Justice News

THE LONG AND WINDING ROAD:
CONVERGENCE IN THE APPLICATION
OF ANTITRUST TO INTELLECTUAL PROPERTY

Remarks by

MAKAN DELRAHIM
Deputy Assistant Attorney General
Antitrust Division
U.S. Department of Justice

Presented at

The George Mason Law Review Symposium

October 6, 2004

I. Introduction

Good afternoon. It is a pleasure to be here on this beautiful day, and on this distinguished panel, to discuss the path toward convergence between the US and the EU in the application of antitrust law to intellectual property. The title of my speech today refers to the Beatles' song­"The Long and Winding Road"­because I believe the path toward convergence in this area is such a road, and both the US and the EU have made tremendous progress along this road in recent years. I also thought it would be fitting to name this speech in a way that recognizes the tremendous contribution that intellectual property makes to our lives every day, not just in the sense of amazing technological innovations that are protected by patents and the laws regarding trade secrets, but also the creations of artists­like the Beatles­who create poetry, literature, music and movies­the things that feed the soul. Speeches about the application of antitrust to intellectual property are often dry because they tend to focus on patents and software, but it is important to remember that we are really talking about encouraging innovation and protecting creativity in all its dimensions. "The Long and Winding Road" is a perfect example of that creativity.

When I talk about "convergence" in the application of antitrust to intellectual property, I am referring to the goal of reaching consensus on antitrust enforcement strategies that are grounded in sound economic theory, not mere coincidence in the application of antitrust law to specific cases. Because antitrust authorities' enforcement policies help shape international business practices, consensus-based antitrust enforcement is vital to global business and consumer welfare. Fortunately, the United States and European Union have made some real progress toward convergence in this area, although there is still work to be done. Today, I would like to focus on one practice in particular­intellectual property licensing­and discuss several specific examples that illustrate areas of convergence and divergence between the US and EU. I will also suggest how, in some areas, a period of "constructive divergence" may ultimately help us reach consensus in the future.

II. The Antitrust Analysis of IP Licensing in the US and EU: A Brief Overview

Of course, it has become widely recognized that intellectual property ("IP") has in recent years become one of the most valuable assets in the global economy. In the Department's view, antitrust enforcement policies must be carefully designed so they do not interfere with or discourage the legitimate exploitation of intellectual property rights through technology licensing. After all, the economic return from technology licensing is what encourages innovative firms to produce new products and feed the global economy.

These days, many firms have recognized the tremendous potential of licensing their IP. These firms are investing a great deal of effort into valuing their IP portfolios in order to license more efficiently. At the same time, firms are also taking measures to protect their IP assets from piracy by aggressively enforcing their IP rights. During the 2002 Antitrust-Intellectual Property Hearings, which the Department conducted along with the Federal Trade Commission ("FTC"), we learned that many firms are willing to spend substantial resources as necessary to enforce their IP rights; as a result, would-be infringers are entering into licensing arrangements in order to avoid costly litigation.⁽¹⁾ "Defensive patenting," which refers to the practice whereby firms patent their technology specifically to ensure that they will have a seat at the licensing negotiation table, is also a common practice, according to some hearing panelists.⁽²⁾ Moreover, the advent of digital rights management ("DRM") technology, which more effectively than before limits unauthorized access to certain IP and thereby enables firms to control who uses their IP, is also strengthening the licensing trend.⁽³⁾

Technology licensing, as you might imagine, reaches across borders and touches consumers all over the world. Indeed, technology licensing has truly "gone global."⁽⁴⁾ The Department's business review letters in the late 1990s illustrate this point well. Of all the patent pools that we reviewed and approved during that period, each one involved multinational corporations.⁽⁵⁾ The reason is obvious: in analyzing these pooling agreements, one efficiency that the Department found compelling was that pooling can significantly reduce transaction costs by introducing "one-stop shopping" and thereby facilitating the manufacture of products according to industry standards. Needless to say, this efficiency would be eroded if a patent pool were deemed anti-competitive in a foreign jurisdiction and individual licenses needed to be negotiated. Increased licensing costs might also result in increased downstream prices for consumers, a real no-win situation.

