This document is available in two formats: this web page (for browsing content) and PDF (comparable to original document formatting). To view the PDF you will need Acrobat Reader, which may be downloaded from the Adobe site. For an official signed copy, please contact the Antitrust Documents Group.
 
HP Invent logo
Hewlett-Packard Company
3000 Hanover Street
Mail Stop 1051
Palo Alto, California 95034
www.hp.com

Jeffery Fromm
Vice President, Deputy General Counsel &
Director of Intellectual Property
650.857.2472 Tel
650.852.8194 Fax
jeff.fromm@hp.com

April 25, 2006

Legal Policy Section
Antitrust Division
U.S. Department of Justice
Suite 3234
950 Pennsylvania Avenue, N.W.
Washington, D.C. 20530
Donald S. Clark
Office of the Secretary
Federal Trade commission
Room H-135 (Annex Z)
600 Pennsylvania Avenue, N.W.
Washington, D.C. 20580

Re: Comments Regarding, Section 2 Hearings, Project No. PO62106

Ladies and Gentlemen:

Hewlett-Packard Company ("HP") is pleased to submit this comment in the above-referenced proceeding for the purpose of proposing one addition to the published list of particular business practices that raise issues under Section 2 of the Sherman Act and that will be the focus of discussion at the upcoming hearings. HP's proposed addition is as follows: patent owners' conduct in standard-setting proceedings that results in incorporation of their patented technologies into adopted standards followed by imposition of exclusionary license terms upon locked-in industry rivals.

The practices in question and HP's concerns with them are summarized in the attached paper that HP submitted with four other companies to the Department of Justice and the Federal Trade Commission in June of last year (exhibit A). HP's perspectives on this situation were also presented in two papers submitted in connection with the 2002 FTC/DOJ hearings on the intersection between competition and intellectual property law and policy (attached as exhibits B and C). We appreciate the fact that both agencies are already familiar with anticompetitive effects of what has come to be described generally as standards-related "patent holdup conduct." Assistant Attorney General Hewitt Pate discussed it in his speech of June 3, 2005; FTC Chairman Deborah Majores discussed it in her speech of September 23, 2005.

The FTC has challenged one dimension of this conduct, based on allegations that patent owners have completely hidden the existence of their patent claims, in its Dell, Unocal and Rambus cases. As the attachments to this letter reflect, however, similar anticompetitive consequences arise even when patent claims are disclosed during a standard-setting proceeding; patent owners induce acceptance of their patented solutions on the basis of generalized assurances of reasonable and non-discriminatory ("RAND") license terms while withholding information on actual license terms until well after the standard is adopted and prospective licensees are heavily invested in the development of compliant products. The patent owner thereby leverages the standard into monopoly power over the market in which prospective licensees seek to participate, monopoly power the owner would not otherwise have obtained if the intended license terms were disclosed in the course of the standard-setting process when participants could have chosen available, less costly, competitive solutions.

HP and a growing number of other standard-setting participants have urged standards development organizations ("SDOs") to address this problem by adopting new policies to require or at least encourage patent owners to disclose license terms before votes are taken on a proposed standard and to make RAND commitment more meaningful and enforceable after a standard has been adopted. Most SDOs have declined to move in these directions despite a proliferation of anticompetitive patent holdup situations leading to litigation that stalls acceptance of the affected standards. These situations undermine fundamental "open" standards objectives, morphing what would otherwise be robustly competitive markets into monopolized markets controlled by a patent owner's proprietary technology.

HP's proposed addition of this subject to the agenda for the upcoming hearings would enable the agencies to develop and publicize guidance on the circumstances under which patent owners' practices of various kinds during and after a standard-setting proceeding may be considered "exclusionary" conduct reachable under Section 2 of the Sherman Act. This issue is directly presented in at least one pending private suit, Broadcom v. Qualcomm, Civil Action No. 05-3350 (D.N.J., filed July 1, 2005), arid can be expected to arise in a growing number of similar cases over the years to come. Several members of the FTC staff have co-authored an article suggesting how practices in this area could be challenged under Section 2 principles as an "abuse of standard-setting process" through "opportunistic behavior that confers market power"; their analysis provides a promising platform for a broad dialogue on this subject. Creighton, et al, Cheap Exclusion, 72 Antitrust LJ. 975, 987-89 (2005).

If the practices discussed hereinabove are included in the hearings agenda, HP will be pleased to provide a representative to participate and contribute to the hearings record on them. We believe several other companies would similarly welcome the opportunity to be heard on these issues.

 

Sincerely,



Jeffery B. Fromm
Vice President, Deputy General Counsel
and Director of Intellectual Property
Hewlett-Packard Company

cc: Gail Kursh, Esq.
Frances E. Marshall, Esq.
Susan S. DeSanfi, Esq.


TAB A

Disclosure of Licensing Terms During Standard Setting:
The Need for Antitrust Agency Guidance

June 2005

Submitted by:

Apple Computer
Cisco Systems
Hewlett-Packard Company
International Business Machines
Sun Microsystems

To:

The Antitrust Division of the United States Department of Justice
The Federal Trade Commission


Disclosure of Licensing Terms During Standard Setting:
The Need for Antitrust Agency Guidance

  1. Introduction

Both federal antitrust agencies are familiar with the "patent thicket" confronting innovators in high technology industries such as semiconductors, software, imaging, and communications. The proliferation of patents in high technology means that the creation of industry standards will often result in the disclosure of dozens of intellectual property rights owned by numerous companies, universities, government research organizations, and individuals that participate in a particular standards exercise.

An example of the proliferation of IP rights in standard setting is the 802.11 series of standards for wireless Ethernet data networking, which have been created under the auspices of the Institute of Electrical and Electronics Engineers ("IEEE"). In the decade since the 802.11 standard-setting effort began, over sixty different entities have made over 100 separate disclosures of patents that the disclosers believe to be necessary to implement one of the family of 802.11 standards.1

The rules of standards development organizations ("SDOs") typically request participants in standard setting to disclose prior to adoption of a standard only that they are willing to make available IP rights necessary to practice a standard either without compensation or "under reasonable rates, with reasonable terms and conditions that are demonstrably free of any unfair discrimination."2 At least one standards organization - the IEEE - goes further, and advises participants in standard setting as follows: "You must not discuss subjects like the pricing for use of a patent, how a patent should be licensed, validity or interpretation of a patent claim, or any terms or conditions of use. These are not appropriate topics for discussion in a standards developing committee."3

The IEEE bases this prohibition on licensing disclosures that go beyond an offer to license IP rights on free or reasonable and non-discriminatory ("RAND") terms in significant part on the need to avoid antitrust liability for the IEEE and its members.4 Guideline 10 of the IEEE's "Guide to Standards Meetings Policies" directs participants in IEEE standard-setting activities to "Refrain from discussions that violate anti-trust laws," including the "validity of patents or the cost of using them." 5

This concern is not unique to IEEE. As Assistant Attorney General Pate observed in a speech he gave earlier this month, "Some standards development organizations have reported to the Department of Justice that they currently avoid any discussion of actual royalty rates, due in part to fear of antitrust liability."6 Explaining the American National Standards Institute's patent policy, which limits disclosure of licensing intentions to free or RAND, Amy Marasco, then ANSI's General Counsel, likewise cited antitrust concerns, stating that "[m]any believe that the discussion of licensing issues among competitors in a standards-setting context imposes a risk that the [SDOs] and the participants will become targets of allegations of improper antitrust conduct."7

  1. SDO rules that limit disclosures of licensing terms to RAND suppress competition.

As Assistant Attorney General Pate observed in his Florence speech earlier this month, "[tjhere is a possibility of anticompetitive effects from ex ante license fee negotiations, but it seems only reasonable to balance that concern against the inefficiencies of ex post negotiation and licensing hold up."8 From the perspective of five leading technology companies that each participate in standard setting on a daily basis, we believe that rules preventing detailed disclosures of licensing terms harm competition in at least three ways:

  • During the standard-setting process, these rules force participants to choose between alternative technologies based on technical merit alone, with no consideration given - or permitted - to the royalties the providers of those technologies will seek if their alternative is selected for inclusion.

