Antitrust and Intellectual Property: Recent Highlights and Uncertainties

The American Law Institute-American Bar Association, "Antitrust/Intellectual Property Claims in High Technology Markets", Ritz-Carlton Hotel

Boston, Massachusetts

Date:
By: 
Mary L. Azcuenaga, Former Commissioner

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 with you this morning and to begin this discussion of antitrust and intellectual property claims in high technology markets. I plan to talk about developments at the FTC from my perspective as a commissioner and an antitrust lawyer. As a courtesy to my colleagues on 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 for those who follow developments in antitrust and intellectual property law. From the perspective of the federal antitrust enforcement agencies, 1996 was more subdued than 1995, which was a blockbuster year, at least in terms of headlines. In 1995, the Federal Trade Commission and the Department of Justice released the Antitrust Guidelines for the Licensing of Intellectual Property ("Intellectual Property Guidelines") and the Court of Appeals finally resolved the long-running Microsoft saga.(1) Both those events now seem long ago.

In Microsoft, the challenge for the Department of Justice was to craft a remedy that eliminated the potentially exclusionary effect of per processor licenses without also inhibiting volume discounts, or, looked at another way, to develop a remedy that allowed volume discounts but was more than a cosmetic solution that would leave the company free to accomplish the same result in another manner. Some of the commenters, both named and anonymous, and Judge Sporkin found the result unacceptable, but, as you know, ultimately the settlement was approved.

The primary legacy of the Microsoft case is procedural. First, the opinion of the Court of Appeals provided clarification regarding the appropriate role of the district court in a Tunney Act proceeding. Second, the case demonstrated that even if one of the federal antitrust agencies decides not to challenge a particular course of conduct, a company may be subject to a second investigation and challenge by the other federal antitrust agency.

As most of you know, the Commission investigated the Microsoft matter first and decided, on a 2-2 vote, not to pursue the case. Whether the Commission's decision not to sue Microsoft was right or wrong, the tie vote was a decision, not an inability to reach a decision. As in other deliberative bodies, when a matter is presented and voted on, as the Microsoft case was, a tie vote at the Commission is considered a formal decision, not a failure to reach a decision. Like other procedures, this one is best understood and respected in an outcome-neutral context.

One may well question whether as a matter of policy, efficient use of government resources or fundamental fairness both agencies should investigate the same conduct, but in the meantime, it is a risk you should keep in mind. Having said that, I also should note that I think the risk of that occurring again is extremely small.

The major event in 1995 that is of continuing significance was the issuance of the Intellectual Property Guidelines.(2) The guidelines reaffirmed some important fundamental principles and set forth a framework intended "to assist those who need to predict whether the [Department or the Commission] will challenge a practice as anticompetitive."(3) In my view, the guidelines adopt the right principles and achieve the right balance with respect to competition policy and the policy underlying intellectual property rights, and they are quite successful in accomplishing what they were intended to accomplish in terms of providing analytical predictability.

Despite the guidelines, the subtitle of this program is "Litigating and Advising in an Era of Uncertainty." This provides my segue to developments subsequent to issuance of the guidelines. More recent developments, although perhaps not as showy as some of their immediate predecessors, have been intriguing and important. To some extent, they illustrate or even introduce areas of uncertainty. In the past, I have defended our ability to predict outcomes in debating the degree of uncertainty in this area of the law with Ron Katz, one of the co-chairs of this program. I still believe that we can predict how the agencies will treat most antitrust/intellectual property cases, but reluctantly and in view of certain recent developments, I have come closer to his view and have to acknowledge that there is a little more uncertainty in the area than I would like to see and more than the guidelines, standing alone, would suggest. With certainties and uncertainties in mind, today I plan to discuss a number of developments bearing on the nexus of antitrust and intellectual property law. First, I will mention some general developments in antitrust analysis that apply to cases involving intellectual property as well as other cases. Then I will discuss an important development in the area of standards-setting and antitrust, offer a few observations about remedies and, finally, so-called innovation markets, including the most recent example of an innovation market case, the Ciba/Sandozcase. One topic I will not cover today is the extent to which antitrust involves judgments about the scope and validity of intellectual property and the extent to which antitrust provides a basis for changing the scope of those rights. Nevertheless, I think it is a topic worth mentioning because there are a number of nascent signs of change in this area and, depending on what these signs presage, it could become one of the most, if not the most, important topics of all. But that is the subject of a speech for another day.

