Business Testimony

1

1	UNITED STATES FEDERAL TRADE COMMISSION
2	and
3	UNITED STATES DEPARTMENT OF JUSTICE
4
5
6
7	SHERMAN ACT SECTION 2 JOINT HEARING
8	BUSINESS TESTIMONY
9	TUESDAY, JANUARY 30, 2007
10
11
12
13
14	HELD AT:
15	UNIVERSITY OF CALIFORNIA AT BERKELEY
16	2220 PIEDMONT AVENUE
17	WELLS FARGO ROOM
18	BERKELEY, CALIFORNIA
19	9:30 A.M. TO 4:35 P.M.
20
21
22
23
24	Reported and transcribed by:
25	Kathleen Carr Meheen, CSR 8748

2

1	MODERATORS
2	Morning Session:
3	WILLIAM E. COHEN
4	Deputy General Counsel for Policy Studies
5	Federal Trade Commission
6	and
7	JOSEPH J. MATELIS
8	Attorney, Legal Policy Section
9	Antitrust Division, U.S. Department of Justice
10
11	PANELISTS
12	Morning Session:
13	Michael D. Hartogs
14	David A. Heiner
15	Scott K. Peterson
16	Robert A. Skitol
17
18
19
20
21
22
23
24
25

3

1	MODERATORS
2	Afternoon Session:
3	KAREN GRIMM
4	Assistant General Counsel for Policy Studies
5	Federal Trade Commission
6	and
7	JOSEPH J. MATELIS
8	Attorney, Legal Policy Section
9	Antitrust Division, U.S. Department of Justice
10
11	PANELISTS
12	Afternoon Session:
13	David A. Dull
14	Michael E. Haglund
15	Thomas M. McCoy
16
17
18
19
20
21
22
23
24
25

4

1	C O N T E N T S
2	MORNING SESSION (BUSINESS):
3	Introduction
4	Presentations:
5	David A. Heiner
6	Scott K. Peterson
7	Robert A. Skitol
8	Michael D. Hartogs
9	Moderated Discussion
10	Lunch Recess
11
12	AFTERNOON SESSION (BUSINESS):
13	Introduction
14	Presentations:
15	Thomas McCoy
16	Michael E. Haglund
17	David A. Dull
18	Moderated Discussion
19
20
21
22
23
24
25

5

1	P R O C E E D I N G S
2	* * * * *
3	MR. COHEN: Good morning. I'm Bill Cohen,
4	Deputy General Counsel for Policy Studies at the Federal
5	Trade Commission. I'm going to be one of the moderators
6	at this session. My co-moderator, who is sitting next to
7	me, is Joe Matelis, an attorney in the Legal Policy
8	Section of the Antitrust Division of the U.S. Department
9	of Justice.
10	Before we start I need to make a few
11	housekeeping announcements. As a courtesy to our
12	speakers, we'll urge you all to be sure that you've turned
13	off your cell phones, Blackberries, and any other devices
14	that might ring, vibrate, play music or anything like
15	that.
16	The other point that I need to make is that
17	these panels are being run as hearings involving the
18	moderators and the participants. So, consequently, we
19	request that the audience not make comments or ask
20	questions during the sessions. Thank you on that.
21	Before introducing our speakers, what I'd like
22	to do is first thank the University of California at
23	Berkeley for hosting the FTC/DOJ Section 2 hearings on
24	business testimony. And in particular I'd like to thank
25	Howard Shelanski and his colleagues, Richard Gilbert and

6

1	Paul Shapiro, for offering us their facilities and for
2	making the necessary arrangements for these hearings to go
3	forward.
4	I'd also like to thank the Competition and
5	Policy Center, the Berkeley Center for Law and Technology,
6	and the Haas Business School, for providing the
7	facilities, refreshments, videotaping, and webcasting
8	capabilities, and for working with the agency staffs to
9	provide other logistical support. Arranging hearings like
10	this takes quite a bit of that and we thank you.
11	Others who provided tremendous help with the
12	additional details include Bob Barde, Louise Reed, and
13	Dana Lund in the audiovisual crew. Our thanks to them as
14	well.
15	Finally I would like to thank the FTC and the
16	DOJ Section 2 team members. And within the FTC
17	delegation, Pat Schultheiss and Jim Taronji in particular,
18	who I know have worked very hard to put together these
19	sessions and all the other sessions that we've held to
20	date, and the FTC's San Francisco Regional Office for
21	their help and support on this occasion.
22	We're honored to have assembled the various
23	members of the panel from a number of companies that have
24	agreed to offer their testimony in connection with the
25	hearing sessions. These panelists have broad perspectives

7

1	on how the companies operate within the complex and
2	globally diverse realm of Section 2 jurisprudence. We
3	anticipate that they will help us to identify and better
4	understand areas where single-firm conduct may cause
5	competitive harm, areas where desirable, procompetitive
6	behavior may be being chilled, and areas where additional
7	antitrust guidance would be useful.
8	Our panelists, and I'll name them in the order
9	that they'll be speaking this morning, are David Heiner,
10	who is the Vice President and the Deputy General Counsel
11	for Antitrust at Microsoft Corporation; Scott Peterson,
12	who is Senior Counsel at Hewlett-Packard Company; Robert
13	Skitol, who is the Senior Partner in the Antitrust
14	Practice Group at Drinker Biddle & Reath in Washington,
15	D.C. and counsel to the VMEbus International Trade
16	Association; and Michael Hartogs, who is the Senior Vice
17	President and Division Counsel at QUALCOMM Technology
18	Licensing.
19	Detailed bios for all of our speakers are in a
20	packet on the table in the back of the room, as well as on
21	the agencies' websites.
22	As to format for this morning, what we're going
23	to do is we're going to allow each speaker some time,
24	about twenty to thirty minutes if they wish, for a
25	presentation. Then after all the presentations are

8

1	finished, we'll likely take a break for around fifteen
2	minutes. After the break, we'll reconvene for a moderated
3	discussion with our panelists.
4	The sessions today are an extremely important
5	component of the Section 2 hearings overall. FTC Chairman
6	Deborah Majora made it clear at the opening session that
7	she hoped to learn from the presentations of businesses
8	through testimony of their executives and their advisers.
9	As Chairman Majoras noted, "The hearings will
10	that have panels that will focus on specific types of
11	conduct that at least to date, can implicate liability. We want
12	the panels to discuss the conduct from the market perspective
13	from the ground up, that is, examine why and when firms
14	engage in it, how they do it, and what effects it produces
15	for the firm, for other firms (customers and competitors),
16	and for consumers. We should look at whether firms in
17	competitive markets engage in the same conduct and, if so,
18	examine why they do it. We want these discussions, to the
19	extent possible, to include knowledgeable business people
20	or at least their advisers."
21	Well, I think over the last seven months or so,
22	we have held conduct specific hearings on predatory
23	pricing, refusals to deal, tying, exclusive dealing,
24	bundled and loyalty discounts, and misleading and
25	deceptive conduct. Some of these panels include business

9

1	executives or their legal advisers. Today we're going to
2	have them talk.
3	The sessions will bring together a number of
4	panelists who are able to speak with a business
5	perspective, in keeping with our goal of obtaining as much
6	practical insight and real world experience as possible.
7	We look forward to our panelists' remarks and a
8	round-table discussion
9	I want to thank all of today's panelists for
10	their participation. We appreciate it. It takes a great
11	deal of time to prepare for and participate in hearings
12	like this. And we know that you're all extremely busy
13	individuals. So, again, thank you for your time and your
14	efforts.
15	What I'd like now to do is to turn this over to
16	my DOJ co-moderator, Joe Metalis, for any remarks he'd
17	like to add.
18	MR. MATELIS: Thanks, Bill. The Department of
19	Justice's Antitrust Division is extremely pleased to
20	participate in these hearings. In the single-firm conduct
21	hearings we have held to date, we have benefitted from the
22	insights of many highly skilled antitrust attorneys and
23	economists.
24	Today's hearings, and the hearings to be held
25	next month in Chicago, grow out of the belief that we can

10

1	also learn much about single-firm conduct from the
2	perspective of businesses themselves. Our panelists today
3	are people who must help devise and implement business
4	plans, aware that their firm's unilateral conduct may be
5	challenged in private or government litigation or by
6	foreign competition authorities. Their companies are also
7	directly affected by the conduct of other firms.
8	Whether you have had occasion to view Section 2
9	of the Sherman Act as a sword directed at the heart of
10	your business or as a shield protecting you from
11	anticompetitive conduct, we look forward to hearing from
12	you and about your perspectives today.
13	On behalf of the Antitrust Division, I would
14	like to take this opportunity to thank the Berkeley Center
15	for Law and Technology and the Competition Policy Center
16	at the University of California Berkeley for hosting these
17	hearings today.
18	And I'd also like to thank on behalf of the
19	Antitrust Division all of our panelists. I know it takes
20	a lot of time and thought to prepare for these and we're
21	truly appreciative of your efforts to improve our efforts
22	of protecting consumers.
23	Finally, I'd like to thank Bill and his
24	colleagues at the FTC for all of their hard work in
25	organizing today's hearing and assembling the fine

11

1
2	MR. COHEN: Our first speaker this morning will
3	be David Heiner, who I just mentioned is the Vice
4	President/Deputy General Counsel for antitrust at
5	Microsoft Corporation. Mr. Heiner is responsible for
6	antitrust counseling and representation of the company
7	before antitrust agencies and compliance with agency
8	rulings.
9	Since joining Microsoft in 1994, Mr. Heiner has
10	played a leading role in Microsoft's response to
11	government antitrust proceedings in the United States,
12	Europe and Asia.
13	Mr. Heiner is a graduate of Cornell University,
14	with a bachelor's degree in physics, and a graduate of the
15	University of Michigan Law School. He's the author of a
16	2005 article, "Assessing Tying Claims in the Context of
17	Software Integration: A suggested framework for Applying
18	the Rule of Reason Analysis."
19	So, now we'll turn it over to David.
20	MR. HEINER: Thank you very much, Bill and Joe,
21	for the opportunity to present here today. My colleagues
22	at Microsoft and I really appreciate the opportunity to
23	contribute to these proceedings.
24	We were asked to provide a business perspective
25	on living under Section 2 of the Sherman Act. I think

12

1	it's fair to say that Microsoft has considerable
2	experience in this area, probably more than most companies
3	might wish for, to be honest. And not only Section 2 of
4	the Sherman Act, but also Article 82 in Europe and
5	comparable provisions around the world.
6	Section 2 issues are potentially relevant to a
7	broad range of Microsoft's business: product design
8	issues, as well as more traditional subjects of antitrust
9	analysis, such as packaging, pricing and IP licensing.
10	One point comes through loud and clear from the
11	business people when you ask them about their experience
12	under Section 2, as I did in preparation for the
13	presentation today. And that is, as business people, you
14	just want to know what are the rules. If you could
15	provide it to them in clearer fashion than we're able to
16	today, they'd be happy to go devise business strategies,
17	to live within those rules and still be successful.
18	What's really challenging in the Section 2 area,
19	as opposed to, say, Section 1 cartel behavior, is that so
20	often advice has to be provided in shades of gray. That's
21	of course the reality we live with, but this can be
22	challenging for business executives, especially I would
23	say mid-level people and below, who just aren't used to
24	getting that kind of advice, who are busy with their own
25	planning and strategizing, and they look to the law

13

1	department of a company such as Microsoft to give a green
2	light or a red light. And all too often it's a yellow
3	light.
4	You might say, what's new in all of this? It's
5	always been this way. And that's certainly true. But, as
6	the Antitrust Modernization Commission has commented in
7	its draft report, as we move toward a more flexible
8	approach to antitrust analysis over the past thirty years,
9	one side effect has been, less predictability. And
10	it's of course a positive thing that we move to a more
11	flexible approach. But it seems that the combination of
12	that, plus a range of other factors that I'll discuss, are
13	really building upon one another to move to such a level
14	of difficulty in predicting the outcome of various
15	antitrust issues as to create a significant problem.
16	Part of this arises from the rule of reason.
17	And obviously it's a balancing test. So, any time you
18	have a balancing test, it's a fair question as to how a
19	typical judge or agency will do the balance.
20	I think we've got something even deeper going on
21	here, though, in the Section 2 context, in that lawyers
22	and economists often disagree as to whether particular
23	conduct is procompetitive or anticompetitive in the first
24	place, before you even get to any analysis. And that
25	obviously is a really fundamental kind of point.

14

1	Two examples here that I found kind of striking,
2	one is from the Department of Justice case against Microsoft
3	back in 1998. That case, as many of you will remember,
4	primarily concerned the development of Windows 95 and
5	Windows 98 and the inclusion of web browsing functionality
6	in that time frame. There were additional allegations as
7	well.
8	And the DOJ had as its expert economist, world
9	renowned economist, defender of IBM, Frank Fisher. And
10	Professor Fisher came in and looked at the range of
11	conduct, which was a substantial subset of everything
12	Microsoft had done in competing with Netscape, and said,
13	it's all anticompetitive, you know, it doesn't make
14	business sense except for its tendency to exclude and
15	therefore it's anticompetitive.
16	Now, Microsoft got expert testimony from another
17	renowned economist, also from Boston, Dean Schmalensee of
18	the MIT Sloan School of Management. Dean Schmalensee
19	looked at the very same set of practices. And there was not
20	much dispute as to facts. There was some, but basically the
21	facts were understood. He looked at the same set of conduct,
22	and said, not only is it not anticompetitive, this conduct
23	is procompetitive. This is a firm building better
24	products and distributing them broadly to consumers.
25	So, fundamental disagreement among two very

15

1	respected people. Before you get to any balance just is
2	the conduct procompetitive or not?
3	Another example is pertinent today. After the
4	Department of Justice proceedings, there was a proceeding
5	in Europe that also concerned the same issue, which is the
6	integration of new features into a product, again in this
7	case Windows. The European case concerns media play back
8	software. So, this is Windows Media Player.
9	And Microsoft has explained to the European
10	Commission that the purpose of Windows is to be a platform
11	for running applications. So, there's a set of software
12	services in that product. They're exposed to the
13	development community through application programming
14	interfaces. Developers can write to those interfaces and
15	it saves them a great deal of work in creating their
16	applications.
17	And what we said to the Commission is that, part
18	of the value, a big part of the value that Windows
19	provides, is that it's a kind of compatibility layer
20	across hardware from many different computer manufacturers,
21	hundreds of different manufacturers. So, if these
22	manufacturers install Windows, a software developer can
23	run an application, it will run on Windows, and therefore
24	it runs on an HP machine or a Dell machine or Gateway or
25	anything else.

16

1	And the Commission said, you know, we think of
2	the media play back functionality is something separate
3	from the operating system. We don't think it should be
4	there and therefore we think you should offer multiple
5	versions of Windows with and without that functionality.
6	And we said, well, if we do that, it's going to make that
7	functionality less valuable to the developers because if
8	they write to those APIs and a customer has a version of
9	Windows installed where those APIs are not present, the
10	application will not function properly.
11	So, from our perspective, we're saying that
12	maintaining the uniformity of Windows across all these
13	different systems is key to the value it provides and
14	therefore it's procompetitive.
15	And the Commission came back and said, the very
16	thing you're talking about, that's what we see as
17	anticompetitive because only you Microsoft have the
18	ability to add functionality to Windows since you're the
19	only developer of Windows and therefore be able to get it
20	out on virtually every PC since so many PCs are shipped
21	with Windows.
22	And here the competitor was Real Networks. And
23	the Commission's decision was, they will always be on less
24	than the number of machines that Windows is on and
25	therefore they will have a disadvantage that's unfair and

17

1	it's illegal.
2	So, here again, a very fundamental question: Is
3	that conduct procompetitive or not? This case is on
4	appeal to the Court of First Instance in Europe. We
5	expect a ruling perhaps within the next six months, so we
6	might have some decision on that particular point, which
7	will be interesting.
8	So, as I think about the development of
9	antitrust law, especially over the past ten years or so, I
10	think a range of factors are coming together to make the
11	job of an in-house counsel or outside counsel providing
12	antitrust advice even more challenging than it's been in
13	the past.
14	One of these is the development of new business
15	models. Business models with which the law has relatively
16	little experience so far and business models that lead
17	firms to engage in business strategies that wouldn't make
18	sense in traditional brick-and-mortar-type industries.
19	I'm thinking here, for instance, of the development of
20	compatible ecosystems, businesses with network effects,
21	businesses that, as the economists would say, are
22	multi-sided, multiple players involved that a firm is trying
23	to satisfy. With Windows, it's computer manufacturers who
24	license it from Microsoft, and software developers who
25	write applications. Or the Apple iTunes services, where

18

1	you've got the record labels, artists and consumers. Or
2	the Google ad platform, where they're serving websites and
3	developing advertising systems for those websites,
4	advertisers and consumers.
5	In these kinds of markets, it's often the case
6	that it makes sense to give away something that's very
7	valuable, which a competitor might not be giving away,
8	in order to attract users early on and thereby try to
9	generate a network effect.
10	It often makes sense to give something away,
11	again, that someone else might not be giving away, in
12	order to attract one set of players to a market where
13	there's multiple players involved.
14	Interesting questions arise as to business
15	strategy between ecosystems and the compatibility between
16	those systems. So, iTunes, for instance, is I think
17	incompatible by design with other media play back systems.
18	Apple has developed an end-to-end system that works very
19	well. And kind of part of the beauty is they own
20	everything. They own the device, the iPod, the software,
21	the client software, and the service. And they're able to
22	design it to work very well.
23	Well, in Europe at least, they're under attack
24	for that in a very significant way. Very interesting
25	questions that are not really handled in the case books.

19

1	Then we have the fact that in many of the
2	emerging businesses today, business models, characterized
3	by products with very low margin of costs and that soon
4	leads to a range of new business strategies.
5	Bundled pricing, pricing a collection of
6	products or services for significantly less than the sum
7	of the stand-alone pricing. Often highly efficient and
8	valuable for consumers in the case where it costs the firm
9	very little because the marginal cost is little and it adds
10	more value for consumers.
11	In these businesses, based on information and
12	goods, it's often the case that a competitor can very
13	quickly ramp up to satisfy one hundred percent of demand.
14	And that means that when we look at the market share at
15	any given point in time, it doesn't necessarily reflect
16	productive capacity like in the old days, and so that firm
17	doesn't need to build new factories or anything like that
18	in order to satisfy all demand.
19	How do you analyze that in the context of giving
20	antitrust advice?
21	We also see that in these new business models
22	and low marginal cost products many different ways in
23	which you can modify your business. And you end up in a
24	situation where different firms are competing directly
25	with one another but with very different business models.

20

1	So, in the case of Microsoft Windows, the model is quite
2	clear that you primarily earn revenue by licensing the
3	product to computer manufacturers for a royalty. And it's
4	essentially free to software developers who can build
5	applications.
6	Along comes the open source movement and Linux,
7	and here we have essentially a direct competitor, on both
8	the client side and server computers, and that product is
9	free. And we have firms that just -- Red Hat and Novell
10	and others, making a business out of providing service for
11	the software once it's provided to customers. Very
12	different model.
13	Similarly, with Apple, they're making their
14	money by selling the iPod device and they're making money
15	by selling the subscription service to music over the
16	Internet.
17	Many of these new models lead to complex
18	relationships between firms. And that's a point that I'll
19	return to.
20	Another aspect that I think is interesting in
21	terms of predictability is how technology based so many
22	businesses are today. Many of these technologies are very
23	much IP-based, as Windows is. It's nothing but IP.
24	Copyright license that we're providing to computer
25	manufacturers. So, right off the bat in analyzing these

21

1	issues, we are at the always difficult IP/antitrust
2	intersection.
3	Here we are in 2007 and the debate is still
4	going on about whether a patent confers market power.
5	It's a fundamental question that still needs to be
6	resolved.
7	With the focus on new technology, we're seeing
8	an increasing focus on product design. And that again is
9	not something we've seen in the past. Questions regarding
10	integration of new features, not just Windows, but in
11	other contexts as well. How features work; how third
12	parties can connect.
13	And this is an area where, given the complexity
14	of the technology, it can be quite challenging for lawyers
15	and economists to work through these issues. And that
16	complexity of course makes it then an additional degree of
17	uncertainty, with the adviser trying to provide advice to
18	his client.
19	In many cases, technology is so complex we have
20	to turn to experts, to technical experts. They may have a
21	religious view about some of these topics. They may have
22	an axe to grind.
23	And when you have technology, at least in the
24	case of software, which I'm familiar with, it is so often
25	the case that any design can be second guessed because

22

1	there's always a different way something could have been
2	done. So that too adds a degree of uncertainty.
3	When you get into product design, you have the
4	antitrust agencies, or whoever else is enforcing the
5	antitrust laws, having to look at engineering tradeoffs.
6	So, you have a tradeoff between some benefit from an
7	engineering perspective and a competition effect. That can
8	be hard to assess. And you may want to consider the risk
9	that a competition agency, by its very nature, may place
10	much greater weight on a competition concern that is
11	relativity minor, compared to some engineering concern
12	that quite significant.
13	Then you have the challenge of time lags. The
14	development cycle of some of these products is quite long.
15	I mean, it has been famously long for Vista. You have a
16	situation where the engineers need to be told what they're
17	going to build very early on. You know, they're
18	black-and-white people, what are the specifications for
19	what we're building. So, from day one they're looking at
20	what will this product be. And that's when you have to
21	give the antitrust advice. It will be assessed perhaps
22	many years later.
23	Two other factors that I think are making
24	predictions more challenging than in the past. Multiple
25	constituencies involve multiple enforcers. One way to

23

1	reduce antitrust risk from a practical perspective is to
2	try to address concerns before they arise. And we're very
3	much on that path at Microsoft. In connection with a
4	product like Windows, there's a lot of people involved.
5	There's computer manufacturers, there's software
6	developers, there's consumers, there's peripheral
7	manufacturers, there's websites, and others. And everyone
8	has an idea about how it should be built. And, as part of
9	the product design process, we're out there to a great
10	extent getting feedback.
11	We now try to get the legal concerns out early
12	in the process as well and address them. One of the
13	things we find is that different groups may have very
14	different interests. So, the interests of a computer
15	manufacturer such as HP may differ in some cases from the
16	interests of a software developer.
17	We've seen cases recently where even similarly
18	situated firms may have different views about how some
19	things ought to be done. And these views are expressed to
20	Microsoft and agencies in the language of antitrust.
21	I can give you an example here. We released
22	Internet Explorer 7 recently. So, this is a version of
23	the web browser that gets installed on existing Windows XP
24	systems. And this browser, if you used it, has a box up
25	in the corner for searching the web. The design is as

24

1	open as it can possibly be. You can set that box to use
2	any web search engine, you can have multiple web search
3	engines, you can add search engines, you can delete search
4	engines. So, it's all very open.
5	A question arose about what the initial setting
6	would be. So, a customer asks his or her computer to
7	install Internet Explorer 7. The very first time you
8	conduct a search, will it go to Google or Yahoo or AOL or
9	Microsoft, where will it go?
10	And one firm said, you ought to just look at
11	what the existing settings are in Internet Explorer 6.
12	And that would be Microsoft's normal practice in upgrading
13	Windows, you just carry over the settings.
14	Another firm said, you know, the settings are
15	kind of a hard to find within Internet Explorer 6, so they
16	don't necessarily reflect a consumer preference. Why
17	don't you just ask, just say, what would you like the
18	initial setting to be?
19	Both firms felt very strongly about their
20	respective positions. They both expressed their views in
21	the language of antitrust. And we couldn't satisfy both
22	of them. Eventually it was worked out and we have what we
23	think is a compromise solution that we hope they're both
24	satisfied with. But it illustrates the point about the
25	challenges one can face.