  • Once a standard is adopted, companies seeking to commercialize products that implement the standard face uncertainty regarding what they will ultimately have to pay in royalties to the numerous holders of IP rights that might seek a RAND royalty on each unit they sell. This creates the risk that pricing decisions premised on particular assumptions regarding royalty costs will - years later - turn out to be wildly incorrect.

  • Once products implementing a standard begin to enjoy commercial success, each holder of IP rights that are necessary to implement the standard obtains significant market power, market power that it might not have attained during the standard setting process if the cost of the technology had been known to other participants. Each may be tempted to exploit that market power in an exclusionary manner by seeking a royalty that is limited by nothing more than its own conception of "reasonableness."

The anti-competitive consequences of standards organization rules that prohibit disclosures of licensing intentions that go beyond RAND are heightened by the sheer number of entities that disclose IP rights in standards proceedings like 802.11. In any standard setting exercise with so many participants, it is likely that the great majority of participants disclosing IP rights will never successfully implement the standard in commercial products. For those that do not, the only way to benefit from their contribution will be to monetize the IP right they have disclosed by exploiting the market power that comes with inclusion in a successful standard. The potential stacking of royalty claims made by each of the numerous parties that disclose IP rights may make potential royalty payments a significant barrier to successful commercialization of products that implement the standard.9

The stacking issue is further magnified by the "entire market value" rule of patent damages, which permits a patentee to base damages "on unpatented components typically sold with a patented item."10 The effect of the "entire market value" rule is to permit each of potentially dozens of parties that disclosed IP rights in the course of a standard-setting exercise to claim damages based on the entire value of the products implementing the standard that the putative infringer sells.

On the other side of the ledger of competitive harms is the concern that permitting disclosures of royalty rates and other licensing terms creates the increased potential for licensor or licensee coordination or collusion. With respect to licensee coordination, we note that both agencies recognize that group purchasing agreements can have powerful pro-competitive effects.11 A group of licensees negotiating collectively with owners of competing technologies, only one of which will be included in a standard that will be available on non-discriminatory terms to the licensees and their competitors, with no downstream restriction on the licensees' output of standards compliant products, does not implicate the antitrust concerns typically associated with oligopsony or buyers' cartels.12 Put simply, it is not the case that licensee coordination is inevitably anticompetitive.

Moreover, should a party disclosing its intellectual property in standard setting come to believe that pricing disclosures in standard setting had resulted in anticompetitive behavior by potential licensors or potential licensees, it can address this behavior using the tools of government and private antitrust enforcement. Claims of this kind have been a staple of private antitrust litigation.13 Given the existence of government and private enforcement, SDO rules that prohibit any disclosure of licensing terms are a grossly overbroad reaction to the risk that such disclosure might give rise to increased antitrust concerns in standard setting. We are unable to identify another context in which antitrust concerns with illegal agreements between buyers or between sellers have been addressed by banning sellers or licensors from disclosing the price they seek.

In short, the benefits of more detailed disclosures of pricing information in standard setting far outweigh the small risk that collusive behavior between prospective licensors or licensees will go undetected and unpunished. Making meaningful pricing information available during the course of standard setting will:

  • Improve the choices participants in standard setting make between technological alternatives by permitting participants to make tradeoffs between cost and technical sophistication;
  • Increase the accuracy of product pricing decisions by removing the uncertainty around what companies implementing a standard will wind up paying holders of IP rights necessary to practice that standard; and
  • Prevent the exploitation of market power by parties that disclose IP rights during the course of standard-setting but later seek purportedly "reasonable" royalties from companies that successfully implement the standard.

  1. Agency Guidance Should Favor the Ability of SDOs to Permit Disclosures of Royalty Terms that Go Beyond RAND.

Like other companies that make standards-compliant products, our five companies face increasing threats of patent infringement litigation from entities that disclose IP rights to standards bodies but later exploit the amorphous concept of RAND. We have responded by encouraging standards bodies in which we regularly participate to amend their rules to permit more specific and informative disclosures of proposed licensing terms. We have been met with resistance, stemming in part from the standards organizations' concerns with antitrust liability.

The Antitrust Division and the Federal Trade Commission are uniquely well situated to offer guidance to address the concerns standards organizations have with the consequences of moving away from their current rules. In view of the competitive harm that rules limiting disclosure to RAND are causing, both agencies should state clearly and unequivocally that standards body rules that permit disclosure of actual licensing terms, including royalty rates and other relevant terms, benefit competition.14

In announcing a clear position in favor of standards organization rules that permit disclosure of royalties and other license terms, the Antitrust Division and FTC would be doing no more than applying to the standards arena the position they have long taken in opposing rules adopted by professional associations that discourage or prevent the disclosure of pricing terms by members.15 Standards organization rules that prohibit disclosures of licensing terms beyond RAND have the effect of barring parties that are proposing substitute technologies from disclosing at what price they will make their IP rights available to potential licensees. As in Professional Engineers, such rules operate "as an absolute ban on competitive bidding" between technology suppliers vying to have their invention included in a standard.

* * *

As Assistant Attorney General Pate noted in his Florence speech, standards organization policies that prohibit detailed royalty disclosures during the course of standard setting present the "strange result" of "antitrust policy ... being used to prevent price competition."16 As the nation's preeminent authorities on antitrust law and policy, the Antitrust Division and Federal Trade Commission can correct this "strange result" by assuring standards organizations that they have the freedom to change their rules and bring the benefits of competition to the standard setting process.


TAB B

Patents and Standard-Setting P
rocesses Statement of Scott K. Peterson, Hewlett-Packard Company
for

FTC/DOJ Hearings on Competition and Intellectual Property
Law and Policy in the Knowledge-Based Economy

April 18, 2002

Introduction

I am pleased to have this opportunity to share perspectives from my experience on behalf of Hewlett-Packard Company (HP) as a participant in a wide array of standard-setting processes in the information technology sector. My focus is on standards that enable interoperability among both competing and complementary products employing new technologies, presenting some challenges beyond those involved in more traditional safety or related kinds of standards. Your agencies' interest in standards of this kind should be welcomed in many quarters in light of their central contribution to the evolution of open, competitive and innovative markets across the information technology landscape.

I propose to discuss the increasing role of patents in this area and some difficult issues presented by this trend. All of us involved in these standards processes should by this stage be sensitive to the combination of both positive and negative effects that can occur when technology subject to patent protection ends up in a final standard: it can enhance the quality of the standard and thereby promote both competition and innovation in affected new markets; but it can also enable the patent holder to obtain and exercise market power and to act opportunistically against its rivals. Before delving into these concerns, however, it is important to delineate the diversity of processes, contexts and circumstances under which IT standards evolve. An appreciation of this diversity should then elucidate the need for considerable flexibility and experimentation in approaches to addressing the issues presented.

