One of the most challenging issues in the modern law of economic regulation concerns the much-discussed overlap, if not conflict, between patent and antitrust law. The source of this problem lies in two short texts, each of which bristles with its own theoretical and interpretive difficulties. The first text is Article I, Section 8, Clause 8, of the United States Constitution, under which Congress has the power:
To promote the Progress of Science and useful Arts, by securing for limited Times to Authors and Inventors the exclusive Right to their respective Writings and Discoveries.
The second canonical text is the 1890 Sherman Act, specifically Sections 1 and 2, which reads:
Every contract, combination in the form of trust or otherwise, or conspiracy, in restraint of trade or commerce among the several states, or with foreign nations, is declared to be illegal. . . .
Every person who shall monopolize, or attempt to monopolize, or combine or conspire with any other person or persons to monopolize any part of the trade or commerce among the several States, or with foreign nations shall be deemed guilty of a felony.
This chapter explores the interrelationship between these two basic provisions, both as a matter of general theory and through their development in case law over the past 130 years—which spans multiple eras of technological innovation. It is easy to find cases where patent law appears to move in one direction and antitrust law in the opposite. But as a general matter, this chapter defends the thesis that, as the Federal Circuit has written, “[t]he patent and antitrust laws are complementary, the patent system serving to encourage invention and the bringing of new products to market by adjusting investment-based risk, and the antitrust laws serving to foster industrial competition.” As a descriptive matter, today, this thesis is largely, but not uniformly, respected.
More specifically, the central task of this chapter is to note how the concern with monopolization—explicit in the antitrust laws—plays a powerful, if somewhat concealed, role in the articulation of patent law as well. As is always the case, any concern with monopolization is a two-edged sword: It is always important to make sure that monopoly practices do not go undetected, but it is equally important that the doctrines of both patent and antitrust law do not impose penalties for supposed monopolistic practices that ultimately turn out to be procompetitive.
To discharge this task, this chapter begins with some fundamentals of intellectual property law before moving on to discuss the interaction of patent law with antitrust law. Section 1 lays the groundwork for the enterprise by articulating the standard rationales for patent protection and copyright protection. Section 2 then explores the relationship between patents and trade secrets as sources of protection. Section 3 examines the restrictions imposed on patent eligibility to deal with natural substances and abstract ideas and argues that these two concepts have been stretched too far—that, in their effort to prevent monopoly, they frustrate the emergence of useful technologies.
Section 4 explores the concept of an “exclusive right” to inventions and writings—whereby the simple assertion of an exclusive right is never sufficient to find a monopoly in the market to which the patent applies. Section 5 then looks at the way in which the antitrust law deals with certain unilateral patentee practices. Section 6 does the same for coordinated efforts among different patentees—noting that there are vastly different consequences to bringing complementary patents under single control in order to overcome coordination problems than there are for bringing substitute patents into a single pool in order to create monopolization problems. Section 7 examines the saga of Federal Trade Commission v. Qualcomm Incorporated, in which liability was imposed outside the standard theories described in sections 5 and 6, before the Ninth Circuit Court of Appeals reversed the decision. A brief conclusion follows.
I. The Rationale for Patents and Copyrights
The text of the patent clause offers its own tight, theoretical justifications for the use of patents. The exclusive rights afforded to patents under Article I, Section 8, Clause 8, are stated as an explicit reward intended to promote science and the useful—read technological—arts. The logic behind that grant of exclusive right is that people would not invest sufficient time and energy to create “any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof” if rivals were able to produce unlimited units of that same technology for commercial sale—without having to bear any of the heavy, front-end costs to design, make, and perfect the device or process. That ultimate demise would be known to the putative inventor, who would not invent if appropriation of that invention at lower cost is a direct threat. Thus, the main use for patent protection is between merchants. Nothing is more common than for a purchaser of a patented device to share it with his friends, neighbors, or coworkers, without seeking to recover any additional charge. That kind of extra use is within the limits welcomed by the patentee because it increases both the scope of the market and the value of the good to any individual purchaser. But expansion of the market quickly turns negative if the purchaser seeks to go into competition with the patentee by selling versions of the same product at a lower price than what the inventor or his assignee charges. The resource effects are exactly the opposite as rival sales to strangers would no longer expand the market but contract it.
The same is true with copyrighted materials. The seller of a recording or of sheet music is delighted if his purchaser shares it with others, but rightly turns hostile when the copyrighted work is reproduced in any form that is then resold to third parties. The failure to understand this simple illustration can lead to dangerous results if innocent instances of customary sharing of patented or copyrighted works are said to stand for the far larger proposition that removing all intellectual property (IP) protections would cause no losses at all. The so-called “artificial scarcity” of intellectual property is justified—within the limits just stated—for commercial transactions. This scarcity can coexist happily with the noncommercial sharing of different forms of IP. The obvious differences between tangible property—which can be used by only a single person—and information, which can be shared, does not make intellectual property a misnomer or a dangerous form of monopolization. It only justifies—given the tradeoffs between common and private use of the different asset classes—a regime that limits the duration of intellectual property, without leading to its total demise.
II. Patents and Trade Secrets
It follows from the above analysis that some patent law is necessary for innovation markets to operate effectively, but the hard, theoretical question lies in determining just how these markets will operate when the exclusive right is either eliminated or, more often, unduly circumscribed. It is not the case that we should expect no inventive activity, even in the extreme case where all patent protection is removed. Rather, we should expect that the level of innovation would be lower than it would be if a sound form of patent protection were afforded, and that this reduction would translate into some loss in overall productivity. Even in the absence of patent protection, it is possible that an inventor would gain a first mover advantage that lets him establish a toehold (or footprint) before the imitators could enter the fray. More importantly, an inventor may be able to obtain limited protection for various inventions under trade secret law, which is defined today as “any information that can be used in the operation of a business or other enterprise and that is sufficiently valuable and secret to afford an actual or potential economic advantage over others.” Those trade secrets have customarily been held to include “a formula for a chemical compound, a process of manufacturing, treating or preserving materials, a pattern for a machine or other device, or a list of customers.” The protection for the trade secret is limited in that it only stops people from stealing the secret. Unlike patents, it does not stop other individuals from independently discovering the same, or similar secret—so two competitors may, unbeknownst to each other, possess the same form of intellectual property. For the most part, the critics of patent protection treat the prospect of such discovery as protection against the monopoly risk that they perceive—wrongly—to be created by all forms of patent or copyright protection.
It should be apparent that there is an enormous overlap between the types of intellectual property that potentially fall under both patent and trade secret law. Chemical compounds, for example, are compositions of matter that can be fashioned by machines, by techniques of manufacturing; they are therefore patentable. The know-how required to make such compounds may quite plausibly be valuable and secret; they are therefore within the domain of trade secret law. Fortunately, the Supreme Court has held that trade secret law covers both patentable and nonpatentable subject matters, on the grounds that “the patent policy of encouraging invention is not disturbed by the existence of another form of incentive to invention.”
The legal protection afforded to trade secrets does not require the holder of a trade secret to limit its use to the internal operations of the firm. It has been universally held that any holder of a trade secret is entitled to license that secret, on either an exclusive or nonexclusive basis, to other individual firms—who in turn must be able to share those secrets with their workforce and their subcontractors—so long as the holder of the trade secret makes reasonable efforts to keep it confidential. All of these arrangements put the recipient under a general obligation to keep the trade secret confidential, which is necessarily lost with public disclosure at any point; universal usage can negate any economic advantage to the creator of that secret. This evident vulnerability means that the trade secret must be protected in all jurisdictions. Its disclosure in one location necessarily allows for its use worldwide—one reason why the forced disclosure of trade secrets in any one jurisdiction is treated as a taking of the secret. And, in general, it appears that the set of legal, contractual, and informal sanctions has met this condition of universal secrecy. Within this general framework, the licensing provisions between commercial parties are normally theirs to decide, so that gains from trade increase the value of the trade secret, just as licenses increase the value of patents.
Any truncation of the class of patentable inventions surely enlarges the role of trade secrets. But it also remains the case that trade secrets are often—especially for nascent firms—a poor substitute for patent protection. These observations quickly lead to the question of which form of protection should be preferred when both are available; and that question in turn usually leads to a split verdict. For processes, firms typically prefer trade secret protection because they know that if the process is made public, it will be exceedingly difficult for them to determine which products manufactured or sold by rival firms will have resorted to the use of a particular process— routinely, multiple processes are available to do a particular job. On the other hand, trade secrets cannot protect any product for sale where imitation is far easier. In these cases, patents are far preferable because they allow for the dissemination of the product without loss of the right to exclusivity. This exclusivity comes with the condition, as has always been required under Section 112, that the patentee, as part of his or her application, supplies information as to the best mode of production that rival inventors can use to design their own devices or compositions of matter that do not infringe existing patents. Section 112 requires this disclosure, notwithstanding that the similar functionality of these other products allows them to provide competition for the products sold under alternative patents, or, indeed, with once-patented products that have already entered the public domain.
III. Patent Eligibility and Monopoly Power
How, then, is the scope of patent protection determined? In practice, there is a prominent public choice dimension to the debate over what form intellectual property protections should take. There are strong reasons to believe, as Jonathan Barnett has argued, that large, established, and integrated firms are in a better position to protect their intellectual property portfolios, through internal use on the one side, and trade secret licensing (including cross-licensing agreements) on the other. In contrast, small start-up firms with a single product must be able to sell in the open market to a large number of unrelated parties, with whom they do not have prior, dense relationships. Their common business plan requires strong patent rights. Hence, the striking conclusion—it is weak patent rights that act as a barrier to entry. Barnett’s hypothesis is borne out by the legislative position that major high-tech firms have taken in connection with debates over patent strength, most recently in connection with several recent decisions on patent eligibility—which requires courts, under Section 101 of the Patent Act, to draw a line between patentable inventions on the one side, and laws of nature or abstract propositions on the other. These large, established firms consistently favor weakening patent rights in at least two dimensions: Narrowing of the class of patentable inventions, and making it more difficult to obtain injunctive relief for patent infringement.