The first set of developments I would like to mention relates to the methods of antitrust analysis employed by the federal enforcement agencies. These are developments in general modes of antitrust analysis and apply with equal force to any restraint of competition including those that involve intellectual property issues. In March 1996, the Commission abandoned its so-called "Mass. Board" analysis, named for the 1988 case in which it was first announced.(4) In the decade before the Commission decided the Mass. Boardcase, the Supreme Court, in the NCAA(5) and BMI(6) cases, had opened the way to an antitrust analysis that focuses more on competitive effects and efficiencies rather than categorization into per se or rule of reason analysis. The Commission's Mass. Boardopinion was a considered attempt to further that trend. Under Mass. Board, the Commission's analysis proceeded by asking a series of questions rather than focusing on whether conduct is per se unlawful or subject to a rule of reason analysis.

In the California Dental Association case, the Commission overruled Mass. Board.(7) The Commission did not specifically say that it overruled Mass. Board, but there can be no other reasonable reading of the California Dental Association opinion. If there was any lingering doubt about the new position, in February 1997, the Commission reaffirmed its abandonment of the Mass. Board approach in the International Association of Conference Interpreters case.(8)

What does all this mean? I think it is a little too early to say. Mass. Board analysis emphasized consideration of efficiencies. In overrulingMass. Board, I doubt that the Commission intended to de-emphasize efficiencies, but only time will tell. I do anticipate that the Commission, having returned to per se or rule of reason categorization, will look for opportunities to abbreviate the analysis in rule of reason cases. I predict that, among other things, we will see debate about how much proof of market power is necessary to establish liability in a rule of reason case. In the California Dental Association case, the Commission majority relied on a surprisingly truncated analysis as a basis for liability, so truncated in fact that if that were the last word on the subject, there might be some reason for alarm.

But in the International Association of Conference Interpreters decision last February, the Commission seemed to return to a somewhat more traditional and rigorous standard for establishing liability under the rule of reason. Which of the two approaches to rule of reason analysis ultimately will prevail, California Dental or Conference Interpreters, is uncertain. So that is the first uncertainty I will reveal. I hope not to provide many more. My own views in this area are not so uncertain, but I must confess to some uncertainty about where the Commission is going analytically.

Ironically, at just about the same time that the Commission was abandoning Mass. Board, the Acting Assistant Attorney General for the Antitrust Division announced that the Department of Justice "reject[ed] the notion that there should only be two methods of analysis -- per se or full-blown market analysis."(9) In a speech last November, Acting Assistant Attorney General Klein described a step-wise analytical approach that sounds very similar to the old FTC Mass. Board analysis.

A second development in general antitrust analysis is the announcement earlier this month of new Section 4 of the Horizontal Merger Guidelines regarding efficiencies. In 1982, when the Department of Justice issued the progenitor of the current guidelines, the Department announced that efficiencies would not be considered as "a mitigating factor" in deciding whether to challenge an anticompetitive merger. In 1982, the Commission took the opposite position and recognized efficiencies as part of the balance in deciding whether to challenge a merger.(10) Subsequently, the 1984 and 1992 versions of the Guidelines, although somewhat cryptic, seemed to adopt the FTC's 1982 approach to the issue.

A number of people have observed that the 1997 revision is an improvement over the 1992 version. Although one view is that the 1997 changes are more of style than substance, I can see several potential advantages that may result from the publication of the new guidelines. First, the 1997 version may help clarify some of the questions the agencies ask about efficiencies. Second, new terminology and other changes likely will provoke debate that may further illuminate the subject. Third, particularly in light of the overruling of Mass. Board, issuance of new Section 4 is a welcome reaffirmation by the antitrust agencies of the importance of considering efficiencies.