25

1	Then we have multiple enforcers. So, when
2	you're making a prediction, it usually is kind of an
3	academic, theoretical question: How would a judge, when
4	presented with all the facts, rule on this. At a much
5	more practical level, you're really saying, how would the
6	Department of Justice look at this? How would the State
7	Attorneys General look at this? How would the European
8	Commission look at this? How would the Fair Trade
9	Commissions in Taiwan, Australia, Japan and others look at
10	this? How would competitors look at this? And competitors
11	are clearly not in a position of a judge applying -- coming
12	up with a perfect result. They have their own parochial
13	interest of course. And consumers. You know, class
14	action lawsuits, we faced two hundred of them in the past
15	ten years or so, many consolidated, but still a big
16	number.
17	So, there's a lot of different enforcers to look
18	at. This is especially significant given globalization.
19	We have a situation today where increasingly firms are
20	running their businesses on a worldwide basis and it's the
21	same business worldwide. These are typically American
22	firms.
23	So, in the case of Microsoft, it is very much
24	the case that it's the same Windows every place in the
25	world. And, again, that's part of the beauty and the

26

1	value of the product: that it is the same. We license it
2	to multinational corporations, so they're taking a license
3	to install it in America and Europe and Asia. They want
4	one licensing paradigm. So, it's very much in Microsoft's
5	interest to have one set of rules that govern all of that.
6	Increasingly we see foreign agencies stepping up
7	their antitrust enforcement, partly as a result of some
8	efforts by the U.S. agencies over the years to have
9	foreign countries adopt and apply antitrust laws.
10	And while that's of course a useful thing, we
11	may find that some of these agencies have differing
12	interests, differing views as to how the antitrust
13	laws ought to be applied. They come from different legal
14	systems. So, in Europe, the development of antitrust law
15	is very much influenced by German thought and French
16	thought, which is somewhat alien to U.S. lawyers coming
17	out of the UK tradition.
18	And then we go overseas where we have matters
19	pending in Japan and Korea, and here you're outside
20	western culture altogether. And we have China developing
21	antitrust laws. That's interesting to think about. How
22	will this Communist country apply the set of rules that
23	really goes to the essence of capitalism.
24	With the stepped up enforcement, we have the
25	prospect of forum shopping. And that clearly is going

27

1	on. So, just this morning, there's an interview with a
2	Brussels-based lawyer, who points out that he's actually
3	from Seattle, who has filed a complaint on behalf of
4	leading American firms against Microsoft in Brussels. And
5	the reason the complaint is filed in Brussels is that it
6	probably wouldn't get very far under U.S. law. But
7	they're hoping for a better, more favorable hearing in
8	Brussels.
9	Another challenge is the broad scope of
10	prosecutorial discretion. When you look at the range of
11	antitrust laws, again, especially in Europe, one can see
12	that there's quite a range of practices that might
13	actually be subject to challenge and yet they're not
14	challenged. So, the counselor has to think about what
15	actually would be the enforcement agenda of these
16	different agencies.
17	In Europe at least, we see the European
18	Commission going after practices for which, in our view, a
19	consensus does not exist that the practices are actually
20	anticompetitive. And I'm thinking here of the discussion
21	paper that came out six months or a year ago.
22	We have, considering how prosecutors and
23	enforcement agencies overseas will exercise their
24	discretion, to focus on their different views of antitrust
25	law. We have the consumer welfare standard in the United

28

1	States pretty well established. In Europe, not so well
2	established. Much more a sense over there that the
3	antitrust laws are designed to protect the small fish from
4	the big fish. The small fish may well be little firms.
5	Mainly in the cases with Microsoft, it turns out they're
6	not. They're the large firms based in the U.S. But in
7	some cases, they may be local small fish. This raises the
8	specter of protectionism.
9	To what extent will trade policy come into play
10	in the application of antitrust law overseas?
11	And then one has to consider the interaction
12	between enforcement agencies. In the United States, Chris
13	raised the perfect discussion about the relationship
14	between the respective rules of the DOJ and the FTC and
15	the states. And here at least we have federalism that
16	moderates that to some extent. There's nothing really
17	comparable going on at the level of Washington, Brussels
18	and other foreign capitals.
19	And what we can see from time to time is people
20	who believe in competition competing very vigorously with
21	one another. So, competition between enforcement
22	agencies.
23	Hew Pate gave a speech a few years ago where he
24	talked about multiple agencies taking a whack at the
25	pinata. And I thought that was really quite apt. In

29

1	Microsoft's case, the central issue we've been dealing
2	with for more than ten years is this question of how the
3	integration of new function into Windows over time ought
4	to be thought about from an antitrust perspective.
5	And we had a major trial on that in the United
6	States. And there was an outcome. And an approach came
7	out of that outcome which focuses on trying to balance the
8	interests of all concerned. And it's an approach where
9	Microsoft is including functionality in Windows, but at
10	the same time, doing so in such a way that opportunities
11	are preserved for third parties to write software that
12	runs on top and can be broadly distributed. So, that's
13	the U.S. approach.
14	Now, the Commission said -- and we tried to
15	explain that approach to the Commission and said the
16	problem is being largely addressed. The Commission said,
17	everything you've done here is all well and good, but it's
18	not enough, and we want you to take it to the next level.
19	And their solution was, do everything under the U.S.
20	consent decree, which was the outcome of this U.S. case,
21	and make multiple versions of Windows with and without key
22	features. Then we get to the point where it's troublesome
23	from a business perspective in providing value.
24	In the case of Media Player, they said
25	explicitly that it's a precedent to be applied in the

30

1	future. So, now we have that additional step where we're
2	talking about multiple versions. And we do have Windows
3	in Europe without Media Player, although no one has
4	purchased it to speak of, less than two thousand units
5	sold.
6	Korea then came along next and said, everything
7	you did in the U.S. is well and fine, and so is everything
8	you did in Europe, but you should take an additional step.
9	And that is, any version that has all the functionality,
10	you should include links to your competitors' products.
11	So, we've done that, too. So, in Korea, the Korean
12	version of Windows, when you boot it up, right there
13	there's a promotion for third party products on the
14	screen. Three difference approaches, each one adding to
15	the other.
16	So, you might say, again, you know, what's new,
17	it's sort of always been this way. And I think it is
18	getting to be a more challenging issue, as I say,
19	particularly how the law will be applied. But then adding
20	to that is really the stakes are higher than ever for a
21	couple of reasons.
22	One is, since we are focused now on product
23	design, we've got a situation where engineers really need
24	to know what we're building. And you saw in my slide,
25	we're having to make decisions. And at that time it may

31

1	be the case that you don't even know as a firm whether you
2	have competitors, much less what their concerns might be
3	for some functionality that you're building. Your
4	competitors may be at the same stage of development as you
5	are, which is it isn't released yet, it's the next
6	generation kind of thing. But you have to make decisions
7	anyway.
8	Years later it will be assessed with a set of
9	facts that didn't exist when you made the decision. This
10	is especially sort of challenging because it's often quite
11	difficult to undo a design decision. It's unlike the
12	traditional stuff of antitrust where you have got a
13	contract, if someone decides the contract is improper, you
14	can change the contract. Well, once the cake is baked and
15	it's on the cooling rack, it's baked. You can bake a
16	different cake next time, but that cake is done.
17	And when it comes to complex products, like
18	microprocessors or cell phone technologies, different
19	parts of the system will rely upon particular features
20	that might have been the subject of antitrust defense.
21	You can change them, but other parts of the system will
22	fail.
23	Third parties, the software developers, may rely
24	on that functionality. If you change it, their products
25	will not work. An example here that I think is quite

32

1	telling is the development of Windows 95. So, in the days
2	before Windows 95, you might remember, we had MS-DOS,
3	which was the character-based operating system then,
4	running on top of that, Windows 3.1. And in about 1990,
5	when those products were really just getting to critical
6	mass at that time, Microsoft set out in its plans to
7	develop Windows 95. Windows 95 was released in 1995, and
8	attacked at that time by some as an unlawful tie of MS-DOS
9	and Windows 3.1.
10	So, what some said was, this product really
11	should be called MS-DOS 7.0. I think seven was the next
12	number in Windows 3.2 or Windows 4.0. Now, the Department
13	of Justice looked at that in connection with a consent
14	decree we were negotiating at that time and it was
15	recognized in those discussions that Windows 95 was an
16	example of good integration. This was a real step
17	forward. It was really building something new. It would
18	not be regarded as a tie of these two separate products.
19	And Windows 95 was released and it was probably
20	one of the most successful products in the history of
21	commerce. Tremendous value provided to customers and the
22	very best of times for the PC industry. Sales of HP and
23	other manufacturers took off, and then we moved right into
24	the Internet era in the late '90s. So, a terrific
25	outcome.

33

1	But still there were claims that that product
2	which was so successful and so valuable could be thought
3	of as a tie. And even today in 2007, as we sit here
4	today, that claim is on trial in a courtroom in Iowa. So,
5	one of our consumer class action cases is pending today
6	and this very issue is being discussed in 2007, twelve
7	years down the road. Now, if the Iowa view were
8	correct, in the view of those plaintiffs, we wouldn't
9	have had Windows 95.
10	Another aspect in which the stakes are higher
11	than ever is the focus on IP licensing. I think we're
12	increasingly seeing firms around the world seeking access
13	to the technology of their rivals on favorable terms. And
14	here again, it's kind of like the product design case
15	where it's an either/or situation.
16	So, your technology is either licensed and made
17	available or it's not. And if it's made available, it's
18	out there, it's gone, you probably won't be able to get it
19	back.
20	In the computer industry context, the IP is
21	often based on trade secrets. Once you have licensed that
22	technology, you can try for protectionism on the use of
23	it, but the trade secrets are out in the world. And once
24	it's licensed, the point of licensing it obviously is for
25	third parties to use it and rely upon it, and if you do

34

1	rely upon it, it would be hard to get it back. So, when
2	you make these decisions, the stakes are high.
3	The rise of global antitrust enforcement is
4	quite significant here. In the European Commission case,
5	a decision was taken against Microsoft relating not only
6	to the product integration issues but also IP licensing.
7	And here the Commission made a decision that Microsoft
8	would have to license protocol technology to third
9	parties. And the Commission observed that it's
10	essentially a global market for this kind of IP and
11	therefore this technology ought to be licensed on a global
12	basis. So, Microsoft is doing that.
13	The Commission has also taken the position that
14	Microsoft ought to license this technology in a way that
15	it can be taken in practice by open source developers.
16	And that's quite troublesome for a commercial firm such as
17	Microsoft because that means that the trade secrets will
18	be revealed to the world. Once the technology is
19	licensed, it will be built into open source products, the
20	source code can be seen, and therefore the trade secrets.
21	Similarly, it's very hard to maintain the value
22	of IP once it's licensed under an open source model
23	because, again, every copy of the product will be made
24	available for free. It's hard having this kind of
25	limitation on sublicensing and royalties coming back.

35

1	Now, it's not the view of the U.S. enforcement
2	agencies that Microsoft should have to make this
3	technology available essentially for free and disclose the
4	trade secrets. This comes up under the consent decree
5	where we have protocol licensing as well.
6	And this is before the European Commission and
7	Microsoft is contesting it at this point and the outcome
8	is yet to be seen. But if the European Commission
9	prevails, then we'll have a situation where you have a
10	split of authority essentially between the U.S. and EU and
11	the EU version will prevail because it's more restrictive
12	because they're seeking greater licensing.
13	In case after case, I think we may see kind of a
14	race to the bottom from the perspective of the target firm
15	in IP licensing. And all of this of course in an economy
16	that is increasingly IP based creates a specter of reduced
17	innovation around IP, and a greater uncertainty as to
18	whether the IP can be properly monetized.
19	So, what are the consequences of all of this?
20	Well, I think we do have a risk at least of over
21	deterrence arising from a combination of the difficulty in
22	predicting the outcomes, the difficulty in changing course
23	later, the variety and number of possible claims, and the
24	desire to avoid controversy.
25	What are the consumer welfare effects of all

36

1	this? Well, we may see limitations on the products'
2	improvement. And there have been cases in the context of
3	both Windows and Office, Microsoft's flagship products,
4	where decisions were made not to include particular
5	features that would have been valuable to consumers based
6	at least in part on antitrust advice. And one might say
7	it was the right outcome or maybe it wasn't the right
8	outcome, but the bottom line is, those features are not in
9	those products.
10	We see antitrust advice from time to time to
11	raise prices. And I always kind of pause, as an antitrust
12	counselor, before saying the price is too low for that
13	collection of products or services. But it's a judgment
14	call based on the state of the law on a worldwide basis,
15	the range of possible claims, that we better raise prices.
16	And clients sometimes get quite confused about
17	that because when we do antitrust training, we usually
18	start at a 101 kind of point that the purpose of
19	antitrust law is more innovation, more output and lower
20	prices. So, they receive this advice with a bit of
21	skepticism, but it's given nonetheless.
22	And I think we're seeing increased R&D costs.
23	For something like Windows, there are six billion dollars
24	of R&D in that product. That's obviously an extreme case.
25	But the amount of time that's spent by executives trying

37

1	to pick through how this shades-of-gray antitrust advice
2	fits with engineering decisions is really considerable.
3	And, finally, I would note that, because of the
4	challenges of predicting how antitrust law will be applied
5	by the multiple agencies and other enforcers, we may see
6	some work that's being undertaken that is of really
7	questionable value but done in order to satisfy a
8	regulatory concern.
9	So, suggestions on how to move forward. I think
10	it's a very hard problem and there probably aren't any
11	easy answers. In trying to move toward greater clarity in
12	the law, I do think it would be helpful if we had a
13	stronger presumption that conduct that is widely practiced
14	by firms without market power is efficient.
15	This is a concept that I think finds some basis
16	in U.S. law. It's referenced in the U.S. Court of Appeals
17	decision in the Microsoft case in a helpful way, from
18	Microsoft's perspective, on the integration issues. It
19	doesn't really resonate overseas, I have to say. And
20	there's been cases where I've been sitting across the
21	table trying to make the point that every firm in the
22	industry is engaging in some particular practice,
23	therefore they must think it's valuable aside from the
24	ability to exclude because they are excluding anybody
25	because they have low share.

38

1	And the reaction on the other side is often
2	really just a blank stare. And so what are you saying,
3	it's obvious that the firms -- that the rules are
4	different for high share firms, so we really don't
5	understand the point you're making.
6	Convergence, it's been much discussed. I think
7	it would be helpful to see a redoubled effort by U.S.
8	agencies to evangelize the U.S. approach.
9	And for everything I've said about
10	predictability, U.S. law is more predictable than European
11	law and the law of other countries with their emerging
12	antitrust regimes. A great deal has been said about this
13	through the years. Given globalization, I think it is
14	increasingly important to find some way to allocate
15	responsibility among multiple agencies. And certainly a
16	kind of common sense approach would seem to me a greater
17	deference to the rules of the defendant's home country. And
18	I would say from Microsoft's perspective, we really haven't
19	seen much of that in the cases that we've been involved
20	in.
21	So, again, thank you very much for the
22	opportunity to present here today.
23	(Applause.)
24	MR. COHEN: Thank you, David
25	Our next speaker will be Scott Peterson, who is

39

1	senior counsel at Hewlett-Packard Company. Mr. Peterson
2	has practiced as an intellectual property attorney for a
3	number of years, focusing on information technologies. He
4	joined HP in 1991 and provided intellectual property
5	support for a wide range of HP's businesses, as well as in
6	the context of standards development.
7	Along with his law degree from Franklin Pierce
8	Law Center, Mr. Peterson holds bachelor's and master's
9	degrees in electrical engineering from MIT.
10	So, we'll hand it over to Scott
11	MR. PETERSON: Thank you very much. Thank you
12	and I appreciate the opportunity to be here.
13	I am going to be talking on the topic of the
14	intersection between intellectual property and standards
15	and the competition implications.
16	And I want to say I really appreciate the
17	attention that the agencies have been paying to this topic
18	over the years. And, in fact, the guidance that the
19	agencies have been giving in recent years I think has been
20	very helpful and has played a role in some of the changes
21	that we are actually beginning to see. So, I really thank
22	you for your attention to this area.
23	I really have one core message throughout this
24	presentation. You are actually going to see it on every
25	slide. It was the title: Transparency of patent

40

1	licensing information during development of standards
2	facilitates efficiency in markets for technologies and
3	standards. That's the message. I am going to talk about
4	it. I'm going to elaborate on it a little bit. But
5	that's the core.
6	And a kind of corollary to that or related is to
7	recommend that guidance on application -- further guidance
8	beyond what we have -- on application of Section 1 to
9	collective action during standard setting regarding
10	licensing terms for patents essential to standard,
11	facilitates behavior that reduces the likelihood of
12	conduct in violation of Section 2
13	So, this is a hearing where the focus is on
14	Section 2. My message is actually for guidance on
15	Section 1 because the behavior that can be beneficial in
16	reducing the Section 2 risks is behavior that's
17	potentially chilled by concern about Section 1.
18	So, in fact we see significant value in what we
19	think of as sort of a voluntary industry-led approach to
20	reducing the risk of anticompetitive use of patents
21	essential to standards. We recommend proactive action
22	that would operate to reduce the need for after-the-fact
23	corrective agency enforcement actions of a Section 2 type.
24	But this desirable procompetitive behavior that
25	could operate to reduce this potential for the

41

1	anticompetitive use is being chilled to some extent by
2	concern that that collective action poses some Section 1
3	liability to the participants in the standard activity.
4	So, let me say a little about some background,
5	myself and Hewlett-Packard.
6	My particular background is that of an
7	intellectual property attorney. I have given advice to a
8	range of HP businesses. But over the last decade in
9	particular, I have given advice on the topic of patents
10	and standards. And in the last half of that decade or so,
11	I've -- I guess initially that advice was in the context
12	of particular transactions, particular standards,
13	development activities from people with business
14	activities -- and then in the latter half of that decade of
15	activity that I have been involved with this, has been in
16	trying to coordinate at HP our policy level considerations
17	of these questions that arise about intellectual property and
18	standard setting.
19	HP is -- to turn to the company that I'm talking
20	about -- fundamentally in the information technologies business,
21	a business which depends enormously on standards, a business which
22	has enormous network effects. So, standards are something that HP
23	is extremely familiar with. We participate in hundreds of
24	standards development activities. We have products that implement
25	dozens and dozens of standards. This is not an area where a

42

1	product implements a standard. This is an area where
2	products implement many, many standards. So, we have
3	developed a great deal of experience with the challenges
4	of standards development.
5	HP is also active as an innovator. HP has
6	invested -- let's see -- in the last fiscal year, we
7	reported 3.6 billion dollars investment in R&D. HP has
8	long invested in R&D. That investment has been reflected
9	in an extensive patent portfolio. Again, at the end of
10	the last fiscal year, that was reported as about 30,000
11	patents.
12	So, innovation and the patents that reflect that
13	innovation are also very important to HP. So, to give you
14	a sense of the perspective of where I'm coming from, it's
15	one where an effective standards environment is extremely
16	important because it's critical to the nature of the
17	products. It enlarges markets for products that HP makes.
18	And yet on the other side, patents are also
19	something that are an important part of HP's business.
20	So, with that background on HP, let me go back
21	then through the message, which you have seen here again:
22	transparency of patent licensing information during
23	development of standards facilitates efficiency in markets
24	for technologies and standards.
25	Let me start off by saying that there is

43

1	potential for anticompetitive use of the patents. This
2	was discussed in particular at the December 6th hearing.
3	And my goal is not going to be to replow this ground that they
4	talked about, but rather -- the fact that a patent that is
5	essential to standards can be employed in anticompetitive
6	ways is particularly important to recognize. And this
7	flows from the fact that once the patent is -- once a
8	standard is set and a patent is essential to it -- if the
9	standard becomes successful in the sense that there is a
10	lock-in effect such that participation in that marketplace
11	requires that you implement the standard -- then implementing
12	-- and implementing the standard requires a license, then that
13	patent now takes on a leverage that goes potentially beyond
14	the innovation that underlies it.
15	And it's that combination of factors -- there
16	is the leverage that one obtains from the innovation itself,
17	and yet there's also leverage that could come from the
18	lock-in effect of the standard. It's that combination that
19	leads to the challenge of potential anticompetitive uses of
20	patents that are essential to standards.
21	In my 2002 testimony -- I testified in April and
22	in November of that year on essentially this same topic --
23	I expressed some concern that there was a trend that
24	patents essential to standards were going to become an
25	increasing problem in the success of standards, and the

44

1	potential for abuse was a growing one.
2	And I have to say that our observations in the
3	intervening years have confirmed our concern about that
4	trend. And let me offer one example of something that
5	illustrates the trend.
6	There is, I think, a fairly increased mobility
7	of patents over what we would have seen ten or twenty
8	years ago. For example, the concept of patent auctions is
9	far more conventional now than it was a decade ago.
10	And I am not suggesting there's anything
11	inappropriate about this mobility of patents. I think the
12	ability to transfer intellectual property rights can be
13	extremely valuable. So, I'm not criticizing the trend as
14	such, but I simply want to point out that there is a
15	substantial change in the dynamic for how a patent gets
16	employed and what the licensing and enforcement
17	implications might be when the patent moves from the place
18	where it started to some other place, in particular for a
19	patent that is essential for the standard. It may well
20	have begun in a company that was working on technologies,
21	and had products, in the area of that particular standard
22	and would have certain motivations and expected a business
23	behavior. When that patent moves elsewhere, the
24	expectations and dynamics are going to be different.
25	So, this sort of increase in the mobility of

45

1	patents is an example of why I think we have to be more
2	careful about paying attention to patents during the
3	development of standards, because the opportunity for
4	aggressive behavior that may employ or exploit the
5	leverage from the standard -- not just the leverage from
6	the patent, but the leverage from the standard -- has been
7	increasing over the last decade or so.
8	So, there is a market which I think is sometimes
9	overlooked in talking about licensing of patents in
10	connection with standards. It is important to recognize
11	that there's a market for technologies in standards, and
12	there should be competition in this market for
13	technologies in standards. And there are -- in the
14	process of making choices as to what will go into the
15	standards -- in some cases there are a variety of relatively
16	equivalent choices in terms of the capabilities that they
17	offer, and yet in other technologies, in other settings,
18	sometimes one stands out dramatically above the others
19	because the nature of the technology is such that, you
20	know, there is opportunity for the standard to make a
21	substantially better choice in that particular area.
22	The license fees in those cases ought to reflect
23	that underlying reality. If in development of a standard one
24	is selecting one of many alternatives that are essentially
25	comparable in their end result, comparable in the

46

1	performance, characteristics and so forth, one would
2	expect the license fees to be substantially smaller than
3	when one is in a situation where the selected technology
4	is in fact head and shoulders above the alternatives, in
5	which case the license fees ought to reflect that
6	contribution to the standard.
7	Once the standard has been selected, however,
8	that distinction is easily lost because, again, if there's
9	a lock-in effect from the standard, it won't matter that
10	there were alternatives at that earlier stage. The
11	competition -- the effect of that competition is active at
12	the time that the standard is selected. It is either
13	effective then or the value of the competition is lost
14	because the lock-in effect later would mean that.
15	Suppose you had ten different alternatives that were
16	fundamentally equivalent. Once that one is anointed as
17	the way that you're going to agree among competitors to
18	build products in that domain, having a license to that
19	patent, if there was a patent, is vastly more valuable
20	than it would have been in another case.
21	In any case, I think it's important to realize
22	that this process of selecting, there is essentially, a market,
23	but it's a market that has this odd characteristic. There
24	is the collection of people, oftentimes competitors, who
25	are selecting what the standard will be. And there will

47

1	be a single decision -- in a sense, a single buy decision. And
2	the technology that is put in the standard at that point now
3	has been selected, in some sense, as if it was purchased. So,
4	now if you think about the subsequent licensing transactions,
5	these are not really a family of separate independent
6	transactions. For those who wish to implement the standard
7	and need to have a license to the patent that's essential,
8	their licensing transactions are not independent. They're
9	already -- they've already fixed the buy decision. There's
10	no walk-away for them. In that sense, these aren't
11	independent transactions. These are all flowing from
12	the single decision which was made as a part of the
13	standard's selection.
14	So, I guess my point here is that efficiency in
15	the market for technologies in standards -- the result of that
16	selection -- is very important because the technology selections
17	have implications for all of the subsequent licensing
18	transactions. Those later transactions may appear in some
19	sense as separate, but they're not because the buy
20	decision was made once. It was made in the selection of the
21	standard.
22	Efficiency, market efficiency. So, I make my
23	point, you know -- inadequacy of information is preventing
24	some efficiency. Well, let me talk about the inefficiency
25	which is worthy of some -- being made more efficient.

48

1	The inefficiency in the market for the
2	technology that goes into the standards is essentially the
3	information problem associated with the licensing terms
4	for patents that would be required by the various
5	alternative choices.
6	So, I talked about a market for technologies and
7	standards. A choice is going to be made among potentially
8	alternative technical choices. One of the factors which one
9	would normally consider when making an economic choice is
10	price or other terms that might be associated with the
11	decision. And, oddly enough, instandard setting, that
12	information is not circulated, is not readily available to
13	those who are making this decision. So, you have a group of
14	participants in a standard setting activity who are talking
15	about a wide range of characteristics of the technologies
16	and choices that they are choosing among, and yet this
17	topic of what the licensing implications would be is oddly
18	excluded from that conversation. And, in fact, the mechanics
19	by which anyone comes to know that is, by and large vastly more
20	obscure. And the flow of that information is inhibited by
21	the concern that, because it involves a dollar amount
22	there must be price fixing concern of some sort. And
23	therefore this is the Section 1 concern that I referred to
24	that is inhibiting the sharing of this information, which
25	is in fact important in making a rational and fully

49

1	informed decision in this market for technologies.
2	Let me talk about -- so, markets for
3	technologies in standards. I think it's important to
4	realize I have been focused on patents in the sense of
5	essential patents -- those patents which you must have a
6	license to because of how the standard was conceived.
7	The competition in products that employ
8	standards and the innovation in those products
9	predominately takes place outside of what's specified in
10	the standard. So, in general, as I say on the slide here,
11	standard setting should seek to enable technology and not
12	to specify or require it.
13	Now, many times the nature of the problem being
14	addressed, there may be somewhat limited constraints or
15	constraints that make a range of behaviors possibly not as
16	great as one would like. But I think that in many cases
17	inadequate imagination has been applied to the problem of,
18	"Let's make sure that we specify as little as possible
19	because we want to foster competition and we want to
20	foster ongoing competition." And yet choosing a standard
21	essentially freezes a particular technological point.
22	There ceases to be competition to the extent that there's
23	-- that there's lock-in on the standard. And from the
24	time that that standard is important, there ceases to be
25	competition on that particular set of things which is

50

1	specified in the standard.
2	There are technological decisions that can be
3	made as to how you define the specification, what is
4	needed to achieve the network effects that the standard is
5	trying to accomplish.
6	I think that the environment that we presently
7	have, which excludes to a large extent from consideration
8	the licensing concerns, results in, to some extent, a
9	motivation to incorporate as much technology and
10	innovation into the standard as possible. And, in fact,
11	that's the wrong motivation. We want to motivate people
12	to keep technology out of the standard. You want to keep
13	the technology from being specified. You want the
14	standard to enable the non-required technology which
15	continues to be the subject of further evolution and
16	competition among even the preexisting alternatives.
17	So, I think that the present environment,
18	where the licensing considerations are not considered, has
19	an interesting adverse effect in this regard.
20	And then finally -- transparency of patent
21	licensing decisions during development of standards. This
22	procompetitive behavior of considering that information
23	while the standard is being selected -- as I pointed out,
24	people are concerned and have a longstanding concern that
25	there's some kind of a price fixing type environment that

51

1	will be created if in fact the license terms are
2	considered.
3	I think that in fact, in this environment, that's
4	a misunderstanding of the situation. In fact, there will
5	be a single group buy decision in the sense of the group
6	will select a final specification. The problem is that it
7	won't be informed by this information.
8	So, the idea of looking at this as leaving the
9	door open for a multitude of independent later licensing
10	decisions, I think it's failing to understand that the
11	reality is that there is one decision that's going to be
12	made. It is deciding whether a particular thing is
13	essential or not essential. The question is whether
14	that's going to be informed by license terms.
15	So, I go back to the beginning slide, and let me
16	make some comments in sort of the recommendation category.
17	It can be difficult to separate, after a
18	standard has been selected and after a patent is
19	essential -- it can be difficult to separate the legitimate
20	aggressive enforcement of patent rights from the use of a
21	patent that is being leveraged to essentially leverage the
22	value that was created by the collective work of the
23	competitors.
24	So, those are very difficult to keep apart after
25	the fact. There is no market, really that you can rely on

52

1	in the ex post world. So, I think it's very important to
2	foster a proper attention to this issue while the
3	standard is being selected.
4	A couple of -- let' see -- one problem -- two
5	particular problems that I want to point out that merit
6	some attention going forward.
7	One is the -- I mentioned mobility of patents
8	is increasing patents are increasingly mobile. So, one
9	challenge is that licensing commitment typically you cannot
10	-- under the regime of many standards development activities,
11	you cannot rely on those licensing commitments passing
12	through as the patents move from one owner to another. This
13	is a problem meriting attention. And organizations may strive
14	to do something about that in the context of standard setting.
15	They may ask people to make commitments or something. It's
16	a problem of increasing concern because of the likelihood
17	that patents are moving.
18	And another problem is that of the injunctions
19	in the face of licensing commitments. So, again, this is
20	another sign the commitments are of a fairly tenuous
21	nature. So, there may be licensing commitments. On the
22	other hand, the ability to turn off someone's ability to
23	practice a particular standard can be an incredibly large
24	negotiating lever. And the fact that that lever could be
25	available even in the case of a licensing commitment is a

53

1	very troubling one.
2	I guess I'll close there. And I guess I'll once
3	again thank the agencies for continuing to pay attention
4	to this topic. I appreciate the guidance that's been
5	offered so far, but I think there's lot more. As the
6	world changes and begins to pay more attention to patents
7	during the development of standards, we're going to learn
8	more about what the issues are and perhaps more guidance
9	will be needed
10	Thank you very much.
11	(Applause.)
12	MR. COHEN: Our next speaker is going to give us
13	some insights from the perspective of a standard setting
14	organization. He is Robert Skitol, who is senior partner
15	in the Antitrust Practice Group at Drinker Biddle & Reath
16	in Washington. And he is counsel to the VMEbus
17	International Trade Association, know as VITA.
18	Mr. Skitol is a graduate of Hobart College and
19	NYU Law School. He has over 35 years experience in all
20	facets of antitrust and trade regulation, and written and
21	lectured extensively in the antitrust and trade regulation
22	field
23	At this point, we'll give the podium to Bob.
24	Do you have slides, Bob, or not?
25	MR. SKITOL: I do have slides.

