The views expressed are those of the Commissioner and do not necessarily reflect those of the Federal Trade Commission or any other commissioner.
Good morning. I am delighted to be here today and to begin this two day program on antitrust and intellectual property claims in high technology markets. As a courtesy to my colleagues at the Commission, I will begin with my customary disclaimer: the views that I express today are my own and not necessarily those of the Commission or any other commissioner.
It has been an interesting year in antitrust enforcement that relates somehow to intellectual property. The world watched as the most talked about antitrust case of the decade proceeded through an astonishing Tunney Act review in the courts, which proved that in law, as in other aspects of life, the truth can be stranger than fiction. I am referring, of course, to the Microsoft case, in which the Court of Appeals for the District of Columbia Circuit reversed and remanded the district court's rejection under the Tunney Act of the proposed consent settlement,(1) and the settlement ultimately was approved.(2) I won't address the merits of the case. Suffice it to say that as a prosecutor, I was happy that the Court of Appeals made clear in its decision that the responsibility for deciding which conduct merits prosecution will remain with the prosecutor.
Although Microsoft was the most visible case, several other interesting cases that are equally if not more notable have been decided or settled in the past year. The Commission no doubt fueled the controversy over innovation markets by issuing several more consent orders alleging product markets based on research and development in merger cases. The Commission also broke new ground in the Dell Computer case, and in the Silicon Graphics case, it chose a theory that also may be a harbinger of things to come.
Last but not least, the last twelve months have been unusually active in terms of policy discussions. Last fall, a selection of preeminent business leaders, industry experts, economists and lawyers were invited to gather at the Commission to discuss competition and innovation issues during a series of hearings on global and innovation-based competition. Although today would have been a prime opportunity for me to offer some of my observations and reactions to the hearings, unfortunately, I missed most of the sessions. For reasons over which I have had no control, I have missed virtually all of the hearings on innovation. Since early November, I have been sitting as a juror in a very long civil trial in D.C. Superior Court. Indeed, I am still on the jury, and the judge has adjourned the case for two days to enable me to address another group yesterday(3) and to come here and be with you today. The staff are preparing a report based on the hearings, and I, along with many others, look forward to reading what they have to say.
The application of competition principles in areas characterized by high technology and innovation is increasingly important to the Commission's law enforcement program, and I will focus today on recent enforcement developments. In the last year, the Commission has brought several cases involving intellectual property. First, I will talk about two recent Commission cases, the Silicon Graphics merger case and the Dell Computer case, which has to do with standard setting. Then I plan to talk about our analysis of research and development markets in merger cases. I understand this topic was discussed extensively at the hearings on competition and innovation, and the government's approach has attracted a great deal of attention in the private bar and among the commentators. I will conclude with a few general observations about the intersection of antitrust and intellectual property and the application of competition policy arguments to issues of intellectual law and policy.
The first case I want to mention is Silicon Graphics,(4) in which the Commission challenged the acquisition by Silicon Graphics of two independent producers of entertainment graphics software, Alias Research and Wavefront Technologies. Silicon Graphics produced entertainment graphics workstations and, according to the complaint, had over 90 per cent of that market. The entertainment graphics software market also was highly concentrated, and the only competitor of Alias and Wavefront recently had been acquired by Microsoft. In short, in a market having only three firms, Silicon Graphics acquired two of them. Plainly, the effect was the same as if two firms in a three firm market had merged.
One allegation of competitive harm in the complaint was the elimination of horizontal competition between Alias and Wavefront in the entertainment graphics software market. This is an obvious theory of violation that is clear, simple and, given the history of merger enforcement under Section 7 of the Clayton Act, exactly what we would expect to see. It also is a theory for which there was an obvious, simple and effective remedy, and that was to allow Silicon Graphics to acquire only one of the two software firms, not both.
The Commission, however, did not frame the case solely as a horizontal merger but also alleged vertical foreclosure theories of competitive harm. The allegations were that the acquisition might foreclose workstation producers other than Silicon Graphics from access to entertainment graphics software and software producers other than Alias and Wavefront from the workstation market. To address these vertical foreclosure concerns, the order required Silicon Graphics to enter a porting agreement that would permit Alias software to run on the porting partner's computers. It also required Silicon Graphics to maintain an open architecture and publish its application programming interfaces to preserve access for software developers.