Tellingly, both of these points are cardinal elements of the position of the powerful High Tech Inventors Alliance, whose membership includes Adobe, Amazon, Cisco, Dell, Google, Intel, Oracle, and Salesforce. Its mission statement reads: “HTIA supports a balanced patent system that fixes the problems of low quality patents, baseless assertions and patent troll litigation while promoting investment in new technologies and American jobs.” In order to advance that position, the Alliance commissioned a paper from Paul C. Clement to advance the view that patent rights are, as his title suggests, the antithesis of property rights, such that their protection is more a matter of legislative grace than constitutional entitlement.
An emphasis on the tension between property rights and patents has profound implications for the structure of patent law and the scope of its exclusive right. The language of Section 101 talks about “any new and useful process, machine, manufacture, or composition of matter.” Although this collection of terms may not appear to have any antitrust implications, the connection between these definitional provisions and the traditional antitrust concern with monopoly is key. A narrow reading of these four terms necessarily reduces the scope of any exclusive right, thus increasing the area of the public domain. Two key items are exempted from the scope of patent protection: Natural substances and abstract propositions. In the conventional sense, both of these limitations make good sense. Could one imagine how the progress of mathematics would take place if, for example, James Napier could obtain a patent on the use of logarithms, or charge subsequent users a royalty for the use of the decimal point in mathematics? But then he would be required to pay royalties to any mathematician whose theorems he relied on in his own work. The blockade on these matters would rise to the level of insanity if each advance had to pay royalties to others. It is a clear that everyone is better off from the free exchange of information rather than from a system of exclusive rights in abstract propositions. Hence it has always been the case that support of mathematics has taken the form of salaries and prizes to the people who make these key advances, precisely because these rewards give incentives for productivity, while keeping all the building blocks of knowledge firmly within the public domain. Similarly, it would make no sense to give Madam Curie the exclusive right to the use of radium, solely because she was the first person to isolate it in the laboratory. A Nobel Prize works fine for these discoveries.
At this generalized level, there is no opposition to these two limitations of patentability; but, in recent years, both of these conceptions have been expanded greatly to cover forms of knowledge where holdout problems are the exception and not the rule. Indeed, the broader the definition of these two areas, the narrower the scope for patents and the smaller the risk of monopolization—a benefit that comes at the cost of potentially inhibiting innovation. So, drawing the proper line becomes critical. In two recent cases, Mayo v. Prometheus and Alice Corp. v. CLS Bank International, the Supreme Court has narrowed the scope of patentable subject matter, both for abstract propositions and for natural substances: Mayo gave a narrow construction to medical devices, and Alice did the same with respect to algorithms for financial models.
This development has not gone unnoticed. In his statement before the Senate Judiciary Committee, the Executive Director of HTIA, David Jones, insisted that the high levels of current investment justified the continuation of the Alice/Mayo regime. In contrast, the rival statement by Professor Adam Mossoff, with no backing whatsoever from powerful institutions, examined and criticized the line of cases that applied the Alice/Mayo definitions to both abstract ideas and natural laws, denying coverage to small inventors with major innovations. He writes: “As a result of the Alice-Mayo framework, the U.S. is now denying patent applications or invalidating issued patents in cutting-edge discoveries in medical care, such as treatments for breast cancer, diabetes, and strokes, among others.”
In examining this issue, it is important to look at this granular level to see whether the decisions are sound—which I believe that they are not—and then assess their economic impact. It takes hard work and constant testing to figure out these interrelationships, and the earlier decisions of Judge Giles Rich in such cases as In re Alappat, dealing with medical devices, and State Street Bank & Trust Co. v. Signature Financial Group, dealing with business method patents, represent a sound approach that should still be followed today. In both these situations, mathematical transformations took input data from outside the system and turned it into commands that could be executed by the system, yielding specific instructions on how to behave. The key element here, which both courts correctly recognized, is that the middle, computational element of an invention should never be looked at in isolation. To do so puts virtually all claims at high risk of invalidation. As the State Street Bank & Trust Co court put it, “the mere fact that a claimed invention involves inputting numbers, calculating numbers, outputting numbers, and storing numbers, in and of itself, [does] not render it nonstatutory subject matter.”
These rules require that all patentable subject matter involves some substance or device. It is not possible, for example, to obtain a patent for the exclusive use of some natural resource because of the massive preclusive effect on other inventors who seek to develop alternative devices utilizing that same natural resource. The most vivid statement of that principle is found in O’Reilly v. Morse, where Chief Justice Taney writes in no uncertain terms:
“Eighth. I do not propose to limit myself to the specific machinery or parts of machinery described in the foregoing specification and claims; the essence of my invention being the use of the motive power of the electric or galvanic current, which I call electro-magnetism, however developed for marking or printing intelligible characters, signs, or letters, at any distances, being a new application of that power of which I claim to be the first inventor or discoverer.”
It is impossible to misunderstand the extent of this claim. He claims the exclusive right to every improvement where the motive power is the electric or galvanic current, and the result is the marking or printing intelligible characters, signs, or letters at a distance.
If this claim can be maintained, it matters not by what process or machinery the result is accomplished. For aught that we now know some future inventor, in the onward march of science, may discover a mode of writing or printing at a distance by means of the electric or galvanic current, without using any part of the process or combination set forth in the plaintiff’s specification. His invention may be less complicated—less liable to get out of order—less expensive in construction, and in its operation. But yet if it is covered by this patent the inventor could not use it, nor the public have the benefit of it without the permission of this patentee.
The ambiguity in this passage is whether Taney’s characterization that this patent was an attack on abstract principles fairly lines up with the would-be patentee’s claim. Historically, there is some doubt about this issue because the nineteenth century style of claiming tended to treat each count as part of the full application, and not in isolation from the other counts. Looked at as a whole, listing eight elements of the claim do nothing out of the ordinary in comparison to a modern legal claim. Thus, this claim could be read as an early anticipation of the doctrine of equivalents—a key doctrine of patent law intended to make sure that the exclusive right conferred by a grant is not too narrow, lest it be evaded by the inconsequential substitution of one kind of motor or battery for another that allows a so-called “new” device to freeload off the original inventor. It is too costly to insist that a patent application list every possible permutation of a given design in advance; the general rule allows for close substitutes to receive the same protection as the covered invention, unless of course a particular extension is disclaimed in the patent application. One could read “claim eight” as stating only that proposition. But if “claim eight” is read more broadly, as asserting an exclusive right to any use of electromagnetism to create markings, then Taney has a point, applicable both in his own time and today. For in that case, the claimed invention will block out serious competitors from using the natural world, impeding an important feature of scientific inquiry. Such preclusion matters because the state of technology is often such that multiple parties are homing in on the same invention at the same time. Indeed, Taney did make the explicit reference that, “in this state of things it ought not to be a matter of surprise that four different magnetic telegraphs, purporting to have overcome the difficulty [of preventing loss of signal power over distance], should be invented and made public so nearly at the same time that each has claimed a priority.”
Two general conclusions, both consistent with the basic economic preference for competition over monopoly, are at work here. First, preventing close substitution increases the value of acquiring and protecting property rights, which encourages the initial innovation. Second, preventing the patenting of natural objects means that patent law does not become a barrier to entry by allowing only one person to use what could be described as part of the common heritage of all humankind. One notable illustration of this principle is found in Association for Molecular Pathology v. Myriad Genetics, which concerned the patentability of the BRCA1 and BRCA2 genes. These genes are positively correlated with the occurrence of breast cancer in women. Their structure was first identified by Myriad, which then sought to patent the genes itself.
At issue are two distinct questions that should receive very different treatment. The first asks whether Myriad could have any exclusive rights over the BRCA genes in situ, to which the answer should be in the negative. These are natural occurring substances, which by definition have not been isolated from their natural environment. It would be an incredible limitation on the advancement of knowledge—the creation of, as it were, an undeserved monopoly—if the party that isolated the gene could prevent anyone else from treating its associated diseases in a patient without the consent of the patentee.
The second question is whether Myriad should be able to patent the isolated gene. The restriction-on-treatment issue could not arise with the isolated gene, given that the gene is, again by definition, never isolated inside any human being. But at the same time, it is much more defensible to allow for the patenting of the gene outside of the human body. In dealing with this issue, Justice Thomas insisted that it was proper to obtain a process patent that would prevent other individuals’ use of Myriad’s method for the reproduction of the BRCA genes; namely, the creation of a complementary form of DNA (cDNA), which offers a template for DNA found in nature to be produced in a cheap and reliable fashion. cDNA is patent-eligible precisely because it is created by using materials that are not found in nature. But it is a far more difficult question to decide whether the process patent should be sufficient for the BRCA genes, given that Myriad’s work has made using other methods to produce the gene far easier.
That exact issue was raised in the earlier decision of Parke-Davis & Co. v. H.K. Mulford Co., which concerned the patentability of adrenaline, which had been isolated and purified by Joiichi Takamine, a Japanese chemist. Learned Hand baldly stated: “Takamine was the first to make it available for any use by removing it from the other gland-tissue in which it was found, and, while it is of course possible logically to call this a purification of the principle, it became for every practical purpose a new thing commercially and therapeutically. That was a good ground for a patent.” Without any analysis, Learned Hand conferred on that “new thing” a substance patent, not just a patent for the process of its isolation. He offered no systematic analysis as to why he applies the stronger form substantive protection (which precludes the synthesis of adrenaline by any other technique during the patent period), and that debate rages on today.