As many of you know, the Department of Justice and the Federal Trade Commission have set forth the antitrust principles they will apply in analyzing licensing agreements in the 1995 Antitrust Guidelines for the Licensing of Intellectual Property ("Antitrust-IP Guidelines").⁽⁶⁾ Many of these principles are reflected in the European Commission's recently-revised Technology Transfer Block Exemption regulation (which I will refer to as the "TTBE") and the accompanying Technology Guidelines. The revised TTBE addresses the competitive significance of bilateral technology licensing agreements involving software copyright, patent, and know-how. The Technology Guidelines not only interpret the TTBE, but also provide more transparent guidance on multilateral licensing agreements as well.⁽⁷⁾ Together, the TTBE and Technology Guidelines follow an approach that is much more flexible and effects-based than the old block exemption.⁽⁸⁾ They also confirm that the US and EU have reached general consensus in a number of areas that relate to technology licensing. Here are a few examples:

• Both the TTBE and the US 1995 Antitrust-IP Guidelines start from the assumption that IP licensing is generally procompetitive.⁽⁹⁾
• Both put the onus on the parties to determine whether their actions will cause competitive harm, and both try to provide adequate guidance that allows firms to assess the legality of their conduct.⁽¹⁰⁾
• Both have safe harbors for agreements that fall below a certain market share.⁽¹¹⁾

Of course, it has not always been this way. There was a time when neither the EC nor the US adhered to a flexible analysis based on economic effects. Indeed, the Department once considered a number of licensing practices to be per se illegal without regard to economic effect.⁽¹²⁾ These prohibited licensing practices­known as the Nine No-Nos⁽¹³⁾­were later abandoned when new economic thinking began to challenge these strict prohibitions as economically unsound. This new economic analysis focused on whether an IP licensing agreement inhibited competition that would have been present but for license.⁽¹⁴⁾ That "but for" approach continues to form the basis for the Department's analysis today.⁽¹⁵⁾

Our experience with the Nine No-Nos is similar to the EC's more recent experience revising the TTBE. In its original form, the TTBE regulation favored rigid rules over a flexible antitrust analysis. Then a review of the TTBE caused the Commission to abandon its "formalistic" rules for a more economics-based approach­thus, making the TTBE consistent with other block exemptions, which are also economically-centered.⁽¹⁶⁾ And in my view, outgoing Commissioner Mario Monti deserves a lot of credit for his leadership of these changes. Both the US and EU experiences demonstrate that reassessing policy positions based on new thinking can have a positive effect on competition and consumer welfare.

III. Intellectual Property Licensing: Three Examples of Potential Divergence

Although the US and EU approaches to the antitrust analysis of IP licensing agreements have converged in many ways, there are still areas where the analysis differs in important respects. But even where their approaches diverge, the ultimate goal is the same: to promote consumer welfare while preserving the incentive to innovate. To achieve that goal, US and EU antitrust authorities should pay particularly close attention to those areas where our enforcement approaches diverge, so that we may constructively assess the strengths and weaknesses of our own respective enforcement policies and continue to learn from those assessments. This process of "constructive divergence" holds the promise of improving consumer welfare, whereas divergence alone­without reflection­could have the potential to harm consumer welfare, by, for example, needlessly stifling the efficient distribution of technology-related products to consumers.

To help advance that process, I would like to discuss several examples of licensing practices that the US and EU may analyze differently. These examples involve areas where learning through constructive divergence may prove useful.

1. Field-of-use and Territorial Restraints

The first example involves vertical restraints:

Consider a firm that develops new software for the inventory and management of electronic documents. The firm decides to license its software to end users, but it incorporates field-of-use and territorial restrictions into its licenses. As a result of these restrictions, some licensees may use the software only in their small businesses, while other licensees may use the software only in connection with the management of large-scale corporations. The licenses are also restricted by territory, so that licensees may use the software only in certain parts of the United States and certain foreign countries. The firm charges different royalty rates depending on the type of license, which allows it to price efficiently. Nothing in the licenses prevent the licensees from developing, using, or selling their own software programs.