54

1	MR. COHEN: We just have to find them.
2	MR. SKITOL: I can proceed without the slides.
3	There is a slide set, but I'm happy to speak without it.
4	Well, thank you for your indulgence. I am
5	delighted to be here on behalf of the VITA standards
6	organization. I'll be offering VITA's perspectives on
7	some of the same points and issues and concerns that Scott
8	spoke about.
9	My comments are complimentary to Scott's in many
10	respects. Scott spoke about patents and standards from
11	the standpoint of a major technology innovation intensive
12	company that participates in standard setting proceedings.
13	My client VITA is a major standards development
14	organization that is the flip side of the concerns. But
15	for VITA certainly, Scott's transparency theme resonates
16	quite a bit. And so I want to use my time today to offer
17	VITA's perspectives on how the antitrust agencies should
18	assist SDOs in protecting their processes from
19	exclusionary patent hold up conduct.
20	Of course VITA appreciates and has been a major
21	beneficiary of steps in this direction that the agencies
22	have already undertaken. My remarks concern desirable
23	next steps along this path.
24	I think the logical place to begin is with the
25	definition of exclusionary patent hold up conduct. And I

55

1	want to propose one broad enough to encompass an array of
2	patent related practices that subvert or can subvert open
3	standards and produce anticompetitive market outcomes.
4	So, my proposed definition for the agency's
5	consideration is as follows. A patent owner's inducement
6	of an SDO's adoption of a standard that implicates the
7	owner's patent claims without other participants'
8	awareness of that fact or without their awareness of the
9	cost and other impacts of it, thereby enabling the owner
10	to acquire and exercise monopoly power that it would not
11	otherwise have obtained.
12	Now, this is not news to the antitrust agencies,
13	this general concept. The FTC has been active in
14	challenging hold up conduct of this kind for about twelve
15	years. The Dell, Unocal and Rambus cases collectively
16	delineate a framework for treating hold up conduct as a
17	Section 2 violation in circumstances involving deliberate
18	deception regarding the existence of patent claims
19	implicated by a draft standard under development.
20	These cases also support the idea that the
21	requisite deception need not be overt. Mere silence about
22	essential patent claims can be unlawful when that behavior
23	actually misleads other participants in light of
24	expectations generated by the organization's rules or
25	established practices.

56

1	But hiding the existence of essential patent
2	claims is not the only way that exclusionary outcomes can
3	occur. There are other ways that patents can be used to
4	morph or subvert an open standards process into the
5	practical equivalent of market monopolization.
6	And I want to suggest three examples for your
7	consideration, all involving situations where the
8	existence of essential patent claims may well be
9	disclosed, may well be known, but patent hold up conduct
10	of an anticompetitive nature can nonetheless occur.
11	And the first example is one that entails
12	inducing reliance on a generalized commitment to license
13	essential claims on reasonable and nondiscriminatory
14	terms, the so-called RAND assurance that is in widespread
15	use, without the patent owner's acceptance of any
16	meaningful constraint on what it demands as actual license
17	terms after the standard has been adopted and a whole
18	industry is locked into sunk investments in compliant
19	products.
20	This is the essence of the allegations in
21	Broadcom versus QUALCOMM. We don't know the facts. We
22	know the allegations. And the allegations tell a story of
23	how generalized undefined RAND commitments can end up
24	bringing about monopolization.
25	The second example entails inducing reliance on

57

1	that kind of RAND assurance followed by seeking
2	injunctive relief to enforce the applicable claims. This is
3	a situation Scott also commented upon.
4	From my standpoint, from VITA's standpoint, the
5	injunction threat is fundamentally contrary to the whole
6	idea of the RAND assurance and the intended reliance upon
7	it. The only legitimate issue in any ensuing litigation,
8	once that assurance has been given and relied upon, should
9	be what those promised reasonable terms are, the patent
10	owner having effectively given up the right to exclude
11	under the patent code in return for what will often be
12	mega benefits from incorporation of that owner's
13	technology into the standard being developed.
14	The third example entails the transfer of ownership
15	of an implicated patent without binding the new owner of
16	it to the original owner's license commitment, the patent
17	owner having induced the whole industry into employing the
18	patented technology in the belief that acceptable license
19	terms were assured. The owner then transfers the patent
20	in a manner allowing the new owner to repudiate the
21	assurance and exploit the resulting new monopoly power.
22	Scott talked about the recent and increasing
23	trend of patent mobility, which seems to me to underline
24	the danger that this particular kind of hold up conduct is
25	something we need to worry more about in the time ahead.

58

1	So, all of these kinds of exploitive conduct and
2	the resulting hold up outcomes from them are today's
3	version of monopolization through highjacking an industry
4	standards development project, much as did the conduct at
5	issue in the Supreme Court's Allied Tube and Hydrolevel
6	decisions of two decades ago. Those cases involved different
7	kinds of conduct, but with essentially the same kind of effect
8	as patent hold up conduct can have today. This is really all
9	about proprietary capture of what is intended to be an
10	open standards process with market-wide effects of the same
11	nature as those condemned in those past cases of the Supreme
12	Court.
13	Now, there is disagreement in the standards
14	development community about the extent or prevalence of
15	these kinds of hold up situations, as I will explain in a
16	few minutes. My client, VITA, has some relevant
17	experience in this regard and knows from its own
18	experience that this is far from an isolated event.
19	But two developments, at least two developments,
20	strongly suggest increasing exposure to it. One is the
21	vast proliferation of patent grants that we are witnessing
22	within standards intensive technology spaces.
23	And the other development is what we're
24	seeing as the emergence of new business models of some
25	technology companies that depend on maximization of

59

1	licensing revenues from the use of their patents in
2	standards specifications.
3	In this environment with these developments,
4	SDOs' inattention to the problems that do surface invites
5	proliferation of these hold up situations in the years
6	ahead.
7	Now let me tell you more specifically -- let
8	me catch up on the slides. Let me tell you more
9	specifically about VITA and VITA's role in this story.
10	VITA develops standards for modular embedded
11	computer systems in a wide range of products. Members and
12	participants in its working groups include a broad cross
13	section of builders and users of these systems for such
14	applications as medical imaging, aviation and navigation
15	devices for military defense and space exploration.
16	VITA's management, particularly its
17	distinguished executive director Ray Alderman, have come
18	to acquire some rather deep expertise and experience in
19	patent hold up. In its own proceedings, VITA has
20	encountered no less than four major patent hold up
21	episodes within the past six years, each one causing major
22	delay in the implementation of foundation standards
23	critical to members' technology advancement needs, and
24	imposing on the organization major expenses to address and
25	counter the asserted claims.

60

1	These episodes are described in some detail in
2	VITA's application for a DOJ business review letter that
3	I'll talk about shortly.
4	VITA recognized one year ago that it was exposed
5	to more such episodes and encounters of this sort in the
6	immediate years ahead, in light of a considerable patent
7	thicket surrounding a planned technology transition that
8	would need to drive the upcoming standards development
9	activity.
10	It also recognized, and its members recognized,
11	that VITA's longstanding patent policy actually enabled
12	and facilitated rather than protecting against hold up
13	conduct of this sort given reliance on wholly undefined
14	RAND assurances with no information on actual license
15	terms until after a standard was adopted or at a very
16	advanced stage of the VITA development process.
17	So, VITA devised a new patent policy designed to
18	ensure greater transparency earlier in the proceeding in
19	all of these respects. There are several elements of the new
20	policy revolving around disclosure obligations of working
21	group members at each of four stages of the working
22	group process, including the very beginning and midpoints
23	of it.
24	Required disclosures of all potentially
25	essential patent claims, including those set forth in

61

1	pending applications, based on good faith and reasonable
2	inquiry into the members' patent positions; required
3	disclosures of a maximum royalty rate and incentives for
4	disclosure of other license terms; clear acknowledgment
5	that the proffered disclosures will be legally enforceable
6	by prospective licensees against not only the disclosing
7	member company but also successors and assigns and
8	transferees of the underlying patents; and, finally, an
9	arbitration procedure for compliance disputes.
10	In June of last year, VITA applied to the
11	Department of Justice for advice on the new policy under
12	the business review procedure. On October 30, 2006, the
13	DOJ issued a favorable letter, and it provides a
14	considerable amount of analysis and insight on DOJ's
15	perspectives about the patent hold up problem in general
16	and about how disclosure requirements of the sort
17	described in VITA's new policy can be an effective
18	safeguard against that kind of conduct and outcome.
19	The letter concluded that the new VITA policy
20	would be an efficiency enhancing contribution to VITA's
21	standards development processes. DOJ characterized the
22	policy as an attempt to preserve competition and thereby
23	avoid unreasonable patent licensing terms that might
24	threaten the success of future standards; avoiding
25	disputes over licensing terms that can delay adoption and

62

1	implementation after standards are set; and, thus, a
2	sensible effort by VITA to address a problem created by
3	the standard setting process itself.
4	Needless to say, VITA very much welcomes and
5	appreciates the guidance that this letter provided and
6	believes it has a tremendous value to the standards
7	development community as a whole.
8	With the DOJ letter in hand, the VITA membership
9	on January 17, 2007 overwhelmingly approved and adopted
10	the new patent policy and it's now undergoing the
11	requisite review by the ANSI Executive Standards Council.
12	Now, at this point -- hold on one second. That
13	is where I am. I'd like to offer four reasons why the
14	agencies should now affirmatively encourage other SDOs
15	to follow VITA's lead by experimenting with new patent
16	policies of their own.
17	And the first reason is that the DOJ's VITA letter,
18	as well as several speeches by officials of both agencies
19	in the last two years, recognize that SDO policies of
20	this general kind are not just okay from an antitrust
21	standpoint but can be procompetitive in their protection
22	against hold up outcome. In short, these policies serve
23	the public interest in protecting and promoting a robust
24	competition throughout standards driven technology
25	markets.

63

1	Second, the FTC's Rambus decision suggests that
2	the viability of any Section 2 case against hold up
3	conduct in this context may depend on a showing that the
4	patent owner's actions were contrary to SDO participants'
5	reasonable expectations in light of SDO policies in place.
6	So, in short, in this respect, if an SDO fails
7	to implement effective protection against abuse of its
8	processes in this manner, then participants will be in an
9	awfully weak position, if any position at all, to complain
10	about the resulting injury to them. And the government
11	will be in a weak position or no position to mount an
12	attack upon the situation, even though the public is
13	adversely affected by an anticompetitive market outcome.
14	Third, effective SDO self-policing or
15	self-regulation through policies of this sort will reduce the
16	need for agency enforcement actions, as well as reducing
17	all participants' exposure to disruptive private suits
18	over license terms. And self-regulation is a far
19	more efficient solution to this problem than any reliance
20	on litigation. This should be obvious to all concerned,
21	to everyone that's ever participated in a standards
22	development process.
23	SDO and its members may spend several years
24	developing a new standard, bringing it to completion and
25	ultimate adoption but then seeing the whole effort fail

64

1	because hold up conduct blocks implementation.
2	Now, even if the government at that point steps
3	in with a Section 2 enforcement action that results in an
4	order, four or five or six years later the damage is done
5	and there is no real remedy for the resulting harm to the
6	public. So much, much better to prevent the conduct from
7	happening in the first place than ever needing to try to
8	undo it.
9	So, finally, the fourth -- reason number four,
10	is that there's no reason to think that VITA's new policy
11	is the perfect solution or one suitable for SDOs
12	generally. Lessons learned from other SDOs'
13	experimentation with variations upon it will resound to
14	the benefit of all SDOs and participants in them. There's
15	no one size fits all in this area. VITA itself may well
16	want to revise, and in all likelihood will want to refine
17	in some respects, its new policy a year or so from now
18	after experience with it in several working groups.
19	VITA will be at least as interested in following
20	innovations by other SDOs as they may be interested in
21	VITA's experience under its new policy. The enforcement
22	agencies, I would suggest, should want to encourage
23	information sharing and benchmarking efforts among SDOs
24	along these lines.
25	Now, allow me to conclude with some specific

65

1	suggestions for what the agencies can do in the months and
2	years ahead to promote desirable SDO initiatives in this
3	area.
4	First, the agencies should affirmatively encourage
5	more requests for DOJ letters or FTC advisory opinions on
6	patent policy proposals of various kinds to provide more
7	and deeper guidance for the SDO community in general. And
8	one specific example I'd like to suggest of where
9	additional guidance and more specific guidance would be
10	highly desirable is on the extent to which and manner in
11	which a policy might go beyond requiring a disclosure of
12	licensing terms, as the VITA policy does, and beyond that
13	allowing discussions or even collective negotiation of
14	those license terms during SDO meetings.
15	I personally believe that these further steps
16	going beyond mere disclosure and actually letting the
17	working group do something collectively with the
18	information would be desirable; it is logical; it makes
19	sense in the context of the core mission of an SDO's
20	working group, which is to make collective decisions about
21	choosing one solution over another; and it makes eminent
22	sense for costs or relevant costs between competing
23	solutions to be part of the equation.
24	I've actually done a whole article on this
25	subject, which appeared in the Antitrust Law Journal,

66

1	and I understand it's being placed in the record of
2	today's hearing. So, now I've plugged my own article.
3	But I am convinced that resistance to these
4	further steps, anything beyond pure disclosure, rests on
5	unfounded antitrust concerns. And there's at least the
6	beginning of indication, more than a beginning, that the
7	agencies are seeing the matter that way. The latest word
8	on this is footnote 27 in DOJ's VITA letter, indicating
9	the likelihood that DOJ would address the discussion or
10	collective negotiations scenario as a rule of reason
11	question because it could actually be procompetitive.
12	FTC Chairman Majoras expressed that same view in
13	her Stanford speech of September 2005. I hope that one or
14	both of the agencies will get an opportunity to provide
15	more definitive guidance on this front in the near future.
16	Second specific suggestion, I believe the
17	agencies should consider undertaking an industry-wide
18	study of SDOs' experience with various kinds of hold up
19	situations and how existing SDO policies either address or
20	fail to address any problems thereby encountered. A study
21	of this sort could certainly help to resolve the
22	disagreements to which I referred a little while ago over
23	whether the hold up threat is or is not prevalent and
24	growing. Such a study could also provide a valuable
25	information base for suggested solutions or new proposals

67

1	for SDO policy reforms.
2	Third, the agencies should help to shape case
3	law development in this general area by entering private
4	suits, by filing Amicus briefs in private cases
5	challenging SDO-related conduct and practices where
6	unfortunate and harmful decisions are sprouting up.
7	Examples of private cases of this sort where DOJ or FTC
8	Amicus input could have been valuable are Golden Bridge
9	Technology versus Nokia, last year's decision in Texas,
10	with its holding of per se illegality against conduct
11	appearing to be a common feature of standards development
12	activity; and also last year's Broadcom versus QUALCOMM,
13	with its ruling that breach of an SDO rule that results in
14	monopoly power that would not otherwise be obtained cannot
15	ever state an antitrust claim.
16	And, fourth and finally, I would respectfully
17	encourage both of the agencies to support enactment of
18	legislation enabling SDOs to implement desirable patent
19	policies without fear of private antitrust claims.
20	There's no doubt that that fear has inhibited SDOs from
21	considering policies to address patent hold up problems.
22	Again, prime examples of private suits having
23	exactly that kind of chilling effect and that get talked
24	about all the time at SDO meetings as why we better err on
25	the side of caution, stay away from any new kind of idea of

68

1	that sort, etc., etc., would be the Golden Bridge Technology
2	case that I already mentioned, and Sony versus Soundview from
3	six years ago.
4	VITA is only one of several parties with a lot
5	at stake in open standard setting processes and that are now
6	exploring the opportunities for legislation in this area.
7	I hope DOJ and FTC officials will be interested in
8	dialoguing about this possibility with us over the weeks
9	ahead.
10	Thank you very much.
11	(Applause.)
12	MR. COHEN: One of the cases you mentioned
13	toward the end of your talk was the Broadcom v. QUALCOMM
14	case. We have on this panel a representative from
15	QUALCOMM and our afternoon session will have a
16	representative from Broadcom.
17	Our fourth and final speaker is Michael Hartogs,
18	Senior Vice President and Division Counsel at QUALCOMM's
19	Technology Licensing. Mr. Hartogs has spent his career
20	handling intellectual property and competition matters for
21	companies that compete in dynamic industries.
22	He's been with QUALCOMM since December of 1999.
23	Like so many of our other panelists, he brings a diverse
24	background: an undergraduate degree in engineering
25	physics from the University of Arizona and a law degree

69

1	from The George Washington University and registration to
2	practice before the United States Patent and Trademark
3	Office.
4	We turn to Mike.
5	MR. HARTOGS: I also want to thank the
6	Department of Justice and Federal Trade Commission for
7	inviting us to participate in these proceedings today, as
8	well as the Berkeley Center for Law and Technology for
9	hosting these important discussions.
10	I am going to primarily focus my discussions on
11	the issues raised by Scott Peterson and Bob Skitol today
12	relating to standards setting organizations and the
13	diverse membership of those entities.
14	I would like to comment quickly on Dave Heiner's
15	presentation about the challenges facing in-house counsel
16	in addressing antitrust and competition issues in the face
17	of disparate regimes that exist in various jurisdictions.
18	I think he addressed all of those very well, so I won't be
19	focusing on those topics today.
20	First I want to give a little bit of background
21	about QUALCOMM and its business model. It has recently
22	come under fairly close scrutiny and examination and I
23	think it's important to understand that in the context of
24	where QUALCOMM came from to where it is today as a
25	technology innovator and enabler.

70

1	As is fairly well known, the company was founded
2	in the mid-80s by several retired professors who had vast
3	interest in wireless communications technology Doctors Irwin
4	Jacobs and Andrew Viterbi, as well as five others. They
5	founded the company in Doctor Jacob's living room.
6	After setting up a company, they realized that
7	there were ways to vastly improve cellular technology as
8	used by terrestrial consumers that could take advantage of
9	a lot of work that they had looked into previously, both for
10	military and satellite applications.
11	To say that their proposals were met with some
12	level of skepticism is a vast understatement. There were
13	actual nay sayers who said that the technology proposals
14	they had would never work and would cost too much. There
15	was a professor across the bay at Stanford who actually
16	said the proposals defied the laws of physics.
17	Notwithstanding the proclamation of violation of
18	laws, they actually were able to demonstrate a viable and
19	working cellular system based on the technology called
20	CDMA, code division multiple access technology. I promise
21	not to go into too many technical acronyms today and stay
22	on topic.
23	But the efforts then following by Doctors Jacobs
24	and Viterbi and the others at QUALCOMM to proselytize this
25	technology, to find adopters for the technology and the

70

71

1	willingness to take the risk of deploying a
2	generation-leaping innovation were hardly trivial.
3	To seed the industry with the technology the
4	company had to develop its own cellular handset business,
5	a business that was filled with tremendously large
6	multinational corporate players at the time.
7	Infrastructure equipment was even more
8	complicated. In order to provide CDMA cellular base
9	stations for trial systems, they then asked incumbent
10	cellular operators to take a risk on a small company in
11	San Diego, which was primarily known for surfing and blue
12	skies, to trust for the deployment of their next-generation
13	networks.
14	In the face of all these challenges, the company
15	actually did manage to find some cellular operators who
16	were facing serious capacity constraints in their analog
17	networks at the time and were able to convince them that
18	CDMA technology was actually far more advantageous to
19	competing digital technology than was emerging in Europe,
20	which was GSM technology. And I won't go into it, other
21	than to say QUALCOMM had and still has a very firm
22	conviction about the superiority of CDMA technology over
23	GMS technology.
24	As part of having the operators' willingness to
25	embrace QUALCOMM's technology, they actually placed a

72

1	requirement on the company that the company make other
2	vendors of equipment available. There was concern that
3	QUALCOMM would not be able to satisfy all of the needs for
4	these wireless operators or have anywhere near the skill
5	necessary to support the adoption and proliferation of
6	these technologies.
7	So, very early in QUALCOMM's history, QUALCOMM
8	entered into its first licensing agreements. Those were
9	with Motorola and AT&T, who at that time were two of the
10	largest companies operating in the cellular industry.
11	QUALCOMM was a very small company at that time and was in
12	a much weaker position with respect to negotiating
13	leverage and strength as compared to those larger
14	companies.
15	As I will discuss a little bit later, it was
16	actually those early licensing deals that set the
17	framework for QUALCOMM's future licensing activities and
18	its efforts in licensing that continue to this day.
19	Having succeeded in seeing widespread adoption
20	of QUALCOMM's technology, the company very quickly
21	determined that it was actually not the company best
22	suited to either be in the cellular infrastructure
23	business or the cellular handset business. Vast
24	manufacturing companies with tremendous expertise were far
25	more suited. And, frankly, QUALCOMM didn't prove to be

73

1	the most competent manufacturer of these kinds of
2	products.
3	So, in late '99 and early 2000, QUALCOMM
4	actually sold its businesses for infrastructure equipment
5	and handsets to companies far more able to run with those
6	businesses.
7	QUALCOMM did retain its business of developing
8	chipsets and software solutions for use in cellular
9	handsets and maintained its licensing program, which it
10	had started in the very early days through the deals with
11	Motorola and AT&T, and through all of the '90s continued
12	signing up licensees for manufacturing of wireless
13	handsets and infrastructure equipment.
14	As a licensor of technology, there are some
15	concerns I guess that need to be recognized. It was
16	stated today that there are efforts by some licensing
17	companies to maximize licensing revenue. And while there
18	may be some goal in achieving maximal revenue from the
19	licensing side, you have to recognize that in order to do
20	that, your downstream licensees, the producers of
21	handsets and infrastructure equipment that are paying you
22	royalties, need to maximize their sales volumes.
23	We're not actually interested in seeing any one
24	or two companies maximize their profit at the downstream
25	level. We're looking at a downstream industry we want to

74

1	see as fiercely competitive as possible to drive price
2	reductions and increase volumes. The total revenues
3	generated that way will be higher licensing revenues at
4	the upstream licensing level.
5	So, QUALCOMM's business model from the beginning
6	and on the licensing side has been focused on
7	proliferation of technology and enabling companies
8	downstream to compete aggressively. We are able to take
9	our licensing revenues that are generated, pump them into
10	an R&D system, with now thousands of engineers producing
11	chip and software solutions for use in handsets, and
12	continuing development and improvement of the very
13	wireless standards upon which our lifeblood depends.
14	We then make these products, our chips and
15	software solutions and our patentable inventions,
16	available to a very broad downstream industry, which then
17	we've seen aggressively competing on introduction of new
18	products, new features and rapid price reductions.
19	Last year we spent one-and-a-half billion
20	dollars on research and development and we also have
21	thousands of patents pending patent applications.
22	One of the interesting benchmarks we've seen
23	some companies use at the handset level, with a different
24	view of the universe than QUALCOMM, is their own vast R&D
25	expenditures and patenting activities. What they don't

75

1	disclose, after suggesting that they spend billions of
2	dollars on R&D and have many thousands of patents is that
3	they don't make those products available to their competitors
4	of handset technology, licensing only that small body of
5	patents that they declared may be essential, and even then
6	only in some instances.
7	I want to turn from the background of QUALCOMM's
8	business model to the topic we've been focusing on today.
9	The intersection of intellectual property and antitrust
10	policies has been looked at closely for many years. It's
11	often described as a conflict. But I think most recently
12	Tom Barnett at the George Mason conference in September
13	made a much clearer statement that strong intellectual
14	property protection is not separate from competition,
15	rather it is an integral part of antitrust policy and
16	intellectual property rights and should not be viewed as
17	protecting their owners from competition, but rather
18	should be viewed as encouragement to engage in
19	competition.
20	There's no debate on the incentives to innovate
21	provided by a strong patent system. And it's in the light
22	of the innovation incentives generated by the patent
23	system that I want to speak today. I believe there are
24	efforts to consolidate a number of attacks with respect to
25	standard setting, legislative challenges, and lobbying the

