LICENSING FREEDOM AND THE LIMITS OF ANTITRUST
June 3, 2005
Defining the relationship of intellectual property rights and competition law is an important economic issue in Europe and the United States. This paper attempts to outline some bedrock principles of intellectual property and antitrust policy in the United States, then discuss how they explain, and in some cases require, the current U.S. approach to a series of specific licensing practices. The basic U.S. approach, reflected in the 1995 DOJ/FTC Guidelines for the Licensing of Intellectual Property, calls for flexible application of economic analysis to licensing practices. And the recent trend has been one of increasing convergence in U.S. and European approaches to IP licensing questions, as seen in the new revisions to the Technology Transfer Block Exemption and accompanying guidelines.
The opening question for this workshop asks whether intellectual property is like other property. This question has been discussed to death many times over in recent years, without much improvement on the answer given ten years ago in the 1995 Guidelines. In short, for competition law purposes, intellectual property should be treated in essentially the same way as other forms of property, though this does not mean that it is in all respects the same as other forms of property. "Intellectual property is thus neither particularly free from scrutiny under the antitrust laws, nor particularly suspect under them."(1)
This answer means rejection of the hostility toward intellectual property that held sway in the U.S. during the 1970's. During this era, the Antitrust Division had a section devoted to attacking IP licensing practices that we routinely applaud today. This was the era of the "Nine No Nos," during which we applied per se rules of illegality to many licensing practices. The contention that IP should be treated essentially like other forms of property at that time was meant as a call to curtail hostility toward IP rights, a call for the end of disfavored status for IP.
Today, in contrast, our policy is animated by the recognition that IP licensing is generally procompetitive. But the modern answer to the question whether IP is like other forms of property also requires rejection of extreme claims of privilege on the part of IP owners. Today, the statement that IP is essentially like other forms of property is often heard in arguments against claims for complete exemption from antitrust scrutiny. The mere presence of an IP right that somehow figures in a course of otherwise anticompetitive conduct does not act as a talisman that wards off all antitrust enforcement. The classic statement on this point is contained in United States v. Microsoft Corp., 253 F.3d 34 (D.C. Cir. 2001) ("Microsoft's primary copyright argument borders upon the frivolous. The company claims an absolute and unfettered right to use its intellectual property as it wishes. . . . That is no more correct than the proposition that use of one's personal property, such as a baseball bat, cannot give rise to tort liability.").
II. First Principles of U.S. Intellectual Property Law and Antitrust
Sound antitrust enforcement condemns anticompetitive conduct. It does not attempt to regulate the amount of competition in a general sense or address vague questions of fairness. It does not attempt to create an affirmative incentive for procompetitive conduct, by promising any specific reward or legal recognition for competitors who play by the rules. It focuses on specific anticompetitive actions, as judged by their effects on markets and consumer welfare. Although this narrow focus is a limitation, at the same time it is a great strength--it makes possible objectivity, predictability, and transparency.
Intellectual property laws, by contrast, provide a complex system of affirmative rewards for an important type of procompetitive behavior--innovation. They take consumer welfare into account, but in different ways than does antitrust. First, they reward innovators with exclusive rights that serve as an incentive to bring new and improved goods and services to market. The hope is that such innovations will lead to increased competition and increased consumer welfare in the long term. Second, they strike a balance between these rights and certain types of public access, such as fair use under copyright law(2) or the disclosure requirement and the limited term of patents.(3) They also include a fail-safe procedure under which a rival or a customer can sue to declare an intellectual property right noninfringed or unenforceable for a number of reasons. So the legislature, via the IP laws, has struck a balance between the rights of IP owners, the rights of consumers, and concerns for a competitive marketplace. This may or may not be the correct balance; nevertheless, it is the one the legislature has chosen.
It is important to understand precisely what reward is offered by the IP laws. Each type of IP right provides "exclusivity" for its owner. What does this exclusivity mean? It does not mean a right to commercialize any invention or creation. The owner of an improvement patent, for example, may find itself blocked from practicing its own patent if it cannot secure permission from the original patentee. Instead, what IP rights provide is the right to exclude others. The right to exclude is not simply one of the rights provided by intellectual property, it is the fundamental right, the foundation upon which the entire IP system is built.
III. Specific Practices and the Freedom to License
These bedrock principles of antitrust and intellectual property law inform the proper approach to specific licensing and IP-related practices. A decade's experience with the Guidelines, together with subsequent judicial precedent, provide reliable guidance on several issues in the U.S. On many, but not all, of these issues, it is also possible to rely on continued transatlantic convergence.
