This document is available in three formats: this web page (for browsing content), PDF (comparable to original document formatting), and WordPerfect.

1

1UNITED STATES FEDERAL TRADE COMMISSION

2and

3UNITED STATES DEPARTMENT OF JUSTICE

4

5

6

7SHERMAN ACT SECTION 2 JOINT HEARING

8BUSINESS TESTIMONY

9TUESDAY, JANUARY 30, 2007

10

11

12

13

14HELD AT:

15UNIVERSITY OF CALIFORNIA AT BERKELEY

162220 PIEDMONT AVENUE

17WELLS FARGO ROOM

18BERKELEY, CALIFORNIA

199:30 A.M. TO 4:35 P.M.

20

21

22

23

24Reported and transcribed by:

25Kathleen Carr Meheen, CSR 8748

2

1MODERATORS

2Morning Session:

3WILLIAM E. COHEN

4Deputy General Counsel for Policy Studies

5Federal Trade Commission

6and

7JOSEPH J. MATELIS

8Attorney, Legal Policy Section

9Antitrust Division, U.S. Department of Justice

10

11PANELISTS

12Morning Session:

13Michael D. Hartogs

14David A. Heiner

15Scott K. Peterson

16Robert A. Skitol

17

18

19

20

21

22

23

24

25

3

1MODERATORS

2Afternoon Session:

3KAREN GRIMM

4Assistant General Counsel for Policy Studies

5Federal Trade Commission

6and

7JOSEPH J. MATELIS

8Attorney, Legal Policy Section

9Antitrust Division, U.S. Department of Justice

10

11PANELISTS

12Afternoon Session:

13David A. Dull

14Michael E. Haglund

15Thomas M. McCoy

16

17

18

19

20

21

22

23

24

25

4

1C O N T E N T S

2MORNING SESSION (BUSINESS):

3Introduction

4Presentations:

5   David A. Heiner

6   Scott K. Peterson

7   Robert A. Skitol

8   Michael D. Hartogs

9Moderated Discussion

10Lunch Recess

11

12AFTERNOON SESSION (BUSINESS):

13Introduction

14Presentations:

15   Thomas McCoy

16   Michael E. Haglund

17   David A. Dull

18Moderated Discussion

19

20

21

22

23

24

25

5

1P R O C E E D I N G S

2* * * * *

3MR. COHEN: Good morning. I'm Bill Cohen,

4Deputy General Counsel for Policy Studies at the Federal

5Trade Commission. I'm going to be one of the moderators

6at this session. My co-moderator, who is sitting next to

7me, is Joe Matelis, an attorney in the Legal Policy

8Section of the Antitrust Division of the U.S. Department

9of Justice.

10Before we start I need to make a few

11housekeeping announcements. As a courtesy to our

12speakers, we'll urge you all to be sure that you've turned

13off your cell phones, Blackberries, and any other devices

14that might ring, vibrate, play music or anything like

15that.

16The other point that I need to make is that

17these panels are being run as hearings involving the

18moderators and the participants. So, consequently, we

19request that the audience not make comments or ask

20questions during the sessions. Thank you on that.

21Before introducing our speakers, what I'd like

22to do is first thank the University of California at

23Berkeley for hosting the FTC/DOJ Section 2 hearings on

24business testimony. And in particular I'd like to thank

25Howard Shelanski and his colleagues, Richard Gilbert and

6

1Paul Shapiro, for offering us their facilities and for

2making the necessary arrangements for these hearings to go

3forward.

4I'd also like to thank the Competition and

5Policy Center, the Berkeley Center for Law and Technology,

6and the Haas Business School, for providing the

7facilities, refreshments, videotaping, and webcasting

8capabilities, and for working with the agency staffs to

9provide other logistical support. Arranging hearings like

10this takes quite a bit of that and we thank you.

11Others who provided tremendous help with the

12additional details include Bob Barde, Louise Reed, and

13Dana Lund in the audiovisual crew. Our thanks to them as

14well.

15Finally I would like to thank the FTC and the

16DOJ Section 2 team members. And within the FTC

17delegation, Pat Schultheiss and Jim Taronji in particular,

18who I know have worked very hard to put together these

19sessions and all the other sessions that we've held to

20date, and the FTC's San Francisco Regional Office for

21their help and support on this occasion.

22We're honored to have assembled the various

23members of the panel from a number of companies that have

24agreed to offer their testimony in connection with the

25hearing sessions. These panelists have broad perspectives

7

1on how the companies operate within the complex and

2globally diverse realm of Section 2 jurisprudence. We

3anticipate that they will help us to identify and better

4understand areas where single-firm conduct may cause

5competitive harm, areas where desirable, procompetitive

6behavior may be being chilled, and areas where additional

7antitrust guidance would be useful.

8Our panelists, and I'll name them in the order

9that they'll be speaking this morning, are David Heiner,

10who is the Vice President and the Deputy General Counsel

11for Antitrust at Microsoft Corporation; Scott Peterson,

12who is Senior Counsel at Hewlett-Packard Company; Robert

13Skitol, who is the Senior Partner in the Antitrust

14Practice Group at Drinker Biddle & Reath in Washington,

15D.C. and counsel to the VMEbus International Trade

16Association; and Michael Hartogs, who is the Senior Vice

17President and Division Counsel at QUALCOMM Technology

18Licensing.

19Detailed bios for all of our speakers are in a

20packet on the table in the back of the room, as well as on

21the agencies' websites.

22 As to format for this morning, what we're going

23to do is we're going to allow each speaker some time,

24about twenty to thirty minutes if they wish, for a

25presentation. Then after all the presentations are

8

1finished, we'll likely take a break for around fifteen

2minutes. After the break, we'll reconvene for a moderated

3discussion with our panelists.

4The sessions today are an extremely important

5component of the Section 2 hearings overall. FTC Chairman

6Deborah Majora made it clear at the opening session that

7she hoped to learn from the presentations of businesses

8through testimony of their executives and their advisers.

9As Chairman Majoras noted, "The hearings will

10that have panels that will focus on specific types of

11conduct that at least to date, can implicate liability. We want

12the panels to discuss the conduct from the market perspective

13from the ground up, that is, examine why and when firms

14engage in it, how they do it, and what effects it produces

15for the firm, for other firms (customers and competitors),

16and for consumers. We should look at whether firms in

17competitive markets engage in the same conduct and, if so,

18examine why they do it. We want these discussions, to the

19extent possible, to include knowledgeable business people

20or at least their advisers."

21Well, I think over the last seven months or so,

22we have held conduct specific hearings on predatory

23pricing, refusals to deal, tying, exclusive dealing,

24bundled and loyalty discounts, and misleading and

25deceptive conduct. Some of these panels include business

9

1executives or their legal advisers. Today we're going to

2have them talk.

3The sessions will bring together a number of

4panelists who are able to speak with a business

5perspective, in keeping with our goal of obtaining as much

6practical insight and real world experience as possible.

7We look forward to our panelists' remarks and a

8round-table discussion

9I want to thank all of today's panelists for

10their participation. We appreciate it. It takes a great

11deal of time to prepare for and participate in hearings

12like this. And we know that you're all extremely busy

13individuals. So, again, thank you for your time and your

14efforts.

15What I'd like now to do is to turn this over to

16my DOJ co-moderator, Joe Metalis, for any remarks he'd

17like to add.

18MR. MATELIS: Thanks, Bill. The Department of

19Justice's Antitrust Division is extremely pleased to

20participate in these hearings. In the single-firm conduct

21hearings we have held to date, we have benefitted from the

22insights of many highly skilled antitrust attorneys and

23economists.

24Today's hearings, and the hearings to be held

25next month in Chicago, grow out of the belief that we can

10

1also learn much about single-firm conduct from the

2perspective of businesses themselves. Our panelists today

3are people who must help devise and implement business

4plans, aware that their firm's unilateral conduct may be

5challenged in private or government litigation or by

6foreign competition authorities. Their companies are also

7directly affected by the conduct of other firms.

8Whether you have had occasion to view Section 2

9of the Sherman Act as a sword directed at the heart of

10your business or as a shield protecting you from

11anticompetitive conduct, we look forward to hearing from

12you and about your perspectives today.

13On behalf of the Antitrust Division, I would

14like to take this opportunity to thank the Berkeley Center

15for Law and Technology and the Competition Policy Center

16at the University of California Berkeley for hosting these

17hearings today.

18And I'd also like to thank on behalf of the

19Antitrust Division all of our panelists. I know it takes

20a lot of time and thought to prepare for these and we're

21truly appreciative of your efforts to improve our efforts

22of protecting consumers.

23Finally, I'd like to thank Bill and his

24colleagues at the FTC for all of their hard work in

25organizing today's hearing and assembling the fine

11

1

2MR. COHEN: Our first speaker this morning will

3be David Heiner, who I just mentioned is the Vice

4President/Deputy General Counsel for antitrust at

5Microsoft Corporation. Mr. Heiner is responsible for

6antitrust counseling and representation of the company

7before antitrust agencies and compliance with agency

8rulings.

9Since joining Microsoft in 1994, Mr. Heiner has

10played a leading role in Microsoft's response to

11government antitrust proceedings in the United States,

12Europe and Asia.

13Mr. Heiner is a graduate of Cornell University,

14with a bachelor's degree in physics, and a graduate of the

15University of Michigan Law School. He's the author of a

162005 article, "Assessing Tying Claims in the Context of

17Software Integration: A suggested framework for Applying

18the Rule of Reason Analysis."

19So, now we'll turn it over to David.

20MR. HEINER: Thank you very much, Bill and Joe,

21for the opportunity to present here today. My colleagues

22at Microsoft and I really appreciate the opportunity to

23contribute to these proceedings.

24We were asked to provide a business perspective

25on living under Section 2 of the Sherman Act. I think

12

1it's fair to say that Microsoft has considerable

2experience in this area, probably more than most companies

3might wish for, to be honest. And not only Section 2 of

4the Sherman Act, but also Article 82 in Europe and

5comparable provisions around the world.

6Section 2 issues are potentially relevant to a

7broad range of Microsoft's business: product design

8issues, as well as more traditional subjects of antitrust

9analysis, such as packaging, pricing and IP licensing.

10One point comes through loud and clear from the

11business people when you ask them about their experience

12under Section 2, as I did in preparation for the

13presentation today. And that is, as business people, you

14just want to know what are the rules. If you could

15provide it to them in clearer fashion than we're able to

16today, they'd be happy to go devise business strategies,

17to live within those rules and still be successful.

18What's really challenging in the Section 2 area,

19as opposed to, say, Section 1 cartel behavior, is that so

20often advice has to be provided in shades of gray. That's

21of course the reality we live with, but this can be

22challenging for business executives, especially I would

23say mid-level people and below, who just aren't used to

24getting that kind of advice, who are busy with their own

25planning and strategizing, and they look to the law

13

1department of a company such as Microsoft to give a green

2light or a red light. And all too often it's a yellow

3light.

4You might say, what's new in all of this? It's

5always been this way. And that's certainly true. But, as

6the Antitrust Modernization Commission has commented in

7its draft report, as we move toward a more flexible

8approach to antitrust analysis over the past thirty years,

9one side effect has been, less predictability. And

10it's of course a positive thing that we move to a more

11flexible approach. But it seems that the combination of

12that, plus a range of other factors that I'll discuss, are

13really building upon one another to move to such a level

14of difficulty in predicting the outcome of various

15antitrust issues as to create a significant problem.

16Part of this arises from the rule of reason.

17And obviously it's a balancing test. So, any time you

18have a balancing test, it's a fair question as to how a

19typical judge or agency will do the balance.

20I think we've got something even deeper going on

21here, though, in the Section 2 context, in that lawyers

22and economists often disagree as to whether particular

23conduct is procompetitive or anticompetitive in the first

24place, before you even get to any analysis. And that

25obviously is a really fundamental kind of point.

14

1Two examples here that I found kind of striking,

2one is from the Department of Justice case against Microsoft

3back in 1998. That case, as many of you will remember,

4primarily concerned the development of Windows 95 and

5Windows 98 and the inclusion of web browsing functionality

6in that time frame. There were additional allegations as

7well.

8And the DOJ had as its expert economist, world

9renowned economist, defender of IBM, Frank Fisher. And

10Professor Fisher came in and looked at the range of

11conduct, which was a substantial subset of everything

12Microsoft had done in competing with Netscape, and said,

13it's all anticompetitive, you know, it doesn't make

14business sense except for its tendency to exclude and

15therefore it's anticompetitive.

16Now, Microsoft got expert testimony from another

17renowned economist, also from Boston, Dean Schmalensee of

18the MIT Sloan School of Management. Dean Schmalensee

19looked at the very same set of practices. And there was not

20much dispute as to facts. There was some, but basically the

21facts were understood. He looked at the same set of conduct,

22and said, not only is it not anticompetitive, this conduct

23is procompetitive. This is a firm building better

24products and distributing them broadly to consumers.

25So, fundamental disagreement among two very

15

1respected people. Before you get to any balance just is

2the conduct procompetitive or not?

3Another example is pertinent today. After the

4Department of Justice proceedings, there was a proceeding

5in Europe that also concerned the same issue, which is the

6integration of new features into a product, again in this

7case Windows. The European case concerns media play back

8software. So, this is Windows Media Player.

9And Microsoft has explained to the European

10Commission that the purpose of Windows is to be a platform

11for running applications. So, there's a set of software

12services in that product. They're exposed to the

13development community through application programming

14interfaces. Developers can write to those interfaces and

15it saves them a great deal of work in creating their

16applications.

17And what we said to the Commission is that, part

18of the value, a big part of the value that Windows

19provides, is that it's a kind of compatibility layer

20across hardware from many different computer manufacturers,

21hundreds of different manufacturers. So, if these

22manufacturers install Windows, a software developer can

23run an application, it will run on Windows, and therefore

24it runs on an HP machine or a Dell machine or Gateway or

25anything else.

16

1And the Commission said, you know, we think of

2the media play back functionality is something separate

3from the operating system. We don't think it should be

4there and therefore we think you should offer multiple

5versions of Windows with and without that functionality.

6And we said, well, if we do that, it's going to make that

7functionality less valuable to the developers because if

8they write to those APIs and a customer has a version of

9Windows installed where those APIs are not present, the

10application will not function properly.

11So, from our perspective, we're saying that

12maintaining the uniformity of Windows across all these

13different systems is key to the value it provides and

14therefore it's procompetitive.

15And the Commission came back and said, the very

16thing you're talking about, that's what we see as

17anticompetitive because only you Microsoft have the

18ability to add functionality to Windows since you're the

19only developer of Windows and therefore be able to get it

20out on virtually every PC since so many PCs are shipped

21with Windows.

22And here the competitor was Real Networks. And

23the Commission's decision was, they will always be on less

24than the number of machines that Windows is on and

25therefore they will have a disadvantage that's unfair and

17

1it's illegal.

2So, here again, a very fundamental question: Is

3that conduct procompetitive or not? This case is on

4appeal to the Court of First Instance in Europe. We

5expect a ruling perhaps within the next six months, so we

6might have some decision on that particular point, which

7will be interesting.

8So, as I think about the development of

9antitrust law, especially over the past ten years or so, I

10think a range of factors are coming together to make the

11job of an in-house counsel or outside counsel providing

12antitrust advice even more challenging than it's been in

13the past.

14One of these is the development of new business

15models. Business models with which the law has relatively

16little experience so far and business models that lead

17firms to engage in business strategies that wouldn't make

18sense in traditional brick-and-mortar-type industries.

19I'm thinking here, for instance, of the development of

20compatible ecosystems, businesses with network effects,

21businesses that, as the economists would say, are

22multi-sided, multiple players involved that a firm is trying

23to satisfy. With Windows, it's computer manufacturers who

24license it from Microsoft, and software developers who

25write applications. Or the Apple iTunes services, where

18

1you've got the record labels, artists and consumers. Or

2the Google ad platform, where they're serving websites and

3developing advertising systems for those websites,

4advertisers and consumers.

5In these kinds of markets, it's often the case

6that it makes sense to give away something that's very

7valuable, which a competitor might not be giving away,

8in order to attract users early on and thereby try to

9generate a network effect.

10It often makes sense to give something away,

11again, that someone else might not be giving away, in

12order to attract one set of players to a market where

13there's multiple players involved.

14Interesting questions arise as to business

15strategy between ecosystems and the compatibility between

16those systems. So, iTunes, for instance, is I think

17incompatible by design with other media play back systems.

18Apple has developed an end-to-end system that works very

19well. And kind of part of the beauty is they own

20everything. They own the device, the iPod, the software,

21the client software, and the service. And they're able to

22design it to work very well.

23Well, in Europe at least, they're under attack

24for that in a very significant way. Very interesting

25questions that are not really handled in the case books.

19

1Then we have the fact that in many of the

2emerging businesses today, business models, characterized

3by products with very low margin of costs and that soon

4leads to a range of new business strategies.

5Bundled pricing, pricing a collection of

6products or services for significantly less than the sum

7of the stand-alone pricing. Often highly efficient and

8valuable for consumers in the case where it costs the firm

9very little because the marginal cost is little and it adds

10more value for consumers.

11In these businesses, based on information and

12goods, it's often the case that a competitor can very

13quickly ramp up to satisfy one hundred percent of demand.

14And that means that when we look at the market share at

15any given point in time, it doesn't necessarily reflect

16productive capacity like in the old days, and so that firm

17doesn't need to build new factories or anything like that

18in order to satisfy all demand.

19How do you analyze that in the context of giving

20antitrust advice?

21We also see that in these new business models

22and low marginal cost products many different ways in

23which you can modify your business. And you end up in a

24situation where different firms are competing directly

25with one another but with very different business models.

20

1So, in the case of Microsoft Windows, the model is quite

2clear that you primarily earn revenue by licensing the

3product to computer manufacturers for a royalty. And it's

4essentially free to software developers who can build

5applications.

6Along comes the open source movement and Linux,

7and here we have essentially a direct competitor, on both

8the client side and server computers, and that product is

9free. And we have firms that just -- Red Hat and Novell

10and others, making a business out of providing service for

11the software once it's provided to customers. Very

12different model.

13Similarly, with Apple, they're making their

14money by selling the iPod device and they're making money

15by selling the subscription service to music over the

16Internet.

17Many of these new models lead to complex

18relationships between firms. And that's a point that I'll

19return to.

20Another aspect that I think is interesting in

21terms of predictability is how technology based so many

22businesses are today. Many of these technologies are very

23much IP-based, as Windows is. It's nothing but IP.

24Copyright license that we're providing to computer

25manufacturers. So, right off the bat in analyzing these

21

1issues, we are at the always difficult IP/antitrust

2intersection.

3Here we are in 2007 and the debate is still

4going on about whether a patent confers market power.

5It's a fundamental question that still needs to be

6resolved.

7With the focus on new technology, we're seeing

8an increasing focus on product design. And that again is

9not something we've seen in the past. Questions regarding

10integration of new features, not just Windows, but in

11other contexts as well. How features work; how third

12parties can connect.

13And this is an area where, given the complexity

14of the technology, it can be quite challenging for lawyers

15and economists to work through these issues. And that

16complexity of course makes it then an additional degree of

17uncertainty, with the adviser trying to provide advice to

18his client.

19In many cases, technology is so complex we have

20to turn to experts, to technical experts. They may have a

21religious view about some of these topics. They may have

22an axe to grind.

23And when you have technology, at least in the

24case of software, which I'm familiar with, it is so often

25the case that any design can be second guessed because

22

1there's always a different way something could have been

2done. So that too adds a degree of uncertainty.

3When you get into product design, you have the

4antitrust agencies, or whoever else is enforcing the

5antitrust laws, having to look at engineering tradeoffs.

6So, you have a tradeoff between some benefit from an

7engineering perspective and a competition effect. That can

8be hard to assess. And you may want to consider the risk

9that a competition agency, by its very nature, may place

10much greater weight on a competition concern that is

11relativity minor, compared to some engineering concern

12that quite significant.

13Then you have the challenge of time lags. The

14development cycle of some of these products is quite long.

15I mean, it has been famously long for Vista. You have a

16situation where the engineers need to be told what they're

17going to build very early on. You know, they're

18black-and-white people, what are the specifications for

19what we're building. So, from day one they're looking at

20what will this product be. And that's when you have to

21give the antitrust advice. It will be assessed perhaps

22many years later.

23Two other factors that I think are making

24predictions more challenging than in the past. Multiple

25constituencies involve multiple enforcers. One way to

23

1reduce antitrust risk from a practical perspective is to

2try to address concerns before they arise. And we're very

3much on that path at Microsoft. In connection with a

4product like Windows, there's a lot of people involved.

5There's computer manufacturers, there's software

6developers, there's consumers, there's peripheral

7manufacturers, there's websites, and others. And everyone

8has an idea about how it should be built. And, as part of

9the product design process, we're out there to a great

10extent getting feedback.

11We now try to get the legal concerns out early

12in the process as well and address them. One of the

13things we find is that different groups may have very

14different interests. So, the interests of a computer

15manufacturer such as HP may differ in some cases from the

16interests of a software developer.

17We've seen cases recently where even similarly

18situated firms may have different views about how some

19things ought to be done. And these views are expressed to

20Microsoft and agencies in the language of antitrust.

21I can give you an example here. We released

22Internet Explorer 7 recently. So, this is a version of

23the web browser that gets installed on existing Windows XP

24systems. And this browser, if you used it, has a box up

25in the corner for searching the web. The design is as

24

1open as it can possibly be. You can set that box to use

2any web search engine, you can have multiple web search

3engines, you can add search engines, you can delete search

4engines. So, it's all very open.

5A question arose about what the initial setting

6would be. So, a customer asks his or her computer to

7install Internet Explorer 7. The very first time you

8conduct a search, will it go to Google or Yahoo or AOL or

9Microsoft, where will it go?

10And one firm said, you ought to just look at

11what the existing settings are in Internet Explorer 6.

12And that would be Microsoft's normal practice in upgrading

13Windows, you just carry over the settings.

14Another firm said, you know, the settings are

15kind of a hard to find within Internet Explorer 6, so they

16don't necessarily reflect a consumer preference. Why

17don't you just ask, just say, what would you like the

18initial setting to be?

19Both firms felt very strongly about their

20respective positions. They both expressed their views in

21the language of antitrust. And we couldn't satisfy both

22of them. Eventually it was worked out and we have what we

23think is a compromise solution that we hope they're both

24satisfied with. But it illustrates the point about the

25challenges one can face.

25

1Then we have multiple enforcers. So, when

2you're making a prediction, it usually is kind of an

3academic, theoretical question: How would a judge, when

4presented with all the facts, rule on this. At a much

5more practical level, you're really saying, how would the

6Department of Justice look at this? How would the State

7Attorneys General look at this? How would the European

8Commission look at this? How would the Fair Trade

9Commissions in Taiwan, Australia, Japan and others look at

10this? How would competitors look at this? And competitors

11are clearly not in a position of a judge applying -- coming

12up with a perfect result. They have their own parochial

13interest of course. And consumers. You know, class

14action lawsuits, we faced two hundred of them in the past

15ten years or so, many consolidated, but still a big

16number.

17So, there's a lot of different enforcers to look

18at. This is especially significant given globalization.

19We have a situation today where increasingly firms are

20running their businesses on a worldwide basis and it's the

21same business worldwide. These are typically American

22firms.

23So, in the case of Microsoft, it is very much

24the case that it's the same Windows every place in the

25world. And, again, that's part of the beauty and the

26

1value of the product: that it is the same. We license it

2to multinational corporations, so they're taking a license

3to install it in America and Europe and Asia. They want

4one licensing paradigm. So, it's very much in Microsoft's

5interest to have one set of rules that govern all of that.

6Increasingly we see foreign agencies stepping up

7their antitrust enforcement, partly as a result of some

8efforts by the U.S. agencies over the years to have

9foreign countries adopt and apply antitrust laws.

10And while that's of course a useful thing, we

11may find that some of these agencies have differing

12interests, differing views as to how the antitrust

13laws ought to be applied. They come from different legal

14systems. So, in Europe, the development of antitrust law

15is very much influenced by German thought and French

16thought, which is somewhat alien to U.S. lawyers coming

17out of the UK tradition.

18And then we go overseas where we have matters

19pending in Japan and Korea, and here you're outside

20western culture altogether. And we have China developing

21antitrust laws. That's interesting to think about. How

22will this Communist country apply the set of rules that

23really goes to the essence of capitalism.