76

1	Supreme Court to undermine the vitality of patents in
2	the patent system today. And I think it should be
3	recognized that these are primarily not driven by
4	so-called desires for transparency of information, as has
5	been suggested, but actually is purely an effort to shift
6	bargaining power away from patent holders, to drive prices
7	down, and which I believe will have the result of actually
8	driving innovative companies and patent holders out
9	altogether, robbing ultimately consumers of choice and
10	opportunities for innovative technologies.
11	On the standards side, there are very few people
12	that I think would challenge the procompetitive effects
13	that standardization can bring. The interoperability
14	between many companies' products, welfare-enhancing
15	cooperation among many different kinds of firms, increases
16	in choice, reductions in costs, broadening the size of the
17	markets, all are procompetitive benefits of standards
18	setting.
19	But one thing that needs to be remembered and
20	recognized is that in general the standard setting
21	activity is a participation of competitors in a market
22	cooperating in a way that needs to be carefully watched.
23	The suggestion that you can then take the step of
24	technical development, which is the purpose of standards,
25	and then move one more step toward collective price

77

1	discussions doesn't seem like a very big leap. But it is
2	if you look at it from the context of the accommodations
3	that have already been made by the antitrust laws and
4	enforcement agencies to allow competitive companies to
5	work together in concert for their procompetitive
6	aspirations.
7	I will get to the reasons for concern, but there
8	is a risk of undermining the very benefits provided by
9	standardization through an anticompetitive result.
10	One of the reasons I gave a little more
11	background on QUALCOMM than I might otherwise have is I
12	think it's important to understand that the benefits of
13	standardization do require cooperative industry efforts,
14	but that all of the participants in standards setting
15	activities don't wear the same hats. In some simple types
16	of standards, you may only have participants who are
17	producers of products who strictly need to ensure that
18	their products all work together. That isn't the most
19	common standards activity in QUALCOMM's experience,
20	where we find that development standards involve very
21	complex technologies, very long-term iterations of
22	contributions of technical proposals, a process which
23	benefits greatly not just from the participation of the
24	end product manufacturers being in the process, but also
25	innovative companies, companies like QUALCOMM, who don't

78

1	participate in the handset space or infrastructure
2	equipment space. But we have a very significant
3	interest in seeing optimal wireless technologies developed
4	and employed for those industries.
5	Now, in the context of the development of
6	wireless technologies, we do produce chips and software to
7	be used in the downstream products such as handsets and
8	wireless modems, but the bulk of our earnings is actually
9	driven from our ability to license the technologies that
10	come out of the innovations both in the standards settings
11	and the innovative research and development.
12	So, in addition, you have the manufacturers that
13	are clearly interested in developing their products, but
14	you also want to have companies like QUALCOMM who are
15	primarily motivated by improving and enhancing
16	technologies.
17	QUALCOMM is not the typical type of company you
18	think of in this capacity. Frequently you will think of
19	start-ups, sole inventors, universities, other companies
20	for whom valuable contributions can be made in advancing
21	the technological frontiers.
22	Then there are companies that really are hybrids
23	or vertically integrated firms, companies who do sell
24	significant products downstream, which may incorporate
25	their own innovations and the innovation of others, but

79

1	who also are contributors of innovation in the development
2	of the underlying industry standards. These companies may
3	have multiple interests in seeing both technology advance,
4	but also assuring that their products benefit from
5	early development opportunities.
6	Then, finally, we also are seeing more
7	participation in the standards setting by a group of
8	companies that are ultimately consumers of products. In
9	our industry, that would be the wireless operators. They
10	have an interest in seeing wireless standards developed
11	that meet certain specifications and so they would
12	participate in driving technical solutions or technical
13	requests for innovators and early participants to try to
14	solve.
15	It's important to recognize that these diverse
16	firms who participate in the standards setting process
17	have asymmetrical interests. Innovators who seek to
18	promote and advance technology through the proliferation
19	of their technology most often receive their return on
20	investment in the form of licensing revenues. They don't
21	sell products necessarily downstream and aren't able to
22	extract any return on their investment from the end
23	customers of the standards-implementing products that are
24	involved.
25	Manufacturing companies, on the other hand, see

80

1	their returns on investment coming from their downstream
2	sales. In our industry, sales of handsets is a good
3	example.
4	Again, the vertically-integrated firms, the ones
5	that are both manufacturers downstream and those that
6	contribute technology in the standards setting process
7	have mixed incentives. Now, on the one hand, they may be
8	very interested in high licensing costs in order to keep
9	their competitors out of their market. On the other hand,
10	they find themselves exposed to licensing needs from other
11	innovative companies and would like to see low royalty
12	overhead in order to drive costs down, recognizing that
13	they can recover their investments through sales of
14	downstream products.
15	All of these business models have their
16	advantages and their disadvantages. It's a little obvious
17	to state, but I will, the choice made by each company
18	should be where its strength lies. QUALCOMM clearly
19	demonstrated itself as not having strength in the
20	manufacturing of handsets and infrastructure equipment,
21	but those businesses were necessary to start the
22	proliferation of its technology. Having succeeded at
23	that, QUALCOMM quickly divested itself of those businesses
24	in order to increase efficiency in focusing on innovative
25	developments and making available enabling technology

81

1	solutions.
2	I guess the caution or concern I request of the
3	enforcement agencies is to tread cautiously in making
4	decisions that favor one business model over another. The
5	risk of driving certain kinds of companies out of
6	standards setting bodies probably comes at a societal risk
7	that isn't measurable, in that if that company is not
8	participating, you don't know what contributions are lost
9	and what welfare-enhancing solutions may have been
10	foregone.
11	There may be some standards where there isn't a
12	particularly high level of innovation wanted or needed,
13	and in those instances, nothing is lost. And in other
14	areas, the need for non-manufacturing companies to
15	participate and provide, in some cases, phenomenal
16	innovative solutions is something to be encouraged and I
17	think guarded carefully.
18	One of the points I want to get back to which I
19	raised before is the efforts that are going on in a
20	variety of arenas today with the stated goal of
21	transparency or the stated goals of avoiding certain
22	types of hold up, which I am going to address further as
23	to whether there really is a serious problem of hold up.
24	Recognizing that there are efforts going on to
25	rewrite IPR policies in standards setting processes, I

82

1	think it is with a pretty simple goal: just to reduce
2	costs for technologies included in the standards.
3	In order to reach those objectives, a number of
4	proposals have been made in a variety of standards bodies
5	in the last couple of years. There was an effort recently
6	that went on at ETSI where a transparent effort to
7	redefine the IPR policy was proposed which would establish
8	royalty capping set by the standards body and established
9	rules to share royalties on some sort of pro rata
10	allocation basis. There was a lot of interesting debate
11	that went on regarding those proposals which were
12	ultimately not adopted.
13	Other approaches call for ex ante disclosure of
14	licensing terms. While I appreciate the simplicity of the
15	proposals and apparent requests for knowledge, it is
16	firmly our belief that either compulsory ex ante
17	disclosure of licensing terms, or voluntary disclosures
18	with so-called strong encouragement, as some are calling
19	it, more than run the risk of resulting in exercises that
20	end in collective action. I think it's inevitable.
21	If you look at the very basis of standards
22	activity, it is about collective action, but for the
23	purpose of establishing technical specifications, adding
24	cost and price information into the mix would inevitably
25	be a factor which leads to collective discussions about

83

1	those topics which are not the purpose of standards
2	setting.
3	Now, it may be that policies are explicit in
4	their statements that such things shouldn't happen in the
5	context of the standards working groups, but that leaves
6	the sort of negative inference that there may be somewhere
7	else where such discussions may occur.
8	Those arguing in favor of compulsory ex ante
9	licensing disclosures typically make three criticisms of
10	the present regime: lack of predictability or
11	transparency; risk of hold up; and then somewhat related
12	to that, the problem of royalty stacking.
13	In our experience, the alleged criticisms are
14	not convincing and certainly don't prove that it's
15	reasonably necessary to scuttle an existing system that
16	has actually worked very well in favor of a system that
17	brings with it inherent risk of collective price
18	discussions, which could ultimately lead to disincentives
19	to participate by those who seek to earn their returns
20	from licensing.
21	The environment that gets created, as I
22	indicated, in the standards setting, is one of cooperative
23	development. Introducing price information will likely
24	lead to efforts of price setting by strong buyers.
25	One important thing to understand with respect

84

1	to calls for compulsory ex ante licensing disclosure is
2	that in fact ex ante licensing negotiations go on today.
3	This notion that participants in a standard are unable to
4	obtain sufficient information regarding price information
5	of technology incorporated in standards are not correct.
6	Voluntary ex ante disclosure and negotiation of licensing
7	terms on a bilateral basis prior to setting standards
8	are entirely consistent with the current FRAND regimes.
9	They certainly don't prevent potential licensees from
10	asking potential licensors about their planned licensing
11	terms and conditions. This isn't a theoretical
12	possibility. It actually goes on today and it frequently
13	goes on.
14	As I indicated, QUALCOMM's own licensing program
15	long predates standardization of any new technologies that
16	we worked on in the wireless industry. We consistently
17	engaged in licensing discussions before the beginning of
18	the standardization process, during standardization, and
19	long after, and are well aware that many other companies
20	do too.
21	There's an argument that it's inefficient for a
22	prospective implementer of a technology to ask prospective
23	licensors what their licensing terms are. I don't fully
24	understand that. The number of prospective licensors is
25	typically dwarfed by the number of standards implementers,

85

1	and in all but the most complicated technologies there
2	aren't that many licenses that need to be negotiated.
3	The second criticism is with respect to
4	so-called patent hold up. There are a number of
5	allegations made about what constitutes patent hold up.
6	And I think there is recognition that some activities such
7	as intentional withholding of patent disclosures has been
8	decided. However, there are those that suggest patent
9	hold-up also includes the case where a prospective
10	licensor of an essential patent seeks a royalty rate that
11	is surprisingly high.
12	In reality, licensees frequently claim to find
13	licensing rates surprisingly high. It's part of the
14	negotiation process. You start somewhere, you end
15	somewhere, and that's the nature of the business. There
16	are many give-and-takes in the licensing negotiation. So,
17	to suggest that the rate information or lack of
18	information on licensing terms, which would have been
19	readily available if a prospective licensee asked, I fail
20	to see how that justifies a need for mandatory ex ante
21	disclosure rules.
22	Another argument to support notions of patent
23	hold up is that essential patents gives a licensor the
24	ability to impose unconstrained licensing terms on the
25	licensees. And this just isn't the case. You have to

86

1	recognize even as a licensor of essential patents, there
2	are a number of constraints that exist. There are
3	horizontal constraints, constraints about wanting to see
4	the market develop downstream, impacted by what other
5	competitors are doing in the licensing community.
6	Vertical constraints with respect to the licensor and
7	licensee.
8	As I said, QUALCOMM is a licensor of technology.
9	If its licensees succeed, then QUALCOMM succeeds. So,
10	imposing onerous or technology-chilling licensing terms is
11	not in our interest and it's not a reason to participate
12	in the standards setting process.
13	And then there are dynamic constraints. The
14	development of standards is not a single function in time
15	in most cases. The standards continue to evolve. Other
16	participants join standards setting groups. And the
17	pressures and, shall we say, discipline that come upon
18	companies participating in the standard setting process by
19	other companies who have a history of not playing by the
20	rules is a real threat.
21	The final point I wanted to touch on is -- and
22	it's closely related to hold up arguments and the way they
23	have been used recently -- is the issue of royalty
24	stacking. The argument is fairly simple.
25	If there are multiple patent holders with

87

1	multiple essential patents in a standard, then the
2	potential royalty burden that can be imposed on licensees
3	may add up to some cumulative amount that's unreasonable.
4	First, it's important to recognize that many of
5	the companies participating in the standards setting
6	process have diverse incentives that I talked about
7	before, and subject to the various constraints that I just
8	talked about as well.
9	In some empirical research that's going on,
10	despite the claims of royalty stacking, there have
11	actually been very few instances identified. And several
12	years ago in the biotech industry, a paper was written on
13	the tragedy of the anti-commons in biotech. But twenty
14	years later, a paper on the fallacy of the anti-commons
15	came out. Royalty stacking is just not something that has
16	manifested itself. There is a lot of public rhetoric and
17	misinformation that's being spread, particularly in our
18	industry, that cumulative royalty rates are going to
19	amount to hundreds of percentage points.
20	And yet even some of the companies that QUALCOMM
21	is fiercely at odds with have publicly stated that they
22	don't think that anybody is paying double digit rates.
23	And there are a lot of factors to explain that. There's a
24	lot of cross-licensing that goes on. A lot of companies
25	maintain patents for defense purposes. There are many

88

1	dynamics that work together that result in the limiting of
2	royalty stacking despite the sort of argument that if
3	there's lots of patents, there's lots of royalties.
4	So, the proposals for compulsory ex ante that
5	are being proposed are being proposed to fix a problem
6	that either doesn't exist or certainly doesn't exist in
7	the widespread extent to which it has been attributed.
8	And the fact is these proposals run severe risks of
9	driving anticompetitive results and provoking the
10	elimination of innovators willing to participate in
11	the process.
12	There were a few comments that Bob made on the
13	efforts of VITA to revise its IPR policy that I feel I
14	ought to respond to. Having the benefit of going last and
15	having heard them, I will take that opportunity.
16	As I said, there may be standards in which
17	fairly low technology proposals are made. Complete
18	solutions are brought in by each company and they're
19	weighed on their respective merits and a selection among
20	them is made.
21	And I don't profess to know much about what VITA
22	does or its technologies. I've read descriptions that
23	it's focused on plugs and connectors and bus signaling
24	protocols. And I don't know the level of significant
25	innovation that goes on in those areas, but it may in fact

89

1	be an organization in which little harm would be done in
2	the face of compulsory disclosure of cost information with
3	technical solutions.
4	But the notion that such a solution would fit
5	all standards is deeply concerning. One size doesn't fit
6	all. I think in the vast majority of cases such a
7	disclosure regime will actually lead to the things that
8	I've expressed concern about, which is that there will be
9	collective discussion of price by large groups of
10	purchasers who produce product for the downstream market,
11	leading to some form of concerted purchasing power, the
12	end result being the driving out of innovative companies
13	who seek a return on investment based on licensing.
14	And I do note that a significant founding member
15	of VITA, very soon after the passage of the approval of
16	the policy by the board, withdrew its membership from
17	VITA. That company is Motorola, who I think is one of the
18	more innovative companies in America today.
19	The advice -- not advice, but the request I
20	would make of the enforcement agencies when asked to look
21	at revisions to IPR policies, and Bob's suggestion would
22	actually encourage such guidance, is to pay particular
23	attention to the facts and circumstances that exist in
24	each situation.
25	Efforts should be taken to avoid taking as

90

1	gospel allegations of hold up and royalty stacking. The
2	evidence isn't there. And there's a lot of research
3	coming out now in the last year combatting -- addressing
4	these many years of literature that's stated sort of the
5	contrary.
6	I will submit a bibliography with some notes
7	that can be included on the FTC's website identifying some
8	of the recent efforts to challenge these premises with
9	robust analysis.
10	Thank you.
11	(Applause.)
12	MR. COHEN: Well, we're a bit behind on our
13	schedule. We had talked about doing a 15 minute break. I
14	suggest that we take about two or three minutes in our own
15	seats to give us an opportunity to stand up, then we're
16	going to go forward so we can try to get as much of a
17	moderated discussion as possible. So, in about three
18	minutes I'm going to start again.
19	(A brief recess was taken.)
20	MR. COHEN: I am one of the belief that with the
21	panel as the meat and any questions that we have as the
22	gravy, we're going to try and get as much of the meat as
23	we can.
24	And probably the way to do that is to divide
25	our remaining time into two segments. One will be more of

91

1	the general issues which David Heiner was so good to
2	raise. And following that, the other segment would deal
3	specifically with some of the standard setting issues that
4	have been discussed already.
5	And what I'd like to do is begin and see if any
6	of the other panelists have comments or responses to
7	anything raised by David in particular, because you get a
8	chance to respond to the standard setting issues in about
9	15 minutes.
10	Anything you want to say? No? Okay, then I
11	will pick some questions to get you going.
12	You all have been people who have received or
13	watched others receive over the years antitrust counsel of
14	the various kinds of single-firm conduct.
15	I'm wondering if anything strikes you as having
16	been an area where advice or the legal tests that you're
17	trying to articulate has been particularly easy to
18	understand or particularly difficult to understand, any
19	recurring problems that you're facing?
20	MR. SKITOL: I will take a shot.
21	In my experience over the last couple of years,
22	I think the single most difficult area of Section 2 law to
23	advise on has been the loyalty rebate and bundled pricing
24	area. And you had an excellent panel on that subject a
25	couple of months ago, with a number of competing

92

1	suggestions for what the standards should be.
2	It's a tangled mess. It's been a tangled mess
3	in particular ever since the LePage's decision. And the
4	world is divided between those who think Lepage's is about
5	the right approach and those who think it isn't.
6	It's extremely difficult to give clear advice to
7	business people on what kinds of loyalty discounts are and
8	are not okay, what is the legal standard.
9	And so I would certainly urge special attention
10	and priority to the agencies in giving advice to the
11	courts because this is an area that's gotten terribly
12	muddled, not because of anything the government has done
13	but because of conflicting decisions in private
14	litigation.
15	MR. HEINER: I would agree with Bob that that's
16	a pretty tough area and one that I think gets all the more
17	challenging when you overlay the European focus on top as
18	well, as articulated in the Draft Article 82 Discussion
19	Paper.
20	More broadly to your question, I think I'd say
21	that it's a clear divide between Section 1 and Section 2,
22	where the Section 1 counseling is pretty easy, frankly,
23	and Section 2 is pretty hard.
24	MR. HARTOGS: I will agree that the issues on
25	joint conduct out participation and cooperation, I

93

1	think is fairly clear. I particularly echo the sentiment
2	about needing some measure of global harmonization in knowing
3	what the rules are for multinational companies
4	participating with other multinational companies in the
5	face of enforcement agencies and regimes in which they are
6	not in agreement on an application of a particular
7	standard.
8	We find ourselves trying to determine what is
9	the most restrictive set of rules under which we should do
10	our analysis and guide our conduct.
11	MR. COHEN: Okay. That leads me to some
12	questions on the international situation.
13	We just had one view of trying to find the sort
14	of the least common denominator. Have you found that your
15	businesses -- in general, have you tried to decentralize
16	to adapt to local competition rules, or do you find that
17	most of you are being forced in one way or another to fly
18	with the most restrictive laws potentially applicable to
19	you in different jurisdictions?
20	MR. HARTOGS: I think, unfortunately, localizing
21	is an idea that wouldn't work for us. We develop product
22	in the U.S., Europe, India, Korea and Japan. We sell
23	products to companies everyone in the word. They sell
24	their products further downstream everywhere else in the
25	world.

94

1	Agreements with respect to various related
2	entities with affiliates that are not U.S. entities
3	probably render it still necessary to look for the most
4	restrictive set of rules in guiding our conduct.
5	MR. COHEN: And we heard from Microsoft that
6	some of these -- the way this works with licensing.
7	Did similar issues arise with regard to your
8	contract practices?
9	MR. HEINER: It's very much a global business.
10	So, the answer is kind of the same as what Mike was
11	saying. We have looked at whether in particular cases you
12	can try to localize the business practices to the local
13	jurisdiction. The issues that come up are mostly not
14	around local facts, however. It's not as if the issue is
15	relations with a retailer in any particular country. The
16	issue, rather, is of a global nature, what is the design
17	of Windows around the world, what is the licensing
18	paradigm of Windows around the world?
19	And so we do find ourselves kind of looking to
20	what's the most restrictive set of rules. And that's what
21	we have to adhere to.
22	We have given some thought to whether it would
23	be possible -- notwithstanding the costs that it would
24	entail -- would it be possible to have different
25	products, different licensing plans in one part of the

95

1	world versus another. And it may come to that some day.
2	But if it does come to that, it would certainly be with a
3	certain loss of efficiency, and for customers as well.
4	MR. COHEN: Bringing us back to the United
5	States, one of our concerns at this hearing is to find out
6	the degree or whether, and if so the degree to which,
7	uncertainties about antitrust analysis of single-firm
8	conduct have been chilling potentially procompetitive
9	conduct.
10	We heard some examples and a discussion of that
11	in David's talk this morning.
12	Have any of you others found similar experiences
13	where business practices that may have been beneficial to
14	consumers have been put on hold because of uncertainty
15	about antitrust exposure?
16	MR. HARTOGS: I guess I would just quickly say,
17	Bob's comment before that guidance on pricing is
18	particularly difficult where you lack clarity here, you
19	lack clarity in Europe. And again not having sort of
20	flexibility to always choose what may be the most price
21	friendly, consumer friendly result, is a risk.
22	MR. SKITOL: There are lots of situations
23	involving Kodak aftermarket kinds of issues. We've all
24	been living with the difficulties of Kodak aftermarket
25	Section 2 as well as Section 1 problems for fifteen

96

1	years now. There are lots of situations I find where a
2	client has in mind doing X, Y, Z with its consumables,
3	which would be of significant consumer value, would
4	enhance the product, and it looks great. But because
5	of Kodak and all of the law that's built up around it,
6	this is problematic, and Trinko doesn't do that much to
7	help. There is hesitation and sometimes desirable
8	developments are canned because of concern about what
9	aftermarket rivals might be able to stir up by way of
10	mischief about it.
11	I think the whole Kodak aftermarket area is one
12	that could benefit from agency guidance. Where are we on
13	legitimate versus illegitimate aftermarket practices
14	fifteen years after Kodak and three years after Trinko?
15	Because the courts in private cases still don't get it
16	right. We still have not gotten the rules.
17	MR. COHEN: And just per a request for more
18	agency guidance, guidance can take different forms. And
19	because of time constraints, I'm going to throw three of
20	them out at once and see how you react to them and see if
21	they're suggestions you might want in one of these areas.
22	Guidance can take the form of explanatory text
23	such as we often give through reports on hearings and some
24	business review letters. It can take the form of safe
25	harbors, which can be announced. And it can take the form

97

1	of presumption. And we heard one suggestion for
2	presumptions today about conduct that's used by firms with
3	particularly great market power in competitive situations.
4	Would any of these three forms be particularly
5	useful to you? Do any of you have ideas of things that
6	you would like us to provide in any of these areas?
7	MR. HEINER: I think all three can be very
8	helpful. With respect to the text, of course it depends
9	what the text is.
10	MR. COHEN: Right.
11	MR. HEINER: There's always the possibility of
12	obfuscation instead of the intended fact. As one of my
13	colleagues pointed out to me before I came down here
14	today, we could have very predictable antitrust law in a
15	way that wouldn't be at all favorable to our firm. That's
16	the risk as well, I suppose.
17	MR. COHEN: Beware of what you ask for because
18	you might not like it when you get it.
19	I guess I should ask questions directed to the
20	other side of things, too.
21	We looked at the chilling as procompetitive
22	conduct. But do any of you have issues which you haven't
23	already touched on in which conduct involving dominant
24	firms has hurt you and that you think the agency should be
25	looking at but hasn't been paying full attention to or

98

1	much -- close enough attention to that might be desirable?
2	Anybody have anything in that area? Already
3	touched on.
4	Okay, let's go to the standard setting area.
5	And I think probably the way to begin would be to give an
6	opportunity for Scott and Bob to offer responses to what
7	they've heard. We had a response to them, so I guess you
8	should have a rebuttal opportunity. And we'll probably
9	open it up to a third rebuttal as well.
10	MR. PETERSON: I am going to yield my time to
11	the agencies. I'd much rather hear your questions.
12	MR. SKITOL: Can I just make a couple of
13	comments? I listened closely to Michael's discussion
14	about the QUALCOMM business model and the importance of
15	there being respect for diversity of business models and
16	that there shouldn't be a thumb on the scale against one
17	business model in favor of another. I agree with all of
18	those points.
19	I think from the standpoint of an organization
20	like my client VITA, from the standpoint of anybody who
21	supports open standards processes, competing business
22	models are good. But that's on the assumption, on the
23	premise, that all of the competing business models should
24	play by the same free-market rules and the same transparency
25	rules. All business models are subject to the same

99

1	antitrust laws. No business model should be imposed on a
2	group of standard setting participants.
3	It's good if all of the cards are up rather than
4	down. It's good for standard setting participants to have
5	choices. It's good for standard setting participants
6	sitting around in a working group with multiple possible
7	solutions to the specification writing, one of which may
8	well come from a business model that emphasizes licensing
9	revenue, and another comes from a business model that
10	enables the solution to be offered royalty free. It's
11	good to have that choice as long as everyone knows what
12	the respective costs are as well as what the respective
13	differences in quality and performance will be. And then
14	performance-cost tradeoffs can be collectively made and
15	there can be informed decision-making. That's all to the
16	good.
17	So, those of us who believe that ex ante license
18	terms disclosures and similar transparency policies are
19	good are not anti-licensing business models. We're not
20	anti-patent. We are pro free market, pro choice.
21	MR. COHEN: Any rebuttal?
22	MR. HARTOGS: To the extent Bob agreed with me,
23	I don't have any comments.
24	On the -- just a cautionary comment. In his
25	talk he suggested that the next step actually ought to be

100

1	a sanctioning for group discussions. And I do believe
2	that the ultimate result of that would be a chilling of
3	willingness of participants in the standard setting
4	organizations who do rely on licensing.
5	I think it should be recognized that the bulk
6	of participants in standards setting activities are
7	prospective licensees and the impact the proposed changes
8	can have is on more than transparency, but directed toward
9	driving pricing down where there is no return on investment.
10	That is something that needs to be watched and watched
11	carefully.
12	MR. HEINER: One time on this.
13	I think all of the speakers on this topic
14	identified the threshold question of how great a problem
15	is it this so-called hold up problem.
16	And from Microsoft's perspective, and we're a
17	company that's involved in dozens, I am sure hundreds, of
18	standard setting endeavors, and from our perspective, we
19	do not have a business model of really trying to make any
20	significant revenue licensing of IP into standards.
21	In our experience in participating in standard
22	setting bodies, we really have not experienced these sort
23	of hold up situations in standards that we wish to
24	implement in Windows and Office and other products. And
25	these products do implement huge number of standards.