I dissented from the order in Silicon Graphics on the ground that the acquisitions should have been treated as a routine horizontal merger. A remedy that permitted Silicon Graphics to acquire only one of the two competing entertainment graphics software houses would have resolved the horizontal overlap as well as any vertical foreclosure concerns.
What does the Silicon Graphics case tell us about future enforcement? First, it tells us that the Commission is paying attention to the computer hardware and software industries. Second, it tells us that the Commission will challenge anticompetitive mergers in industries characterized by high technology and relatively rapid product change. I should add, however, that the rapidity of product change in a particular case may cause the Commission to forgo an enforcement action.
Third, Silicon Graphics suggests that the Commission is unusually open to new uses of and applications for vertical merger theory. We can draw this inference because Silicon Graphics was in my view obviously a horizontal case. Why the Commission chose the vertical remedy is puzzling, yet surely the case was more than an academic exercise or an experiment. Whatever the reason, the remedy has potentially significant policy implications. Like you, I await further developments with interest.
Another of the more interesting recent cases is Dell Computer Corporation, a (5)negotiated settlement that was announced in November for public comment. The complaint alleges an abuse of a private standard setting process by Dell, a patentee, and the Commission's order bars Dell from enforcing an otherwise valid patent. Before I describe the case, let me offer a preliminary observation. The complaint in the Dell case, like those in other negotiated settlements that are based on novel antitrust theories, can alert you to enforcement concerns and intentions, but they are negotiated settlements, not legal precedents. In the interests of full disclosure, I should tell you that I dissented from the Commission's decision in the Dell case, because of doubts about the theory of liability, which I will describe in a few minutes.
First, the facts of the case. Dell belonged to the Video Electronics Standards Association, known as VESA, a private standard setting organization of computer hardware and software manufacturers. In 1992, a VESA committee developed a standard for a computer bus design to carry information between the computer central processing unit and peripheral devices. The members of VESA, through their representatives, voted to approve the VESA Local Bus or "VL-bus" standard. On the VESA ballot, the representative of each member was required to sign a statement that "to the best of my knowledge" the proposed standard did not infringe the member's intellectual property rights. Dell's representative to VESA signed the requisite statement and voted to approve the standard.
After the VL-bus standard was adopted and incorporated in the design of many computers, according to the complaint, Dell claimed that "implementation of the VL-bus is a violation of Dell's exclusive rights." Dell's claim allegedly was based on a patent granted a year before the vote on the standard. The complaint does not allege that Dell intentionally misled VESA concerning its patent or the potential infringement by the VL-bus standard, nor does it allege that the Dell representative had knowledge of any patent claims or that his certification was untruthful. I am unaware of any reliable evidence that Dell's representative had any such knowledge.
If the evidence showed that Dell intentionally misled VESA to adopt a standard that was patented by Dell, thereby acquiring market power, the conduct could be examined under established monopolization theory. But there is no claim that Dell engaged in exclusionary conduct (except, of course, to the extent that asserting its lawfully acquired, valid patent could be viewed as exclusionary), and the market power element of monopolization also may be lacking. Indeed, the anticompetitive effects alleged in the complaint appear to be inconsistent with the notion that Dell possessed market power. According to the complaint, Dell's conduct hindered industry acceptance of the VESA bus standard, as firms avoided Dell's patent. To the extent that firms were able to switch to alternative bus designs, Dell apparently lacked market power.
The complaint also alleges that Dell's conduct chilled participation in private industry standard setting. Any chilling effect would seem far more likely to result from the Commission's order than from Dell's conduct, to the extent that the order creates a novel and higher standard of conduct for participants in voluntary private standard setting than previously has existed. Even assuming that Dell's conduct violated the FTC Act, a remedy more closely related to the alleged harm might bar Dell, for example, from enforcing its patent until VESA could adopt a different standard.