I should mention one other general development. Within the next day or so, the Commission will publish in the Federal Register a notice of an opportunity to comment and appear at public hearings relating to joint ventures. The project involves a reexamination of the antitrust rules relating to joint ventures and other means of collaboration. For those of you who are interested in or concerned about the possibility of federal antitrust guidelines on joint ventures, this is your opportunity to be heard. The hearings will be in early June but written comments will be accepted until the first of August.

Now I would like to turn to an important development in the analysis of standard setting under the antitrust laws, the final resolution in the Dell Computer Corp. case.(11) I discussed the Dell case at this program last January in San Francisco.(12) The final Commission order in Dell was issued in May 1996. Along with the final order, the Commission majority issued a statement that if not designed to create uncertainty in this area of the law certainly did a good job of creating that impression. For purposes of full disclosure, I should mention that the statement of the majority was written in response to my dissenting statement.

The facts of the case are simple. After the Video Electronics Standard Association (VESA) issued a standard for the VL-bus, Dell asserted a patent against some firms that were implementing the VL-bus standard. When the standard was under consideration, Dell's voting representative to VESA, an engineer, had signed two ballot certifications that "to the best of my knowledge," the proposed standard did not infringe any firm's intellectual property rights. The complaint against Dell was heralded as "precedent-setting" principally because the charge against Dell did not involve any allegation of intentional misconduct. If Dell had intentionally deceived VESA into issuing a standard covered by a Dell patent, liability would have been clear. It was the absence of an allegation of intent that made the consent agreement interesting.

Although the Commission infrequently receives lengthy, substantive comments during the 60-day comment period for consent agreements, eleven thoughtful comments were filed in Dell. Several commenters, including ANSI and the American Intellectual Property Law Association, expressed concern about the possibility that the decision would chill participation in the standards-setting process because the consent agreement appeared to impose an affirmative duty to disclose on participants in standards-setting. That is, by participating in a standard-setting, a firm assumes the duty to identify and disclose any relevant intellectual property. Other commenters, including VESA and the American Committee for Interoperable Standards, supported the idea that there should be a general antitrust duty on any participant in the standards-setting process to identify and disclose its relevant intellectual property rights to the standards-setting organization.

After considering the comments, the Commission majority issued the final complaint and order without significant change. Read within the four corners of the documents, the complaint and order suggest that the Commission agreed with VESA and the American Committee for Interoperable Standards that by participating in standards-setting, a firm assumes the duty to identify and disclose all relevant intellectual property. The Commission majority, however, took the unusual step of issuing a statement purporting to explain its decision in which it explicitly disclaimed the imposition of a general duty to search. The majority said that the order should not be read to mean that "inadvertence in the standard-setting process provides a basis for enforcement action," and that "there is reason to believe that Dell's failure to disclose the patent was not inadvertent." The majority statement raises two problems. First, Dell signed a consent agreement after reviewing a complaint and order that did not allege intentional (or "not inadvertent") misconduct. The majority stated that it found reason to believe that Dell engaged in "not inadvertent" conduct, which I would normally take to mean "advertent" conduct. Since "reason to believe" is the standard for issuing a complaint, the allegation that the conduct was "not inadvertent" could have been, but was not, included in the complaint. Dell's counsel presumably negotiated to avoid any allegation of intentional misconduct in the complaint, but he could not control what the majority said in a separate, explanatory statement. The case should sound a note of caution for counsel negotiating consent agreements with the Commission.

Second, the statement of the majority creates complete uncertainty regarding the existence or extent of a duty on the part of participants in the standards process to search for and disclose relevant intellectual property. Read in isolation, the complaint and order suggest that an absolute duty to search and disclose arises from participation in the process regardless of intent, but the majority statement suggests that it arises only for "not inadvertent" conduct, whatever that means. Reasonable people can debate the policy question of the nature and scope of the duty of a participant in the standards-setting process. Rather than deciding the question, the Dell case introduced a new element of uncertainty into this area of the law.

