OBJECTIVE STANDARDS AND THE ANTITRUST
ANALYSIS OF SDO AND PATENT POOL CONDUCT
GERALD F. MASOUDI
address at the
Annual Comprehensive Conference on Standards Bodies and Patent Pools
October 11, 2007
The topic about which I have been asked to speak today -- the interplay among standardization, intellectual property licensing, and the antitrust laws -- is a familiar one to this audience. I am sure that you are already aware of major recent developments such as the IP2 Report(1) and the Department of Justice VITA and IEEE business review letters,(2) and you may know that I have published remarks on these issues twice before.(3) In addition, Kenneth Glazer of the Federal Trade Commission and the Department's Special Counsel for Intellectual Property, Frances Marshall, are scheduled to speak later today regarding some of the specifics of these developments. What I plan to do, therefore, is touch only briefly on these developments themselves, and devote the bulk of my remarks to explaining how these developments fit within the larger enforcement philosophy of the U.S. antitrust agencies and the development of U.S. antitrust law generally.
Antitrust analysis in the U.S. emphasizes liability tests and remedies that are objective, as opposed to subjective, and ex ante in focus, versus ex post. In addition, U.S. law establishes structural thresholds that are treated as safe harbors in many areas to allow businesses the ability to make decisions without facing serious threat of an antitrust challenge for every business decision. This is as true for technology pooling and standard setting as it is for any other area. And the beginning assumption under U.S. antitrust law is that if a market inefficiency becomes apparent, a market-based solution -- if one is available -- is generally superior to a government-imposed solution. Accordingly, I will divide my remarks today into three parts:
first, a discussion of the objective, structural safe harbor, ex ante trend in antitrust law;
I. Objective, Structural Safe Harbor, Ex Ante:
Let's start with the concept of structural thresholds treated as safe harbors. There are two leading examples in guidance issued by the Antitrust Division. One is the joint DOJ and FTC Horizontal Merger Guidelines,(4) which establish certain thresholds based on market definition and merging parties' combined market shares, which indicate that transactions falling under those thresholds ordinarily are not problematic. Transactions falling above such thresholds are examined closely for their ultimate competitive effects. Another guidance document is the joint DOJ and FTC Antitrust Guidelines for the Licensing of Intellectual Property.(5) The licensing guidelines explain that the agencies will begin by examining market structure in much the same way that they do in the merger context, will examine the extent to which a grantor-licensee relationship is horizontal as well as vertical, and will apply an "antitrust safety zone" where a licensing restraint is not facially anticompetitive and the grantor and licensee collectively account for no more than twenty percent of the affected relevant market.(6) Whether in the merger or licensing context, such structural approaches generally provide a quick, predictable mode of analysis for determining whether additional effects-based competitive inquiry is necessary.
The United States also strongly prefers structure in the remedy phase, when possible and justified by the violation. This is exemplified by the Division's Merger Remedies Guide, which makes this point explicitly. The classic structural remedy is a divestiture: an unconditional sale, in the case of a physical asset, or an irrevocable license, in the case of an intangible asset, both of which avoid the need to monitor later actions by the affected party or parties. There are a number of reasons to avoid behavioral "conduct" remedies, if possible: they create direct costs associated with monitoring the target's adherence to the decree; they may restrain procompetitive behavior by the target or otherwise prevent it from responding efficiently to changing market conditions; and they have the potential to involve courts in costly, ongoing entanglement in the market, something that is good for neither the market nor the courts. There are some situations where conduct remedies such as ongoing sharing of know-how cannot be avoided, but generally such compulsory cooperation and compulsory licensing should be, as we have said, a "rare beast."(7) U.S. antitrust agencies do not "manage" competition -- the preferred approach, where intervention is necessary, is to create a competitive structure and then step back to let the market work.