The subject of unilateral refusals to license intellectual property is one in which the premise that IP is essentially like other forms of property has sometimes been stretched beyond sensible limits. Because, outside the area of IP, antitrust law holds out the possibility of rare exceptions to the principle that parties are free unilaterally to refuse to deal with others, the argument is that there must therefore be some circumstance in which the unilateral, unconditional refusal to license a patent must constitute an antitrust violation. With a single much-criticized exception, this is an argument that has never found support in any U.S. legal decision. At this point in the development of U.S. law, it is safe to say that this argument is without merit.
A unilateral, unconditional refusal to license a valid patent cannot, by itself, result in antitrust liability under U.S. law. It is instructive that the very notion of such liability was not even discussed in the 1995 Guidelines. Instead, the Guidelines unequivocally state that, even in the case of IP that conveys market or monopoly power, that power does not "impose on the intellectual property owner an obligation to license the use of that property to others."(4) This is hardly surprising, as the right to choose whether the license has long been recognized by the U.S. Supreme Court as the core of the patent right.(5) Although the Supreme Court decisions are not directly on point, lower courts have correctly held that the unilateral, unconditional refusal to license a valid patent does not give rise to liability as an improper refusal to deal under Section 2 of the Sherman Act.(6) But of course, while an intellectual property owner has the right to decide not to license its technology, the owner does not have the right to impose conditions on licensees that would effectively extend an intellectual property right beyond the limits of the Patent Act.(7)
The clarity of U.S. law on unilateral refusals was enhanced by last year's Supreme Court decision in Verizon Communications Inc. v. Law Offices of Curtis V. Trinko, LLP.(8) In Trinko, the Supreme Court found that private plaintiffs did not state an antitrust claim when they alleged a failure by a communications provider, Verizon, to provide adequate assistance to its rivals. The Court showed great skepticism about expanding liability for the refusal to deal because such liability "may lessen the incentive for the monopolist, the rival, or both to invest in . . . economically beneficial facilities" and "also requires antitrust courts to act as central planners . . . a role for which they are ill-suited."(9) The Court posed the question as being whether the narrow list of exceptions to the general rule against liability should be expanded.(10) Although Trinko was not an intellectual property case--the rights in that case were governed by the Telecommunications Act--the Supreme Court would apply similar logic under the Patent Act. Given the many cases indicating that the right to exclude is a fundamental right embodied in the patent grant, it is safe to say that liability for the unilateral, unconditional refusal to license a valid patent is not going to be added to the narrow list of exceptions the Court mentioned.
When analyzing the effects of a unilateral refusal to deal, one cannot merely consider the effect on a rival that is refused a license; one must also consider the alternative world in which the IP owner would have had less of an incentive to innovate because he could not be assured of the right to refuse to license. Would that IP owner have chosen to innovate less? If so, would competition or consumer welfare have been better off with the present state of affairs, including the right to refuse? In the short term, it will always be more efficient to disregard the IP right and allow duplication. The IP system rests on the idea of long-term innovation incentives, so we must think about the long-term effects of a rule imposing liability in this context. That is entirely consistent with antitrust policy related to exclusionary conduct, which also focuses on dynamic competition and long-term effects. Where we cannot reliably predict the effects of enforcement decisions, false positives are likely, and the increased uncertainty itself will raise costs to businesses and enforcers.
It is useful to remember that the creation of intellectual property tends to add to consumer choices, rather than to reduce them. The development of intellectual property for new technological solutions usually does not cause older solutions to be withdrawn from a marketplace; instead, it increases competition, which tends to erode the prices of the old solutions over time, increasing choice and consumer welfare. Of course, a patent sometimes issues for an obvious or previously-known solution to a problem, but such a patent should be invalidated, and the proper remedy is to seek invalidation under the patent laws.
Does this mean that the policy on unilateral refusals conflicts with EU law as stated in IMS Health?(11) At this time, that it is difficult to tell. The European Court of Justice decision, issued a year ago, began by stating that a refusal to license a copyright "cannot in itself" constitute an abuse of a dominant position. That seems to match the U.S. view on unilateral refusals to license. But the court added that liability might occur if: (1) the refusal prevents the emergence of a new product for which consumer demand exists; (2) the refusal is not justified by any objective considerations; and (3) the refusal excludes competition in a "secondary market." It is not clear how these three factors will be interpreted, or whether the same reasoning would apply to other contexts such as a refusal to license a patent. (Some have observed that the IP right asserted in IMS was relatively weak, and that the lack of a unified European system of IP rights may explain differing attitudes toward antitrust liability in this context.) It will be interesting to see how the IMS Health decision is applied, for example in the Microsoft appeal. While the Justice Department required Microsoft to make certain IP available to its competitors as part of the agreed remedy for antitrust violations, the European Commission imposed liability for the failure to make IP available. It will be up to the Court of First Instance to determine whether this was permissible under EU law.