At the back of the argument is an empirical disagreement over the optimal strength of the exclusive protection that should be given to the inventor who takes the first critical step. The more alternative pathways for synthesis, the less valuable the process protection—which could be insufficient to induce the optimal amount of innovative effort. But the broader substance protection could give enormous breadth to a patent when multiple paths to discovery exist, thereby conferring monopoly power on the holder of that patent. The empirical debate can vary from substance to substance, and it is quite possible that a greater level of protection was preferable in 1911, and a lesser level is preferable today. But by the same token, it should be recalled that, according to the Jonathan Barnett thesis, weaker levels of patent protection will have a greater impact on small firms seeking to break into the market than on larger firms that are more likely to have alternative means to protect their intellectual property. Hence it is not surprising that no consensus has emerged on this troublesome issue. But, even if patent protection is extended to the isolated DNA, nothing should prevent finding BRCA in situ as a nonpatentable natural substance, not created by either human wit or invention.
If Myriad raises hard questions, the same is not true of the Alice/Mayo line of cases that goes far beyond both Morse v. O’Reilly and Parke-Davis. There is no plausible resemblance between the discovery of the electromagnetic spectrum or the identification of adrenaline on the one side, and the ad hoc determinations, obtained typically by trial and error, to determine a method for calculating proper dosage of thiopurine, which is about as specific as possible—as if there could be any general principle to determine the proper dosage of any particular chemical. These recent decisions are an unhappy throwback to the 1948 case of Funk Brothers Seed Co. v. Kalo Inoculant Co., which refused to grant a patent to an inventor, who first discovered a way to mix distinctive inoculants for six different leguminous plants into a common package. The new process saved both time and money and depended on detailed, specific research to achieve the correct mix. Once again, it is a mistake to think that this combination should be treated, as Justice Douglas believes, as a law of nature, given that the mix cannot be deduced from first principles: it was discovered in a unique, context-specific way. Indeed, the Seventh Circuit got it correct when it wrote that “the evidence is clear that what he discovered was that certain existing bacteria do not possess the mutually inhibitive characteristics which had previously prevented a successful commercial composite inoculant and that those uninhibitive species may be successfully combined. It was this contribution of noninhibitive strains which successfully combine that brought about a new patentable composition.”
The economic advantages of a six-in-one package are commendable because it both reduces error and increases the rate of application. It is hard to think of how any system of salaries and prizes that work for natural substances could provide the needed incentive here, when the risks of freeriding are so manifest. Speaking more generally, the economic consequences here are serious, because if the property rights are too narrow, the first entrant will not come into the field; just as if they are too broad, no new entrant could ever enter the field.
IV. Exclusive Right
At this point, it becomes necessary to tie this general discussion back into the discussion of what is meant by the phrase “exclusive right to their respective inventions,” which is “secured” by the constitutional protection of patents. One common argument is that this phrase confirms the obvious fact that all patents are monopolies, which should in turn be uniformly and necessarily subject to regulation under the antitrust laws, like other forms of monopoly. That extension has been uniformly rejected in nineteenth century cases, even if it has gained some more traction in recent years. In his 1833 Circuit Court opinion in Ames v. Howard, Joseph Story neatly linked the appropriate rule of the construction of patents to the view that patents were granted in the public interest:
Patents for inventions are not to be treated as mere monopolies odious in the eyes of the law, and therefore not to be favored; nor are they to be construed with the utmost rigor, as strictissimi juris. The constitution of the United States, in giving authority to congress to grant such patents for a limited period, declares the object to be to promote the progress of science and useful arts, an object as truly national, and meritorious, and well founded in public policy, as any which can possibly be within the scope of national protection. Hence it has always been the course of the American courts, (and it has latterly become that of the English courts also,) to construe these patents fairly and liberally, and not to subject them to any over-nice and critical refinements.
That sentiment was followed in other cases. Thus, in 1855, in Allen v. Hunter, dealing with the setting of mineral teeth on metallic plates, Circuit Justice McLean wrote:
Patentees are not monopolists. . . . No exclusive right can be granted for anything which the patentee has not invented or discovered. . . . [T]he law repudiates a monopoly. The right of the patentee entirely rests on his invention or discovery of that which is useful. And which was not known before. And the law gives him the exclusive use of the thing invented or discovered, for a few years, as a compensation for “his ingenuity, labor, and expense in producing it.”
Similarly, in Birdsall v. McDonald, involving a combined thrashing and hulling machine, Justice Swayne wrote:
Inventors are a meritorious class of men. They are not monopolists in the odious sense of that term. They take nothing from the public. They contribute largely to its wealth and comfort. Patent laws are founded on the policy of giving to them remuneration for the fruits, enjoyed by others, of their labor and their genius. Their patents are their title deeds, and they should be construed in a fair and liberal spirit, to accomplish the purpose of the laws under which they are issued.
Note that there is no hesitation in treating the patent as a form of property, protected by “title deeds”—and not as an “odious” monopoly that restricts output, raises prices, and generates social loses. After all, these patents are not just gratuitously awarded as a favor or a political payback from the Crown or the government. The exclusive right to a patentable invention is on a model that is an antithesis to the standard economic monopoly. A patent is no more of a monopoly than giving a title deed that gives an owner of a plot of land exclusive rights to its possession, use and disposition. It is well understood, in both the Roman and the Anglo-American tradition, that the notion of property in land carries with it, as Thomas Merrill has noted, the idea of exclusivity, which the Supreme Court has often said—most notably in Kaiser Aetna v. United States— counts as “one of the most essential sticks in the bundle of rights that is commonly characterized as property.” Indeed, Merrill goes one step further and insists that, without the right to exclude, there is no such thing as property at all.
It is important to sort out the cross currents in these various claims. First, it is incorrect to claim generally that, without exclusivity, there is no property right. Indeed, the sentence, as quoted, does not distinguish between private and common property. The right to exclude is part of the former but not of the latter. But the latter is still governed by property rights, albeit of a different sort. A property right in the commons—both for real and intellectual property—means that all persons have the right not to be excluded from the access to a particular resource, such that it should be a taking (perhaps even of property) if someone has been excluded from access to the commons, as by building a wall that separates a riparian from access to water in a river to which he or she would otherwise have access. The argument in favor of treating this denial of access as a taking is that the same action of blocking access, if done by a private party, would subject that party to damages for past losses and an injunction against future walls. The wall could well be for public use, but that fact alone should not dispense with the need to supply just compensation to the aggrieved riparian. That result has been avoided in the decided cases, solely on the false claim that in dealing with water rights, the rules that apply to disputes between private parties are said not to carry over to actions brought by the government, which is claimed, without explanation, to have a “paramount” or “dominant” easement over all public waters, which functions like an absolute claim. That sovereign claim is in stark contrast to the usual analysis of riparian rights, whose typical refrain talks about correlative duties, for the obvious reason that a river, stream or lake—which can be used for transportation, recreation, fishing, watering homes, cattle, and crops—is worth more as a “going concern” than as still water poured into a large barrel.
This issue, moreover, has an exact parallel in the intellectual property space, where there is a similar commons consisting of abstract ideas and natural substances, which are from the outset open to all. In addition, since intellectual property grants are made only for a limited period of time, a patented invention (but not subsequent improvements thereon)—the original technology—passes into the public domain at the expiration of the initial term of years. Setting the length of that term is critical to the overall success of the system. Make it too short, and the inventor will not have sufficient time to recoup the front-end costs of development, before the technology enters the public domain. Make it too long, and the exclusion could lead to the perpetuation of undue rents to the patentee, by denying use to those who value the technology at more than zero, but less than the prices charged.
Of equal importance is the scope of patentability, given that laws of nature and abstract propositions are excluded from their scope. Here, it is possible to confuse the definition of an exclusive right with that of a monopoly. But that comparison is flawed. Professor Merrill, for example, stresses the element of exclusivity as part of the definition of a property right, but not once in his article does he equate exclusivity to monopoly power. In this connection, the same rules apply to real property and to patents, notwithstanding that the former are typically of indefinite duration and the latter are only for limited terms. As a matter of first principle, it is not possible to create a competitive market unless owners of land and other forms of property have some exclusive right. The right, of course, must be more than the right to exclude, for such a right would be of little, if any, value, if the property owner labors under a disability whereby he cannot enter his own property, let alone use, develop, sell, lease, mortgage, or otherwise dispose of the property in question.
The point becomes clear when one looks at the additional value conferred as each of these rights are added to a system of property. The right to exclude, even standing alone, is not without value. It allows, for example, the owner of some adjacent property to preserve views over the vacant lot, and it also allows that owner the ability to keep out competition from rivals to his own business. But the added value to ownership from being able to enter the land is surely great, and, as a rough empirical generalization, the losses to others—especially if they get the same set of augmented rights with respect to their own properties—will be trivial in comparison. As with all great social generalizations, the source here is common social understanding, which long antedates any supposedly empirical studies on the question. Similarly, rights to use and develop land can be added to the bundle, because they confer great value on the user with, in most instances, relatively small losses by others. Given the reciprocal nature of these rights across all users, the gains are large. One way to test the possibility of reciprocal gains is to note that when a single-owner creates a condominium or gated community, rights of use and development are always included in the mix. Yet, at the same time, the risk of negative externalities grows from incompatible uses and developments; thus, the set of restrictions on use and development are almost always larger than those on occupation. These risks are less evident in cases of intellectual property than they are with physical property, because the class of nuisance-like activities is far smaller—which explains why police power limitations on patents and copyrights are much harder to justify.