There may be remedies for nondisclosure of patents in the standard setting context, if the theory that a patentee should not profit from his own mistakes to the detriment of innocent infringers has any validity. If Dell attempted to enforce its patent, for example, the nondisclosure might provide the basis for a defense based on the doctrine of "unclean hands" or on patent misuse, although the courts generally have required some element of willfulness, an(6) element that is conspicuously absent from the Commission's complaint. But not all business torts violate the FTC Act. The penumbra of Section 5 can reach conduct that violates the spirit but not the letter of the Sherman Act, but it is not clear to me that the allegations in the Dell complaint establish an appropriate case in which to use that authority.
I would like to move now from the misadventures of Dell to a new issue in merger analysis, innovation markets. Last January, in remarks to this group, I talked about research and development markets, and the question of innovation markets occupied center stage in the antitrust arena during the last year. This still recent development in competition analysis has continued to confound and fascinate the commentators and the antitrust bar. Meanwhile, the number of antitrust cases in which the complaint alleges a research and development market, separate and distinct from other markets, continues to increase. The cases in which innovation markets have been alleged have involved mergers, but the same theory might apply in any antitrust rule of reason analysis.
Since the topic of research and development markets remains timely and controversial, I propose to talk about it again this year. Because I was not at the hearings last fall, I will not attempt a serious response or any response to the scholarly work that was presented there. Rather, I will offer a few practical and fundamental observations. After summarizing a few points I have made before, I will use a series of cases in the pharmaceutical industry to explain a little more about what the Commission is doing in alleging research and development markets and why the allegations make sense particularly in the drug industry context.
Let me take a moment and define my terms. In many industries, particularly high technology industries, we can identify three types of competition. O(7)ne is competition with respect to existing goods and services. I call these present products, and competition in producing and selling these products is the subject of most antitrust analysis. The second is competition with respect to products that are not yet on the market but that are in development and can be foreseen with reasonable certainty. I call this competition in future products. Most of the Commission's complaints alleging research and development markets are in this category. Finally, there is competition to develop products that have only been envisioned or perhaps not yet that. I will call this competition in pure research.
In discussing competition in future products (called research and development markets in the Commission's complaints), last year at this time I looked to three examples from the drug and medical device industries. A(8)ll three cases involved allegations of a lessening of competition in a market for a drug that was in the development pipeline but had not reached the market. Another characteristic of the cases is that they all involved competition among only two or three firms that were close to bringing a product to market and that were far ahead of other firms that might be capable of entering the market. Given the very small number of firms involved, we could be reasonably confident that the alleged markets were highly concentrated with a large increase in concentration resulting from the merger without the need to calculate the level of and increase in market share and concentration, as in analyzing markets for existing products.
The Commission's careful focus on highly concentrated markets for future products should help to allay the concern expressed by some commentators that predictions about effects on research and development cannot be inferred from market structure.(9) The competition among firms to be the first to bring to market an important future product, as I have defined the term, is important and ought to be preserved against a merger that would reduce the number of participants in the market from, for example, three to two.(10) Even for pure research, economic theory provides a range of predictions(11) and, although the relationship may not be well understood, there is support for the theory that increasing concentration can be harmful to research and development. As we gain experience in this area, it is important to remember that a sound economic foundation is important for the credibility of antitrust enforcement and that there may still be novel and difficult questions to be resolved.
Since last January, the Commission has accepted three more consent agreements in merger cases in which it has relied on allegations of a research and development market. All three were in the pharmaceutical industry.(12) It is striking that almost every Commission case alleging a research and development market has been in the pharmaceutical or medical device industry. This fact is explained partly by the luck of the draw, because there have been many mergers of drug manufacturers in recent years. Luck, however, is only a small part of the equation. In addition, certain characteristics make the pharmaceutical industry particularly likely both to entail research and development markets for specific, well-defined future products and to provide a persuasive evidentiary base in support of such markets. The large investments by drug companies in research and development and the long development time before new drugs are approved for market are factors that make antitrust analysis of future product markets in the pharmaceutical industry particularly important.