Next, I would like to talk for a moment about remedies. Increasingly, the Commission has used licensing of intellectual property as a remedy in merger cases. In the past, when the Commission challenged a horizontal merger, the typical remedy was either to block the merger or to require divestiture of one of the competing businesses. This approach worked well when the divested entity could be operated as a separate business unit or when a package of tangible assets, such as a plant, could form the core of a new competitor. In the 1980's, however, we began to encounter industries in which the core assets were not physical property, but instead were intellectual property.

The traditional divestiture of a business unit may not work as well in industries where the principal assets are intangibles. In the pharmaceutical industry, for example, the Commission has frequently relied on intellectual property licensing as a merger remedy.(13)With this increased reliance on licensing, we are facing the problems that many of you deal with on a daily basis, that is, the need to draft tight, enforceable licensing agreements, and we have encountered drafting problems. For example, several Commission orders have required the divesting firm to grant a right of reference to certain data on file with the Food and Drug Administration. As an alternative to conducting new clinical studies, an applicant for approval of a new drug may sometimes obtain permission to refer to safety and efficacy data on file with the FDA for a similar drug. A question has arisen regarding the scope of a right of reference in one or our consent orders. Whether we prevail or not in our reading of the order, no doubt it is a point with respect to which we will be even more attentive in the future.

We have encountered other problems in implementing remedies other than traditional divestitures. In Silicon Graphics, Inc.,(14) the Commission accepted a settlement of Section 7 charges arising from Silicon Graphic's acquisition of two of the three leading developers of entertainment graphics software (Alias and Wavefront). On the merits, I thought we should have resolved the horizontal problem by requiring divestiture of one of the two software developers, and the vertical problems would have taken care of themselves. But I do not mention it for that reason. The Commission took another approach. The order required Silicon Graphics/Alias, no later than March 31, 1996, to enter a "porting agreement" with another computer hardware manufacturer. Under the porting agreement, Silicon Graphics/Alias would agree to keep its software compatible with the hardware produced by another manufacturer. The porting agreement was intended to resolve vertical concerns that Silicon Graphics' ownership of both software developers might foreclose other hardware manufacturers from this market.

On March 28, 1997, almost a full year after the application to approve the porting agreement was filed, the Commission finally approved a porting agreement between Silicon Graphics/Alias and IBM. As indicated in the Commission letter approving the agreement, Silicon Graphics/Alias submitted "commitments to market the ported software products and to continue to support the ported products until March 12, 1999." It seems fair to ask whether the Commission's endorsement of a continuing entanglement between IBM and Silicon Graphics, which are horizontal competitors in hardware, is better or worse for competition than the vertical concerns being remedied.

As we rely increasingly on licensing as a remedy, we may be drawn into questions regarding the scope and validity of the intellectual property in question. In American Home Products Corp.,(15), the Commission accepted for public comment an order arising from American Home Products' acquisition of the animal health business of Solvay, S.A. and requiring divestiture of three veterinary vaccines. Since Solvay and American Home Products were important suppliers of the vaccines in highly concentrated markets, the need for an order seemed apparent. The order not only requires divestiture of Solvay's canine corona virus vaccine, which was newly developed to compete with American Home Products canine corona virus vaccine, it also requires American Home Products to agree not to sue the acquirer for patent infringement. This latter provision means that the order does more than return matters to the status quo before the merger because it implicitly decides the patent dispute.

Although the Commission effectively resolved the patent dispute in Solvay's favor, it did not take evidence or hear argument directly on the patent issue. As a practical matter, resolving the patent dispute may have been necessary to induce a company to purchase the Solvay vaccine assets. I do not know. In theory, a purchaser stepping into Solvay's shoes could defend a patent suit, but finding such a purchaser might have been difficult. I voted in favor of the order, but it seems to me that whether and how such disputes regarding patent infringement should be resolved is a question for further study. The Commission may be able to impose this kind of remedy and force the company to pay the price as part of the cost of the transaction. In crafting this kind of remedy, it should go without saying that an agency charged with protecting competition should not tax a transaction beyond what is necessary to accomplish that protection.

