Understanding Single-Firm Behavior: Remedies

1

1	UNITED STATES FEDERAL TRADE COMMISSION
2	and
3	UNITED STATES DEPARTMENT OF JUSTICE
4
5
6
7	SHERMAN ACT SECTION 2 JOINT HEARINGS
8	UNDERSTANDING SINGLE-FIRM BEHAVIOR:
9	REMEDIES
10	THURSDAY, MARCH 29, 2007
11
12
13
14
15	HELD AT:
16	UNITED STATES FEDERAL TRADE COMMISSION
17	6TH & PENNSYLVANIA AVENUE, N.W.
18	WASHINGTON, D.C.
19
20
21
22
23
24
25

2

1	A P P E A R A N C E S
2
3
4
5	REMEDY IN THE FACE OF TECHNOLOGICAL CHANGE
6
7
8	MODERATORS:
9	Douglas Hilleboe, Federal Trade Commission
10	Ed Eliasberg, U.S. Department of Justice
11
12
13	PANELISTS:
14	Michael Cunningham, Red Hat, Inc.
15	Renata B. Hesse, Wilson Sonsini
16	Marina Lao, Seton Hall Law School
17	William H. Page, University of Florida
18	Howard A. Shelanski, UC Berkeley
19
20
21
22
23
24
25

3

1	P R O C E E D I N G S
2	- - - - -
3	MR. HILLEBOE: Good morning, everyone, thank you
4	for coming. I'm Doug Hilleboe, attorney with the
5	Federal Trade Commission, Office of the General Counsel,
6	I'm going to be one of the moderators here today for
7	this third session on remedies. My co-moderator is Ed
8	Eliasberg, he's an attorney with the U.S. Department of
9	Justice, Legal Policy Section of the Antitrust Division.
10	Before we start, I need to go over a few
11	housekeeping matters. As a courtesy to our speakers,
12	please turn off your cell phones, Blackberries and other
13	devices that make a noise, and I'll ask the speakers to
14	do the same, they actually interfere with the
15	microphones and we had a little problem with that.
16	Second, the restrooms are located down the hall,
17	through the double doors that you came through. Third,
18	in the unlikely event that the building alarms go off,
19	please proceed calmly and quickly, as instructed. If we
20	must leave the building, take the stairway which is to
21	the right, on Pennsylvania -- on the Pennsylvania side,
22	and after leaving the building, follow the stream of FTC
23	people and meet at the sculpture garden, which is across
24	from the intersection of Constitution Avenue and 7th
25	Street.

4

1	Also, we must enforce our rule that there's no
2	questions or comments that come from the audience during
3	the session. Thank you.
4	We're honored today to have assembled a
5	distinguished group of panelists that have agreed to
6	offer their testimony in connection with this hearing on
7	remedies in the face of technology change.
8	Howard Shelanski is an associate dean and
9	professor of law at the University of California,
10	Berkeley, and the director of the Berkeley Center For
11	Law and Technology.
12	Renata Hesse is a partner at Wilson Sonsini
13	Goodrich and Rosati, and formerly was a chief of the
14	Networks and Technology Enforcement Section At the
15	Antitrust Division.
16	Michael Cunningham is general counsel at Red
17	Hat, Inc.
18	William Page is a Marshall M. Criser eminent
19	scholar at the University of Florida's Levin College of
20	Law.
21	And Marina Lao is a professor of law at Seton
22	Hall Law School.
23	We plan to hear from each of the speakers for
24	about 15 minutes each and then take a ten-minute break
25	and then we'll hear from the remaining speakers. We

5

1	will then have the speakers comment upon what they've
2	heard, and then have a moderated discussion among the
3	speakers with Ed and I leading the discussion.
4	Before starting, I would just like to state by
5	way of introduction that many of the product markets in
6	which the United States enjoys a comparative advantage,
7	vis-a-vis the rest of the world, are fast-changing
8	dynamic markets, including high technology markets.
9	Some critics of the antitrust laws have claimed that the
10	laws, including Section 2, are not nimble enough for
11	effective use in these types of markets. Others
12	disagree. We will explore this issue and others in this
13	session.
14	Some commentators have suggested that the
15	potential for error in antitrust enforcement may be
16	greater in these dynamic markets; however, other
17	commentators have suggested that due to network effects
18	and other possible factors, these markets may tend
19	towards monopolization to a greater agree and therefore
20	perhaps deserve particular antitrust scrutiny.
21	We are interested to learn what these panelists
22	believe about these and other issues, and their
23	implications for antitrust enforcement in Section 2
24	cases.
25	Before beginning with the speakers, my

6

1	co-moderator, Ed Eliasberg has some words about the
2	hearing.
3	MR. ELIASBERG: Thank you, Doug. I very briefly
4	on behalf of the Antitrust Division plan to welcome our
5	panelists, thank you for coming and we look -- we're
6	very much looking forward to hearing what you have to
7	say.
8	So, with that, Ed, let me turn back to you.
9	MR. HILLEBOE: Thank you, Doug. Howard
10	Shelanski is the Associate Dean and Professor of Law,
11	Boalt Hall, University of California, Berkeley and the
12	Director of the Berkeley Center for Law and Technology.
13	From 1999 to 2000, he served as chief economist of the
14	Federal Trade Commission -- Federal Communications
15	Commission, excuse me, and from 1998 to 1999, he served
16	as senior economist for the President's Council of
17	Economic Advisors At the White House.
18	Howard?
19	MR. SHELANSKI: Thanks, Doug, and I appreciate
20	the promotion. Well, I have a few main points that I
21	want to make and the points that I am going to make I
22	hope connect to what my co-panelists are going to say.
23	We had a call a week ago and I just want to set
24	up a few ideas here about the implications of the
25	implementation of remedies for monopolization in a

7

1	high-tech or technologically dynamic markets. And I
2	think my main point, my overall point would be this:
3	Remedies are hard in the best of circumstances, and I
4	think they become more complicated in technologically
5	dynamic settings, but I also think that innovation and
6	the presence of ongoing innovation in a market may
7	affect remedies in somewhat unpredictable ways, and may
8	create opportunities along with the challenges.
9	In particular, I think while innovation makes
10	structural remedies more difficult, it may in some cases
11	make conduct remedies particularly valuable. So, I
12	think while innovative markets are cause for agencies
13	and courts to be more cautious about remedies, I think
14	innovation is not cause for systematic retreat from
15	enforcement or from behavioral injunctions.
16	So, let me explain a little bit why I think this
17	is the case. You'll hear, and I think one often hears
18	that structural remedies are preferable to conduct
19	remedies or behavioral remedies in monopolization cases.
20	But, there are some caveats to this. First I would say
21	that structural remedies are not always available.
22	Where a firm is so integrated that there are not obvious
23	divisions, it's very hard to know how to implement a
24	structural remedy. Just as a classic example, the
25	District Court's second opinion in the United Shoe

8

1	machinery case would be an example.
2	The second caveat I would have is that
3	structural remedies are not always easier than conduct
4	or behavioral remedies, and in fact must often include
5	some supporting behavioral remedies, and as an example,
6	I would talk about the AT&T vertical divestiture that
7	had to be implemented by open access regulations
8	enforced by the FCC and overseen by the District Court.
9	And then, finally, I would say as a general
10	caveat, the effectiveness of structural remedies in
11	Section 2 cases is not assured and there's certainly
12	quite a bit of debate of effectiveness historically over
13	structural remedies. I'll give you a couple of
14	examples. One early quotation, "In administering the
15	antitrust acts, a number of great and powerful defenses
16	against them have been dissolved. So far as is possible
17	to judge the consuming public has not yet greatly
18	profited by their dissolution." That's Judge Rose in
19	United States against American Can in 1916.
20	Okay, now, we haven't had a lot of experience in
21	enforcing Section 2 by 1916, so maybe things have
22	changed, at least some people disagree. Bob Crandell in
23	2003 writes, divestitures are "costly exercises in
24	futility," but I would point you to the excellent work
25	of John Baker and Greg Werden in 2003 providing some

9

1	counter arguments. Just a way of saying effective
2	remedies structurally offer no guarantee of success.
3	Now, I think the structural remedies may
4	actually be even harder in technologically dynamic
5	markets, and let me offer a couple of reasons. First,
6	where a firm or industry is driven by R&D, it may do no
7	good to divest a given division or to leave a company in
8	two without sending the R&D operations with the divested
9	portions of the entity, but R&D operations are often,
10	perhaps even likely, to be more integrated and
11	inter-dependent within the firm and not susceptible to
12	clean lines of separation.
13	The second reason why I think the presence of
14	ongoing technological change may make structural
15	remedies difficult is that even if divestiture is
16	possible, high-tech firms may require more monitoring of
17	conduct during after the divestiture, because key assets
18	in such divestiture are likely to be intellectual
19	property, IP that in some cases may provide joint uses,
20	uses across the lines of the new or divested entities,
21	disputes are likely to be offered over what items to
22	transfer and whether all IP has been disclosed to the
23	new entity.
24	Moreover, because of the cooperative nature of
25	research and development, and in production, in markets

10

1	where product life cycles are short, some post
2	divestiture monitoring of relationships between newly
3	distinct entities may be needed because there may be a
4	natural incentive to favor each other as business
5	partners, and that was something that came up in the
6	wake of the AT&T divestiture, for example.
7	The third reason I think that fast technological
8	change renders structural remedies more challenging is
9	that firm and market structure may be less of an issue,
10	in some technologically dynamic markets. To the extent
11	that the so-called Schumpeterian School is correct, that
12	dynamic markets often display competition that occurs
13	sequentially, through periodic waves of creative
14	destruction, rather than concurrently, through
15	simultaneous production, divestitures may be less
16	effective or necessary such markets, although this is
17	probably more true for horizontal than for vertical
18	divestitures.
19	Okay, and my final reason that structural
20	remedies are tough in technologically dynamic markets,
21	is that where network effects are at issue, structural
22	issues might harm consumers by dissipating positive
23	network externalities. The fact that it might have been
24	better not to have monopoly in the first place does not
25	always mean it is better to break up the monopoly later,