A little perspective on the drug industry may help to explain the Commission’s actions. According to data developed by the Tufts Center for the Study of Drug Development, spending on R&D by the U.S. pharmaceutical industry climbed from about $2 billion in 1963 to more than $12 billion in 1993. During the same thirty year period, the mean total development time for new drug approvals (including clinical trials and government review) climbed from six to more than 12 years.(13) The large investment in new drug research suggests both that pharmaceutical firms are competing to develop new products and that competition in these future product markets will be important to the national economy. The long lead times suggest that as firms get closer to the end of the process, the nature of the relevant markets and the extent of competition becomes more clear.
The pharmaceutical industry is unusual, if not unique, because promising new discoveries cannot be rushed to market. Instead, there are clinical trials and regulatory reviews designed to ensure the safety and efficacy of the new drug before it is allowed on the market. As a proposed new drug moves further along the path of development, the likelihood that it will make it to market increases. Only about one in 1000 drugs tested in pre-clinical studies survives to clinical studies. Of the drugs that enter clinical studies, during which the effect of a drug on humans is studied, only 20 percent ultimately are approved.(14) The drop out rate for proposed new drugs is particularly high in the early phases of clinical studies. Sixty-seven per cent of the new drugs that are eliminated from further development drop out between the start of clinical trials (typically, testing for one year on non-patient volunteers) and the beginning of phase III clinical studies (large-scale testing for three years on patients).(15)
This background provides a context for evaluating the allegations of research and development markets in the pharmaceutical industry. In the Glaxo-Wellcome case,(16) which arose out of Glaxo's $15 billion acquisition of Wellcome, the Commission's complaint alleged a market comprising "the research and development of non-injectable 5HT1D agonists," which is a class of drugs used to treat migraine headaches. The alleged competitive harm was the elimination of competition between the two firms in the relevant market and the resulting decrease in the number of R&D tracks for non-injectable 5HT1D agonists. The consent order requires Glaxo to divest Wellcome's worldwide assets relating to research and development of 311C90, Wellcome's drug in the class of 5HT1D agonists, and provides for divestiture of Glaxo's Sumatriptan assets, also a 5HT1D agonist, if the Wellcome assets are not divested. Although the complaint does not provide details, Sumatriptan and 311C90 were well along in the development and approval process at the time the merger was challenged.
The Commission relied on potential competition theory in the Hoechst case,(17) involving the $7 billion acquisition of Marion Merrill Dow. The complaint alleged, among other things, that the relevant product markets were the "research, development, manufacture and sale" of three drugs: Hoechst was developing generic products to compete with Marion Merrill Dow's Pentasa, an oral form of meslamine for treatment of ulcerative colitis, and Marion Marion Dow's Rifadin, an oral form of rifampin drug for the treatment of tuberculosis. Marion Merrill Dow was developing a drug to compete with Hoechst's Trental, used in the treatment of intermittent claudication, which, for your information, means severe leg cramps suffered by patients with arteriosclerosis.
The Commission's order in the Upjohn case arose out of the $13.9 billion merger of Upjohn and Pharmacia Aktiebolag.(18) The market identified in the complaint is "the research, development, manufacture and sale of topoisomerase I inhibitors for the treatment of colorectal cancer." Upjohn was "expected to be" the first to the market with the product, and Pharmacia planned to seek FDA approval "within the next few years." The complaint does not disclose the precise status of the two R&D projects, but Upjohn and Pharmacia allegedly were two of only a few firms in the advanced stages of developing topoisomerase I inhibitors, sales of which were anticipated to exceed $100 million by 2002. The complaint in Upjohn alleges both the elimination of horizontal competition in a research and development market, as in Glaxo, and the elimination of potential competition, as in Hoechst.
All three of these cases seem to fall within my second category, namely competition to develop new products that are not available but that can now be foreseen with a reasonable degree of certainty. The analysis in such cases is similar to merger analysis in present product markets and, as many commentators have pointed out, has much in common with potential competition analysis.(19)Competition in pure research, in contrast, poses an entirely different set of challenges, and the Commission has not ventured far into this difficult terrain. I am going to take the radical step, which I hope I will not regret later, of predicting that it is unlikely to do so any time soon.