101

1	So I offer that comment on the extent of the
2	problem we had about weighing against the collusive kind
3	of risk that [unintelligible].
4	MR. COHEN: You called that a threshold
5	question, but it was my first.
6	Let me direct toward the end of the table,
7	anything you might want to say as to the frequency of hold
8	up? I know you have identified four instances within
9	VITA. But how about the consideration that reputational
10	considerations and a desire to see downstream success of
11	the product is going to put a real limit on the likelihood
12	of hold up activity?
13	MR. PETERSON: So, yes, I think my discussion
14	earlier about patent mobility goes directly to that point.
15	And that decades ago where there was more stability in a
16	particular industry and much less patent movement, those
17	kind of reputational effects could have been more valuable
18	than they are likely to be in the future because the fact
19	is that patents have become separated from the reputation
20	that once was associated with them and thus that constraint
21	is no longer as strong.
22	MR. SKITOL: I would just add a comment that the
23	interest in growing the market and in the market being
24	successful is a factor in any monopoly, any monopolization
25	case. Every monopoly has its limits. A monopoly price

102

1	which is not limitless. It's got a limit.
2	So, in this respect, Section 2 monopolization
3	through patent hold up is no different than Section 2
4	monopolization through any other kind of predatory
5	conduct.
6	MR. COHEN: Let's lead into some of the
7	predicates for the ex ante disclosure rules. I guess
8	there's some other alternatives to that which I'd like to
9	get reactions to first.
10	I'm wondering whether a mere disclosure of
11	relevant patents, not disclosure of licensing terms,
12	followed by an opportunity for bilateral ex ante
13	negotiations would be sufficient? Why or why not?
14	MR. SKITOL: The point made about bilateral
15	negotiation is always out there and possible. That's
16	inviting secret behind closed doors bilateral special
17	deals between the big guys at the expense of new entrants
18	and smaller players.
19	Why isn't it preferable to do the negotiation
20	out in the open as part of the open standards development
21	deliberation process itself that is available to all
22	parties that want to participate? After all, this is all
23	in the context of the traditional RAND commitment which
24	has a nondiscriminatory as well as a reasonable component
25	to it.

103

1	So, the idea that we should stay away from more
2	transparency for everyone because we already have
3	bilateral opportunities, it doesn't make sense.
4	MR. HARTOGS: I guess in answer, what you
5	describe actually is the system that does exist today
6	about disclosure and bilateral negotiations. And it's
7	worked well. We had descriptions relabeling of things
8	today as hold up, which wouldn't have been viewed as hold
9	up previously.
10	I didn't hear any suggestion about
11	discrimination being part of the motivation of
12	licensors prior to the discussion. But to the
13	extent that companies are committed to licensing on a
14	nondiscriminatory basis, there are structural remedies and
15	opportunities to fix abuses there as well.
16	So, I don't see how ex ante disclosures of
17	licensing terms and collective negotiation or licensing
18	agreements fixes that. As indicated before, the large
19	number of potential licensees for any essential patent will
20	greatly exceed the single licensor.
21	MR. COHEN: I notice you talk about the
22	nondiscriminatory aspects of RAND. Let's focus on the
23	reasonable for just a moment.
24	What's the feeling of the panel as to whether
25	that has a well-defined meaning? And to what degree has

104

1	arbitration procedures of the type that VITA has talked
2	about been applied in the past? We have a history to go
3	on as to whether this is really successful in resolving
4	disputes in the area.
5	MR. SKITOL: Nobody knows what RAND means. I
6	defy anybody on this panel to tell us what reasonable
7	means and what the standard for it is. It's a meaningless
8	term that facilitates deception and facilitates hold up
9	for the very reason that it fools everyone involved into
10	thinking that it's a real limitation on what the patent
11	owner will do, when in fact it isn't.
12	MR. HARTOGS: I would strongly disagree. If you
13	look at the origins of IPR policies that call for RAND
14	declarations, the purpose is directed at eliminating outright
15	refusals to make licenses available for patents that
16	become essential for standards.
17	What RAND intended is an important flexibility
18	that recognizes that licensors and licensees are almost
19	always differently situated. And having the ability to
20	bilaterally determine mutually agreeable solutions that
21	satisfy both is probably the best test of reasonableness.
22	In some cases you might be able to look to
23	pre-standardization licensing activity. I am not suggesting
24	that there will always be circumstances where we can point
25	to ex ante licensing results as a benchmark to compare to

105

1	post standardization licensing to demonstrate reasonableness
2	or at least confirm that standardization didn't lead to a
3	change in licensing terms. Certainly they do exist. In some
4	cases and when they do exist, they seem a fair benchmark as to
5	establishing reasonableness.
6	MR. COHEN: Moving now to the idea of ex ante
7	disclosure of relevant terms, you need to tie this of
8	course to perhaps essential patents under the standard,
9	some concept along those lines.
10	I'm wondering if anybody has a sense of what the
11	impediments are to giving meaningful -- to even
12	identifying in advance what's likely to be in a standard
13	and what's likely to evolve out of the patent application
14	process in order to determine what you have and before you
15	can explain what the terms would be on it. Anybody want
16	to comment?
17	MR. PETERSON: Yes. So, it may be an evolving
18	thing over the course of a standard. It shouldn't be an
19	expectation that this is something that should be known up
20	front.
21	On the other hand, people are making judgments
22	about other aspects in the standard on an ongoing basis,
23	and this is information that ought to be brought forward in
24	that same spirit -- as it becomes apparent what will be
25	needed, information will be made available about it.

106

1	And, I'm sorry, there was another point I was going to
2	make.
3	Well, I'm sorry, go ahead.
4	MR. HARTOGS: So, I think it's an important
5	question because it goes back to my comment that the
6	proposal for VITA's policy may well work for VITA. But
7	that if you look at our experience in some very complex
8	wireless standards, there are multiple years of
9	development, multiple iterations of contribution of
10	technology and of innovation. And being forced to place a
11	stake in the ground from which you can't retract your
12	position or change it, it really is an important timing
13	question as to when you would do that. You have to make
14	an assumption on sort of the most optimistic view about
15	how successful you have been in providing your innovations
16	in developing the standard, and then make your proposals
17	based on that, on the assumption that if you have
18	something more valuable to contribute, you no longer
19	retain the right to price that effectively.
20	MR. PETERSON: The thought that I was missing a
21	moment ago is that one doesn't necessarily always have to
22	wait until a patent is matured into a patent, or even a
23	patent application, because it's often in many cases
24	possible to make judgements about what one's licensing
25	intentions would be, even not knowing what the particular

107

1	patents might ultimately be, because the judgments are in
2	many cases informed by other factors.
3	MR. COHEN: Assuming now that we've reached the
4	point that we're talking about some form of ex ante
5	activity that type of term requirement.
6	Perhaps the requirement of disclosing terms may
7	be at one end of the spectrum. You might then go a little
8	farther and have some provision for discussion or
9	clarification of the term, sort of at the middle of the
10	spectrum. And then go all the way to the far end and
11	actually have clear joint negotiation of the term.
12	Does anybody see -- or could you give your
13	thoughts on whether going beyond mere announcement of the
14	terms is necessary? What are the considerations?
15	MR. PETERSON: So, I think there will be
16	different -- this is an area where there should be
17	diversity and variety could be explored. So, I think
18	there may be certain kinds of product or technology areas
19	in which the exploration of the license term issue might
20	profitably go farther than in others. In others, it may
21	be that very little needs to be done. It may be simple
22	disclosure needs to be done.
23	So, I think there is a variety -- there are a
24	variety of different kinds of cases. Some are worthy of
25	more detailed attention than others.

108

1	One thing I would point out is that, if there is
2	a perception of this cliff that you step off of after
3	disclosure and that if you embark on anything beyond
4	disclosure that there's some kind of interactive
5	discussion is a very serious matter, then that chills even
6	the value of the disclosure.
7	So, I think, although I see the need for the
8	more collective action regarding the terms as being
9	perhaps very much the unusual case, to say that -- to make
10	it clear that's it's only disclosure which is
11	procompetitive and the discussion of the terms is a high
12	risk activity, it has that chilling effect. As I have seen
13	already in organizations that have been toying with
14	introducing more consideration of license terms, the
15	idea -- the steps that they feel they need to take in
16	order to assure themselves that nobody will ever talk
17	about them is seriously chilling just that first step
18	about getting information made available.
19	MR. SKITOL: I think the time has come to
20	recognize that a lot of the information technology and
21	communications technology standard setting processes that
22	we are talking about are really indistinguishable from an
23	antitrust analysis standpoint from all kinds of joint
24	product development, joint technology development
25	ventures. That's essentially what this kind of standards

109

1	development activity is. It is a group of companies
2	getting together, combining their resources and their IP
3	and collectively developing something new.
4	It is standard joint venture law today that when
5	you have a lawful joint venture, it is lawful for the
6	participants in that venture to make collective decisions
7	about which input to buy for this and which input to buy
8	for that. There are collective decisions and collective
9	negotiations over cost as well as other features of one
10	versus the other. That's what standard setting is about
11	today.
12	Now, there could be lots of situations where the
13	result of ex ante license terms disclosure is that the
14	parties sitting around the table in the working group
15	recognize that they've got two main good proposals. One
16	comes with a two percent royalty disclosure and the other
17	comes with a five percent royalty disclosure. And they
18	all agree that the latter is technically superior to the
19	former, but five percent is too much to pay.
20	What is wrong with a non-coercive negotiation
21	process, arms length process, in which the group
22	collectively discusses with patent owner B that we really
23	prefer your solution, we would go with your solution if
24	you could reduce that rate somewhat. And if that patent
25	owner decides to do so, to go ahead and accommodate that

110

1	interest, then what's wrong with that? That's an arm's
2	length decision and everybody ends up all the better for
3	it except for the solution A guy whose solution ends up
4	being excluded. But exclusion of one or the other is
5	inherent in the process.
6	MR. HARTOGS: I'd like to comment on two points.
7	One, I think the joint venture analogy breaks
8	down when you look at the sort of absence of certain kind
9	of participants you want involved in standard setting.
10	You wouldn't typically have the nonproduct companies such
11	as the universities. You may engage them to do contract work,
12	but the kind of joint venture activity you're suggesting is
13	very different from the standard setting, where in fact
14	your very customer may be a participant in the standard
15	setting process. In the joint venture context, you wouldn't
16	condone discussions collectively with our co-developers with
17	respect to dictating the price that each can ask of its
18	customers.
19	On the collective discussions that aren't
20	diversified, I had trouble sort of parsing that because I
21	think the effect is going to be exactly what I suggested
22	that we would fear, which was a shift to strong buyer power
23	by a much larger group of prospective licensees. It may be
24	that in an idealized simple A versus B scenario where
25	there are pure substitutes available, and it really is

111

1	distinguished on price, there may be an effect that selection
2	of one over the other will be determined by pricing and
3	it's a fair discussion. But the reality is that in none of
4	the groups that I am familiar with do such black and white
5	distinctions arise in practice. There's always
6	tradeoffs on performance, abilities, time to market, and
7	costs being one additional factor, but one additional
8	factor that if pressed would lead to potentially alienating
9	the very participants making the proposals.
10	MR. HEINER: I guess I too wonder if the joint
11	venture analogy is really right. In a joint venture
12	context, the parties to a venture are not competing with one
13	another. That's the essence of it. Whereas in the
14	standard setting context, the implementers typically will
15	be competing with each other in the implementation of the
16	standard. And that's very important.
17	So in one you're trying to preserve competition
18	and in the other you're not. In the standard setting
19	area, as you said, Bob, it's already something that raises
20	some concern in antitrust law since it's essentially a
21	group of firms coming together and agreeing on how
22	something should be done, rather than competing about how
23	that should be done. So, I think there is a legitimate
24	risk here.
25	I could then take it to the next level and say,

112

1	let's also have discussions about agreeing on pricing of
2	the technology that is the input to that standard.
3	MR. SKITOL: Well, see --
4	MR. PETERSON: Let me respond to that.
5	So, the pricing discussion that would be -- that
6	should be undertaken is only that pricing discussion that
7	is related to the cost of where they have agreed they're
8	not competing. So, in fact these are competitors as to
9	products which include implementations of standards. But
10	as to the standard, they're not competing. That's what the
11	exercise is about.
12	And I think too -- it's important to
13	realize that the decision to select the standard is the
14	relevant decision to which the price needs to be a factor.
15	And to suggest that the price can somehow efficiently, in
16	a market sense, be determined later is -- you know, the
17	prices of products, the prices of other cost components
18	will absolutely need to be determined later -- but the
19	decision on what this particular feature will be is being
20	made collectively.
21	And if that was not a procompetitive thing to
22	do, then that's a problem. There is a collective choice
23	of a particular thing where there will be no competition.
24	And it's entirely appropriate to consider the full
25	economic scenario of what will be the costs associated

113

1	with making that.
2	MR. HEINER: That is a little bit of a strong
3	statement because you may often have standards competing
4	with one another.
5	MR. PETERSON: And I agree. I make it a strong
6	statement in the extreme case. But there are a range.
7	But in the case where there is lock-in, yes.
8	MR. COHEN: Well, let me see if there's a
9	consensus on that.
10	The joint negotiation could in theory represent
11	al la monopsy with effects that might impede innovation
12	incentives.
13	MR. SKITOL: Well, that is a potential problem
14	that should be recognized but would rarely occur in the
15	real world. It's an antitrust problem only to the extent
16	that it would have the likely effect of reducing output or
17	reducing innovation, and that's a real stretch.
18	I would refer you to the extensive discussion on
19	the monopsony issue in Sony versus Soundview, where I think
20	the district court got it about right and made it clear
21	that the plaintiff's attack on the collective negotiation
22	that went on in that case involving the consumer
23	electronic players that the viability of the attack, the
24	antitrust claim against the collective negotiation that
25	occurred there, would depend on a showing of actual output

114

1	restraint or reduction. And it's a real stretch.
2	To my mind, it's a potential anticompetitive
3	effect of small likelihood, balanced against a major
4	procompetitive benefit that is very likely to occur in
5	many circumstances where negotiation would occur.
6	MR. HARTOGS: I probably already answered this
7	question. I clearly view that not only is it not a rare
8	occurrence, but it would be a frequent occurrence and
9	potentially one debilitating to the willingness of some
10	companies to participate in setting the standards.
11	To the extent that ex ante licensing already
12	does occur in certain instances, there's no prohibitions
13	on seeking licensing terms on a bilateral basis prior to
14	the setting of a standard. It does occur. When we look
15	at ourselves, we actually do provide transparencies to
16	all of the companies in the industry that we deal with.
17	We do deals ex ante, as probably many do.
18	MR. PETERSON: So, on this point, again, all
19	that we're talking about is a discussion of the cost of a
20	choice which is going to be made. And in fact to decide
21	the price of that later is not to postpone competition,
22	but in fact to make the choice without it having been
23	informed by the price information. So, in other words,
24	the idea that there is some -- the choice is whether or
25	not a particular technology is going to be collectively

115

1	decided to be put into the standard or not.
2	If the owner of the technology doesn't like the
3	price, at the end of the day, they can walk away at that
4	point. In other words, that's the power of the patent.
5	The patent has the power to be able to say, this is what I
6	have to offer. And so that's their walk-away opportunity
7	after the standard has been set.
8	The flip side walk-away opportunity, it seems,
9	in this event, is a collective one. And it in general
10	would be procompetitive because the value of creating
11	these standards is so useful. But it is a collective
12	event and it should include the economics associated with
13	it.
14	MR. COHEN: You touched on my last question,
15	whether there's an ability of the patentholders to
16	discipline a standard setting organization which too
17	aggressively pursues a price negotiation by either
18	withholding its technology or entirely leaving the
19	standard setting organization.
20	MR. SKITOL: On an ex ante basis, everybody has
21	got choice. The participants who are the potential
22	licensees have choices, but the patent owners who would
23	like to see their patented solutions adopted also more often
24	than not have choices.
25	So, if there's any dissatisfaction with what the

116

1	willing buyers seem willing to pay, then those patent
2	owners have the ability to go off and productize their
3	technology on their own or find some way to turn it into a
4	proprietary standard.
5	MR. COHEN: Is that realistic?
6	MR. HARTOGS: I think it's rarely realistic.
7	There are scenarios. I look at Motorola's now withdrawal
8	from their participation with VITA. But where you have
9	an organization like IEEE where you have such a broad
10	spectrum of standards and technologies, that viability of
11	not participating, not being a member severely handicaps
12	your ability to participate in business for the technologies
13	they address.
14	MR. PETERSON: I think this is an area where we,
15	as we said before, have many different experiences going
16	forward. There will be different sets of rules explored
17	and we'll develop experience with that in going forward.
18	In the past we had something that was a fairly
19	extreme policy, the W3C introduced a policy that requires
20	royalty free -- a royalty free result in a sense that they
21	don't want to issue a standard to which they're aware
22	there's some non-free patent. And the world has continued
23	to work with that. I don't think that that approach
24	applies to a wide range of other technologies, but that's
25	an example of where I think we need to try some things to

117

1	see where we actually stand.
2	MR. COHEN: Unless I hear an objection from any
3	of my panelists, I think we've covered the topic.
4	I want to thank all of you for your interesting
5	and insightful remarks. And I'd like to encourage the
6	audience to join me in a round of applause for our
7	speakers today.
8	(Applause.)
9	MR. COHEN: Our afternoon session will begin at
10	2:00. There's going to be a speaker luncheon at the
11	Berkeley Women's Faculty Club. Thank you.
12	(Whereupon, at 12:46 p.m., a lunch recess was
13	taken.)
14
15
16
17
18
19
20
21
22
23
24
25

118

1	AFTERNOON SESSION
2	(2:10 p.m.)
3	MS. GRIMM: Good afternoon everyone. I would
4	like to welcome you all to this session of our business
5	testimony hearings and I'm glad that you all could join
6	us.
7	I am Karen Grimm. I am Assistant General
8	Counsel for Policy Studies at the Federal Trade
9	Commission, and I am also one of the moderators of this
10	session.
11	My co-moderator is Joe Matelis, who you met this
12	morning, an attorney in the Legal Policy Section of the
13	Antitrust Division, U.S. Department of Justice.
14	Before we start, I just have to cover two
15	housekeeping matters. As a courtesy to our speakers,
16	please turn off your cell phones, your Blackberries, any
17	other devices you may have. And also we request that you
18	not make any comments or ask questions during the session.
19	We are honored to have a distinguished group of
20	panelists from the business community with us this
21	afternoon. They are, in order, Thomas McCoy, who is the
22	Executive Vice President of Legal Affairs and Chief
23	Administrative Officer at AMD; Michael Haglund, who is a
24	partner in Haglund Kelley Horngren Jones & Wilder in
25	Portland, Oregon, and counsel to Ross-Simmons, the

119

1	Weyerhaeuser -- and counsel to Ross-Simmons in the
2	Weyerhaeuser/Ross-Simmons predatory buying case; finally,
3	we have David Dull, who is the Vice President of Business
4	Affairs, General Counsel and Secretary of Broadcom
5	Corporation.
6	Our format this afternoon will be essentially
7	the same as this morning's. Each speaker will make a 20
8	to 30 minute presentation. After the presentations are
9	finished, we will take about a 15-minute break. And after
10	the break, we will reconvene and have a moderated
11	discussion with two of our panelists. Unfortunately,
12	David, who has a scheduling conflict, will not be able to
13	join us for the roundtable discussion. We are, however,
14	very grateful that he is still able to participate as a
15	presenter here this afternoon.
16	As Bill Cohen said this morning, these business
17	sessions are an extremely important component of the
18	Section 2 hearings overall. Over the last seven months or
19	so, we have held conduct specific hearings on predatory
20	pricing and buying, refusals to deal, tying, exclusive
21	dealing, bundled and royalty rebates and discounts, and
22	misleading and deceptive conduct.
23	Some of these prior panels have included
24	business executives or their in-house attorneys who are
25	typically heavily involved in the company's business

120

1	decision-making processes.
2	The sessions today are designed to further our
3	goal of obtaining as much real world insight as possible
4	into Section 2 issues from a business perspective and
5	basically from business executives and their counsel.
6	To that end, we have invited our business panel
7	to address whatever Section 2 issues they consider
8	important to their respective businesses and to share with
9	us any views they may have on how we at the FTC and the
10	Justice Department can better address those issues from an
11	enforcement perspective.
12	We heard a number of helpful suggestions this
13	morning. We look forward to our panelists' remarks in the
14	roundtable discussion this afternoon.
15	I want to thank all of today's panelists for
16	their participation. We appreciate all of them taking
17	time out of their very busy schedules to prepare for and
18	participate in these hearings.
19	I would now like to turn the podium over to my
20	DOJ colleague and co-moderator, Joe Matelis, for any
21	remarks he would want to make.
22	MR. MATELIS: Thank you, Karen.
23	I just have brief additional remarks to make in
24	addition to what Karen said.
25	On behalf of the Antitrust Division, I just

121

1	want to thank the Berkeley Center for Law and Technology
2	and the Competition Policy Center at the University of
3	California Berkeley for hosting these hearings today.
4	And also on behalf of the Antitrust Division, I
5	want to thank all of the panelists for volunteering your
6	time and sharing your insights with us
7	And finally I'd like to thank Karen and her
8	colleagues at the FTC for all of their hard work in
9	organizing this hearing and assembling such a fine panel.
10	MS. GRIMM: Our first speaker today is Tom
11	McCoy. He is Executive Vice President of Legal Affairs
12	and Chief Administrative Officer of AMD. Tom joined AMD
13	in January 1995 and was Senior Vice President, General
14	Counsel and Secretary until 2003.
15	Tom's current leadership responsibilities
16	include legal, business development, employee
17	communications, international policy, government and
18	community affairs, corporate secretary, environmental
19	health and safety, and global real estate. He's busy.
20	Mr. McCoy holds an undergraduate degree in
21	history from Stanford University and a law degree from the
22	University of Southern California.
23	Prior to coming to AMD, Tom spent 17 years
24	practicing law at O'Melveny & Myers, where he specialized
25	in business litigation. Tom.

122

1	MR. McCOY: Karen, thank you very much. And
2	thanks to everybody in the room. Thank you for having me
3	here today to share my thoughts and experience on this
4	very important topic. I'm particularly pleased to join my
5	fellow representatives in the technology industry in
6	presenting here today.
7	I believe our presence is a testament to a
8	common belief in the critical role that enforcement of
9	Section 2 plays in ensuring innovation and competition in
10	high technology sectors
11	Technology is often cited, and I believe
12	correctly so, as the driver of our new economy in a
13	rapidly globalizing world.
14	As Federal Reserve Chairman Ben Bernanke
15	emphasized in a speech just last August, the innovation
16	that technology companies produce spurs economic growth
17	and innovation, not only within the sector itself, but
18	outside the IP sector as well. His remarks cited numerous
19	economic studies demonstrating that information technology
20	was the single greatest impetus for the tremendous rise in
21	productivity our national economy experienced in the late
22	1990s.
23	So, what was behind the innovation surge? Ask
24	economic experts and their answer is simple:
25	competition. And, not coincidentally, more competition in

123

1	the microprocessor market, which produces the brains of
2	computers.
3	I've been with AMD for over a decade now and I
4	was a business and antitrust lawyer for nearly twenty
5	years before that, as was mentioned. I believe the
6	competitive dynamic within the microprocessor market provides
7	a particularly important example as we discuss Section 2.
8	Look at the late 1990s and the impact of the
9	speed of innovation in this market, before and after AMD
10	transformed from a second source follower to an innovation
11	leader. As Professor Michael Scherer testified in an
12	earlier hearing, that difference was dynamic. When
13	competition arrived, the pace of innovation quickened.
14	But as the Japanese Fair Trade Commission ruled
15	in 2005, that innovation of AMD did not go unpunished.
16	Because of the critical importance of the
17	technology sector to the strength of our national economy,
18	there is perhaps no market in which the committed
19	enforcement of antitrust law and competition policies are
20	more important.
21	But if we are to do so effectively, we must
22	first dispel the most common myths about the technology
23	marketplace. Namely, myth number one: Market power is
24	inherently transient in high tech industries. Myth number
25	two: Section 2 is not equipped to deal with the special

124

1	characteristics of high tech markets. And myth number
2	three: Consumers are not harmed if technology solution
3	prices are coming down.
4	The fact is, each of these myths is simply wrong
5	and acceptance would stand in the way of fair and open
6	competition in technology markets. Indeed, these myths
7	would empower a monopoly of use and consumer harm of the
8	very kind Section 2 is intended to stop.
9	Accordingly, we must rigorously consider how
10	firms, and dominant firms in particular, actually behave
11	in real markets. When we do, we will discover the
12	provable truths that should inform this discussion of
13	Section 2.
14	So, allow me to address these myths one by one.
15	First, myth number one, market power is
16	inherently transient in high technology industries. The
17	truth? In many high tech industries, just as in low and
18	no tech firms, customers are tied to the dominant firm for
19	a very large percentage of their requirements, at least in
20	the intermediate term. With their customers at their
21	mercy, dominant firms can use a combination of
22	exclusionary tactics, monopoly to both price and nonprice
23	behaviors in order to deter competition and preserve their
24	position in the marketplace. Monopoly tactics signal the
25	customers and the marketplace that other actors should

125

1	play ball.
2	This disrupts the natural balance of a free
3	market as innovators are no longer rewarded for building a
4	better mouse trap and selling it at a better price. I can
5	think of no better example then the global market in
6	microprocessors in which AMD competes. In its March 2005
7	ruling that I noted above, the JFDC cited evidence that
8	showed quite clearly from the beginning of this decade
9	until it was able to fend off competitive technologies
10	from AMD, which had been gaining market share, by using
11	its entrenched position in Japanese OEMs to crack down
12	through anticompetitive tactics, level of those that would
13	strive to bring differentiation and choice to endusers
14	around the world.
15	AMD has competed against a persistent monopolist
16	in a global market. We've confronted a variety of
17	exclusionary abuses, including payments for exclusivity;
18	rebates to make it too costly to ship to a rival even a
19	small share of the customer's business; threats to
20	withhold road maps, technical information and support;
21	discriminatory allocations and scarce parts; and delay or
22	reduced marketing share or substance.
23	In a vacuum, with names and faces attached, the
24	damaging impact of each of these individual acts may seem
25	less obvious. While the FTC and DOJ appropriately have

126

1	been examining specific practices one by one that occurred
2	previously, it is important not to lose sight of the fact,
3	as business firms competing against dominant firms know,
4	that dominant firms can and do use a combination of
5	practices, seldom just one, to maintain dominance. They
6	can modulate the mix of practices as rivals try to adjust
7	and react to maintain the marketplace in a prisoner's
8	dilemma.
9	What's important to understand is the collective
10	impact. These bad acts often add up to a pattern of
11	conduct that sends very strong signals to the marketplace,
12	signals that are direct and punitive and that have a
13	chilling effect on competition and the innovation process.
14	Once a monopolist has injected enough fear into
15	the marketplace, the need to explicitly threaten rivals
16	every time is eliminated. It becomes understood and all
17	too often accepted as the natural condition of the market.
18	This is how our rival, even when lagging behind
19	on the technological innovation front, manages to always
20	maintain more than eighty percent revenue share for more
21	than a decade. In other words, the dominant firm is
22	perfectly capable of maintaining its market share through
23	abusive conduct, even in a high technology market, for
24	indefinite periods of time. This is particularly true in
25	markets where the barriers to entry, including

127

1	intellectual property and capital, are so very, very high.
2	Which leads me to myth number two: Section 2 is
3	not equipped to deal with the special characteristics of
4	high tech markets. The truth? There is general agreement
5	among global regulatory bodies as to what constitutes bad
6	conduct on the part of dominant players in the market.
7	And under those standards, bad conduct is bad
8	conduct, plain and simple, no matter the industry in
9	question. There is nothing unique about technology,
10	whether it's the oil business, the pharmaceutical
11	business, the chemical business or the computer business.
12	The microprocessor market, once again, provides
13	an example. In 2002, when AMD set out to earn its place
14	in HP's commercial desktop product road map, AMD agreed to
15	provide HP with one million processors for free, not just
16	any processors, but the most advanced chips in its
17	portfolio. HP was able to use only 140,000 and left
18	860,000 units, free units, on the table. We believe
19	because, had it taken more, its AMD-related savings would
20	have been cancelled out several times over because of
21	penalty Intel would have exacted in the form of higher
22	prices on HP's Intel purchases.
23	The result? Customers paid more; were forcibly
24	deprived of an AMD alternative that might have been more
25	suitable for their needs.