11

1	and if such divestitures are to preserve network
2	externalities, they may have to be accompanied by
3	conduct remedies related to interconnection and
4	interoperability, doing away with those clean properties
5	of structural remedies.
6	Okay, let me turn now to conduct remedies, talk
7	a little bit about how they might work in high-tech
8	markets. As a general matter, we often hear that
9	conduct remedies are difficult, but there are some
10	caveats here as well. Not all conduct remedies are
11	created equal, and as many people have pointed out,
12	negative prohibitions, thou shalt not have exclusive
13	deals, for example, are probably easier to implement
14	than affirmative obligations, thou shall deal with your
15	rivals. In part because the negative prohibitions
16	entail less involvement of courts or agencies in
17	regulating terms of trade.
18	The second caveat that I would add is that
19	conduct remedies can have beneficial prospective impact,
20	even if they cannot roll back illegally accumulated or
21	prolonged market power. Some people say, look, conduct
22	remedies are closing the barn doors after the cows are
23	out, but if there are still some cows inside the barn,
24	it's not a bad idea to shut the door.
25	Third, even if a conduct remedy is ineffective

12

1	or weak in a given case, I think conduct remedies can
2	have important deterrent effects on others contemplating
3	the illegal behavior, and it's -- in a point that's
4	often made, some people say, if you can't be sure that
5	your conduct remedy is going to be effective, why bring
6	the case? Another reason to bring the case beyond
7	deterrence is I think as we get more experience with
8	different kinds of conduct, it can become clearer what
9	is good and what is bad, and it enables agencies to move
10	more quickly in subsequent cases, and perhaps get a
11	remedy implemented while the harm is still able to be --
12	to be nipped in the bud, so I would not let lack of a
13	clearly successful conduct remedy -- I think one needs
14	to be clearly articulable at the start of a case, but if
15	you can't be sure it will be implemented in time or it
16	will be successful in remedying the market power, there
17	may be some reasons to go ahead with the case anyway in
18	terms of establishing precedent and creating deterrence
19	effects.
20	And finally, just an observation, I think that
21	the effectiveness of conduct remedies are likely to --
22	the effectiveness is likely to be tied to the precision
23	with which one can define the cause of anticompetitive
24	harm, and in some cases, this can be done quite clearly,
25	and in those cases, I think behavioral injunctions can

13

1	be quite effective.
2	So, the overall lesson about conduct remedies, I
3	think that it is right to be weary of behavioral
4	remedies, particularly those in which the enjoined
5	conduct has ambiguous welfare effects, or in which
6	courts or agencies will have to become involved that
7	were doing terms of trade, but in the right context,
8	conduct remedies can work and can send valuable
9	deterrent signals.
10	I would just say that inability to articulate a
11	structural remedy therefore should not be decisive in
12	whether or not to prosecute an argument that is
13	sometimes heard.
14	Okay. Well, I think that technologically
15	dynamic markets create both challenges and opportunities
16	for implementing conduct remedies. The first challenge
17	is this: If one accepts that remedies may deter
18	marginal innovation, and I'll assume for the moment that
19	all innovation is good, because private returns are less
20	than social returns to innovation. Let's just take that
21	as a working assumption, it need not be true in all
22	cases, but if one accepts that, and one accepts that
23	remedies can marginally deter innovation, then the
24	deterrence risk and the costs of such deterrence may be
25	much greater in dynamic markets. It needn't be the

14

1	case, but I think innovation deterrence becomes a more
2	salient issue and a more salient concern in
3	technologically dynamic markets.
4	The second challenge is that in fast-changing
5	markets, it is more likely than it is in more static
6	settings that the conduct at issue in the case will be
7	moot by the time antitrust liability is established.
8	And in such cases, neither conduct nor structural
9	remedies are likely to be effective, and perhaps
10	something else like disgorgement might be called for if
11	such a remedy can be created.
12	But there are also opportunities in high
13	technology settings, I think, for conduct remedies to be
14	particularly effective. In some cases, technological
15	dynamics can render conduct remedies effective where
16	they would not be in more static markets.
17	In some cases, monopoly once obtain may not be
18	easily eroded, even if exclusionary or predatory conduct
19	that contributed to that monopoly is stopped. Whether
20	because of brand recognition, economies of scale, or
21	customer switching costs, new entrants will be slow to
22	appear or succeed, even when other barriers to entry,
23	such as the exclusionary or predatory conduct at issue
24	in the case, even when those barriers are eliminated,
25	you might not see competition arising.

15

1	But I think where competition is more innovation
2	based and where product life cycles are short, an
3	injunction against the behavior that led to the
4	establishment or maintenance of monopoly power may prove
5	very effective, as it is the latter set of barriers,
6	rather than any brand or economic advantage, that might
7	have kept the incumbent dominant.
8	As new waves of innovation come forward, how did
9	they stop someone else from being the innovator who came
10	in with the new product? Well, through the exclusionary
11	or predatory conduct, and branded here and switching
12	costs, other things like that, may be very, very
13	different in the high-tech environment. So, merely
14	eliminating the harmful conduct may open the door for
15	new entry and the conduct or remedy, particularly
16	negative injunctions, I think, can be very successful
17	and very helpful.
18	I would like to just raise an additional point
19	about the overall question of whether or not the cycles
20	of innovation move so quickly and the innovation process
21	moves in such different a way from the standard
22	competitive process that we should step back generally
23	from antitrust enforcement, and this is an argument that
24	one hears quite often.
25	I think when one looks at the kinds of behavior

16

1	that limit innovation, and that stop people -- that stop
2	competitors from innovating, it's very unclear to me
3	whether or not monopoly has anything particular to
4	recognize it, nor is it clear to me that new waves of
5	innovation are always going to be sufficiently powerful
6	to overcome artificial barriers to entry like
7	exclusionary -- exclusionary kinds of behavior like
8	exclusive deals when it is a monopolist that has that
9	exclusive deal, contractual terms that bar competitors'
10	products from ever being used, tying that prevents
11	consumers from ever having access to products.
12	It's unclear to me no innovation will always be
13	so great that it can overcome those barriers, those
14	barriers can lead to slower product life cycles, and
15	greatly harm consumers, and I think that there's a lot
16	of evidence of benefits from antitrust enforcement in
17	high-tech areas. And when one looks at the studies that
18	have said there are no benefits to Section 2
19	enforcement, or in a more nuance way, no benefits to
20	Section 2 enforcement in technologically dynamic
21	markets, there's a counterfactual, all of these papers
22	acknowledge the counterfactual, and we can't tell what
23	would have happened absent the antitrust enforcement, we
24	can't tell what would have happened in other markets had
25	there been antitrust enforcement, and then those

17

1	arguments are sort of dismissed, tucked under the
2	carpet.
3	I wouldn't dismiss them so easily. And, so, my
4	overall argument would be, be very cautious, be very
5	case-by-case in the application of Section 2 remedies in
6	high-tech markets, I think structural remedies are
7	likely to be harder to implement, but there may be good
8	opportunities for conduct remedies to be very effective.
9	Thanks.
10	(Applause.)
11	MR. HILLEBOE: Thank you very much, Howard. Our
12	next speaker, excuse me, is Renata Hesse, who is a
13	partner at Wilson Sonsini Goodrich and Rosati. Prior to
14	joining Wilson Sonsini, Renata served as the chief of
15	the Networks and Technology Enforcement Section at the
16	Antitrust Division and oversaw much of the division's
17	technology litigation, including the Oracle/Peoplesoft
18	and First Data/Concord matters. In addition, Renata
19	worked extensively on both the American Airlines and the
20	Microsoft case.
21	Renata?
22	MS. HESSE: Getting myself around is a little
23	harder these days.
24	So, Howard covered a lot of ground which I think
25	fundamentally I agree with almost everything he said.

18

1	In fact, I think I probably agree with everything he
2	said, but wanted to pick up where he was leaving off,
3	which was I think in talking about the notion that you
4	shouldn't back away from Section 2 enforcement in high
5	technology markets, and the main reason why I think
6	that's true is that despite all of the innovation and
7	the fast pace of change in those markets, there is an
8	opportunity for durable market power to exist in them,
9	and you do want to make sure that you're not overlooking
10	that possibility and potentially addressing it.
11	So, I wanted to start with just a few basic
12	points about Section 2 remedies that I think are
13	important, and some of these overlap with some of the
14	things that Howard said and I'm sure that will happen as
15	we go along down the line of speakers, but the first
16	thing that I wanted to talk about is the importance of
17	focusing on remedy early, and the main reason -- there
18	are several reasons for that, but the biggest reason is
19	that it helps you try to figure out what your goal is.
20	What's the violation that you're really thinking about,
21	what do you think has really happened that's harmful,
22	and how can you address it? That isn't to say that if
23	you can't come up with a perfect solution to the problem
24	that you shouldn't go ahead and try and do something
25	about it.

19

1	I think Howard is right that there's a good
2	deterrent effect in enforcing the law, even if you're
3	not 100 percent sure that the way that you think you can
4	fix it will be successful, but I do think it will -- it
5	helps you focus your investigation, and here again, I'm
6	speaking as if I were a government lawyer, but focus
7	your investigation and theories so that you can really
8	figure out whether or not you've got a case that is
9	worth allocating resources to, and pursuing.
10	And I just think it gives you a much better
11	sense of the definition of the harm that you're trying
12	to alleviate.
13	The second point is that I think when you start
14	with thinking about remedy, or at least you think about
15	remedy relatively early in the process, you can get a
16	better sense for whether or not you actually can come up
17	with a remedy that is really going to leave the
18	marketplace in a better place than it was when you
19	started.
20	And I would sort of call this the first do no
21	harm rule, and it is one of these things which you
22	always need to bear in mind, which is that you don't
23	always want to make things worse, you don't want to
24	deter innovation or take an action in the marketplace
25	which stifles productivity, and I think in technology

20

1	markets, that's something that you really need to keep
2	in mind.
3	But if you were stepping back and thinking about
4	that early, you can think about whether or not there are
5	ways to achieve the goal that you want to achieve
6	without having at least a large countervailing harmful
7	effect.
8	The third point is related to the resource
9	allocation point that I made. I think fundamentally
10	it's just a basic responsibility that particularly
11	government enforcers have to think about how you're
12	going to fix the problem, and whether or not the problem
13	is subject to a fix that's worth the investment of
14	resources in not only the investigation and prosecution
15	of the matter, but also the compliance and enforcement
16	activities that will happen post judgment, and those
17	are, I think, much more complicated when you're talking
18	about conduct remedies and structural remedies, but,
19	again, Howard correctly notes that when you do a
20	structural remedy in these markets, very often there are
21	going to be conduct remedies associated with it in any
22	event.
23	But I think you really do want to have in your
24	mind whether or not the consumption of the resource is
25	likely to result in some improvement to the competitive