Much of the criticism of research and development markets appears to be based on the incorrect perception that the Commission is evaluating competition in pure research among laboratories, but this is not the case. I think that I and perhaps others have been remiss in the past by accepting and using the terms "innovation markets" and "research and development markets" without qualification. This may be the source of a great deal of the confusion about and opposition to so-called innovation markets or research and development markets. By definition, we cannot predict true innovation. But that is not what our cases are about. Our casual use of the term "research and development market" actually refers to cases that allege research and development in a particular product that we have been able to define with some degree of specificity and that can be foreseen with reasonable certainty. This is what I have called the market for future products. Our "research and development" complaints always contain a definition of research and development that is tied to a very specific product market.
Analyzing a merger that involves a future product market is not very different from analyzing one that involves a market for existing products. In the pharmaceutical industry, for example, when a market is defined as a product that is approaching the end of development and regulatory review, the participants in the market are the companies that have those product-specific research & development projects. In the cases that the Commission has brought, only a small number of firms participated in the market, and other firms that were likely to have the same research capabilities lagged many years behind the market participants in the specific product markets at issue. Under these circumstances, it is easy to understand what the market is and to see that the market is highly concentrated. In contrast, it may be much more difficult even to identify who is doing pure research in a field. Some chemist in China may have just realized an incredible scientific breakthrough, the implications of which could change an entire industry in the next ten years, and we may have no way of knowing that. But we can find out which firms are working through the FDA approval process for particular drugs.
The ability to predict with some confidence which firms are way ahead of the pack in terms of developing and obtaining regulatory approval for a new drug is important to the Commission's research and development market cases. The time and effort needed to develop a product and obtain regulatory approval is the barrier or impediment to entry that distinguishes firms in the market from other possible competitors. Identifying barriers to entry into pure research markets might prove more difficult. I am not suggesting that such markets are impossible. For example, on the entry issue, it is possible that in some industries only a few firms may have the technological capability to mount a serious research effort, and technological capability may be decisive. Nonetheless, the analysis necessary to such a case would differ considerably from the research and development market cases that the Commission has brought so far.
The regulatory requirements to obtain approval to market a new drug in the United States limit the prospective participants to domestic or foreign firms that are seeking approval. Since intellectual property is freely transferrable across international borders by license, pure research by foreign firms would need to be included in such a market.
I haven't talked about the analysis of efficiencies in the context of an research and development merger. Some consolidations of R&D projects can be efficient by, for example, eliminating redundancies or combining complementary projects. A (20)concern is that firms may be insufficiently informed about overlapping R&D projects before the merger to make persuasive arguments to the enforcement agencies about efficiencies. Given the substantial lessening of competition in the R&D mergers that we have challenged, arguably any such efficiencies would be insufficient to outweigh the anticompetitive effects. This is an issue to which we should be sensitive in our analysis.
The Commission's cases alleging research and development markets have been settled, not litigated. If such a case were litigated, a legal question might be raised whether the protection of Section 7 of the Clayton Act extends to this kind of competition. Section 7 prohibits anticompetitive acquisitions "in any line of commerce." Some might say that there can be no line of commerce if the product is not yet on the market.(21) Analysis of future effects, of course, is fundamental to Section 7, but we usually mean future effects in markets for existing products. We have a theory of liability based on potential competition, but potential competition theory traditionally applies to existing lines of commerce in which some firms are actually selling a product. In an R&D market, no one is selling a product yet, and there is no guarantee that any of the participants ultimately will sell a product. There is no question as a policy matter about the significance of this kind of competition, but the question of when a line of commerce comes into existence for Section 7 purposes may need to be litigated one day.
In merger cases, the Commission alleges a violation of Section 7 and also adds Section 5 of the FTC Act as a backup count that includes the more specific Clayton Act count. If the government's use of Section 7 should ever be challenged and found wanting for research and development markets, then the Commission could challenge the same transaction under the penumbra of Section 5 of the FTC Act. Indeed, one of the more promising uses I have seen of the Commission's authority under the penumbra of Section 5 was in a case that involved a course of conduct, including an acquisition, that posed a serious threat to competition but in which a serious question could have been raised, had the case been litigated, regarding the applicability of Section 7.(22)
I would like to turn now from how we adapt our antitrust analysis to issues and industries involving intellectual property to offer a few observations about what antitrust can contribute to developments in intellectual property law and policy. The focus on the intersection of intellectual property and antitrust is at an all time high. Some recent developments and a few rumblings in the enforcement community suggest to me that we may see greater attention and even activity arising from the antitrust community on issues of intellectual property law and policy. Assuming that my prediction is informed and at least reasonably reliable, this might be a good time to think about what competition analysis can bring to the debate and what, if any, dangers this involvement might entail.