In the US, we believe that these kinds of territorial and field-of-use restrictions are generally procompetitive.⁽¹⁷⁾ Indeed, the Antitrust-IP Guidelines analyze a similar example favorably, finding competitive harm unlikely because the licensing agreement did not impede competition among firms that were potential or actual competitors, and nothing prevented the licensees from using different software or creating their own.⁽¹⁸⁾ Moreover, allowing the firm that developed the software to price discriminate may increase social welfare by promoting the efficient commercialization of the asset. In contrast, requiring a uniform per-unit royalty might actually result in under-commercialization of the software because the licensor might end up charging some businesses higher royalties (inefficient pricing) than it might otherwise have charged. As a result, some of these same firms might forgo licensing entirely.

The EC and national competition authorities in Europe may analyze my example differently because their position on vertical restraints is quite different from ours. The EC engages in a more searching analysis of vertical restraints by analyzing how a particular restrictive provision might hinder competition and whether an objectively less restrictive solution for that provision exists.⁽¹⁹⁾ The "but-for" or counterfactual analysis that we conduct in the US, by contrast, examines only whether competition under the licensing agreement as a whole would be less than that which would occur in the absence of any licensing agreement at all.

Critics of the EC's approach, both lawyers and economists, argue that it should be more tolerant of vertical restraints­such as territorial restrictions­because restricted licenses can increase an IP owner's profits, thereby fostering further innovation and creating new competition.⁽²⁰⁾ According to some commentators, uncertainty created by the EC's approach may cause IP owners to avoid licensing their technology in this environment, choosing instead "non-licensing solutions," such as vertical integration, which may not always be as efficient, or simply not fully exploiting the technology.⁽²¹⁾ The EU recognizes in their Technology Guidelines, however, that there are legitimate, objective reasons for a licensor to impose territorial restraints on intra-brand competition, such as when the territorial restraint would allow a licensee to penetrate the market.⁽²²⁾ Thus, in our hypothetical, the agreement may be justified in the Commission's view if the software developer's licensees produced and marketed the software. But the EC's Technology Guidelines do not approve or offer other objective justifications­such as price discrimination­that the parties may rely upon in defending their agreement. Indeed, the EC may reject price discrimination as a justification in light of its desire for market integration, which is not a concern in the US.⁽²³⁾

2. Exclusivity Provisions

My next example concerns exclusivity provisions:

Imagine that Firm A is the inventor of a new medical device, but it lacks the capability to bring the new device to market. Instead of marketing the device itself, Firm A decides to license Firm B­which is not its competitor­to do so. Several other firms offer competing medical devices and their manufacture and distribution is easy. Suppose also that demand for the new medical device is uncertain and its success depends on a significant promotional effort by Firm B. Accordingly, Firm A negotiates a provision in the licensing agreement that prohibits Firm B from marketing competing devices.

The Antitrust-IP Guidelines analyze a similar hypothetical favorably because the agreement encouraged the efficient commercialization of the product and no competition concerns were apparent.⁽²⁴⁾ The exclusivity provision gives Firm B a greater incentive to invest in the commercialization, distribution, and improvement of licensed technology because it means Firm B will not have to worry that other licensees will free-ride on its investments. The exclusivity provision will thus allow the licensor (Firm A in my example) to exploit its IP rights efficiently and thus preserve its incentive to innovate in the first place.

Under the EC's old TTBE regulation, the exclusive dealing provision of the licensing agreement between Firm A and Firm B would have been black-listed. The revised TTBE, however, is much more permissive on exclusive dealing arrangements and exempts such agreements when they fall below certain market share thresholds.⁽²⁵⁾ If the market thresholds are not met, the parties may have to come up with an objective justification for their agreement, and the desire to encourage market penetration by Firm B may provide such a justification. In addition, the EC's Technology Guidelines indicate that the Commission will intervene in an exclusive licensing arrangement between non-competitors only in exceptional circumstances "irrespective of the territorial scope of the license."⁽²⁶⁾ Thus, the US and EU appear to be fairly close in their analysis of exclusive licensing, although it remains to be seen what kinds of objective justifications will be deemed acceptable by the EC in practice.

