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