  1. Diversify of the Standard-Setting Universe

Proposals to develop new information technology standards emerge in a myriad of ways and are pursued through a great variety of organizational structures. One can begin to appreciate the diversity by recognizing three different but common kinds of approaches: promoters' groups, consortia and standards development organizations ("SDOs").

A promoter's group may arise when a single firm seeks to develop a set of specifications around its own technology or piece of technology in a manner that facilitates widespread deployment so that the specifications ultimately becomes an industry standard. The firm -- or promoter -- may invite a small number of other firms, selected on the basis of their particular capabilities and incentives to contribute to the objective, to join as a group in developing the specifications. The group may proceed on a fast track and cease to exist upon completion of initial specifications. Its work product may thereupon be submitted to an SDO for formal adoption as an industry standard.

A consortium may arise when several firms involved in a technology market share a view on the need for standards that promote interoperability among their products and that can thereby expand their market to the advantage of all suppliers and users alike. They create an organization and agree on procedures for it as a vehicle for their collective development of the envisioned standards. It will typically encompass a larger number of firms, advance a broader agenda and remain at work for a significantly longer period than a promoter's group. It may, for example, contemplate developing a relatively broad array of specifications and successive generations of specifications, promoting their adoption over the course of several years.

An SDO may be an established trade association with a broad and diverse membership. One of its longstanding functions may be development of many different kinds of standards at the request of its members. It will typically pursue standard-setting in accordance with detailed ANSI-compliant procedures and policies aimed at ensuring maximum openness, due process and "consensus" decision-making among all affected or interested parties. It has both advantages and disadvantages as compared to what I have described as promoters' groups and consortia. On the one hand, the broader and more fullsome participation of affected interests can result in a better, more open and useable standard. On the other hand, the process in accordance with those procedures can take much longer to conclude than processes employed by smaller groups in less formal settings.

This tripolar picture I've outlined is a considerable oversimplification of the real world. There are many hybrids that fall, for example, somewhere between what may look like a promoter's group and what may look like a consortium. Many consortia, moreover, follow rules capturing the substance and spirit if not all procedures of an SDO. And, as noted, the work product of a promoter's group (as well as that of a consortium) may end up before an SDO for ultimate certification as an ANSI standard.

There is great value in this diversity. Some technologies are more complex and difficult than others in terms of the facility of their translation into open standards. Marketplace dynamics may call for particularly expedited processes in some instances but can tolerate longer, more deliberate incubation periods in other instances. Some standards may affect the competitive opportunities of more classes of parties and in more fundamental ways than will be the case with other standards, thus calling for different kinds or degrees of participation rights. Some of these processes entail joint development of the technology in question, thus looking more like what could be called a "joint production venture" than a traditional standard-setting process, or a hybrid of both kinds of undertakings, thus calling for different degrees of tolerance for development efforts driven by one or a small number of firms. This last example is now common and is often aimed at building a market for what at the beginning may well include technology subject to patent protection. Therein lies one among several reasons why patents are playing an increasing role in standard-setting. I will shortly outline other reasons as well.

Consider the following examples that illustrate (i) the need for different types of standard-setting rules and (ii) the need to pay close attention to patents in some cases but less so in other cases:

Case A: Success of standard A depends on its adoption in products from hundreds of different sources; some of these products are distributed within products for which customers are not presently charged any fee and for which the producer has essentially a zero marginal cost (as with many electronically distributed software products).

Case B: Success of standard B depends on adoption by only a handful of manufacturers who will make the component embodying the standard and the component will then be used in many other products. Assume that all of those component manufacturers already have established licensing arrangements.

Assume in each case that there are several options from which the standard can be selected and consider the implications of waiting to address licensing of essential patents until after the standard has been selected. In case A, waiting to confront the licensing of essential patents until after the standard has been selected could be fatal to the standard's success. In case B, there is far less danger in this respect.

  1. The Increasing Rnle nf Patents

Technologies subject to patent protection are often exceptionally valuable and sometimes essential inputs into IP standards. Many technologies either protected by outstanding patents or subject to pending patent applications may embody innovations holding the potential to create new markets and to benefit both suppliers and consumers across many industries. For these reasons, standards-groups should welcome and encourage the availability of these technologies as inputs into the standards they develop. This, of course, does not mean that acceptance of such contributions should be without some strings attached or without conditions designed to ensure that standards create open and robustly competitive markets. More on that subject in a few minutes.

A related factor is a more general IT-sector-wide shift in emphasis from defensive use of patents and broad cross-licenses toward more aggressive exploitation of patents for revenue generation. Developers of technologies that are either already protected by patents or subject to pending patent applications increasingly see the incorporation of their technologies into proposed specifications as a way of leveraging their patents into positions to extract financial and competitive benefit from widespread adoption of the resulting industiy standards. In this environment, innovators are highly incented to promote the consideration of their inventions in all promising standard-setting venues.

Another factor is the increasing rate of patent grants in recent years. The result is that there may be a multitude of patents shadowing or potentially burdening any particular standard-setting effort. Standard-setting participants usually include companies with large patent portfolios; a company's representative in a standard-setting process may not in fact know whether a proposed standard will or may implicate a patent within his or her company's portfolio. There may be no nefarious intent to hide the ball in this respect. Use of the ultimately adopted standard nonetheless may well require use of one or more of that company's patents. Again, this is a consequence of patent proliferation and an unavoidable part of the climate in which standard-setting occurs.

In considering the tools that might be used for factoring patents into the standards development process, one might focus on the following three categories: disclosure policies; endorsement thresholds; and licensing policies. Each of these tools can be used in a variety of ways, as I will now briefly explain.

Disclosure policies. Different groups employ different approaches to encouraging or requiring timely disclosures of information about essential patents. Variables include (a) scope of knowledge triggering the disclosure (whether, for example, any search or other inquiry beyond the participant's own awareness may be expected); (b) nature of the disclosed information (whether, for example, it includes only issued patents or also pending patent applications); and (c) point or points in time when disclosure is to be made (early in the process, shortly before balloting, or at several different stages).

Endorsement threshold. Many groups use a rule under which the forum will not endorse or publish a standard for which the fomm is aware of an essential patent unless certain conditions are met. Most frequently, the condition is an indication by the patent owner of a willingness to offer licenses. I refer to these rules as endorsement thresholds. Variables include: the licensing commitment that the forum might expect as a condition, extent of the necessary forum "knowledge" and retraction (or not) of previously endorsed standards.

Licensing. In some groups, participants agree to commit to certain licensing terms with respect to essential patents they might hold. Variables include: to whom the licensing obligation extends (for example, whether the obligation is undertaken only by those actively involved in the process or is undertaken by broader classes of interested parties); license terms to which the commitment is made (RAND, free, other); to what patents does the obligation extend (already issued patents or also future patents based on then-pending applications).

Many groups impose on themselves the requirement that they will not select a standard that would require use of a patent that will not be available on reasonable and nondiscriminatory terms. This policy (an endorsement threshold) is widely employed.

  1. Consideration of Patents During Standard-Setting

There is no rational argument in favor of "blissful ignorance" of patent implications during the course of a standard-setting process. The more that is known before a standard is adopted, the better from the standpoint of anticipating and protecting against the post-adoption exercise of market power that a patent may confer if it is essential to the standard's use.