3. Maximum Resale Price Maintenance

My last example concerns resale price maintenance. Of course, minimum resale price maintenance is still per se illegal in the US. Patentees, for example, cannot control the prices at which dealers sell their patented products, and they certainly cannot coordinate with others in order to control downstream prices.⁽²⁷⁾ Licensing agreements that set maximum resale prices, on the other hand, are not per se illegal. Economic theory has taught us that consumers can benefit from such restraints because they can prevent a licensee in a given territory from charging a monopoly price, thereby keeping prices down for consumers. Thus, in State Oil Co. v. Khan,⁽²⁸⁾ the Supreme Court declared that maximum resale price maintenance is no longer a per se violation of the antitrust laws. Khan involved an agreement between State Oil and its supplier that essentially obligated the supplier to charge no more than the suggested retail price for the gasoline, which was set by State Oil. The Court saw no economic justification for applying a per se rule to such an agreement, finding rather that a "rule of reason analysis will effectively identify those situations in which vertical maximum price fixing amounts to anticompetitive conduct."⁽²⁹⁾

The EC takes a somewhat different approach. While Article 4(2)(a) of the TTBE does permit a licensor to impose a maximum sale price or recommended sale price on a licensee when the parties are non-competitors, what we would call a vertical relationship, the TTBE is less clear on the outcome when the licensor and licensee are competitors in a given market, but are actually in a vertical relationship with respect to the licensed IP. To illustrate, Firms A and B may be competitors in the market for licensing and manufacturing semiconductor technology, but Firm A and B may also be in a vertical relationship because Firm B licenses some of Firm's A patents to produce the technology. The issue becomes more complicated if the parties are in a cross-licensing relationship implicating both the horizontal and vertical relationships between the firms.

The US approach permits consideration of competitive effect for both the vertical and horizontal aspects of the licensing agreement. By contrast, the TTBE's "competitor" and "non-competitor" dichotomy might subject firms to harsher rules if they are classified as competitors without regard to the vertical aspects of their agreements.⁽³⁰⁾ Article 4(1)(a) of the TTBE, for example, prohibits all agreements between competitors that restrict "a party's ability to determine its prices when selling products to third parties," including products that incorporate the licensed technology.⁽³¹⁾ Therefore, where the US would consider Firms A and B above in a vertical relationship with respect to the licensed technology, and analyze any maximum resale price restraint in their agreement under the rule of reason, the EC may treat such a restraint (depending on the facts) as a hard core restraint between competitors­equivalent to a per se prohibition in the United States. But again, with the TTBE in its infancy, we'll have to wait and see how these agreements actually will be analyzed in practice.

IV. Conclusion

What I hope these examples demonstrate is that, while extraordinary strides have been made towards convergence between the US and EU in the application of antitrust law to intellectual property rights, there are still particular areas around the edges where differences remain. Some of these differences may be traceable to the EU's desire for a common market,⁽³²⁾ some may be the result of different policy choices, and some may simply be accidental. Whatever the cause, we might rely on a process of "constructive divergence" to bring us closer together. By recognizing our differences, paying close attention to the economic consequences of our respective enforcement decisions over time, and using those observations to test the assumptions that underlie our analyses, we might be able to come together and achieve even greater levels of convergence in the future.

1. See, e.g., Nov.6, 2002 Hr'g Tr., Antitrust Law and Patent Landscapes, at http://www.ftc.gov/opp/intellect/021106ftctrans.pdf. Transcripts of the Antitrust-IP Hearings are available on the FTC's website, at http://www.ftc.gov/opp/intellect/index.htm.

2. See, e.g., Peter Grindley, IP, Cross-Licensing and Patent Pools (Apr. 17, 2002 Hr'g R.) (slides) at 9, at http://www.ftc.gov/opp/intellect/020417petergrindley.pdf; see also, Carl Shapiro & Joseph Farrell, Intellectual Property, Competition, and Information Technology, Working Paper No. CPC04-45, Competition Policy Center UC Berkeley (2004), available at http://repositories.cdlib.org/iber/cpc/CPC04-045 [hereinafter "Shapiro & Farrell"].