Finally, extensive rights of use and development, standing alone, fall short in yet another way; namely, they do not permit gains from trade in either the outright transfer, or the pooling of both intellectual and physical assets. Here, the gains from trade call for allowing these efforts to take place, but again, there is with intellectual property, as with other forms of assets, an implicit limitation that systematic losses to third parties will, in some cases, eclipse the gains to the parties from cooperative assets. Ordinary competition always results in losers who cannot match the price, but that form of competitive harm should never be restrained, because otherwise there would be enormous harm imposed on consumers who are deprived of buying goods at a lower price. The traditional legal systems understood this—along with the harm from blocked views—and treated them as damnum absque iniuria, or harm without legal injury, which could easily be dismissed for doubletalk, when it is not. The net positive social gains from a system of competition can only be achieved if disappointed competitors try to use their resources somewhere else, which is why the nineteenth century cases called these instances “noncognizable” or “nonactionable harms.” But the situation is very different when the agreement calls for the destruction of physical (or intellectual) property of third parties, for now the choices for consumers are reduced, not enhanced. The case of monopoly is less compelling but more common, for with contracts in restraint of trade, the gains to the contracting parties, systematically, are more than offset by the social losses from restricting output and raising prices, which explains the social objection to cartelization. That restraint applies with equal force to intellectual property and other forms of assets, once again proving there is a continuity between intellectual property and other forms of assets. The right to dispose receives a strong general endorsement, subject to a limited antitrust exception.
V. Unilateral Conduct
At this point, it is possible to make a smooth transition from intellectual property rules to explicit antitrust concerns. Definitionally, “unilateral” conduct does not mean that a party enters into no agreements at all. Rather, it means that all the agreements are made with parties on the opposite side of the market—sellers to buyers, landlords to tenants, lenders to borrowers—and none with potential competitors. In these circumstances, the general rule is that the owner of a patent, like the owner of a plot of land, is free to use or dispose of his property however he sees fit, without any independent constraint under the antitrust laws. There are three relevant variations on this theme: pricing, common carriers, and tie-in arrangements.
An early statement of the pricing rule is found in E. Bement & Sons v. National Harrow Co., where the National Harrow Company, the plaintiff below, had entered into a licensing agreement for its patented “float spring tooth harrows, their frames, and attachments applicable thereto.” When the plaintiff sued to recover liquidated damages and obtain injunctive relief for breach of contract, the defendants sought to avoid enforcement by showing that the contract was in violation of the Sherman Act, due to a “combination on the part of all the dealers in patented harrows to control their manufacture, sale, and price in all portions of the United States,” by such devices as manufacturing or sale quotas, territorial restrictions, line of use restrictions and the like, all of which would be the source of per se antitrust liability if done by dealers as part of a horizontal cartel. But in this context, the dealers’ role was a makeweight, for all the power lay exclusively in the hands of National Harrow, the patentee, who selected the dealers and instructed them on all these terms. Hence the defense failed. Justice Peckham, in line with earlier authority, upheld the unilateral actions of National Harrow:
The plaintiff, according to the finding of the referee, was at the time when these licenses were executed the absolute owner of the letters patent relating to the float spring tooth harrow business. It was, therefore, the owner of a monopoly recognized by the Constitution and by the statutes of Congress. An owner of a patent has the right to sell it or to keep it; to manufacture the article himself or to license others to manufacture it; to sell such article himself or to authorize others to sell it.
The impeccable economic logic behind this opinion is as sound today as it was on the day it was written. The owner of any form of property is of course entitled to keep or sell it. There is nothing which says that the sale of a particular property has to be made to one, and only one, person. It is clear that the patent confers some ability to raise price over marginal cost, but this advantage is exactly what patents are intended to produce. Or, to put the point in general terms: anyone who owns an especially desirable asset can reap extra gains by selling it into a competitive market, without running afoul of the antitrust laws. In standard economic terms, the seller reaps Ricardian rents. By way of example, these rents are enjoyed by the owner of coal deposits that lie close to the surface and thus can be mined at lower cost than other market participants, whose coal lies deeper underground. Thus, if the market price of coal is $25 per ton, everyone who can produce at below that price will remain in the market. A coal owner who can extract at a cost of $10 will earn a Ricardian rent of $15 per ton and will remain in the market, so long as the price of coal exceeds that level, even as those with higher costs of production will exit first. If the first party sells its mine to another, the sales price will reflect the present discounted value of the future Ricardian rents, allowing the late entrant only the normal competitive rate of return. Accordingly, Ricardian rents do not create an umbrella for new entry, but monopoly rents do. Put otherwise, the violation of the antitrust law comes only when two or more independent sellers come together and raise the competitive price, at which point the seller whose coal is near the surface gets some mixture of a legal Ricardian rent and an illegal monopoly return.
Exactly the same situation applies to patented goods. The patent (or trade secret) creates a Ricardian rent but does not create a market distortion. It follows, therefore, that it should be allowed to use whatever means it sees fit to exploit that position. Ordinarily, any restriction on the mode of its disposition will reduce monopoly rents, but it will render the system of distribution less efficient that it was before. Hence Justice Peckham is entirely correct in noting that the patentee can act by itself or through one or more agents, and can sell or license, absolutely or on condition.
B. Common Carriers
To this particular pricing rule, moreover, there is an important exception dealing with common carriers. Much of the late-nineteenth and early-twentieth century law dealt with the difficult questions of rate regulation that arose when new technology was discovered; important industries like railroads and public utilities were often serviced by a single supplier of the new technology. In some instances, these suppliers obtained a natural monopoly position because they received an exclusive government franchise. In others, there were physical limitations that prevented the entry of rival suppliers. Starting with the important English case of Allnut v. Inglis, the English Courts held that a common carrier was restricted to charging a competitive rate, which could be easily determined on the facts of that case. The defendant operated a customs house that was exempted from inland taxes because the goods stored there were to be shipped overseas. The tax exemption was an effort of the British government to steer international transportation through British ports. The proprietor attempted to raise his rates to capture some of the tax reduction, which impaired the success of the two-part system. The government had to come in to correct the situation by limiting the proprietor’s rates to match the competitive warehouses located nearby.
The incorporation of the common carrier doctrine into the United States rested on the explicit reliance on Allnut in the great American rate case of Munn v. Illinois, which gave rise to a far more difficult question of how to apply the principles of rate regulation when there was no competitive benchmark lying nearby. At such a point, the entire enterprise requires rate of return regulation, which was one of the main topics of the late-nineteenth and early-twentieth century Supreme Court case law. The complex system is hard to describe in a few lines, but the challenge faced by the courts was to find the middle path between unacceptable extremes: monopoly profits from unregulated rates and asset confiscation from low rates determined by a public utility commission, which could result in the inability of a firm to recover its invested capital. For our purposes, the precise formulas for determining reasonable rates are not in issue. What does matter is that in Bement, Justice Peckham rightly held that the same power to regulate the rates of common carriers that held a monopoly position also applied to patent holders in that same position. Thus, in Bement, Peckham wrote that service duties applied
because a telephone company, being in the nature of a common carrier, was bound to render an equal service to all who applied and tendered the compensation fixed by law for the service; that while the patentees were under no obligation to license the use of their inventions by any public telephone company, yet, having done so, they were not at liberty to place restraints upon such a public corporation which would disable it to discharge all the duties imposed upon companies engaged in the discharge of duties subject to regulation by law.
That formula has, in rate regulation circles, been called service on (F)RAND, or (fair), reasonable and nondiscriminatory terms. The argument for using this formula is the same today as it was back in the early 20th century. The term “fair” was intended to show that the rates in question could not be left to the appetites of the parties, but rather had to meet some external standard of validity. The term “reasonable” was included to imitate the overall rate of return reduced to a competitive-market rate, adjusted for risk under the particular regulatory environment, and the term “nondiscriminatory” meant that certain forms of price discrimination, namely, any practice that charged different rates to customers who had equal cost of service had to be banned. It is exactly these basic conditions that are used today in dealing with so-called standard essential patents. The exact formulas that are used to resolve these cases can vary, owing to the fact that there is no unique price that is mutually advantageous for all parties.  But these are common problems in all rate regulation cases, and the sole purpose of these observations is to stress that these difficulties are not unique to the patents.
Another form of unilateral conduct is tie-in sales or required purchases accompanying leases of patented goods. Under the general logic of Bement, there is no sharp distinction between the various kinds of restrictions that the patentee can impose on the use of its object. Thus, tie-in arrangements for a patented product, in which a company forces the purchase of an accompanying non-patented product, would have been subject to a rule of per se legality that applies to price so long as there was no cooperative arrangement with other parties. That position was taken by the Supreme Court in Henry v. A.B. Dick Co., where the requirement to use unpatented ink was attached to the sale of a patented rotary mimeograph machine. The Court held that the patentee could sue a rival supplier of ink to one of its customers as a form of contributory infringement. That decision, in turn, raised a dissent from Chief Justice White, who objected to the tying arrangement, in his view, as the abuse of “legislative power” to extend the patent to cover things that are not within its original scope. The dissent proved influential with Congress, which two years later passed Section 3 of the Clayton Act. That section applied to the sale or lease of both “patented or unpatented” merchandise of all kinds and description, on condition that the lessee or purchaser cannot use that product in dealing with the product of some competitor, if that restriction substantially lessens competition or tends to create a monopoly in any line of commerce.