128

1	Or take the recent revelation in the "Financial
2	Times Deutschland" that Intel has entered into an
3	exclusive contract Germany Media-Saturn-Holding,
4	stipulating that competitors of Intel such as chipmaker
5	AMD are not allowed to sell their products in Germany's
6	dominant PC retail.
7	The result? While consumers elsewhere in Europe
8	favor AMD-powered computers, because they get a better
9	equipped system for the same number of Euros, any German
10	customers don't get to choose. The product in the
11	marketplace in question are indeed complex, but the abuse
12	of that should be a question for [unintelligible].
13	Nor are these examples unique. Consider the
14	Rambus 2006 Federal Trade Commission order, which stated
15	that, quote, "Rambus engaged in exclusionary conduct which
16	significantly contributed to its acquisition of monopoly
17	power in four-related markets." Or the often overlooked
18	original Microsoft decree that banned Microsoft from
19	requiring its OEMs to pay the same licensing fees whether
20	they installed the Windows operating system or not,
21	thereby forcing the buyers and substitute operating
22	systems to give their product away for free.
23	In fact, if we take a moment to consider the
24	fundamental considerations underlying the most high
25	profile technology industry cases that come before the

129

1	courts, we find at their core anticompetitive conduct that
2	is almost universally recognized as impermissible under
3	antitrust standards around the globe, which clearly falls
4	within the band of Section 2.
5	Perceptions like these exist around the industry
6	and they cloud our ability to protect consumers.
7	But none is more damaging than the industry myth
8	that I'd like to address here today. Myth number three:
9	Consumers aren't harmed if system prices are coming down.
10	The truth? Apparent discounts are not always real
11	discounts. Exclusionary conduct by monopolies keeps
12	prices higher, slows innovation and limits consumer
13	choice.
14	There's plenty of real precedent from around the
15	world from every industry to support this point. Consider
16	"The United States vs. Dentsply International, Inc.,"
17	Third Circuit case. The Third Circuit recognized that
18	Dentsply's exclusive dealing arrangement improperly
19	limited the ability of its rivals to compete, thus denying
20	customer choice.
21	And in its decision in LePage's, Inc, which is
22	3M, the Third Circuit similarly made the claim that the
23	application of Section 2 to exclusionary conduct,
24	explaining that, quote, "Even the foreclosure of one
25	significant competitor from the market may lead to higher

130

1	prices and reduced output."
2	And the European Commission acted recently in
3	the "Tomra" decision to make plain that, as its
4	Competition Commissioner explained, quote, "I will not
5	tolerate dominant companies hindering competition or
6	excluding other players from the market as this harms
7	innovation and consumers. Rebates and discounts cannot be
8	used by a dominant company as part of the strategy to
9	exclude actual and potential competitors."
10	For instance, industry analysts have recently
11	suggested that if the x86 microprocessor market were fully
12	competitive, it would have allowed AMD to gain a greater
13	share of the market and far more benefits would have been
14	delivered to consumers in the form of lower prices and
15	better and faster innovation.
16	In recent economic analysis by Cal Tech
17	Professor Preston McAfee shows that the U.S. Government
18	pays higher prices and squanders taxpayer dollars when
19	procurement prices are curbed by brand-specific
20	specifications and contracts that foreclose competition
21	and the benefits that open procurement policies promote.
22	As with the aforementioned Microsoft decree,
23	often what passes for pricing is just the imposition of a
24	legal condition and the veiled threat of yet higher prices
25	to exclude competition.

131

1	That's why I believe these hearings are so
2	important. And I commend the Department of Justice and
3	the Federal Trade Commission for bringing them here to
4	Berkeley, so close to the heart of the U.S. technology
5	industry in Silicon Valley. Because, while rigorous
6	enforcement of Section 2 is important, it is absolutely
7	vital to the continued success of the United States
8	technology industries.
9	As we look to craft sound competition policy to
10	govern our industries, we must consider the way in which
11	these markets function in the real world. We cannot get
12	caught ignoring tangible truths in favor of marketplace
13	myths. We must send a strong deterrent message to all
14	industries, including technology, Section 2 applies to
15	what we do and who you harm. And crossing the line into
16	illegality will not be permitted, no matter how cool the
17	product, how familiar the logo or how high tech the
18	industry.
19	In the technology market, the stakes are
20	particularly high because the progress of innovation and
21	the health of our broader national economy in a
22	globalizing world requires both robust competition and
23	robust and enforced competition policy
24	Thank you very much.
25	MS. GRIMM: Thank you, Tom.

132

1	Our next speaker is Michael Haglund. Can you
2	hear me?
3	Mike is a partner in Haglund Kelley Horngren
4	Jones & Wilder in Portland, Oregon, and counsel to
5	Ross-Simmons. He graduated in 1973 from Western Oregon
6	University with a B.A. in Education, and he received his
7	law degree from Boston University in 1977
8	Mr. Haglund has primarily practiced in natural
9	resources, admiralty and general business law throughout
10	his career, and is experienced in a wide range of legal
11	representation, including antitrust
12	In 2003, he acted as lead counsel for the
13	largest antitrust verdict in the history of the Pacific
14	Northwest, a $79 million dollar judgment against
15	Weyerhaeuser.
16	In November of 2006, Mr. Haglund argued in the
17	U.S. Supreme Court on behalf of Ross-Simmons in
18	"Ross-Simmons v. Weyerhaeuser," a Section 2 case involving
19	allegations of predatory bidding or buying
20	Mike.
21	MR. HAGLUND: Thank you.
22	I wish to thank the Federal Trade Commission and
23	the U.S. Department of Justice Antitrust Division for the
24	invitation to present testimony today as part of this
25	series of hearings on Section 2 of the Sherman Act

133

1	I offer this testimony, not on behalf of any
2	individual business or client, but from the perspective of
3	the many small and medium-sized businesses, mostly family
4	owned, that I have been privileged to represent throughout
5	the course of my career in the resource-based industries
6	of the Pacific Northwest.
7	I am the exception on the program today. I'm
8	more of a bricks-and-mortar or in-the-ground kind of
9	antitrust practitioner. I'm in my thirtieth year of
10	law practice and have devoted most of that to the
11	representation of the small and medium-sized participants
12	in the forest products, fishing and agricultural
13	industries.
14	One of the common threads of this client base
15	has been the production or is the production of
16	commodities derived from the rich natural resources of our
17	region in the Pacific Northwest: logs, lumber and plywood
18	in the forest products industry; salmon and crab in the
19	fishing industry; and essential oils like peppermint or
20	spearmint, in agriculture.
21	The application of Section 2 to these types of
22	markets is important and must be analyzed within the
23	context of the unique market realities that govern those
24	markets, where in many cases there is the potential for a
25	dominant buyer to exercise monopsony power to the

134

1	detriment of its small competitors, input or commodity
2	sellers generally, and ultimately consumers.
3	These markets may be localized in that they're
4	confined to a region of the United States and they are
5	often exemplified by what Professor Warren Grimes refers
6	to as, quote, "small atomistic sellers," unquote, who are
7	more vulnerable to market abuses than consumers.
8	There are multiple such markets in the Pacific
9	Northwest, where a large and diverse number of small
10	players are selling their commodity products to firms that
11	process the logs, the fish, or the agricultural product
12	into a host of other products.
13	In some markets, the processor base may be quite
14	small and dominated by one or a few large firms. As
15	Professor Roger Noel has observed, "Local monopsony in
16	conditions where the monopsonist does not have market
17	power at the output level in a national or regional
18	market, causes harm to consumers by misallocating
19	production across regions or across localities."
20	Antitrust cases associated with input markets
21	have received very little attention until quite recently.
22	In fact, a good share of the scholarship on the subject
23	that exists today is found in this quarterly 2005 issue of
24	the "Antitrust Law Journal," which contains a symposium
25	collection of nine articles, including the two I've

135

1	referenced from Professors Grimes and Noel a moment ago.
2	The application of Section 2 to input markets is
3	an area of antitrust law deserving of more attention, in
4	my view, and it is about to receive it from the United
5	States Supreme Court in its forthcoming decision in
6	"Weyerhaeuser vs. Ross-Simmons Hardwood Lumber Company,"
7	which will likely be handed down in March or April of this
8	year.
9	I argued the Weyerhaeuser case on behalf of
10	respondent Ross-Simmons before the Supreme Court the end
11	of November. Although it is difficult, and some would say
12	dangerous, to make predictions based upon the briefs and
13	the oral argument, but having been with the case since its
14	inception and lead counsel at trial, and arguing counsel
15	both in the Ninth Circuit and the Supreme Court, I believe
16	the result is going to surprise people.
17	When cert was granted, all of the pundits
18	predicted that the court had taken the case to reverse it.
19	And that view is still being expressed post argument on
20	various blogs that follow the Supreme Court docket.
21	For those of you who may not be fully aware, the
22	Weyerhaeuser case as to predatory bidding or buying in
23	input markets presents two issues. The first, whether the
24	Brooke Group Price Cost Test, which was adopted in 1993,
25	should be extended from the sell side to the buy side,

136

1	first issue. And, second, whether the jury instruction
2	regarding predatory bidding was flawed on grounds other
3	than Brooke Group.
4	The first issue, based upon the briefing and
5	based upon the tenor of the oral argument, we are
6	optimistic that the Supreme Court is going to affirm the
7	Ninth Circuit in its decision that the safe harbor for
8	pricing behavior that exists on the sell side through
9	Brooke Group does not apply with the same force and should
10	not be extended at least to inelastic input markets like
11	the alder saw market at issue in the Weyerhaeuser case.
12	Over the last quarter century, except for Brooke
13	Group, the Supreme Court has eliminated or narrowed per se
14	rules that did not have a sound economic foundation in the
15	market realities of the individual case.
16	The wisdom of Brooke Group most I think would
17	say is its protection of inherently procompetitive price
18	cutting in output markets. In the context of input
19	markets, the challenged conduct involves price raising,
20	bidding, resource prices up. Very few cases in the last
21	fifty years and scholarship in its infancy. Conditions
22	that are the exact opposite of those that prevail when
23	Brooke Group's per se rule was developed.
24	In these circumstances, the correct approach is
25	the one that has always been the gold standard of

137

1	antitrust rules, the rule of reason.
2	The rationale underlying Brooke Group was also
3	rounded substantially in concern about false positives,
4	based in large part upon a sizable body of literature to
5	that effect.
6	In the predatory bidding context, there is no
7	similar body of economic literature offering a similar
8	warning. In point of fact, the very few cases of
9	overbidding that do exist show that it is a rational
10	strategy that does work. And I'm referring here to just a
11	very few cases: American Tobacco from the Supreme Court;
12	the Ross-Simmons case about to be decided; and the Reed
13	Brothers case also out of the timber market that was
14	decided by the Ninth Circuit in 1983.
15	There are two reasons underlying my optimism
16	that the Supreme Court will refuse to extend Brooke Group
17	from the predatory selling context to immunize bidding
18	conduct by a dominant buyer.
19	First, the position of Weyerhaeuser and its many
20	big business amici is based upon the notion of symmetry,
21	that a rule that works for predatory selling and output
22	markets should apply equally in predatory bidding to input
23	markets by the sheer force of logic alone.
24	The law, however, is no slave to symmetry. As
25	Justice Holmes has written in what has been characterized

138

1	by Judge Posner as the single most famous sentence in
2	American legal scholarship, quote, "The life of the law
3	has not been logic; it has been experience."
4	In the past, notions of symmetry have influenced
5	the antitrust juris prudence of the U.S. Supreme Court.
6	However, in the last twenty-five years, market realities
7	have consistently trumped symmetry and the per se rules
8	which were sometimes developed as a result.
9	The Supreme Court embraced symmetry, for
10	example, in equating maximum and minimum vertical resale
11	price constraints as per se illegal in "Albrecht vs.
12	Harold Company" in 1968, but relied on market realities in
13	overruling Albrecht's prohibition against maximum resale
14	pricing agreements nearly thirty years later in "State Oil
15	vs. Khan" in 1997.
16	The other half of that rule, by the way, now
17	appears in some jeopardy with the Supreme Court's recent
18	decision to reexamine whether vertical minimum resale
19	price maintenance agreements should be deemed per se
20	illegal under Section 1 of the Sherman Act, or whether
21	they should instead be evaluated under the rule of reason.
22	I refer here to "Leegin Creative Leather Products vs.
23	PSKS," a decision out of the Fifth Circuit on which cert
24	was granted just last month.
25	In my view, the Supreme Court is clearly focused

139

1	on eliminating per se rules or presumptions in antitrust
2	which are not justified by market realities or which
3	distort the fact-finding process at trial in a way that
4	unfairly disadvantages one party or the other.
5	The Independent Ink case of last term, in which
6	the court abandoned the per se rule that patent equals
7	market power in a tie-in case is the most recent example
8	of this trend.
9	My second reason for optimism on the Brooke
10	Group issues comes from the oral argument. We were struck
11	by the apparent lack of enthusiasm among the Supreme Court
12	Justices for extending Brooke Group from the sell side to
13	the buy side. Several justices, including Justice
14	Kennedy, who wrote the 6-3 majority opinion in Brooke
15	Group, expressed concern about the workability of
16	converting the Brooke Group price cost test into a
17	price revenue test on the buy side.
18	There was record evidence that Weyerhaeuser used
19	below-market transfers of all their saw logs from its
20	company fee lands to subsidize its bidding up of saw log
21	prices in the so-called open market in which it competed
22	with Ross-Simmons. Weyerhaeuser argued that such bidding
23	was immune from antitrust scrutiny so long as its alder
24	division was not losing money overall.
25	Adoption of such a rule, however, in this type

140

1	of resource market would put a large company that had
2	amassed low cost raw materials in a position to eliminate
3	its competition by bidding up scarce supplies of open
4	market sources and subsidizing that predation with below
5	market transfer prices from its own captive supplies.
6	The result would be under-deterrence of
7	predatory bidding behavior, while impeding the most
8	efficient allocation of scare resources.
9	Another administrability problem not found with
10	Brooke Group on the sell side is associated with the fact
11	that the relevant input in the Weyerhaeuser case, alder saw
12	logs, are used to produce very different products. In an
13	alder saw mill those are chips; pallet lumber, which is a
14	low-grade type of lumber which you see underneath products
15	in various Costcos and elsewhere; and kiln-dried finish
16	lumber. But Weyerhaeuser actually had 25 to 50 different
17	lumber grades in the finished lumber category
18	Each of the saw logs that went through any
19	given alder mill produces products in all three of these
20	categories, but the larger the diameter of the log, the
21	even more higher grade lumber you're going to produce.
22	Applying Brooke Group is extremely difficult in
23	this sort of single input but multiple product output
24	environment. And there is no comparable corollary
25	on the buy side to the commonly utilized average variable

141

1	cost or marginal cost formulation used in the sell side
2	predatory pricing case.
3	In sum, regarding the primary question in
4	Weyerhaeuser, whether to extend Brooke Group to the buy
5	side, we are guardedly optimistic that the Supreme Court
6	will decline to do so because of the court's consistency
7	over the last quarter century in refusing to create new
8	per se rules or to extend old ones unless justified by the
9	market realities of the particular industry or the
10	particular type of antitrust claim.
11	And, also, because of the TENOR of the oral
12	argument. Brooke Group really was an exceptional case.
13	Today, 14 years after it was decided, the rule of reason
14	shines even more brightly as the gold standard of
15	antitrust analysis.
16	Now, assuming the Supreme Court does not extend
17	Brooke Group to the buy side in Weyerhaeuser, it must then
18	examine a second issue, whether the district court's
19	instructions defining when predatory bidding will
20	constitute anticompetitive conduct were flawed on some
21	other basis.
22	This was the instruction in which the district
23	judge, having given the standard ABA model instructions
24	for monopolization and anticompetitive conduct, instructed
25	the jury that it could find that Weyerhaeuser engaged in

142

1	anticompetitive conduct if it bought more logs than it
2	needed or, quote, "paid a higher price than necessary in
3	order to prevent plaintiffs from obtaining the logs that
4	they needed at a fair price," unquote.
5	This formulation was pounced upon by
6	Weyerhaeuser and its amicis as, in their words, "standard
7	gibberish," which constituted an independent ground beyond
8	Brooke Group for reversal of the Ninth Circuit opinion.
9	However, as pointed out in our merits brief, Weyerhaeuser
10	never preserved any such alternative objection to the
11	instruction. Attacking a pair of sentences in the jury
12	instructions as unduly subjective or as an invitation for
13	unguided speculation, proved an effective springboard for
14	a grant of certiorari. But deciding the case on the
15	merits requires an assessment of the instructions as a
16	whole in light of the evidence, the closing arguments and
17	the other instructions.
18	In the trial court, Weyerhaeuser's counsel
19	actually invited the formulation of the two sentences that
20	have been so criticized in the commentary about this case.
21	But in opening statements, and again in closing argument,
22	Weyerhaeuser's counsel told the jury that multiple
23	witnesses would be called who would and then did testify
24	that the company never bought more than it needed and
25	never pushed log prices up in order to hurt its

143

1	competition. And a litany of two questions was put to 13
2	different witnesses, obtaining denials on each of those
3	same two points.
4	It's worth noting that the Supreme Court has
5	already decided the case from the very first one of this
6	term involving a challenge to ambiguous language in a jury
7	instruction. In "Aires vs. Del Montes," the court
8	examined California's catch-all mitigation instruction and
9	using the instructions in the penalty phase of a capital
10	murder case.
11	Based upon the way the case was tried and the
12	evidence presented, a 5-4 majority found no reasonable
13	likelihood that the jury had applied the admittedly
14	ambiguous instruction in a way that prevented
15	consideration of constitutionally relevant evidence.
16	If the type of common sense -- and I put that
17	word in quotes because that was the court's term. If that
18	type of common sense approach is to apply in a capital
19	murder case to consideration of ambiguous instruction,
20	it's hard to see how there is a reason for a stricter
21	approach in antitrust, especially in a case where the
22	defendant tried the case in a manner that invited the very
23	formulation of that jury instruction.
24	In fairness, however, it should be noted that I
25	was pressed at oral argument, particularly by Justice

144

1	Souter, regarding the vagueness of the instruction on
2	predatory bidding and the need for the Supreme Court to
3	say something about that instruction. I conceded that the
4	instruction was not perfect, but emphasized that neither
5	the district judge nor plaintiff's counsel was given any
6	chance through a defense objection on that ground to
7	consider whether the instruction could be made more
8	precise with other language.
9	At trial, we in fact never attempted to exploit
10	the nature of that couple of sentences and urged the jury
11	to just award whatever they considered was fair. Instead,
12	through economists, forest economists, we presented
13	detailed market evidence to show how much the market for
14	alder saw logs was artificially elevated above where it
15	would have been but for the mix of anticompetitive
16	practices, including manipulative bidding by the defendant
17	Ultimately, the jury in Weyerhaeuser selected to
18	the dollar one of the three damages scenarios presented by
19	these forest economists. Had Weyerhaeuser challenged the,
20	quote," paid a higher price than necessary," unquote,
21	language, we would have had no problem adding precision to
22	that instruction by linking the higher log prices to
23	market factors tied to Weyerhaeuser's manipulative
24	behavior as opposed to the normal operation of the market
25	In fact, we could have accepted the suggestion

145

1	made by the eight amicus states that filed a brief
2	supporting Ross-Simmons, including Oregon and California,
3	that the instruction that defined predatory bidding as
4	having anticompetitive effect, quote, if the conduct
5	raised the price that the buyers' rivals had to pay for
6	the input beyond the level that could be justified or
7	explained by other market factors and substantially
8	affected the ability of the buyers' rivals to compete for
9	the input.
10	Because our evidence was designed to show how
11	the historical relative equilibrium between finished
12	lumber prices and log prices had been distorted by
13	Weyerhaeuser's behavior in order to kill off rivals, I'm
14	confident that there would have been no change in the
15	result at trial with a more precise formulation for
16	defining when bidding conduct in an input market can be
17	found anticompetitive.
18	What happens, you might ask, however, if my
19	admittedly optimistic view is wrong and the Supreme Court
20	reaches the vague instruction issue and reverses on that
21	basis. In all likelihood, a retrial will then be
22	necessary, but we are confident of a similar plaintiff's
23	verdict for two reasons.
24	First, the Ross-Simmons verdict generated
25	several follow-on cases in which Weyerhaeuser produced

146

1	thousands of additional incriminating documents,
2	demonstrating the deliberate character of its multi-tactic
3	plan to monopsonize the alder saw log market in the
4	Pacific Northwest.
5	By the way, the Pacific Northwest is the only
6	place west of the Mississippi where there is a hardwood
7	industry, in stark contrast to the east, where hardwood
8	species predominate and there's a substantial hardwood
9	industry.
10	In other words, we're even stronger on liability
11	in the retrial than we were the first time around, and
12	perhaps that's why Weyerhaeuser settled three follow-on
13	cases we handled on behalf of ten other plaintiffs for a
14	total of $62 million.
15	Provided we are not saddled with a Brooke Group
16	test, we believe our damages theory can easily be matched
17	up with a more objective formulation of the market
18	distorting bidding conduct than the two-sentence
19	formulation now at issue before the Supreme Court.
20	But however it turns out, the Weyerhaeuser case
21	will be important for all resource space input markets,
22	particularly those at the inelastic end of the spectrum.
23	Section 2 has a real role to play in these markets. If
24	you are a tree farmer, you want to have a healthy number
25	of saw mills competing for your log production within a

147

1	reasonable distance of your tree farm. And even if
2	you happen to sell your logs of a particular species to a
3	rising or emerging monopsonist, paying premium prices
4	during this period of predation, you're concerned about
5	the long-term health of your input market for that
6	particular species and will likely cause you not to
7	replant it if you fear that there will only be a single
8	buyer 30 to 50 years down the road when those seedlings
9	are now mature and ready for harvest. And we have
10	evidence to that effect.
11	It was precisely this type of real market
12	consideration that caused most of the log seller community
13	in the U.S., represented by the National Woodland Owners
14	Association and the American Loggers Council, to support
15	Ross-Simmons in an amicus brief in the Supreme Court.
16	Avoiding expansion of Brooke Group from the sell
17	side to the buy side is important in other input markets
18	as well. Most U.S. fish markets are classically inelastic
19	because the total catch is fixed by state and federal
20	regulators. The crab fishermen plying U.S. waters off the
21	coast of Oregon, Washington and Alaska need a healthy mix
22	of seafood processors to ensure market prices that sustain
23	the crab industry and its U.S. fleet.
24	A flexible rule of reason approach to
25	exclusionary conduct in this type of market is vital both

148

1	to deterring illegal conduct and to ensuring fair results
2	at trial
3	Also, many agricultural markets, especially
4	those like peppermint where production is regulated by
5	federal marketing orders, are susceptible to abuse in the
6	form of artificially low prices dictated by a dominant
7	buyer, or oligopolistic behavior in a highly concentrated
8	processor market
9	I would like to take this opportunity to thank
10	the FTC and the DOJ Antitrust Division for holding this
11	hearing out on the west coast rather than in Washington,
12	D.C. I believe it is critically important for federal
13	antitrust enforcers to be out in the field regularly to
14	have a full appreciation of the importance of local and
15	regional markets
16	Indeed, the lack of consideration of local and
17	regional markets in the Solicitor General's brief
18	supporting Weyerhaeuser was one of the primary reasons, I
19	am told by state officials, that eight states on short
20	notice submitted their amicus briefs on Ross-Simmons' side
21	in this case.
22	In its antitrust jurisprudence, the Supreme
23	Court has repeatedly emphasized that antitrust analysis,
24	quote, "must be attuned to the particular structure and
25	circumstance of the industry at issue," unquote.