24With the stepped up enforcement, we have the

25prospect of forum shopping. And that clearly is going

27

1on. So, just this morning, there's an interview with a

2Brussels-based lawyer, who points out that he's actually

3from Seattle, who has filed a complaint on behalf of

4leading American firms against Microsoft in Brussels. And

5the reason the complaint is filed in Brussels is that it

6probably wouldn't get very far under U.S. law. But

7they're hoping for a better, more favorable hearing in

8Brussels.

9Another challenge is the broad scope of

10prosecutorial discretion. When you look at the range of

11antitrust laws, again, especially in Europe, one can see

12that there's quite a range of practices that might

13actually be subject to challenge and yet they're not

14challenged. So, the counselor has to think about what

15actually would be the enforcement agenda of these

16different agencies.

17In Europe at least, we see the European

18Commission going after practices for which, in our view, a

19consensus does not exist that the practices are actually

20anticompetitive. And I'm thinking here of the discussion

21paper that came out six months or a year ago.

22We have, considering how prosecutors and

23enforcement agencies overseas will exercise their

24discretion, to focus on their different views of antitrust

25law. We have the consumer welfare standard in the United

28

1States pretty well established. In Europe, not so well

2established. Much more a sense over there that the

3antitrust laws are designed to protect the small fish from

4the big fish. The small fish may well be little firms.

5Mainly in the cases with Microsoft, it turns out they're

6not. They're the large firms based in the U.S. But in

7some cases, they may be local small fish. This raises the

8specter of protectionism.

9To what extent will trade policy come into play

10in the application of antitrust law overseas?

11And then one has to consider the interaction

12between enforcement agencies. In the United States, Chris

13raised the perfect discussion about the relationship

14between the respective rules of the DOJ and the FTC and

15the states. And here at least we have federalism that

16moderates that to some extent. There's nothing really

17comparable going on at the level of Washington, Brussels

18and other foreign capitals.

19And what we can see from time to time is people

20who believe in competition competing very vigorously with

21one another. So, competition between enforcement

22agencies.

23Hew Pate gave a speech a few years ago where he

24talked about multiple agencies taking a whack at the

25pinata. And I thought that was really quite apt. In

29

1Microsoft's case, the central issue we've been dealing

2with for more than ten years is this question of how the

3integration of new function into Windows over time ought

4to be thought about from an antitrust perspective.

5And we had a major trial on that in the United

6States. And there was an outcome. And an approach came

7out of that outcome which focuses on trying to balance the

8interests of all concerned. And it's an approach where

9Microsoft is including functionality in Windows, but at

10the same time, doing so in such a way that opportunities

11are preserved for third parties to write software that

12runs on top and can be broadly distributed. So, that's

13the U.S. approach.

14Now, the Commission said -- and we tried to

15explain that approach to the Commission and said the

16problem is being largely addressed. The Commission said,

17everything you've done here is all well and good, but it's

18not enough, and we want you to take it to the next level.

19And their solution was, do everything under the U.S.

20consent decree, which was the outcome of this U.S. case,

21and make multiple versions of Windows with and without key

22features. Then we get to the point where it's troublesome

23from a business perspective in providing value.

24In the case of Media Player, they said

25explicitly that it's a precedent to be applied in the

30

1future. So, now we have that additional step where we're

2talking about multiple versions. And we do have Windows

3in Europe without Media Player, although no one has

4purchased it to speak of, less than two thousand units

5sold.

6Korea then came along next and said, everything

7you did in the U.S. is well and fine, and so is everything

8you did in Europe, but you should take an additional step.

9And that is, any version that has all the functionality,

10you should include links to your competitors' products.

11So, we've done that, too. So, in Korea, the Korean

12version of Windows, when you boot it up, right there

13there's a promotion for third party products on the

14screen. Three difference approaches, each one adding to

15the other.

16So, you might say, again, you know, what's new,

17it's sort of always been this way. And I think it is

18getting to be a more challenging issue, as I say,

19particularly how the law will be applied. But then adding

20to that is really the stakes are higher than ever for a

21couple of reasons.

22One is, since we are focused now on product

23design, we've got a situation where engineers really need

24to know what we're building. And you saw in my slide,

25we're having to make decisions. And at that time it may

31

1be the case that you don't even know as a firm whether you

2have competitors, much less what their concerns might be

3for some functionality that you're building. Your

4competitors may be at the same stage of development as you

5are, which is it isn't released yet, it's the next

6generation kind of thing. But you have to make decisions

7anyway.

8Years later it will be assessed with a set of

9facts that didn't exist when you made the decision. This

10is especially sort of challenging because it's often quite

11difficult to undo a design decision. It's unlike the

12traditional stuff of antitrust where you have got a

13contract, if someone decides the contract is improper, you

14can change the contract. Well, once the cake is baked and

15it's on the cooling rack, it's baked. You can bake a

16different cake next time, but that cake is done.

17And when it comes to complex products, like

18microprocessors or cell phone technologies, different

19parts of the system will rely upon particular features

20that might have been the subject of antitrust defense.

21You can change them, but other parts of the system will

22fail.

23Third parties, the software developers, may rely

24on that functionality. If you change it, their products

25will not work. An example here that I think is quite

32

1telling is the development of Windows 95. So, in the days

2before Windows 95, you might remember, we had MS-DOS,

3which was the character-based operating system then,

4running on top of that, Windows 3.1. And in about 1990,

5when those products were really just getting to critical

6mass at that time, Microsoft set out in its plans to

7develop Windows 95. Windows 95 was released in 1995, and

8attacked at that time by some as an unlawful tie of MS-DOS

9and Windows 3.1.

10So, what some said was, this product really

11should be called MS-DOS 7.0. I think seven was the next

12number in Windows 3.2 or Windows 4.0. Now, the Department

13of Justice looked at that in connection with a consent

14decree we were negotiating at that time and it was

15recognized in those discussions that Windows 95 was an

16example of good integration. This was a real step

17forward. It was really building something new. It would

18not be regarded as a tie of these two separate products.

19And Windows 95 was released and it was probably

20one of the most successful products in the history of

21commerce. Tremendous value provided to customers and the

22very best of times for the PC industry. Sales of HP and

23other manufacturers took off, and then we moved right into

24the Internet era in the late '90s. So, a terrific

25outcome.

33

1But still there were claims that that product

2which was so successful and so valuable could be thought

3of as a tie. And even today in 2007, as we sit here

4today, that claim is on trial in a courtroom in Iowa. So,

5one of our consumer class action cases is pending today

6and this very issue is being discussed in 2007, twelve

7years down the road. Now, if the Iowa view were

8correct, in the view of those plaintiffs, we wouldn't

9have had Windows 95.

10Another aspect in which the stakes are higher

11than ever is the focus on IP licensing. I think we're

12increasingly seeing firms around the world seeking access

13to the technology of their rivals on favorable terms. And

14here again, it's kind of like the product design case

15where it's an either/or situation.

16So, your technology is either licensed and made

17available or it's not. And if it's made available, it's

18out there, it's gone, you probably won't be able to get it

19back.

20In the computer industry context, the IP is

21often based on trade secrets. Once you have licensed that

22technology, you can try for protectionism on the use of

23it, but the trade secrets are out in the world. And once

24it's licensed, the point of licensing it obviously is for

25third parties to use it and rely upon it, and if you do

34

1rely upon it, it would be hard to get it back. So, when

2you make these decisions, the stakes are high.

3The rise of global antitrust enforcement is

4quite significant here. In the European Commission case,

5a decision was taken against Microsoft relating not only

6to the product integration issues but also IP licensing.

7And here the Commission made a decision that Microsoft

8would have to license protocol technology to third

9parties. And the Commission observed that it's

10essentially a global market for this kind of IP and

11therefore this technology ought to be licensed on a global

12basis. So, Microsoft is doing that.

13The Commission has also taken the position that

14Microsoft ought to license this technology in a way that

15it can be taken in practice by open source developers.

16And that's quite troublesome for a commercial firm such as

17Microsoft because that means that the trade secrets will

18be revealed to the world. Once the technology is

19licensed, it will be built into open source products, the

20source code can be seen, and therefore the trade secrets.

21 Similarly, it's very hard to maintain the value

22of IP once it's licensed under an open source model

23because, again, every copy of the product will be made

24available for free. It's hard having this kind of

25limitation on sublicensing and royalties coming back.

35

1Now, it's not the view of the U.S. enforcement

2agencies that Microsoft should have to make this

3technology available essentially for free and disclose the

4trade secrets. This comes up under the consent decree

5where we have protocol licensing as well.

6And this is before the European Commission and

7Microsoft is contesting it at this point and the outcome

8is yet to be seen. But if the European Commission

9prevails, then we'll have a situation where you have a

10split of authority essentially between the U.S. and EU and

11the EU version will prevail because it's more restrictive

12because they're seeking greater licensing.

13In case after case, I think we may see kind of a

14race to the bottom from the perspective of the target firm

15in IP licensing. And all of this of course in an economy

16that is increasingly IP based creates a specter of reduced

17innovation around IP, and a greater uncertainty as to

18whether the IP can be properly monetized.

19So, what are the consequences of all of this?

20Well, I think we do have a risk at least of over

21deterrence arising from a combination of the difficulty in

22predicting the outcomes, the difficulty in changing course

23later, the variety and number of possible claims, and the

24desire to avoid controversy.

25What are the consumer welfare effects of all

36

1this? Well, we may see limitations on the products'

2improvement. And there have been cases in the context of

3both Windows and Office, Microsoft's flagship products,

4where decisions were made not to include particular

5features that would have been valuable to consumers based

6at least in part on antitrust advice. And one might say

7it was the right outcome or maybe it wasn't the right

8outcome, but the bottom line is, those features are not in

9those products.

10We see antitrust advice from time to time to

11raise prices. And I always kind of pause, as an antitrust

12counselor, before saying the price is too low for that

13collection of products or services. But it's a judgment

14call based on the state of the law on a worldwide basis,

15the range of possible claims, that we better raise prices.

16And clients sometimes get quite confused about

17that because when we do antitrust training, we usually

18start at a 101 kind of point that the purpose of

19antitrust law is more innovation, more output and lower

20prices. So, they receive this advice with a bit of

21skepticism, but it's given nonetheless.

22And I think we're seeing increased R&D costs.

23For something like Windows, there are six billion dollars

24of R&D in that product. That's obviously an extreme case.

25But the amount of time that's spent by executives trying

37

1to pick through how this shades-of-gray antitrust advice

2fits with engineering decisions is really considerable.

3And, finally, I would note that, because of the

4challenges of predicting how antitrust law will be applied

5by the multiple agencies and other enforcers, we may see

6some work that's being undertaken that is of really

7questionable value but done in order to satisfy a

8regulatory concern.

9So, suggestions on how to move forward. I think

10it's a very hard problem and there probably aren't any

11easy answers. In trying to move toward greater clarity in

12the law, I do think it would be helpful if we had a

13stronger presumption that conduct that is widely practiced

14by firms without market power is efficient.

15This is a concept that I think finds some basis

16in U.S. law. It's referenced in the U.S. Court of Appeals

17decision in the Microsoft case in a helpful way, from

18Microsoft's perspective, on the integration issues. It

19doesn't really resonate overseas, I have to say. And

20there's been cases where I've been sitting across the

21table trying to make the point that every firm in the

22industry is engaging in some particular practice,

23therefore they must think it's valuable aside from the

24ability to exclude because they are excluding anybody

25because they have low share.

38

1And the reaction on the other side is often

2really just a blank stare. And so what are you saying,

3it's obvious that the firms -- that the rules are

4different for high share firms, so we really don't

5understand the point you're making.

6Convergence, it's been much discussed. I think

7it would be helpful to see a redoubled effort by U.S.

8agencies to evangelize the U.S. approach.

9And for everything I've said about

10predictability, U.S. law is more predictable than European

11law and the law of other countries with their emerging

12antitrust regimes. A great deal has been said about this

13through the years. Given globalization, I think it is

14increasingly important to find some way to allocate

15responsibility among multiple agencies. And certainly a

16kind of common sense approach would seem to me a greater

17deference to the rules of the defendant's home country. And

18I would say from Microsoft's perspective, we really haven't

19seen much of that in the cases that we've been involved

20in.

21So, again, thank you very much for the

22opportunity to present here today.

23(Applause.)

24MR. COHEN: Thank you, David

25Our next speaker will be Scott Peterson, who is

39

1senior counsel at Hewlett-Packard Company. Mr. Peterson

2has practiced as an intellectual property attorney for a

3number of years, focusing on information technologies. He

4joined HP in 1991 and provided intellectual property

5support for a wide range of HP's businesses, as well as in

6the context of standards development.

7Along with his law degree from Franklin Pierce

8Law Center, Mr. Peterson holds bachelor's and master's

9degrees in electrical engineering from MIT.

10So, we'll hand it over to Scott

11MR. PETERSON: Thank you very much. Thank you

12and I appreciate the opportunity to be here.

13 I am going to be talking on the topic of the

14intersection between intellectual property and standards

15and the competition implications.

16And I want to say I really appreciate the

17attention that the agencies have been paying to this topic

18over the years. And, in fact, the guidance that the

19agencies have been giving in recent years I think has been

20very helpful and has played a role in some of the changes

21that we are actually beginning to see. So, I really thank

22you for your attention to this area.

23I really have one core message throughout this

24presentation. You are actually going to see it on every

25slide. It was the title: Transparency of patent

40

1licensing information during development of standards

2facilitates efficiency in markets for technologies and

3standards. That's the message. I am going to talk about

4it. I'm going to elaborate on it a little bit. But

5that's the core.

6And a kind of corollary to that or related is to

7recommend that guidance on application -- further guidance

8beyond what we have -- on application of Section 1 to

9collective action during standard setting regarding

10licensing terms for patents essential to standard,

11facilitates behavior that reduces the likelihood of

12conduct in violation of Section 2

13So, this is a hearing where the focus is on

14Section 2. My message is actually for guidance on

15Section 1 because the behavior that can be beneficial in

16reducing the Section 2 risks is behavior that's

17potentially chilled by concern about Section 1.

18So, in fact we see significant value in what we

19think of as sort of a voluntary industry-led approach to

20reducing the risk of anticompetitive use of patents

21essential to standards. We recommend proactive action

22that would operate to reduce the need for after-the-fact

23corrective agency enforcement actions of a Section 2 type.

24But this desirable procompetitive behavior that

25could operate to reduce this potential for the

41

1anticompetitive use is being chilled to some extent by

2concern that that collective action poses some Section 1

3liability to the participants in the standard activity.

4So, let me say a little about some background,

5myself and Hewlett-Packard.

6My particular background is that of an

7intellectual property attorney. I have given advice to a

8range of HP businesses. But over the last decade in

9particular, I have given advice on the topic of patents

10and standards. And in the last half of that decade or so,

11I've -- I guess initially that advice was in the context

12of particular transactions, particular standards,

13development activities from people with business

14activities -- and then in the latter half of that decade of

15activity that I have been involved with this, has been in

16trying to coordinate at HP our policy level considerations

17of these questions that arise about intellectual property and

18standard setting.

19HP is -- to turn to the company that I'm talking

20about -- fundamentally in the information technologies business,

21a business which depends enormously on standards, a business which

22has enormous network effects. So, standards are something that HP

23is extremely familiar with. We participate in hundreds of

24standards development activities. We have products that implement

25dozens and dozens of standards. This is not an area where a

42

1product implements a standard. This is an area where

2products implement many, many standards. So, we have

3developed a great deal of experience with the challenges

4of standards development.

5HP is also active as an innovator. HP has

6invested -- let's see -- in the last fiscal year, we

7reported 3.6 billion dollars investment in R&D. HP has

8long invested in R&D. That investment has been reflected

9in an extensive patent portfolio. Again, at the end of

10the last fiscal year, that was reported as about 30,000

11patents.

12So, innovation and the patents that reflect that

13innovation are also very important to HP. So, to give you

14a sense of the perspective of where I'm coming from, it's

15one where an effective standards environment is extremely

16important because it's critical to the nature of the

17products. It enlarges markets for products that HP makes.

18And yet on the other side, patents are also

19something that are an important part of HP's business.

20So, with that background on HP, let me go back

21then through the message, which you have seen here again:

22transparency of patent licensing information during

23development of standards facilitates efficiency in markets

24for technologies and standards.

25Let me start off by saying that there is

43

1potential for anticompetitive use of the patents. This

2was discussed in particular at the December 6th hearing.

3And my goal is not going to be to replow this ground that they

4talked about, but rather -- the fact that a patent that is

5essential to standards can be employed in anticompetitive

6ways is particularly important to recognize. And this

7flows from the fact that once the patent is -- once a

8standard is set and a patent is essential to it -- if the

9standard becomes successful in the sense that there is a

10lock-in effect such that participation in that marketplace

11requires that you implement the standard -- then implementing

12-- and implementing the standard requires a license, then that

13patent now takes on a leverage that goes potentially beyond

14the innovation that underlies it.

15And it's that combination of factors -- there

16is the leverage that one obtains from the innovation itself,

17and yet there's also leverage that could come from the

18lock-in effect of the standard. It's that combination that

19leads to the challenge of potential anticompetitive uses of

20patents that are essential to standards.

21In my 2002 testimony -- I testified in April and

22in November of that year on essentially this same topic --

23I expressed some concern that there was a trend that

24patents essential to standards were going to become an

25increasing problem in the success of standards, and the

44

1potential for abuse was a growing one.

2And I have to say that our observations in the

3intervening years have confirmed our concern about that

4trend. And let me offer one example of something that

5illustrates the trend.

6There is, I think, a fairly increased mobility

7of patents over what we would have seen ten or twenty

8years ago. For example, the concept of patent auctions is

9far more conventional now than it was a decade ago.

10And I am not suggesting there's anything

11inappropriate about this mobility of patents. I think the

12ability to transfer intellectual property rights can be

13extremely valuable. So, I'm not criticizing the trend as

14such, but I simply want to point out that there is a

15substantial change in the dynamic for how a patent gets

16employed and what the licensing and enforcement

17implications might be when the patent moves from the place

18where it started to some other place, in particular for a

19patent that is essential for the standard. It may well

20have begun in a company that was working on technologies,

21and had products, in the area of that particular standard

22and would have certain motivations and expected a business

23behavior. When that patent moves elsewhere, the

24expectations and dynamics are going to be different.

25So, this sort of increase in the mobility of

45

1patents is an example of why I think we have to be more

2careful about paying attention to patents during the

3development of standards, because the opportunity for

4aggressive behavior that may employ or exploit the

5leverage from the standard -- not just the leverage from

6the patent, but the leverage from the standard -- has been

7increasing over the last decade or so.

8So, there is a market which I think is sometimes

9overlooked in talking about licensing of patents in

10connection with standards. It is important to recognize

11that there's a market for technologies in standards, and

12there should be competition in this market for

13technologies in standards. And there are -- in the

14process of making choices as to what will go into the

15standards -- in some cases there are a variety of relatively

16equivalent choices in terms of the capabilities that they

17offer, and yet in other technologies, in other settings,

18sometimes one stands out dramatically above the others

19because the nature of the technology is such that, you

20know, there is opportunity for the standard to make a

21substantially better choice in that particular area.

22The license fees in those cases ought to reflect

23that underlying reality. If in development of a standard one

24is selecting one of many alternatives that are essentially

25comparable in their end result, comparable in the

46

1performance, characteristics and so forth, one would

2expect the license fees to be substantially smaller than

3when one is in a situation where the selected technology

4is in fact head and shoulders above the alternatives, in

5which case the license fees ought to reflect that

6contribution to the standard.

7Once the standard has been selected, however,

8that distinction is easily lost because, again, if there's

9a lock-in effect from the standard, it won't matter that

10there were alternatives at that earlier stage. The

11competition -- the effect of that competition is active at

12the time that the standard is selected. It is either

13effective then or the value of the competition is lost

14because the lock-in effect later would mean that.

15Suppose you had ten different alternatives that were

16fundamentally equivalent. Once that one is anointed as

17the way that you're going to agree among competitors to

18build products in that domain, having a license to that

19patent, if there was a patent, is vastly more valuable

20than it would have been in another case.

21In any case, I think it's important to realize

22that this process of selecting, there is essentially, a market,

23but it's a market that has this odd characteristic. There

24is the collection of people, oftentimes competitors, who

25are selecting what the standard will be. And there will

47

1be a single decision -- in a sense, a single buy decision. And

2the technology that is put in the standard at that point now

3has been selected, in some sense, as if it was purchased. So,

4now if you think about the subsequent licensing transactions,

5these are not really a family of separate independent

6transactions. For those who wish to implement the standard

7and need to have a license to the patent that's essential,

8their licensing transactions are not independent. They're

9already -- they've already fixed the buy decision. There's

10no walk-away for them. In that sense, these aren't

11independent transactions. These are all flowing from

12the single decision which was made as a part of the

13standard's selection.

14So, I guess my point here is that efficiency in

15the market for technologies in standards -- the result of that

16selection -- is very important because the technology selections

17have implications for all of the subsequent licensing

18transactions. Those later transactions may appear in some

19sense as separate, but they're not because the buy

20decision was made once. It was made in the selection of the

21standard.

22Efficiency, market efficiency. So, I make my

23point, you know -- inadequacy of information is preventing

24some efficiency. Well, let me talk about the inefficiency

25which is worthy of some -- being made more efficient.

48

1The inefficiency in the market for the

2technology that goes into the standards is essentially the

3information problem associated with the licensing terms

4for patents that would be required by the various

5alternative choices.

6So, I talked about a market for technologies and

7standards. A choice is going to be made among potentially

8alternative technical choices. One of the factors which one

9would normally consider when making an economic choice is

10price or other terms that might be associated with the

11decision. And, oddly enough, instandard setting, that

12information is not circulated, is not readily available to

13those who are making this decision. So, you have a group of

14participants in a standard setting activity who are talking

15about a wide range of characteristics of the technologies

16and choices that they are choosing among, and yet this

17topic of what the licensing implications would be is oddly

18excluded from that conversation. And, in fact, the mechanics

19by which anyone comes to know that is, by and large vastly more

20obscure. And the flow of that information is inhibited by

21the concern that, because it involves a dollar amount

22there must be price fixing concern of some sort. And

23therefore this is the Section 1 concern that I referred to

24that is inhibiting the sharing of this information, which

25is in fact important in making a rational and fully

49

1informed decision in this market for technologies.

2Let me talk about -- so, markets for

3technologies in standards. I think it's important to

4realize I have been focused on patents in the sense of

5essential patents -- those patents which you must have a

6license to because of how the standard was conceived.

7The competition in products that employ

8standards and the innovation in those products

9predominately takes place outside of what's specified in

10the standard. So, in general, as I say on the slide here,

11standard setting should seek to enable technology and not

12to specify or require it.

13Now, many times the nature of the problem being

14addressed, there may be somewhat limited constraints or

15constraints that make a range of behaviors possibly not as

16great as one would like. But I think that in many cases

17inadequate imagination has been applied to the problem of,

18"Let's make sure that we specify as little as possible

19because we want to foster competition and we want to

20foster ongoing competition." And yet choosing a standard

21essentially freezes a particular technological point.

22There ceases to be competition to the extent that there's

23-- that there's lock-in on the standard. And from the

24time that that standard is important, there ceases to be

25competition on that particular set of things which is

50

1specified in the standard.

2 There are technological decisions that can be

3made as to how you define the specification, what is

4needed to achieve the network effects that the standard is

5trying to accomplish.

6I think that the environment that we presently

7have, which excludes to a large extent from consideration

8the licensing concerns, results in, to some extent, a

9motivation to incorporate as much technology and

10innovation into the standard as possible. And, in fact,

11that's the wrong motivation. We want to motivate people

12to keep technology out of the standard. You want to keep

13the technology from being specified. You want the

14standard to enable the non-required technology which

15continues to be the subject of further evolution and

16competition among even the preexisting alternatives.

17So, I think that the present environment,

18where the licensing considerations are not considered, has

19an interesting adverse effect in this regard.

20And then finally -- transparency of patent

21licensing decisions during development of standards. This

22procompetitive behavior of considering that information

23while the standard is being selected -- as I pointed out,

24people are concerned and have a longstanding concern that

25there's some kind of a price fixing type environment that

51

1will be created if in fact the license terms are

2considered.

3I think that in fact, in this environment, that's

4a misunderstanding of the situation. In fact, there will

5be a single group buy decision in the sense of the group

6will select a final specification. The problem is that it

7won't be informed by this information.

8So, the idea of looking at this as leaving the

9door open for a multitude of independent later licensing

10decisions, I think it's failing to understand that the

11reality is that there is one decision that's going to be

12made. It is deciding whether a particular thing is

13essential or not essential. The question is whether

14that's going to be informed by license terms.