The Antitrust Division sometimes hears complaints about demands for large royalties. Most frequently, although not always, the complaints arise in the context of a technical standard. According to the complainants, one or more patent holders can "hold up" licensees by waiting until participants are locked into the standard, then charging an allegedly "excessive" royalty for patents that cover the standard. The U.S. Federal Trade Commission has brought antitrust enforcement actions related to this issue in two recent cases, Rambus and Unocal. Both cases are ongoing.
Bringing a complaint to the Antitrust Division about "excessive" royalties, without more, is a losing strategy. Antitrust enforcers are not in the business of price control. We protect a competitive process, not a particular result, and particularly not a specific price. In fact, if a monopoly is lawfully obtained, whether derived from IP rights or otherwise, we do not even object to setting a monopoly price. A high patent royalty rate, after all, might just reflect that the Patent Act is functioning correctly and the market is rewarding an inventor for a pioneering invention. When a complainant begins a presentation by telling the Antitrust Division that a royalty rate is "excessive," the staff responds that the complainant is putting the cart before the horse. A complaining party must first identify some anticompetitive conduct beyond a mere unilateral refusal to license and beyond the mere attempt to charge, where a lawful monopoly exists, a monopoly price.
Many situations of standard setting "hold up" can be mitigated by disclosure in the ex ante phase, before the standard is set. For example, if all participants are required to disclose their financial interest in any version of the standard--including any patents they own or are seeking on the technology--other participants can adjust their behavior accordingly. If a participant agrees to disclose but then fails to do so, it can be liable for breach of contract or fraud. Such liability would hinge on a pattern of breaches, frauds, or other unlawful conduct. If antitrust liability is also contemplated, it would require, in addition, proof of market effects.
Increasingly, standards development organizations are requiring "reasonable and non-discriminatory" (RAND) licensing, which is a partial solution. A difficulty of RAND, however, is that the parties tend to disagree later about what level of royalty rate is "reasonable." It would be useful to clarify the legal status of ex ante negotiations over price. Some standards development organizations have reported to the Department of Justice that they currently avoid any discussion of actual royalty rates, due in part to fear of antitrust liability.(12) It would be a strange result if antitrust policy is being used to prevent price competition. There is a possibility of anticompetitive effects from ex ante license fee negotiations, but it seems only reasonable to balance that concern against the inefficiencies of ex post negotiations and licensing hold up. It is interesting to note that the EU licensing guidelines already address this point: in their Paragraph 225, the guidelines state that firms normally should be allowed to negotiate royalty rates before a standard setting effort, as well as after a standard is set.
Barriers to discussing licensing rates may not be entirely law-related. Some standard setting participants do not want the distraction of considering licensing terms. Engineers and other technical contributors may prefer to leave the lawyers at home and limit discussions to technical issues alone. So there may be powerful incentives to keep the status quo. If that is the case, this may be yet another area where the outcomes can be imperfect but antitrust does not provide a solution.
Compulsory licensing is another place where enforcers need to be fully aware of antitrust's limitations.(13) Licensing can be an effective remedy in some contexts; for example, for merger cases, it can serve as a less drastic alternative to a divestiture. But in the first instance, there must be conduct that warrants a remedy--licensing is only a remedy, not a liability theory. And there are practical reasons to tread carefully when considering compulsory licensing: designing and enforcing such licenses is complex and can be an invitation to endless ancillary compliance litigation. As explained in the Trinko case, an enforcement agency should not impose a duty to deal that it cannot reasonably supervise, since this risks assuming the day-to-day controls characteristic of a regulatory agency. For these and other reasons, compulsory licensing of intellectual property as an antitrust remedy should be a rare beast.
There has been much talk in recent years, and perhaps worldwide, about whether there is a problem of "excessive patenting," meaning patents being granted too easily or in too great a number. Of course, it is the job of the U.S. Patent and Trademark Office in the Department of Commerce--not the Department of Justice--to make and regulate awards of patent rights. The PTO has mechanisms for reconsidering specific patents and hearing complaints about the patent system as a whole, and it employs untold hundreds of patent experts. The Federal Trade Commission, an independent agency, has issued a useful report on possible improvements to the patent system.(14) The National Academies have also issued a report.(15)
It is open to question whether antitrust analysis, which is specific and effects-based, can be applied to a question as broad as "excessive patenting." To know whether patenting is excessive, we would first have to make a conclusion about the "but-for" world. If fewer patents were granted, would innovation have decreased? Would firms have reduced their research and development in areas that currently are covered by patents, and would the result have been fewer benefits for consumers? Antitrust enforcement is not well suited to answering such questions. These questions should be directed, instead, to the patent authorities or to legislators.