The FTC's Dell Computer action of six years ago called attention to the manner in which anticompetitive "patent hold-up" or "patent ambush" situations can arise when standard-setting bodies go about their business of fashioning and voting upon proposed standards without knowledge that patents may be infringed by the use of the standards they adopt. That action, however, opened a virtual Pandora's Box of follow-on issues over how to address and minimize exposure to post-adoption opportunistic conduct by holders of patents required for a standard's use. In some situations, as alleged in several private lawsuits as well as in reports of now-pending FTC investigations, the problem may arise from deliberate deception during the standards development period. In other cases, however, no premeditation may be involved; it may indeed be that the existence of the patent coverage could not reasonably have been known when the standard was written. Nonetheless in those situations as well, anticompetitive consequences can emerge from the post-adoption discovery and assertion of the patent that the standard encompasses.

There are many ways that patent license terms revealed only after the standard is adopted can generate conflict and impair many parties' ability to compete in the affected market. Permit me to offer several examples of the possibilities in this regard:

  • Patentee seeks a royalty that is "fair" from the patentee's perspective but greater than the average profit margin of all of the parties who will need licenses.

  • Patentee seeks a broad grantback that appears even handed but with significantly disparate effects on different parties, perhaps forcing particular licensees to forfeit the value of their own major innovation investments, but patentee refuses to deviate from its "standard" agreement for any reason.

  • Patentee demands a minimum annual royalty based on "administrative costs" but with the effect of locking out smaller rivals and new entrants.

  • Patentee seeks royalties from downstream providers (e.g., manufacturers of finished goods) and refuses to license suppliers of upstream inputs such as IC vendors. The purpose is to increase the income to the patentee (3% of a computer is more than 3% or even 5% of an IC within the computer), but this practice also greatly increases rivals' costs and time to market.

  • Patentee requires admissions of infringement and validity, and/or retains the right to immediately terminate a license if the licensor challenges infringement or validity.

  • Patentee requires acceptance of venue in a "home court" which might be fine for large companies but a major problem for small companies or foreign competitors.1

There is no single, ideal solution to this problem or combination of problems that would be appropriate for all of the different kinds of standard-setting going on in so many different contexts. There is no neat one-size-fits-all remedy that could be effective across the whole universe. A particular set of disclosure obligations or advance license commitments may be fine for some promoters' groups or particular consortia while being impractical and unacceptable for SDO proceedings involving large numbers of diverse participants. Different groups are now employing or considering different approaches, involving a variety of pre-adoption disclosure and license commitment policies. Everyone concerned about these issues, including your agencies, should welcome this diversity of experience and of experimentation with methods of addressing the problems at hand.

HP has in some circumstances favored an approach that we believe should be encouraged but that is often opposed by others upon what we believe is a misapprehension of antitrust risks. If a party promoting use of its patented technology for incorporation in a proposed standard states it is willing to offer a license on terms that are "reasonable and nondiscriminatory" (but terms that are not otherwise specified), consideration of the impact of the patent on the proposed standard often ends at that point -- indeed some participants insist any further or more specific discussion about it would invite antitrust trouble. But all potentially affected parties have a legitimate interest in knowing before a standard decision is made what the economic effects will be of accepting a patent into the standard. Nonetheless, when suggesting that licensing terms be considered, we have encountered the objection that doing so could invite antitrust challenge. Indeed, some standards organizations expressly prohibit consideration of license terms in their rules.2 These objections are unfounded. To the contrary, disclosures of the sort we have suggested would be procompetitive by foreclosing opportunistic hold-up situations that are all too easy to arise when a patent holder's view of "reasonable" license terms remains secret until after a standard has been adopted.

To be more specific, consider a circumstance where a consortium is developing what is expected to be a critically important standard that defines the infrastmcture for a whole array of next-generation products that all consortium members and other parties as well need to develop to remain in the affected market space. One consortium member promotes specifications based on technology that it has patented and the patent is appropriately disclosed during the process. Other consortium members see technical benefits to those specifications but also recognize alternative approaches that would entail countervailing benefits as well as avoiding any need for users to obtain patent licenses.

The patent holder's proposal might be considered the "best technical solution" but that does not necessarily make it the "best solution" when the overriding objective is a standard that ensures a level playing field and robust competition in the new market that the standard is designed to foster. A standard that enables one user to extract exorbitant royalties from all other users could, in this light, be the "best technical solution" but not by a long shot the "best solution" either for the industry generally or from the broader standpoint of the public interest.

In that scenario, why should not all of the consortium members have the right to ask the patent holder, before any decision is made on which approach to adopt, to specify the royally and other terms the holder would impose if its technology is selected? A truly informed and intelligent decision on which among these alternative approaches would best serve all parties' interests -- including the public's interest in competitive market conditions -- cannot be made without knowing what the patent holder would extract from all users as the price for admission into the affected market.

Now let's assume that that same consortium proposes to go beyond requesting disclosure of the holder's planned license terms and embarks on a concerted negotiation over mutually acceptable terms under which they will adopt the holder's specifications for the new standard. One could characterize that scenario as a form of "joint buying" of an input into a new "product" that the parties are jointly developing. The agencies have blessed joint buying scenarios in many contexts, and this is one where agency approval would be appropriate as well.3 We are worlds away from case law condemning "buyer cartels"; we are talking about collaborations for creating new standards that advance new technologies, and these collaborations fit well within case law applicable to "joint ventures" of many kinds.4

In short, the sort of "joint negotiation" or joint consideration of license terms during a standard-setting process that I have described in my example should be unassailable under the governing antitrust rule of reason. Indeed, as I have already suggested, it may often be the most efficient if not the only practical way of avoiding patent holdup or ambush problems and should be considered presumptively procompetitive on that ground. Limitations on the scope of or other safeguards attaching to this activity can be fashioned to obviate concern over any countervailing anticompetitive risk. The bottom line, in my view, is that it is desirable and in the public interest that standard-setting groups be not only permitted but encouraged to experiment with various mechanisms for consideration of specific license terms before decisions on the content of a standard are engraved in stone. Thoughtful guidance from your agencies on this point should be welcomed by all quarters of the standard-setting community.5

* * * *

Finally, as your agencies review issues presented by standard-setting processes in the information technology space, it is important to be sensitive to the thoroughly international nature of the standards that we are talking about. This characteristic highlights why conflicts or inconsistencies in applicable public policies among jurisdictions could significantly impede the progress of standard-setting groups and thereby slow innovation and technology development generally. For this reason, your agencies could play a valuable role in promoting policy harmonization on a global basis.


TAB C

Consideration of Patents
during the Setting of Standards

Scott K. Peterson
Hewlett-Packard Company

for

FTC and DOJ Roundtable on Morning of November 6, 2002:
Standard Setting Organizations: Evaluating the Anticompetitive Risks Of Negotiating IP
Licensing Terms and Conditions Before A Standard Is Set

The world of standards setting and patents is changing.1 I want to report on what I see, why things are changing, where I think the intersection of standards setting and patents is headed, and why attention should be paid to patents while standards are being set.

There are standards of various types, and not all standards play the same role.2 My perspective is that of a participant in the information technology ("IT") industry. My comments will focus on interoperability standards, which play a vital role in fostering growth in many IT markets.

Cooperate on Standards. Compete on Implementations.

The IT industry produces products that form complex, interconnected and interdependent systems. Interoperability adds critical value to IT products, as is reflected in the network economics that is so characteristic of these systems. Although interoperable products can provide great value for customers, that value may not be realized unless standards exist to foster the availability of a network of related, interoperable products; many innovative products might never have existed without standards. As a result, HP's longstanding approach has been to cooperate on standards and compete on implementations.

Cooperate on standards. The use of common protocols and interfaces can expand markets for networks of products that implement those protocols and interfaces.