15So, I go back to the beginning slide, and let me

16make some comments in sort of the recommendation category.

17It can be difficult to separate, after a

18standard has been selected and after a patent is

19essential -- it can be difficult to separate the legitimate

20aggressive enforcement of patent rights from the use of a

21patent that is being leveraged to essentially leverage the

22value that was created by the collective work of the

23competitors.

24So, those are very difficult to keep apart after

25the fact. There is no market, really that you can rely on

52

1in the ex post world. So, I think it's very important to

2foster a proper attention to this issue while the

3standard is being selected.

4A couple of -- let' see -- one problem -- two

5particular problems that I want to point out that merit

6some attention going forward.

7One is the -- I mentioned mobility of patents

8is increasing patents are increasingly mobile. So, one

9challenge is that licensing commitment typically you cannot

10-- under the regime of many standards development activities,

11you cannot rely on those licensing commitments passing

12through as the patents move from one owner to another. This

13is a problem meriting attention. And organizations may strive

14to do something about that in the context of standard setting.

15They may ask people to make commitments or something. It's

16a problem of increasing concern because of the likelihood

17that patents are moving.

18And another problem is that of the injunctions

19in the face of licensing commitments. So, again, this is

20another sign the commitments are of a fairly tenuous

21nature. So, there may be licensing commitments. On the

22other hand, the ability to turn off someone's ability to

23practice a particular standard can be an incredibly large

24negotiating lever. And the fact that that lever could be

25available even in the case of a licensing commitment is a

53

1very troubling one.

2I guess I'll close there. And I guess I'll once

3again thank the agencies for continuing to pay attention

4to this topic. I appreciate the guidance that's been

5offered so far, but I think there's lot more. As the

6world changes and begins to pay more attention to patents

7during the development of standards, we're going to learn

8more about what the issues are and perhaps more guidance

9will be needed

10Thank you very much.

11(Applause.)

12MR. COHEN: Our next speaker is going to give us

13some insights from the perspective of a standard setting

14organization. He is Robert Skitol, who is senior partner

15in the Antitrust Practice Group at Drinker Biddle & Reath

16in Washington. And he is counsel to the VMEbus

17International Trade Association, know as VITA.

18Mr. Skitol is a graduate of Hobart College and

19NYU Law School. He has over 35 years experience in all

20facets of antitrust and trade regulation, and written and

21lectured extensively in the antitrust and trade regulation

22field

23At this point, we'll give the podium to Bob.

24Do you have slides, Bob, or not?

25MR. SKITOL: I do have slides.

54

1MR. COHEN: We just have to find them.

2MR. SKITOL: I can proceed without the slides.

3There is a slide set, but I'm happy to speak without it.

4Well, thank you for your indulgence. I am

5delighted to be here on behalf of the VITA standards

6organization. I'll be offering VITA's perspectives on

7some of the same points and issues and concerns that Scott

8spoke about.

9My comments are complimentary to Scott's in many

10respects. Scott spoke about patents and standards from

11the standpoint of a major technology innovation intensive

12company that participates in standard setting proceedings.

13My client VITA is a major standards development

14organization that is the flip side of the concerns. But

15for VITA certainly, Scott's transparency theme resonates

16quite a bit. And so I want to use my time today to offer

17VITA's perspectives on how the antitrust agencies should

18assist SDOs in protecting their processes from

19exclusionary patent hold up conduct.

20Of course VITA appreciates and has been a major

21beneficiary of steps in this direction that the agencies

22have already undertaken. My remarks concern desirable

23next steps along this path.

24I think the logical place to begin is with the

25definition of exclusionary patent hold up conduct. And I

55

1want to propose one broad enough to encompass an array of

2patent related practices that subvert or can subvert open

3standards and produce anticompetitive market outcomes.

4So, my proposed definition for the agency's

5consideration is as follows. A patent owner's inducement

6of an SDO's adoption of a standard that implicates the

7owner's patent claims without other participants'

8awareness of that fact or without their awareness of the

9cost and other impacts of it, thereby enabling the owner

10to acquire and exercise monopoly power that it would not

11otherwise have obtained.

12Now, this is not news to the antitrust agencies,

13this general concept. The FTC has been active in

14challenging hold up conduct of this kind for about twelve

15years. The Dell, Unocal and Rambus cases collectively

16delineate a framework for treating hold up conduct as a

17Section 2 violation in circumstances involving deliberate

18deception regarding the existence of patent claims

19implicated by a draft standard under development.

20These cases also support the idea that the

21requisite deception need not be overt. Mere silence about

22essential patent claims can be unlawful when that behavior

23actually misleads other participants in light of

24expectations generated by the organization's rules or

25established practices.

56

1But hiding the existence of essential patent

2claims is not the only way that exclusionary outcomes can

3occur. There are other ways that patents can be used to

4morph or subvert an open standards process into the

5practical equivalent of market monopolization.

6And I want to suggest three examples for your

7consideration, all involving situations where the

8existence of essential patent claims may well be

9disclosed, may well be known, but patent hold up conduct

10of an anticompetitive nature can nonetheless occur.

11And the first example is one that entails

12inducing reliance on a generalized commitment to license

13essential claims on reasonable and nondiscriminatory

14terms, the so-called RAND assurance that is in widespread

15use, without the patent owner's acceptance of any

16meaningful constraint on what it demands as actual license

17terms after the standard has been adopted and a whole

18industry is locked into sunk investments in compliant

19products.

20This is the essence of the allegations in

21Broadcom versus QUALCOMM. We don't know the facts. We

22know the allegations. And the allegations tell a story of

23how generalized undefined RAND commitments can end up

24bringing about monopolization.

25The second example entails inducing reliance on

57

1that kind of RAND assurance followed by seeking

2injunctive relief to enforce the applicable claims. This is

3a situation Scott also commented upon.

4From my standpoint, from VITA's standpoint, the

5injunction threat is fundamentally contrary to the whole

6idea of the RAND assurance and the intended reliance upon

7it. The only legitimate issue in any ensuing litigation,

8once that assurance has been given and relied upon, should

9be what those promised reasonable terms are, the patent

10owner having effectively given up the right to exclude

11under the patent code in return for what will often be

12mega benefits from incorporation of that owner's

13technology into the standard being developed.

14The third example entails the transfer of ownership

15of an implicated patent without binding the new owner of

16it to the original owner's license commitment, the patent

17owner having induced the whole industry into employing the

18patented technology in the belief that acceptable license

19terms were assured. The owner then transfers the patent

20in a manner allowing the new owner to repudiate the

21assurance and exploit the resulting new monopoly power.

22Scott talked about the recent and increasing

23trend of patent mobility, which seems to me to underline

24the danger that this particular kind of hold up conduct is

25something we need to worry more about in the time ahead.

58

1So, all of these kinds of exploitive conduct and

2the resulting hold up outcomes from them are today's

3version of monopolization through highjacking an industry

4standards development project, much as did the conduct at

5issue in the Supreme Court's Allied Tube and Hydrolevel

6decisions of two decades ago. Those cases involved different

7kinds of conduct, but with essentially the same kind of effect

8as patent hold up conduct can have today. This is really all

9about proprietary capture of what is intended to be an

10open standards process with market-wide effects of the same

11nature as those condemned in those past cases of the Supreme

12Court.

13Now, there is disagreement in the standards

14development community about the extent or prevalence of

15these kinds of hold up situations, as I will explain in a

16few minutes. My client, VITA, has some relevant

17experience in this regard and knows from its own

18experience that this is far from an isolated event.

19But two developments, at least two developments,

20strongly suggest increasing exposure to it. One is the

21vast proliferation of patent grants that we are witnessing

22within standards intensive technology spaces.

23And the other development is what we're

24seeing as the emergence of new business models of some

25technology companies that depend on maximization of

59

1licensing revenues from the use of their patents in

2standards specifications.

3In this environment with these developments,

4SDOs' inattention to the problems that do surface invites

5proliferation of these hold up situations in the years

6ahead.

7Now let me tell you more specifically -- let

8me catch up on the slides. Let me tell you more

9specifically about VITA and VITA's role in this story.

10VITA develops standards for modular embedded

11computer systems in a wide range of products. Members and

12participants in its working groups include a broad cross

13section of builders and users of these systems for such

14applications as medical imaging, aviation and navigation

15devices for military defense and space exploration.

16VITA's management, particularly its

17distinguished executive director Ray Alderman, have come

18to acquire some rather deep expertise and experience in

19patent hold up. In its own proceedings, VITA has

20encountered no less than four major patent hold up

21episodes within the past six years, each one causing major

22delay in the implementation of foundation standards

23critical to members' technology advancement needs, and

24imposing on the organization major expenses to address and

25counter the asserted claims.

60

1These episodes are described in some detail in

2VITA's application for a DOJ business review letter that

3I'll talk about shortly.

4 VITA recognized one year ago that it was exposed

5to more such episodes and encounters of this sort in the

6immediate years ahead, in light of a considerable patent

7thicket surrounding a planned technology transition that

8would need to drive the upcoming standards development

9activity.

10It also recognized, and its members recognized,

11that VITA's longstanding patent policy actually enabled

12and facilitated rather than protecting against hold up

13conduct of this sort given reliance on wholly undefined

14RAND assurances with no information on actual license

15terms until after a standard was adopted or at a very

16advanced stage of the VITA development process.

17So, VITA devised a new patent policy designed to

18ensure greater transparency earlier in the proceeding in

19all of these respects. There are several elements of the new

20policy revolving around disclosure obligations of working

21group members at each of four stages of the working

22group process, including the very beginning and midpoints

23of it.

24Required disclosures of all potentially

25essential patent claims, including those set forth in

61

1pending applications, based on good faith and reasonable

2inquiry into the members' patent positions; required

3disclosures of a maximum royalty rate and incentives for

4disclosure of other license terms; clear acknowledgment

5that the proffered disclosures will be legally enforceable

6by prospective licensees against not only the disclosing

7member company but also successors and assigns and

8transferees of the underlying patents; and, finally, an

9arbitration procedure for compliance disputes.

10In June of last year, VITA applied to the

11Department of Justice for advice on the new policy under

12the business review procedure. On October 30, 2006, the

13DOJ issued a favorable letter, and it provides a

14considerable amount of analysis and insight on DOJ's

15perspectives about the patent hold up problem in general

16and about how disclosure requirements of the sort

17described in VITA's new policy can be an effective

18safeguard against that kind of conduct and outcome.

19The letter concluded that the new VITA policy

20would be an efficiency enhancing contribution to VITA's

21standards development processes. DOJ characterized the

22policy as an attempt to preserve competition and thereby

23avoid unreasonable patent licensing terms that might

24threaten the success of future standards; avoiding

25disputes over licensing terms that can delay adoption and

62

1implementation after standards are set; and, thus, a

2sensible effort by VITA to address a problem created by

3the standard setting process itself.

4Needless to say, VITA very much welcomes and

5appreciates the guidance that this letter provided and

6believes it has a tremendous value to the standards

7development community as a whole.

8With the DOJ letter in hand, the VITA membership

9on January 17, 2007 overwhelmingly approved and adopted

10the new patent policy and it's now undergoing the

11requisite review by the ANSI Executive Standards Council.

12 Now, at this point -- hold on one second. That

13is where I am. I'd like to offer four reasons why the

14agencies should now affirmatively encourage other SDOs

15to follow VITA's lead by experimenting with new patent

16policies of their own.

17And the first reason is that the DOJ's VITA letter,

18as well as several speeches by officials of both agencies

19in the last two years, recognize that SDO policies of

20this general kind are not just okay from an antitrust

21standpoint but can be procompetitive in their protection

22against hold up outcome. In short, these policies serve

23the public interest in protecting and promoting a robust

24competition throughout standards driven technology

25markets.

63

1Second, the FTC's Rambus decision suggests that

2the viability of any Section 2 case against hold up

3conduct in this context may depend on a showing that the

4patent owner's actions were contrary to SDO participants'

5reasonable expectations in light of SDO policies in place.

6So, in short, in this respect, if an SDO fails

7to implement effective protection against abuse of its

8processes in this manner, then participants will be in an

9awfully weak position, if any position at all, to complain

10about the resulting injury to them. And the government

11will be in a weak position or no position to mount an

12attack upon the situation, even though the public is

13adversely affected by an anticompetitive market outcome.

14Third, effective SDO self-policing or

15self-regulation through policies of this sort will reduce the

16need for agency enforcement actions, as well as reducing

17all participants' exposure to disruptive private suits

18over license terms. And self-regulation is a far

19more efficient solution to this problem than any reliance

20on litigation. This should be obvious to all concerned,

21to everyone that's ever participated in a standards

22development process.

23SDO and its members may spend several years

24developing a new standard, bringing it to completion and

25ultimate adoption but then seeing the whole effort fail

64

1because hold up conduct blocks implementation.

2Now, even if the government at that point steps

3in with a Section 2 enforcement action that results in an

4order, four or five or six years later the damage is done

5and there is no real remedy for the resulting harm to the

6public. So much, much better to prevent the conduct from

7happening in the first place than ever needing to try to

8undo it.

9So, finally, the fourth -- reason number four,

10is that there's no reason to think that VITA's new policy

11is the perfect solution or one suitable for SDOs

12generally. Lessons learned from other SDOs'

13experimentation with variations upon it will resound to

14the benefit of all SDOs and participants in them. There's

15no one size fits all in this area. VITA itself may well

16want to revise, and in all likelihood will want to refine

17in some respects, its new policy a year or so from now

18after experience with it in several working groups.

19VITA will be at least as interested in following

20innovations by other SDOs as they may be interested in

21VITA's experience under its new policy. The enforcement

22agencies, I would suggest, should want to encourage

23information sharing and benchmarking efforts among SDOs

24along these lines.

25Now, allow me to conclude with some specific

65

1suggestions for what the agencies can do in the months and

2years ahead to promote desirable SDO initiatives in this

3area.

4First, the agencies should affirmatively encourage

5more requests for DOJ letters or FTC advisory opinions on

6patent policy proposals of various kinds to provide more

7and deeper guidance for the SDO community in general. And

8one specific example I'd like to suggest of where

9additional guidance and more specific guidance would be

10highly desirable is on the extent to which and manner in

11which a policy might go beyond requiring a disclosure of

12licensing terms, as the VITA policy does, and beyond that

13allowing discussions or even collective negotiation of

14those license terms during SDO meetings.

15I personally believe that these further steps

16going beyond mere disclosure and actually letting the

17working group do something collectively with the

18information would be desirable; it is logical; it makes

19sense in the context of the core mission of an SDO's

20working group, which is to make collective decisions about

21choosing one solution over another; and it makes eminent

22sense for costs or relevant costs between competing

23solutions to be part of the equation.

24I've actually done a whole article on this

25subject, which appeared in the Antitrust Law Journal,

66

1and I understand it's being placed in the record of

2today's hearing. So, now I've plugged my own article.

3But I am convinced that resistance to these

4further steps, anything beyond pure disclosure, rests on

5unfounded antitrust concerns. And there's at least the

6beginning of indication, more than a beginning, that the

7agencies are seeing the matter that way. The latest word

8on this is footnote 27 in DOJ's VITA letter, indicating

9the likelihood that DOJ would address the discussion or

10collective negotiations scenario as a rule of reason

11question because it could actually be procompetitive.

12FTC Chairman Majoras expressed that same view in

13her Stanford speech of September 2005. I hope that one or

14both of the agencies will get an opportunity to provide

15more definitive guidance on this front in the near future.

16Second specific suggestion, I believe the

17agencies should consider undertaking an industry-wide

18study of SDOs' experience with various kinds of hold up

19situations and how existing SDO policies either address or

20fail to address any problems thereby encountered. A study

21of this sort could certainly help to resolve the

22disagreements to which I referred a little while ago over

23whether the hold up threat is or is not prevalent and

24growing. Such a study could also provide a valuable

25information base for suggested solutions or new proposals

67

1for SDO policy reforms.

2Third, the agencies should help to shape case

3law development in this general area by entering private

4suits, by filing Amicus briefs in private cases

5challenging SDO-related conduct and practices where

6unfortunate and harmful decisions are sprouting up.

7Examples of private cases of this sort where DOJ or FTC

8Amicus input could have been valuable are Golden Bridge

9Technology versus Nokia, last year's decision in Texas,

10with its holding of per se illegality against conduct

11appearing to be a common feature of standards development

12activity; and also last year's Broadcom versus QUALCOMM,

13with its ruling that breach of an SDO rule that results in

14monopoly power that would not otherwise be obtained cannot

15ever state an antitrust claim.

16And, fourth and finally, I would respectfully

17encourage both of the agencies to support enactment of

18legislation enabling SDOs to implement desirable patent

19policies without fear of private antitrust claims.

20There's no doubt that that fear has inhibited SDOs from

21considering policies to address patent hold up problems.

22Again, prime examples of private suits having

23exactly that kind of chilling effect and that get talked

24about all the time at SDO meetings as why we better err on

25the side of caution, stay away from any new kind of idea of

68

1that sort, etc., etc., would be the Golden Bridge Technology

2case that I already mentioned, and Sony versus Soundview from

3six years ago.

4VITA is only one of several parties with a lot

5at stake in open standard setting processes and that are now

6exploring the opportunities for legislation in this area.

7I hope DOJ and FTC officials will be interested in

8dialoguing about this possibility with us over the weeks

9ahead.

10Thank you very much.

11(Applause.)

12MR. COHEN: One of the cases you mentioned

13toward the end of your talk was the Broadcom v. QUALCOMM

14case. We have on this panel a representative from

15QUALCOMM and our afternoon session will have a

16representative from Broadcom.

17Our fourth and final speaker is Michael Hartogs,

18Senior Vice President and Division Counsel at QUALCOMM's

19Technology Licensing. Mr. Hartogs has spent his career

20handling intellectual property and competition matters for

21companies that compete in dynamic industries.

22He's been with QUALCOMM since December of 1999.

23Like so many of our other panelists, he brings a diverse

24background: an undergraduate degree in engineering

25physics from the University of Arizona and a law degree

69

1from The George Washington University and registration to

2practice before the United States Patent and Trademark

3Office.

4We turn to Mike.

5MR. HARTOGS: I also want to thank the

6Department of Justice and Federal Trade Commission for

7inviting us to participate in these proceedings today, as

8well as the Berkeley Center for Law and Technology for

9hosting these important discussions.

10I am going to primarily focus my discussions on

11the issues raised by Scott Peterson and Bob Skitol today

12relating to standards setting organizations and the

13diverse membership of those entities.

14I would like to comment quickly on Dave Heiner's

15presentation about the challenges facing in-house counsel

16in addressing antitrust and competition issues in the face

17of disparate regimes that exist in various jurisdictions.

18I think he addressed all of those very well, so I won't be

19focusing on those topics today.

20First I want to give a little bit of background

21about QUALCOMM and its business model. It has recently

22come under fairly close scrutiny and examination and I

23think it's important to understand that in the context of

24where QUALCOMM came from to where it is today as a

25technology innovator and enabler.

70

1 As is fairly well known, the company was founded

2in the mid-80s by several retired professors who had vast

3interest in wireless communications technology Doctors Irwin

4Jacobs and Andrew Viterbi, as well as five others. They

5founded the company in Doctor Jacob's living room.

6After setting up a company, they realized that

7there were ways to vastly improve cellular technology as

8used by terrestrial consumers that could take advantage of

9a lot of work that they had looked into previously, both for

10military and satellite applications.

11To say that their proposals were met with some

12level of skepticism is a vast understatement. There were

13actual nay sayers who said that the technology proposals

14they had would never work and would cost too much. There

15was a professor across the bay at Stanford who actually

16said the proposals defied the laws of physics.

17Notwithstanding the proclamation of violation of

18laws, they actually were able to demonstrate a viable and

19working cellular system based on the technology called

20CDMA, code division multiple access technology. I promise

21not to go into too many technical acronyms today and stay

22on topic.

23But the efforts then following by Doctors Jacobs

24and Viterbi and the others at QUALCOMM to proselytize this

25technology, to find adopters for the technology and the
70

71

1willingness to take the risk of deploying a

2generation-leaping innovation were hardly trivial.

3To seed the industry with the technology the

4company had to develop its own cellular handset business,

5a business that was filled with tremendously large

6multinational corporate players at the time.

7Infrastructure equipment was even more

8complicated. In order to provide CDMA cellular base

9stations for trial systems, they then asked incumbent

10cellular operators to take a risk on a small company in

11San Diego, which was primarily known for surfing and blue

12skies, to trust for the deployment of their next-generation

13networks.

14In the face of all these challenges, the company

15actually did manage to find some cellular operators who

16were facing serious capacity constraints in their analog

17networks at the time and were able to convince them that

18CDMA technology was actually far more advantageous to

19competing digital technology than was emerging in Europe,

20which was GSM technology. And I won't go into it, other

21than to say QUALCOMM had and still has a very firm

22conviction about the superiority of CDMA technology over

23GMS technology.

24As part of having the operators' willingness to

25embrace QUALCOMM's technology, they actually placed a

72

1requirement on the company that the company make other

2vendors of equipment available. There was concern that

3QUALCOMM would not be able to satisfy all of the needs for

4these wireless operators or have anywhere near the skill

5necessary to support the adoption and proliferation of

6these technologies.

7So, very early in QUALCOMM's history, QUALCOMM

8entered into its first licensing agreements. Those were

9with Motorola and AT&T, who at that time were two of the

10largest companies operating in the cellular industry.

11QUALCOMM was a very small company at that time and was in

12a much weaker position with respect to negotiating

13leverage and strength as compared to those larger

14companies.

15As I will discuss a little bit later, it was

16actually those early licensing deals that set the

17framework for QUALCOMM's future licensing activities and

18its efforts in licensing that continue to this day.

19Having succeeded in seeing widespread adoption

20of QUALCOMM's technology, the company very quickly

21determined that it was actually not the company best

22suited to either be in the cellular infrastructure

23business or the cellular handset business. Vast

24manufacturing companies with tremendous expertise were far

25more suited. And, frankly, QUALCOMM didn't prove to be

73

1the most competent manufacturer of these kinds of

2products.

3So, in late '99 and early 2000, QUALCOMM

4actually sold its businesses for infrastructure equipment

5and handsets to companies far more able to run with those

6businesses.

7QUALCOMM did retain its business of developing

8chipsets and software solutions for use in cellular

9handsets and maintained its licensing program, which it

10had started in the very early days through the deals with

11Motorola and AT&T, and through all of the '90s continued

12signing up licensees for manufacturing of wireless

13handsets and infrastructure equipment.

14As a licensor of technology, there are some

15concerns I guess that need to be recognized. It was

16stated today that there are efforts by some licensing

17companies to maximize licensing revenue. And while there

18may be some goal in achieving maximal revenue from the

19licensing side, you have to recognize that in order to do

20that, your downstream licensees, the producers of

21handsets and infrastructure equipment that are paying you

22royalties, need to maximize their sales volumes.

23We're not actually interested in seeing any one

24or two companies maximize their profit at the downstream

25level. We're looking at a downstream industry we want to

74

1see as fiercely competitive as possible to drive price

2reductions and increase volumes. The total revenues

3generated that way will be higher licensing revenues at

4the upstream licensing level.

5So, QUALCOMM's business model from the beginning

6and on the licensing side has been focused on

7proliferation of technology and enabling companies

8downstream to compete aggressively. We are able to take

9our licensing revenues that are generated, pump them into

10an R&D system, with now thousands of engineers producing

11chip and software solutions for use in handsets, and

12continuing development and improvement of the very

13wireless standards upon which our lifeblood depends.

14 We then make these products, our chips and

15software solutions and our patentable inventions,

16available to a very broad downstream industry, which then

17we've seen aggressively competing on introduction of new

18products, new features and rapid price reductions.

19Last year we spent one-and-a-half billion

20dollars on research and development and we also have

21thousands of patents pending patent applications.

22One of the interesting benchmarks we've seen

23some companies use at the handset level, with a different

24view of the universe than QUALCOMM, is their own vast R&D

25expenditures and patenting activities. What they don't

75

1disclose, after suggesting that they spend billions of

2dollars on R&D and have many thousands of patents is that

3they don't make those products available to their competitors

4of handset technology, licensing only that small body of

5patents that they declared may be essential, and even then

6only in some instances.

7I want to turn from the background of QUALCOMM's

8business model to the topic we've been focusing on today.

9The intersection of intellectual property and antitrust

10policies has been looked at closely for many years. It's

11often described as a conflict. But I think most recently

12Tom Barnett at the George Mason conference in September

13made a much clearer statement that strong intellectual

14property protection is not separate from competition,

15rather it is an integral part of antitrust policy and

16intellectual property rights and should not be viewed as

17protecting their owners from competition, but rather

18should be viewed as encouragement to engage in

19competition.