21

1	conditions in the marketplace.
2	And then there's a fourth point which is that
3	sort of the question of if you have a good idea of what
4	you think the remedy that you want to put into place is,
5	then I think you'll have a better idea of whether or not
6	the -- again, the pursuit of the investigation or
7	prosecution is worth while, and by that I mean that
8	there are some kinds of Section 2 violations that are
9	easier to remedy than others.
10	So, one example might be you can think of
11	exclusive dealing or vertical foreclosure, for example,
12	where you have fairly easily identifiable concrete types
13	of conduct that you can undo. I think monopoly
14	maintenance, to a certain degree, monopoly acquisition
15	cases are much harder.
16	So, if you're in the situation where you're
17	balancing these things out, and you've got a choice
18	between two matters that you want to devote your
19	resources to and one of them has a reasonably good
20	likelihood of being able to be fixed, and the other is a
21	little tougher, then you've got to figure out how to
22	allocate your resources, then you might want to think
23	about going towards the one that actually has a solution
24	that you can identify and that you think will be likely
25	to result in an improvement in the competitive

22

1	conditions.
2	And this just goes back to something that I
3	think people often think about in the context of -- of
4	the -- when you're trying to come up with a remedy, what
5	is it that you're trying to achieve, are you looking at
6	a monopoly that you believe has been illegally created
7	and are you trying to undo that, or are you looking at
8	conduct that has maintained a monopoly and are you
9	trying to restore the conditions of the competitive
10	marketplace to the pre-exclusionary conduct state? And
11	depending on which of those two things you're looking
12	at, you're going to have a pretty different, I think,
13	idea about what's the right way to go about recommending
14	the harm.
15	The second thing I wanted to talk about was just
16	the point that Howard started with, which is structural
17	remedies and the general point that generally I think
18	structural remedies should be preferred. I think it's
19	clearly true that they are not always possible, and
20	that's certainly more true in Section 2 cases than in
21	other kinds of cases, but I wouldn't advise sort of
22	ignoring them as possible ways of recommending harm,
23	because I think they do have a number of benefits.
24	One of the benefits is that developing a
25	functional set of conduct restrictions that are likely

23

1	to have a beneficial effect, without having this sort of
2	countervailing, potentially negative effect on the
3	marketplace is an extremely complicated and resource
4	intensive process. It took a really long time to come
5	up with the conduct restrictions that we developed in
6	the Microsoft case, and I think, you know, you can --
7	it's open for debate whether or not those were worked
8	well or not well, but it took a long time to figure them
9	out, and to just evaluate all the different
10	possibilities and try to develop language that's
11	concrete enough and understandable enough in a legal
12	document for people to actually then be able to
13	implement it and understand it and understand what the
14	rules of the road are. It's just an inherently
15	difficult process to do, and I think that isn't just
16	Microsoft, that's any time when you're trying to come up
17	with a set of conduct restrictions where you're dealing
18	with complex technology.
19	It's also hard to judge their success, I think,
20	and that's also true in structural remedies, in some
21	situations, but it's very hard to know when conduct
22	restrictions have succeeded. I think you can know when
23	they've failed, but I don't think you can know as easily
24	when they've succeeded. How do you measure success with
25	conduct restrictions?

24

1	I think structural remedies generally eliminate,
2	although not entirely, the need for ongoing enforcement
3	in compliance activity, which also can be an extremely
4	time consuming and resource intensive process. It can
5	require, and this is something else I can talk about a
6	little bit later, but it can require a lot of assistance
7	from people who know more about technology and business
8	and licensing and all these things that come up in
9	technology markets work, and structural remedies tend to
10	need a lot less of that.
11	I think structural remedies are generally less
12	easy to evade. It's pretty clear what you're supposed
13	to do, and you've either done it or you haven't done it.
14	You've either divested the plant or the asset or
15	whatever it is, or you haven't. You know, there are
16	issues associated with those kinds of things, whether or
17	not you found an adequate buyer and all of those other
18	sorts of issues, but at least there's a very clear line
19	about what you are supposed to have done.
20	I think they have a potentially greater
21	deterrent effect, because they have the capability at
22	least of really restructuring a business in a way that
23	most businesses don't want to have happen. So, that can
24	discourage people from engaging in conduct that folks
25	think violates Section 2.

25

1	And I think generally, again with some of the
2	caveats that Howard laid out, they're more likely to
3	work. The lines are clearer, and if you've actually
4	proven a violation where you can support imposition of a
5	structural remedy, I think the likelihood of that
6	structural remedy having an effect is probably higher.
7	So, those are some kind of basic points. A few
8	points that are more directly connected, just to sort of
9	the technology markets, and the first is, you know,
10	everybody always talks about technology markets are fast
11	changing and innovation changes everything, and as
12	Howard said, sometimes people say, maybe you don't need
13	to worry about them because they're just going to be
14	self correcting. I tend not to agree with that latter
15	viewpoint, for the reason that I started with, which is
16	that it's clear that there's a possibility for the
17	existence of durable market power in these markets, so I
18	think just leaving them alone and hoping that the
19	exclusionary conduct somehow magically stops and things
20	correct themselves is not likely to lead to a lot of
21	success.
22	I do think that the fact that they can sometimes
23	be slow and that the antitrust enforcement process can
24	sometimes be slow is a down side in these markets, a
25	greater down side in these markets than in other

26

1	markets, because sometimes you feel like you get to the
2	end and you're addressing the problem when it's actually
3	a little bit too late.
4	As a consequence, I think you need, when you're
5	thinking about conduct remedies in technology markets,
6	to be a little bit more flexible about how you think
7	about them. And to address categories or types of
8	conduct relating to types or categories of products or
9	services as opposed to saying, well, this -- you did
10	this particular thing with this particular kind of
11	product, and you should do that -- you shouldn't do that
12	anymore. This is the negative prohibition point versus
13	an affirmative obligation point.
14	If the conduct remedy is too narrowly focused,
15	it runs the risk of being ineffective, and I think in
16	most cases is likely to be ineffective, particularly,
17	again, if you're talking about undoing some sort of harm
18	that has occurred.
19	You know, Microsoft is a simple example of this,
20	the consent decree doesn't just talk about browsers,
21	which was the primary focus of the case, but it talks
22	about other products which were potential platform
23	threats and has some construct restrictions in it that
24	are designed to try to go after those particular -- or
25	not go after them, but to try and make sure that the

27

1	conduct relating to those other kind of potential
2	platform threats were restrained.
3	There's a possibility in technology markets that
4	they should be of shorter duration. Again, Microsoft is
5	another example, it was a five-year consent decree, it's
6	now been extended in some pieces for longer than that,
7	but I think there's a reasonable basis for at least
8	looking at the question of whether or not you really
9	need something to last ten, 20, some decrees in the past
10	have lasted for hundreds of years, some of them very
11	perpetual, and whether or not that makes sense
12	particularly in the context of technology markets is I
13	think something that people -- it's worth looking at.
14	I also think if you're going to think about
15	decrees of shorter durations, or remedies of shorter
16	durations, that including some mechanism for revisiting
17	that question before the term of the decree expires is a
18	good idea. I think it's just these markets are
19	inherently unpredictable, and given the complication of
20	structuring conduct provisions in them, that giving
21	yourself an opportunity to take a second look and having
22	a standard for how you would be able to convince a court
23	that you need to extend a decree in these kinds of
24	markets is something that should be given some
25	consideration.

28

1	And the final point on this area is that I think
2	conduct remedies in Section 2, Section 2 remedies in
3	technology markets may need to be more forward looking,
4	and this is a little slightly basically the same thing
5	with a slightly different pitch on it, but you do have
6	to think about what it is that you can predict about the
7	marketplace and changes in the marketplace going forward
8	and whether or not what you've devised in the context of
9	the conduct remedy is adequate to address the changing
10	technology in the marketplace.
11	The last piece about technology markets that I
12	think makes them different is that they're hard, and
13	it's hard to understand them, and they're particularly
14	hard for people who are not educated in technology.
15	And, so, compliance monitoring enforcement can be a
16	difficult thing to do.
17	As a consequence, I think if you're looking at
18	these markets and you're looking at behavioral
19	restrictions, particularly ones that relate to licensing
20	of intellectual property or access to technology or
21	just, you know, you're requiring a company to stop doing
22	a particular activity with a particular type of
23	technology, that you really need to anticipate getting
24	some technical help, and when I think of technical help
25	in this context, I don't think just of software

29

1	engineers or hardware engineers, but I also think of
2	licensing expertise, business expertise, you know,
3	trying to figure out whether a royalty ran is a
4	difficult problem, and it's not a problem that most
5	antitrust lawyers deal with on a day-to-day basis.
6	And having the ability to have access to people
7	who actually do that kind of work for a living, who know
8	what particular types of technologies, what kinds of
9	royalties particular types of technologies command is, I
10	think, critical to the ability to actually do an
11	adequate job of monitoring and enforcing compliance.
12	Again, I started with sort of a more broad
13	definition of technical assistance, but a narrow
14	definition of technical assistance, which is just
15	actually having somebody who knows how software code is
16	written, and what to look for and how to evaluate
17	whether or not something has been done in the code is
18	very important. I think one of the really unusual and
19	innovative things that was in the Microsoft decree was
20	the technical committee provision, which allowed the
21	Department of Justice and the states to have access to
22	basically a full-time group of technical consultants who
23	were hired to work for those people and the cost of
24	which was borne and continues to be borne by Microsoft.
25	I think it was an unusual idea, but it really