Compete on implementations. Producers compete by innovating on top of the standardized functions. Although the functional characteristics specified in the standard may be the same in all products, innovation builds on those points of commonality (such as by implementing those characteristics with innovative techniques and by adding innovative features to products that incorporate the standardized elements).

Formulation of Patent Policy for Development of Standards is an Exercise in Balance

Formulation of a policy for how to address the interplay between patents and standards is not a simple maximization exercise -- it requires balancing among a variety of interests:

  • standards development organizations ("SDOs")
  • participants in the standards development process (companies and individuals)
  • patent owners
  • users of standards (implementers)
  • customers (ultimate consumers)
These interests sometimes have competing concerns and other times share concerns, although often with differing priorities. The following are some examples of the concerns to factor into a patent policy: liability; complexity, speed, and effectiveness of the standardization process; control over use of patent rights (including whether rights are used, and if so what the compensation or other implications might be); burden on participants; competitive implications of resulting solutions, and impact on innovativeness and cost.

Like many companies, HP finds itself in many of these roles (patent owner, implementer, participant, consumer). HP invests in technology innovation and develops a large patent portfolio that reflects this investment. HP is a very large user of standards; it builds products that operate according to a multitude of standards, including products that will be successful only if there is market-wide acceptance of the standards to which they conform. HP's own operations rely on many products and services that depend on standards. Given this degree of importance to HP's business, HP has long been an active participant in the process of developing standards and has a strong interest in patent policies that foster success of standards.

Diversity in Standards Development Forums

Standards in the IT industry are developed in a wide range of types of forums. This range can be illustrated by considering three regions of the range: promoters groups, consortia, and SDOs. At the SDO end are the long-lived more formal forums. At the other end are the short-lived promoters groups formed to launch initial versions of one or a small family of standards. The SDOs excel in breadth of participation; the promoters groups often act more quickly. An intermediate balance in this speed/participation trade-off is found in consortia.

This variety of forums is valuable. Marketplace dynamics may call for particularly expedited processes in some instances but may value longer, more deliberate incubation periods in other instances. Some standards may affect the competitive opportunities of more classes of parties and in more fundamental ways than will be the case with some other standards, thus calling for different kinds or degrees of participation. Some standards may be based on emerging technologies that entail joint development of the technology in question, thus looking more like what could be called a "joint production venture" than a traditional standard-setting process, or a hybrid of both kinds of undertakings, and thus calling for different degrees of tolerance for development efforts driven by one or a small number of firms.

There Should be Variety in Patent Policies

The variety in characteristics of technologies, markets, business situations, and standards development forums should lead to variety in patent policies.

There is a wide range of the extent to which technical solutions vary in "performance". This variation is sometimes due to variation in the nature of the constraints in the problem area. This affects the likelihood that a patented solution will offer significant advantage over alternatives. Often a protocol can be implemented in many ways that have similar performance; on the other hand, for example, where the degree of data compression is critical, it may be important to use one of the more well-refined algorithms.

The complexity and other implications of licensing vary depending on the industry and on product characteristics. For example, consider the differences between the situation where only a small number of companies need to implement the standard3 and the situation where success of the standard depends on implementation by a very large number of different implementers. Also, the state of existing licensing or cross-licensing arrangements among the expected implementers (keeping in mind the possibility of new entrants) can also affect the degree of benefit in considering the licensing implications as a part of the standard setting process.

The following contrasting examples illustrate the need for different types of standard-setting rules, and the need to pay close attention to patents in some cases but less so in other cases:

Case A: Success of standard A depends on its implementation in products from hundreds of different sources; some of these are products that are distributed without charging any fee and for which the producer has essentially a zero marginal cost (as with many electronically distributed software products).

Case B: Success of standard B depends on adoption by only a handful of manufacturers who will make the component embodying the standard (a component that will then be used in other products), where all of the component manufacturers already have established licensing arrangements; the standard is implemented in products with significant unit manufacturing cost.

In Case A, the impact that any royalties would have on product economics is more dramatic; in addition, the transaction costs of licensing are significantly higher than in Case B. This suggests a greater value in attending to patent issues in Case A than in Case B.

Essential Patents

In talking about patents and standards, the focus is on what have come to be known as "essential patents", since they have the most direct impact on standards. In general terms, essential patents are those to which one will need a license in order to implement the standard in products.4 These essential patents are distinguished from other patents that, although possibly useful in implementing the standard, are not patents to which one must have a license, such as because of the availability of alternative ways of implementing the functions specified in the standard.5

Because a license to an essential patent is (by definition) essential to practicing the standard, that license can have significant impact on the competitive viability of participation of many parties who may seek to participate in markets for products that depend on the standard (flowing from timing, cost, and other factors associated with obtaining the necessary license).

Tools for Considering Patents in Development of Standards

How might a standards development forum "take patents into consideration" when developing a standard? I find it useful to think of the tools (the rules that a standards development forum might use) for such consideration as falling into several categories:

  • endorsement thresholds
  • disclosure of essential patents
  • licensing commitments
  • disclosure of license terms
Endorsement Threshold. Many forums use a rule under which the forum will not endorse or publish a standard for which the forum is aware of an essential patent unless certain conditions are met. Most frequently, the condition is an indication by the patent owner of a willingness to offer licenses on reasonable and non-discriminatory ("RAND") terms. I refer to such rules as endorsement thresholds. Variables include: the licensing commitment that the forum might expect as a condition; and retraction (or not) of previously endorsed standards when an essential patent is identified after adoption of the standard. Note that, in the case of an endorsement threshold, the obligation or requirement is imposed on the forum, not on the patent owner; the patent owner has no obligation to make a licensing commitment.

Disclosure of Essential Patents. Various approaches have been taken to encouraging or requiring timely identification of essential patents. Variables include: scope of knowledge triggering the disclosure (whether, for example, any search or other inquiry beyond the individual participant's own awareness may be expected -- most commonly not requiring a search); nature of the disclosed information (whether, for example, it includes only issued patents or also pending patent applications, often making a distinction between published and unpublished applications); and point or points in time when disclosure is to be made (e.g., early in the process, shortly before balloting, or at several different stages).

Licensing Commitments. In some forums, participants agree to commit to certain licensing terms with respect to essential patents they might hold. This tool is commonly used in promoters groups, and is used less widely as one moves toward the SDO end of the range of types of standards development forums. Variables include: who undertakes the licensing obligation (for example, whether the obligation is undertaken only by those participating in the development of the particular standard, or is undertaken by broader classes of interested parties); license terms to which the commitment is made (is a fee permitted? are particular terms specified? or is this a more generalized commitment?); to what patents does the obligation extend (already issued patents, future patents based on then-pending applications).

Disclosure of License Terms. Rather than expecting commitment to particular license terms, a forum might encourage or require that the terms under which essential patents will be licensed are disclosed. Variables include: the level of detail of the terms to be disclosed; what "required" means (such as a condition of participation in development of the standard, or as a condition of the forum's endorsement of the standard). In factoring into the consideration of a proposed standard the implications of the required license, it may be appropriate to have an iterative process where license terms are disclosed, implications considered, and revised terms disclosed and considered; in its most extended form this could amount to a kind of license negotiation, as discussed further below.

None of these tools represents "the solution". The goal is one of balance, where different forums may strike different balances. To achieve its particular balance of interests, a forum's patent policy will typically configure two or more of these tools and rely on that combination, rather than rely on one tool.

License Terms Categorized

In considering the implications that a license to an essential patent may have on a standard, I have found it useful to group license terms into the following categories:

(A) scope of the grant: essential claims; for the purpose of implementing the standard.

