should be interpreted in an “absurd” manner, which Justice Scalia described as something that was “obviously a technical or ministerial” drafting mistake that was “so monstrous, that all mankind would, without hesitation, unite in rejecting” its application. In fact, the possibility of criminal sanctions does not disprove that Section 2 is a no-fault statute, for three reasons.
First, it might strike many as absurd to imprison owners of tiny firms in a large city who fix prices if they have no market power. Nevertheless, we certainly do want to imprison the executives of large companies who fix prices. Similarly, it might be absurd to imprison a small-town monopolist, one that by 1890 standards might not even have affected interstate commerce, under a no-fault approach. But in 1890 Congress might not have thought it absurd to imprison a “robber baron” like John D. Rockefeller without examining whether he committed anticompetitive acts. During the relevant period many politicians called, not for his trial, but simply his imprisonment.
Second, enforcers have attempted to prosecute some other antitrust violations criminally, even though today most believe these criminal prosecutions were incredibly unwise on policy grounds. Few today believe that violations of the Robinson-Patman Act or the prohibition against resale price maintenance should result in criminal sanctions, yet prosecutors sought them.
Third, there are plausible reasons why Congress could have included criminal penalties in the Sherman Act, yet never intended criminal enforcement. After all, no corporate official has ever been imprisoned for any Section 2 offense requiring anticompetitive conduct, yet the statute unquestionably is valid in civil cases that require anticompetitive conduct. Section 2’s framers might well have included criminal provisions without giving them much thought, only as a threat that might deter improper conduct, or for political posturing purposes.
As Justice Scalia wrote, “once the meaning is plain, it is not the province of a court to scan its wisdom or its policy.” Indeed, if a court were to do so, this would be the antithesis of textualism. A straightforward textualist interpretation of Section 2 demonstrates that a violation does not require anticompetitive conduct, even though in theory a prosecution could result in criminal penalties.
Effects on Monopolization Law. Justice Scalia’s opinion in Trinko held that anticompetitive conduct is required for a Section 2 violation, but his opinion did not undertake a textualist analysis of the statute. Nor did it explain how or why the word “monopolize” should be interpreted to contain a conduct requirement. Scalia instead simply cited precedent for this holding.
A textualist no-fault interpretation of Section 2 would overturn this and the other precedent that has been controlling for at least 55 years. In the Leegin case, however, the Supreme Court overturned its own antitrust precedent after 96 years. Justice Scalia voted with the majority because of their belief that economic learning concerning how often this practice is anticompetitive had changed.
Surely the changing state of economic knowledge should count for less than a fair reading of the relevant statutory language. The Supreme Court that decided to revise antitrust law in Leegin should be even more willing to overturn Trinko and related cases.
Because Section 2 does not explicitly require anticompetitive conduct, its violation should only require defendants to engage in a serious “attempt” to gain a monopoly. This should require only the intent to take over a market, and one concrete, significant act in furtherance of this intent. The act could be required to be “sufficient both in magnitude and in proximity to the fact intended… [and not merely] trivial and small.”
The attempted monopolization doctrine should be applied only to firms attempting to take over an entire market (or since, as noted, a textualist should not construe the word “monopoly” literally or strictly, to a firm attempting to take over virtually all of a market). Suppose, for example, a firm with 40 percent of a relevant market conceived of and was attempting to implement a serious plan that would, if successful, give it 95 percent of that market. This should violate Section 2. Although courts might draw a minimum market share line or range and include an intent requirement, this should encompass more cases than the requirements courts impose today, which include both anticompetitive conduct and defendants having a “dangerous probability” of acquiring monopoly power.
This article will not attempt to analyze whether a textualist analysis of Section 5 of the FTC Act should conclude that no-fault monopolization violates this statute. However, it is noteworthy that the Supreme Court has interpreted this statute to prohibit not only every violation of the Sherman Act, but also (1) incipient violations of this law, (2) conduct violating the spirit or policy of the Sherman Act, and (3) conduct violating recognized standards of business behavior.
Using this precedent, no-fault should be found to violate Section 5 if the Court undertakes and then builds upon a textualist analysis of Section 2. The Court could, moreover, decide that no-fault violates Section 5 even if it is reluctant to overturn Section 2 precedent, by declaring no-fault monopolization to be a violation of the spirit or policy, but not the letter, of the Sherman Act. Further, only government enforcers can bring cases under the FTC Act, and violations do not result in criminal penalties. If the Court wants to be certain that no-fault cases cannot lead to criminal penalties or private suits, it can decide that no-fault violates Section 5 because it violates the spirit or policy, but not the letter, of Section 2.
As noted, Justice Scalia cautions that a textualist should not interpret a statute in a manner that would reach an “absurd” conclusion. Significantly, he advocates a narrow version of this doctrine that would limit it to technical or drafting errors that “all mankind” would agree would lead to an absurd result. Even a brief overview of the economic effects of no-fault monopolization demonstrates, however, that this doctrine would not be absurd. Indeed, its overall economic effects are uncertain, potentially offsetting, and depend upon empirical issues whose net effect is speculative or ambiguous. There is, moreover, substantial reason to believe its effects are likely to be beneficial on the whole.