Turning next to objective tests, a leading example comes from the Supreme Court's decision in Weyerhaeuser Co. v. Ross-Simmons Hardwood Lumber Co., Inc.,(8) issued in February of this year. The case involved an allegation of "predatory bidding": a smaller lumber mill alleged that its larger rival paid high prices for lumber inputs in a scheme to prevent the smaller mill from obtaining lumber at competitive rates, leading the smaller mill to go out of business. The trial court gave a jury instruction that emphasized subjective "fairness":
One of [the plaintiff's] contentions in this case is that [the larger mill] purchased more logs than it needed or paid a higher price for logs than necessary, in order to prevent [the plaintiff] from obtaining the logs [it] needed at a fair price. If you find this to be true, you may regard it as an anti-competitive act.(9)
After the jury found for the plaintiff under Section 2 of the Sherman Act and the Court of Appeals for the Ninth Circuit affirmed, the larger mill sought Supreme Court review. The United States in an amicus brief supported the larger mill and asserted that the proper test should, instead, employ the objective principles that the Supreme Court set forth in its 1993 decision in Brooke Group Ltd. v. Brown & Williamson Tobacco Corp.(10) The Court agreed. In Brooke Group, which was a predatory pricing (sell-side) case, the Court set forth an objective two-part test: first, the plaintiff must show that the defendant's prices are below its costs; second, the plaintiff must show that the defendant has a "dangerous probability" of recouping those losses, after the predation causes the victim to stop competing.(11) In Weyerhaeuser, the Court applied the same test, adapting the first prong to the bidding (buy-side) context -- a plaintiff in a predatory bidding case must show that the defendant made bids so high in the input market that they led to below-cost pricing in the output market -- and applying the same second, recoupment prong.(12) Thus, whether in the upstream or downstream side of a market, analysis of allegedly predatory conduct must be objective.
The Supreme Court rejected subjective tests in Weyerhaeuser because subjective tests do not provide a useful screen: they do not help to answer the question of whether conduct harms competition as a whole, as opposed to harming only a particular competitor. Raising prices for inputs or lowering them for outputs is core competitive activity -- "the very essence of competition," according to the Court(13) -- and "too lax a liability standard" runs a high "risk of chilling procompetitive behavior,"(14) which would harm the very market participants that the law seeks to help. As the United States explained in its brief to the Supreme Court in Weyerhaeuser:
If allowed to stand, such a subjective standard would have the effect of deterring procompetitive conduct by large firms. . . . antitrust rules "must be clear enough for lawyers to explain them to clients" and "must be designed with the knowledge that firms ultimately act, not in precise conformity with the literal language of complex rules, but in reaction to what they see as the likely outcome of court proceedings." If the line between lawful competitive bidding and unlawful predatory bidding were to turn on a jury's ex post assessment of whether the price paid for an input was excessive, large firms competing for inputs would rationally err on the side of caution, pull their competitive punches, and bid less aggressively.(15)
That quote leads me nicely into a discussion of the ex ante focus of U.S. antitrust law, as opposed to the ex post perspective.
Under U.S. law, the key inquiry of antitrust liability -- the question, "is a firm's conduct harmful to competition?" -- is asked based on the state of the world as it existed at the time the conduct was undertaken. The ex ante inquiry applies to analysis of structure: for example, in a licensing context, the question of whether grantor and licensee are horizontal competitors is asked as of the time before the license was signed.(16) The ex ante inquiry also applies to effects: for example, a merger between two of the ten firms in a competitive market does not become retroactively a violation of the antitrust laws merely because the eight other firms independently decide later to exit. To summarize the ex ante focus of U.S. antitrust law, it is difficult to do better than Judge Learned Hand's famous quote from the 1945 Alcoa case: "[t]he successful competitor, having been urged to compete, must not be turned upon when he wins."(17)
So, to recap, the trend in U.S. antitrust law is toward objective, ex ante analysis. U.S. antitrust law disdains any rule that would depend on subjective concepts such as fairness or the intent of a firm's employees, continuing government "management" of the marketplace, or the market impact of future unknown events, since these types of rules are likely to chill businesses from the aggressive, hard-nosed competition that leads them to invest, innovate, and ultimately benefit consumers.
II. Antitrust Law Does Not Prevent Competitive Business Solutions
Let's evaluate how these principles fit the United States' approach to antitrust enforcement in the areas of patent pools and standards development organizations.