3. See Shapiro & Farrell at 2, 10 (discussing schools for and against the use of DRM to protect copyrighted works of authorship); Arizona Cartridge Remanufacturers Assoc., Inc. v. Lexmark International, Inc., 290 F. Supp. 2d 1034 (N.D. Cal. 2003) (rejecting manufacturers claims that the use of DRM technology violated state law).

4. Cf. Philip Lowe, How Different is EU Antitrust?: A Route Map for Advisors, An Overview of EU Competition Law and Policy on Commercial Practices, ABA Fall Meeting, Brussels 1 (16 October 2003) (hence, competition authorities are also "going global").

5. 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.pdf (video compression technology proposal); 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.pdf (3 company DVD proposal); 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 (6 company DVD proposal); see also Letter from Charles A. James, Assistant Attorney General, U.S. Dep't of Justice, to Ky P. Ewing, Esq. (Nov. 12, 2002), at http://www.usdoj.gov/atr/public/busreview/200455.pdf (3G Patent Platform Partnership).

6. U.S. Dep't of Justice & Federal Trade Comm'n, Antitrust Guidelines for the Licensing of Intellectual Property (Apr. 6, 1995) at http://www.usdoj.gov/atr/public/guidelines/ipguide.htm [hereinafter "Antitrust-IP Guidelines"].

7. See Commission Notice, Guidelines on the application of Article 81 of the EC Treaty to technology transfer agreements, available at http://europa.eu.int/eur-lex/pri/en/oj/dat/2004/c_101/c_10120040427en00020042.pdf [hereinafter "EC Technology Guidelines"]. Article 81(1) of the EC Treaty prohibits agreements that prevent, restrict or distort competition in the European Union. Article 81(3) permits the EC to exempt certain agreements that fall within Article 81(1), if, on balance, the agreement is procompetitive. Article 81 of the EC Treaty is available on the Commission's website at, http://europa.eu.int/comm/competition/legislation/treaties/ec/art81_en.html. The TTBE exempts a whole class of intellectual property licensing agreements from the reach of Article 81.

8. Commission Regulation (EC) No. 772/2004 of 27 April 2004 on the application of Article 81 (3) of the Treaty to certain categories of technology transfer agreements, 2004 O.J. (L 123/11), available at http://europa.eu.int/eur-lex/pri/en/oj/dat/2004/l_123/l_12320040427en00110017.pdf [hereinafter "TTBE"]. The TTBE had the force and effect of law and is binding on the EC as well as member states. By contrast, the Technology Guidelines only have persuasive influence.

9. See generally Philip Lowe, Current Issues of the EU Competition Law: The New Competition Enforcement Regime, 24 NW. J. Int'l L. & Bus. 567, 581 (2004) ("Licensing, also when it contains restrictions on licensee or licensor, will therefore often be procompetitive as it allows the integration of complementary assets, allows for more rapid entry, helps disseminating the technology and provides a reward for what really is risky investment.").

10. Some commentators criticize the EC for providing firms with less certainty (the old TTBE strictly defined legal and illegal practices); however, the revised TTBE is much more tolerant of once black-listed restraints, such as exclusive grantbacks on improvements, giving firms the freedom to license more efficiently. See generally Vanessa Turner, Reform of the European Antitrust Rules on Technology Transfer ­ A "Safe harbor" for Small Tugs and Fishing Boats?, Discussion Paper Before the ABA Antitrust Law Section Annual Spring Meeting (2004) [hereinafter "Turner Discussion Paper"].

11. Antitrust-IP Guidelines § 4.3; TTBE ¶¶ 10, 11; EC Technology Guidelines ¶ 24; see also, Turner Discussion Paper at 12 (but questioning whether this is a true safe harbor when it is not explicitly set forth in the TTBE and merely appears in the Guidelines).

12. See Bruce B. Wilson, Patent and Know-How License Agreements, Field of Use, Territorial, Price and Quantity Restrictions, Remarks Before the Fourth New England Antitrust Conference 19 (Nov. 6, 1970); see also Edward G. Biester, III, Finding Reason Through Extremes: From Nine No-Nos to Patent "Immunity," 3 Antitrust and Intellectual Property 26 (Spring 2002).