Of course, this point must not be overstated. Part of the patent system is court review of patent enforceability.(16) In the appropriate case the Antitrust Division will examine enforceability and, if necessary, challenge the validity or scope of a patent as part of an antitrust claim. This is not necessary where a patent-related practice will be lawful (or at least, does not violate the antitrust laws) or unlawful regardless of the patent's enforceability. But if the conduct would have violated the antitrust laws in the absence of patent rights, is difficult to address fundamental questions about the but-for world--here, meaning the world that would have existed without the allegedly anticompetitive patent-related practice--unless one knows whether the patent owner could have won an infringement claim. If the patent is valid, all entry before its expiration is a competitive "gift," but if it is invalid, any delay in entry due to threatened patent enforcement is a competitive harm. Just three months ago, an appellate court asserted this need to examine the but-for world in a case involving the antitrust analysis of a patent settlement. According to the court, it is impossible to measure a patent settlement's effect on competition unless one first makes a conclusion about the validity and enforceability of the patent.(17) A petition for rehearing in that case is pending.
Last on my list of specific issues is the concept of market power. Intellectual property cannot be presumed to establish market power. While intellectual property grants exclusive rights, these rights are not monopolies in the economic sense: they do not necessarily provide a large share of any commercial market and they do not necessarily lead to the ability to raise prices in a market. A single patent, for example, may have dozens of close substitutes. The mere presence of an intellectual property right does not permit an antitrust enforcer to skip the crucial steps of market definition and determining market effects.
In the view of the Department of Justice and the Federal Trade Commission, the idea that IP rights cannot be presumed to create market power is a settled question. Interestingly, however, there is still some debate in courts that decide private party antitrust claims. In the January 2005 case Independent Ink,(18) the Federal Circuit--which handles all direct patent appeals in the United States--held that Supreme Court precedent(19) compelled it to conclude that a patent does raise a presumption of market power in an IP tying case. But even the Federal Circuit disagreed with the presumption; in fact, the Federal Circuit's opinion invited the Supreme Court to reverse. The patentees in this case filed a petition for Supreme Court review. If the Supreme Court agrees to take the case, it would provide a good opportunity to settle the question once and for all.
Many other IP issues arise at the competition law interface. With respect to patent pools, the Antitrust Division has issued several "Business Review Letters" analyzing proposed licensing arrangements.(20) Package licensing, bundling, and tying all receive some coverage in our Guidelines. Our general approach is to avoid rigid tests and instead rely on a review of the likely economic effects to the marketplace as a whole, both in the short term and over the long term, factoring in incentives for procompetitive innovation. Both IP law and competition law seek to maintain dynamic, robustly innovative markets far into the future, and to that end they properly are willing to tolerate--or rather, offer the inducement of--a degree of private reward and market power in the present day.
We have made great strides in the United States in bringing sound economics to the antitrust analysis of intellectual property. Europe is doing the same with the newly revised Technology Transfer Block Exemption and its accompanying licensing guidelines, both of which embrace an effects-based analysis for licensing transactions.(21) We have experienced significant international convergence in this area and we have every reason to expect more of the same. While some differences remain between the U.S., the EU, and our other important trading partners, the general trend toward convergence is continuing.
1. U.S. Dep't of Justice & Fed. Trade Comm'n, Antitrust Guidelines for the Licensing of Intellectual Property (April 6, 1995), at http://www.usdoj.gov/atr/public/guidelines/ipguide.pdf.
2. 17 U.S.C. § 107.
3. 35 U.S.C. § 271(e)(1).
4. Guidelines Section 2.2.
5. See, e.g., Bement v. National Harrow Co., 186 U.S. 70, 90 (1902) ("[The patentee's] title is excleusive, and so clearly within the constitutional provisions in respect of private property that he is neither bound to use his discovery himself nor permit others to use it."); United States v. United Shoe Mach. Co., 247 U. 32, 57 (1918) (reasoning that the exercise of "the right to exclude others from the use of the invention . . . is not an offense against the Anti-Trust Act."); Hartford-Empire Co. v. United States, 323 U.S. 386, 432 (1945) ("A patent owner is not in the position of quasi-trustee for the public or under any obligation to see that the public acquires the free right to use the invention. He has no obligation either to use it or to grant its use to others."); Simpson v. Union Oil Co., 377 U.S. 13, 24 (1964) ("[t]he patent laws[,] which give a 17-year monopoly on 'making, using, or selling the invention[,]' are in pari materia with the antitrust laws and modify them pro tanto").