For pools, antitrust analysis begins, as always, with the ex ante state of the world -- the state, in other words, in which a pool does not yet exist. In some industries, that world faces a number of problems, among them: a "patent thicket" of multiple blocking positions, defined as a situation where several patent holders can each block a product in the absence of a license from each; high transaction costs, since a licensee must negotiate (or litigate) separately with the multiple patent holders; and no mechanism for sharing useful non-patented information such as manufacturing secrets or medical dosing. This is a highly inefficient ex ante world. Patent pools have the potential to solve each of these problems by eliminating multiple blocking positions, reducing transaction costs, and providing an efficient mechanism for sharing useful non-patented information. The potential for increased efficiency is tremendous, which ultimately translates into a benefit for consumers. Of course, U.S. antitrust law bases its ultimate conclusion on specific inquiries regarding a particular pool -- not pools in the abstract -- but as a general matter, this ex ante analysis leads to a strong argument in favor of patent pools.
There also are risks of anticompetitive effects from patent pools: pools could potentially exclude new innovation; they could reduce competition if they were to prevent patent owners from competing for licensees; they could reduce a potential licensee's incentives to challenge invalid patents; and they could provide a forum for collusion between or among competitors. In response to these concerns, the Department of Justice issued several patent pool business review letters(18) in the late 1990s stating that it did not intend to bring antitrust enforcement against specific proposed pools. Taken together, these letters identify a number of pool features that act as partial safeguards against anticompetitive effects, five of the most important being: clarifying which patents are in the pool; limiting pools to complementary patents and avoiding substitutes; requiring licenses into and out of(19) the pool to be nonexclusive, and licenses out to be nondiscriminatory; limits on the scope of grantback demands; and strict, written limits on the collection of and access to competitively sensitive proprietary information of pool members and licensees, to prevent downstream coordination.
Notice that each of these features can be judged ex ante, when the pool is formed, and each is objective, structural or, if behavioral, relatively easy to administer. Clarifying which patents are in the pool is objective: either the patent is listed, or it is not. The determination of whether two patents are substitutes or complements, although it may require expertise, is also objective. Including only complementary patents in the pool is structural. Strict written limitations on information sharing are behavioral, but relatively easy to administer: instead of permitting information sharing and then attempting to police what pool members do with it, the pool simply prohibits the practice completely. And licenses that are nonexclusive, nondiscriminatory, and limited as to grantbacks are both objective and relatively easy to administer: they are objective in the sense that observers should be able to agree as to how they are defined -- is the grant nonexclusive, yes or no? -- and they are administrable in that they can be clean transfers of value with a minimum of contingencies and coordination between grantor and licensee.
None of the safeguards in the Department's patent pool letters depends on a fairness inquiry. None depends upon aspects of the marketplace that cannot be determined on the day the pool is first organized. None attempts to regulate purely business issues, such as price. As a pool is implemented and maintained, other issues may arise that will not be susceptible to such clean resolution, but the structure discussed in the Department's patent pool letters can significantly cabin those issues so that disagreements, if they occur, should occur within a narrow range. The business and antitrust risks of such a pool, therefore, should be reasonably predictable.