30

1	has become, I think, a key component to the United
2	States enforcement and monitoring, compliance monitoring
3	efforts of the Microsoft decree, and it was essentially
4	copied by the European Commission in the work that
5	they're doing in Microsoft as well.
6	And it had not been done before. There were
7	lots of times where in complicated markets people had
8	used monitoring trustees, I shouldn't say there were
9	lots of times, but there were examples of monitoring
10	trustees being used, usually they were in things like
11	prison condition litigation, where there was some pretty
12	complicated oversight that was needed, but hiring
13	technical experts to help out was an innovative thing to
14	do and I think has proven to be a pretty successful
15	component of the Microsoft decree.
16	Now, you also may need technical assistance when
17	you're trying to figure out whether or not somebody has
18	violated the decree and you actually want to go after
19	them for contempt. I think the Microsoft model doesn't
20	quite fit so well in that context, because it's a little
21	hard to see how you can justify the party who you're
22	going to be pursuing in contempt actually paying for the
23	expert that you're going to be using, to go after them
24	in contempt, but it's something that people -- you want
25	to think about, and at least have the resources and

31

1	capability to get that kind of help on board.
2	So, I have probably 30 seconds at this point
3	left. The last thing I would say is that licensing
4	remedies are incredibly common in technology markets.
5	They can be useful, and I think can work well, but I
6	think they work particularly well in the context where
7	you know or have a very good idea of what the
8	intellectual property is or what the asset is that needs
9	to be licensed, are there particular patents who needs
10	them, and again, if you go back at the very beginning,
11	to those are things that you can think about early on
12	and figure out and they'll help you determine whether or
13	not a licensing remedy is likely to be successful.
14	And of course when you're doing that, you need
15	to think about the policy issues that are associated
16	with compulsory licensing of intellectual property,
17	which is a hot topic these days.
18	(Applause.)
19	MR. HILLEBOE: Thank you so much, Renata, for
20	those comments.
21	Michael Cunningham is general counsel at Red
22	Hat, Inc. Prior to joining Red Hat, he served as
23	associate general counsel at IBM, where he had legal
24	advisory responsibilities for the Business Consulting
25	Services Division for Europe, the Middle East and

32

1	Africa. He was also a partner and associate general
2	counsel at PricewaterhouseCoopers.
3	Michael?
4	MR. CUNNINGHAM: Thank you, and good morning.
5	I'm pleased to have the opportunity to participate in
6	this important consideration of Section 2 remedies, to
7	do so before distinguished representatives of the
8	government, as well as with this particularly
9	knowledgeable panel.
10	I'm the general counsel of Red Hat. I'm going
11	to make a little disclaimer, I'm a technology lawyer,
12	I'm not principally an antitrust lawyer. I hope that I
13	can offer some comments, however, as an executive of a
14	technology company that are relevant to these inquiries.
15	With your indulgence, I would like to describe a
16	bit about our business that I think is relevant
17	innovation, given the debate about antitrust remedies
18	stifling innovation, I think it's particularly
19	appropriate this morning.
20	The software solutions that Red Hat offers, and
21	for which we provide services, are developed by very
22	broad horizontal communities that are without
23	geographic, organizational or political boundaries. The
24	community of innovators that unleash the value of open
25	source are not contained within Red Hat. Some of its

33

1	contributors are, but it's not.
2	The contributors include the customers and
3	vendors of hardware and software. It includes
4	academics, it includes many, many motivated individuals
5	that we call hackers, it includes persons from every
6	continent and from multiple political subdivisions.
7	The development environment is also not
8	controlled by any single individual company or political
9	entity, it is instead a free, meritocratic marketplace
10	of ideas. Individuals take these ideas and they place
11	these ideas with their individual name and reputation
12	into the marketplace in a particular software
13	development project to which their idea is relevant.
14	There are literally thousands of these projects
15	out there. In one of our offerings, Red Hat Enterprise
16	Linux, hundreds of projects are represented. These
17	ideas are then reviewed by that development community,
18	for that project, and only those ideas that can handle
19	the open scrutiny of this open source community are then
20	adopted.
21	In this way, the best ideas and the bets bits of
22	ideas bubble up. Moreover, if there happen to be a
23	serendipitous discovery that is made in one of those
24	projects that's relevant to another project or might be
25	an entirely new approach, the contributor or any other

34

1	person is free to contribute it to that project or
2	indeed to go out and start a new project to take the
3	technology in a new direction.
4	This model has produced and continues to produce
5	copious innovation. It also accelerates and multiplies
6	innovation, I would argue, by providing tools of
7	innovation, such as information ideas to a broader and
8	more diverse community than development within any one
9	firm is possible could provide.
10	The open exchange of information and ideas is an
11	innovation force multiplier. For example, sophisticated
12	business and other users of software frequently take the
13	modular pieces of well crafted software that's developed
14	in the open source community, cobble bits and pieces of
15	it together, modify it, append to it and create
16	solutions for problems that heretofore were not solved,
17	or new problems that arise in their business.
18	Similarly, the creative juices of the lone
19	teenager in North Dakota in some remote location can
20	contribute to that process, so can a Cal Tech physicist
21	who is wondering why there hasn't been a software
22	development that would help in his or her research. And
23	so are many, many others unleashed in the creative
24	process through this open development and collaboration
25	model.

35

1	The modular and open nature of open source
2	software has fueled much innovation, but it is by no
3	means limited to software. It is not a software-only
4	phenomena. No, I would submit to you that the relative
5	ubiquity and low cost of the Internet, and collaboration
6	tools like email and dedicated web sites portends for
7	joint collaboration that is unleashing all sorts of
8	innovation across the world.
9	If you've read the best selling book by Tom
10	Friedman, The World is Flat, you will get a very good
11	sense of some of these trends, I think. I would also be
12	happy to comment on some other areas where that
13	innovation is being unleashed in the questioning, if
14	that's helpful.
15	With that bit of an introduction, maybe I should
16	turn my attention now more directly to remedies. First,
17	I believe that in the software space at least, the
18	relevance of the antitrust law hangs on the issue of
19	remedies. I can think of no way as a practitioner and
20	an executive in a company in the industry to more
21	starkly illustrate that point than to disclose my actual
22	advice to my client in pursuing whether to participate
23	in or pursue any monopoly-related case, whether that be
24	in a government-related case or in private litigation.
25	I would tell my client, it's too expensive for

36

1	you to fully embrace and do that. You cannot do it.
2	You don't have enough money to pursue it, it's certainly
3	over $10 million, it will be a long time, and it is
4	likely, I would submit to you, at least this would be my
5	advice, it is likely and substantially likely that the
6	remedy that will result will be of limited utility. So,
7	therefore, those sorts of expenditures would not be
8	justified.
9	And guess what? Those that the government
10	representatives seek to regulate know this, and they
11	know it well. By way of illustration, a high-ranking
12	representative, indeed a very high-ranking
13	representative of a party found to have market power by
14	multiple international competitive authorities has
15	aggressively and indeed smugly advised Red Hat that
16	there is no competition authority in the world that this
17	firm will not outspend, outlast, and seek to thwart.
18	In short, the system seems broken in terms of
19	speed, cost, and effectiveness of remedies, at least
20	from my little corner of the world. You know, why is
21	this the case? Well, as others have said, technological
22	change is very rapid and litigation is not. The rate of
23	change at least in information technology is in very
24	short cycles, three to five years, maybe six to eight
25	years, certainly not longer than that in many, many

37

1	areas of information technology.
2	Remedies that only address a particular market
3	complained of, and established at great expense, will
4	often be too late to provide meaningful relief. A
5	remedy focused on future conduct would address some of
6	those limitations and in many instances I think is
7	necessary.
8	I also am intrigued by the idea of smaller
9	simpler cases with speedier trial times that would focus
10	on future contact to make the law more relevant.
11	Clearly cost and delay undermine the perceived and
12	actual effectiveness of the antitrust laws in our
13	competitive zone.
14	In that way, some of Professor's Lao's writing
15	on the role of the intent in finding liability seem a
16	fruitful avenue for further inquiry to me.
17	Second, technology can be manipulated. The
18	speed with which information technology moves and can be
19	molded provides real opportunity for conscious
20	manipulation by the monopolist away from the market
21	complained of. The government enforcement actions
22	against Microsoft are an example of the timing
23	challenges, I'm thinking now about the European Union,
24	even the most aggressive threats by the EC are mired in
25	delay, seemingly extended without limit.

38

1	According to the most recent statistics we've
2	seen, Microsoft continues to gain in the operating
3	system worker group server market, meanwhile the market
4	continues its very rapid evolution, probably reducing
5	the relevance of any remedy that may eventually be
6	enforced and/or issued.
7	I guess I should also point out that private
8	enforcement actions have not solved the problem either,
9	this won't be a surprise from my earlier comment. The
10	antitrust law, like the Ritz Carlton, is open to the
11	rich and poor alike. The most entrepreneurial and the
12	most innovative firms, the small fledgling ones are
13	without means to mount private antitrust cases.
14	Let me turn my attention for a few moments to
15	innovation. Protecting competition does not mean
16	stifling innovation, I don't believe. While there is an
17	inevitable tension between the intellectual property law
18	and the antitrust law, competition law cannot achieve
19	its purpose if regulators and courts are preoccupied
20	with a concern that remedies affecting some intellectual
21	property rights will necessarily stifle innovation.
22	That focus on IP, that is intellectual property,
23	a legal concept, is misguided. The focus should be on
24	true innovation, not patents and copyrights, public
25	grants of a monopoly.