(B) reciprocal license: not just any grantback, but the same license as the licensor is offering (such as limited to essential claims), possibly including a commitment to offer licenses to others.

(C) administrative terms: customary, non-substantive terms such as choice of law.

(D) fees.

(E) other compensation: beyond any fees and the reciprocal license listed above, such as a broader grantback or requiring use of licensor's products or services.

(F) control terms: licensor-defined requirements imposed on implementations, such as technical requirements (beyond that specified in the standard), or requirements on how implementations are to be licensed or distributed.

(G) patent non-assert conditions.

(H) anything else.

The first three categories are to be expected in most cases.

Fees may or may not be required, depending on the situation. The extent to which standards developers should seek to find free solutions is context-dependent, and is a significant matter of debate in some forums.

In general, the last four categories should be empty. Terms in these categories should be treated with skepticism and given careful scrutiny, based on the full context of the particular standard. Because of the tight connection between a standard and the license for any patents essential to it, control terms, non-fee compensation, and patent non-assert provisions are particularly susceptible to being anticompetitive. As the other categories are generally adequate to do the job, anything in the catchall category at the end should be carefully considered.

Ex Ante Consideration of License Terms

The time at which license terms for an essential patent are determined can have very significant implications, with the most dramatic difference being between the situation before ("ex ante") and the situation after ("expost") the standard is set.

Traditionally, determination of license terms has been left until after the standard was set, when implementers were beginning to build products that met the standard. Typically, a statement by the patent owner of a willingness to offer licenses on RAND terms was all that was expected of the patent holder prior to the setting of the standard.

Yet, a commitment to offer a license on terms that are merely specified as "RAND" is not an adequate safeguard against abusive use of a patent that has become essential to a standard. The value of a patent is so dependent on the specific case (including both the particulars of the invention and of the context in which it is to be used) that different parties could have wildly different views of what RAND should mean in the context of any given standard or market environment. In particular, what a patent holder considers to be a "reasonable" royalty rate will inevitably be considerably higher than what licensees believe such a rate to be, particularly at the ex post stage when the patent holder has the added leverage flowing from the lock-in effect of the industry adoption of the standard, referred to below as 'Y'.

As a result, I have long advocated increased attention to license terms before a standard is set. Such ex ante consideration of license terms (1) facilitates factoring the patent license implications into the decision on selection of the best standard, and (2) reduces the risk of anticompetitive implications of selecting a standard that requires a license for which the terms have yet to be set.6 Moving issues to ex ante consideration should operate to reduce the extent to which litigation is needed to resolve issues relating to patents and standards: ex ante, there are more options (both in technical decisions and in other business options for resolving the issues); ex post, more has become locked-in.

Some participants in standards development activities have refused to permit license terms to be taken into consideration in the selection of a standard because of a concern about antitrust risks. However, this reflects what I believe to be a misapprehension of such risk. In fact, ex ante consideration of license terms is fundamentally procompetitive, as it forecloses opportunistic holdup situations that are all too easy to arise when a patent holder's view of "reasonable" license terms remains secret until after a standard has been adopted. Clarification of this risk issue could facilitate use of ex ante consideration to help make better choices in selecting standards and to reduce the potential for ex post adverse impact of essential patents.

Antitrust concerns should not inhibit experimentation with various means of considering patent license terms during a standard-setting process, including multilateral negotiation processes, because risks are outweighed by distinctly procompetitive benefits of the sort that render such activity eminently defensible under the antitrust laws.

Risks

Some have expressed the concern that collective ex ante consideration of license terms could be challenged as "buyer price-fixing" or a "buyers' cartel." This concern is unfounded. Price-fixing and cartel appellations are reserved for naked kinds of collusive activity devoid of any valid business justification.

Collective consideration of terms for licenses (i) for use in implementing a standard that the parties are jointly producing (ii) of patents that would be essential to that contemplated standard would be an activity intimately connected to and supportive of the participants' entirely lawful undertaking to fashion a standard designed to grow a market that will be open to all comers on a level playing field.7 Such license negotiation activity is the same in purpose and effect as various kinds of negotiation activity the parties to a joint production or technology development venture engage in when acquiring an input into the new product or technology they are creating. Antitrust's permissive rule of reason applies to these kinds of situations.8

Even under the rule of reason, concerted buying activity can give rise to "oligopsony" concerns. A necessary condition to anticompetitive oligopsony effects, however, is a reduction in the quantity of purchases, thereby reducing upstream output.9 Collective ex ante consideration of license terms in the standard-setting context may well result in royalties that are lower than what they would be if unilaterally specified at the ex post stage, but there will be no reduction in the quantity of licenses sought and issued. Indeed, lower royalties can only increase the number of applicants and ultimate competitors in the affected product markets, an output-enhancing effect.

Could collective negotiation reduce "just rewards" from a patent holder's innovation investment and thereby anticompetitively reduce innovation incentives? This would appear highly implausible under any proper definition of a patent holder's just rewards. One could think of X as representing the value or leverage that the patent provides in a competitive environment when routes other than adoption of the standard to which that patent would be essential remain open. However, once there has been agreement on selection of a particular standard that requires use of that patent, the value or leverage that the patent provides is increased to X+Y. X is related to the inventive contribution of the patent. Y is related to the collective action of the standards setting activity. As the value of a patent is highly context dependent, X and Y are not readily susceptible of determination in the abstract. Instead, the value is best determined by conducting negotiation at the appropriate stage. Ex ante consideration of license terms is most likely to approximate X, while ex post consideration is most likely to result in X+Y.

One anticompetitive risk that should be considered is the possibility of "spillover" effects -- exchange of information on or discussions about impermissible matters such as the parties' expected product prices, product development plans, or marketing strategies. Those developing standards already have extensive experience managing this risk.

Benefits

The fundamental and entirely procompetitive business justification supporting ex ante consideration of license terms is protection of all participants -- indeed nonparticipating but future new entrants in affected markets as well -- from ex post holdup by the party whose patent becomes essential to implement the standard. Absent ex ante agreement on terms, the standard-setting process may artificially hand the patent holder market or monopoly power that the patent by itself would not confer, enabling the patent owner to impose royalties that may bear little relation to the value of the patented aspect, and impose other anticompetitive conditions such as overbroad grantbacks and non-assert provisions. Excessive license royalties and other terms reflecting ex post artificial market power raise barriers to entry into affected markets, raise product costs and thereby raise product prices.

Anticompetitive effects may well be exacerbated by broad cross-licenses between major incumbents and the patent holder; in that event, the competitive disadvantage of supracompetitive royalties could be overwhelming to smaller rivals and could block new entry altogether. Conversely, ex ante consideration in connection with an open contest between competing solutions for the proposed standard can be expected to generate competitive license terms that enable new entry and maximally open competition in affected markets, avoiding supracompetitive product prices and creating an environment in which innovation rivalry can flourish.

In short, avoidance of the exercise of the "Y" leverage discussed above is a procompetitive benefit that can be achieved by ex ante consideration of license terms.

Indeed, one could argue that standards setting participants act irresponsibly in fashioning standards that incorporate patented technology without taking account in the course of the standard-setting process itself of the potential for anticompetitive effects if the patent holder subsequently imposes anticompetitive terms. The Supreme Court's antitrust jurisprudence on standard-setting activities has consistently admonished SDOs and SDO participants to adopt safeguards against misuse of a standard by any participant that will be a position to employ it in an anticompetitive manner.10 For all of the reasons already mentioned, ex ante consideration of license terms qualifies as precisely the kind of safeguard envisioned and encouraged by those decisions.