My next topic is the so-called innovation market, its origins and its critics. It seems to me that the term "innovation market" has spawned a great deal of unnecessary confusion. In drafting Section 7 complaints, the Federal Trade Commission does not even use the term. Instead, we have alleged research and development markets with respect to specifically identified future products. In the earliest cases, in addition to alleging this kind of market, the Commission's complaints also included allegations regarding elimination of "potential competition" and of a "potential entrant."(16) In the Roche Holdings case, the Commission alleged a market of CD4-based therapeutics for use in the treatment of AIDS/HIV and alleged that Roche and Genentech were among the limited number of firms pursuing research on this drug.(17) The notice pleading of the complaint did not signal the adoption of a novel theory, and in fact, Commissioner Owen dissented on the ground that the elements of a potential competition case had not been satisfied.(18) Other early cases also had the flavor of the potential competition doctrine. For example, the Wright Medical Technology case involved the acquisition of a firm engaged in the research and development of an improved orthopaedic implant by the firm producing the current generation of implants. The Commission required licensing of the intellectual property related to the implant under development.(19) Although the new orthopaedic implant was a long way from market and needed to go through the FDA approval process, the complaint alleged that the developer was a potential entrant into the implant market and was a competitor in a research and development market. Put another way, the complaint alleged both the traditional potential competition theory and the novel research and development market. At about the same time, the complaint in the American Home Products case, which alleged as a market "the research and development of a vaccine against Rotavirus infection in humans," did not include potential competition allegations.(20) Clearly this was a transitional period and the Commission was still feeling its way along.

The adoption of the Intellectual Property Guidelines in April 1995, brought increased legitimacy to the concept of the innovation market. The Guidelines defined an innovation market as the "research and development directed to particular new or improved goods or processes and the close substitutes for that research and development." The Guidelines set forth the caveat that such markets can be defined only when the capability to do the research is "associated with the specialized assets or characteristics of specific firms."

Although the Intellectual Property Guidelines endorse antitrust analysis of innovation markets, they provide little insight into how the analysis is to be done, particularly in comparison to the method of analysis set forth for product markets in the Merger Guidelines. TheIntellectual Property Guidelines say that the agencies will consider "all relevant evidence," but the only concrete example of what evidence is relevant is the statement that if available, the agencies "will include market share data in this assessment." This implies a relationship between concentration and innovation, but provides little guidance on how to measure a share of an innovation market. The guidelines say that market shares can be based on the share of assets or characteristics needed for innovation, on the share of expenditures, or shares of a related product.

That the innovation market theory should attract critics came as no surprise, but the nature and breadth of the criticism is impressive. A fundamental criticism is that we have no theoretical or empirical showing for the proposition that increased concentration yields decreased innovation. Richard Rapp, who is one of the speakers at this program, points out that innovation is not bought and sold in the marketplace and that we cannot measure innovation. What we can measure is research and development spending, which is an input in the process of innovation, but that is not necessarily a reliable predictor of the amount of innovation that the spending will yield. Citing work by F.M. Scherer and others, Dr. Rapp says that firm size and market concentration do not appear to be determinative of innovation.(21) As he points out, Jorde and Teece have argued that successful innovation requires cooperation among firms.(22) Cooperation can help spread the risk of failure, provide access to complementary technologies and permit economies of scale and scope. This favorable view of collective research seems inconsistent with the premise of the innovation market theory that increased concentration leads to diminished innovation. Other informed economists have joined the criticism.(23) One experienced lawyer pointed out that as a practical matter, it may be impossible to make a reliable assessment of the concentration in an innovation market.(24) He pointed out that researchers in pursuit of scarce funding for their project may tend toward the optimistic and that third-party competitors, not involved in a particular merger, have incentives to downplay their own research efforts.