39

1	Why is that the case? Well, first I think
2	equating innovation to the accumulation of intellectual
3	property is suspect, at least in the software world.
4	The software patent approach in the United States is
5	being broadly questioned, and that's the case for at
6	least two or three different reasons.
7	First of all, the software industry in
8	particular survived for almost 20 years with very
9	limited forms of software patents, not the broad range
10	that we now see following State Street and other court
11	decisions.
12	Second, I would submit to you the relationship
13	of software patents to innovation is suspect. I
14	regularly review the academic literature in this area
15	and I am aware of no convincing argument that software
16	patents have unleashed -- and no empirical study --
17	that they have unleashed and spurred additional
18	innovation.
19	Third, the news is regularly filled with stories
20	of highly suspect software patents, patents that are not
21	new and innovative, ones that are anticipated by prior
22	art and ones that common sense tell us lack sufficient
23	novelty to warrant 20 years of protection.
24	Of course that shouldn't be surprising, there
25	are well publicized challenges in the Patent & Trademark

40

1	Office, there's no effective and searchable database on
2	prior art for software. There's also serious challenges
3	in retracting and retaining the kinds of experts that
4	Renata talked about to actually evaluate what is seeking
5	to be patented.
6	I say that just to suggest that the innovation
7	reflected in software patents is questionable at times.
8	Therefore, giving, you know, complete deference to
9	intellectual property in that context seems misguided.
10	Even more important to this debate, as my
11	opening remarks sought to illustrate, there are broad
12	communities of collaboration that are massively
13	innovative. Please note that their style of
14	collaboration is not readily or naturally susceptible to
15	patent protection, given the open and collaborative
16	nature of their exchanges.
17	Thus, innovation of the firm is not the only or
18	even the most effective form of innovation to be
19	considered or protected when facing the market
20	disruptive effects of monopolists. Powerful new
21	innovation paradigms are upon us now and they're growing
22	and they need to be considered and measured in balance.
23	But even if we were to assume that the firm is
24	the epicenter of innovation, the smallest and perhaps
25	most innovative are without the means to challenge the

41

1	innovation of the monopolist that is purported to be
2	reflected in intellectual property. The combination of
3	suspect software patent quality and the disparity of the
4	cost to acquire a patent versus the cost to defend
5	against it skew IP protection in favor of larger
6	enterprises with market power.
7	Cost of acquiring a patent, let's say, is
8	$25,000 to $35,000. It absolutely pales in contrast to
9	the cost of a proper infringement defense. That is
10	variously $3 to $5 to $7 million, and by all accounts is
11	growing at present.
12	Moreover, the monopolist can disrupt the
13	business of smaller competitors merely by suggesting to
14	consumers that its IP is infringed, without any proof
15	whatsoever. If you consider Steven Bommer's recent
16	statements that the users of Linux have an undisclosed
17	off balance sheet liability to Microsoft, which were
18	offered without any substantiation whatsoever. And the
19	SCO litigation that is ongoing I think offers some
20	interesting and vicarious variance on the same theme,
21	which I would also be happy to comment on in the
22	question and answer period.
23	Keeping on the intellectual property theme, an
24	effective remedy needs to prevent the extension of
25	market power. A company who has acquired market power

42

1	through anticompetitive conduct shall not be permitted
2	to be able to hide behind intellectual property
3	protection to reinforce and extend its market power. I
4	think there is an interesting lesson in history on this
5	that deals with data formats.
6	In particular, I would like to contrast how
7	Microsoft came to compete in word processing, versus how
8	it now competes. The background is as follows:
9	Software products manipulate and ultimately store
10	customer data after that manipulation. To the extent
11	this data is then placed into storage formats, that are
12	claimed as either proprietary or protected by
13	intellectual property of the software vendor, then the
14	ability of a competing product to make effective use of
15	the stored customer data and break into and compete in
16	that market, which is likely reinforced by very strong
17	network effects, can be precluded.
18	Take, for example, Microsoft's word processor
19	competition against the then-important market position
20	of the WordPerfect product in the 1980s. Because the
21	data format's inability to represent the data with
22	substantial fidelity was possible, Microsoft could
23	compete at the enterprise level by saying, give me a try
24	in parallel with WordPerfect. If I do better, then
25	incur the cost of switching out your old technology and

43

1	taking on our technology.
2	In contrast today, I would submit to you the
3	formats of Microsoft alphus data have been and are
4	increasingly being obscured by Microsoft and cannot be
5	presented, that is the data cannot be presented with
6	true fidelity by any competitor, like OpenOffice, which
7	thereby extends the time of their dominant position and
8	permits extension of power into adjacent markets.
9	It is the case that Red Hat cannot effectively
10	compete with open source personal productivity
11	applications, like word processors and other things, at
12	the enterprise level against Microsoft, it can't get its
13	foot in the door. If a client wants to give someone a
14	try and you can't render their existing data in a
15	meaningful fashion, that prevents anyone from entering
16	into that market, I would submit to you, or doing so
17	easily, anyway.
18	Microsoft controls, I would submit to you, a
19	facility of competition through the extension of IP and
20	proprietary formats that is needed to meaningfully
21	render and manipulate customer data. I have no doubt
22	that's why you're seeing states like Massachusetts
23	aggressively consider the open document format, a truly
24	open standard in format in its procurement processes.
25	The mono type litigation of Red Hat is another

44

1	example that illustrates that that I would be happy to
2	comment on later.
3	In summary, I guess I would say that innovation
4	does not equate to intellectual property, and therefore
5	greater focus on preserving and promoting true
6	innovation in the marketplace is warranted. Further,
7	there are numerous ways in which the use and assertion
8	of intellectual property rights can be a pretext that
9	chills competition and extends monopoly power.
10	Thank you.
11	(Applause.)
12	MR. HILLEBOE: Thank you very much, Michael, for
13	that, and I think we will take about a ten-minute break
14	now.
15	(Whereupon, there was a recess in the
16	proceedings.)
17	MR. HILLEBOE: Thank you, everyone. William
18	Page is a Marshall M. Criser eminent scholar at the
19	University of Florida Levin College of Law and he is
20	also an alumnus of the Antitrust Division, where he
21	served as a trial attorney in the 1970s.
22	Bill?
23	MR. PAGE: Thank you. Rather than speak in
24	generalities about Section 2 remedies in high-tech
25	markets, I want to zero in on one highly technical and

45

1	seemingly obscure provision in the final judgments in
2	the government's Microsoft case that has turned out to
3	be the most difficult and the most problematic in its
4	enforcement.
5	The provision requires Microsoft to license to
6	software developers communications protocols that
7	Microsoft uses in its Windows Client operating systems
8	to interoperate with Microsoft server operating systems,
9	either in corporate networks or over the Internet.
10	Communications protocols are the rules for transmitting
11	information between different devices.
12	So, in a computer network, the protocols allow a
13	user of a client computer, for example, to store
14	information on a network drive or send an email or
15	display a web page, among many other things.
16	This sort of interoperation is relatively easy
17	when the client computer's operating system and the
18	server operating system share a common base in code.
19	It's like they speak the same language, so they can
20	interoperate easily.
21	Where the client computer, usually a Windows
22	client, has to interoperate with servers from other
23	vendors, then the problem with interoperability becomes
24	much more difficult, but there are ways of solving them.
25	There are recognized ways of solving them. Some involve

46

1	installing a client on Windows that would allow
2	interoperation with the non-Windows server and
3	applications running on it.
4	There are also standard protocols that are
5	available and supported in Windows. This provision
6	requires another way of assuring interoperation, that is
7	requires Microsoft to disclose its proprietary
8	protocols, to license them to software developers so
9	that they can interoperate. The near-term goal would be
10	for them to be able to write programs that will
11	interoperate as well with Windows clients as
12	applications running on Microsoft servers.
13	The long-term goal is to allow -- is to preserve
14	in this network context the so-called middleware threat
15	that was the focus of the government case. The
16	middleware applications running on servers, the concern
17	is, may eventually evolve into platforms that could
18	rival the Windows desktop and thereby erode the
19	application's barrier to entry. Essentially the theory
20	of the government case.
21	In spite of its apparent obscurity, this
22	provision has been given an unusual amount of importance
23	by the District Court enforcing the Microsoft judgment.
24	She's referred to it as the most forward looking
25	provision in the final judgments and as necessary to

47

1	assure that the other provisions don't become
2	prematurely obsolete. It's now being implemented by the
3	two sets of plaintiffs in the Microsoft litigation, the
4	Antitrust Division and the nine settling states, and
5	also by the group of non-settling plaintiffs who were
6	awarded essentially the same relief, but there are
7	different enforcement mechanisms.
8	There's the technical committee that Renata
9	referred to in the Antitrust Divisions's consent decree
10	and there's a technical consultant to the non-settling
11	states under their decree, but they're coordinating
12	their enforcement efforts. Both of these judgments went
13	into effect in 2002.
14	And the plaintiffs in both cases and Microsoft
15	has been filing status reports every two months about
16	the enforcement of both of the judgments, and I have
17	studied these reports with the help of a research
18	assistant, who was also a software developer and a
19	management consultant, and so he has been sort of my
20	technical consultant. He provided all of the technical
21	expertise in this study, because I certainly claim none.
22	The enforcement of this provision, this one
23	provision in these judgments has dominated these
24	reports, particularly in recent years. It by far
25	occupies most of the reports and certainly most of the

48

1	time of the technical committee. And I'll argue that
2	this provision has not accomplished its purpose, and
3	that we can draw some lessons from that experience.
4	So, I want to first describe what I take to be
5	the principles of Section 2 remedies, I'll then suggest
6	that most of the provisions in the Microsoft judgments
7	adhere to these principles, but that this provision, the
8	protocol licensing provision, departs from the
9	principles and that is part of the reason why it has not
10	been successful.
11	I'll describe briefly how it has been
12	implemented and then in the end I'll try to draw some
13	lessons. And incidentally, this is a very brief summary
14	of a much longer article which I hope to post on SSRN
15	shortly.
16	The goals of Section 2 remedies should be to
17	restore competitive conditions that would have existed
18	but for the illegal conduct. They should not be to try
19	to restore or to create some sort of ideal competitive
20	condition or to supervise market outcomes. I take the
21	primary antitrust remedy to be deterrence, through fines
22	and covered damages. If deterrence can be effective, if
23	an optimal penalty can be imposed, that's always going
24	to be preferable to having an administrative structure
25	imposing remedies. It's simply the direct costs of

49

1	imposing those remedies will be -- will impose a greater
2	cost than effective deterrence.
3	Assuming that some sort of injunctive relief is
4	required, I would suggest that injunctions should be
5	limited to preventing reoccurrence of proven
6	anticompetitive behavior. The Sherman Act, unlike
7	sector-specific regulation, I believe reflects the
8	assumption that if specific impediments to competition
9	are removed, then private contracting within the market
10	will lead to the efficient outcome. And if that would
11	not be the case, then that would argue that the market
12	should be regulated.
13	Beyond that, I would suggest that injunctions
14	are problematic. First, divestiture, at least in the
15	case of a unitary company, should be a last resort,
16	primarily appropriate to dissolve recent combinations.
17	Regulatory decrees also, as many have observed, should
18	be avoided. As the Supreme Court said in Trinko, they
19	require antitrust courts to act as central planners,
20	identify improper price policy and other terms of
21	dealing in roles for which they are well suited.
22	Most of the Microsoft final judgment provisions
23	reflect these principles. They do not require any form
24	of divestiture, and most provisions respond more or less
25	directly to the liability holdings in the case that were