20There's no debate on the incentives to innovate

21provided by a strong patent system. And it's in the light

22of the innovation incentives generated by the patent

23system that I want to speak today. I believe there are

24efforts to consolidate a number of attacks with respect to

25standard setting, legislative challenges, and lobbying the

76

1Supreme Court to undermine the vitality of patents in

2the patent system today. And I think it should be

3recognized that these are primarily not driven by

4so-called desires for transparency of information, as has

5been suggested, but actually is purely an effort to shift

6bargaining power away from patent holders, to drive prices

7down, and which I believe will have the result of actually

8driving innovative companies and patent holders out

9altogether, robbing ultimately consumers of choice and

10opportunities for innovative technologies.

11On the standards side, there are very few people

12that I think would challenge the procompetitive effects

13that standardization can bring. The interoperability

14between many companies' products, welfare-enhancing

15cooperation among many different kinds of firms, increases

16in choice, reductions in costs, broadening the size of the

17markets, all are procompetitive benefits of standards

18setting.

19But one thing that needs to be remembered and

20recognized is that in general the standard setting

21activity is a participation of competitors in a market

22cooperating in a way that needs to be carefully watched.

23The suggestion that you can then take the step of

24technical development, which is the purpose of standards,

25and then move one more step toward collective price

77

1discussions doesn't seem like a very big leap. But it is

2if you look at it from the context of the accommodations

3that have already been made by the antitrust laws and

4enforcement agencies to allow competitive companies to

5work together in concert for their procompetitive

6aspirations.

7I will get to the reasons for concern, but there

8is a risk of undermining the very benefits provided by

9standardization through an anticompetitive result.

10One of the reasons I gave a little more

11background on QUALCOMM than I might otherwise have is I

12think it's important to understand that the benefits of

13standardization do require cooperative industry efforts,

14but that all of the participants in standards setting

15activities don't wear the same hats. In some simple types

16of standards, you may only have participants who are

17producers of products who strictly need to ensure that

18their products all work together. That isn't the most

19common standards activity in QUALCOMM's experience,

20where we find that development standards involve very

21complex technologies, very long-term iterations of

22contributions of technical proposals, a process which

23benefits greatly not just from the participation of the

24end product manufacturers being in the process, but also

25innovative companies, companies like QUALCOMM, who don't

78

1participate in the handset space or infrastructure

2equipment space. But we have a very significant

3interest in seeing optimal wireless technologies developed

4and employed for those industries.

5Now, in the context of the development of

6wireless technologies, we do produce chips and software to

7be used in the downstream products such as handsets and

8wireless modems, but the bulk of our earnings is actually

9driven from our ability to license the technologies that

10come out of the innovations both in the standards settings

11and the innovative research and development.

12So, in addition, you have the manufacturers that

13are clearly interested in developing their products, but

14you also want to have companies like QUALCOMM who are

15primarily motivated by improving and enhancing

16technologies.

17QUALCOMM is not the typical type of company you

18think of in this capacity. Frequently you will think of

19start-ups, sole inventors, universities, other companies

20for whom valuable contributions can be made in advancing

21the technological frontiers.

22Then there are companies that really are hybrids

23or vertically integrated firms, companies who do sell

24significant products downstream, which may incorporate

25their own innovations and the innovation of others, but

79

1who also are contributors of innovation in the development

2of the underlying industry standards. These companies may

3have multiple interests in seeing both technology advance,

4but also assuring that their products benefit from

5early development opportunities.

6Then, finally, we also are seeing more

7participation in the standards setting by a group of

8companies that are ultimately consumers of products. In

9our industry, that would be the wireless operators. They

10have an interest in seeing wireless standards developed

11that meet certain specifications and so they would

12participate in driving technical solutions or technical

13requests for innovators and early participants to try to

14solve.

15It's important to recognize that these diverse

16firms who participate in the standards setting process

17have asymmetrical interests. Innovators who seek to

18promote and advance technology through the proliferation

19of their technology most often receive their return on

20investment in the form of licensing revenues. They don't

21sell products necessarily downstream and aren't able to

22extract any return on their investment from the end

23customers of the standards-implementing products that are

24involved.

25Manufacturing companies, on the other hand, see

80

1their returns on investment coming from their downstream

2sales. In our industry, sales of handsets is a good

3example.

4Again, the vertically-integrated firms, the ones

5that are both manufacturers downstream and those that

6contribute technology in the standards setting process

7have mixed incentives. Now, on the one hand, they may be

8very interested in high licensing costs in order to keep

9their competitors out of their market. On the other hand,

10they find themselves exposed to licensing needs from other

11innovative companies and would like to see low royalty

12overhead in order to drive costs down, recognizing that

13they can recover their investments through sales of

14downstream products.

15All of these business models have their

16advantages and their disadvantages. It's a little obvious

17to state, but I will, the choice made by each company

18should be where its strength lies. QUALCOMM clearly

19demonstrated itself as not having strength in the

20manufacturing of handsets and infrastructure equipment,

21but those businesses were necessary to start the

22proliferation of its technology. Having succeeded at

23that, QUALCOMM quickly divested itself of those businesses

24in order to increase efficiency in focusing on innovative

25developments and making available enabling technology

81

1solutions.

2I guess the caution or concern I request of the

3enforcement agencies is to tread cautiously in making

4decisions that favor one business model over another. The

5risk of driving certain kinds of companies out of

6standards setting bodies probably comes at a societal risk

7that isn't measurable, in that if that company is not

8participating, you don't know what contributions are lost

9and what welfare-enhancing solutions may have been

10foregone.

11There may be some standards where there isn't a

12particularly high level of innovation wanted or needed,

13and in those instances, nothing is lost. And in other

14areas, the need for non-manufacturing companies to

15participate and provide, in some cases, phenomenal

16innovative solutions is something to be encouraged and I

17think guarded carefully.

18One of the points I want to get back to which I

19raised before is the efforts that are going on in a

20variety of arenas today with the stated goal of

21transparency or the stated goals of avoiding certain

22types of hold up, which I am going to address further as

23to whether there really is a serious problem of hold up.

24Recognizing that there are efforts going on to

25rewrite IPR policies in standards setting processes, I

82

1think it is with a pretty simple goal: just to reduce

2costs for technologies included in the standards.

3In order to reach those objectives, a number of

4proposals have been made in a variety of standards bodies

5in the last couple of years. There was an effort recently

6that went on at ETSI where a transparent effort to

7redefine the IPR policy was proposed which would establish

8royalty capping set by the standards body and established

9rules to share royalties on some sort of pro rata

10allocation basis. There was a lot of interesting debate

11that went on regarding those proposals which were

12ultimately not adopted.

13Other approaches call for ex ante disclosure of

14licensing terms. While I appreciate the simplicity of the

15proposals and apparent requests for knowledge, it is

16firmly our belief that either compulsory ex ante

17disclosure of licensing terms, or voluntary disclosures

18with so-called strong encouragement, as some are calling

19it, more than run the risk of resulting in exercises that

20end in collective action. I think it's inevitable.

21If you look at the very basis of standards

22activity, it is about collective action, but for the

23purpose of establishing technical specifications, adding

24cost and price information into the mix would inevitably

25be a factor which leads to collective discussions about

83

1those topics which are not the purpose of standards

2setting.

3Now, it may be that policies are explicit in

4their statements that such things shouldn't happen in the

5context of the standards working groups, but that leaves

6the sort of negative inference that there may be somewhere

7else where such discussions may occur.

8Those arguing in favor of compulsory ex ante

9licensing disclosures typically make three criticisms of

10the present regime: lack of predictability or

11transparency; risk of hold up; and then somewhat related

12to that, the problem of royalty stacking.

13In our experience, the alleged criticisms are

14not convincing and certainly don't prove that it's

15reasonably necessary to scuttle an existing system that

16has actually worked very well in favor of a system that

17brings with it inherent risk of collective price

18discussions, which could ultimately lead to disincentives

19to participate by those who seek to earn their returns

20from licensing.

21The environment that gets created, as I

22indicated, in the standards setting, is one of cooperative

23development. Introducing price information will likely

24lead to efforts of price setting by strong buyers.

25One important thing to understand with respect

84

1to calls for compulsory ex ante licensing disclosure is

2that in fact ex ante licensing negotiations go on today.

3This notion that participants in a standard are unable to

4obtain sufficient information regarding price information

5of technology incorporated in standards are not correct.

6Voluntary ex ante disclosure and negotiation of licensing

7terms on a bilateral basis prior to setting standards

8are entirely consistent with the current FRAND regimes.

9They certainly don't prevent potential licensees from

10asking potential licensors about their planned licensing

11terms and conditions. This isn't a theoretical

12possibility. It actually goes on today and it frequently

13goes on.

14 As I indicated, QUALCOMM's own licensing program

15long predates standardization of any new technologies that

16we worked on in the wireless industry. We consistently

17engaged in licensing discussions before the beginning of

18the standardization process, during standardization, and

19long after, and are well aware that many other companies

20do too.

21There's an argument that it's inefficient for a

22prospective implementer of a technology to ask prospective

23licensors what their licensing terms are. I don't fully

24understand that. The number of prospective licensors is

25typically dwarfed by the number of standards implementers,

85

1and in all but the most complicated technologies there

2aren't that many licenses that need to be negotiated.

3The second criticism is with respect to

4so-called patent hold up. There are a number of

5allegations made about what constitutes patent hold up.

6And I think there is recognition that some activities such

7as intentional withholding of patent disclosures has been

8decided. However, there are those that suggest patent

9hold-up also includes the case where a prospective

10licensor of an essential patent seeks a royalty rate that

11is surprisingly high.

12In reality, licensees frequently claim to find

13licensing rates surprisingly high. It's part of the

14negotiation process. You start somewhere, you end

15somewhere, and that's the nature of the business. There

16are many give-and-takes in the licensing negotiation. So,

17to suggest that the rate information or lack of

18information on licensing terms, which would have been

19readily available if a prospective licensee asked, I fail

20to see how that justifies a need for mandatory ex ante

21disclosure rules.

22Another argument to support notions of patent

23hold up is that essential patents gives a licensor the

24ability to impose unconstrained licensing terms on the

25licensees. And this just isn't the case. You have to

86

1recognize even as a licensor of essential patents, there

2are a number of constraints that exist. There are

3horizontal constraints, constraints about wanting to see

4the market develop downstream, impacted by what other

5competitors are doing in the licensing community.

6Vertical constraints with respect to the licensor and

7licensee.

8As I said, QUALCOMM is a licensor of technology.

9If its licensees succeed, then QUALCOMM succeeds. So,

10imposing onerous or technology-chilling licensing terms is

11not in our interest and it's not a reason to participate

12in the standards setting process.

13And then there are dynamic constraints. The

14development of standards is not a single function in time

15in most cases. The standards continue to evolve. Other

16participants join standards setting groups. And the

17pressures and, shall we say, discipline that come upon

18companies participating in the standard setting process by

19other companies who have a history of not playing by the

20rules is a real threat.

21The final point I wanted to touch on is -- and

22it's closely related to hold up arguments and the way they

23have been used recently -- is the issue of royalty

24stacking. The argument is fairly simple.

25If there are multiple patent holders with

87

1multiple essential patents in a standard, then the

2potential royalty burden that can be imposed on licensees

3may add up to some cumulative amount that's unreasonable.

4First, it's important to recognize that many of

5the companies participating in the standards setting

6process have diverse incentives that I talked about

7before, and subject to the various constraints that I just

8talked about as well.

9In some empirical research that's going on,

10despite the claims of royalty stacking, there have

11actually been very few instances identified. And several

12years ago in the biotech industry, a paper was written on

13the tragedy of the anti-commons in biotech. But twenty

14years later, a paper on the fallacy of the anti-commons

15came out. Royalty stacking is just not something that has

16manifested itself. There is a lot of public rhetoric and

17misinformation that's being spread, particularly in our

18industry, that cumulative royalty rates are going to

19amount to hundreds of percentage points.

20And yet even some of the companies that QUALCOMM

21is fiercely at odds with have publicly stated that they

22don't think that anybody is paying double digit rates.

23And there are a lot of factors to explain that. There's a

24lot of cross-licensing that goes on. A lot of companies

25maintain patents for defense purposes. There are many

88

1dynamics that work together that result in the limiting of

2royalty stacking despite the sort of argument that if

3there's lots of patents, there's lots of royalties.

4So, the proposals for compulsory ex ante that

5are being proposed are being proposed to fix a problem

6that either doesn't exist or certainly doesn't exist in

7the widespread extent to which it has been attributed.

8And the fact is these proposals run severe risks of

9driving anticompetitive results and provoking the

10elimination of innovators willing to participate in

11the process.

12 There were a few comments that Bob made on the

13efforts of VITA to revise its IPR policy that I feel I

14ought to respond to. Having the benefit of going last and

15having heard them, I will take that opportunity.

16 As I said, there may be standards in which

17fairly low technology proposals are made. Complete

18solutions are brought in by each company and they're

19weighed on their respective merits and a selection among

20them is made.

21And I don't profess to know much about what VITA

22does or its technologies. I've read descriptions that

23it's focused on plugs and connectors and bus signaling

24protocols. And I don't know the level of significant

25innovation that goes on in those areas, but it may in fact

89

1be an organization in which little harm would be done in

2the face of compulsory disclosure of cost information with

3technical solutions.

4But the notion that such a solution would fit

5all standards is deeply concerning. One size doesn't fit

6all. I think in the vast majority of cases such a

7disclosure regime will actually lead to the things that

8I've expressed concern about, which is that there will be

9collective discussion of price by large groups of

10purchasers who produce product for the downstream market,

11leading to some form of concerted purchasing power, the

12end result being the driving out of innovative companies

13who seek a return on investment based on licensing.

14And I do note that a significant founding member

15of VITA, very soon after the passage of the approval of

16the policy by the board, withdrew its membership from

17VITA. That company is Motorola, who I think is one of the

18more innovative companies in America today.

19The advice -- not advice, but the request I

20would make of the enforcement agencies when asked to look

21at revisions to IPR policies, and Bob's suggestion would

22actually encourage such guidance, is to pay particular

23attention to the facts and circumstances that exist in

24each situation.

25Efforts should be taken to avoid taking as

90

1gospel allegations of hold up and royalty stacking. The

2evidence isn't there. And there's a lot of research

3coming out now in the last year combatting -- addressing

4these many years of literature that's stated sort of the

5contrary.

6I will submit a bibliography with some notes

7that can be included on the FTC's website identifying some

8of the recent efforts to challenge these premises with

9robust analysis.

10Thank you.

11(Applause.)

12MR. COHEN: Well, we're a bit behind on our

13schedule. We had talked about doing a 15 minute break. I

14suggest that we take about two or three minutes in our own

15seats to give us an opportunity to stand up, then we're

16going to go forward so we can try to get as much of a

17moderated discussion as possible. So, in about three

18minutes I'm going to start again.

19(A brief recess was taken.)

20MR. COHEN: I am one of the belief that with the

21panel as the meat and any questions that we have as the

22gravy, we're going to try and get as much of the meat as

23we can.

24And probably the way to do that is to divide

25our remaining time into two segments. One will be more of

91

1the general issues which David Heiner was so good to

2raise. And following that, the other segment would deal

3specifically with some of the standard setting issues that

4have been discussed already.

5And what I'd like to do is begin and see if any

6of the other panelists have comments or responses to

7anything raised by David in particular, because you get a

8chance to respond to the standard setting issues in about

915 minutes.

10Anything you want to say? No? Okay, then I

11will pick some questions to get you going.

12 You all have been people who have received or

13watched others receive over the years antitrust counsel of

14the various kinds of single-firm conduct.

15 I'm wondering if anything strikes you as having

16been an area where advice or the legal tests that you're

17trying to articulate has been particularly easy to

18understand or particularly difficult to understand, any

19recurring problems that you're facing?

20MR. SKITOL: I will take a shot.

21In my experience over the last couple of years,

22I think the single most difficult area of Section 2 law to

23advise on has been the loyalty rebate and bundled pricing

24area. And you had an excellent panel on that subject a

25couple of months ago, with a number of competing

92

1suggestions for what the standards should be.

2It's a tangled mess. It's been a tangled mess

3in particular ever since the LePage's decision. And the

4world is divided between those who think Lepage's is about

5the right approach and those who think it isn't.

6It's extremely difficult to give clear advice to

7business people on what kinds of loyalty discounts are and

8are not okay, what is the legal standard.

9And so I would certainly urge special attention

10and priority to the agencies in giving advice to the

11courts because this is an area that's gotten terribly

12muddled, not because of anything the government has done

13but because of conflicting decisions in private

14litigation.

15MR. HEINER: I would agree with Bob that that's

16a pretty tough area and one that I think gets all the more

17challenging when you overlay the European focus on top as

18well, as articulated in the Draft Article 82 Discussion

19Paper.

20More broadly to your question, I think I'd say

21that it's a clear divide between Section 1 and Section 2,

22where the Section 1 counseling is pretty easy, frankly,

23and Section 2 is pretty hard.

24 MR. HARTOGS: I will agree that the issues on

25joint conduct out participation and cooperation, I

93

1think is fairly clear. I particularly echo the sentiment

2about needing some measure of global harmonization in knowing

3what the rules are for multinational companies

4participating with other multinational companies in the

5face of enforcement agencies and regimes in which they are

6not in agreement on an application of a particular

7standard.

8We find ourselves trying to determine what is

9the most restrictive set of rules under which we should do

10our analysis and guide our conduct.

11MR. COHEN: Okay. That leads me to some

12questions on the international situation.

13We just had one view of trying to find the sort

14of the least common denominator. Have you found that your

15businesses -- in general, have you tried to decentralize

16to adapt to local competition rules, or do you find that

17most of you are being forced in one way or another to fly

18with the most restrictive laws potentially applicable to

19you in different jurisdictions?

20 MR. HARTOGS: I think, unfortunately, localizing

21is an idea that wouldn't work for us. We develop product

22in the U.S., Europe, India, Korea and Japan. We sell

23products to companies everyone in the word. They sell

24their products further downstream everywhere else in the

25world.

94

1Agreements with respect to various related

2entities with affiliates that are not U.S. entities

3probably render it still necessary to look for the most

4restrictive set of rules in guiding our conduct.

5MR. COHEN: And we heard from Microsoft that

6some of these -- the way this works with licensing.

7Did similar issues arise with regard to your

8contract practices?

9MR. HEINER: It's very much a global business.

10So, the answer is kind of the same as what Mike was

11saying. We have looked at whether in particular cases you

12can try to localize the business practices to the local

13jurisdiction. The issues that come up are mostly not

14around local facts, however. It's not as if the issue is

15relations with a retailer in any particular country. The

16issue, rather, is of a global nature, what is the design

17of Windows around the world, what is the licensing

18paradigm of Windows around the world?

19And so we do find ourselves kind of looking to

20what's the most restrictive set of rules. And that's what

21we have to adhere to.

22We have given some thought to whether it would

23be possible -- notwithstanding the costs that it would

24entail -- would it be possible to have different

25products, different licensing plans in one part of the

95

1world versus another. And it may come to that some day.

2But if it does come to that, it would certainly be with a

3certain loss of efficiency, and for customers as well.

4MR. COHEN: Bringing us back to the United

5States, one of our concerns at this hearing is to find out

6the degree or whether, and if so the degree to which,

7uncertainties about antitrust analysis of single-firm

8conduct have been chilling potentially procompetitive

9conduct.

10We heard some examples and a discussion of that

11in David's talk this morning.

12Have any of you others found similar experiences

13where business practices that may have been beneficial to

14consumers have been put on hold because of uncertainty

15about antitrust exposure?

16MR. HARTOGS: I guess I would just quickly say,

17Bob's comment before that guidance on pricing is

18particularly difficult where you lack clarity here, you

19lack clarity in Europe. And again not having sort of

20flexibility to always choose what may be the most price

21friendly, consumer friendly result, is a risk.

22MR. SKITOL: There are lots of situations

23involving Kodak aftermarket kinds of issues. We've all

24been living with the difficulties of Kodak aftermarket

25Section 2 as well as Section 1 problems for fifteen

96

1years now. There are lots of situations I find where a

2client has in mind doing X, Y, Z with its consumables,

3which would be of significant consumer value, would

4enhance the product, and it looks great. But because

5of Kodak and all of the law that's built up around it,

6this is problematic, and Trinko doesn't do that much to

7help. There is hesitation and sometimes desirable

8developments are canned because of concern about what

9aftermarket rivals might be able to stir up by way of

10mischief about it.

11I think the whole Kodak aftermarket area is one

12that could benefit from agency guidance. Where are we on

13legitimate versus illegitimate aftermarket practices

14fifteen years after Kodak and three years after Trinko?

15Because the courts in private cases still don't get it

16right. We still have not gotten the rules.

17MR. COHEN: And just per a request for more

18agency guidance, guidance can take different forms. And

19because of time constraints, I'm going to throw three of

20them out at once and see how you react to them and see if

21they're suggestions you might want in one of these areas.

22Guidance can take the form of explanatory text

23such as we often give through reports on hearings and some

24business review letters. It can take the form of safe

25harbors, which can be announced. And it can take the form

97

1of presumption. And we heard one suggestion for

2presumptions today about conduct that's used by firms with

3particularly great market power in competitive situations.

4Would any of these three forms be particularly

5useful to you? Do any of you have ideas of things that

6you would like us to provide in any of these areas?

7MR. HEINER: I think all three can be very

8helpful. With respect to the text, of course it depends

9what the text is.

10MR. COHEN: Right.

11MR. HEINER: There's always the possibility of

12obfuscation instead of the intended fact. As one of my

13colleagues pointed out to me before I came down here

14today, we could have very predictable antitrust law in a

15way that wouldn't be at all favorable to our firm. That's

16the risk as well, I suppose.

17MR. COHEN: Beware of what you ask for because

18you might not like it when you get it.

19I guess I should ask questions directed to the

20other side of things, too.

21We looked at the chilling as procompetitive

22conduct. But do any of you have issues which you haven't

23already touched on in which conduct involving dominant

24firms has hurt you and that you think the agency should be

25looking at but hasn't been paying full attention to or

98

1much -- close enough attention to that might be desirable?

2Anybody have anything in that area? Already

3touched on.

4Okay, let's go to the standard setting area.

5And I think probably the way to begin would be to give an

6opportunity for Scott and Bob to offer responses to what

7they've heard. We had a response to them, so I guess you

8should have a rebuttal opportunity. And we'll probably

9open it up to a third rebuttal as well.

10MR. PETERSON: I am going to yield my time to

11the agencies. I'd much rather hear your questions.

12MR. SKITOL: Can I just make a couple of

13comments? I listened closely to Michael's discussion

14about the QUALCOMM business model and the importance of

15there being respect for diversity of business models and

16that there shouldn't be a thumb on the scale against one

17business model in favor of another. I agree with all of

18those points.

19I think from the standpoint of an organization

20like my client VITA, from the standpoint of anybody who

21supports open standards processes, competing business

22models are good. But that's on the assumption, on the

23premise, that all of the competing business models should

24play by the same free-market rules and the same transparency

25rules. All business models are subject to the same

99

1antitrust laws. No business model should be imposed on a

2group of standard setting participants.

3It's good if all of the cards are up rather than

4down. It's good for standard setting participants to have

5choices. It's good for standard setting participants

6sitting around in a working group with multiple possible

7solutions to the specification writing, one of which may

8well come from a business model that emphasizes licensing

9revenue, and another comes from a business model that

10enables the solution to be offered royalty free. It's

11good to have that choice as long as everyone knows what

12the respective costs are as well as what the respective

13differences in quality and performance will be. And then

14performance-cost tradeoffs can be collectively made and

15there can be informed decision-making. That's all to the

16good.

17So, those of us who believe that ex ante license

18terms disclosures and similar transparency policies are

19good are not anti-licensing business models. We're not

20anti-patent. We are pro free market, pro choice.

21MR. COHEN: Any rebuttal?

22MR. HARTOGS: To the extent Bob agreed with me,

23I don't have any comments.

24On the -- just a cautionary comment. In his

25talk he suggested that the next step actually ought to be

100

1a sanctioning for group discussions. And I do believe

2that the ultimate result of that would be a chilling of

3willingness of participants in the standard setting

4organizations who do rely on licensing.

5I think it should be recognized that the bulk

6of participants in standards setting activities are

7prospective licensees and the impact the proposed changes

8can have is on more than transparency, but directed toward

9driving pricing down where there is no return on investment.