Section 3 reflects the basic view of Chief Justice White: that ties-ins allow parties to “lever” their monopoly position in one market into a monopoly position in the sales of the tied good. For many years, the Supreme Court went even further by taking the position that “[t]ying agreements serve hardly any purpose beyond the suppression of competition,” in holding that Standard Oil could neither make nor enforce an exclusive dealing contract for petroleum products and automobile accessories. In the course of his opinion, Justice Frankfurter correctly made reference to tie-in arrangements with patented goods. In this context, the key decision is International Salt. Co v. United States, which carried forward the tradition of per se illegality of tying agreements. International Salt (IS) owned two key patents: “the ‘Lixator,’ [which] dissolves rock salt into a brine used in various industrial processes, [and] the ‘Saltomat,’ [which] injects salt, in tablet form, into canned products during the canning process.” Both of these goods were tied to a non-patented good, salt. IS’s standard business practice required lessees of these patented machines to only use IS’s salt. By way of justification, IS claimed that its salt in question had a higher level of purity than other brands of salt commonly available in the market, and that using lower quality salts could ruin the leased equipment in which IS still had a proprietary interest. Its contract also contained a provision that stated that if a competitor offered an equivalent product at a lower price, IS would have to meet that price or forfeit its exclusivity. The Supreme Court, speaking through Justice Jackson, held that the tying arrangement was a per se violation of the antitrust laws that was not saved by either of the purported justifications.
Given the basic logic of Bement, why is it the case that IS contains complete freedom of pricing over its own product, but none over the use of tied goods? In choosing its marketing strategy, IS could charge the same price for the entire package if it raised the price of its equipment and lowered the price of the salt, in order to ward off further competition in the salt market. But would that be wise, given that the proportions matter? At one extreme, it would not do to charge some fixed price for the machine and nothing for the salt, no matter how much was used. Nor, at the other extreme, would it work to charge nothing for the machine, and obtain all revenues from the sale of salt. The difficulty with these polar arrangements arises whenever (as is always the case) the various lessees use different amounts of salt with the same equipment. High demanders would gravitate to the first arrangement, on which the company could lose money, as the low demanders would drop out. Conversely, on the latter arrangement, the low demanders would stay and the high demanders would drop out. Clearly some intermediate solution would work best to keep the full range of potential customers, and there is no reason to think that any court could find that intermediate position better than IS. Simply put, there is no sensible reason to let a court force IS to change the relative price of its two products.
Justice Jackson was also wrong to pooh-pooh the “meeting competition” provision of the agreement, which allows for price reduction to remain in the market if a competitor offers the same goods or services for less. That term gave the company an advantage in the case of ties, but it did nothing to prevent a more efficient salt producer from offering salt at a price that IS could not match, thereby retaining competition for both the tying and the tied good. The right of first refusal is common in other contexts, and this provision is just a variation on that theme. And, finally, there is a difference between stipulating a standard of salt that outsiders must supply and supplying it yourself. There is less of a quality control problem when a company supplies its own product than when it has to monitor the lessee’s use an inferior product, which might cause equipment damage that is borne by the lessor.
It should be clear that this brief discussion undercuts the once-regnant theory calling for per se illegality in patent tie-in cases—namely “a hostility to use of the statutorily granted patent monopoly to extend the patentee’s economic control to unpatented products.” A more finely attuned understanding of these business arrangements reveals that different arrangements, such as tie-ins or exclusive dealing contracts, might well have efficiency justifications that render per se illegality an inappropriate response to the issue. Hence the modern cases have marked a retreat from the per se rule in favor of a move toward a rule of reason analysis, most notably in Jefferson Parish Hospital District Number 2 v. Hyde.
That shift in attitude has worked its way back into dealing with patent tie-in cases in Illinois Tool Works, Inc. v. Independent Ink, Inc.—yet another patent tie-in case with ink and cartridge that applied the rule-of-reason test from Jefferson Parish. At issue in the case was yet another instance in which the sale of ink was the tied product to a sale of a patented inkjet printer, to which was attached a condition that the ink for that equipment had to be purchased from Illinois Tool Works. The decision rested specifically on the notion that the holding of a patent should not be regarded as holding a monopoly position in the sale of the tying good, and that such claims had to be separately proved, without any presumption in favor of treating the patent as a dominant product. As Justice Stephens had noted, the patent law had come to reject the position that the tie-in condition was a form of market abuse, and he held that the antitrust laws should take the same view with respect to the patented equipment. Indeed, that analysis has to be correct, because otherwise there would be two different bodies of law to respond to the common question of monopoly. At this point, however, the question does arise as to how exactly a rule-of-reason inquiry should work in tie-in cases. First, the printing business is highly competitive, so the entire monopoly issue seems to be misplaced; but second, the decision here—while explicitly rejecting the claim of Standard Oil that tie-in arrangements are used solely to suppress competition—does not stop to answer the question of what kind of efficiency justifications should be allowed once Standard Oil’s major premise is rejected. And so the law seems to come to its resting place, when it appears that a rule of per se legality most accurately captures the situation.
D. Coordinated Conduct
The last question on the intersection between the patent and antitrust laws concerns the cases where two or more patentees agree, between themselves, to some cooperative arrangement. In line with the general thesis of this chapter, general antitrust principles should apply with equal force to patented goods, which was the position taken, in 2017, by the Department of Justice and the Federal Trade Commission in their Antitrust Guidelines for the Licensing of Intellectual Property. At this point, the key distinction in antitrust law is between horizontal and vertical agreements, where the former link together rival producers at the same level of production, and the latter link together producers at the different levels of production. The reason for the sharp distinction between these two cases depends on key, first-order differences. With horizontal arrangements, the coordination between rival producers creates a cartel situation, in which the parties are able to raise prices above marginal cost and obtain a monopoly profit that they could divide among themselves, thereby generating a social loss. The vertical arrangements join together producers at different levels of the process, and ordinarily have strong efficiency justifications: the elimination of the double-marginalization problem, and the coordination of production in technology across different stages of production.
The first of these justifications is the elimination of the blockade problem that arises when the cooperation of all parties is necessary to bring a particular good to market. The situation is similar to the historical blockades along the Rhine River, where each toll could block movement along the entire river—an arrangement that was undone by the Treaty of Westphalia. The greater the number of tolls, the more the impediment, until the cumulative demands are so great that the shipment stops altogether. If it is assumed that some fees have to be collected to maintain the river, a single set of charges to cover the entire course offers a better way to raise those funds, without disrupting what goes on.
A second advantage of the unified vertical production line is that it eases problems in the coordination of production. A closer integration within the firm allows for the tighter integration of the various steps involved and also has the added advantage of making it easier to deal with external parties, be they merchants or consumers, when a finished product fails in some regard. Without vertical integration, disappointed purchasers or users who wish to get relief may sue the maker of the finished product—who, in turn, might find it difficult to defend the case if it does not know how each component functions, or which one went wrong. Worse, even if the original claim is disposed of, actions for contribution and indemnity may quickly follow. These could be litigated by contract on a piecemeal basis, at great cost and uncertainly, especially given that subcontractors might be involved. Vertical integration makes sense when internal controls are more efficient than litigation to achieve the desired ends. That need not be the case; but, if so, separate firms can continue to operate linked together only by contract—which might well happen when many of the key components sell to many independent players. But where these mergers do take place, the efficiency justifications seem to dominate any purported monopoly problem.
These differences are what drove the original merger of three major shoe companies, each with a different specialization, to form the United Shoe Machinery Company in 1899, in a transaction orchestrated by Louis Brandeis. That transaction was upheld in United States v. Winslow, explicitly on efficiency grounds:
On the face of it the combination was simply an effort after greater efficiency. The business of the several groups that combined, as it existed before the combination, is assumed to have been legal. The machines are patented, making them is a monopoly in any case, the exclusion of competitors from the use of them is of the very essence of the right conferred by the patents, Paper Bag Patent Case, 210 U.S. 405, 429  and it may be assumed that the success of the several groups was due to their patents having been the best. As, by the interpretation of the indictment below, and by the admission in argument before us, they did not compete with one another, it is hard to see why the collective business should be any worse than its component parts.
The arguments thus far point again to per se legality, but that is not the position that is taken by the DOJ/FTC guidelines. These guidelines argue that a vertical merger poses two risks. The first is that its position in the upstream market may lead it to foreclose the activities of rivals in a downstream market. That certainly could happen, but mostly in unique situations, given that the firm must sacrifice potential sales in the hope that the lost profits from those ventures could be made up elsewhere. There will always been some reluctance to engage in that activity, but if it takes place, the hard question is why it is wrong for a given company to cease serving a given client if there is no obligation to begin that service in the first place. It was therefore instructive that, in the AT&T/Time Warner merger, the parties responded to this threat by saying that it would not unilaterally cut off any outside distributor from its content, and offered all third party distributors a baseball-style arbitration for seven years, whereby each party would set a price for the transaction and require the arbitrator to pick the number that was closer to the arbitrator’s own assessment of price. That concession only makes sense because it does not impair the efficiency elements of the transaction.
The second risk is that the dominant firm may well acquire valuable information about the trade secrets of one of its trading partners, which it could put to good use in its own business. But leaks of information do not depend on having any market power. They only depend on having access to information that could be valuable, even if neither firm has close to a dominant position. These two risks seem far less important than the identifiable gains relating to double marginalization and supply chain management, so even though the Vertical Guidelines purport to treat them as matters of increased concern, most of these arrangements are likely to pass muster unless some clear inefficiency could be identified.
VI. The Qualcomm Saga: Wipeout and Recovery
There was, however, one complex transaction on which the DOJ and FTC have parted ways. It involves an attack on Qualcomm for alleged antitrust violations by the misuse of its patent portfolio. The story took place in two stages. At the District Court level, Judge Lucy Koh ordered Qualcomm to engage in a complex set of practices that would have required it, among other things, to license its technology to its direct competitors; to renegotiate the prices for its chipsets to its customers in the United States and overseas; to offer its chips without any collateral conditions; and to submit to arbitration of judicial resolution of disputes over reasonable royalties, other patent terms, and standard essential patents. That decision was in turn reversed by a panel of the Court of Appeals for the Ninth Circuit. The district court opinion had attracted extensive commentary, much but not all of it critical.