Another objection to the innovation market theory is that it is very difficult to monopolize innovation. The Intellectual Property GuidelinesSections 3.2.2 and 3.2.3 distinguish between a technology market and an innovation market. Although one can patent a technology, the components in a research and development program, such as laboratories, computers, scientists, engineers, and so forth, can be hired away or replicated.(25) A further problem is that it is very difficult to predict in advance where an innovation will come from. As one witness at the FTC Global Hearings observed, promising leads often go nowhere, and innovation sometimes comes from the unlikeliest places.(26)

The critiques of the innovation market theory raise serious questions regarding how far it should be pursued, at least given our current knowledge. Nonetheless, the valid criticisms of the theory seem to apply to its application in a broad sense to the concept of innovation. They do not seem to undercut our antitrust concerns for future competition in a specific product that is already under development. Almost all the FTC cases have involved research and development by a very few firms of a pharmaceutical product to remedy a particular disease or condition. The Commission has focused on future competition to manufacture and sell the particular drug in question and not the general level of research or innovation in the pharmaceutical industry.

The regulatory framework in the United States that surrounds the development and review of a new drug resolves some of the theoretical criticisms of the innovation market theory. Given the lengthy clinical testing and review phases, an innovation almost certainly will not come out of left field and suddenly render obsolete a research project that is currently in the midst of clinical trials.(27) Once a firm commences clinical trials on a pharmaceutical product, the product is identifiable and fairly well developed. It is also possible to identify competitive products that are as or more developed. At this point, purchasing the competitive product does not consist entirely of purchasing the process of innovation but also, in part, the product of innovation, just as one might purchase an innovative, competing product that is already on the market.

Research and development markets are still relatively new. Overall, I am satisfied that the Commission has applied this theory responsibly and to the benefit of consumers in particular markets. We may have had a few missteps but so far, I think, remarkably few. The early Sensormatic case raised issues about the specific delineation of a research and development product market and the relevant geographic market.(28) I noted those issues at the time and they seem to have come into greater relief in light of subsequent experience and learned critiques. By the way, the Sensormatic case was one that did not involve a pharmaceutical product. Our most recent case in this area, the Ciba/Geigy Sandoz case, also raises some questions about the Commission's use of research and development markets. I will return to that case in a moment. I anticipate that the Commission will continue to apply the law in research and development markets, but it is my hope that to the extent the Commission reaches for new applications in this area, it will do so informed by the criticisms that have been offered.

Finally, is there anything we can learn from the Commission's recent order in the Ciba-Sandoz matter?(29) In March 1997, the Commission issued a consent order, settling charges that the $63 billion merger of Ciba-Geigy Limited and Sandoz Ltd. to form Novartis would violate Section 7 of the Clayton Act. It was an enormous transaction. Since the companies, which together have assets valued at $80 billion, have numerous products, it is not surprising that the Commission identified horizontal competition issues in several markets, including herbicides used by corn growers and animal flea control products. The challenging antitrust/intellectual property issues arose from the combination of two U.S. firms, each affiliated with one of the merging Swiss firms and both engaged in research and development of gene therapy products.

The complaint alleged four research and development product markets relating to gene therapy for specific medical conditions,(30)similar to other recent Commission orders involving research and development markets. The antitrust concern was that in each of four markets the merger combined two firms with competing products in the FDA regulatory pipeline, and the remedy was to require licensing of the intellectual property necessary to continue the research.

One of the product markets alleged in the complaint is "gene therapy technology and research and development of gene therapies, including ex vivo and in vivo gene therapy." The breadth of this research and development market and the fact that it does not apply to a particular product make it somewhat novel. The remedy for the broad market is also somewhat novel; it is to require licensing of certain patents to all comers at what the majority characterize in the Analysis to Aid Public Comment as a "low royalty." One such patent is the Anderson patent, issued to the NIH and licensed by NIH to Sandoz. The Anderson patent broadly covers the technique of ex vivo gene therapy. Since Ciba-Geigy did not have an interest in a substitute gene therapy technology, the requirement to license the Anderson patent does not appear necessary to cure an anticompetitive overlap under Section 7 of the Clayton Act. Not only is there no overlapping drug product in the development pipeline, there is no overlapping technology.