A property right granted under the intellectual property law that is overbroad or improvidently granted is likely to restrain competition more than is warranted by the policy in favor of inventiveness that underlies the intellectual property laws. This is the competition perspective on intellectual property, and it is axiomatic. It is not a recent discovery. Indeed, it is like a law of nature. Much of what antitrust can bring to the intellectual property policy debate is contained in this aphorism. Let us consider two recent examples in which the competition perspective was raised in the intellectual property arena.
Last June, the Patent and Trademark Office published proposed guides for use by its patent examiners in examining patent applications for computer-implemented inventions.(23) The staff of the Commission provided comments about the competitive implications of the PTO proposal under the Commission's advocacy program, through which the staff, when invited and when authorized by the Commission to do so, provides competition analysis on pending legislation or regulations to federal and state government bodies. In its advocacy comment to the PTO, the staff said that "inappropriate or overbroad grants of intellectual property rights may interfere with the competition that often drives innovation." This single (24)sentence was the gravamen of the advocacy filing. The point may seem obvious to those who practice antitrust law, but it is an important point and no doubt worth making from time to time. But was there anything else of use that might have been said?
The PTO notice suggested at least two different areas that might elicit comment. One is whether the law has been applied appropriately. The other is what the law should be. These are important contemporary policy issues,(25) but the first of these largely is outside our institutional expertise. We could assess the patentability of a particular product by reference to the competitive consequences of granting intellectual property protection for the product. A problem with this approach is that the statutory criteria for intellectual property rights do not include competition. This brings to mind a statement attributed to Judge Learned Hand, who, testifying before the Senate Judiciary Committee in 1957, reportedly said, "I know very little about anything, but least of all about patents."(26)
The second issue, what the law should be, is one to which the competition perspective may have much to contribute. In deciding what the law should be, the underlying question is to define the optimal scope of intellectual property rights to ensure the necessary investment to develop the intellectual property. To provide a useful contribution to this issue that is grounded in competition policy could require a significant investment of resources in a large scale, highly sophisticated empirical study. Opinions about when it would be useful for consumers to have access to certain new products are easy to come by and not unique to the Federal Trade Commission. Although I expect that the Commission's highly capable Bureau of Economics could undertake such a study, the magnitude of the undertaking may be greater than warranted.
My second example of an intellectual property question to which a competition perspective has been brought is Lotus v. Borland,(27) a high profile case recently resolved or not resolved, depending on your point of view, by the Supreme Court. Let me start with another disclaimer: I have no intention whatsoever of opining on the merits of the holding of the Court of Appeals that the menu command hierarchy on Lotus 1-2-3 is an uncopyrightable "method of operation" under the applicable statute. Those merits have been and will continue to be debated ad nauseum by people far more knowledgeable than I about both copyrights and computers. Instead, I propose to talk about the perspective that competition policy might bring to the case.
For those of you who are brand new to intellectual property law and might not know the case, the issue as I understand it was whether Borland infringed Lotus's copyright by copying the Lotus 1-2-3 menu command hierarchy in Borland's Quattro and Quattro Pro computer spreadsheet programs. The Court of Appeals for the First Circuit held that the Lotus command hierarchy was not copyrightable and therefore not infringed, and the Supreme Court affirmed on a 4-4 split. In his concurring opinion, Judge Boudin raised some concerns that seem grounded in competition policy. According to Judge Boudin, if the Lotus menu commands were copyrightable, then computer users who were familiar with Lotus and had written macros for Lotus 1-2-3 would be "captive" to the copyrighted menu, "because of an investment in learning."(28) He observed that "Lotus has already reaped a substantial reward for being first" on the market with a computer spreadsheet program,(29) and he concluded that Borland should be able to use the Lotus commands "to enable the old customers to take advantage of a new advance, and to reward Borland in turn for making a better product."