10That is something that needs to be watched and watched

11carefully.

12MR. HEINER: One time on this.

13I think all of the speakers on this topic

14identified the threshold question of how great a problem

15is it this so-called hold up problem.

16And from Microsoft's perspective, and we're a

17company that's involved in dozens, I am sure hundreds, of

18standard setting endeavors, and from our perspective, we

19do not have a business model of really trying to make any

20significant revenue licensing of IP into standards.

21In our experience in participating in standard

22setting bodies, we really have not experienced these sort

23of hold up situations in standards that we wish to

24implement in Windows and Office and other products. And

25these products do implement huge number of standards.

101

1So I offer that comment on the extent of the

2problem we had about weighing against the collusive kind

3of risk that [unintelligible].

4MR. COHEN: You called that a threshold

5question, but it was my first.

6Let me direct toward the end of the table,

7anything you might want to say as to the frequency of hold

8up? I know you have identified four instances within

9VITA. But how about the consideration that reputational

10considerations and a desire to see downstream success of

11the product is going to put a real limit on the likelihood

12of hold up activity?

13MR. PETERSON: So, yes, I think my discussion

14earlier about patent mobility goes directly to that point.

15And that decades ago where there was more stability in a

16particular industry and much less patent movement, those

17kind of reputational effects could have been more valuable

18than they are likely to be in the future because the fact

19is that patents have become separated from the reputation

20that once was associated with them and thus that constraint

21is no longer as strong.

22MR. SKITOL: I would just add a comment that the

23interest in growing the market and in the market being

24successful is a factor in any monopoly, any monopolization

25case. Every monopoly has its limits. A monopoly price

102

1which is not limitless. It's got a limit.

2So, in this respect, Section 2 monopolization

3through patent hold up is no different than Section 2

4monopolization through any other kind of predatory

5conduct.

6MR. COHEN: Let's lead into some of the

7predicates for the ex ante disclosure rules. I guess

8there's some other alternatives to that which I'd like to

9get reactions to first.

10I'm wondering whether a mere disclosure of

11relevant patents, not disclosure of licensing terms,

12followed by an opportunity for bilateral ex ante

13negotiations would be sufficient? Why or why not?

14MR. SKITOL: The point made about bilateral

15negotiation is always out there and possible. That's

16inviting secret behind closed doors bilateral special

17deals between the big guys at the expense of new entrants

18and smaller players.

19Why isn't it preferable to do the negotiation

20out in the open as part of the open standards development

21deliberation process itself that is available to all

22parties that want to participate? After all, this is all

23in the context of the traditional RAND commitment which

24has a nondiscriminatory as well as a reasonable component

25to it.

103

1So, the idea that we should stay away from more

2transparency for everyone because we already have

3bilateral opportunities, it doesn't make sense.

4MR. HARTOGS: I guess in answer, what you

5describe actually is the system that does exist today

6about disclosure and bilateral negotiations. And it's

7worked well. We had descriptions relabeling of things

8today as hold up, which wouldn't have been viewed as hold

9up previously.

10I didn't hear any suggestion about

11discrimination being part of the motivation of

12licensors prior to the discussion. But to the

13extent that companies are committed to licensing on a

14nondiscriminatory basis, there are structural remedies and

15opportunities to fix abuses there as well.

16So, I don't see how ex ante disclosures of

17licensing terms and collective negotiation or licensing

18agreements fixes that. As indicated before, the large

19number of potential licensees for any essential patent will

20greatly exceed the single licensor.

21MR. COHEN: I notice you talk about the

22nondiscriminatory aspects of RAND. Let's focus on the

23reasonable for just a moment.

24What's the feeling of the panel as to whether

25that has a well-defined meaning? And to what degree has

104

1arbitration procedures of the type that VITA has talked

2about been applied in the past? We have a history to go

3on as to whether this is really successful in resolving

4disputes in the area.

5MR. SKITOL: Nobody knows what RAND means. I

6defy anybody on this panel to tell us what reasonable

7means and what the standard for it is. It's a meaningless

8term that facilitates deception and facilitates hold up

9for the very reason that it fools everyone involved into

10thinking that it's a real limitation on what the patent

11owner will do, when in fact it isn't.

12MR. HARTOGS: I would strongly disagree. If you

13look at the origins of IPR policies that call for RAND

14declarations, the purpose is directed at eliminating outright

15refusals to make licenses available for patents that

16become essential for standards.

17What RAND intended is an important flexibility

18that recognizes that licensors and licensees are almost

19always differently situated. And having the ability to

20bilaterally determine mutually agreeable solutions that

21satisfy both is probably the best test of reasonableness.

22In some cases you might be able to look to

23pre-standardization licensing activity. I am not suggesting

24that there will always be circumstances where we can point

25to ex ante licensing results as a benchmark to compare to

105

1post standardization licensing to demonstrate reasonableness

2or at least confirm that standardization didn't lead to a

3change in licensing terms. Certainly they do exist. In some

4cases and when they do exist, they seem a fair benchmark as to

5establishing reasonableness.

6MR. COHEN: Moving now to the idea of ex ante

7disclosure of relevant terms, you need to tie this of

8course to perhaps essential patents under the standard,

9some concept along those lines.

10I'm wondering if anybody has a sense of what the

11impediments are to giving meaningful -- to even

12identifying in advance what's likely to be in a standard

13and what's likely to evolve out of the patent application

14process in order to determine what you have and before you

15can explain what the terms would be on it. Anybody want

16to comment?

17MR. PETERSON: Yes. So, it may be an evolving

18thing over the course of a standard. It shouldn't be an

19expectation that this is something that should be known up

20front.

21On the other hand, people are making judgments

22about other aspects in the standard on an ongoing basis,

23and this is information that ought to be brought forward in

24that same spirit -- as it becomes apparent what will be

25needed, information will be made available about it.

106

1And, I'm sorry, there was another point I was going to

2make.

3Well, I'm sorry, go ahead.

4MR. HARTOGS: So, I think it's an important

5question because it goes back to my comment that the

6proposal for VITA's policy may well work for VITA. But

7that if you look at our experience in some very complex

8wireless standards, there are multiple years of

9development, multiple iterations of contribution of

10technology and of innovation. And being forced to place a

11stake in the ground from which you can't retract your

12position or change it, it really is an important timing

13question as to when you would do that. You have to make

14an assumption on sort of the most optimistic view about

15how successful you have been in providing your innovations

16in developing the standard, and then make your proposals

17based on that, on the assumption that if you have

18something more valuable to contribute, you no longer

19retain the right to price that effectively.

20MR. PETERSON: The thought that I was missing a

21moment ago is that one doesn't necessarily always have to

22wait until a patent is matured into a patent, or even a

23patent application, because it's often in many cases

24possible to make judgements about what one's licensing

25intentions would be, even not knowing what the particular

107

1patents might ultimately be, because the judgments are in

2many cases informed by other factors.

3MR. COHEN: Assuming now that we've reached the

4point that we're talking about some form of ex ante

5activity that type of term requirement.

6Perhaps the requirement of disclosing terms may

7be at one end of the spectrum. You might then go a little

8farther and have some provision for discussion or

9clarification of the term, sort of at the middle of the

10spectrum. And then go all the way to the far end and

11actually have clear joint negotiation of the term.

12Does anybody see -- or could you give your

13thoughts on whether going beyond mere announcement of the

14terms is necessary? What are the considerations?

15MR. PETERSON: So, I think there will be

16different -- this is an area where there should be

17diversity and variety could be explored. So, I think

18there may be certain kinds of product or technology areas

19in which the exploration of the license term issue might

20profitably go farther than in others. In others, it may

21be that very little needs to be done. It may be simple

22disclosure needs to be done.

23So, I think there is a variety -- there are a

24variety of different kinds of cases. Some are worthy of

25more detailed attention than others.

108

1One thing I would point out is that, if there is

2a perception of this cliff that you step off of after

3disclosure and that if you embark on anything beyond

4disclosure that there's some kind of interactive

5discussion is a very serious matter, then that chills even

6the value of the disclosure.

7So, I think, although I see the need for the

8more collective action regarding the terms as being

9perhaps very much the unusual case, to say that -- to make

10it clear that's it's only disclosure which is

11procompetitive and the discussion of the terms is a high

12risk activity, it has that chilling effect. As I have seen

13already in organizations that have been toying with

14introducing more consideration of license terms, the

15idea -- the steps that they feel they need to take in

16order to assure themselves that nobody will ever talk

17about them is seriously chilling just that first step

18about getting information made available.

19MR. SKITOL: I think the time has come to

20recognize that a lot of the information technology and

21communications technology standard setting processes that

22we are talking about are really indistinguishable from an

23antitrust analysis standpoint from all kinds of joint

24product development, joint technology development

25ventures. That's essentially what this kind of standards

109

1development activity is. It is a group of companies

2getting together, combining their resources and their IP

3and collectively developing something new.

4It is standard joint venture law today that when

5you have a lawful joint venture, it is lawful for the

6participants in that venture to make collective decisions

7about which input to buy for this and which input to buy

8for that. There are collective decisions and collective

9negotiations over cost as well as other features of one

10versus the other. That's what standard setting is about

11today.

12Now, there could be lots of situations where the

13result of ex ante license terms disclosure is that the

14parties sitting around the table in the working group

15recognize that they've got two main good proposals. One

16comes with a two percent royalty disclosure and the other

17comes with a five percent royalty disclosure. And they

18all agree that the latter is technically superior to the

19former, but five percent is too much to pay.

20What is wrong with a non-coercive negotiation

21process, arms length process, in which the group

22collectively discusses with patent owner B that we really

23prefer your solution, we would go with your solution if

24you could reduce that rate somewhat. And if that patent

25owner decides to do so, to go ahead and accommodate that

110

1interest, then what's wrong with that? That's an arm's

2length decision and everybody ends up all the better for

3it except for the solution A guy whose solution ends up

4being excluded. But exclusion of one or the other is

5inherent in the process.

6MR. HARTOGS: I'd like to comment on two points.

7One, I think the joint venture analogy breaks

8down when you look at the sort of absence of certain kind

9of participants you want involved in standard setting.

10You wouldn't typically have the nonproduct companies such

11as the universities. You may engage them to do contract work,

12but the kind of joint venture activity you're suggesting is

13very different from the standard setting, where in fact

14your very customer may be a participant in the standard

15setting process. In the joint venture context, you wouldn't

16condone discussions collectively with our co-developers with

17respect to dictating the price that each can ask of its

18customers.

19On the collective discussions that aren't

20diversified, I had trouble sort of parsing that because I

21think the effect is going to be exactly what I suggested

22that we would fear, which was a shift to strong buyer power

23by a much larger group of prospective licensees. It may be

24that in an idealized simple A versus B scenario where

25there are pure substitutes available, and it really is

111

1distinguished on price, there may be an effect that selection

2of one over the other will be determined by pricing and

3it's a fair discussion. But the reality is that in none of

4the groups that I am familiar with do such black and white

5distinctions arise in practice. There's always

6tradeoffs on performance, abilities, time to market, and

7costs being one additional factor, but one additional

8factor that if pressed would lead to potentially alienating

9the very participants making the proposals.

10MR. HEINER: I guess I too wonder if the joint

11venture analogy is really right. In a joint venture

12context, the parties to a venture are not competing with one

13another. That's the essence of it. Whereas in the

14standard setting context, the implementers typically will

15be competing with each other in the implementation of the

16standard. And that's very important.

17So in one you're trying to preserve competition

18and in the other you're not. In the standard setting

19area, as you said, Bob, it's already something that raises

20some concern in antitrust law since it's essentially a

21group of firms coming together and agreeing on how

22something should be done, rather than competing about how

23that should be done. So, I think there is a legitimate

24risk here.

25I could then take it to the next level and say,

112

1let's also have discussions about agreeing on pricing of

2the technology that is the input to that standard.

3 MR. SKITOL: Well, see --

4MR. PETERSON: Let me respond to that.

5So, the pricing discussion that would be -- that

6should be undertaken is only that pricing discussion that

7is related to the cost of where they have agreed they're

8not competing. So, in fact these are competitors as to

9products which include implementations of standards. But

10as to the standard, they're not competing. That's what the

11exercise is about.

12And I think too -- it's important to

13realize that the decision to select the standard is the

14relevant decision to which the price needs to be a factor.

15And to suggest that the price can somehow efficiently, in

16a market sense, be determined later is -- you know, the

17prices of products, the prices of other cost components

18will absolutely need to be determined later -- but the

19decision on what this particular feature will be is being

20made collectively.

21And if that was not a procompetitive thing to

22do, then that's a problem. There is a collective choice

23of a particular thing where there will be no competition.

24And it's entirely appropriate to consider the full

25economic scenario of what will be the costs associated

113

1with making that.

2MR. HEINER: That is a little bit of a strong

3statement because you may often have standards competing

4with one another.

5MR. PETERSON: And I agree. I make it a strong

6statement in the extreme case. But there are a range.

7But in the case where there is lock-in, yes.

8MR. COHEN: Well, let me see if there's a

9consensus on that.

10The joint negotiation could in theory represent

11al la monopsy with effects that might impede innovation

12incentives.

13MR. SKITOL: Well, that is a potential problem

14that should be recognized but would rarely occur in the

15real world. It's an antitrust problem only to the extent

16that it would have the likely effect of reducing output or

17reducing innovation, and that's a real stretch.

18I would refer you to the extensive discussion on

19the monopsony issue in Sony versus Soundview, where I think

20the district court got it about right and made it clear

21that the plaintiff's attack on the collective negotiation

22that went on in that case involving the consumer

23electronic players that the viability of the attack, the

24antitrust claim against the collective negotiation that

25occurred there, would depend on a showing of actual output

114

1restraint or reduction. And it's a real stretch.

2To my mind, it's a potential anticompetitive

3effect of small likelihood, balanced against a major

4procompetitive benefit that is very likely to occur in

5many circumstances where negotiation would occur.

6MR. HARTOGS: I probably already answered this

7question. I clearly view that not only is it not a rare

8occurrence, but it would be a frequent occurrence and

9potentially one debilitating to the willingness of some

10companies to participate in setting the standards.

11To the extent that ex ante licensing already

12does occur in certain instances, there's no prohibitions

13on seeking licensing terms on a bilateral basis prior to

14the setting of a standard. It does occur. When we look

15at ourselves, we actually do provide transparencies to

16all of the companies in the industry that we deal with.

17We do deals ex ante, as probably many do.

18 MR. PETERSON: So, on this point, again, all

19that we're talking about is a discussion of the cost of a

20choice which is going to be made. And in fact to decide

21the price of that later is not to postpone competition,

22but in fact to make the choice without it having been

23informed by the price information. So, in other words,

24the idea that there is some -- the choice is whether or

25not a particular technology is going to be collectively

115

1decided to be put into the standard or not.

2If the owner of the technology doesn't like the

3price, at the end of the day, they can walk away at that

4point. In other words, that's the power of the patent.

5The patent has the power to be able to say, this is what I

6have to offer. And so that's their walk-away opportunity

7after the standard has been set.

8The flip side walk-away opportunity, it seems,

9in this event, is a collective one. And it in general

10would be procompetitive because the value of creating

11these standards is so useful. But it is a collective

12event and it should include the economics associated with

13it.

14MR. COHEN: You touched on my last question,

15whether there's an ability of the patentholders to

16discipline a standard setting organization which too

17aggressively pursues a price negotiation by either

18withholding its technology or entirely leaving the

19standard setting organization.

20MR. SKITOL: On an ex ante basis, everybody has

21got choice. The participants who are the potential

22licensees have choices, but the patent owners who would

23like to see their patented solutions adopted also more often

24than not have choices.

25So, if there's any dissatisfaction with what the

116

1willing buyers seem willing to pay, then those patent

2owners have the ability to go off and productize their

3technology on their own or find some way to turn it into a

4proprietary standard.

5MR. COHEN: Is that realistic?

6MR. HARTOGS: I think it's rarely realistic.

7There are scenarios. I look at Motorola's now withdrawal

8from their participation with VITA. But where you have

9an organization like IEEE where you have such a broad

10spectrum of standards and technologies, that viability of

11not participating, not being a member severely handicaps

12your ability to participate in business for the technologies

13they address.

14 MR. PETERSON: I think this is an area where we,

15as we said before, have many different experiences going

16forward. There will be different sets of rules explored

17and we'll develop experience with that in going forward.

18In the past we had something that was a fairly

19extreme policy, the W3C introduced a policy that requires

20royalty free -- a royalty free result in a sense that they

21don't want to issue a standard to which they're aware

22there's some non-free patent. And the world has continued

23to work with that. I don't think that that approach

24applies to a wide range of other technologies, but that's

25an example of where I think we need to try some things to

117

1see where we actually stand.

2MR. COHEN: Unless I hear an objection from any

3of my panelists, I think we've covered the topic.

4I want to thank all of you for your interesting

5and insightful remarks. And I'd like to encourage the

6audience to join me in a round of applause for our

7speakers today.

8(Applause.)

9MR. COHEN: Our afternoon session will begin at

102:00. There's going to be a speaker luncheon at the

11Berkeley Women's Faculty Club. Thank you.

12(Whereupon, at 12:46 p.m., a lunch recess was

13taken.)

14

15

16

17

18

19

20

21

22

23

24

25

118

1AFTERNOON SESSION

2(2:10 p.m.)

3MS. GRIMM: Good afternoon everyone. I would

4like to welcome you all to this session of our business

5testimony hearings and I'm glad that you all could join

6us.

7I am Karen Grimm. I am Assistant General

8Counsel for Policy Studies at the Federal Trade

9Commission, and I am also one of the moderators of this

10session.

11My co-moderator is Joe Matelis, who you met this

12morning, an attorney in the Legal Policy Section of the

13Antitrust Division, U.S. Department of Justice.

14Before we start, I just have to cover two

15housekeeping matters. As a courtesy to our speakers,

16please turn off your cell phones, your Blackberries, any

17other devices you may have. And also we request that you

18not make any comments or ask questions during the session.

19We are honored to have a distinguished group of

20panelists from the business community with us this

21afternoon. They are, in order, Thomas McCoy, who is the

22Executive Vice President of Legal Affairs and Chief

23Administrative Officer at AMD; Michael Haglund, who is a

24partner in Haglund Kelley Horngren Jones & Wilder in

25Portland, Oregon, and counsel to Ross-Simmons, the

119

1Weyerhaeuser -- and counsel to Ross-Simmons in the

2Weyerhaeuser/Ross-Simmons predatory buying case; finally,

3we have David Dull, who is the Vice President of Business

4Affairs, General Counsel and Secretary of Broadcom

5Corporation.

6Our format this afternoon will be essentially

7the same as this morning's. Each speaker will make a 20

8to 30 minute presentation. After the presentations are

9finished, we will take about a 15-minute break. And after

10the break, we will reconvene and have a moderated

11discussion with two of our panelists. Unfortunately,

12David, who has a scheduling conflict, will not be able to

13join us for the roundtable discussion. We are, however,

14very grateful that he is still able to participate as a

15presenter here this afternoon.

16As Bill Cohen said this morning, these business

17sessions are an extremely important component of the

18Section 2 hearings overall. Over the last seven months or

19so, we have held conduct specific hearings on predatory

20pricing and buying, refusals to deal, tying, exclusive

21dealing, bundled and royalty rebates and discounts, and

22misleading and deceptive conduct.

23Some of these prior panels have included

24business executives or their in-house attorneys who are

25typically heavily involved in the company's business

120

1decision-making processes.

2The sessions today are designed to further our

3goal of obtaining as much real world insight as possible

4into Section 2 issues from a business perspective and

5basically from business executives and their counsel.

6To that end, we have invited our business panel

7to address whatever Section 2 issues they consider

8important to their respective businesses and to share with

9us any views they may have on how we at the FTC and the

10Justice Department can better address those issues from an

11enforcement perspective.

12We heard a number of helpful suggestions this

13morning. We look forward to our panelists' remarks in the

14roundtable discussion this afternoon.

15I want to thank all of today's panelists for

16their participation. We appreciate all of them taking

17time out of their very busy schedules to prepare for and

18participate in these hearings.

19I would now like to turn the podium over to my

20DOJ colleague and co-moderator, Joe Matelis, for any

21remarks he would want to make.

22MR. MATELIS: Thank you, Karen.

23I just have brief additional remarks to make in

24addition to what Karen said.

25On behalf of the Antitrust Division, I just

121

1want to thank the Berkeley Center for Law and Technology

2and the Competition Policy Center at the University of

3California Berkeley for hosting these hearings today.

4And also on behalf of the Antitrust Division, I

5want to thank all of the panelists for volunteering your

6time and sharing your insights with us

7And finally I'd like to thank Karen and her

8colleagues at the FTC for all of their hard work in

9organizing this hearing and assembling such a fine panel.

10MS. GRIMM: Our first speaker today is Tom

11McCoy. He is Executive Vice President of Legal Affairs

12and Chief Administrative Officer of AMD. Tom joined AMD

13in January 1995 and was Senior Vice President, General

14Counsel and Secretary until 2003.

15Tom's current leadership responsibilities

16include legal, business development, employee

17communications, international policy, government and

18community affairs, corporate secretary, environmental

19health and safety, and global real estate. He's busy.

20Mr. McCoy holds an undergraduate degree in

21history from Stanford University and a law degree from the

22University of Southern California.

23Prior to coming to AMD, Tom spent 17 years

24practicing law at O'Melveny & Myers, where he specialized

25in business litigation. Tom.

122

1MR. McCOY: Karen, thank you very much. And

2thanks to everybody in the room. Thank you for having me

3here today to share my thoughts and experience on this

4very important topic. I'm particularly pleased to join my

5fellow representatives in the technology industry in

6presenting here today.

7I believe our presence is a testament to a

8common belief in the critical role that enforcement of

9Section 2 plays in ensuring innovation and competition in

10high technology sectors

11Technology is often cited, and I believe

12correctly so, as the driver of our new economy in a

13rapidly globalizing world.

14As Federal Reserve Chairman Ben Bernanke

15emphasized in a speech just last August, the innovation

16that technology companies produce spurs economic growth

17and innovation, not only within the sector itself, but

18outside the IP sector as well. His remarks cited numerous

19economic studies demonstrating that information technology

20was the single greatest impetus for the tremendous rise in

21productivity our national economy experienced in the late

221990s.

23So, what was behind the innovation surge? Ask

24economic experts and their answer is simple:

25competition. And, not coincidentally, more competition in

123

1the microprocessor market, which produces the brains of

2computers.

3I've been with AMD for over a decade now and I

4was a business and antitrust lawyer for nearly twenty

5years before that, as was mentioned. I believe the

6competitive dynamic within the microprocessor market provides

7a particularly important example as we discuss Section 2.

8Look at the late 1990s and the impact of the

9speed of innovation in this market, before and after AMD

10transformed from a second source follower to an innovation

11leader. As Professor Michael Scherer testified in an

12earlier hearing, that difference was dynamic. When

13competition arrived, the pace of innovation quickened.

14But as the Japanese Fair Trade Commission ruled

15in 2005, that innovation of AMD did not go unpunished.

16Because of the critical importance of the

17technology sector to the strength of our national economy,

18there is perhaps no market in which the committed

19enforcement of antitrust law and competition policies are

20more important.

21But if we are to do so effectively, we must

22first dispel the most common myths about the technology

23marketplace. Namely, myth number one: Market power is

24inherently transient in high tech industries. Myth number

25two: Section 2 is not equipped to deal with the special

124

1characteristics of high tech markets. And myth number

2three: Consumers are not harmed if technology solution

3prices are coming down.

4The fact is, each of these myths is simply wrong

5and acceptance would stand in the way of fair and open

6competition in technology markets. Indeed, these myths

7would empower a monopoly of use and consumer harm of the

8very kind Section 2 is intended to stop.

9Accordingly, we must rigorously consider how

10firms, and dominant firms in particular, actually behave

11in real markets. When we do, we will discover the

12provable truths that should inform this discussion of

13Section 2.

14So, allow me to address these myths one by one.

15First, myth number one, market power is

16inherently transient in high technology industries. The

17truth? In many high tech industries, just as in low and

18no tech firms, customers are tied to the dominant firm for

19a very large percentage of their requirements, at least in

20the intermediate term. With their customers at their

21mercy, dominant firms can use a combination of

22exclusionary tactics, monopoly to both price and nonprice

23behaviors in order to deter competition and preserve their

24position in the marketplace. Monopoly tactics signal the

25customers and the marketplace that other actors should

125

1play ball.