Process

There will be challenges in incorporating greater consideration of patent issues into processes for developing standards. For example, consideration of patent issues requires expertise that is not part of the background of those who are typically most directly involved in the standards setting activities. I see these process challenges as being attacked incrementally -- starting with those situations where the patent issues are likely to have the greatest impact, and then moving forward as appropriate.

I have seen the creation of a diverse array of standard development forums to meet the specific needs of particular standards situations. In a sense, there is an active, competitive market among standards setting forums, within which the developers of standards have demonstrated an ability to innovate in the processes for accomplishing their goals. From this experience, I am confident that standards developers will meet the challenges associated with factoring patent considerations more fully into their processes.

Hot Topic. Why now?

Standards and patents -- a hot topic these days. Why now?

Recent years have seen heightened recognition of the value of intellectual property. This is reflected in more aggressive enforcement of patent rights. For example, there is a broader, more intense interest in patent licensing as a direct revenue opportunity, and there is greater attention being paid to the competitive advantages that can be obtained from patent rights. Thus, when competitors sit down to cooperate on development of standards, it is understood that there is more at stake than might have been the case in the past. This has been particularly evident in the IT industry. With increased opportunistic use of patent rights to obtain competitive advantage, it is becoming less prudent to rely on expectations of ex post cooperative licensing.

Guidelines?

Given the widely acknowledged need for flexibility to handle the great variety of standards setting situations, prescriptive guidelines seem undesirable.

However, the agencies could offer useful guidance in their final report on the hearings record on how they see patent-related behavior to relate to antitrust risk. I see particular value in agency guidance on the both risks and benefits associated with ex ante consideration of essential patents, as mentioned above. Such guidance could remove the impediment that uncertainty about the risk creates -- an impediment to behavior that is actually procompetitive.

View of the Future

I anticipate a future in which greater attention will be paid to the interplay between standards and patents that are essential to those standards. This will include a greater interest in understanding the implications that patents hold for success of standards to which they are essential, and therefore a greater interest in factoring patent licensing considerations into the judgments that lead to selection of standards. Less frequently will statements limited to a willingness to offer a license on unspecified "reasonable and non-discriminatory" terms be the end of the consideration.

The extent of this attention will vary from standard to standard. The likelihood that essential patents will have significant impact on the success of any particular standard depends on a variety of factors. Many standards simply do not implicate patents. In some cases, those developing the standard might conclude that, although patented technologies may be involved, established licensing practices among the anticipated implementers are such that the standards developers can afford to devote a relatively low level of attention to these issues. But, in yet other cases, the terms for licenses to essential patents can have dramatic impact on success of a standard, and the standards developers may choose to pay careful attention to the implications of required patent licenses.

This will not eliminate all ex post patent issues; there is no effective way to identify all essential patents in advance. But this additional advance planning -- thoughtful, timely consideration of the important issue of patent licensing -- will make a positive contribution to the success of standards. This will be good for standards and good for the procompetitive, market-enlarging benefits that standards can offer.

 


FOOTNOTES
TAB A

1. The IEEE's list of 802.11 patent disclosures is available at http://standards.ieee.org/db/patents/pat802 11 .html (reviewed June 11, 2005).

2. Institute of Electrical and Electronics Engineers, Inc., IEEE-SA Standards Board By-Laws, Section 6, available at http://standards.ieee.org/guides/bvlaws/sect6-7.htmltf6 (reviewed June 11,2005). See also JEDEC Solid State Technology Association, JEDEC Manual of Organization and Procedure §8.2 (July 2002), available at http://www.iedec.org/Home/manuals/JM21 L.pdf (reviewed July 11, 2005); Internet Engineering Task Force, Intellectual Property Rights in IETF Technology §6.5 (March 2005), available at http://www.ietf.org/rfc/rfc3979.txt (reviewed July 11,2005).

3. Institute of Electrical and Electronics Engineers, Inc., Understanding Patent Issues During IEEE Standards Development, available at http://standards.ieee.org/board/pat/guide.html (reviewed June 11,2005) (emphasis in original).

4. The liability of a standards organization for the anticompetitive acts of its members was established by the Supreme Court in American Soc y of Mechanical Engineers v. Hydrolevel Corp., 456 U.S. 556 (1982).

5. Institute of Electrical and Electronics Engineers, Inc., Guide to IEEE Standards Meeting Policies, Guideline 10, available at http://standards.ieee.org/resources/meetingguide.html (reviewed July 11,2005) (emphasis supplied).

6. Competition and Intellectual Property in the US: Licensing Freedom and the Limits of Antitrust at 9, delivered by R. Hewitt Pate to the 2005 EU Competition Workshop, Florence, Italy (June 3,2005), available at http://www. usdoj. gov/atr/public/speeches/209359.pdf (reviewed June 14, 2005).

7. Testimony of Amy A. Marasco, Vice President and General Counsel, American National Standards Institute (April 18,2002) at 7. Testimony available at http://www.ftc.gov/opp/intellec1/020418marasco.pdf (reviewed July 11, 2005).

8. Pate speech, cited supra n.6, at 9.

9. There has been industry commentary suggesting that the stacking problem is hindering the introduction of 3G wireless networks based on the WCDMA standard. See Eric Ransdell, What a Tangled Web the World oflPR Can Weave, June 26,2002, available at http://www.thefeature.com/article?articleid=15266&ref=7724316 (reviewed June 12, 2005) ("With nobody to verify who owns which patents and 40 companies currently claiming IPR that is essential to 3G, there is real potential for disaster. Many companies are already asking for a 1 percent royalty. With more than 100 companies potentially entering the fray, the IPR cost element for 3G products could reach 100 percent.")

10. Juicy Whip, Inc. v. Orange Bang, Inc., 382 F. 3d 1367 (Fed. Cir. 2004). See also Rite-Hite Corp. v. Kelley Co., 56 F.3d 1538, 1550 (Fed. Cir. 1995) {en bane).

11. See Robert A. Skitol, Concerted Buying Power: Its Potential For Addressing The Patent Holdup Problem In Standard Setting, 72 ANTITRUST L.J. 727 at 727 fii. 2 (2005) (collecting agency guidelines and business review letters).

12. See generally Gail F. Levine, B2Bs, E-Commerce and the All-Or-Nothing Deal, 28 RUTGERS COMPUTER TECH. L.J. 383 (2002).

13. E.g., Allied Tube & Conduit Corp. v. Indian Head, Inc., 486 U.S. 492 (1988); Radiant Burners, Inc. v. Peoples Gas Light & Coke Co., 364 U.S. 656 (1961); Addamax Corp. v. Open Software Foundation, 152 F.3d 48 (1st Or. 1998); Sony Electronics Inc. v. Soundview Technologies, Inc.. 157 F. Supp. 180 (D. Conn. 2001); Golden Bridge Technology v. Nokia, Inc., Civil Action No. 2:05CV170LED (E.D. Tex., filed May 6, 2005).

14. We note that the European Commission's Competition Directorate has already taken this step. Paragraph 225 of the Guidelines accompanying the EC's 2004 Technology Transfer Block Exemption Regulation states that:

225. Undertakings setting up a technology pool that is compatible with Article 81, and any industry standard that it may support, are normally free to negotiate and fix royalties for the technology package and each technology's share of the royalties either before or after the standard is set. Such agreement is inherent in the establishment of the standard or pool and cannot in itself be considered restrictive of competition and may in certain circumstances lead to more efficient outcomes. In certain circumstances it may be more efficient if the royalties are agreed before the standard is chosen and not after the standard is decided upon, to avoid that the choice of the standard confers a significant degree of market power on one or more essential technologies. On the other hand, licensees must remain free to determine the price of products produced under the licence. Where the selection of technologies to be included in the pool is carried out by an independent expert this may further competition between available technological solutions.