149

1	In my view, this can only be accomplished if one
2	is immersed in the facts and circumstances of a given
3	industry, what I call the who, what, when, where and how
4	that requires extensive use of investigative interviewing
5	in addition to and not as a substitute for analysis of raw
6	data.
7	From my experience in the northwest corner of
8	the United States, I have three suggestions for the FTC
9	and DOJ in its evaluation of antitrust issues to resource
10	space input markets
11	First, please do not discount or dismiss the
12	significance of a local or regional market simply because
13	the dominant buyer/processor may not have the market --
14	may not have market power in the downstream output market.
15	As Professor Noel so convincingly demonstrated
16	in his article, this is an area where input sellers are
17	vulnerable and can be abused by a monopsonist to the
18	detriment of both regional and national economies.
19	Second, please be aware of the influential
20	impact of the extraordinary legal and organizational
21	talent brought to bear by large corporations and their
22	affiliated support organizations on the antitrust issues
23	that come before you. The small, atomistic sellers who
24	make up so many of the local and regional input resource
25	based input markets in the U.S. are no where near as well

150

1	organized and have precious little in the way of financial
2	resources to devote to long-term efforts to influence the
3	direction of Sherman Act jurisprudence.
4	It is therefore particularly important that
5	federal and state antitrust enforcers to look behind the
6	incredibly capable advocacy available to large corporate
7	interests, and to independently investigate the relevant
8	facts of each market and each industry, and I emphasize,
9	in the field.
10	Third and finally, from my perspective,
11	throughout a now 30-year career involved in three resource
12	based sectors of the U.S. economy in the Pacific
13	Northwest, I have been struck by the close match between
14	my own experience and two bedrock principles of antitrust
15	law.
16	One, that the forms of anticompetitive conduct
17	are myriad. And, two, that sound antitrust analysis is
18	joined at the hip with the fact-laden structure of the
19	particular market and industry at issue. This amazing
20	factual variability, in my view, makes the quest for a
21	unitary standard of exclusionary conduct under Section 2
22	illusionary. It is a much sounder policy to embrace the
23	flexibility of the rule of reason standard and to apply it
24	appropriately to the market realities of the industry in
25	the particular antitrust case.

151

1	On this last point, I think it's interesting to
2	note that our own -- excuse me, that own new Chief Justice
3	appears to be no fan of etiological purity in the way the
4	Supreme Court decides its cases. In a very insightful
5	article by Jeffrey Rosen in the January/February issue of
6	"The Atlantic Monthly," Chief Justice Roberts says the
7	following when asked to define the qualities of judicial
8	temperament that he thought successful Chief Justices like
9	Marshall, who was Chief Justice Roberts own personal
10	model, embodied. Quote, "I think judicial temperament is
11	a willingness to step back from your own committed views
12	of the correct jurisprudential approach and evaluate those
13	views in terms of your role as a judge. It's the
14	difference between being a judge and being a law
15	professor," unquote.
16	I think the quest by some in the antitrust
17	division to develop an overarching standard defining all
18	anticompetitive conduct under Section 2 of the Sherman Act
19	is inconsistent with the highly fact-laden and
20	industry-specific character of antitrust. Such a quest is
21	too much of law professor and too little of the practical
22	fact-based enforcer. It should be abandoned and the
23	energy of our antitrust agencies refocused on
24	investigation and enforcement.
25	Thank you for the opportunity to present this

152

1	testimony.
2	(Applause.)
3	MS. GRIMM: Thank you, Mike.
4	Our third and final speaker this afternoon is
5	David Dull, who is Senior Vice President of Business
6	Affairs, General Counsel and Secretary of Broadcom
7	Corporation.
8	Mr. Dull is responsible for the company's
9	acquisition, outside investment and licensing activities,
10	in addition to advising on all legal matters.
11	Mr. Dull joined Broadcom as Vice President of
12	Business Affairs and General Counsel in March 1998, and
13	was elected Secretary of the corporation in April 1998
14	Mr. Dull received a B.A. and a J.D. from Yale
15	university.
16	MR. DULL: Thanks, Karen, for that kind
17	introduction. And thanks to the Haas School and its
18	affiliates here in Berkeley for hosting this event today.
19	I want to compliment the FTC and the Department
20	of Justice for convening these hearings. While like many
21	in the business, we at Broadcom are of course wary of
22	regulation and other governmental and court interventions
23	that may stifle growth and cause inefficiency.
24	We nonetheless recognize the positive role our
25	government has played and can still play in facilitating

153

1	economic growth, efficiency and innovation, which
2	ultimately is what drives our economy.
3	I thank and commend the FTC and the DOJ for
4	taking the time to solicit views from across the spectrum
5	and across the country and hope that what comes out of
6	this process will promote that positive role.
7	Let me begin my remarks by telling you a little
8	bit about the company I've been with since 1998, Broadcom
9	Corporation. In 1991, a graduate student by the name of
10	Henry Nicholas, and his professor, our current chairman,
11	Dr. Henry Samueli, had a vision of an innovative company
12	that would provide semiconductors, computer chips, to
13	facilitate high speed digital communications for business
14	and consumer applications.
15	In a world where television and cell phones were
16	still analog, no one had heard of HD TV, dial-up modems
17	were considered cutting edge technology, and few even
18	contemplated the potential of the internet and today's
19	laptops and hand-held devices. These two visionaries saw
20	that the demand for high bandwidth digital communications
21	would skyrocket. And of course it has.
22	Broadcom's revenue now exceeds three billion
23	dollars a year. We've retained our roots in Southern
24	California, but we now have facilities all over the United
25	States and around the world, including several facilities

154

1	and over 1,250 employees here in the Bay Area.
2	We continue to focus on semiconductors for high
3	speed, high bandwidth applications, such as set-top boxes
4	for television, gigabit ethernet, DSL modems, wireless
5	networking, and cellular phones. We also produce
6	closely-related devices, such as digital TV chips and
7	multimedia chips for iPods and cell phones.
8	Indeed, it is far to say that, as much as any
9	other party or any other factor, Broadcom has enabled the
10	digital communications revolution that touches each of us
11	every day.
12	And we continue to follow the example of our
13	founders. We have built our entire business model around
14	continuing innovation. Our products are state of the art
15	and Broadcom is a technology leader in every market in
16	which we play.
17	Our engineers are top-notch. In fact, of our
18	5,200 or so employees, more than 3,800 are engaged in R&D;
19	439 are Ph.Ds. We spend about 40% of our gross profit on
20	R&D, on innovation.
21	In keeping with the purpose of these hearings,
22	today I plan to talk a bit about real issues that we
23	confront in the high tech industry in which we operate.
24	These are not your father's competition issues.
25	Everyone in this room is keenly aware that the

155

1	antitrust laws date back to the end of the 19th century.
2	So, one overriding theme I hope you will take away from my
3	remarks today is that antitrust laws must not get trapped
4	in traditional analysis or outmoded or dated thinking.
5	They must be dynamic and flexible.
6	With due deference to economic analysis and
7	marketplace realities, our antitrust regime, including
8	that addressing single-firm conduct, must remain robust to
9	deal with the issues of the 21st century. And, as we all
10	know, many of those issues revolve around technologies in
11	the high tech industries.
12	We at Broadcom firmly believe that competition
13	is what makes our innovation economy work. When coupled
14	with a well-educated and highly motivated work force,
15	competition unleashes creative energy and creativity
16	spawns the amazing innovations that we have seen just in
17	the past decade alone.
18	In the semiconductor industry, as Tom knows,
19	competition creates efficiency on a scale greater than
20	anywhere else. The capability of today's high tech
21	products dwarf those of just a few years ago, yet prices
22	continue to drop.
23	The antitrust laws serve their most useful role
24	when they promote competition and prevent companies that
25	have obtained a strong position in one area from

156

1	exploiting it to prevent competition in other areas.
2	Before addressing that in greater detail, let me
3	be clear about two things. First, it is important not to
4	penalize innovation by attacking those companies that have
5	achieved strong market positions solely through
6	innovation. Innovation must be encouraged because it is
7	the key to our country's continued success in the
8	increasingly challenging global economy.
9	Secondly, it is important that the intellectual
10	property rights of innovation be respected. Our patent
11	system encourages innovation by ensuring that its vendors
12	will reap a portion of the economic benefits of their
13	inventions, while at the same time requiring those
14	inventions to be shared with the public. That is a good
15	thing and we must not sacrifice it in the name of
16	competition.
17	At Broadcom, we hold over 1,900 U.S. patents and
18	have another 5,900 U.S. and foreign patent applications
19	pending. We care deeply about intellectual property
20	rights. But companies that use the strong positions they
21	have obtained, even if attained by innovation, to close
22	other markets to competition, or that use deception and
23	false promises to obtain their strong position in the
24	first place, are not innovative, but rather are standing
25	in the way of innovation. The antitrust law must address

157

1	that type of behavior.
2	As I said, Broadcom designs and sells computer
3	chips. In today's highly sophisticated electronic
4	applications, be they computers, cell phones or cable
5	boxes, no one produces all of the systems and components
6	for a particular application. In fact, a typical consumer
7	product incorporates chips and software from a number of
8	different suppliers.
9	In our vernacular, no one company produces all
10	of the silicon on the motherboard. Today, in hardware and
11	software, open systems is the name of the game. Open
12	systems are why we have the PCs and the internet.
13	Interfaces between one component and another are therefore
14	necessary. Some of those interfaces are specified by
15	standards developed with broad industry participation
16	under the auspices of standard setting bodies such as the
17	IEEE and ANSI.
18	The highly successful 802.11B and G wireless
19	networking standards fall into this category. The
20	proliferation of Wi-Fi networking, supported by devices
21	from hundreds of manufacturers, demonstrate the power of
22	industry standards arrived at through non-partisan
23	processes.
24	Other interfaces are de facto industry standards
25	that arose without a formal standard setting process, but

158

1	are generally open for industry participants to use in
2	deploying their own standards compliant price.
3	And some interfaces are entirely proprietary,
4	which is to say they're put into place unilaterally by one
5	or another industry player who claims ownership of that,
6	quote, "standard," unquote, and asserts the right to
7	prevent or control its use by others.
8	Obtaining control of key interfaces through
9	anticompetitive means, or using control of key interfaces
10	to extend a dominant position in one market into other
11	markets is a real danger in our industry. It is of major
12	concern to companies like Broadcom who win through their
13	ability to innovate.
14	It should also be of concern to consumers and to
15	their representatives in the antitrust agencies. That
16	sort of behavior chokes off competition among industry
17	players, which deprives consumers of the innovations and
18	lower prices that come from vigorous competition.
19	At its most extreme, in our industry, interface
20	control could enable a dominant firm in one critical piece
21	of the motherboard to take control of the whole system,
22	even if the quality and cost of its products do not
23	support that result.
24	Those of us of a certain age know what an
25	end-to-end monopolist in a communication space looks like.

159

1	It was the old totally vertically integrated telephone
2	company. One company controlled all of the equipment, all
3	of the connections, all of the interfaces. Indeed,
4	everything from the chips to the telephone repairman.
5	It wasn't simply that they had a lock on the
6	industry. They, not competition, decided what innovations
7	made their way to the consumer and when. That slowed down
8	the transfer of innovation, and as a consequence,
9	telecommunications innovation in this country was outpaced
10	by that in others.
11	In an increasingly competitive global economy,
12	we cannot afford to return to those days. And the
13	antitrust laws governing single-firm conduct were the
14	means by which that situation was remedied.
15	Today different technologies from different
16	companies come together to create a plethora of consumer
17	products, which we all enjoy and to a substantial extent
18	take for granted. This creates an ongoing challenge in
19	defining how those technologies will interconnect and
20	interoperate and the rules that will apply to that
21	endeavor.
22	Even the best technology is of little use in
23	isolation. The antitrust laws have an important role in
24	policing the conduct of firms who would seek to take
25	control of those interconnections so as to eliminate

160

1	competition and thus harm consumers.
2	In my remaining remarks today, I will focus on
3	two areas of concern which, in Broadcom's experience, are
4	particularly important to preserving competition.
5	The first is standard setting. I know there was
6	a fair amount of discussion on that this morning. There
7	will be more of it this afternoon. The second is the use
8	of proprietary interfaces from one market to another.
9	These are not theoretical issues. These are
10	real issues that Broadcom has faced in the past and
11	continues to face today.
12	We come at this from the perspective of a highly
13	innovative company with world-class technology, attempting
14	to break into new markets dominated by entrenched rivals.
15	At the same time, we are an example of a company
16	that has thrived through key contributions to important
17	industry standards and, today, without charging royalties
18	for those innovations.
19	Standard setting refers to the process of
20	creating and implementing a way of doing things. As a
21	simple example known to all of us, there's the standard
22	format for video known as VHS. That standard makes it
23	possible for a variety of competing manufacturers to make
24	the various components that are needed to record and play
25	home video: the camera, the tape, the VCR, and so forth.

161

1	Similar standards exist for CDs, DVDs, as well as
2	standards that allow voice video data and multimedia to be
3	shared among various wired and wireless devices.
4	In addition to facilitating competition by
5	enabling different companies to produce products that will
6	interconnect and interoperate, standard setting, when done
7	properly, can also resolve intellectual property rights or
8	IPR issues that might otherwise impede progress.
9	With the complexity of today's products, often
10	multiple parties own IPR that is needed to implement a
11	particular technology-based application. If Company A
12	owns essential IPR and so do Companies B, C, D and E, each
13	can block the other and everyone else from making a
14	product using the best available technical solutions.
15	In the standard setting process, companies
16	typically are required to agree that they will disclose
17	their IP rights that are essential to practice this
18	standard before the standard is adopted. This gives the
19	standard setting body and the participants in the standard
20	setting process the ability to avoid such IPR or to
21	address the means by which that IPR will be licensed to
22	those who practice the standard.
23	I will get to licensing in a minute, but first a
24	word on IPR disclosure in standards making.
25	There are those who say that disclosure is not a

162

1	significant problem because companies generally play by
2	the disclosure rules. They say that failure to disclose
3	is rare and therefore not really a problem. At Broadcom,
4	we aren't sure whether failure to disclose is in fact rare
5	in all standard setting bodies. But even if that is the
6	case, it can still be a serious problem.
7	Indeed, the fact that participants in standard
8	setting expect disclosure and rely upon it makes those
9	instances of failure to disclose all the more problematic.
10	Without disclosure, the standard is at constant risk of
11	being hijacked by an IPR holder that has hidden in the
12	weeds during the development of the standard or, even
13	worse, has helped steer development toward its own
14	undisclosed proprietary technology only to spring its trap
15	after the standard has been set and millions or even
16	billions of dollars have been invested in its
17	implementation.
18	This risk is not an abstract or a theoretical
19	concern. In fact, these hearings are particularly timely.
20	Just this past Friday, four days ago, the jury in San
21	Diego rejected an attack on my own company by a firm
22	attempting to force us out of certain technology spaces by
23	asserting two patents that it controlled. Its
24	infringement case was based in substantial part on our
25	implementation of an industry standard for video

163

1	compression. The jury found no infringement, thank god.
2	And, perhaps more significantly, also found that our
3	adversary had violated the disclosure rules of the
4	standard setting body by failing to disclose its patents
5	which allegedly covered the standard.
6	Sadly, the company that launched this
7	ill-founded patent assault on an international standard,
8	cynically justified its actions afterwards on the grounds
9	that it had nothing to lose, even though after a nine-day
10	trial, a jury unanimously agreed that the company had used
11	the standards process and had also violated its duty of
12	honesty and fair dealing with the U.S. Patent and
13	Trademark Office.
14	Meanwhile, defending itself against those
15	illegitimate claims cost Broadcom millions of dollars.
16	And the lawsuit created confusion and concern among our
17	customers and the many others who use the H.264 video
18	compression technology.
19	So, this is a very real risk. If an
20	opportunistic company can get away with these tactics, it
21	would be in a position to dominate components for an
22	important ubiquitous video compression technology by
23	asserting its patents against all would be competitors.
24	But disclosure, important as it is, is not
25	enough. Disclosure is only the first step in assuring

164

1	that hijacking will not occur. Disclosure merely allows
2	the standards development body to thwart attempts to
3	insert proprietary technology into the standard.
4	It is at least equally important for industry
5	participants to abide by the rules after the standard is
6	in practice, is in place. A key element of that is
7	licensing terms and conditions.
8	The rules of standards bodies typically provide
9	that IPR that is essential to practice the standard will not
10	be included in the standard unless the owner agrees to
11	license that IPR to those who wish to practice the
12	standard on either a royalty free or fair reasonable and
13	nondiscriminatory, so-called FRAND, sometimes called RAND,
14	terms.
15	What happens when someone fails to live up to
16	these commitments? As I noted, once a standard is set,
17	the industry moves forward and invests millions if not
18	billions of dollars in implementing the standard. That
19	investment is based on the understanding and assumption
20	that IPR issues are resolved. Either there will be no
21	need to take a license to the IPR, or any licensing will
22	be on FRAND terms.
23	If a company with essential IPR seeks to impose
24	non-FRAND licenses, the balance is completely upset.
25	Suddenly the industry which adopted the standard with the

165

1	understanding that licensing costs would be reasonable, is
2	confronted with a monopolist seeking to charge monopoly
3	rates.
4	In industries that are involved in standard
5	setting, there are certain practices that I would venture
6	to say everyone understands are not FRAND terms. For
7	starters, refusing to license at all violates a FRAND
8	commitment. Amazingly, there are some in the industry who
9	take the position that, notwithstanding their commitment
10	to license all who wish to practice the standard,
11	essential IPR holders can pick and choose among potential
12	licensees for any reason, including, it would seem,
13	whether the potential licensee is a downstream competitor
14	Another example: Broadcom has been confronted
15	by a licensor who participated in the standard setting
16	process, insisting that, as a condition to being granted a
17	license to the intellectual property essential to practice
18	the standard, it would have to give back a royalty-free
19	license to a much broader sweep of Broadcom's own
20	intellectual property, including IP-covered features and
21	functions entirely unrelated to the standard.
22	To usurp the blood, sweat, tears and genius of
23	interface companies in such a manner as a condition to
24	practicing an industry standard runs directly contrary to
25	the fundamental objectives of standard setting bodies.

166

1	If this sort of practice is allowed, what
2	incentive will any company have to innovate or invest,
3	knowing that unrelated technology can be appropriated as
4	the price for making standardized products.
5	Another example that we have seen is a company
6	attempting to use access to essential IPR to coerce
7	customers into buying its products, rather than letting
8	the merits of the products determine who gets the sale.
9	And we have examples where a company has thought
10	to stack a standard setting organization with supposedly
11	independent voters to skew the standard towards it own
12	technology or away from the technology of its rivals.
13	To be clear, I do not suggest that a company
14	should be required to share its technology with others.
15	Far from it. Patents are available to protect innovation
16	and Broadcom is a firm believer in the patent system.
17	But it is imperative that, when a company has
18	made a commitment to license on FRAND terms as a condition
19	of getting its technology included in a standard, it must
20	not then be allowed to exploit the market position it
21	gained through incorporation in its IPR and the standard,
22	by reneging on that commitment.
23	And a company, likewise, should not be allowed
24	to subvert the rules that are put into place to ensure
25	that standard setting is a nonpartisan exercise.

167

1	These are very real and contemporaneous examples
2	of the kind of anticompetitive single-firm conduct we at
3	Broadcom believe the antitrust laws are intended to
4	address.
5	Some say that determining what is fair and
6	reasonable is too hard a task. That is a standard that
7	cannot be enforced. We heard some discussion along those
8	lines this morning.
9	Often the firms that say this are the very firms
10	that fail to disclose their patents, have engaged in
11	rampant discrimination that cannot possibly be reconciled
12	with a FRAND obligation, and have engaged in other
13	behavior that demonstrates that it is a lack of will, not
14	a lack of ability, that has resulted in their FRAND
15	violations.
16	Fair and reasonable simply means that the
17	technology will be available on competitive terms, rather
18	than on terms that reflect a market power gain through
19	inclusion of technology in the standard.
20	It also means that no participant will charge a
21	disproportionately high royalty so as to hobble the
22	standard or render it uncompetitive.
23	Technology companies are often engaged in patent
24	litigation where a question before the court is how to
25	assess a reasonable royalty in damages. There's no reason

168

1	to believe that the courts would have a harder time
2	figuring out what reasonable royalty is in the standards
3	context than in any other context. The court can take due
4	account of the competitive goal of the standard setting
5	body in requiring a FRAND commitment up front, and
6	otherwise undertake the same exercise it goes through when
7	evaluating damages and so forth.
8	It has also been suggested that failure to
9	comply with a FRAND obligation is a matter better left to
10	contract than antitrust law. One might ask, if a court
11	applying contract law can figure out what FRAND means, why
12	can't the same court apply antitrust law?
13	Contract law is a private remedy to redress
14	private rights. FRAND violations can eliminate
15	competition and hurt consumers, competitors, innovation
16	and the economy as a whole. Isn't preventing such an
17	injury exactly what the antitrust regime is all about?
18	Moreover, if companies are willing to break
19	their commitment because they conclude they have little or
20	nothing to lose by doing so, the contract remedy is
21	inherently insufficient to protect innovation, competition
22	and consumers. And that becomes the job of antitrust
23	law.
24	The second area I would like to talk about is
25	interfaces. As I noted before, interfaces are the way one

169

1	piece of technology connects to another. By manipulating
2	the interface and making it proprietary, a company with a
3	monopoly over one area of technology can effectively shut
4	out competitors and technology that would connect with the
5	monopoly technology.
6	For example, if a company had a monopoly in
7	amplifiers, it could obtain a monopoly in speakers by
8	creating a proprietary amplifier-to-speaker interface and
9	refusing to license that interface to anyone. The speaker
10	market, which previously enjoyed vigorous competition that
11	fostered innovation and lower prices, would suddenly be
12	controlled by one firm with little incentive to innovate
13	or reduce prices.
14	We've seen this in practice. Broadcom is a
15	communications chip company. Our chips connect devices
16	and systems. We've seen, for example, companies that
17	control the main processor of a particular system, one
18	that was at one time characterized by an open interface,
19	suddenly making that interface proprietary. For no good
20	technological reason, they make it harder to interconnect
21	with that chip, while at the same time launching their own
22	communications chips that competes with Broadcom and
23	others.
24	This two-prong strategy, control the connection
25	with the dominant product and compete in the adjoining

170

1	market, has a predictable result. The dominant firm
2	leverages its monopoly from one area outward into ever
3	greater areas
4	Over time, the dominant firm expands its empire
5	to the entire motherboard, destroying its competitors and
6	the innovation they would bring along the way. There
7	certainly are instances where the development of new
8	interfaces is real innovation.
9	Where there is real innovation in the interface,
10	innovators should have the opportunity to be appropriately
11	compensated. But that compensation should at best take
12	the form of a modest, truly nondiscriminatory royalty. It
13	should not be a vehicle for extending dominance from one
14	kind of chip to another by, for example, the kind of
15	asymmetrical brand back of IPR from the licensee to the
16	licensor that I discussed earlier.
17	And a small improvement in interface technology
18	should not come at the sacrifice of innovations of orders
19	of magnitude more significant in the adjacent
20	communications markets if innovators' chips can no longer
21	communicate with the now closed interface.
22	Of course sometimes the new interface does not
23	even represent an improvement, just a difference. When a
24	company has a history of using open interfaces or of
25	licensing its interfaces to third parties and then stops

171

1	doing so, while at the same time entering the market on
2	the other side of the interface, one ought to become
3	suspicious.
4	We've experienced that in our industry. Again,
5	there is a role for antitrust when such changes provide
6	little or no benefit but substantially hurt innovation and
7	therefore consumers and the economy as a whole.
8	I recognize that today I barely scratched the
9	surface of the issues that I talked about. And of course
10	much depends on the individual facts and circumstances of
11	any particular case and market.
12	That said, the antitrust laws and the courts and
13	agencies that are called upon to enforce them should not
14	shy away. Usually, once the facts are separated from the
15	noise, it is not difficult to separate the procompetitive
16	stories from the anticompetitive ones, particularly in the
17	area of deceptive conduct in standard setting processes
18	there is little risk that procompetitive behavior will be
19	deterred.
20	In closing, I hope the FTC and DOJ and those who
21	are thinking seriously about antitrust in the 21st Century
22	will take away from my remarks three basic concepts
23	First, antitrust, as it relates to single-firm
24	conduct, remains important to ensuring competition in our
25	high technology markets.

172

1	Second, we have seen in recent years the
2	creation and abuse of monopoly positions through conduct
3	that serves no useful purpose and therefore should be
4	counteracted by the antitrust laws.
5	Third, the antitrust laws must remain flexible
6	and responsive to these ever-changing conditions. Blind
7	reliance on outmoded principles, and even more
8	importantly, a refusal to consider the particular facts of
9	a particular case is a terrible mistake that the courts
10	and the agencies should not make.
11	I thank the FTC and Department of Justice for
12	the opportunity to speak today and for your thoughtful
13	consideration of these important issues.
14	(Applause.)
15	MS. GRIMM: Thank you very much.
16	We'll now take a 15-minute break and we'll
17	reconvene here then for the round-table discussion. Thank
18	you.
19	(A brief recess was taken.)
20	MS. GRIMM: I'd like to start this portion of
21	our program by asking our two panelists if they would like
22	to comment in any way on each other's presentations and
23	respond to any questions between them
24	Would either of you like to comment or ask any
25	questions?