2This disrupts the natural balance of a free

3market as innovators are no longer rewarded for building a

4better mouse trap and selling it at a better price. I can

5think of no better example then the global market in

6microprocessors in which AMD competes. In its March 2005

7ruling that I noted above, the JFDC cited evidence that

8showed quite clearly from the beginning of this decade

9until it was able to fend off competitive technologies

10from AMD, which had been gaining market share, by using

11its entrenched position in Japanese OEMs to crack down

12through anticompetitive tactics, level of those that would

13strive to bring differentiation and choice to endusers

14around the world.

15 AMD has competed against a persistent monopolist

16in a global market. We've confronted a variety of

17exclusionary abuses, including payments for exclusivity;

18rebates to make it too costly to ship to a rival even a

19small share of the customer's business; threats to

20withhold road maps, technical information and support;

21discriminatory allocations and scarce parts; and delay or

22reduced marketing share or substance.

23In a vacuum, with names and faces attached, the

24damaging impact of each of these individual acts may seem

25less obvious. While the FTC and DOJ appropriately have

126

1been examining specific practices one by one that occurred

2previously, it is important not to lose sight of the fact,

3as business firms competing against dominant firms know,

4that dominant firms can and do use a combination of

5practices, seldom just one, to maintain dominance. They

6can modulate the mix of practices as rivals try to adjust

7and react to maintain the marketplace in a prisoner's

8dilemma.

9What's important to understand is the collective

10impact. These bad acts often add up to a pattern of

11conduct that sends very strong signals to the marketplace,

12signals that are direct and punitive and that have a

13chilling effect on competition and the innovation process.

14Once a monopolist has injected enough fear into

15the marketplace, the need to explicitly threaten rivals

16every time is eliminated. It becomes understood and all

17too often accepted as the natural condition of the market.

18This is how our rival, even when lagging behind

19on the technological innovation front, manages to always

20maintain more than eighty percent revenue share for more

21than a decade. In other words, the dominant firm is

22perfectly capable of maintaining its market share through

23abusive conduct, even in a high technology market, for

24indefinite periods of time. This is particularly true in

25markets where the barriers to entry, including

127

1intellectual property and capital, are so very, very high.

2Which leads me to myth number two: Section 2 is

3not equipped to deal with the special characteristics of

4high tech markets. The truth? There is general agreement

5among global regulatory bodies as to what constitutes bad

6conduct on the part of dominant players in the market.

7And under those standards, bad conduct is bad

8conduct, plain and simple, no matter the industry in

9question. There is nothing unique about technology,

10whether it's the oil business, the pharmaceutical

11business, the chemical business or the computer business.

12The microprocessor market, once again, provides

13an example. In 2002, when AMD set out to earn its place

14in HP's commercial desktop product road map, AMD agreed to

15provide HP with one million processors for free, not just

16any processors, but the most advanced chips in its

17portfolio. HP was able to use only 140,000 and left

18860,000 units, free units, on the table. We believe

19because, had it taken more, its AMD-related savings would

20have been cancelled out several times over because of

21penalty Intel would have exacted in the form of higher

22prices on HP's Intel purchases.

23The result? Customers paid more; were forcibly

24deprived of an AMD alternative that might have been more

25suitable for their needs.

128

1Or take the recent revelation in the "Financial

2Times Deutschland" that Intel has entered into an

3exclusive contract Germany Media-Saturn-Holding,

4stipulating that competitors of Intel such as chipmaker

5AMD are not allowed to sell their products in Germany's

6dominant PC retail.

7The result? While consumers elsewhere in Europe

8favor AMD-powered computers, because they get a better

9equipped system for the same number of Euros, any German

10customers don't get to choose. The product in the

11marketplace in question are indeed complex, but the abuse

12of that should be a question for [unintelligible].

13Nor are these examples unique. Consider the

14Rambus 2006 Federal Trade Commission order, which stated

15that, quote, "Rambus engaged in exclusionary conduct which

16significantly contributed to its acquisition of monopoly

17power in four-related markets." Or the often overlooked

18original Microsoft decree that banned Microsoft from

19requiring its OEMs to pay the same licensing fees whether

20they installed the Windows operating system or not,

21thereby forcing the buyers and substitute operating

22systems to give their product away for free.

23In fact, if we take a moment to consider the

24fundamental considerations underlying the most high

25profile technology industry cases that come before the

129

1courts, we find at their core anticompetitive conduct that

2is almost universally recognized as impermissible under

3antitrust standards around the globe, which clearly falls

4within the band of Section 2.

5Perceptions like these exist around the industry

6and they cloud our ability to protect consumers.

7But none is more damaging than the industry myth

8that I'd like to address here today. Myth number three:

9Consumers aren't harmed if system prices are coming down.

10The truth? Apparent discounts are not always real

11discounts. Exclusionary conduct by monopolies keeps

12prices higher, slows innovation and limits consumer

13choice.

14There's plenty of real precedent from around the

15world from every industry to support this point. Consider

16"The United States vs. Dentsply International, Inc.,"

17Third Circuit case. The Third Circuit recognized that

18Dentsply's exclusive dealing arrangement improperly

19limited the ability of its rivals to compete, thus denying

20customer choice.

21And in its decision in LePage's, Inc, which is

223M, the Third Circuit similarly made the claim that the

23application of Section 2 to exclusionary conduct,

24explaining that, quote, "Even the foreclosure of one

25significant competitor from the market may lead to higher

130

1prices and reduced output."

2And the European Commission acted recently in

3the "Tomra" decision to make plain that, as its

4Competition Commissioner explained, quote, "I will not

5tolerate dominant companies hindering competition or

6excluding other players from the market as this harms

7innovation and consumers. Rebates and discounts cannot be

8used by a dominant company as part of the strategy to

9exclude actual and potential competitors."

10For instance, industry analysts have recently

11suggested that if the x86 microprocessor market were fully

12competitive, it would have allowed AMD to gain a greater

13share of the market and far more benefits would have been

14delivered to consumers in the form of lower prices and

15better and faster innovation.

16In recent economic analysis by Cal Tech

17Professor Preston McAfee shows that the U.S. Government

18pays higher prices and squanders taxpayer dollars when

19procurement prices are curbed by brand-specific

20specifications and contracts that foreclose competition

21and the benefits that open procurement policies promote.

22As with the aforementioned Microsoft decree,

23often what passes for pricing is just the imposition of a

24legal condition and the veiled threat of yet higher prices

25to exclude competition.

131

1That's why I believe these hearings are so

2important. And I commend the Department of Justice and

3the Federal Trade Commission for bringing them here to

4Berkeley, so close to the heart of the U.S. technology

5industry in Silicon Valley. Because, while rigorous

6enforcement of Section 2 is important, it is absolutely

7vital to the continued success of the United States

8technology industries.

9As we look to craft sound competition policy to

10govern our industries, we must consider the way in which

11these markets function in the real world. We cannot get

12caught ignoring tangible truths in favor of marketplace

13myths. We must send a strong deterrent message to all

14industries, including technology, Section 2 applies to

15what we do and who you harm. And crossing the line into

16illegality will not be permitted, no matter how cool the

17product, how familiar the logo or how high tech the

18industry.

19In the technology market, the stakes are

20particularly high because the progress of innovation and

21the health of our broader national economy in a

22globalizing world requires both robust competition and

23robust and enforced competition policy

24Thank you very much.

25MS. GRIMM: Thank you, Tom.

132

1Our next speaker is Michael Haglund. Can you

2hear me?

3Mike is a partner in Haglund Kelley Horngren

4Jones & Wilder in Portland, Oregon, and counsel to

5Ross-Simmons. He graduated in 1973 from Western Oregon

6University with a B.A. in Education, and he received his

7law degree from Boston University in 1977

8Mr. Haglund has primarily practiced in natural

9resources, admiralty and general business law throughout

10his career, and is experienced in a wide range of legal

11representation, including antitrust

12In 2003, he acted as lead counsel for the

13largest antitrust verdict in the history of the Pacific

14Northwest, a $79 million dollar judgment against

15Weyerhaeuser.

16In November of 2006, Mr. Haglund argued in the

17U.S. Supreme Court on behalf of Ross-Simmons in

18"Ross-Simmons v. Weyerhaeuser," a Section 2 case involving

19allegations of predatory bidding or buying

20 Mike.

21MR. HAGLUND: Thank you.

22I wish to thank the Federal Trade Commission and

23the U.S. Department of Justice Antitrust Division for the

24invitation to present testimony today as part of this

25series of hearings on Section 2 of the Sherman Act

133

1I offer this testimony, not on behalf of any

2individual business or client, but from the perspective of

3the many small and medium-sized businesses, mostly family

4owned, that I have been privileged to represent throughout

5the course of my career in the resource-based industries

6of the Pacific Northwest.

7I am the exception on the program today. I'm

8more of a bricks-and-mortar or in-the-ground kind of

9antitrust practitioner. I'm in my thirtieth year of

10law practice and have devoted most of that to the

11representation of the small and medium-sized participants

12in the forest products, fishing and agricultural

13industries.

14One of the common threads of this client base

15has been the production or is the production of

16commodities derived from the rich natural resources of our

17region in the Pacific Northwest: logs, lumber and plywood

18in the forest products industry; salmon and crab in the

19fishing industry; and essential oils like peppermint or

20spearmint, in agriculture.

21The application of Section 2 to these types of

22markets is important and must be analyzed within the

23context of the unique market realities that govern those

24markets, where in many cases there is the potential for a

25dominant buyer to exercise monopsony power to the

134

1detriment of its small competitors, input or commodity

2sellers generally, and ultimately consumers.

3These markets may be localized in that they're

4confined to a region of the United States and they are

5often exemplified by what Professor Warren Grimes refers

6to as, quote, "small atomistic sellers," unquote, who are

7more vulnerable to market abuses than consumers.

8There are multiple such markets in the Pacific

9Northwest, where a large and diverse number of small

10players are selling their commodity products to firms that

11process the logs, the fish, or the agricultural product

12into a host of other products.

13In some markets, the processor base may be quite

14small and dominated by one or a few large firms. As

15Professor Roger Noel has observed, "Local monopsony in

16conditions where the monopsonist does not have market

17power at the output level in a national or regional

18market, causes harm to consumers by misallocating

19production across regions or across localities."

20Antitrust cases associated with input markets

21have received very little attention until quite recently.

22In fact, a good share of the scholarship on the subject

23that exists today is found in this quarterly 2005 issue of

24the "Antitrust Law Journal," which contains a symposium

25collection of nine articles, including the two I've

135

1referenced from Professors Grimes and Noel a moment ago.

2The application of Section 2 to input markets is

3an area of antitrust law deserving of more attention, in

4my view, and it is about to receive it from the United

5States Supreme Court in its forthcoming decision in

6"Weyerhaeuser vs. Ross-Simmons Hardwood Lumber Company,"

7which will likely be handed down in March or April of this

8year.

9I argued the Weyerhaeuser case on behalf of

10respondent Ross-Simmons before the Supreme Court the end

11of November. Although it is difficult, and some would say

12dangerous, to make predictions based upon the briefs and

13the oral argument, but having been with the case since its

14inception and lead counsel at trial, and arguing counsel

15both in the Ninth Circuit and the Supreme Court, I believe

16the result is going to surprise people.

17When cert was granted, all of the pundits

18predicted that the court had taken the case to reverse it.

19And that view is still being expressed post argument on

20various blogs that follow the Supreme Court docket.

21For those of you who may not be fully aware, the

22Weyerhaeuser case as to predatory bidding or buying in

23input markets presents two issues. The first, whether the

24Brooke Group Price Cost Test, which was adopted in 1993,

25should be extended from the sell side to the buy side,

136

1first issue. And, second, whether the jury instruction

2regarding predatory bidding was flawed on grounds other

3than Brooke Group.

4The first issue, based upon the briefing and

5based upon the tenor of the oral argument, we are

6optimistic that the Supreme Court is going to affirm the

7Ninth Circuit in its decision that the safe harbor for

8pricing behavior that exists on the sell side through

9Brooke Group does not apply with the same force and should

10not be extended at least to inelastic input markets like

11the alder saw market at issue in the Weyerhaeuser case.

12Over the last quarter century, except for Brooke

13Group, the Supreme Court has eliminated or narrowed per se

14rules that did not have a sound economic foundation in the

15market realities of the individual case.

16The wisdom of Brooke Group most I think would

17say is its protection of inherently procompetitive price

18cutting in output markets. In the context of input

19markets, the challenged conduct involves price raising,

20bidding, resource prices up. Very few cases in the last

21fifty years and scholarship in its infancy. Conditions

22that are the exact opposite of those that prevail when

23Brooke Group's per se rule was developed.

24In these circumstances, the correct approach is

25the one that has always been the gold standard of

137

1antitrust rules, the rule of reason.

2The rationale underlying Brooke Group was also

3rounded substantially in concern about false positives,

4based in large part upon a sizable body of literature to

5that effect.

6In the predatory bidding context, there is no

7similar body of economic literature offering a similar

8warning. In point of fact, the very few cases of

9overbidding that do exist show that it is a rational

10strategy that does work. And I'm referring here to just a

11very few cases: American Tobacco from the Supreme Court;

12the Ross-Simmons case about to be decided; and the Reed

13Brothers case also out of the timber market that was

14decided by the Ninth Circuit in 1983.

15There are two reasons underlying my optimism

16that the Supreme Court will refuse to extend Brooke Group

17from the predatory selling context to immunize bidding

18conduct by a dominant buyer.

19First, the position of Weyerhaeuser and its many

20big business amici is based upon the notion of symmetry,

21that a rule that works for predatory selling and output

22markets should apply equally in predatory bidding to input

23markets by the sheer force of logic alone.

24The law, however, is no slave to symmetry. As

25Justice Holmes has written in what has been characterized

138

1by Judge Posner as the single most famous sentence in

2American legal scholarship, quote, "The life of the law

3has not been logic; it has been experience."

4In the past, notions of symmetry have influenced

5the antitrust juris prudence of the U.S. Supreme Court.

6However, in the last twenty-five years, market realities

7have consistently trumped symmetry and the per se rules

8which were sometimes developed as a result.

9The Supreme Court embraced symmetry, for

10example, in equating maximum and minimum vertical resale

11price constraints as per se illegal in "Albrecht vs.

12Harold Company" in 1968, but relied on market realities in

13overruling Albrecht's prohibition against maximum resale

14pricing agreements nearly thirty years later in "State Oil

15vs. Khan" in 1997.

16 The other half of that rule, by the way, now

17appears in some jeopardy with the Supreme Court's recent

18decision to reexamine whether vertical minimum resale

19price maintenance agreements should be deemed per se

20illegal under Section 1 of the Sherman Act, or whether

21they should instead be evaluated under the rule of reason.

22I refer here to "Leegin Creative Leather Products vs.

23PSKS," a decision out of the Fifth Circuit on which cert

24was granted just last month.

25In my view, the Supreme Court is clearly focused

139

1on eliminating per se rules or presumptions in antitrust

2which are not justified by market realities or which

3distort the fact-finding process at trial in a way that

4unfairly disadvantages one party or the other.

5The Independent Ink case of last term, in which

6the court abandoned the per se rule that patent equals

7market power in a tie-in case is the most recent example

8of this trend.

9My second reason for optimism on the Brooke

10Group issues comes from the oral argument. We were struck

11by the apparent lack of enthusiasm among the Supreme Court

12Justices for extending Brooke Group from the sell side to

13the buy side. Several justices, including Justice

14Kennedy, who wrote the 6-3 majority opinion in Brooke

15Group, expressed concern about the workability of

16converting the Brooke Group price cost test into a

17price revenue test on the buy side.

18There was record evidence that Weyerhaeuser used

19below-market transfers of all their saw logs from its

20company fee lands to subsidize its bidding up of saw log

21prices in the so-called open market in which it competed

22with Ross-Simmons. Weyerhaeuser argued that such bidding

23was immune from antitrust scrutiny so long as its alder

24division was not losing money overall.

25Adoption of such a rule, however, in this type

140

1of resource market would put a large company that had

2amassed low cost raw materials in a position to eliminate

3its competition by bidding up scarce supplies of open

4market sources and subsidizing that predation with below

5market transfer prices from its own captive supplies.

6The result would be under-deterrence of

7predatory bidding behavior, while impeding the most

8efficient allocation of scare resources.

9Another administrability problem not found with

10Brooke Group on the sell side is associated with the fact

11that the relevant input in the Weyerhaeuser case, alder saw

12logs, are used to produce very different products. In an

13alder saw mill those are chips; pallet lumber, which is a

14low-grade type of lumber which you see underneath products

15in various Costcos and elsewhere; and kiln-dried finish

16lumber. But Weyerhaeuser actually had 25 to 50 different

17lumber grades in the finished lumber category

18Each of the saw logs that went through any

19given alder mill produces products in all three of these

20categories, but the larger the diameter of the log, the

21even more higher grade lumber you're going to produce.

22Applying Brooke Group is extremely difficult in

23this sort of single input but multiple product output

24environment. And there is no comparable corollary

25on the buy side to the commonly utilized average variable

141

1cost or marginal cost formulation used in the sell side

2predatory pricing case.

3In sum, regarding the primary question in

4Weyerhaeuser, whether to extend Brooke Group to the buy

5side, we are guardedly optimistic that the Supreme Court

6will decline to do so because of the court's consistency

7over the last quarter century in refusing to create new

8per se rules or to extend old ones unless justified by the

9market realities of the particular industry or the

10particular type of antitrust claim.

11And, also, because of the TENOR of the oral

12argument. Brooke Group really was an exceptional case.

13Today, 14 years after it was decided, the rule of reason

14shines even more brightly as the gold standard of

15antitrust analysis.

16Now, assuming the Supreme Court does not extend

17Brooke Group to the buy side in Weyerhaeuser, it must then

18examine a second issue, whether the district court's

19instructions defining when predatory bidding will

20constitute anticompetitive conduct were flawed on some

21other basis.

22This was the instruction in which the district

23judge, having given the standard ABA model instructions

24for monopolization and anticompetitive conduct, instructed

25the jury that it could find that Weyerhaeuser engaged in

142

1anticompetitive conduct if it bought more logs than it

2needed or, quote, "paid a higher price than necessary in

3order to prevent plaintiffs from obtaining the logs that

4they needed at a fair price," unquote.

5This formulation was pounced upon by

6Weyerhaeuser and its amicis as, in their words, "standard

7gibberish," which constituted an independent ground beyond

8Brooke Group for reversal of the Ninth Circuit opinion.

9However, as pointed out in our merits brief, Weyerhaeuser

10never preserved any such alternative objection to the

11instruction. Attacking a pair of sentences in the jury

12instructions as unduly subjective or as an invitation for

13unguided speculation, proved an effective springboard for

14a grant of certiorari. But deciding the case on the

15merits requires an assessment of the instructions as a

16whole in light of the evidence, the closing arguments and

17the other instructions.

18In the trial court, Weyerhaeuser's counsel

19actually invited the formulation of the two sentences that

20have been so criticized in the commentary about this case.

21But in opening statements, and again in closing argument,

22Weyerhaeuser's counsel told the jury that multiple

23witnesses would be called who would and then did testify

24that the company never bought more than it needed and

25never pushed log prices up in order to hurt its

143

1competition. And a litany of two questions was put to 13

2different witnesses, obtaining denials on each of those

3same two points.

4It's worth noting that the Supreme Court has

5already decided the case from the very first one of this

6term involving a challenge to ambiguous language in a jury

7instruction. In "Aires vs. Del Montes," the court

8examined California's catch-all mitigation instruction and

9using the instructions in the penalty phase of a capital

10murder case.

11Based upon the way the case was tried and the

12evidence presented, a 5-4 majority found no reasonable

13likelihood that the jury had applied the admittedly

14ambiguous instruction in a way that prevented

15consideration of constitutionally relevant evidence.

16If the type of common sense -- and I put that

17word in quotes because that was the court's term. If that

18type of common sense approach is to apply in a capital

19murder case to consideration of ambiguous instruction,

20it's hard to see how there is a reason for a stricter

21approach in antitrust, especially in a case where the

22defendant tried the case in a manner that invited the very

23formulation of that jury instruction.

24In fairness, however, it should be noted that I

25was pressed at oral argument, particularly by Justice

144

1Souter, regarding the vagueness of the instruction on

2predatory bidding and the need for the Supreme Court to

3say something about that instruction. I conceded that the

4instruction was not perfect, but emphasized that neither

5the district judge nor plaintiff's counsel was given any

6chance through a defense objection on that ground to

7consider whether the instruction could be made more

8precise with other language.

9At trial, we in fact never attempted to exploit

10the nature of that couple of sentences and urged the jury

11to just award whatever they considered was fair. Instead,

12through economists, forest economists, we presented

13detailed market evidence to show how much the market for

14alder saw logs was artificially elevated above where it

15would have been but for the mix of anticompetitive

16practices, including manipulative bidding by the defendant

17Ultimately, the jury in Weyerhaeuser selected to

18the dollar one of the three damages scenarios presented by

19these forest economists. Had Weyerhaeuser challenged the,

20quote," paid a higher price than necessary," unquote,

21language, we would have had no problem adding precision to

22that instruction by linking the higher log prices to

23market factors tied to Weyerhaeuser's manipulative

24behavior as opposed to the normal operation of the market

25In fact, we could have accepted the suggestion

145

1made by the eight amicus states that filed a brief

2supporting Ross-Simmons, including Oregon and California,

3that the instruction that defined predatory bidding as

4having anticompetitive effect, quote, if the conduct

5raised the price that the buyers' rivals had to pay for

6the input beyond the level that could be justified or

7explained by other market factors and substantially

8affected the ability of the buyers' rivals to compete for

9the input.

10Because our evidence was designed to show how

11the historical relative equilibrium between finished

12lumber prices and log prices had been distorted by

13Weyerhaeuser's behavior in order to kill off rivals, I'm

14confident that there would have been no change in the

15result at trial with a more precise formulation for

16defining when bidding conduct in an input market can be

17found anticompetitive.

18What happens, you might ask, however, if my

19admittedly optimistic view is wrong and the Supreme Court

20reaches the vague instruction issue and reverses on that

21basis. In all likelihood, a retrial will then be

22necessary, but we are confident of a similar plaintiff's

23verdict for two reasons.

24First, the Ross-Simmons verdict generated

25several follow-on cases in which Weyerhaeuser produced

146

1thousands of additional incriminating documents,

2demonstrating the deliberate character of its multi-tactic

3plan to monopsonize the alder saw log market in the

4Pacific Northwest.

5By the way, the Pacific Northwest is the only

6place west of the Mississippi where there is a hardwood

7industry, in stark contrast to the east, where hardwood

8species predominate and there's a substantial hardwood

9industry.

10In other words, we're even stronger on liability

11in the retrial than we were the first time around, and

12perhaps that's why Weyerhaeuser settled three follow-on

13cases we handled on behalf of ten other plaintiffs for a

14total of $62 million.

15Provided we are not saddled with a Brooke Group

16test, we believe our damages theory can easily be matched

17up with a more objective formulation of the market

18distorting bidding conduct than the two-sentence

19formulation now at issue before the Supreme Court.

20But however it turns out, the Weyerhaeuser case

21will be important for all resource space input markets,

22particularly those at the inelastic end of the spectrum.

23Section 2 has a real role to play in these markets. If

24you are a tree farmer, you want to have a healthy number

25of saw mills competing for your log production within a

147

1reasonable distance of your tree farm. And even if

2you happen to sell your logs of a particular species to a

3rising or emerging monopsonist, paying premium prices

4during this period of predation, you're concerned about

5the long-term health of your input market for that

6particular species and will likely cause you not to

7replant it if you fear that there will only be a single

8buyer 30 to 50 years down the road when those seedlings

9are now mature and ready for harvest. And we have

10evidence to that effect.

11It was precisely this type of real market

12consideration that caused most of the log seller community

13in the U.S., represented by the National Woodland Owners

14Association and the American Loggers Council, to support

15Ross-Simmons in an amicus brief in the Supreme Court.

16Avoiding expansion of Brooke Group from the sell

17side to the buy side is important in other input markets

18as well. Most U.S. fish markets are classically inelastic

19because the total catch is fixed by state and federal

20regulators. The crab fishermen plying U.S. waters off the

21coast of Oregon, Washington and Alaska need a healthy mix

22of seafood processors to ensure market prices that sustain

23the crab industry and its U.S. fleet.