European Commission, Guidelines on the Application of Article 81 of the EC Treaty to Technology Transfer Agreements 1225,0.J. C 101/2 2004 (April 27,2004) (emphasis supplied), available at http://europa.eu.int/eur-lex/pri/en/oi/dat/2004/c 101/c 10120040427en00020042.pdf(reviewed June 14,2004).

15. E.g., National Society of Professional Engineers v. US, 435 U.S. 679,692 (1978) ("In this case we are presented with an agreement among competitors to refuse to discuss prices with potential customers until after negotiations have resulted in the initial selection of an engineer. While this is not price fixing as such, no elaborate industry analysis is required to demonstrate the anticompetitive character of such an agreement. It operates as an absolute ban on competitive bidding "); American Medical Ass'n, 94 F.T.C. 701 (1979), affdas modified, 638 F.2d 443 (2d Cir. 1980), affdper curiam by an equally divided Court, 455 U.S. 676 (1982).

16. Pate Speech, supra n.6, at 9.


FOOTNOTES
TAB B

1. For an example now in litigation, see Intersil C.orp v Proxim, Tnc , Civil Action No. 01-266 (U.S.D.C. Del., Complaint Filed April 24, 2001). The allegations are that (a) during an IEEE proceeding to develop a standard for a wireless LAN communications protocol, Proxim provided the requisite letter committing to license its patents (if needed for use of the standard) on reasonable terms; (b) after the standard was adopted in reliance on Proxim's commitment, Proxim filed infringement suits against numerous users and simultaneously sent letters to them with what the complaint describes as a "sham overture to negotiate license terms"; (c) recipients were required to sign an NDA forbidding disclosure of the proposed license terms to anyone else in the industry and were also required to admit infringement of Proxim's patents; and (d) recipients were given 30 days within which to accept the proffered terms and told that those failing to do so by that time would face "lengthy, complex and expensive litigation relating to the infringement" of the patents. The complaint also alleges that Proxim's intent was to disrupt implementation of the IEEE standard in order to benefit a competing protocol based on Proxim-supplied proprietary technology.

2. See, e.g., "Understanding Patent Issues During IEEE Standards Development" at http://standards.ieee.org/board/pat/guide.html: "So what can you discuss about patents at a standards-development meeting? You can cover the content of the patent letter of assurance form, you can discuss the technical merits of using the technology under patent, and you can discuss the way patent information is made available from the IEEE. You must not discuss subjects like the pricing for use of a patent, how a patent should be licensed, validity or interpretation of a patent claim, or any terms or conditions of use. These are not appropriate topics for discussion in a standards developing committee." See. also "What You Need to Know About IEEE Standards and the Law" at http://standards.ieee.org/resources/StdsLaw_Brochure.pdf: "During standards-development meetings, discussions should be confined to technical, engineering, and safety considerations. Commercial considerations are not proper factors for consideration."

3. See, e.g., FTC/DOJ "Antitrust Guidelines for Collaborations Among Competitors" at 14 (April 2000), recognizing that many "agreements jointly to purchase necessary inputs ... do not raise antitrust concerns and indeed maybe procompetitive." The guidelines also recognize that there are situations where such agreements may create or facilitate the exercise of buyers' market power or threaten collusion among the participants. These concerns, however, are quite unlikely to arise from the mere consideration of-- or indeed even negotiation over -- proposed license terms for a patent on technology that may be written into proposed specifications. Agreement on what to include in the standard is a necessary part of every standard-setting process; and the fundamental purpose of considering license terms at an early stage of the process of reaching that agreement is to protect against the patent holder's exercise of market power that it may obtain as a direct result of the standard itself against all other users of the standard that otherwise would be highly vulnerable to opportunistic conduct. There is no reason to assume or expect collaboration about such license terms to spill over into unlawful forms of collusive activity and, in any event, procedures can be fashioned to obviate any concern of that sort.

4. See; e.g., Adriamax Corp. v Open Software Foundation, 152 F.3d 48 (1st Cir. 1998), affirming dismissal of an antitrust challenge to OSF, a consortium to develop specifications for a new industry-standard UNIX platform. The Court of Appeals saw OSF as a "venture . . . producing a new product" with "potential for a productive contribution to the economy . . . ." Id. at 52.

5. Your agencies' guidance in this regard could be particularly valuable as applied to situations involving government-mandated industry standard setting such as that envisioned by the proposed "Consumer Broadband and Digital Television Promotion Act" (S. 2048, 107th Congress, 2d Session, introduced March 21, 2002). That Act would (a) require the information technology and digital content industries to reach agreement on "security system standards for use in digital media devices" (PCs, TVs, etc.) within one year of enactment; and (b) thereafter prohibit the sale of any digital media device unless it "includes and utilizes standard security technologies that adhere to" the adopted security system standards. The one-year deadline would virtually compel development of these standards around already existing technologies that are highly likely to be protected by patents. Given the proposed mandate that all device vendors comply with the standards and without any advance understanding on patent license terms, the patent holders would wield extraordinary power to impose terms and conditions as they see fit. The result could be exceptionally anticompetitive, enabling the patent holders to raise rivals' costs beyond the point of meaningful competitive viability and raise entry barriers beyond penetration.


FOOTNOTES
TAB C

1. Although in this paper I focus on patents, in the broader topic of intellectual property and standards setting, change is not limited to patents. For example, there are issues brewing concerning copyright in relation to those standards where literal code from the standard needs to be incorporated into implementations of the standard, such as is the case with some XML-based standards.

2. One very different topic is that of safety standards.

3. For example, some standards are implemented in components that are then incorporated into other products so that, although the standard is incorporated into products from many different companies, the components with the standardized features may be supplied by a small group of implementers.

4. Actually, the focus should be at a slightly finer level of granularity: essential claims, rather than patents. Also, note that determination of "essentiality" to a standard can involve a variety of nuances beyond the discussion here.

5. The concept of essential patents is also important in discussions of standards-related patent pools.

6. See, for example, my statement submitted for the FTC/DoJ hearing on April 18, 2002, which is published at <http://www.ftc.gov/opp/intellect/020418scottkpeterson.pdf>.

7. Note that, because the licensing considered here is for implementation of the standard, licensing of the same patents for other purposes can remain left to individual bi-lateral negotiations.

8. See, e.g., Addamax Corp. v. Open Software Foundation, 152 F.3d 48 (Is' Cir. 1988), affirming dismissal of an antitrust challenge to OSF, a consortium to develop a new industry-standard operating system platform for which the parties negotiated terms for the purchase of a security software input with competing suppliers. The Court of Appeals saw OSF as a "venture .. . producing a new product" with "potential for a productive contribution to the economy . . .." Id. at 52.

9. See G. Levine, "B2Bs, E-Commerce & The All-Or-Nothing Deal," 28 Rutgers Computer & Technology Law Journal 305 (2002).

10. See Allied Tube & Conduit Corp. v. Indian Head, Inc., 486 U.S. 492, 500, 509 (1988); American Society of Mechanical Engineers v. Hydrolevel Corp., 456 U.S. 556, 571 (1982); Radiant Burner, Inc v. Peoples Gas Light & Coke Co., 364 U.S. 656(1961).