173

1	MR. McCOY: I think I'm going to pass. I think
2	I'm here to answer your questions
3	MS. GRIMM: Okay. What we're going to do is
4	very similar to what we did this morning. We're going to
5	ask some general questions, then we're going to ask some
6	specific questions on predatory buying that Michael will
7	answer and some questions on loyalty discounts that we'll
8	talk about with you, Tom
9	MR. McCOY: Great.
10	MS. GRIMM: So, to begin, we have heard a lot
11	this morning about the lack of uniform standards among and
12	between antitrust enforcement agencies throughout the
13	world. And AMD operates globally, clearly. I believe
14	that you filed a complaint against Intel in Japan, Korea,
15	the EC, and of course the case in District Court in this
16	country.
17	Could you please address the question of
18	standards, whether they are different globally, and also
19	tell us if it does cause a problem for AMD or whether it
20	is not a problem?
21	MR. McCOY: I'd be glad to.
22	We did not file a complaint in Korea --
23	MS. GRIMM: Oh, I'm sorry.
24	MR. McCOY: In fact, we found out about the
25	investigation of Korea in Intel disclosures, so ... But,

174

1	more generally, it's a very interesting time, I think, for
2	Commission authorities around the world, particularly as
3	the world has globalized and the markets are global. And
4	AMD and Intel, for example we are the only two suppliers
5	of X86 processors for the world. The whole world is
6	dependent on us and probably eighty percent of IP runs on
7	X86. And I think more and more we're seeing business
8	conditions like that.
9	My experience is that there is a great
10	opportunity. It shouldn't be viewed as a difficult
11	problem, as Judge Posner has posited in some of his
12	remarks. I think it's an opportunity for the mature
13	competition authorities around the world to establish
14	their common ground.
15	And, in my experience, there is tremendous
16	common ground that I don't see really any outlines out
17	there when it comes to the valuation of unilateral conduct
18	by dominant companies. I see an effort to come together
19	on guiding principles as to what the desired results
20	competition policy are.
21	It's about competition, not about competitors.
22	It's about innovation. It's about competitiveness. You
23	can't have competitiveness without competition. It's
24	about consumer value and consumer choice. It's about a
25	very thoughtful look at barriers to entry and their

175

1	permanency relative to assumptions about their transient
2	nature. And it's about looking for behavior that makes no
3	sense for a long period of time. So, where rational
4	business people are making irrational decisions that
5	suggests that there is a persistent problem.
6	And, in my experience around the world in
7	today's agencies, there is tremendous interaction between
8	those people involved in policy, those people involved in
9	economics, and those people involved in advocacy, in
10	trying to bring together guiding principles where we can
11	all agree that the values of antitrust enforcement have
12	been historically used in this country in terms of
13	promoting efficiency and consumer welfare are far more
14	common.
15	MS. GRIMM: So, just following up on that, you
16	really don't perceive it as a problem. Is that
17	overstating it?
18	MR. McCOY: I have not seen a problem and I have
19	not seen -- and, I'll be honest, in the AMD and Intel, you
20	know, fronts around the world, I don't see a big set of
21	differences in the way that people are looking at this.
22	We may get into a little bit more of that when
23	we look at retrospective rebates. But in terms of what's
24	the appropriate focus, you know, what's happening to the
25	innovation process, what are the barriers to entry, why do

176

1	we have persistent behavior that is out of character for
2	people who are smart business people, why has it endured
3	so long, and what are the effects on the innovation
4	process and the effect on consumers, I think everybody is
5	asking the same questions,.
6	MS. GRIMM: Let's follow up with the loyalty
7	discount and just take rebates and loyalty discounts as
8	one type of conduct that we're looking at.
9	Is there any difference in the standards that
10	you perceive that are being applied in different
11	jurisdictions as to that particular subject?
12	MR. McCOY: I believe that in my experience, and
13	let me make it clear, I don't pretend to be the latest,
14	you know, gift to antitrust academics, but I have been
15	around the block in my career on all these issues.
16	I think the law is pretty settled and policy is
17	pretty settled every where in the world but here in the
18	United States about retrospective rebates. And I think
19	one has to be careful to take a hard look at what really
20	happens in a marketplace, beware of labels. Because we
21	can all agree that price competition is a good thing. And
22	we can all agree, generally speaking, that a discount is a
23	good thing.
24	But a retrospective discount or rebate, and I
25	use those words in quotes, is usually, when deployed by a

177

1	monopolist, not a rebate or discount at all. It's a price
2	coupled with a threat of a price increase it can go to
3	here in demands for market share and monopoly margin.
4	So, there's simply a device, a mechanism, to
5	impose a penalty on capital customers from erring to try
6	to balance out their suppliers.
7	MS. GRIMM: This morning I believe we heard that
8	with respect to discounts there really is no standard
9	that's generally accepted even in this country.
10	Do you agree with that or not?
11	MR. McCOY: I think that (a) the way that most
12	jurisdictions look at this is in terms of exclusion.
13	What's really happening is a matter of fact.
14	What is really happening, which requires a look at
15	relative market share. But I believe that most of the
16	world looks at it in terms of exclusion.
17	In this country, I think the debate is very
18	confused and there are a lot of discussions about words
19	and concepts, but they tend to be -- discussions tend to
20	be somewhat divorced from what really happens in the
21	marketplace, in my experience.
22	So, I don't think we have a settled view on when
23	and if a dominant firm should be permitted to use a
24	retrospective rebate. And I think the debate in the U.S.
25	is far behind some of the more closed debates and

178

1	jurisprudence of other jurisdictions, where they've had a
2	lot of experience in looking at them and actually coming
3	to decisions and enforcement actions. They're coming up
4	with remedies.
5	MS. GRIMM: Let me follow up on that also.
6	What remedies are they coming up with with
7	respect to discounts that are found to be illegal?
8	MR. McCOY: Well, I encourage everybody to
9	actually look at what they do rather than rely on me. As
10	I said, I don't pretend to be a professor.
11	But they're fairly clear remedies in the other
12	jurisdictions about preventing quantity-forcing
13	contractual terms.
14	And, in fact, as I observed in my opening
15	prepared remarks, we have a very clear example coming out
16	of the Microsoft case, where you have a quantity-forcing
17	term that Microsoft had imposed on the world, which is
18	basically you're selling a computer, you're going to pay a
19	royalty to us whether you are selling that computer with
20	an operating system or not.
21	And everybody agreed that was clearly above the
22	line as a quantity-forcing predatory contractual term.
23	And there's no reason why in and out of this context we
24	can't figure out appropriate, clear and fair remedies here
25	as they have elsewhere.

179

1	MS. GRIMM: In your view, are DOJ and the FTC
2	failing to challenge single-firm conduct that they should
3	be challenging? And, if so, what types of conduct?
4	MR. McCOY: Well, I think that we are in a
5	period of having a very healthy and appropriate debate
6	about when there should be regulatory intervention into
7	managed markets where the management is as a result of the
8	unilateral conduct of the dominant firm.
9	And, particularly in a world that is changing
10	rapidly and globalizing, it's very -- I think it's very
11	appropriate to step back and take a look at -- a fresh
12	look at the policy objectives that underlie antitrust law
13	and policy and enforcement, and whether the tools, the
14	analytical tools, are the right tools, whether the right
15	facts are being evaluated, the right priorities being set,
16	and whether enforcement is appropriate and effective.
17	And that is likewise appropriate that that be a
18	global debate. As I said, it shouldn't be viewed as a
19	problem or a burden. I think it should be viewed as an
20	opportunity for competition authorities around the world,
21	particularly in mature jurisdictions and marketplaces to
22	try to find as much common ground as possible, and I
23	believe it can be done. In fact, progress has probably
24	been made.
25

180

1	appears, frankly, that there has been a retreat from
2	Section 2 enforcement, and that not getting the same kind
3	of energetic investigation and enforcement of Section 2 in
4	unilateral conduct, which to me is surprising when we look
5	at the continued investment of resources appropriately.
6	MS. GRIMM: Mike, are you there?
7	MR. HAGLUND: Yes, I'm here.
8	MS. GRIMM: May I ask you the same question?
9	Are the FTC and the DOJ failing to challenge
10	single-firm conduct that they should be challenging? We
11	know about predatory buying. Are there any other forms of
12	conduct that you encountered in counseling your small- to
13	medium-sized clients that we should know about?
14	MR. HAGLUND: Well, I think that there is a --
15	what I've observed in the last five, ten years is a shift,
16	I think, in emphasis at the national levels by the Federal
17	antitrust agencies to having a greater concern with
18	national markets and international markets. And I think
19	that with that -- and some of that is understandable.
20	Some of it I think is a mistake because I think
21	that when one really drills down into some of these lower
22	tech industries that I've been involved in, you find real
23	regionalization and relevant distinct markets that meet
24	the test of that term for purposes of antitrust law and
25	can be significantly hurt in terms of their competitive

181

1	health unless there's significant enforcement of the
2	antitrust laws.
3	And I think that more energy needs to go into
4	knowing the facts of those local and regional markets
5	because the smalls tend not to be able to watch out for
6	themselves because of the level of antitrust expertise out
7	there generally. And I think that the states vary widely
8	in terms of the level of commitment they have to antitrust.
9	So, I think there's more to be in that sector.
10	MR. McCOY: Can I make a positive comment?
11	To give you an example of what the technology
12	industry would view as a very, very good signal. The
13	Federal Trade Commission has obviously invested an
14	incredible amount of time and resources into the Rambus
15	situation. And I am not carrying a brief on either side
16	of those issues, but those issues are very important.
17	They're very important to innovation and
18	competitiveness. They're very important to market entry.
19	And they're very timely. Market standards are a very good
20	thing from the consumer welfare perspective. They drive
21	scale and they drive the entrepreneurial opportunity.
22	And I think that we have a lot of evidence now
23	to evaluate how standards are a very, very positive thing.
24	They drive competitors and innovation, and therefore, the
25	integrity of the standardization process is something that

182

1	should be really looked at very carefully. And when there
2	is not integrity in that process, the world needs to know
3	that there is going to be enforcement.
4	However the Rambus case ultimately comes out, I
5	think the Federal Trade Commission sends a very
6	appropriate signal to the marketplace that this is
7	important and it's strategic, and it's quite clear that
8	there is going to be some behavior that is simply not
9	going to be tolerated.
10	MS. GRIMM: Let me kind of reverse the question
11	and ask the opposite.
12	Based on your experience, are there certain
13	types of conduct that are benign or procompetitive,
14	deserving of more lenient treatment than they are
15	currently afforded?
16	Either one.
17	MR. HAGLUND: I guess I come at it from the
18	standpoint of looking at the forms of anticompetitive
19	conduct being able to take many, many different shapes.
20	One of the interesting things I heard in Tom's
21	talk was his reference to the potential that a mix of acts
22	can work very effectively for a dominant firm. In the
23	Weyerhaeuser case, for example, we had 15 different types
24	of anticompetitive conduct, but all the attention has been
25	showered on predatory buying, but in fact the table was

183

1	set for the price-raising behavior in the log market by
2	exclusive contracts, by a number of other anticompetitive
3	tactics that worked together in combination to become
4	effective overall.
5	But I guess I'm not able to identify conduct
6	that should be benign, other than that I do see some of
7	the rationale for why Brooke Group was decided wanting
8	to immunize price cutting with the price cost test in
9	terms of not trying to hinder or chill price cutting
10	conduct.
11	But where it's beyond that, I have trouble -- my
12	experience doesn't reveal areas where I think there's too
13	much attention or it shouldn't be used.
14	MR. McCOY: Well, I have been practicing law and
15	business for over thirty years now and been through many
16	different seasons of policy views and the relative
17	oversight by competition authorities.
18	And I guess I would say this: In my career, I
19	have never seen a company hold back. I mean, it's a
20	hardball world out there and I've not seen a client in the
21	days I was a law partner or certainly at AMD where
22	businesses were pulling punches because of worry about the
23	activity. So, that's number one.
24	Number two, depending on what side of the bar
25	you sit on, in any particular matter, you always have one

184

1	side that wants to disaggregate all the behavior and just
2	look at everything piecemeal. But the reality, the
3	reality of life in the business world, is that there is a
4	tapestry of activities. That's just the way the world
5	works.
6	And one really does have to be careful of trying
7	to judge the beauty of the picture by just looking at the
8	eye or the ear or the nose. You really have to look at
9	the whole thing.
10	And, finally, I think that the challenge is
11	always going to be pretty much the same because, if a
12	company is fortunate enough to have a dominant position,
13	however they got there -- let's assume they got there
14	through skill -- and they're now enjoying a big market
15	capitalization of software, they're going to do everything
16	that they can to protect that market place. And that's
17	what they're going to do.
18	And, therefore, there's always going to be, in
19	my view, need for a strong antitrust policy articulation,
20	communication and enforcement, because otherwise you're
21	going to end up with cultures, business cultures, that
22	their compliance programs are not going to be able to keep
23	under control.
24	MS. GRIMM: I'd like to turn to a little
25	different subject now.

185

1	As you may know, antitrust lawyers and judges
2	are battling -- I guess that's too strong a word -- but
3	how much weight do you give to business documents
4	containing evidence of bad predatory intent? What
5	consideration in your view should the antitrust enforcers
6	give to intent documents in assessing a firm's conduct?
7	MR. HAGLUND: Well, I think you hear two schools
8	of thought on this. One is that, oh, every good business
9	wants to kill its competition, that's just the way of the
10	world in terms of being a good competitor. You hear
11	experts talk about juries getting too carried away about
12	statements that they think are just characterizations of a
13	robust effort to compete hard.
14	And I think you need to distinguish between
15	cheerleader-type phraseology that somebody might use in an
16	e-mail, which I don't find to be terribly meaningful, and
17	the documents that really help demonstrate what the intent
18	is relative to a particular business practice and its
19	ultimate effect on the structure in the industry.
20	And where the documents really -- where I find
21	intent helpful, and I think this is where the court in
22	Microsoft and a number of Supreme Court cases have said in
23	"Aspen," for example, and "Trinko," what's important,
24	intent can help give one a means of interpreting what are
25	otherwise ambiguous acts and give you a more firm and

186

1	clear view of what the defendant really intended. And
2	especially if they speak to the structure and the change
3	they wish to achieve in the industry. And if they're
4	already above the fifty percent mark, then I think it's
5	very helpful stuff.
6	MS. GRIMM: Tom, any views?
7	MR. McCOY: Well, I think that government
8	officials involved in antitrust enforcement should look at
9	everything. But I think everybody agrees that the
10	documents that a trial lawyer would love on the
11	plaintiff's side have to be looked at objectively and in
12	context. That of course a dominant company is going to
13	try to preserve that dominant position. That's what
14	they're going to do. That's what they're paid to do.
15	That's what their shareholders expect them to do.
16	So, documents that manifest that obvious
17	reality, so what.
18	But I think that it's important, you know, in
19	being a fact-finder, being a dispassionate fact-finder and
20	evaluating, you know, the purpose of a strategy and
21	whether the advocates are credible or not in trying to
22	defend whether the strategy is being pursued for
23	reasons that really relate to growing a market, satisfying
24	a customer, being creative and innovative in products and
25	marketing, or whether it's simply a design, and a heavily

187

1	lawyer design, for a monopoly to use their power to
2	preserve a monopoly.
3	One needs to look at what people say about what
4	it is they're doing, particularly trying to get a hold of
5	the evidence that matches up externally as to what is the
6	marketplace perceiving as to why the dominant company is
7	doing what it is doing.
8	And I think it is the unity of the evidence on
9	those boundaries that can be generally fairly helpful
10	figuring out whether it's just straight forward hardball
11	business or whether it's a monopoly simply trying to
12	protect its position using their power.
13	MS. GRIMM: Thank you. I think you're pretty
14	much in agreement on that question.
15	MR. McCOY: And I believe, by the way, that that
16	is the view of most of the people in the other
17	jurisdictions in terms of when they're looking at
18	evidence. I think your colleagues and sister agencies
19	from around the world all say, look, if we get a document
20	from a lower-level sales employee that says, you know,
21	we're going to go kill those guys, that we would take that
22	document with somewhat of a grain of salt. That, standing
23	alone, doesn't tell us any about structure, about
24	efficiency, and certainly about what's happening in the
25	industry.

188

1	MS. GRIMM: When we were doing some background
2	research, Google research for this panel, we came across a
3	recent article in "Fortune," August of 2006, that quoted you.
4	And it quoted you as saying, "As a matter of economics,
5	the monopolies probably begin somewhere between thirty
6	percent and thirty-five percent," and it then goes on to
7	explain that at this point a rival's rising market share
8	would imperil a dominant firm's hold on a market. You were
9	talking about Intel in this article.
10	Do you have any experience in suggesting that
11	attaining any particular market share, whether it's thirty
12	or thirty-five percent or whatever, has particular
13	significance for competition against a large competitor?
14	MR. McCOY: Well, my comments were in the
15	context of the X86 processor market where Intel has, for
16	more years than I can count, enjoyed a revenue share of at
17	least eighty percent, and there's really no other rival,
18	but that which typically had a revenue share of somewhere
19	in the ten to fifteen percent range.
20	And so in order to think about specific points
21	where monopoly power begins to erode, you need a lot of
22	context, you need to know where the companies are starting
23	from, and you need to know a lot about the various
24	entries, and you need to know a lot about what is the
25	psychology of the marketplace. Because one of the things

189

1	that gets missed in the academic debates is that markets
2	are comprised of real people making human decisions. And
3	so, that psychological, you know, culture of the market
4	explanation has been patterned by monopoly behavior.
5	My comments are taking a look at where we are
6	and where the competitor is and the penalties that are
7	imposed or that have been imposed on customers for
8	incremental market share provided to us, and where we
9	would have to be as a revenue share before we could
10	overcome those kinds of penalties.
11	And one of the examples that I talked about in
12	the prepared remarks is that, in a situation where you go
13	to a very big and powerful company and you say, we're
14	going to give you a million units for free, units where
15	probably your average procurement cost is running at least
16	$150.00, we're going to give you a million of them free.
17	And they can't be used, they can't be used because the
18	penalty, the retaliatory penalty that is imposed for not
19	maintaining market share margin of the incumbent, tells
20	you something about you got a ways to go as a matter of,
21	quote, economic -- economics. Capital markets and
22	psychology you can amass what you need to overcome the
23	barriers that have been erected that you have to get over,
24	particularly in markets where only a small slice of it is
25	contestable in any relatively short term or intermediate

190

1	period.
2	In some markets, a company could wake up on
3	Friday and say, on Monday I'm going to buy twenty percent
4	more of my needs from a different company. But that's not
5	true in technology.
6	In technology, there is -- a lot of switching
7	costs takes time. It can't be done quickly. And,
8	therefore, getting a relevant market share to be able to
9	overcome the power of the tendency is difficult.
10	MS. GRIMM: Let me follow up with just one
11	further question.
12	With respect to loyalty discounts and rebates,
13	does market share provide any kind of useful screening
14	mechanism that we could use for assessing legality?
15	MR. McCOY: Well, yes. But, again, I believe
16	you have to look at market share and I think you have to
17	look at entry, and you have to have in mind the relative
18	margins of a monopoly supplier and the customer base.
19	So, you can have a situation, as we do in
20	technology, where you have an ingredient supplier with
21	margins that are -- operating margins in the forty percent
22	range, serving customers whose operating margins are
23	in the zero to six percent range. And they're public
24	companies, with people who are trying to manage
25	shareholder expectation, capital market expectations,

191

1	employee morale, and their tenures, with a board of
2	directors looking over them.
3	So, I don't think there are any bright lines
4	here. I know everybody wants a bright line and everybody
5	wants to talk about safe harbors. But in the real world,
6	there are a number of factors that I think is a matter of
7	making sure that you're doing the right thing in the right
8	market at the right time.
9	The unfortunate reality is, from a resources
10	vendor standpoint, that a fair amount of homework should
11	be done. But certainly in marketplaces where you have an
12	enduring monopoly that is enjoying fifty, sixty or more
13	percent of the revenue share, that tells you, frankly any
14	time you have a dominant company using a retrospective
15	rebate, it's -- in my experience, the odds are one hundred
16	percent that a retrospective rebate is being used for no
17	other reasons.
18	MS. GRIMM: Mike, I'd like to ask you one more
19	question on predatory pricing, then we're getting pretty
20	close to closing the session.
21	You've practiced, as you pointed out, for many
22	years representing small- and mid-sized resource
23	companies.
24	Is the issue of predatory buying, the type of
25	conduct that we saw in Ross-Simmons, is it rare, or is it

192

1	more common practice than the case law might reflect?
2	MR. HAGLUND: I think it's fairly rare. And it
3	happens only, from what I've seen in these markets -- at
4	least in the resource sector, in markets where the supply
5	of the inputs is fairly elastic and -- I mean, alder, for
6	example, doesn't get harvested except as a byproduct of
7	the much larger softwood harvest in the Pacific Northwest.
8	Fish stocks, for example, that are so rigidly regulated.
9	Those are the kinds of markets where a really predatory
10	dominant buyer can eliminate its processor or sawmill or
11	other competitors.
12	But, in looking at the case law, there are a
13	very, very few number of cases. And in my own experience,
14	there are so many resource markets, you don't see any
15	evidence of it.
16	So, in the big picture of things, it is a
17	relatively rare situation.
18	MS. GRIMM: Joe, would you like to close with
19	any questions that you might have of our panelists?
20	MR. MATELIS: Sure, I'll ask one.
21	I guess this is primarily for Tom, although I'd
22	be interested in Mike's thoughts.
23	One of our panelist at the morning session
24	talked about, in view of the emerging overlapping
25	international enforcement that's taking place, what he

193

1	termed a principle of comity and, in general, it's the
2	notion that there ought to be principles where one
3	enforcement agency presumptively takes the lead on a
4	certain matter. He proposed home jurisdiction and there
5	had been other proposals.
6	I'd be interested in your thoughts on the
7	potential problem of overlapping enforcement across
8	countries.
9	MR. McCOY: Well, as I said, the issue of
10	harmonization across the borders in the competition
11	network, I think that's very important.
12	I think that particular proposal is absurd. If
13	you were to apply that proposal, particularly with any
14	view of the way the world is going to look to AMD and
15	Intel, you would conclude that the dispute should be
16	resolved in the states.
17	And, the fact of the matter is, for AMD and
18	Intel, if you were to take -- our revenues are probably
19	seventy-five percent coming from outside the U.S. We are
20	-- big multinational companies are citizens of the world.
21	We have productive capacity all over the world. We have
22	employees all over the world. The innovation process is
23	one that is built on human resources located around the
24	world, in no particular jurisdiction. And the marketplaces
25	are global.

194

1	So, to look at where a company is chartered or
2	where the CEO sits is not a relevant variable to determine
3	competition policy.
4	MR. MATELIS: Just to press you a little bit on
5	that: Even if we don't like that specific proposal, is
6	overlapping enforcement from different countries something
7	that we ought to be worried about or a healthy thing?
8	MR. McCOY: Well, I think that I'll be -- I
9	think the competition authorities should compete, just to
10	throw out a radical thought.
11	MS. GRIMM: We heard that [laughter].
12	MR. McCOY: No, I'm serious, that there should
13	be intellectual competition. And that's the free flow of
14	ideas, just like free trade in IP. Nobody has a monopoly
15	on these ideas.
16	But be careful when you talk about who ought to
17	take the lead. I don't think it's ever going to, in the
18	practical world, occur, because in a globalized world,
19	what a dominant company does in any particular
20	jurisdiction affects all the other jurisdictions. So, for
21	example, I think one of the reasons why Europe became so
22	active in the Intel investigation after Japan is because
23	it was so clear that the behavior that was judged to be a
24	violation of the antimonopoly laws and the public policies
25	in Japan had a direct effect on consumers in Europe.

195

1	So, when you have these -- when you have a more
2	globalized world where the dominance, you know, extends
3	globally, behavior anywhere can affect consumers
4	everywhere. And in those scenarios, I just don't
5	think it's -- one has to be practical, including
6	politically practical. To think that any jurisdiction is
7	going to advocate or forebear the protection of its own
8	consumers in favor of another jurisdiction, that would be
9	a remarkable thing. And I just don't think it's healthy.
10	MR. HAGLUND: I'd agree with Tom.
11	MS. GRIMM: And on that note, it is a little
12	past 4:30, I believe. Yes.
13	I again want to thank our panelists for
14	participating in our hearings today. I'd like everyone to
15	please join may in a round of applause for them.
16	(Applause.)
17	MS. GRIMM: I'd also add you're all invited to a
18	reception following this hearing. It will be at the
19	Woman's Faculty Club over here
20	You're also invited to join us tomorrow. We're
21	going to have a number of very distinguished faculty
22	members from both Berkeley and Stanford. The session in
23	the morning will be from 9:30 to noon, and the afternoon
24	session will be from 1:30 to 4:30.
25	Thank you all for attending. I think our

196

1	panelists did a remarkable job. Thank you.
2	(Applause.)
3	(Whereupon, at 4:35 p.m., the hearing was
4	concluded.)
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25

197

1

C E R T I F I C A T I O N O F R E P O R T E R

2

3

DOCKET/FILE NUMBER: P062106

4

CASE TITLE: SECTION 2 HEARING, PREDATORY PRICING

5

DATE: JANUARY 30, 2007

6

7

I HEREBY CERTIFY that the transcript contained

8

herein is a full and accurate transcript of the notes

9

taken by me at the hearing on the above cause before the

10

FEDERAL TRADE COMMISSION to the best of my knowledge and

11

belief.

12

DATED: February 17, 2007

13

14

______________________________

15

KATHLEEN CARR MEHEEN, CSR 8748

16

17

18

19

20

21

22

23

24

25

Updated January 2, 2024