24A flexible rule of reason approach to

25exclusionary conduct in this type of market is vital both

148

1to deterring illegal conduct and to ensuring fair results

2at trial

3Also, many agricultural markets, especially

4those like peppermint where production is regulated by

5federal marketing orders, are susceptible to abuse in the

6form of artificially low prices dictated by a dominant

7buyer, or oligopolistic behavior in a highly concentrated

8processor market

9I would like to take this opportunity to thank

10the FTC and the DOJ Antitrust Division for holding this

11hearing out on the west coast rather than in Washington,

12D.C. I believe it is critically important for federal

13antitrust enforcers to be out in the field regularly to

14have a full appreciation of the importance of local and

15regional markets

16Indeed, the lack of consideration of local and

17regional markets in the Solicitor General's brief

18supporting Weyerhaeuser was one of the primary reasons, I

19am told by state officials, that eight states on short

20notice submitted their amicus briefs on Ross-Simmons' side

21in this case.

22In its antitrust jurisprudence, the Supreme

23Court has repeatedly emphasized that antitrust analysis,

24quote, "must be attuned to the particular structure and

25circumstance of the industry at issue," unquote.

149

1In my view, this can only be accomplished if one

2is immersed in the facts and circumstances of a given

3industry, what I call the who, what, when, where and how

4that requires extensive use of investigative interviewing

5in addition to and not as a substitute for analysis of raw

6data.

7From my experience in the northwest corner of

8the United States, I have three suggestions for the FTC

9and DOJ in its evaluation of antitrust issues to resource

10space input markets

11First, please do not discount or dismiss the

12significance of a local or regional market simply because

13the dominant buyer/processor may not have the market --

14may not have market power in the downstream output market.

15As Professor Noel so convincingly demonstrated

16in his article, this is an area where input sellers are

17vulnerable and can be abused by a monopsonist to the

18detriment of both regional and national economies.

19Second, please be aware of the influential

20impact of the extraordinary legal and organizational

21talent brought to bear by large corporations and their

22affiliated support organizations on the antitrust issues

23that come before you. The small, atomistic sellers who

24make up so many of the local and regional input resource

25based input markets in the U.S. are no where near as well

150

1organized and have precious little in the way of financial

2resources to devote to long-term efforts to influence the

3direction of Sherman Act jurisprudence.

4It is therefore particularly important that

5federal and state antitrust enforcers to look behind the

6incredibly capable advocacy available to large corporate

7interests, and to independently investigate the relevant

8facts of each market and each industry, and I emphasize,

9in the field.

10Third and finally, from my perspective,

11throughout a now 30-year career involved in three resource

12based sectors of the U.S. economy in the Pacific

13Northwest, I have been struck by the close match between

14my own experience and two bedrock principles of antitrust

15law.

16One, that the forms of anticompetitive conduct

17are myriad. And, two, that sound antitrust analysis is

18joined at the hip with the fact-laden structure of the

19particular market and industry at issue. This amazing

20factual variability, in my view, makes the quest for a

21unitary standard of exclusionary conduct under Section 2

22illusionary. It is a much sounder policy to embrace the

23flexibility of the rule of reason standard and to apply it

24appropriately to the market realities of the industry in

25the particular antitrust case.

151

1On this last point, I think it's interesting to

2note that our own -- excuse me, that own new Chief Justice

3appears to be no fan of etiological purity in the way the

4Supreme Court decides its cases. In a very insightful

5article by Jeffrey Rosen in the January/February issue of

6"The Atlantic Monthly," Chief Justice Roberts says the

7following when asked to define the qualities of judicial

8temperament that he thought successful Chief Justices like

9Marshall, who was Chief Justice Roberts own personal

10model, embodied. Quote, "I think judicial temperament is

11a willingness to step back from your own committed views

12of the correct jurisprudential approach and evaluate those

13views in terms of your role as a judge. It's the

14difference between being a judge and being a law

15professor," unquote.

16I think the quest by some in the antitrust

17division to develop an overarching standard defining all

18anticompetitive conduct under Section 2 of the Sherman Act

19is inconsistent with the highly fact-laden and

20industry-specific character of antitrust. Such a quest is

21too much of law professor and too little of the practical

22fact-based enforcer. It should be abandoned and the

23energy of our antitrust agencies refocused on

24investigation and enforcement.

25Thank you for the opportunity to present this

152

1testimony.

2(Applause.)

3MS. GRIMM: Thank you, Mike.

4Our third and final speaker this afternoon is

5David Dull, who is Senior Vice President of Business

6Affairs, General Counsel and Secretary of Broadcom

7Corporation.

8Mr. Dull is responsible for the company's

9acquisition, outside investment and licensing activities,

10in addition to advising on all legal matters.

11Mr. Dull joined Broadcom as Vice President of

12Business Affairs and General Counsel in March 1998, and

13was elected Secretary of the corporation in April 1998

14Mr. Dull received a B.A. and a J.D. from Yale

15university.

16MR. DULL: Thanks, Karen, for that kind

17introduction. And thanks to the Haas School and its

18affiliates here in Berkeley for hosting this event today.

19I want to compliment the FTC and the Department

20of Justice for convening these hearings. While like many

21in the business, we at Broadcom are of course wary of

22regulation and other governmental and court interventions

23that may stifle growth and cause inefficiency.

24We nonetheless recognize the positive role our

25government has played and can still play in facilitating

153

1economic growth, efficiency and innovation, which

2ultimately is what drives our economy.

3I thank and commend the FTC and the DOJ for

4taking the time to solicit views from across the spectrum

5and across the country and hope that what comes out of

6this process will promote that positive role.

7Let me begin my remarks by telling you a little

8bit about the company I've been with since 1998, Broadcom

9Corporation. In 1991, a graduate student by the name of

10Henry Nicholas, and his professor, our current chairman,

11Dr. Henry Samueli, had a vision of an innovative company

12that would provide semiconductors, computer chips, to

13facilitate high speed digital communications for business

14and consumer applications.

15In a world where television and cell phones were

16still analog, no one had heard of HD TV, dial-up modems

17were considered cutting edge technology, and few even

18contemplated the potential of the internet and today's

19laptops and hand-held devices. These two visionaries saw

20that the demand for high bandwidth digital communications

21would skyrocket. And of course it has.

22Broadcom's revenue now exceeds three billion

23dollars a year. We've retained our roots in Southern

24California, but we now have facilities all over the United

25States and around the world, including several facilities

154

1and over 1,250 employees here in the Bay Area.

2We continue to focus on semiconductors for high

3speed, high bandwidth applications, such as set-top boxes

4for television, gigabit ethernet, DSL modems, wireless

5networking, and cellular phones. We also produce

6closely-related devices, such as digital TV chips and

7multimedia chips for iPods and cell phones.

8Indeed, it is far to say that, as much as any

9other party or any other factor, Broadcom has enabled the

10digital communications revolution that touches each of us

11every day.

12And we continue to follow the example of our

13founders. We have built our entire business model around

14continuing innovation. Our products are state of the art

15and Broadcom is a technology leader in every market in

16which we play.

17Our engineers are top-notch. In fact, of our

185,200 or so employees, more than 3,800 are engaged in R&D;

19439 are Ph.Ds. We spend about 40% of our gross profit on

20R&D, on innovation.

21In keeping with the purpose of these hearings,

22today I plan to talk a bit about real issues that we

23confront in the high tech industry in which we operate.

24These are not your father's competition issues.

25Everyone in this room is keenly aware that the

155

1antitrust laws date back to the end of the 19th century.

2So, one overriding theme I hope you will take away from my

3remarks today is that antitrust laws must not get trapped

4in traditional analysis or outmoded or dated thinking.

5They must be dynamic and flexible.

6With due deference to economic analysis and

7marketplace realities, our antitrust regime, including

8that addressing single-firm conduct, must remain robust to

9deal with the issues of the 21st century. And, as we all

10know, many of those issues revolve around technologies in

11the high tech industries.

12We at Broadcom firmly believe that competition

13is what makes our innovation economy work. When coupled

14with a well-educated and highly motivated work force,

15competition unleashes creative energy and creativity

16spawns the amazing innovations that we have seen just in

17the past decade alone.

18In the semiconductor industry, as Tom knows,

19competition creates efficiency on a scale greater than

20anywhere else. The capability of today's high tech

21products dwarf those of just a few years ago, yet prices

22continue to drop.

23The antitrust laws serve their most useful role

24when they promote competition and prevent companies that

25have obtained a strong position in one area from

156

1exploiting it to prevent competition in other areas.

2Before addressing that in greater detail, let me

3be clear about two things. First, it is important not to

4penalize innovation by attacking those companies that have

5achieved strong market positions solely through

6innovation. Innovation must be encouraged because it is

7the key to our country's continued success in the

8increasingly challenging global economy.

9Secondly, it is important that the intellectual

10property rights of innovation be respected. Our patent

11system encourages innovation by ensuring that its vendors

12will reap a portion of the economic benefits of their

13inventions, while at the same time requiring those

14inventions to be shared with the public. That is a good

15thing and we must not sacrifice it in the name of

16competition.

17At Broadcom, we hold over 1,900 U.S. patents and

18have another 5,900 U.S. and foreign patent applications

19pending. We care deeply about intellectual property

20rights. But companies that use the strong positions they

21have obtained, even if attained by innovation, to close

22other markets to competition, or that use deception and

23false promises to obtain their strong position in the

24first place, are not innovative, but rather are standing

25in the way of innovation. The antitrust law must address

157

1that type of behavior.

2As I said, Broadcom designs and sells computer

3chips. In today's highly sophisticated electronic

4applications, be they computers, cell phones or cable

5boxes, no one produces all of the systems and components

6for a particular application. In fact, a typical consumer

7product incorporates chips and software from a number of

8different suppliers.

9In our vernacular, no one company produces all

10of the silicon on the motherboard. Today, in hardware and

11software, open systems is the name of the game. Open

12systems are why we have the PCs and the internet.

13Interfaces between one component and another are therefore

14necessary. Some of those interfaces are specified by

15standards developed with broad industry participation

16under the auspices of standard setting bodies such as the

17IEEE and ANSI.

18The highly successful 802.11B and G wireless

19networking standards fall into this category. The

20proliferation of Wi-Fi networking, supported by devices

21from hundreds of manufacturers, demonstrate the power of

22industry standards arrived at through non-partisan

23processes.

24Other interfaces are de facto industry standards

25that arose without a formal standard setting process, but

158

1are generally open for industry participants to use in

2deploying their own standards compliant price.

3And some interfaces are entirely proprietary,

4which is to say they're put into place unilaterally by one

5or another industry player who claims ownership of that,

6quote, "standard," unquote, and asserts the right to

7prevent or control its use by others.

8Obtaining control of key interfaces through

9anticompetitive means, or using control of key interfaces

10to extend a dominant position in one market into other

11markets is a real danger in our industry. It is of major

12concern to companies like Broadcom who win through their

13ability to innovate.

14It should also be of concern to consumers and to

15their representatives in the antitrust agencies. That

16sort of behavior chokes off competition among industry

17players, which deprives consumers of the innovations and

18lower prices that come from vigorous competition.

19At its most extreme, in our industry, interface

20control could enable a dominant firm in one critical piece

21of the motherboard to take control of the whole system,

22even if the quality and cost of its products do not

23support that result.

24Those of us of a certain age know what an

25end-to-end monopolist in a communication space looks like.

159

1It was the old totally vertically integrated telephone

2company. One company controlled all of the equipment, all

3of the connections, all of the interfaces. Indeed,

4everything from the chips to the telephone repairman.

5It wasn't simply that they had a lock on the

6industry. They, not competition, decided what innovations

7made their way to the consumer and when. That slowed down

8the transfer of innovation, and as a consequence,

9telecommunications innovation in this country was outpaced

10by that in others.

11In an increasingly competitive global economy,

12we cannot afford to return to those days. And the

13antitrust laws governing single-firm conduct were the

14means by which that situation was remedied.

15Today different technologies from different

16companies come together to create a plethora of consumer

17products, which we all enjoy and to a substantial extent

18take for granted. This creates an ongoing challenge in

19defining how those technologies will interconnect and

20interoperate and the rules that will apply to that

21endeavor.

22Even the best technology is of little use in

23isolation. The antitrust laws have an important role in

24policing the conduct of firms who would seek to take

25control of those interconnections so as to eliminate

160

1competition and thus harm consumers.

2In my remaining remarks today, I will focus on

3two areas of concern which, in Broadcom's experience, are

4particularly important to preserving competition.

5The first is standard setting. I know there was

6a fair amount of discussion on that this morning. There

7will be more of it this afternoon. The second is the use

8of proprietary interfaces from one market to another.

9These are not theoretical issues. These are

10real issues that Broadcom has faced in the past and

11continues to face today.

12We come at this from the perspective of a highly

13innovative company with world-class technology, attempting

14to break into new markets dominated by entrenched rivals.

15At the same time, we are an example of a company

16that has thrived through key contributions to important

17industry standards and, today, without charging royalties

18for those innovations.

19Standard setting refers to the process of

20creating and implementing a way of doing things. As a

21simple example known to all of us, there's the standard

22format for video known as VHS. That standard makes it

23possible for a variety of competing manufacturers to make

24the various components that are needed to record and play

25home video: the camera, the tape, the VCR, and so forth.

161

1Similar standards exist for CDs, DVDs, as well as

2standards that allow voice video data and multimedia to be

3shared among various wired and wireless devices.

4In addition to facilitating competition by

5enabling different companies to produce products that will

6interconnect and interoperate, standard setting, when done

7properly, can also resolve intellectual property rights or

8IPR issues that might otherwise impede progress.

9With the complexity of today's products, often

10multiple parties own IPR that is needed to implement a

11particular technology-based application. If Company A

12owns essential IPR and so do Companies B, C, D and E, each

13can block the other and everyone else from making a

14product using the best available technical solutions.

15In the standard setting process, companies

16typically are required to agree that they will disclose

17their IP rights that are essential to practice this

18standard before the standard is adopted. This gives the

19standard setting body and the participants in the standard

20setting process the ability to avoid such IPR or to

21address the means by which that IPR will be licensed to

22those who practice the standard.

23I will get to licensing in a minute, but first a

24word on IPR disclosure in standards making.

25There are those who say that disclosure is not a

162

1significant problem because companies generally play by

2the disclosure rules. They say that failure to disclose

3is rare and therefore not really a problem. At Broadcom,

4we aren't sure whether failure to disclose is in fact rare

5in all standard setting bodies. But even if that is the

6case, it can still be a serious problem.

7Indeed, the fact that participants in standard

8setting expect disclosure and rely upon it makes those

9instances of failure to disclose all the more problematic.

10Without disclosure, the standard is at constant risk of

11being hijacked by an IPR holder that has hidden in the

12weeds during the development of the standard or, even

13worse, has helped steer development toward its own

14undisclosed proprietary technology only to spring its trap

15after the standard has been set and millions or even

16billions of dollars have been invested in its

17implementation.

18 This risk is not an abstract or a theoretical

19concern. In fact, these hearings are particularly timely.

20Just this past Friday, four days ago, the jury in San

21Diego rejected an attack on my own company by a firm

22attempting to force us out of certain technology spaces by

23asserting two patents that it controlled. Its

24infringement case was based in substantial part on our

25implementation of an industry standard for video

163

1compression. The jury found no infringement, thank god.

2And, perhaps more significantly, also found that our

3adversary had violated the disclosure rules of the

4standard setting body by failing to disclose its patents

5which allegedly covered the standard.

6Sadly, the company that launched this

7ill-founded patent assault on an international standard,

8cynically justified its actions afterwards on the grounds

9that it had nothing to lose, even though after a nine-day

10trial, a jury unanimously agreed that the company had used

11the standards process and had also violated its duty of

12honesty and fair dealing with the U.S. Patent and

13Trademark Office.

14Meanwhile, defending itself against those

15illegitimate claims cost Broadcom millions of dollars.

16And the lawsuit created confusion and concern among our

17customers and the many others who use the H.264 video

18compression technology.

19So, this is a very real risk. If an

20opportunistic company can get away with these tactics, it

21would be in a position to dominate components for an

22important ubiquitous video compression technology by

23asserting its patents against all would be competitors.

24But disclosure, important as it is, is not

25enough. Disclosure is only the first step in assuring

164

1that hijacking will not occur. Disclosure merely allows

2the standards development body to thwart attempts to

3insert proprietary technology into the standard.

4It is at least equally important for industry

5participants to abide by the rules after the standard is

6in practice, is in place. A key element of that is

7licensing terms and conditions.

8The rules of standards bodies typically provide

9that IPR that is essential to practice the standard will not

10be included in the standard unless the owner agrees to

11license that IPR to those who wish to practice the

12standard on either a royalty free or fair reasonable and

13nondiscriminatory, so-called FRAND, sometimes called RAND,

14terms.

15What happens when someone fails to live up to

16these commitments? As I noted, once a standard is set,

17the industry moves forward and invests millions if not

18billions of dollars in implementing the standard. That

19investment is based on the understanding and assumption

20that IPR issues are resolved. Either there will be no

21need to take a license to the IPR, or any licensing will

22be on FRAND terms.

23If a company with essential IPR seeks to impose

24non-FRAND licenses, the balance is completely upset.

25Suddenly the industry which adopted the standard with the

165

1understanding that licensing costs would be reasonable, is

2confronted with a monopolist seeking to charge monopoly

3rates.

4In industries that are involved in standard

5setting, there are certain practices that I would venture

6to say everyone understands are not FRAND terms. For

7starters, refusing to license at all violates a FRAND

8commitment. Amazingly, there are some in the industry who

9take the position that, notwithstanding their commitment

10to license all who wish to practice the standard,

11essential IPR holders can pick and choose among potential

12licensees for any reason, including, it would seem,

13whether the potential licensee is a downstream competitor

14Another example: Broadcom has been confronted

15by a licensor who participated in the standard setting

16process, insisting that, as a condition to being granted a

17license to the intellectual property essential to practice

18the standard, it would have to give back a royalty-free

19license to a much broader sweep of Broadcom's own

20intellectual property, including IP-covered features and

21functions entirely unrelated to the standard.

22To usurp the blood, sweat, tears and genius of

23interface companies in such a manner as a condition to

24practicing an industry standard runs directly contrary to

25the fundamental objectives of standard setting bodies.

166

1If this sort of practice is allowed, what

2incentive will any company have to innovate or invest,

3knowing that unrelated technology can be appropriated as

4the price for making standardized products.

5Another example that we have seen is a company

6attempting to use access to essential IPR to coerce

7customers into buying its products, rather than letting

8the merits of the products determine who gets the sale.

9And we have examples where a company has thought

10to stack a standard setting organization with supposedly

11independent voters to skew the standard towards it own

12technology or away from the technology of its rivals.

13 To be clear, I do not suggest that a company

14should be required to share its technology with others.

15Far from it. Patents are available to protect innovation

16and Broadcom is a firm believer in the patent system.

17But it is imperative that, when a company has

18made a commitment to license on FRAND terms as a condition

19of getting its technology included in a standard, it must

20not then be allowed to exploit the market position it

21gained through incorporation in its IPR and the standard,

22by reneging on that commitment.

23And a company, likewise, should not be allowed

24to subvert the rules that are put into place to ensure

25that standard setting is a nonpartisan exercise.

167

1These are very real and contemporaneous examples

2of the kind of anticompetitive single-firm conduct we at

3Broadcom believe the antitrust laws are intended to

4address.

5Some say that determining what is fair and

6reasonable is too hard a task. That is a standard that

7cannot be enforced. We heard some discussion along those

8lines this morning.

9Often the firms that say this are the very firms

10that fail to disclose their patents, have engaged in

11rampant discrimination that cannot possibly be reconciled

12with a FRAND obligation, and have engaged in other

13behavior that demonstrates that it is a lack of will, not

14a lack of ability, that has resulted in their FRAND

15violations.

16Fair and reasonable simply means that the

17technology will be available on competitive terms, rather

18than on terms that reflect a market power gain through

19inclusion of technology in the standard.

20It also means that no participant will charge a

21disproportionately high royalty so as to hobble the

22standard or render it uncompetitive.

23 Technology companies are often engaged in patent

24litigation where a question before the court is how to

25assess a reasonable royalty in damages. There's no reason

168

1to believe that the courts would have a harder time

2figuring out what reasonable royalty is in the standards

3context than in any other context. The court can take due

4account of the competitive goal of the standard setting

5body in requiring a FRAND commitment up front, and

6otherwise undertake the same exercise it goes through when

7evaluating damages and so forth.

8It has also been suggested that failure to

9comply with a FRAND obligation is a matter better left to

10contract than antitrust law. One might ask, if a court

11applying contract law can figure out what FRAND means, why

12can't the same court apply antitrust law?

13Contract law is a private remedy to redress

14private rights. FRAND violations can eliminate

15competition and hurt consumers, competitors, innovation

16and the economy as a whole. Isn't preventing such an

17injury exactly what the antitrust regime is all about?

18Moreover, if companies are willing to break

19their commitment because they conclude they have little or

20nothing to lose by doing so, the contract remedy is

21inherently insufficient to protect innovation, competition

22and consumers. And that becomes the job of antitrust

23law.

24The second area I would like to talk about is

25interfaces. As I noted before, interfaces are the way one

169

1piece of technology connects to another. By manipulating

2the interface and making it proprietary, a company with a

3monopoly over one area of technology can effectively shut

4out competitors and technology that would connect with the

5monopoly technology.

6For example, if a company had a monopoly in

7amplifiers, it could obtain a monopoly in speakers by

8creating a proprietary amplifier-to-speaker interface and

9refusing to license that interface to anyone. The speaker

10market, which previously enjoyed vigorous competition that

11fostered innovation and lower prices, would suddenly be

12controlled by one firm with little incentive to innovate

13or reduce prices.

14We've seen this in practice. Broadcom is a

15communications chip company. Our chips connect devices

16and systems. We've seen, for example, companies that

17control the main processor of a particular system, one

18that was at one time characterized by an open interface,

19suddenly making that interface proprietary. For no good

20technological reason, they make it harder to interconnect

21with that chip, while at the same time launching their own

22communications chips that competes with Broadcom and

23others.

24This two-prong strategy, control the connection

25with the dominant product and compete in the adjoining

170

1market, has a predictable result. The dominant firm

2leverages its monopoly from one area outward into ever

3greater areas

4Over time, the dominant firm expands its empire

5to the entire motherboard, destroying its competitors and

6the innovation they would bring along the way. There

7certainly are instances where the development of new

8interfaces is real innovation.

9Where there is real innovation in the interface,

10innovators should have the opportunity to be appropriately

11compensated. But that compensation should at best take

12the form of a modest, truly nondiscriminatory royalty. It

13should not be a vehicle for extending dominance from one

14kind of chip to another by, for example, the kind of

15asymmetrical brand back of IPR from the licensee to the

16licensor that I discussed earlier.

17And a small improvement in interface technology

18should not come at the sacrifice of innovations of orders

19of magnitude more significant in the adjacent

20communications markets if innovators' chips can no longer

21communicate with the now closed interface.

22Of course sometimes the new interface does not

23even represent an improvement, just a difference. When a

24company has a history of using open interfaces or of

25licensing its interfaces to third parties and then stops

171

1doing so, while at the same time entering the market on

2the other side of the interface, one ought to become

3suspicious.

4We've experienced that in our industry. Again,

5there is a role for antitrust when such changes provide

6little or no benefit but substantially hurt innovation and

7therefore consumers and the economy as a whole.

8I recognize that today I barely scratched the

9surface of the issues that I talked about. And of course

10much depends on the individual facts and circumstances of

11any particular case and market.

12That said, the antitrust laws and the courts and

13agencies that are called upon to enforce them should not

14shy away. Usually, once the facts are separated from the

15noise, it is not difficult to separate the procompetitive

16stories from the anticompetitive ones, particularly in the

17area of deceptive conduct in standard setting processes

18there is little risk that procompetitive behavior will be

19deterred.

20In closing, I hope the FTC and DOJ and those who

21are thinking seriously about antitrust in the 21st Century

22will take away from my remarks three basic concepts

23First, antitrust, as it relates to single-firm

24conduct, remains important to ensuring competition in our

25high technology markets.

172

1Second, we have seen in recent years the

2creation and abuse of monopoly positions through conduct

3that serves no useful purpose and therefore should be

4counteracted by the antitrust laws.

5Third, the antitrust laws must remain flexible

6and responsive to these ever-changing conditions. Blind

7reliance on outmoded principles, and even more

8importantly, a refusal to consider the particular facts of

9a particular case is a terrible mistake that the courts

10and the agencies should not make.

11I thank the FTC and Department of Justice for

12the opportunity to speak today and for your thoughtful

13consideration of these important issues.

14 (Applause.)

15MS. GRIMM: Thank you very much.

16We'll now take a 15-minute break and we'll

17reconvene here then for the round-table discussion. Thank

18you.