As the Ninth Circuit eventually concluded, the District Court’s Qualcomm decision involved multiple departures from established antitrust policy. First, Judge Koh attacked Qualcomm’s “no license-no chips” policy, which stated that Qualcomm would not sell its chips to parties that did not take out a technology license from the company for their use. To the FTC, this policy represented an unfair method of competition because it let Qualcomm use its alleged dominance in the market for chipsets to extract elevated royalties from all licensees of its patented technology under standard FRAND licenses. Thus the FTC complaint stated that Qualcomm’s “‘no license-no chips’ policy dramatically increases customers’ costs of challenging Qualcomm’s preferred license terms [for use of its patented technology]” such that “Qualcomm’s customers have accepted elevated royalties and other license terms that do not reflect an assessment of terms that a court or other neutral arbiter would determine to be fair and reasonable.”
To this charge, the simple rejoinder is that anyone who chooses to license the technology need not purchase any Qualcomm chips. At this point, the options are these: Those who want to buy both the chips and the technology from Qualcomm are indifferent to the allocation of the package costs between the two goods, so long as the total paid minimizes their costs. But for those who want to buy only the license, without the chips, the situation is as it was in Bement; namely, the firm is entitled to get the full cash price for the use of its technology when it is sold on a stand-alone basis, at which point any reference to “fair and reasonable” rates is irrelevant to the overall analysis. Commissioner Maureen K. Ohlhausen was correct in her dissent to the FTC decision to prosecute, when she denied that Qualcomm could use its “alleged chipset monopoly” to force higher rates for its technology packages. It is unclear, moreover, where that monopoly is. As the Ninth Circuit observed, the Taiwanese company MediaTek and the Korean firm Samsung have made major moves in the 5G market. The Ninth Circuit hit hard on this point when it wrote:
The FTC’s conclusion that OEM-level licensing does not further competition on the merits is not only belied by MediaTek and Intel’s entries into the modem chip markets in the 2015–2016 timeframe, it also gives inadequate weight to Qualcomm’s reasonable, procompetitive justification that licensing at the OEM and chip-supplier levels simultaneously would require the company to engage in “multi-level licensing,” leading to inefficiencies and less profit.
This second half of this passage offers a powerful efficiency judgment that is too often overlooked in antitrust cases. The patent exhaustion rule makes it clear that a patentee is only able to enforce its patent rights against the initial buyer or licensee. That limitation on freedom of contract in turn makes it unwise for Qualcomm to license any patents to rival chip manufacturers, because the patent exhaustion rule will then make it impossible for them to enforce any patent rights against original equipment manufacturers where the need for product control is clearly greater. Hence, the peculiarities of the patent law offer an additional justification for Qualcomm practices that might not be needed if the company could enter into comprehensive agreements at all levels, without fear of contractual nonenforcement. The limitations on freedom of contract have implications that go beyond any particular contract to cover a full range of interdependent transactions.
It is also worth noting that this particular arrangement does not result in any form of market foreclosure of the sort mentioned in Jefferson Parish. The risk of a tie-in is not present because any company that purchases only the technology is charged the same right as one that purchases both halves of the package, negating any risk of monopoly extraction. That risk would arguably be present if Qualcomm had announced a “no chips-no license” policy, which would have made it impossible to use anyone else’s chips with Qualcomm’s patented technology—itself a risky strategy, given that many people will choose to take neither instead of both. But, far from that possible form of foreclosure, the only asserted claim of foreclosure was the agreement between Qualcomm and Apple for Apple to take its chips exclusively from Qualcomm. There are obvious efficiencies in having a single source of supply, and, even if we put those aside, in Tampa Electric Co. v. Nashville Co., the Supreme Court held that an exclusive-dealing arrangement “does not violate the section unless the court believes it is probable that performance of the contract will foreclose competition in a substantial share of the line of commerce affected”—but no such line of commerce was identified by Judge Koh, and it would have to cover more than a single contract in what is clearly a global market with many versions of 5G technology in place, or quickly coming on board.
On this point, the Ninth Circuit went a long way to clarify the basic market definition when it correctly noted how the District Court strayed from its initial determination that the relevant markers were “the market for CDMA modem chips and the market for premium LTE modem chips,” only to then extend its analysis to cover all instances of cellular service, including the OEMs, who as the customers of Qualcomm are not entitled under the antitrust laws to have any particular prices for their services.
Nonetheless, it is hard to see how Qualcomm had power in some broader market. For example, the supposed preclusive power of Qualcomm’s exclusive agreement with Apple did not stop Apple from jettisoning Qualcomm entirely in 2015 by signing an exclusive dealing agreement with Intel—which was not challenged under the antitrust laws. Next, Apple sued Qualcomm for antitrust violations, after which Qualcomm sued Apple for improperly sharing Qualcomm’s trade secrets with Intel, purportedly to help Intel build its own rival 5G chips to power the next generation of Apple smart devices. Both cases were settled in April 2019, when Apple and Qualcomm dropped all claims against each other. Thereafter, Apple entered a six-year licensing agreement to use Qualcomm chips in its new phones—undoing its 2015 deal with Intel, while agreeing to pay Qualcomm between 4.5 and 4.7 billion dollars for past royalties. Hours after this settlement was announced, Intel exited the 5G market amid rumors that it would be unable to meet its development goals on time. A simple efficiency explanation for the event is that Qualcomm chips were superior—which Judge Koh did not even allude to in her opinion. Yet the Ninth Circuit picked up the irony when it noted that Qualcomm offered Apple “billions of dollars in incentive payments contingent on Apple sourcing its iPhone modem chips exclusively from Qualcomm and committing to purchase certain quantities of chips each year.” Monopolists normally do not have to pay their customers. By this arrangement, Qualcomm sought to retain the loyalty of Apple, which it would hardly have to do if Apple had no other party to whom it could turn.
As the Ninth Circuit powerfully demonstrated, Judge Koh also faltered dramatically with her second argument, which was that Qualcomm had a duty under the antitrust laws to deal with its direct competitors. As a general matter, the antitrust laws are intended to preserve competitive conditions, so it is odd in the extreme to say that one competitor is under a duty to furnish assistance to its rivals, especially because a court will have to foist a complex agreement on an unwilling party to enforce such a duty. It was for these reasons that Justice Scalia, in a widely quoted passage in Verizon Communications v. Law Offices of Curtis V. Trinko, wrote: “We have been very cautious in recognizing such exceptions, because of the uncertain virtue of forced sharing and the difficulty of identifying and remedying anticompetitive conduct by a single firm.”
Nonetheless, Judge Koh found that “[u]nder certain circumstances, a refusal to cooperate with rivals can constitute anticompetitive conduct and violate § 2.” For that proposition, she relied on the much-criticized decision of Aspen Skiing Co. v. Aspen Highlands Skiing Co, which held that it could constitute an antitrust violation for three ski resorts, which previously offered package deals with a fourth, to break off that arrangement. Aspen Skiing did not require the court to force novel obligations on competitors who were never cooperators. Instead, the companies only had to reinstate the earlier agreement—and even that conclusion is dubious, given that the other three ski resorts might have been able to show that they terminated the agreement because the plaintiff’s facilities did not reach the standard found in the other lifts, at which point the pooling arrangement (which helps bring people to the slopes) is offset by an efficiency disadvantage to the three stronger resorts. But here, the only prior relationships were twenty years earlier in unrelated transactions. The huge gap provoked an angry response by Christine Wilson, a new FTC Commissioner who explicitly broke ranks with her predecessors, noting that it was indefensible to impose a duty to deal today based on Qualcomm business practices that “involved licensing different patents, to different competitors, in a different century.” As if on cue, the Ninth Circuit followed this analysis closely in reversing Judge Koh.
What is most striking about Judge Koh’s Qualcomm decision is how it sidesteps the fundamental task of antitrust law, which is to weigh the restrictive practices of a defendant against any efficiency justifications that it can advance in justification of its behavior. The various permutations of tie-ins and exclusive dealing call out for that treatment—but, although the word “efficiency” is mentioned a few times in the opinion, Judge Koh made no sustained effort to understand either the industry or the business practices. Yet on top of all that, Judge Koh ordered a set of world-wide remedies that were as comprehensive and demanding as her opinion was weak. The Ninth Circuit followed its stay with an emphatic reversal of the decision—which did not come a moment too soon. But the saga is not yet over; the FTC is pursuing en banc review before the full Ninth Circuit.
The purpose of this chapter was to tease out the relationships between patent and antitrust law, insofar as they seek to address, in their separate ways, the problem of monopoly. No one should doubt that the mission is a legitimate one; but, as with all such arrangements, it is a task fraught with risk—given that false charges of monopoly are, in their own way, as dangerous as cases of monopoly practices that escape detection and sanction. To get the right result, it is important first to note that property rights are not tantamount to monopolies, solely because they are exclusive to their holder. This basic insight should not only govern such distinctive patent law questions as patent eligibility and patent abuse, but it should also carry over so that at no point does the antitrust law impose liability for monopoly behavior that passes muster under the patent law, or in the reverse. Harmonization is a minimum condition for a successful integration of the two bodies of law, but it is not a sufficient condition.
At the same time, it is important to make sure that there are no sharp discontinuities between the antitrust laws as they apply to patents and as they apply to other areas. In large measure, the doctrinal unification across separate substantive areas should stabilize this entire body of law. But that will only happen to the extent that the standard theories developed to date continue to apply to patent and nonpatent contexts alike. The subject is far too big to be raised in a final paragraph, but it is worth noting that the entire structure will be transformed for the worse if populist versions of antitrust law are allowed to dominate the area. It is a general proposition that bodies of law are good at doing one, and only one, task. The moment they are given two jobs—to promote efficiency and to create income equality, say—goals conflict and the doctrine muddles. In this regard, the Ninth Circuit decision to overturn the District Court is a watershed decision because it prevented what could have been a first major step in dismantling what has been by and large a well-crafted structure that integrates patent and antitrust law in a way that has served this nation well.