50

1	affirmed by the D.C. Circuit in 2001, prohibiting
2	retaliation against computer manufacturers for promoting
3	rival software, requiring uniform licensing terms,
4	giving computer manufacturers the flexibility to remove
5	the visible means of access to Microsoft middleware
6	products and so forth.
7	The protocol licensing provision does not
8	respond directly to any illegal conduct. Server-based
9	applications were mentioned in the findings of fact,
10	only to exclude them from the market.
11	Interoperability in networks was not an issue in
12	the case, and in fact developing and refusing to license
13	incompatible proprietary software was not held illegal,
14	in fact, it was specifically held to be legal, if
15	nothing more than that were shown.
16	So, where did this come from? The idea for this
17	provision actually arose, according to Ken Alletta's
18	book on the Microsoft litigation, after the findings of
19	fact had been issued. In other words, after the record
20	was closed in the case. The feeling was that Microsoft
21	essentially was not going to continue the conduct that
22	was actually the subject of the litigation, the browser
23	wars were over, Microsoft had already stopped the
24	discriminatory pricing, it had gotten rid of the
25	exclusive terms in its contracts, so we needed to be

51

1	more forward looking and what was forward was this
2	network environment.
3	The fear was that in this -- you've got to, you
4	know, as the computer market moved toward networks, both
5	local corporate networks and the Internet, it was
6	necessary to assure that Microsoft would not
7	discriminate in allowing rivals to interoperate with the
8	dominant Windows client.
9	And, so, various proposals for various
10	interfaces by Microsoft were made. After the original
11	judgment was reversed, of course the Antitrust Division
12	reached an agreement with Microsoft on the consent
13	decree and it included a version of this. The protocol
14	licensing provision, which essentially we now have, in
15	both that consent decree and in this -- the states'
16	judgment.
17	Judge Kollar-Kotelly approved this provision,
18	even though she recognized that the government was not
19	strictly entitled to it, because it was not responsive
20	to proven illegality, and she also recognized that there
21	were these other ways in networks of achieving
22	interoperability besides requiring Microsoft to license
23	its proprietary protocols.
24	Nevertheless, she found that -- and here's the
25	key language, it's closely connected to the theory of

52

1	liability in this case, and furthers efforts to prevent
2	future monopolization.
3	So, under this program, Microsoft has developed
4	the Microsoft communications protocol program, which is
5	an extension of its Microsoft developers network, and
6	under this program, it offers a license to these
7	protocols, and technical documentation. In the initial
8	response in August 2002, actually before the consent
9	decree was approved, but nine months after it was
10	originally agreed to by the parties, Microsoft produced
11	5,000 pages of technical information, documentation, on
12	the protocols, which it reported with a product of the
13	work of five technical writers working essentially
14	full-time for nine months.
15	By July 2003, however, eight months after the
16	entry of the final judgments, only four developers had
17	licensed these protocols. And Judge Kollar-Kotelly told
18	the parties in a status conference, this is reported in
19	the report, that she was very, very concerned that
20	nobody was taking these licenses. And both Microsoft
21	and the government responded to this by various efforts
22	to promote them. Microsoft took out ads, they
23	evangelized these protocols, but with very little
24	success. And finally the government conducted a survey
25	of developers asking them why aren't you licensing this

53

1	material, and they gave a list of reasons, some of which
2	focused on the license itself, said it was way too
3	complicated, it was pages of technical terms, and they
4	were too expensive, the technical documentation was
5	insufficient, the royalty was too high, whatever. But
6	some said, we just don't need them for our development
7	efforts.
8	All of these, except that last one, were
9	addressed over the next three years. The license term
10	has been extended, the limitations in it have been
11	relaxed, and simplified, the royalties have been
12	reduced, many of the open standard protocols that
13	Microsoft supports have been made available under the
14	royalty free license. Microsoft has made its source
15	code available to licensees.
16	Now, to become a licensee, you need to show you
17	have a legitimate purpose. So, you can't go and ask to
18	see the source code, but if you are a licensee and you
19	can show that you have need for it, under the license,
20	then they'll show it to you and they'll actually provide
21	support to show you how to use it. It's also provided
22	500 hours of free premier technical support, it's
23	provided a dedicated account manager, it's provided
24	three-day, what they call plug fests, where you can
25	bring your product and test it and Microsoft engineers

54

1	will work with you to try to make sure it interoperates
2	well with Windows. It's created an interoperability
3	lab, and I should mention, when we had the first plug
4	fest, only two licensees signed up for it, no one has so
5	far signed up for the interoperability lab.
6	So, over the years, what's most dramatic about
7	these status reports is the accounts of how Microsoft
8	and the technical committee have tried to improve the
9	technical documentation of the protocols.
10	In July 2004, the technical committee and
11	Microsoft agreed on a 40-page specification that the
12	documentation was supposed to meet. And the technical
13	committee undertook to develop what it calls prototype
14	implementations of each protocol. There are about 100
15	and 120 protocols, and in order to assure that the
16	documentation of them was sufficient, the technical
17	committee has undertaken to try to actually write a
18	little application using the protocol.
19	And, so, if they could do that, then that would
20	show that the documentation, it could actually be put
21	into effect by the developer. Where they run into
22	problems, if they ran into problems, they treated that
23	as an issue, and they reported that to Microsoft as a
24	bug to be addressed, and depending on its importance,
25	they gave them seven days or, you know, longer time

55

1	limits to respond to it.
2	And this was the approach for about a year, but
3	by early 2006, the technical committee had reported to
4	Microsoft about a thousand of these issues, and only
5	about 300 of them -- 300 of them had been resolved, and
6	in May, this is about a year ago, the plaintiffs
7	reported to the judge that the project had reached what
8	it called a watershed, and at that point, someone who I
9	take to be a strong personality, Robert Muglia, who is
10	the senior vice president of Microsoft and formerly was
11	the head of server division, reviewed this program and
12	said that this process of trying to respond to bugs one
13	by one, as they're reported by the technical committee,
14	was just not working, and that we would need to start
15	from scratch and rewrite all of the technical
16	documentation.
17	And, so, last summer, incidentally, it was at
18	this point that the technical committee made contact
19	with the European Commission's monitoring trustee, which
20	is also administering an order to Microsoft to disclose
21	protocols, and in connection with those communications
22	had with Microsoft, agreed on a new overarching
23	specification. This is now the third standard that will
24	be used to judge the documentation.
25	And Microsoft was given a new set of milestones,

56

1	time tables, to complete the project. At this point, it
2	was clear that the decrees were due to expire in the
3	fall, and it was pretty clear that that was not going to
4	be enough time to do all of this, and so that's when the
5	parties agreed to extend the term of the judgment for up
6	to five years.
7	Meanwhile, Microsoft has suspended royalty
8	payments entirely for its licensees, until the
9	documentation is deemed to be sufficient, and the
10	technical committee has continued to develop these
11	protocol implementations, and interestingly, Microsoft
12	has also undertaken to do something similar, developing
13	what they call test suites, which it's one of the
14	practices of software developers when they're working on
15	an application, they come up with suites of testing
16	applications to see if they work, and Microsoft has
17	undertaken sort of a parallel or duplicate testing
18	mechanism.
19	And in this most recent status report, which was
20	issued earlier this month, the plaintiffs reported that
21	although they've had some questions about Microsoft --
22	apparently Microsoft discovered some new protocols that
23	they hadn't identified before, they said that this new
24	documentation is looking better, although significant
25	additional work needed to be done.

57

1	So, Microsoft now has been -- remember the first
2	project, it had a few technical writers working for a
3	certain number of months to produce these 5,000 pages.
4	They now have 313 employees working on this project.
5	And the technical committee also has increased its staff
6	to 40 engineers, and they now have offices both in
7	Redmond, Washington and in Silicon Valley.
8	The bottom line, as of this month's status
9	report, of the thousands of developers writing
10	applications for servers, for server operators, to run
11	on server operating systems, only 27 firms have taken
12	the royalty-based license, and all but four of these,
13	but for very specific purposes, like media streaming or
14	data storage or security, the proxy firewall segment.
15	So, and of those 27, only 14 are producing any products.
16	And none of these products seems likely to have any
17	potential as a platform.
18	So, what are the lessons from this experience?
19	The original rationale for this project was to preserve
20	the middleware threat to the Microsoft monopoly in the
21	network environment. If so, at least so far, the
22	project has not succeeded, because it's attracted very
23	few licensees, despite these enormous efforts, and I
24	think quite admirable, and impressive efforts on both
25	sides.

58

1	What this suggests to me is that the primary
2	reason why we're not seeing more licensees is that
3	licensing Microsoft's proprietary protocols is generally
4	not necessary for these firms to develop software
5	applications to run on non-Microsoft servers. They can
6	use the standard protocols that Microsoft supports in
7	Windows, or they can develop their own windows client
8	which then could run on the Windows client and
9	communicate directly through Microsoft's application
10	programming interfaces.
11	So, to boil it down, what I would say is that
12	what this remedy does is to treat the Microsoft
13	protocols as if they were an essential facility, except
14	that they're not essential. There are other ways of
15	accomplishing the same thing.
16	So, what I would take to be the two primary
17	lessons are first, injunctive relief, particularly in
18	high technology markets, should be limited to responding
19	to a proven need, and the most important proven need is
20	to -- is to interdict and remove anticompetitive
21	practices, proven anticompetitive practices.
22	So, if Microsoft is proven to have engaged in
23	practices that violate the antitrust laws, those should
24	be enjoined. But as we've seen, the protocol licensing
25	provision did not respond to a proven violation, and did

59

1	not even address technology -- and it addressed
2	technologies that were not even the focus of the
3	liability phase.
4	During the remedial proceedings, there was a
5	record developed on network computing and there was
6	evidence introduced of various so-called bad acts, as
7	Judge Kollar-Kotelly characterized them, but she treated
8	them as being essentially irrelevant, because they had
9	not been shown to be anticompetitive, or at least if
10	they were anticompetitive, they may have had
11	pro-competitive justifications that had not been
12	considered.
13	The second, under this heading of only
14	responding to a proven need, I don't want to rule out
15	the possibility that forward-looking or fencing in kinds
16	of provisions may be necessary, but if they are, then I
17	think there should be -- there should be a record built
18	to support the need for them. And I think in this case,
19	for example, we know that the government at one point
20	actually surveyed software developers to see what their
21	needs were in this area.
22	I'm not sure what was done during the
23	negotiation of the consent decree, but perhaps more in
24	that direction could have been done to find out
25	precisely what was needed to ensure adequate