19(A brief recess was taken.)

20MS. GRIMM: I'd like to start this portion of

21our program by asking our two panelists if they would like

22to comment in any way on each other's presentations and

23respond to any questions between them

24Would either of you like to comment or ask any

25questions?

173

1MR. McCOY: I think I'm going to pass. I think

2I'm here to answer your questions

3MS. GRIMM: Okay. What we're going to do is

4very similar to what we did this morning. We're going to

5ask some general questions, then we're going to ask some

6specific questions on predatory buying that Michael will

7answer and some questions on loyalty discounts that we'll

8talk about with you, Tom

9MR. McCOY: Great.

10MS. GRIMM: So, to begin, we have heard a lot

11this morning about the lack of uniform standards among and

12between antitrust enforcement agencies throughout the

13world. And AMD operates globally, clearly. I believe

14that you filed a complaint against Intel in Japan, Korea,

15the EC, and of course the case in District Court in this

16country.

17Could you please address the question of

18standards, whether they are different globally, and also

19tell us if it does cause a problem for AMD or whether it

20is not a problem?

21 MR. McCOY: I'd be glad to.

22We did not file a complaint in Korea --

23MS. GRIMM: Oh, I'm sorry.

24 MR. McCOY: In fact, we found out about the

25investigation of Korea in Intel disclosures, so ... But,

174

1more generally, it's a very interesting time, I think, for

2Commission authorities around the world, particularly as

3the world has globalized and the markets are global. And

4AMD and Intel, for example we are the only two suppliers

5of X86 processors for the world. The whole world is

6dependent on us and probably eighty percent of IP runs on

7X86. And I think more and more we're seeing business

8conditions like that.

9My experience is that there is a great

10opportunity. It shouldn't be viewed as a difficult

11problem, as Judge Posner has posited in some of his

12remarks. I think it's an opportunity for the mature

13competition authorities around the world to establish

14their common ground.

15And, in my experience, there is tremendous

16common ground that I don't see really any outlines out

17there when it comes to the valuation of unilateral conduct

18by dominant companies. I see an effort to come together

19on guiding principles as to what the desired results

20competition policy are.

21It's about competition, not about competitors.

22It's about innovation. It's about competitiveness. You

23can't have competitiveness without competition. It's

24about consumer value and consumer choice. It's about a

25very thoughtful look at barriers to entry and their

175

1permanency relative to assumptions about their transient

2nature. And it's about looking for behavior that makes no

3sense for a long period of time. So, where rational

4business people are making irrational decisions that

5suggests that there is a persistent problem.

6And, in my experience around the world in

7today's agencies, there is tremendous interaction between

8those people involved in policy, those people involved in

9economics, and those people involved in advocacy, in

10trying to bring together guiding principles where we can

11all agree that the values of antitrust enforcement have

12been historically used in this country in terms of

13promoting efficiency and consumer welfare are far more

14common.

15MS. GRIMM: So, just following up on that, you

16really don't perceive it as a problem. Is that

17overstating it?

18MR. McCOY: I have not seen a problem and I have

19not seen -- and, I'll be honest, in the AMD and Intel, you

20know, fronts around the world, I don't see a big set of

21differences in the way that people are looking at this.

22 We may get into a little bit more of that when

23we look at retrospective rebates. But in terms of what's

24the appropriate focus, you know, what's happening to the

25innovation process, what are the barriers to entry, why do

176

1we have persistent behavior that is out of character for

2people who are smart business people, why has it endured

3so long, and what are the effects on the innovation

4process and the effect on consumers, I think everybody is

5asking the same questions,.

6MS. GRIMM: Let's follow up with the loyalty

7discount and just take rebates and loyalty discounts as

8one type of conduct that we're looking at.

9Is there any difference in the standards that

10you perceive that are being applied in different

11jurisdictions as to that particular subject?

12MR. McCOY: I believe that in my experience, and

13let me make it clear, I don't pretend to be the latest,

14you know, gift to antitrust academics, but I have been

15around the block in my career on all these issues.

16I think the law is pretty settled and policy is

17pretty settled every where in the world but here in the

18United States about retrospective rebates. And I think

19one has to be careful to take a hard look at what really

20happens in a marketplace, beware of labels. Because we

21can all agree that price competition is a good thing. And

22we can all agree, generally speaking, that a discount is a

23good thing.

24But a retrospective discount or rebate, and I

25use those words in quotes, is usually, when deployed by a

177

1monopolist, not a rebate or discount at all. It's a price

2coupled with a threat of a price increase it can go to

3here in demands for market share and monopoly margin.

4So, there's simply a device, a mechanism, to

5impose a penalty on capital customers from erring to try

6to balance out their suppliers.

7MS. GRIMM: This morning I believe we heard that

8with respect to discounts there really is no standard

9that's generally accepted even in this country.

10Do you agree with that or not?

11 MR. McCOY: I think that (a) the way that most

12jurisdictions look at this is in terms of exclusion.

13What's really happening is a matter of fact.

14What is really happening, which requires a look at

15relative market share. But I believe that most of the

16world looks at it in terms of exclusion.

17In this country, I think the debate is very

18confused and there are a lot of discussions about words

19and concepts, but they tend to be -- discussions tend to

20be somewhat divorced from what really happens in the

21marketplace, in my experience.

22So, I don't think we have a settled view on when

23and if a dominant firm should be permitted to use a

24retrospective rebate. And I think the debate in the U.S.

25is far behind some of the more closed debates and

178

1jurisprudence of other jurisdictions, where they've had a

2lot of experience in looking at them and actually coming

3to decisions and enforcement actions. They're coming up

4with remedies.

5MS. GRIMM: Let me follow up on that also.

6What remedies are they coming up with with

7respect to discounts that are found to be illegal?

8MR. McCOY: Well, I encourage everybody to

9actually look at what they do rather than rely on me. As

10I said, I don't pretend to be a professor.

11But they're fairly clear remedies in the other

12jurisdictions about preventing quantity-forcing

13contractual terms.

14And, in fact, as I observed in my opening

15prepared remarks, we have a very clear example coming out

16of the Microsoft case, where you have a quantity-forcing

17term that Microsoft had imposed on the world, which is

18basically you're selling a computer, you're going to pay a

19royalty to us whether you are selling that computer with

20an operating system or not.

21And everybody agreed that was clearly above the

22line as a quantity-forcing predatory contractual term.

23And there's no reason why in and out of this context we

24can't figure out appropriate, clear and fair remedies here

25as they have elsewhere.

179

1MS. GRIMM: In your view, are DOJ and the FTC

2failing to challenge single-firm conduct that they should

3be challenging? And, if so, what types of conduct?

4MR. McCOY: Well, I think that we are in a

5period of having a very healthy and appropriate debate

6about when there should be regulatory intervention into

7managed markets where the management is as a result of the

8unilateral conduct of the dominant firm.

9And, particularly in a world that is changing

10rapidly and globalizing, it's very -- I think it's very

11appropriate to step back and take a look at -- a fresh

12look at the policy objectives that underlie antitrust law

13and policy and enforcement, and whether the tools, the

14analytical tools, are the right tools, whether the right

15facts are being evaluated, the right priorities being set,

16and whether enforcement is appropriate and effective.

17And that is likewise appropriate that that be a

18global debate. As I said, it shouldn't be viewed as a

19problem or a burden. I think it should be viewed as an

20opportunity for competition authorities around the world,

21particularly in mature jurisdictions and marketplaces to

22try to find as much common ground as possible, and I

23believe it can be done. In fact, progress has probably

24been made.

25

180

1appears, frankly, that there has been a retreat from

2Section 2 enforcement, and that not getting the same kind

3of energetic investigation and enforcement of Section 2 in

4unilateral conduct, which to me is surprising when we look

5at the continued investment of resources appropriately.

6MS. GRIMM: Mike, are you there?

7MR. HAGLUND: Yes, I'm here.

8MS. GRIMM: May I ask you the same question?

9Are the FTC and the DOJ failing to challenge

10single-firm conduct that they should be challenging? We

11know about predatory buying. Are there any other forms of

12conduct that you encountered in counseling your small- to

13medium-sized clients that we should know about?

14MR. HAGLUND: Well, I think that there is a --

15what I've observed in the last five, ten years is a shift,

16I think, in emphasis at the national levels by the Federal

17antitrust agencies to having a greater concern with

18national markets and international markets. And I think

19that with that -- and some of that is understandable.

20Some of it I think is a mistake because I think

21that when one really drills down into some of these lower

22tech industries that I've been involved in, you find real

23regionalization and relevant distinct markets that meet

24the test of that term for purposes of antitrust law and

25can be significantly hurt in terms of their competitive

181

1health unless there's significant enforcement of the

2antitrust laws.

3And I think that more energy needs to go into

4knowing the facts of those local and regional markets

5because the smalls tend not to be able to watch out for

6themselves because of the level of antitrust expertise out

7there generally. And I think that the states vary widely

8in terms of the level of commitment they have to antitrust.

9So, I think there's more to be in that sector.

10MR. McCOY: Can I make a positive comment?

11 To give you an example of what the technology

12industry would view as a very, very good signal. The

13Federal Trade Commission has obviously invested an

14incredible amount of time and resources into the Rambus

15situation. And I am not carrying a brief on either side

16of those issues, but those issues are very important.

17 They're very important to innovation and

18competitiveness. They're very important to market entry.

19And they're very timely. Market standards are a very good

20thing from the consumer welfare perspective. They drive

21scale and they drive the entrepreneurial opportunity.

22And I think that we have a lot of evidence now

23to evaluate how standards are a very, very positive thing.

24They drive competitors and innovation, and therefore, the

25integrity of the standardization process is something that

182

1should be really looked at very carefully. And when there

2is not integrity in that process, the world needs to know

3that there is going to be enforcement.

4However the Rambus case ultimately comes out, I

5think the Federal Trade Commission sends a very

6appropriate signal to the marketplace that this is

7important and it's strategic, and it's quite clear that

8there is going to be some behavior that is simply not

9going to be tolerated.

10MS. GRIMM: Let me kind of reverse the question

11and ask the opposite.

12Based on your experience, are there certain

13types of conduct that are benign or procompetitive,

14deserving of more lenient treatment than they are

15currently afforded?

16Either one.

17MR. HAGLUND: I guess I come at it from the

18standpoint of looking at the forms of anticompetitive

19conduct being able to take many, many different shapes.

20One of the interesting things I heard in Tom's

21talk was his reference to the potential that a mix of acts

22can work very effectively for a dominant firm. In the

23Weyerhaeuser case, for example, we had 15 different types

24of anticompetitive conduct, but all the attention has been

25showered on predatory buying, but in fact the table was

183

1set for the price-raising behavior in the log market by

2exclusive contracts, by a number of other anticompetitive

3tactics that worked together in combination to become

4effective overall.

5But I guess I'm not able to identify conduct

6that should be benign, other than that I do see some of

7the rationale for why Brooke Group was decided wanting

8to immunize price cutting with the price cost test in

9terms of not trying to hinder or chill price cutting

10conduct.

11But where it's beyond that, I have trouble -- my

12experience doesn't reveal areas where I think there's too

13much attention or it shouldn't be used.

14MR. McCOY: Well, I have been practicing law and

15business for over thirty years now and been through many

16different seasons of policy views and the relative

17oversight by competition authorities.

18And I guess I would say this: In my career, I

19have never seen a company hold back. I mean, it's a

20hardball world out there and I've not seen a client in the

21days I was a law partner or certainly at AMD where

22businesses were pulling punches because of worry about the

23activity. So, that's number one.

24Number two, depending on what side of the bar

25you sit on, in any particular matter, you always have one

184

1side that wants to disaggregate all the behavior and just

2look at everything piecemeal. But the reality, the

3reality of life in the business world, is that there is a

4tapestry of activities. That's just the way the world

5works.

6And one really does have to be careful of trying

7to judge the beauty of the picture by just looking at the

8eye or the ear or the nose. You really have to look at

9the whole thing.

10And, finally, I think that the challenge is

11always going to be pretty much the same because, if a

12company is fortunate enough to have a dominant position,

13however they got there -- let's assume they got there

14through skill -- and they're now enjoying a big market

15capitalization of software, they're going to do everything

16that they can to protect that market place. And that's

17what they're going to do.

18And, therefore, there's always going to be, in

19my view, need for a strong antitrust policy articulation,

20communication and enforcement, because otherwise you're

21going to end up with cultures, business cultures, that

22their compliance programs are not going to be able to keep

23under control.

24MS. GRIMM: I'd like to turn to a little

25different subject now.

185

1 As you may know, antitrust lawyers and judges

2are battling -- I guess that's too strong a word -- but

3how much weight do you give to business documents

4containing evidence of bad predatory intent? What

5consideration in your view should the antitrust enforcers

6give to intent documents in assessing a firm's conduct?

7MR. HAGLUND: Well, I think you hear two schools

8of thought on this. One is that, oh, every good business

9wants to kill its competition, that's just the way of the

10world in terms of being a good competitor. You hear

11experts talk about juries getting too carried away about

12statements that they think are just characterizations of a

13robust effort to compete hard.

14And I think you need to distinguish between

15cheerleader-type phraseology that somebody might use in an

16e-mail, which I don't find to be terribly meaningful, and

17the documents that really help demonstrate what the intent

18is relative to a particular business practice and its

19ultimate effect on the structure in the industry.

20And where the documents really -- where I find

21intent helpful, and I think this is where the court in

22Microsoft and a number of Supreme Court cases have said in

23"Aspen," for example, and "Trinko," what's important,

24intent can help give one a means of interpreting what are

25otherwise ambiguous acts and give you a more firm and

186

1clear view of what the defendant really intended. And

2especially if they speak to the structure and the change

3they wish to achieve in the industry. And if they're

4already above the fifty percent mark, then I think it's

5very helpful stuff.

6MS. GRIMM: Tom, any views?

7MR. McCOY: Well, I think that government

8officials involved in antitrust enforcement should look at

9everything. But I think everybody agrees that the

10documents that a trial lawyer would love on the

11plaintiff's side have to be looked at objectively and in

12context. That of course a dominant company is going to

13try to preserve that dominant position. That's what

14they're going to do. That's what they're paid to do.

15That's what their shareholders expect them to do.

16So, documents that manifest that obvious

17reality, so what.

18But I think that it's important, you know, in

19being a fact-finder, being a dispassionate fact-finder and

20evaluating, you know, the purpose of a strategy and

21whether the advocates are credible or not in trying to

22defend whether the strategy is being pursued for

23reasons that really relate to growing a market, satisfying

24a customer, being creative and innovative in products and

25marketing, or whether it's simply a design, and a heavily

187

1lawyer design, for a monopoly to use their power to

2preserve a monopoly.

3One needs to look at what people say about what

4it is they're doing, particularly trying to get a hold of

5the evidence that matches up externally as to what is the

6marketplace perceiving as to why the dominant company is

7doing what it is doing.

8And I think it is the unity of the evidence on

9those boundaries that can be generally fairly helpful

10figuring out whether it's just straight forward hardball

11business or whether it's a monopoly simply trying to

12protect its position using their power.

13MS. GRIMM: Thank you. I think you're pretty

14much in agreement on that question.

15MR. McCOY: And I believe, by the way, that that

16is the view of most of the people in the other

17jurisdictions in terms of when they're looking at

18evidence. I think your colleagues and sister agencies

19from around the world all say, look, if we get a document

20from a lower-level sales employee that says, you know,

21we're going to go kill those guys, that we would take that

22document with somewhat of a grain of salt. That, standing

23alone, doesn't tell us any about structure, about

24efficiency, and certainly about what's happening in the

25industry.

188

1MS. GRIMM: When we were doing some background

2research, Google research for this panel, we came across a

3recent article in "Fortune," August of 2006, that quoted you.

4And it quoted you as saying, "As a matter of economics,

5the monopolies probably begin somewhere between thirty

6percent and thirty-five percent," and it then goes on to

7explain that at this point a rival's rising market share

8would imperil a dominant firm's hold on a market. You were

9talking about Intel in this article.

10Do you have any experience in suggesting that

11attaining any particular market share, whether it's thirty

12or thirty-five percent or whatever, has particular

13significance for competition against a large competitor?

14MR. McCOY: Well, my comments were in the

15context of the X86 processor market where Intel has, for

16more years than I can count, enjoyed a revenue share of at

17least eighty percent, and there's really no other rival,

18but that which typically had a revenue share of somewhere

19in the ten to fifteen percent range.

20And so in order to think about specific points

21where monopoly power begins to erode, you need a lot of

22context, you need to know where the companies are starting

23from, and you need to know a lot about the various

24entries, and you need to know a lot about what is the

25psychology of the marketplace. Because one of the things

189

1that gets missed in the academic debates is that markets

2are comprised of real people making human decisions. And

3so, that psychological, you know, culture of the market

4explanation has been patterned by monopoly behavior.

5My comments are taking a look at where we are

6and where the competitor is and the penalties that are

7imposed or that have been imposed on customers for

8incremental market share provided to us, and where we

9would have to be as a revenue share before we could

10overcome those kinds of penalties.

11And one of the examples that I talked about in

12the prepared remarks is that, in a situation where you go

13to a very big and powerful company and you say, we're

14going to give you a million units for free, units where

15probably your average procurement cost is running at least

16$150.00, we're going to give you a million of them free.

17And they can't be used, they can't be used because the

18penalty, the retaliatory penalty that is imposed for not

19maintaining market share margin of the incumbent, tells

20you something about you got a ways to go as a matter of,

21quote, economic -- economics. Capital markets and

22psychology you can amass what you need to overcome the

23barriers that have been erected that you have to get over,

24particularly in markets where only a small slice of it is

25contestable in any relatively short term or intermediate

190

1period.

2In some markets, a company could wake up on

3Friday and say, on Monday I'm going to buy twenty percent

4more of my needs from a different company. But that's not

5true in technology.

6In technology, there is -- a lot of switching

7costs takes time. It can't be done quickly. And,

8therefore, getting a relevant market share to be able to

9overcome the power of the tendency is difficult.

10MS. GRIMM: Let me follow up with just one

11further question.

12With respect to loyalty discounts and rebates,

13does market share provide any kind of useful screening

14mechanism that we could use for assessing legality?

15MR. McCOY: Well, yes. But, again, I believe

16you have to look at market share and I think you have to

17look at entry, and you have to have in mind the relative

18margins of a monopoly supplier and the customer base.

19So, you can have a situation, as we do in

20technology, where you have an ingredient supplier with

21margins that are -- operating margins in the forty percent

22range, serving customers whose operating margins are

23in the zero to six percent range. And they're public

24companies, with people who are trying to manage

25shareholder expectation, capital market expectations,

191

1employee morale, and their tenures, with a board of

2directors looking over them.

3So, I don't think there are any bright lines

4here. I know everybody wants a bright line and everybody

5wants to talk about safe harbors. But in the real world,

6there are a number of factors that I think is a matter of

7making sure that you're doing the right thing in the right

8market at the right time.

9The unfortunate reality is, from a resources

10vendor standpoint, that a fair amount of homework should

11be done. But certainly in marketplaces where you have an

12enduring monopoly that is enjoying fifty, sixty or more

13percent of the revenue share, that tells you, frankly any

14time you have a dominant company using a retrospective

15rebate, it's -- in my experience, the odds are one hundred

16percent that a retrospective rebate is being used for no

17other reasons.

18MS. GRIMM: Mike, I'd like to ask you one more

19question on predatory pricing, then we're getting pretty

20close to closing the session.

21You've practiced, as you pointed out, for many

22years representing small- and mid-sized resource

23companies.

24Is the issue of predatory buying, the type of

25conduct that we saw in Ross-Simmons, is it rare, or is it

192

1more common practice than the case law might reflect?

2MR. HAGLUND: I think it's fairly rare. And it

3happens only, from what I've seen in these markets -- at

4least in the resource sector, in markets where the supply

5of the inputs is fairly elastic and -- I mean, alder, for

6example, doesn't get harvested except as a byproduct of

7the much larger softwood harvest in the Pacific Northwest.

8Fish stocks, for example, that are so rigidly regulated.

9Those are the kinds of markets where a really predatory

10dominant buyer can eliminate its processor or sawmill or

11other competitors.

12But, in looking at the case law, there are a

13very, very few number of cases. And in my own experience,

14there are so many resource markets, you don't see any

15evidence of it.

16So, in the big picture of things, it is a

17relatively rare situation.

18MS. GRIMM: Joe, would you like to close with

19any questions that you might have of our panelists?

20MR. MATELIS: Sure, I'll ask one.

21I guess this is primarily for Tom, although I'd

22be interested in Mike's thoughts.

23One of our panelist at the morning session

24talked about, in view of the emerging overlapping

25international enforcement that's taking place, what he

193

1termed a principle of comity and, in general, it's the

2notion that there ought to be principles where one

3enforcement agency presumptively takes the lead on a

4certain matter. He proposed home jurisdiction and there

5had been other proposals.

6I'd be interested in your thoughts on the

7potential problem of overlapping enforcement across

8countries.

9MR. McCOY: Well, as I said, the issue of

10harmonization across the borders in the competition

11network, I think that's very important.

12I think that particular proposal is absurd. If

13you were to apply that proposal, particularly with any

14view of the way the world is going to look to AMD and

15Intel, you would conclude that the dispute should be

16resolved in the states.

17And, the fact of the matter is, for AMD and

18Intel, if you were to take -- our revenues are probably

19seventy-five percent coming from outside the U.S. We are

20-- big multinational companies are citizens of the world.

21We have productive capacity all over the world. We have

22employees all over the world. The innovation process is

23one that is built on human resources located around the

24world, in no particular jurisdiction. And the marketplaces

25are global.

194

1So, to look at where a company is chartered or

2where the CEO sits is not a relevant variable to determine

3competition policy.

4MR. MATELIS: Just to press you a little bit on

5that: Even if we don't like that specific proposal, is

6overlapping enforcement from different countries something

7that we ought to be worried about or a healthy thing?

8MR. McCOY: Well, I think that I'll be -- I

9think the competition authorities should compete, just to

10throw out a radical thought.

11MS. GRIMM: We heard that [laughter].

12MR. McCOY: No, I'm serious, that there should

13be intellectual competition. And that's the free flow of

14ideas, just like free trade in IP. Nobody has a monopoly

15on these ideas.

16But be careful when you talk about who ought to

17take the lead. I don't think it's ever going to, in the

18practical world, occur, because in a globalized world,

19what a dominant company does in any particular

20jurisdiction affects all the other jurisdictions. So, for

21example, I think one of the reasons why Europe became so

22active in the Intel investigation after Japan is because

23it was so clear that the behavior that was judged to be a

24violation of the antimonopoly laws and the public policies

25in Japan had a direct effect on consumers in Europe.

195

1So, when you have these -- when you have a more

2globalized world where the dominance, you know, extends

3globally, behavior anywhere can affect consumers

4everywhere. And in those scenarios, I just don't

5think it's -- one has to be practical, including

6politically practical. To think that any jurisdiction is

7going to advocate or forebear the protection of its own

8consumers in favor of another jurisdiction, that would be

9a remarkable thing. And I just don't think it's healthy.

10MR. HAGLUND: I'd agree with Tom.

11MS. GRIMM: And on that note, it is a little

12past 4:30, I believe. Yes.

13I again want to thank our panelists for

14participating in our hearings today. I'd like everyone to

15please join may in a round of applause for them.

16(Applause.)

17MS. GRIMM: I'd also add you're all invited to a

18reception following this hearing. It will be at the

19Woman's Faculty Club over here

20You're also invited to join us tomorrow. We're

21going to have a number of very distinguished faculty

22members from both Berkeley and Stanford. The session in

23the morning will be from 9:30 to noon, and the afternoon

24session will be from 1:30 to 4:30.

25Thank you all for attending. I think our

196

1panelists did a remarkable job. Thank you.

2(Applause.)

3(Whereupon, at 4:35 p.m., the hearing was

4concluded.)

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

197

1C E R T I F I C A T I O N O F R E P O R T E R

2

3DOCKET/FILE NUMBER: P062106

4CASE TITLE: SECTION 2 HEARING, PREDATORY PRICING

5DATE: JANUARY 30, 2007

6

7I HEREBY CERTIFY that the transcript contained

8herein is a full and accurate transcript of the notes

9taken by me at the hearing on the above cause before the

10FEDERAL TRADE COMMISSION to the best of my knowledge and

11belief.

12
 DATED: February 17, 2007


13

14
 ______________________________


15
 KATHLEEN CARR MEHEEN, CSR 8748


16

17

18

19

20

21

22

23

24

25