It is unclear what links the compulsory licensing requirement to the violation alleged in the complaint. Under the circumstances, the compulsory licensing requirement can only contribute to uncertainty about the Commission's enforcement intentions. One possibility is that the requirement to license to all comers at low rates reflects a judgment that the patent is too broad. The complaint does not specifically allege that having an unduly broad patent constitutes a violation of Section 7 of the Clayton Act, and it is unlikely that that theory would prevail in court. Nevertheless, the language of the complaint and the remedy suggest that the breadth of the patent may have been a concern. Of course, it would be unusual for one government agency to attack a patent issued by the Patent and Trademark Office to another federal agency. Another possibility is that the Commission has decided to extend the innovation market theory quite beyond what we have seen before. Perhaps the Commission identified a narrow and specific competitive concern related to the Anderson patent. This alternative is a possibility because the complaint alleges increased barriers to entry and altered incentives to license patents, but the precise concern is not well described in the complaint.

The case raises some important questions. For example, how far is the Commission willing to pursue the concept of an innovation market? How far removed from an actual product may an alleged product market be? To what extent is the Commission willing to limit existing intellectual property rights to encourage follow-on invention? To impose that kind of remedy, how far is the Commission willing to depart from traditional Section 7 law? Is the Commission an appropriate agency to evaluate the scope of intellectual property rights? Several questions come to mind but that, too, is the basis for another speech.

Having concluded my tour of recent highlights at the FTC, let me close by returning to the question of the predictability of government enforcement action in antitrust/intellectual property cases. Although the 1995 Intellectual Property Guidelines are a serious and, I think, generally successful effort to reveal the agencies' intentions, not every question can be answered in a set of guidelines. Enough uncertainties remain to keep us interested for some time to come. I hope we will be up to the challenge.

1. United States v. Microsoft Corp., 56 F.3d 1448 (D.C. Cir. 1995).

2. Antitrust Guidelines for the Licensing of Intellectual Property, 4 CCH. Trade Reg. Rep. 13,132 (April 6, 1995) ("Intellectual Property Guidelines").

3. Intellectual Property Guidelines, Section 1.0.

4. Massachusetts Board of Registration in Optometry, 110 F.T.C. 549 (1988).

5. NCAA v. Board of Regents of the University of Oklahoma, 468 U.S. 85 (1984).

6. Broadcast Music, Inc. v. CBS, 441 U.S. 1 (1979).

7. California Dental Association, Docket No. 9259 (March 25, 1996) (Commissioners Azcuenaga dissenting and Commissioner Starek concurring in part and dissenting in part), appeal pending, overruling Massachusetts Board of Registration in Optometry, 110 F.T.C. 549 (1988).

8. International Association of Conference Interpreters, Docket No. 9270 (Commissioner Starek concurring in part and dissenting in part)(February 19, 1997).

9. Joel Klein, "A Stepwise Approach to Antitrust Review of Horizontal Agreements," Speech Before the ABA Antitrust Section Semi-Annual Fall Policy Program, November 7, 1996, at 5.

10. Statement of Federal Trade Commission Concerning Horizontal Mergers (June 14, 1982).

11. Dell Computer Corp., Docket No. C-3658 (May 20, 1996) (Commissioner Azcuenaga dissenting).

12. Mary Azcuenaga, "The Intersection of Antitrust and Intellectual Property: Adaptations, Aphorisms and Advancing the Debate," Speech Before the ALI-ABA Program on "Antitrust/Intellectual Property Claims in High Technology Markets," January 25, 1996.