60

1	interoperation.
2	And also I would just add that the Court of
3	Appeals in the 2001 decision cautioned that remedies
4	should be proportional to the strength of the proof that
5	Microsoft's illegal actions actually reduced
6	competition, and that was why the Court of Appeals said
7	that divestiture was probably not going to be an
8	appropriate remedy, because as they put it, the harm to
9	competition for Microsoft's actions, in other words,
10	whether they had actually prevented Netscape's browser
11	or Java from evolving into a rival platform, that was
12	established by only -- as they put it -- by inference,
13	in other words, there was no evidence that that actually
14	would have happened. And where you have that relatively
15	weak evidence of likely anticompetitive effect, then you
16	need more evidence to support more Draconian remedies.
17	And divestiture is certainly that, but I also
18	think regulatory relief is also a Draconian remedy, and
19	that brings me to my second lesson, and that is to avoid
20	regulatory decrees, especially in high technology
21	markets. And this was recognized, Judge Kolar-Kotelly
22	rejected one principle during the remedial proceedings,
23	on the grounds that it would result in too regulatory of
24	a decree.
25	Well, the protocol licensing has become highly

61

1	regulatory and direct government supervision of price
2	and other terms of dealing and especially quality.
3	Direct government supervision of quality that's being
4	produced. And the device of the technical committee
5	certainly has provided a high level of expertise, but in
6	effect, what its created is a regulatory body, and I'm
7	not sure that the structure of the technical committee
8	and its relationship to the plaintiffs and the court
9	establishes an effective regulatory agency.
10	So, just to conclude, if in the future cases
11	have these characteristics, those should be treated as
12	warning signs, and addressed in the -- in the relief.
13	And with that I'll sit down.
14	(Applause.)
15	MR. HILLEBOE: Thank you, Bill. Marina Lao is a
16	professor of law at Seton Hall Law School. She
17	currently serves on the executive board of the section
18	on antitrust law of the American Association of Law
19	Schools, and she's an alumna of the Antitrust Division,
20	where she was a trial attorney. She has published
21	numerous articles on antitrust law and trade regulation,
22	and somewhat surprisingly on this high-tech panel, she
23	is the only speaker with slides.
24	Marina?
25	MS. LAO: I guess it's even more surprising

62

1	given that I am usually the least high-tech person on
2	the panel. Thank you very much for inviting me and I'm
3	happy to have the opportunity to participate in this
4	hearing.
5	I agree with a number of the speakers who have
6	gone before me who have said that remedies are often
7	treated as an after thought. Unfortunately, that's not
8	a very good idea, because success in proving liability
9	often does not translate into success in remedying the
10	anticompetitive situation, and so it's often best to
11	work your vision of remedy into the case development
12	much earlier on.
13	What I'm going to do, since I'm bringing up the
14	rear, is to try not to overlap too much with what has
15	been said; I'm going to focus on three main points in my
16	comments and I will be skipping over some of the slides.
17	First, where network effects are substantial in
18	the industry that's affected by Section 2 violation, I
19	probably differ from Bill, in that I think that there's
20	a need for broader rather than narrower remedies for
21	some of the reasons that I'll talk about later.
22	Second, again, I guess on this issue I differ a
23	bit from Bill as well. I'm going to talk about the
24	importance of forward-looking remedies. I would call
25	them affirmative remedies that reduce rivals' costs and

63

1	some of the problems in crafting them. I do agree that
2	tayloring these remedies to the problem is a bit
3	difficult.
4	And lastly, I'm going to discuss whether there's
5	any value in bringing Section 2 enforcement action if
6	there is no effective judicial remedy. My conclusion is
7	that there is deterrent value to bringing an enforcement
8	action, even if it is irremediable, so to speak.
9	Let me start with a few words about the ongoing
10	debate among antitrust commentators on the application
11	of antitrust in the dynamic high technology markets.
12	The question that is often raised is: Do we need more
13	rigorous antitrust enforcement or do we need a more
14	hands-off approach? Those who say that less
15	intervention is necessary generally argue that because
16	there is rapid innovation, product cycles are short, and
17	so dominance is fleeting. And there are continuous
18	opportunities for fringe firms to overtake the
19	incumbent. The Microsofts of the world will have to
20	constantly innovate or they're going to be left in the
21	dust.
22	And so for that reason, there's really not that
23	much of a need for antitrust intervention in order for
24	markets to remain robust. In fact, too much antitrust
25	intervention could stifle innovation and competition.

64

1	While there's obviously some truth to that
2	argument, I think the Microsoft case itself tells us
3	that rapid technological change can cut the other way,
4	especially when you have substantial network effects
5	which tend to operate as significant barriers to entry.
6	If these are substantial network barriers to entry, a
7	clearly dominant firm can much more easily exclude even
8	superior technologies, up to only a certain point, of
9	course, if it can ensure that the rival technologies
10	remain incompatible.
11	And, the dominant firm can also control research
12	avenues, up to a certain point. What's more, even
13	without any antitrust violations, there are natural
14	benefits, that flow from network effects of those
15	natural benefits, I think dominant firms can more easily
16	use tying and other exclusionary strategies to preserve
17	their dominance and to exclude competitors
18	anticompetitively.
19	So, my conclusion is that antitrust intervention
20	is not only not redundant, but there is perhaps an even
21	stronger need for it when you have markets with strong
22	network effects.
23	With respect to remedies, there's a similar
24	ongoing debate among commentators. There are those who
25	say that with fast moving technologies, you need milder

65

1	remedies, remedies that are less severe, because of
2	several reasons. First, there is the self correcting
3	market rationale, which postulates that the market is
4	going to correct itself much faster than antitrust
5	intervention can correct it. Second, advocates of mild
6	remedies warn of the possibility of unintended
7	consequences, that is where market conditions in the
8	future are uncertain, one may not know what to prohibit
9	and what not to prohibit, and so the remedies adopted
10	today may not be sensible a few years hence.
11	And, so, they argue it is probably safer to
12	adopt milder forms of remedy in order to lessen the risk
13	of chilling innovation and competition from the dominant
14	firm.
15	First of all, I happen to think that high-tech
16	markets do not that easily, at least self correct, not
17	if network externalities exist, because by definition, a
18	self correcting market, requires innovation and new
19	entry, but network effects raise entry barriers and
20	reduce access to the network.
21	Obviously easy entry markets are not going to
22	easily self correct.
23	As to the argument that uncertainty about future
24	market conditions means that we should perhaps take a
25	more hands-off approach and apply the mildest remedy

66

1	possible, I also do not completely agree with that. I
2	think that if market conditions are uncertain, we have
3	to exercise more care in defining the future boundaries
4	of the relevant market, and in identifying the
5	participants in this future market, and in crafting the
6	remedy.
7	But we should not overlook the danger of doing
8	too little too late, which carries its own risk as well.
9	Another possible solution to the uncertain market
10	condition problem is to have a continuing jurisdiction
11	clause in the remedial order, which I know is not a
12	common practice. With a continuing jurisdiction clause
13	either party can go back to the court for modification
14	if it turns out that the remedies agreed upon do not
15	work because of changing market conditions.
16	As to the "potential chilling effects" argument,
17	it's often said by advocates of milder remedies that
18	compulsory licenses of IP rights and other affirmative
19	remedies tend to chill innovation on the part of the
20	dominant firm, that's basically one of the points
21	Justice Scalia made in Trinko.
22	What is often lost in this discussion, though,
23	is that competition and innovation from fringe firms are
24	also very important, and if remedies for an antitrust
25	violation are insufficient, innovation and competition

67

1	from fringe firms could be chilled. The AT&T
2	divestiture experience is very instructive. Few would
3	disagree that the structural remedy in the AT&T case
4	unleashed innovation from smaller telecommunications
5	firms on an unprecedented scale, which enhanced consumer
6	welfare.
7	Another point that we should not lose sight of
8	is that with high technology markets, it's extremely
9	difficult to resuscitate a competitor, after the
10	competitor has been crushed. The convergence of factors
11	that produced a competitive challenge before it was
12	anticompetitively excluded, may never re-appear, not in
13	the same fashion, anyway.
14	The factors together call for a solution that is
15	less hands-off.
16	They also lead me to conclude that narrowly
17	focusing the remedy on the specific conduct found to be
18	unlawful, will not return competition to the status quo;
19	thus drafting or crafting forward-looking remedies is
20	quite important.
21	Of course I do realize that forward-looking
22	remedies have to be carefully tailored.
23	The problem one faces in crafting
24	forward-looking remedies is that you have to understand
25	the market. You've got to analyze the likely evolution

68

1	of the market, predict which way the market is headed,
2	the innovations will likely emerge, what will be the
3	next generation of innovations, and how these
4	innovations might change the path of the market.
5	Unless you have a pretty good grip on these
6	issues, it's very difficult to predict what remedial
7	actions would work to break down entry barriers and
8	facilitate competition, and what would not.
9	If we do not know what is going to work, then we
10	risk adopting an injunction that constrains conduct that
11	no longer needs to be constrained, but does not
12	constrain conduct that needs to be constrained. Perhaps
13	the prime example of this is the first Microsoft consent
14	decree, which prohibited Microsoft from "per processor"
15	licensing which it had engaged in. But by the time of
16	the decree, Microsoft no longer needed to engage in that
17	strategy, because its competitors in the operating
18	systems market were already defunct and the prohibition
19	accomplished nothing.
20	Another problem, I think, that is rather
21	peculiar to high-tech markets is having to anticipate
22	how dominant firms might circumvent the judicial
23	constraints imposed and still achieve their
24	anticompetitive ends, and then block these alternative
25	paths in the in the decree as well. Fast-changing