Now let's apply the same lens to the Department's business review letters regarding patent policies for the VITA and IEEE standards development organizations (SDOs).(20) The core mission of these SDOs is to establish functional and interoperability standards so that different firms' products will work together. This has the potential to vastly increase efficiency. There is a potential business problem, however, in the form of patent hold up: where firms, after a costly selection process, have committed to a particular standard, a patent that was previously substitutable may become essential, giving the patent owner the ability to hold up the standard and demand a higher rate than would have been the case if it had to compete, before the standard setting, with alternative technologies. To prevent such hold up, VITA and IEEE proposed patent policies, and the Department of Justice issued favorable business review letters to both.(21)
VITA's patent policy had five key provisions: disclosure of patents or patent applications; disclosure of maximum royalty rates and other most-restrictive terms; limited application, meaning that these commitments apply only to implementation of the VITA standard being developed, not to other uses of the technology; prohibition on horizontal negotiations of licensing terms; and an arbitration and consequences provision to police compliance -- in essence, SDO participants who fail to disclose their patents risk having them licensed on a royalty-free basis. IEEE proposed a somewhat different design, giving potential patent holders five options as to whether and how to provide assurance about the existence of patents and maximum licensing terms, and then post the licensing assurances, or lack thereof, on IEEE's website. If a patent holder commits to maximum price and non-price terms, the IEEE working groups proposed to use this information to assess the relative costs of alternative technologies. And the IEEE policy permits its members to consider economic terms only in generalized or non-collusive ways. The policy "prohibits discussion of specific licensing terms within . . . standards development meetings" and, based on statements by IEEE's counsel, the Department understood that "this prohibition extends to joint negotiations of licensing terms within standards development meetings."(22)
As with patent pools, the Department's antitrust analysis of SDO patent policies begins with the ex ante state of the world. In the case of VITA and IEEE, these particular SDOs had actual experience with the ex ante world, and asserted that a lack of patent policies left their standard setting efforts vulnerable to inefficiencies, and perhaps could cause some standards on the margin to never be issued or, once issued, to fail. A potential objection to these SDO patent policies is that they could drive down the rewards to patent holders, thereby reducing innovation incentives, which is a serious argument; however, VITA and IEEE's policies were meant to ensure that prices would be pushed down (if at all) only through competition, not collusive activity by licensees. The Department could hardly object to that strategy; as former AAG Hew Pate said in this context as early as 2005, "[i]t would be a strange result if antitrust policy [were] used to prevent price competition."(23) For this reason and for others explained in the letters and Chapter 2 of the IP2 Report,(24) the Department issued favorable business reviews, with appropriate expressions of caution.
As with the features of patent pools that I discussed earlier, each of the features of SDO patent policies can be judged ex ante, when the patent policy is formed, and each is objective, structural, or a behavioral provision that is relatively easy to administer. Disclosure of patents and maximum terms is objective: when later asked to examine a patent or licensing demand, the question, "was that patent or term disclosed?," should be answerable. The prohibitions on joint negotiations of licensing terms within standards development meetings is relatively easy to administer. While the prohibition does require some ongoing policing of behavior, it does not require a monitor to perform the more difficult task of judging whether some joint negotiations are good and some are bad; instead, both SDOs drew a bright line and prohibited joint negotiations altogether, which is a much easier rule to enforce.
I am often asked to explain the larger message from Department decisions such as those reflected in the VITA and IEEE letter. My first reaction is that the Department was evaluating specific facts or proposed facts, which is how we always operate. The Department is not in the business of endorsing particular approaches to intellectual property licensing and it does not rule out any other, different approach to SDO patent policies; one size does not fit all. That said, there is a message to be drawn from the letters, and it is the same message set forth in the IP2 Report's Chapter 2, regarding standards. The message is that the Department and the FTC intervene only where a practice imposes a restraint on competition and is likely to harm efficiency and the competitive process itself. Businesses should be confident that antitrust enforcement will not stand in the way of conduct that can fairly be described, as the VITA and IEEE plans were, as competition related to intellectual property.
III. RAND/FRAND Provisions
This brings me to another common feature of SDO policies, the RAND or FRAND license. I will use the remainder of my remarks to identify what is subjective about the RAND-FRAND approach, and to explain how these subjective elements clash with the modern trend of objective tests for antitrust liability and remedy.
Let's apply our objective-ex ante lens to reasonable and nondiscriminatory (RAND) royalty obligations and their close cousin FRAND (Fair RAND). Imagine an SDO that does not have a patent policy such as VITA or IEEE's, but instead simply asks participants to commit to licensing their technology to standards-implementers on RAND terms. This is an ex ante commitment, of a sort, but quite a vague one. The term "nondiscriminatory" is a stab toward an administrable behavioral obligation, but there is much detail left to be decided: must all licensees receive the same royalty and other terms on a per-unit basis, or does the nondiscrimination apply only as compared to groups of similarly-situated companies? Does RAND mean that volume discounts are banned, since they may be seen to discriminate in favor of more efficient high-volume producers? What royalty price is "reasonable," and where does a neutral observer go to find the comparator and answer the question, "reasonable as compared to what?"