(30) Judge Boudin's observations are particularly interesting in light of concerns today about first-mover advantages, networks and switching costs. You can see the implications for competition. When the early innovator achieves a large installed base of committed users, new users are attracted to the program by those same features and the installed base grows. If compatibility is barred by intellectual property rights, entry may be difficult because consumers will not want to undertake the costs of switching.
An important underlying concern is whether the market success attained by the first innovator should make way for later entrants. Unless the first innovator has engaged in conduct that would violate the antitrust laws, such as unlawful monopolization, applying the antitrust laws to require it to grant access to late comers penalizes the successful innovator for its success. Focusing too much on the interests of the immediate consumer may result in perverse incentives. Another quotation from Judge Learned Hand is on point: "The successful competitor, having been urged to compete, must not be turned upon when he wins. . . ." (31)
This brings us back to why we have intellectual property rights in the first place. We should remember that firms compete not only to sell their existing products: another important aspect of competition is the rivalry to be first to the market. To the extent that innovators are denied the benefits of their invention, the incentive to innovate, to be first to the market, will be diminished. Recently, in a conversation with me, the general counsel of a major consumer products corporation offered a telling observation: Firms that invest significant resources in research and development favor the intellectual property laws. Firms that don't, favor antitrust. This brings me to another aphorism: Denying intellectual property protection may facilitate competition today, but denying such protection could discourage research and development that provide future products. It also is worth noting that providing access to late entrants is not necessarily a guarantee of competition in the near term, because enabling access may permit the late entrant to lessen other aspects of competition, such as price competition.
These are important contemporary policy issues to which it is useful to bring a competition perspective. As Judge Boudin pointed out in his concurring opinion, "the choices are important ones of policy," to be made by Congress not the courts, and the choices "should be made with the underlying considerations in view."(32) In this context, we would do well to return to some of the fundamental principles of antitrust where it intersects with intellectual property: The possession of intellectual property rights does not necessarily imply market power and, to the extent that market power is conferred by intellectual property rights, it does not necessarily violate the antitrust laws.(33)
Maintaining an environment in which innovation can flourish is a fundamental concern in antitrust, just as the encouragement of innovation is a fundamental concern in intellectual property law. Antitrust and intellectual property are and should be viewed as complementary policies. In our enthusiasm for antitrust and its ultimate focus on the consumer, we should take care not to undermine the protections of intellectual property law and thereby diminish the incentives to innovate. In the next few years, as we hear more about competitive solutions to intellectual property questions, I suggest you keep these principles in mind and ask some serious and exacting questions about the appropriate placement and measures of the intersection between intellectual property and antitrust.
Endnotes:
(1) United States v. Microsoft Corp., 1995-1 Trade Cas. (CCH) ¶ 71,027 (D.C. Cir. 1995).
(2) 1995-2 Trade Cas. (CCH) ¶ 71,096 (D.D.C. 1995).
(3) Azcuenaga, "Antitrust and Intellectual Property in the Year 2000," Remarks before the American Intellectual Property Law Association 19th Mid-Winter Institute, La Quinta, California, January 24, 1996.
(4) Silicon Graphics, Inc., Docket C-3626 (Nov. 14, 1995).
(5) File No. 931-0097 (press release November 2, 1995).
(6) See, e.g., SmithKline Diagnostics, Inc. v. Helena Laboratories, 859 F.2d 878, 891 (Fed. Cir. 1988); Demaco Corp. v. F. Von Langsdorff Licensing Ltd., 851 F.2d 1387, 1394 (Fed. Cir.), cert. denied, 488 U.S. 956 (1988).
(7) See Baxter, "The Definition and Measurement of Market Power in Industries Characterized by Rapidly Developing and Changing Technologies," 53 Antitrust L.J. 717 (1984) ("today's products," R&D and "tomorrow's products").