6. See, e.g., In re Indep. Serv. Orgs. Antitrust Litig., 203 F.3d 1322, 1325-28 (Fed. Cir. 2000), cert. denied, 531 U.S. 1143 (2001); Miller Instituform of N. Am., Inc., 830 F.2d 606, 609 (6th Cir. 1987), cert. denied, 484 U.S. 1064 (1988); SCM Corp. v. Xerox Corp., 645 F.2d 1195, 1204-07 (2d Cir. 1981), cert. denied, 455 U.S. 1016 (1982); but cf. Image Tech. Servs., Inc., v. Eastman Kodak Co., 125 F.3d 1195, 1219 (9th Cir. 1997) (permitting antitrust liability if refusal to license is "pretextual"), cert. denied, 523 U.S. 1094 (1998).
7. See Mercoid Corp. v. Mid-Continent Investment Co., 320 U.S. 661, 666 (1944) ("The fact that the patentee has the power to refuse a license does not enable him to enlarge the monopoly of the patent by the expedient of attaching conditions to its use.").
8. 540 U.S. 398 (2004).
9. Id. at 407, 414-15.
10. Id. at 408.
11. See IMS Health GmbH & Co. OHG, Case C-418/01 (April 29, 2004) at ¶¶ 34, 38, 53, at http://europa.eu.int/smartapi/cgi/sga_doc?smartapi!celexplus!prod!CELEXnumdoc&lg=en&numdoc=62001J0418.
12. Standards development organizations have identified Sony Electronics, Inc. v. Soundview Technologies, Inc., 157 F. Supp. 2d 180 (D. Conn. 2001), as a case that raises the possibility of antitrust liability for ex ante negotiations. In that decision, a district court refused to dismiss an antitrust claim based on the allegation that standards-setters made a group decision, after a standard had been adopted, to refuse to license a patent and to sue to have the patent invalidated. Although the court refused to dismiss the antitrust claim in an initial pretrial ruling, it later dismissed the claim when the patent was found to be invalid.
13. See Makan Delrahim, Forcing Firms to Share the Sandbox: Compulsory Licensing of Intellectual Property Rights and Antitrust, address before the British Institute of International and Comparative Law (May 10, 2004), at http://www.usdoj.gov/atr/public/speeches/203627.pdf.
14. Fed. Trade Comm'n, To Promote Innovation: The Proper Balance of Competition and Patent Law and Policy (2003), at http://www.ftc.gov/os/2003/10/innovationrpt.pdf.
15. National Academy of Sciences, A Patent System for the 21st Century (2004), at http://books.nap.edu/catalog/10976.html.
16. Although the terms are often used interchangeably, "enforceability" is a broader concept than "patent validity." Patents may be unenforceable against a particular alleged infringer for many reasons, including lack of validity, lack of infringement, fraud in the procurement of the patent, misuse, and other inequitable conduct.
17. See Schering-Plough Corp. v. Federal Trade Commission, 402 F.3d 1056 (11th Cir. 2005).
18. Independent Ink, Inc. v. Illinois Tool Works, Inc., 396 F.3d 1342 (Fed. Cir. 2005).
19. See, e.g., Jefferson Parish Hospital Dist. No. 2 v. Hyde, 466 U.S. 2, 16 (1984); International Salt Co. v. U.S., 332 U.S. 392 (1947).
20. Letter from Joel I. Klein, Assistant Attorney General, U.S. Dep't of Justice, to Carey R. Ramos, Esq. (June 10, 1999), at http://www.usdoj.gov/atr/public/busreview/2485.pdf; Letter from Joel I. Klein, Assistant Attorney General, U.S. Dep't of Justice, to Garrard R. Beeney, Esq. (Dec. 16, 1998), at http://www.usdoj.gov/atr/public/busreview/2121.wpd; Letter from Joel I. Klein, Acting Assistant Attorney General, U.S. Dep't of Justice, to G[a]rrard R. Beeney, Esq. (June 26, 1997), at http://www.usdoj.gov/atr/public/busreview/1170.wpd.
21. See Commission publications regarding the TTBE and guidelines at http://europa.eu.int/comm/competition/antitrust/legislation/entente3_en.html#technology.