Under RAND, answers to such questions can be unstructured and subjective, and invite all manner of ex post evaluation. And it is difficult to see how the situation is improved greatly by FRAND. FRAND adds the concept of "fair," in an attempt to prevent predatory conduct by patent holders, but recall the Supreme Court's dim view of a similar concept, "fair price," in the predation context: it found in Weyerhaeuser that the inclusion of a "fair price" concept in a jury instruction was grounds for reversal in an antitrust case, because "too lax a liability standard" runs a high "risk of chilling procompetitive behavior."(25)
Given the vagueness of RAND-FRAND licensing obligations, they can be difficult to enforce in an action for breach of contract. For the same reasons, using alleged breaches of RAND-FRAND licensing obligations as a basis for antitrust liability does not comport well with the modern trends of antitrust law as exemplified by Weyerhaeuser.
So what is the proper role of antitrust with respect to RAND obligations? Antitrust protects the competitive process, not particular competitors. It is not a tool for limiting patent rights or for reflexively reining in the power of whatever player has developed pricing power at a particular time. Since at least 1911, the U.S. rule is that antitrust does not ban the mere possession of monopoly power, but instead bans only the acquisition or maintenance of monopoly through improper means.(26)
At one time, patent holders and other SDO participants may not have realized the potential vagueness and business problems associated with RAND obligations. However, I submit that in the wake of well-publicized ex post disagreements over the meaning of RAND, it is becoming difficult to argue that SDOs and their members reasonably expect that RAND alone guarantees smooth license sailing. As the potential problems with RAND-FRAND become more clear and the potential options for addressing those problems are developed, I anticipate that, over time, antitrust complaints will play a declining role in this arena.
Where genuine fraud, predation, or other objectively unreasonable conduct occurs, antitrust likely always will have a role in standard setting. But where conflict can be fairly characterized merely as a foreseeable disagreement over price, it would be odd and inefficient to federalize such disagreements through use of the antitrust laws and the imposition of treble damages.
1. U.S. Dep't of Justice & Fed. Trade Comm'n, Antitrust Enforcement and Intellectual Property Rights: Promoting Innovation and Competition (April 17, 2007), http://www.usdoj.gov/atr/public/hearings/ip/222655.pdf [hereinafter IP2 Report]. The shorthand title "IP2" is used to distinguish this from an earlier report issued by the FTC. See Fed. Trade Comm'n, To Promote Innovation: The Proper Balance of Competition and Patent Law and Policy (2003), http://www.ftc.gov/os/2003/10/innovationrpt.pdf.
2. Letter from Thomas O. Barnett, Assistant Attorney Gen., U.S. Dep't of Justice, to Michael A. Lindsay, Esq. (Apr. 30, 2007), http://www.usdoj.gov/atr/public/busreview/222978.pdf [hereinafter "IEEE Business Review Letter"]; Letter from Thomas O. Barnett, Assistant Attorney Gen., U.S. Dep't of Justice, to Robert A. Skitol, Esq. (Oct. 30, 2006), http://www.usdoj.gov/atr/public/busreview/219380.pdf [hereinafter "VITA Business Review Letter"].
3. See Gerald F. Masoudi, Deputy Assistant Attorney Gen., U.S. Dep't of Justice, Efficiency in Analysis of Antitrust, Standard Setting, and Intellectual Property, Address at the High-Level Workshop on Standardization, IP Licensing, and Antitrust 15-16 (Brussels, Belgium, Jan. 18, 2007), http://www.usdoj.gov/atr/public/speeches/220972.pdf; see also Masoudi, Antitrust Enforcement and Standard Setting, infra n. 21.
4. U.S. Dep't of Justice & Fed. Trade Comm'n, Horizontal Merger Guidelines (1992 & rev. ed. 1997), http://www.usdoj.gov/atr/public/guidelines/hmg.pdf.
5. U.S. Dep't of Justice & Fed. Trade Comm'n, Antitrust Guidelines for the Licensing of Intellectual Property (1995), http://www.usdoj.gov/atr/public/guidelines/ipguide.pdf [hereinafter Antitrust-IP Guidelines].