* Laurence A. Tisch Professor of Law, The New York University School of Law; The Peter and Kirsten Bedford Senior Fellow, The Hoover Institution; The James Parker Hall Distinguished Service Professor of Law Emeritus and Senior Lecturer, The University of Chicago. I should like to thank Ethan Hoffman and Alexander Krzepicki of the George Mason Law School Class of 2021, and Kenneth Lee, Micah Quigley, and Riley Walters of the University of Chicago, Class of 2021, and Christian McGuire of the University of Chicago, Class of 2022, for their valuable research assistance. Adam Mossoff provided useful historical references to explain the overlap between patent and antitrust law.
 For my more exhaustive account of the historical origins of the patent clause, see Richard A. Epstein, Patent Originalism, 71 Case Western L. Rev. (forthcoming 2021).
 Intergraph Corp. v. Intel Corp., 195 F.3d 1346, 1362 (Fed. Cir. 1999).
 For one such inefficiency, see Quanta Computer, Inc. v. LG Electronics, Inc., 553 U.S. 617 (2008).
 411 F.Supp.3d 658, 818 (N.D. Cal. 2019), rev’d 969 F.3d 974 (9th Cir. 2020).
 Karl B. Lutz, Patents and Science: A Clarification of the Patent Clause of the U.S. Constitution, 32 J. Pat. Off. Soc’y 83, 87 (1950). “Useful arts” originally referred to the practical skills and methods of manufacture and craftsmanship taught as vocational subjects (i.e., ways of making) as distinct from “liberal arts” which were academic subjects taught for intellectual development (i.e., ways of thinking).
 The phrase comes from 35 U.S.C. § 101, which reads in full: “Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.”
 For the mistake, see Brink Lindsey & Daniel Takash, Niskanen Center, Why ‘Intellectual Property’ is a Misnomer (2019) (noting that it is common for a woman who hears a song whistled by a neighbor in the field to sing it at night to put her baby to sleep. The song in question, however, was not copyrighted in the first place, and there was no commercial use. This makes Lindsey and Takash’s example inapposite.).
 For a systematic treatment of these issues, see Richard A. Epstein, What Is So Special About Intangible Property? The Case for Intelligent Carryovers, in Competition Policy and Patent Law Under Uncertainty: Regulating Innovation 42 (G. Manne & J. Wright, eds., 2011).
 Restatement (Third) of Unfair Competition §39 (Am. Law Inst. 1995).
 Restatement (First) of Torts § 757, cmt. b (Am. Law Inst. 1934).
 See Kewanee Oil Co. v. Bicron Corp., 416 U.S. 470, 484 (1974).
 Rockwell Graphic Systems, Inc. v. DEV Industries, Inc, 925 F.2d 174 (7th Cir. 1991).
 See Ruckelshaus v. Monsanto, 467 U.S. 986 (1984); Philip Morris v. Reilly, 312 F.3d 24 (1st Cir. 2002). Monsanto first recognizes that trade secrets are property, id. at 1000–03, only to reverse field and hold that the government can require the release of trade secret information to a competitor, whose own secret formula must be approved for use on health and safety grounds by the United States. Id. at 1005–06. That argument necessarily rejected the application of the unconstitutional conditions doctrine as it applied to this form of monopoly power. Philip Morris departed from the Supreme Court precedent on this point and by a divided vote held that Massachusetts could not force the limited publication of trade secret information as a condition for letting a cigarette company market its product within the state. There are other ways to make sure that tobacco additives do not contain harmful substances, without requiring release of valuable trade secret information. For example, states could mandate the publication of a list, in order of concentration, of key additives—without revealing the full trade secret to Philip Morris competitors. For discussion, see Richard A. Epstein, The Constitutional Protection of Trade Secrets under the Takings Clause, 71 U. Chi. L. Rev. 57 (2004).
 35 U.S.C. § 112 (“The specification shall contain a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the same, and shall set forth the best mode contemplated by the inventor or joint inventor of carrying out the invention.”).
 See Jonathan M. Barnett, Innovators, Firms, and Markets: The Organizational Logic of Intellectual Property (forthcoming, Oxford University Press). “Certain firms—especially older, larger, and more integrated firms—have the ability to earn returns on innovation without recourse to IP rights, while other types of firms—especially, younger, smaller, and less integrated firms—do not. Where the largest, most established, and most highly integrated firms can mimic and even outperform the IP rights delivered by the state, reducing IP rights can raise entry costs and shelter incumbents against the competitive threat posed by more efficient innovators.“
 Paul C. Clement, Patent Rights vs. Property, The Framer’s Understanding of Patents (2019). For my response, see Richard A. Epstein, Patent Originalism, Case Western L. Rev. (forthcoming 2020).
 566 U.S. 66 (2012) (holding a method for calculating proper dosage of thiopurine to treat autoimmune diseases should be treated as an unpatentable law of nature).
 573 U.S. 208 (2014) (holding that techniques to minimize settlement risk—i.e. nonpayment—by using third parties are unpatentable abstract ideas).
 The State of Patent Eligibility in America: Part II hearing before the Subcomm. On Intellectual Property of the S. Comm. on the Judiciary 116th Cong. (2019) (statement of David W., Jones) (“although medical diagnostics is often cited as the area of life sciences most impacted, according to Crunchbase, almost $400 million was invested in bio-diagnostics during the month of May, 2019, alone.”).
 The State of Patent Eligibility in America: Part II hearing before the Subcomm. On Intellectual Property of the S. Comm. on the Judiciary 116th Cong. (2019) (statement of Adam Mossoff).
 Id. at 2, where he instances these cases: Cleveland Clinic Found. v. True Health Diagnostics LLC, 760 Fed. Appx. 1013 (Fed. Cir. 2019) (nonprecedential) (invalidating a patent on a biotech-based medical test for detecting heart disease as covering a patent-ineligible law of nature); Athena Diagnostics, Inc. v. Mayo Collaborative Servs., LLC, 915 F.3d 743 (Fed. Cir. 2019) (invalidating a patent on a biotech-based medical test for identifying neurological disorders as covering a patent-ineligible law of nature); Ariosa Diagnostics, Inc. v. Sequenom, Inc., 788 F.3d 1371 (Fed. Cir. 2015) (invalidating a patent on non-invasive prenatal test using fetal DNA found in the blood of the mother as covering a patent-ineligible law of nature and natural phenomenon).
 33 F.3d 1526 (Fed. Cir. 1994) (upholding a patent for “creating a smooth waveform display in a digital oscilloscope”).
 149 F.3d 1368 (Fed. Cir. 1998) (use of a data processing system for implementing an investment structure).
 Id. at 1374.
 56 U.S. 62 (1853).
 Id. at 112–13.
 For an exhaustive review of these earlier practices, see Adam Mossoff, O’Reilly v. Morse and Claiming a “Principle” in Antebellum Era Patent Law, Case Western Reserve Law Rev. (forthcoming 2021): “[t]he early nineteenth-century legal practice in patent law in which patent claims secured the essential ‘principle’ of an invention, as distinguished from the peripheral claims today that defined the boundaries of the property right in an invention.”
 For an early recognition of the doctrine of equivalents, see Graver Tank & Mfg. Co. v Linde Air Prods. Inc., 339 U.S. 605, 607 (1948) (Courts “have also recognized that to permit imitation of a patented invention which does not copy every literal detail would be to convert the protection of the patent grant into a hollow and useless thing.”).
 O’Reilly, 56 U.S. at 107–108.
 569 U.S. 576 (2013).
 189 F. 95 (C.C.S.D.N.Y. 1911), aff’d in part, rev’d in part, 196 F. 496 (2d Cir. 1912). For a discussion of the connection between Parke-Davis & Myriad written before the Supreme Court decision, see Jon M. Harkness, Dicta on Adrenalin(e): Myriad Problems with Learned Hand’s Product-of-Nature, 93, J. Patent & Trademark Office Soc’y 363 (2011).
 See supra at note 15.
 333 U.S. 27 (1948).
 Kalo Inoculant Co. v. Funk Bros. Seed Co., 161 F.2d 981, 986 (7th Cir. 1947).
 1 F. Cas. 755 (1833).
 Id. at 756
 1 F. Cas. 476, 477 (C.C.D. Ohio 1855) (No. 225).
 3 F. Cas. 441, 444 (C.C.D. Ohio 1874) (No. 1,434) (Swayne, C.J.).
 Thomas W. Merrill, Property and the Right to Exclude, 77 Neb. L. Rev. 730 (1998).
 Kaiser Aetna v. United States, 444 U.S. 164, 176 (1979).
 See, e.g., Rands v. United States, 389 U.S. 121 (1967) (sadly held the opposite).
 For discussion, see Richard A. Epstein, Playing by Different Rules? Property Rights in Land and Water, in Property in Land and other Resources 317 (Daniel H. Cole & Elinor Ostrom, eds., 2012).
 See United States v. Willow River, 324 U.S. 499, 510 (1945) (“Rights, property or otherwise, which are absolute against all the world are certainly rare, and water rights are not among them. Whatever rights may be as between equals such as riparian owners, they are not the measure of riparian rights on a navigable stream relative to the function of the Government in improving navigation. Where these interests conflict they are not to be reconciled as between equals, but the private interest must give way to a superior right, or perhaps it would be more accurate to say that as against the Government such private interest is not a right at all.”).