13. The practices known as the "Nine No-Nos" were: (1) mandatory package licensing; (2) tying of unpatented supplies to the licensing of patented products; (3) compulsory assignment of grantbacks; (4) vertical distribution restraints, such as post-sale restraints on resale by purchasers; (5) compulsory payment of royalties of amounts unrelated to the sales of the patented product; (6) restrictions on the licensee's freedom to deal in products or services outside the scope of the patent; (7) grants to the licensee of veto power over further licenses; (8) restraints in sales of unpatented products made by a patented process; and (9) minimum price maintenance.

14. See, e.g., Charles F. Rule, Technology Licensing and the [']Second American Revolution[']: Storming the Ramparts of Antitrust and Misuse, Address Before the John Marshall Law School (Feb. 22, 1985) (quoting the State of the Union Address by President Ronald Regan (Feb. 6, 1985)).

15. See Antitrust-IP Guidelines § 3.4.

16. Lowe, 24 NW. J. Int'l L. & Bus. at 580. The old TTBE adopted a much more rigid approach setting forth specific "white-listed" "grey-listed" and "black-listed" conduct. See Commission Regulation (EC) No. 240/96 of 31 January 1996 on the application of Article 85 (3) of the Treaty to certain categories of technology transfer agreements, 1996 O.J. (L 31) 2, at http://europa.eu.int.

17. Of course, vertical restraints can cause competition concerns if they foreclose access to competing technologies, prevent research and development, or facilitate price-fixing or other cartel-like behavior. See Antitrust-IP Guidelines § 2.3 & ex. 1.

18. Id.

19. EC Technology Guidelines ¶ 12(b) (discussing the "general framework for applying Article 81").

20. See, e.g., Richard Gilbert, Converging Doctrines? US and EU Antitrust Policy for the Licensing of Intellectual Property (Feb. 16 2004) (unpublished Discussion Paper) at 13, available at http://repositories.cdlib.org/iber/cpc/CPC04-044 [hereinafter "Gilbert Discussion Paper"] ("The Commission should presume that technology licensing, even if territorially restricted, promotes rather than lessens competition.").

21. See generally Robert C. Lind and Paul Muysert, The European Commission's Draft Technology Transfer Block Exemption Regulation and Guidelines: A Significant Departure from Accepted Competition Policy Principles, CRA Competition Policy Discussion Papers (Nov. 2003), at http://www.crai.com/pubs/pub_3563.pdf; see also Maurits Dolmans & Anu Piilola, The New Technology Transfer Block Exemption, 4 Intell. Prop. L. Newsl. 32, 42 (2003) (tightening territorial restrictions with limitations on passive and active sale bans could "backfire" making it more difficult to recover sunk costs which could reduce licensing in member states or increase royalties from their current level).

22. See The EC Technology Guidelines ¶ 12(b).

23. See, e.g., Lowe, 24 NW. J. Int'l L. & Bus. at 583 ("Territorial restrictions are paid more attention to because of the additional market integration objective which EC competition policy has."); Gilbert Discussion Paper at 13 (recognizing that the reason the EU treats intra-technology differently is that the EU is concerned about agreements that prohibit trade between Member states, whereas the US does not consider the promotion of interstate trade as a specific policy objective).

24. Antitrust-IP Guidelines § 4.1.2 ex. 8.

25. The EC Technology Guidelines ¶ 197.

26. See The EC Technology Guidelines ¶ 165, accord, ¶ 12(b).

27. See § 5.2 of the Antitrust-IP Guidelines & n. 33; see also 2 Herbert Hovenkamp, Mark D. Janis, & Mark A. Lemely, IP and Antitrust: An Analysis of Antitrust Principles Applied to Intellectual Property Law § 31.2 (2002) (citing cases).

28. 522 U.S. 3 (1997), overruling Albrecht v. Herald Co., 390 U.S. 145 (1968).

29. Id. at 22, accord, id. at 15.

30. Cf. EC Technology Guidelines at ¶ 28 (stating that when two parties are both active on the same product or technology market, without one or both parties infringing the other's IP rights, they are actual competitors).

31. TTBE Art. 4(1)(a) & EC Technology Guidelines ¶ 79.

32. See e.g., Gilbert Discussion Paper at 13.