6. Id. §§ 4.1, 3.3, & 4.3, respectively.
7. Makan Delrahim, Deputy Assistant Attorney Gen., U.S. Dep't of Justice, Forcing Firms to Share the Sandbox: Compulsory Licensing of Intellectual Property Rights and Antitrust, address before the British Institute of International and Comparative Law 2 (May 10, 2004), http://www.usdoj.gov/atr/public/speeches/203627.pdf.
8. 127 S. Ct. 1069 (2007); see also Brief for the United States as Amicus Curiae Supporting Petitioners (No. 05-381), http://www.usdoj.gov/atr/cases/f217900/217988.pdf.
9. Confederated Tribes of Siletz Indians v. Weyerhaeuser Co., 411 F.3d 1030, 1036 n.8 (9th Cir. 2005).
10. 509 U.S. 209 (1993).
11. Id. at 222, 224.
12. See id.
13. Weyerhaeuser, 127 S. Ct. at 1077 (citing Brooke Group, 509 U.S. at 226).
14. Id. at 1078.
15. Brief for the United States, supra note 8, at 29 (quoting Town of Concord v. Boston Edison Co., 915 F.2d 17, 22 (1st Cir. 1990)).
16. See Antitrust-IP Guidelines, supra note 5, § 3.3.
17. United States v. Aluminum Co. of Am., 148 F.2d 416, 430 (2d Cir. 1945).
18. Letter from Joel I. Klein, Assistant Attorney Gen., U.S. Dep't of Justice, to Carey R. Ramos, Esq. (June 10, 1999), http://www.usdoj.gov/atr/public/busreview/2485.pdf (6C DVD Business Review Letter); Letter from Joel I. Klein, Assistant Attorney Gen., U.S. Dep't of Justice, to Garrard R. Beeney, Esq. (Dec. 16, 1998), http://www.usdoj.gov/atr/public/busreview/2121.pdf (3C DVD Business Review Letter); Letter from Joel I. Klein, Acting Assistant Attorney Gen., U.S. Dep't of Justice, to G[a]rrard R. Beeney, Esq. (June 26, 1997), http://www.usdoj.gov/atr/public/busreview/215742.pdf (MPEG-2 Business Review Letter); see also Letter from Charles A. James, Assistant Attorney Gen., U.S. Dep't of Justice, to Ky P. Ewing, Esq. (Nov. 12, 2002), http://www.usdoj.gov/atr/public/busreview/200455.pdf (regarding a "patent platform," similar to a pool, for third-generation (3G) wireless communication technologies).
19. In licensing parlance, a party licenses "in" when it is the licensee. A party licenses "out" when it is the grantor.
20. See supra note 2.
21. For a more detailed discussion of the VITA and IEEE business review letters, see my previous remarks: Gerald F. Masoudi, Deputy Assistant Attorney Gen., U.S. Dep't of Justice, Antitrust Enforcement and Standard Setting: the VITA and IEEE Letters and the "IP2" Report, Address at the Spring Meeting of the American Intellectual Property Law Association (Boston, May 10, 2007), http://www.usdoj.gov/atr/public/speeches/223363.pdf.
22. IEEE Business Review Letter, supra note 2, at 11.
23. R. Hewitt Pate, Assistant Attorney Gen., U.S. Dep't of Justice, Competition and Intellectual Property in the U.S.: Licensing Freedom and the Limits of Antitrust, Address at the 2005 EU Competition Workshop 9 (Florence, Italy, June 3, 2005), http://www.usdoj.gov/atr/public/speeches/209359.pdf. The FTC's Chairman has made similar statements. See Deborah Platt Majoras, Chairman, Fed. Trade Comm'n, Recognizing the Procompetitive Potential of Royalty Discussions in Standard Setting, Address at Standardization and the Law: Developing the Golden Mean for Global Trade 7 (Stanford University, Sept. 23, 2005), http://www.ftc.gov/speeches/majoras/050923stanford.pdf, ("[T]ransparency on price can increase competition among rival technologies striving for incorporation into the standard at issue.").
24. IP2 Report, supra note 1, at 33-56.
25. Weyerhaeuser, 127 S.Ct. at 1078.
26. United States v. Am. Tobacco Co., 221 U.S. 106, 180-82 (1911).