No-fault monopolization and attempted monopolization could improve economic welfare in many ways. It should increase innovation and international competitiveness. It should prevent the allocative inefficiency effects of monopoly pricing and the form of exploitation that arises when monopolies acquire wealth from consumers. It would be likely to decrease the inefficiencies that result from monopolists enjoying a “quiet life,” and also the waste that arises as firms attain and protect their monopolies. It should significantly reduce the time and costs of Section 2 litigation. It should improve privacy and decrease income inequality.
The standard admittedly would also lead to costs and difficulties. Imposing sanctions on all monopolies could sometimes send a confusing or perverse signal to firms engaging in hard but fair competition, especially as a firm’s market share in some plausible relevant market neared the ambiguous level required for a violation. It could enable competitors to file baseless lawsuits. The transaction costs involved in imposing sanctions on monopolies could be significant. It also could lead to difficult remedy issues in cases involving natural or patent monopolies where conduct or regulatory orders, rather than break-up, would be ordered. There are reasons to believe, however, that the benefits of no-fault are likely to outweigh the costs.
A related but non-economic objection to no-fault is that it is simply “unfair” to sanction a firm unless it engaged in anticompetitive conduct. This idea was perhaps best expressed in Alcoa: “The successful competitor, having been urged to compete, must not be turned upon when he wins.” To many it seems “unfair” to find someone guilty of something that was “not their fault,” and to impose even a mild remedy that might discourage other firms from innovating and engaging in hard but fair competition.
There is a lot of merit to these concerns. Nevertheless, the “unfairness” argument also weighs against monopolies because they usually “unfairly” raise consumer prices above the competitive level. And although no-fault could discourage firms from innovating to become a monopolist, monopolies also tend to innovate less overall than firms operating in competitive markets. So the innovation argument cuts both ways. Moreover, some dislike monopolies for non-economic reasons, believing their “social, political, and economic effect on the nation can only be detrimental.” Considering all these conflicting arguments, a reasonable person could conclude that on the whole it would not be “unfair” to impose sanctions on all monopolies.
Imposing sanctions on all monopolies can best be characterized by analogy to antitrust’s per se rule. When Congress made it illegal “to monopolize” it decided to impose sanctions on every monopoly, without examining how the monopoly arose. It would be more accurate to call this approach to Section 2 “monopolization per se” rather than “no-fault monopolization.”
The use of “monopolization per se” would signify that the court has not undertaken a rule of reason balancing of the procompetitive and anticompetitive effects of each monopoly. Rather, the court would be implementing the decision Congress made to impose sanctions on all monopolies. When it enacted Section 2, Congress performed an overall assessment and determined that, on the whole, the harms from monopolies usually outweigh their benefits. To save time and resources, to give notice and enhance predictability, the courts should give effect to this Congressional desire by adopting the “monopolization per se” label.
Regardless of the name this approach is given, the antitrust field has not seriously undertaken a full policy analysis of it in half a century. This would be a huge undertaking that would require considerations of a large amount of theory and considerable empirical work. Even this article’s brief overview demonstrates, however, that this is a topic that deserves careful analysis and debate by the antitrust community.
It should be emphasized, however, that the standard under which Section 2 should be interpreted to be a no-fault statute is not whether this is the best possible policy option. Because it follows from a textualist reading of the Sherman Act, the standard is whether a no-fault interpretation would be so “absurd” that it must be the result of a “technical or ministerial” drafting mistake that was “so monstrous, that all mankind would, without hesitation, unite in rejecting” its application. Because no-fault instead is an issue upon which reasonable people can disagree, the absurdity doctrine does not apply. Section 2 should be considered a no-fault monopolization (or “monopolization per se”) statute.
Where did the requirement that anticompetitive conduct is necessary for a Section 2 violation originate? Shouldn’t we admit that it arose simply because some judges thought it would be a good policy position and simply added it to the statute? This is the opposite of textualism, whose essence is to start and end with the text of a statute. Since Section 2 does not contain an anticompetitive conduct requirement, a textualist interpretation of Section 2 should not require one.
Textualist analysis does, however, require consideration of the literalness doctrine and the absurdity doctrine. Justice Scalia cautions that these doctrines should be interpreted narrowly, and should not be used by a court to impose its policy judgement or its notion of reasonableness on the statute. Indeed, if these doctrines are used over-expansively, they would negate textualism.
If courts employ textualist analysis and conclude that Congress wanted to impose sanctions on all monopolies, they also should re-think the appropriate name for this cause of action. The term “no-fault” implies the reviewing court examined whether the firm did anything wrong, and concluded that it did not. But this is not what should happen in a no-fault case. A court should not inquire into whether defendant engaged in anticompetitive conduct, because this is not relevant to a violation. For this reason a better name for the offense is “monopolization per se.” If the court finds that defendant is a monopoly, it should hold that Section 2 has been violated. Period. Without examining defendant’s conduct.
The author thanks Neil Averitt, John Kirkwood, and Randy Stutz for wise advice, and Cassandra Brumback, Laura Grant, Michael Hart, and Kristin McManus for excellent research assistance. All errors and opinions are those of the author.