13. Ciba-Geigy Ltd., Docket No. C-3725 (March 4, 1997) (Commissioner Azcuenaga concurring in part and dissenting in part); American Home Products Corp., File No. 971-0009 (Feb. 25, 1997) (accepted for public comment) (Commissioner Azcuenaga concurring); Baxter International Inc., Docket No. C-3726 (March 24, 1997); Glaxo plc, Docket No. C-3586 (June 14, 1995); Hoechst AG, Docket No. C-3629 (Dec. 5, 1995); The Upjohn Co., Docket No. C-3638 (Feb. 8, 1996); Sensormatic Electronics Corp., Docket No. C-3572 (April 18, 1995) (Commissioner Azcuenaga concurring in part and dissenting in part); Wright Medical Technology, Inc., Docket No. C-3564 (March 23, 1995); American Home Products Corp., Docket No. C-3557 (Feb. 14, 1995 ) (Commissioner Azcuenaga concurring); Roche Holdings Ltd., 113 F.T.C. 1086 (1990).

14. Docket No. C-3626 (November 14, 1995) (Commissioners Azcuenaga and Starek dissenting).

15. File No. 971-0009 (Feb. 25, 1997) (Commissioner Azcuenaga concurring).

16. For an early discussion of innovation markets, see Mary Azcuenaga, "Intellectual Property and Antitrust: A Perspective From the FTC," Speech before the ALI-ABA Program on "Antitrust/Intellectual Property Claims in High Technology Markets," January 26, 1995.

17. Roche Holding Ltd., 113 F.T.C. 1086 (1990).

18. Commissioner Owen's dissent relied on the requirement in B.A.T. Industries, Ltd., 104 F.T.C. 916 (1984), that there be "clear proof" that entry would occur in the "near future" to highlight the difference between the allegations in the complaint and the potential competition doctrine.

19. Wright Medical Technology, Inc., Dkt. No. C-3564 (March 23, 1995).

20. American Home Products Corp., Docket No. C-3557 (February 14, 1995)(Commissioner Azcuenaga concurring).

21. Rapp, "The Misapplication of the Innovation Market Approach to Merger Analysis," 64 Antitrust L. J. 19 (1995), citing, F.M. Scherer, "Schumpeter and Plausible Capitalism," 30 J. Econ. Lit. 1416 (1992).

22. Jorde and Teece, "Innovation, Cooperation and Antitrust," 4 High Technology L.J. 1 (1989).

23. G. Hay, "Innovations in Antitrust Enforcement," 64 Antitrust L.J. 7, 16 (1995); Testimony by Professor Dennis Carlton, University of Chicago, FTC Hearings on Global and Innovation-Based Competition at 931 (Oct. 25, 1995) ("In summary, neither theory nor empirical work provides any general justification for an antitrust policy aimed at preserving competition in innovation markets.").

24. Testimony by Michael Sohn, FTC Global Hearings, 993-994 (Oct.25, 1995).

25. Rapp, "The Misapplication of the Innovation Market Approach to Merger Analysis," 64 Antitrust L.J. 19, 36 (1995).

26. Sumanth Addanki, FTC Global Hearings Transcript 944 (October 25, 1995).

27. One unusual characteristic of the pharmaceutical industry is that the research and development of a new drug is conducted under the auspices of the Food and Drug Administration. Before any human testing of a new drug can be done, a firm must file an investigational new drug application. The testing of a new pharmaceutical product is done in discrete phases, followed by a New Drug application filed with the FDA, all of which takes some five to seven years. Kaitin and Houben, "The Process of New Drug Development: Worthwhile Persistence," 1 Odyssey 3-4 (June 1995).

28. Sensormatic Electronics Corp., Docket No. C-3572 (April 18, 1995)(Commissioner Azcuenaga concurring in part and dissenting in part).

29. Ciba-Geigy Ltd., Docket No. C-3725 (March 24, 1997) (Commissioner Azcuenaga concurring in part and dissenting in part).

30. These include herpes simplex virus-thymidine kinase ("HSV-tk") therapy for treatment of cancer and separately for treatment of graft versus host disease, gene therapy for hemophilia, and chemoresistance gene therapy.