 This is yet another version of the so-called marginal cost controversy. See Ronald H. Coase, The Marginal Cost Controversy, 13 Economica 169 (1946); William Vickery, Some Objections to Marginal-Cost Pricing, 56 J.P. Econ. 218 (1948); John Duffy, The Marginal Cost Controversy in Intellectual Property, 71 U. Chi. L. Rev. 37 (2004).
 For a discussion on how this works, see Richard A. Epstein, Positive and Negative Externalities in Real Estate Development, 102 Minn. L. Rev. 1493 (2018).
 186 U.S. 70 (1902).
 Id. at 84.
 See Adam Mossoff, A Simple Conveyance Rule for Complex Innovation, 44 Tulsa L. Rev. 714-15 (2009) (citations omitted).
 National Harrow, 186 U.S at 88–89.
 See Joseph Shaanan, Ricardian or Monopoly Rents? The Perspective of Potential Entrants, 32 Eastern Econ. J. 19 (2006).
 104 Eng. Rep. 206 (K.B. 1810). For a detailed account of the case, see Richard A. Epstein, Principles for a Free Society: Reconciling Individual Liberty with the Common Good 282–286 (1998).
 94 U.S. 113, 127 (1876) (citing Allnut v. Inglis 104 Eng. Rep. 206 (K.B. 1810)).
 See, e.g., Chicago, Mil. & St. Paul R. Co., (The Minnesota Rate Cases), 134 U.S. 418 (1890) (cautioning against the false equivalence of regulation and confiscation); Smyth v. Ames, 169 U.S. 466 (1898) (same).
 Bement, 186 U.S. at 91.
 See, e.g., Mark A. Lemley & Carl Shapiro, A Simple Approach to Setting Reasonable Royalties for Standard-Essential Patents, 28 Berk. Tech. L. J. 1135 (2013) (using baseball final offer arbitration to resolve rate disputes); Richard A. Epstein & Kayvan Noroozi, Why incentives for “Patent Holdout” Threaten to Dismantle FRAND, and Why it Matters, 32 Berk. Tech. L. J. 1381 (2017) (requiring licensees to make offers before attacking FRAND rates, given a bilateral risk of holdouts).
 For a careful background discussion, see Keith Hylton, Antitrust Law: Economic Theory & Common Law Evolution 278–310 (2003).
 224 U.S. 1 (1912) (where a tie-in of the sale of ink was tied to the use of a patented product).
 Id. at 51.
 38 Stat. 731(1914). This section was held to override A.B. Dick. See Motion Picture Patents Co. v. Universal Film Mfg. Co., 243 U.S. 502, 517–518 (1917).
 Standard Oil of California v. United States, 337 U.S. 293, 305–306 (1949) (striking down exclusive dealing contracts).
 332 U.S. 392 (1947).
 Meg Prater, What is a Right of First Refusal? Absolutely Everything You Need to Know, HubSpot (Oct. 8, 2019), https://blog.hubspot.com/sales/right-first-refusal.
 United States v. Loew’s Inc, 371 U.S. 38, 46 (1962) (citing International Salt, 332 U.S. at 392).
 466 U.S. 2 (1984).
 547 U.S. 28 (2006).
 Department of Justice & Federal Trade Commission, Antitrust Guidelines for the Licensing of Intellectual Property (2017), https://www.ftc.gov/system/files/documents/public_statements/
1049793/ip_guidelines_2017.pdf (“These Guidelines embody three general principles: (a) for the purpose of antitrust analysis, the Agencies apply the same analysis to conduct involving intellectual property as to conduct involving other forms of property, taking into account the specific characteristics of a particular property right; (b) the Agencies do not presume that intellectual property creates market power in the antitrust context; and (c) the Agencies recognize that intellectual property licensing allows firms to combine complementary factors of production and is generally procompetitive.”).
 For more on vertical arrangements, see John M. Yun, Vertical Mergers and Integration in Digital Markets, in The GAI Report on the Digital Economy (2020), and Daniel P. O’Brien, The Economics of Vertical Restraints in Digital Markets, in The GAI Report on the Digital Economy (2020).
 See James L. Hamilton & Ibrahim Mqasqas, Double Marginalization and Vertical Integration: New Lessons from Extensions of the Classic Case, 62 S. Econ. J. 567, 567 (1996) (“Suppose that an upstream industry sells an intermediate good to a downstream industry, which in turn produces a final product that it sells to consumers. Then, because the upstream and downstream industries independently engage in noncompetitive pricing, the firms in each industry only see the effect on their output restriction on their own profits, and do not see that their output restriction also affects the profits of the firms in the other industry. This myopia creates a ‘vertical externality’ that vertical integration would internalize. In the simplest case, when the products are homogeneous and final production has fixed input proportions, the conventional wisdom about double marginalization is that ‘. . . the integrated industry makes more profit than the nonintegrated industry, and the consumer price is lower in the case of the integrated firm.’”).
 Treaty of Westphalia, art. LXX, Oct. 1648, Art. LXX, https://avalon.law.yale.edu/17th_century
 For the early history see United States v. Winslow, 227 U.S. 202 (1913). For discussion, see Richard A. Epstein, Antitrust Consent Decrees in Theory and Practice: Why Less Is More 40–54 (2007).
 Winslow, 227 U.S. at 217.
 U.S. Dep’t of Justice & Fed. Trade Comm’n, Vertical merger Guidelines, at 4 (2020), https://www.ftc.gov/system/files/documents/reports/us-department-justice-federal-trade-commission-vertical-merger-guidelines/vertical_merger_guidelines_6-30-20.pdf.
 Tony Romm, AT&T pledged that CNN and TBS channels won’t go dark on any pay TV services if it’s allowed to buy Time Warner, Vox (Nov. 28, 2017), https://www.vox.com/2017/11/28/16712632/att-cnn-tbs-wont-go-dark-pay-tv-buy-time-warner-merger.
 411 F.Supp. 3d 658, 818 (N.D. Cal. 2019).
 Fed. Trade Comm’n v. Qualcomm, Inc., 969 F.3d 974 (9th Cir. 2020).
 See, e.g., Richard A. Epstein, Judge Koh’s Monopolization Mania: Her Novel Assault Against Qualcomm is an Abuse of Antitrust Theory, 97 Neb. L. Rev. 241 (2019) (by way of disclosure, I advised Qualcomm during the course of the litigation); Douglas H. Ginsburg, Joshua D. Wright & Lindsey M. Edwards, Section 2 Mangled: FTC v. Qualcomm on the Duty to Deal, Price Squeezes, and Exclusive Dealing, 8 J. Antitrust Enforcement 335 (2020). But see Brief of Amicus Curiae Open Markets Institute in Support of Plaintiff-Appellee, Federal Trade Commission v. Qualcomm Incorporated, 969 F.3d 974 (9th Cir. 2020).
 Federal Trade Commission’s Complaint for Equitable Relief, Fed. Trade Comm’n v. Qualcomm, Inc., No. 5:17-CV-00220, 2017 WL 242848 (N.D. Cal. Jan. 17, 2017).
 See Jeremy Horwitz, Samsung and Mediatek Emerge as Qualcomm’s top 5G Chip Rivals in 2020, VentureBeat (Nov. 8, 2019), https://venturebeat.com/2019/11/08/samsung-and-mediatek-emerge-as-qualcomms-top-5g-chip-rivals-in-2020/.
 FTC v. Qualcomm, 969 F.3d at 996.
 See id. at 984, quoting from Quanta Computer v. LG Electronics, 553 U.S. at 625.
 365 U.S. 320 (1961).
 Id. at 327.
 FTC v. Qualcomm 411 F. Supp. 3d at 683.
 See Qualcomm Inc. v. Apple Inc., Trade Secrets Institute, Docket No. 37-2017-00041389-CU-BC-NC (Sept. 24, 2018), http://tsi.brooklaw.edu/cases/qualcomm-inc-v-apple-inc.
 For the public announcement, see Qualcomm and Apple Agree to Drop All Litigation, BusinessWire (Apr. 16, 2019), https://www.businesswire.com/news/home/20190416005931/en/.
 See Jean Baptiste Su, Analysis: Apple To Drop Qualcomm Modems From Next iPhones For Intel, Samsung Wireless Chips, Forbes (July 26, 2018), https://www.forbes.com/sites/jeanbaptiste/2018/07/26/analysis-apple-to-drop-qualcomm-modems-from-next-iphones-for-intel-samsung-chips/#3d9e06fc4960.
 For one brief account of the supposed terms of settlement, see Qualcomm Got $4.7 Billion from Apple Settlement According to Earnings Release, MacRumors (May 1, 2019), https://www.macrumors.com
 Chaim Gartenberg, Intel says Apple and Qualcomm’s surprise settlement pushed it to exit mobile 5G, The Verge (Apr. 25, 2019), https://www.theverge.com/2019/4/25/18516830/intel-apple-qualcomm-surprise-settlement-pushed-exit-mobile-5g-modems.
 Qualcomm v. FCC, 969 F.3d at 986.
 Id. at 993-95.
 540 U.S. 398 (2004).
 Id. at 408.
 FTC v. Qualcomm, 411 F.Supp.3d at 758 (quoting Trinko, 540 U.S. at 411).
 See Dennis W. Carlton, A General Analysis of Exclusionary Conduct And Refusal to Deal—Why Aspen and Kodak Are Misguided (NBER Working Paper 8105, 2001), https://www.nber.org/papers/w8105.pdf.
 472 U.S. 585 (1985).
 Christine Wilson, A Court’s Dangerous Antitrust Overreach, Wall St. J. (May 28, 2019).
 FTC v. Qualcomm, 969 F.3d at 994-95.