(8) Roche Holdings, Ltd. 113 F.T.C. 1086 (1990); Wright Medical Technology, Inc., File 951-0015 (Dec. 8, 1994); American Home Products Corp., File 941-0116 (Nov. 10, 1994).
(9) See, e.g., Rapp, "The Misapplication of the Innovation Market Approach to Merger Analysis," 64 Antitrust L.J. 19 (1995).
(10) The joint agency guidelines suggest that five comparable participants in the market would preserve competition. U.S. Department of Justice & Federal Trade Commission, Antitrust Guidelines for the Licensing of Intellectual Property ¶ 3.2.3 (1995). Professor Baxter suggested that four or five actors would be sufficient. Baxter, "The Definition and Measurement of Market Power in Industries Characterized by Rapidly Developing and Changing Technologies," 53 Antitrust L.J. 717, 724 (1984).
(11) See Gilbert & Sunshine, "The Use of Innovation Markets: A Reply to Hay, Rapp, and Hoerner," 64 Antitrust L.J. 75, 76-78 (1995); Gilbert & Sunshine, "Incorporating Dynamic Efficiency Concerns in Merger Analysis: The Use of Innovation Markets," 63 Antitrust L.J. 569, 574-81 (1994).
(12) Glaxo plc, File 951-0054 (March 16, 1995); Hoechst AG, Docket C-3629 (Dec. 5, 1995); The Upjohn Co., File 951-0140 (Oct. 27, 1995).
(13) Kaithin & Houben, "The Process of New Drug Development: Worthwhile Persistence," reprinted from Odyssey, Glaxo Wellcome Journal of Innovation in Healthcare, Vol. 1, issue 3 (June 1995).
(14) Id.
(15) Id.
(16) Glaxo plc, File 951-0054 (March 16, 1995).
(17) Hoechst AG, Docket C-3629 (Dec. 5, 1995).
(18) The Upjohn Co., File 951-0140 (Oct. 27, 1995).
(19) See Azcuenaga, "Mergers: A View from the Federal Trade Commission," Remarks before the Practising Law Institute 25th Annual Advanced Antitrust Seminar, Coral Gables, Florida, March 15, 1995.
(20) See Rapp, "The Misapplication of the Innovation Market Approach to Merger Analysis," 64 Antitrust L.J. 19, 34 (1995).
(21) Cf. SCM Corp. v. Xerox Corp., 645 F.2d 1195, 1210-11 (2d Cir. 1981), cert. denied, 455 U.S. 1016 (1982) (finding acquisition of patents not unlawful when relevant product not introduced until 4 years later and relevant market did not exist until 8 to 13 years after acquisition).
(22) The Vons Companies, Inc., Docket C-3391 (Aug. 7, 1992), Commissioner Azcuenaga concurring on Section 5 but not Section 7; see Azcuenaga, "Shimmers in the Penumbra of Section 5 and Other News," Remarks before the National Economic Research Associates, Inc., 13th Annual Antitrust Trade Regulation Seminar, Santa Fe, New Mexico, July 9, 1992.
(23) Request for Comments on Proposed Examination Guidelines for Computer-Implemented Inventions, 60 Fed. Reg. 28,778 (June 2, 1995).
(24) Comment of the Staff of the Federal Trade Commission on Proposed PTO Examination Guidelines for Computer-Implemented Inventions (Sept. 26, 1995).
(25) See, e.g., "Symposium: Toward a Third Intellectual Property Paradigm," 94 Colum. L. Rev. 2307 (1994) (examining legal protection appropriate for computer software when technology enables easy imitation).
(26) See Panel Discussion, 53 Antitrust L.J. 523, 543 (1984).
(27) Lotus Development Corp. v. Borland International, Inc., 49 F.3d 807 (1st Cir. 1995), aff'd by an equally divided Court, (U.S. January 16, 1996).
(28) 49 F.3d at 821.
(29) Id.
(30) Id.
(31) United States v. Aluminum Co. of America, 145 F.2d 416 (2d Cir. 1945).
(32) 49 F.3d at 822.
(33) FTC and Department of Justice Antitrust Guidelines for the Licensing of Intellectual Property ¶ 2.2 (April 6, 1995).