69

1	markets tend to be pretty malleable, thus giving the
2	dominant firm myriad ways to achieve its anticompetitive
3	objective.
4	To understand how Microsoft or any dominant firm
5	might sidestep an injunction and still achieve its end,
6	we need to know what the possible alternative strategies
7	are. But dominant firms generally have an information
8	asymmetries advantage over the government that's quite
9	natural.
10	That is, the government knows much less than the
11	dominant firm about what the potential new innovations
12	and the possible alternative strategies to achieving the
13	anticompetitive objective are. So how can the
14	government overcome the information asymmetries problem?
15	I think the simplest solution is to just enlist the
16	assistance of the dominant firm's competitors or
17	potential competitors, who probably are in a much better
18	position than any outsider, including government
19	enforcers, to know about the industry, to know what
20	remedies might work and what might not work, and what is
21	the innovation trend, et cetera.
22	Oftentimes, when this is mentioned as a possible
23	solution, you hear the argument that, well, then, the
24	department or agency might be subject to capture. I
25	think that simply relying on competitors to educate

70

1	government enforcers on the market is not equivalent to
2	capture, and is also entirely consistent with the
3	principle that we should protect competition and not
4	competitors.
5	Let me turn, briefly, to the importance of
6	implementing creative affirmative obligations. The
7	problem with conduct remedies and I'm not discussing
8	structural remedies at all, because it's been discussed
9	in detail already is that generally speaking, if the
10	dominant firm has already successfully excluded its
11	competitor and potential competitors, simply stopping
12	the conduct and preventing its recurrence is not going
13	to be enough to restore competition. That is because
14	stopping the exclusionary conduct will not unravel the
15	dominant firm's accumulated market power.
16	Instead, what would be helpful would be to
17	impose affirmative duties on the dominant firm. I call
18	it lowering rivals' cost as opposed to raising rivals'
19	cost. The Post-Chicago school has said that dominant
20	firms can exclude competition anticompetitively by
21	engaging in strategies that raise rivals' costs. For
22	remedy purposes, we need to go a little bit beyond
23	prohibiting acts that raise rivals' costs; we need to
24	impose some obligation on the part of the dominant firm
25	to reduce rivals' costs.

71

1	Some affirmative duties are pretty well
2	established in antitrust jurisprudence, and are not very
3	controversial.
4	One is compulsory licensing of IP rights, with
5	or without royalty fees. The case that springs to mind
6	involving forced licensing is the Xerox case brought by
7	the FTC in 1975. The FTC in that case imposed a
8	compulsory licensing obligation on Xerox. In Microsoft,
9	as Bill just mentioned, there was also a compulsory
10	disclosure of information component in the decree as
11	well Microsoft was required to disclose its APIs and
12	also its communications protocol.
13	Another typical affirmative duty is the
14	obligation to sell to all customers on a
15	non-discriminatory basis, and that was part of the order
16	in the Ninth Circuit Kodak case.
17	The third example that I have listed on the
18	slide is also not terribly controversial, and that is
19	unbundling. For example, in United Shoe, the defendant
20	was required to unbundle its machinery and its repair
21	service.
22	The fourth category is probably the most
23	controversial, and that is requiring the defendant to
24	create products to comply with industry standards and
25	not just with its own proprietary standard. This is the

72

1	remedy that the State of Massachusetts asked the court
2	to impose in Microsoft, in the case that Massachusetts
3	continued to pursue after Microsoft settled with the
4	DOJ. Incidentally, the District Court did not grant
5	that request.
6	I was going to talk about the Korean Microsoft
7	case, which I found very interesting, but I don't think
8	I will have time for that, so let me just end with two
9	points. I have alluded to the first point earlier, and
10	that is the usefulness of a continuing jurisdiction
11	clause in a remedial order. Perhaps those of you who
12	are still in government can enlighten me as to why the
13	government does not seem to want to include these
14	jurisdiction clauses in their remedies anymore, back in
15	the 1950s and 1960s.
16	Having a continuing jurisdiction clause is
17	helpful in a dynamic high technology market because it
18	allows the court to assess the success of the remedy,
19	and to assess future development. The purpose of
20	assessment is not so much to ensure that strict
21	compliance with the decree itself is occurring, although
22	that is very important too, but to ensure that there's
23	movement toward the ultimate objective set by the court.
24	I think Professor Hovenkamp in one of his articles
25	suggested that perhaps a continuing jurisdiction clause

73

1	would be very, very helpful, because it would allow the
2	court to look at whether the decree has been successful
3	or not. I think of success as not simply whether the
4	defendant has complied with the specific terms of the
5	decree, although that is obviously a part of it, but
6	whether the decree is doing anything at all to make the
7	market more competitive.
8	One final note, and that is I think there is
9	value to Section 2 enforcement even if no effective
10	judicially-imposed remedy is available, on two
11	conditions: if there is really an egregious violation
12	of the antitrust laws, and if there is substantial harm
13	to consumer welfare. The reason enforcement is
14	important even if the violation is judicially
15	irremediable is that I think the defendants would
16	moderate their behavior somewhat, simply because
17	litigation has been brought. And they may even
18	voluntarily discontinue some of the challenged
19	practices.
20	I think it is commonly acknowledged and commonly
21	known that Microsoft relaxed enforcement of its
22	exclusive dealing contracts with the OEMs during the
23	process of the litigation. And, as far as I can tell,
24	Microsoft does not seem to be using against the type of
25	tactics that it had engaged in against Netscape and

74

1	Java.
2	I am not a very tech savvy person, but it would
3	seem to me that there must be strategies similar to the
4	kinds that Microsoft had employed against Netscape and
5	Java, and yet they have not engaged in them against
6	Google. Of course we will never know how much of their
7	reticence is the result of the deterrent effect of the
8	government's enforcement action.
9	Finally, for public policy reasons the
10	government should not just step back and say, well,
11	there is no effective remedy, so what's the point of
12	bringing a lawsuit? If consumer harm is substantial,
13	and if the act is egregious, I think it is bad policy to
14	take no action because it sends a wrong signal. Taking
15	enforcement action can deter the Microsofts of the
16	world. Who knows, it might deter Google at some point.
17	With that, I hope I haven't repeated too much of
18	what has been said.
19	(Applause.)
20	MR. HILLEBOE: Thanks, Marina. This is the
21	portion of the hearing where we allow each of the
22	speakers to comment with what they've heard before, and
23	I'll start with Howard, please.
24	MR. SHELANSKI: Well, I thought a number of the
25	presentations raised provocative, extremely provocative

75

1	issues.
2	Let me start with Michael Cunningham's comments
3	about the problems that companies like Red Hat still
4	face, even in the wake of the decree.
5	I found his comments extremely interesting,
6	because they suggested both at the same time a need to
7	be very aggressive against anticompetitive behavior,
8	because it has lasting effects, but also to raise real
9	questions about what can be done about those effects,
10	and if one were to translate that into a recommendation
11	about remedies, it would be hard to know -- it would be
12	hard to know exactly what the result is.
13	On one hand, it might be taken to suggest that
14	we need very aggressive kinds of remedies of the kinds
15	that Professor Lao just suggested, with continuing
16	supervision, and more creative solutions to lowering
17	rivals' costs.
18	On the other hand, I think that Bill Page raised
19	very good reservations that I share about pursuing that
20	kind of aggressive oversight.
21	So, where I come out from Michael's comments is
22	to say that we do need to pursue these cases. We need
23	to pursue these cases to understand what kind of conduct
24	is likely to lead down the road to problems that are
25	very hard to uproot. And in concert, I think, with what

76

1	Professor Lao just suggested, even if we're not sure
2	that the remedy will work, pursue the case so that next
3	time around, we can uproot the conduct earlier and have
4	a remedy that will be effective, but I think, Michael,
5	you pointed to some really very difficult challenges.
6	With regard to Renata Hesse's comments, I think
7	I shared very, very much your point of view. I think
8	you were a little bit more cautious about the likelihood
9	of success of injunctive remedies, I thought you raised
10	some very good points there, but I continue to think
11	that particularly in the high-tech sector, injunctive
12	remedies will take the form of a negative prohibition of
13	thou shalt not are likely to be the most fruitful
14	remedial avenue overall.
15	Professor Page, I found that story fascinating,
16	but I think the detail was extremely instructive, and
17	very helpful. And I guess on one hand, I might be
18	inclined to say, well, does that mean we shouldn't go
19	deep into these kinds of continuing remedies; on the
20	other hand, I might say, well, maybe this is very costly
21	to Microsoft, with little benefit to competitors, but
22	maybe costly to Microsoft in and of itself, isn't so
23	bad.
24	But maybe costly to Microsoft in and of itself
25	isn't so bad. Maybe it's a very back-handed form of

77

1	disgorgement remedy through the front door.
2	I say that partly tongue in cheek, because I
3	don't know that they really notice that kind of spare
4	change over there.
5	(Laughter.)
6	MR. SHELANSKI: No, but it does raise some very
7	serious questions about how even the most carefully
8	wrought and technologically sophisticated attempt at an
9	affirmative remedy can be very difficult, and that's a
10	lesson that I take very much to heart. So, I've learned
11	a lot from all of you. Thanks, very interesting.
12	MR. HILLEBOE: Thank you very much, Howard.
13	Renata?
14	MS. HESSE: Sure. I think -- I don't think the
15	mic' is on. I think the thing that I took away from
16	everyone's comments was very similar to what Howard just
17	said, was that there seems to be a sort of inherent
18	conflict between these two views of both the difficulty
19	and in some cases I think impossibility of imposing
20	remedies in technology markets, and yet at the same time
21	the view that we really need to keep trying, even though
22	we're not likely to be successful.
23	And I haven't come up with a good way of
24	bringing those two points of view together, other than
25	to say that I think, you know, courts, and not in the

78

1	antitrust context, but in lots of other contexts, have
2	over the years dealt with a lot of very difficult
3	issues, which people, I think, over time, have thought,
4	well, you know, how could a court ever figure out how
5	to -- I'll use, you know, prison conditions litigation,
6	which I think I talked about before, you know, school
7	desegregation is another one.
8	Difficult problems that are not within the core
9	competency of either courts or lawyers, and everybody, I
10	think, has thought that a social benefit derives from
11	intervention in those areas, and at least an attempt to
12	try to solve them in some way.
13	And I don't really see technology markets as
14	being different in any -- I mean, they're obviously
15	different in terms of the substance that they deal with,
16	but not different in terms of the importance of the
17	issues that you're dealing with, in terms of the
18	importance of markets to both not just America's
19	economy, but the world economy, and to the every day
20	consumer. I mean, these products and services are
21	things that we all use on a daily basis, and spending
22	time thinking about, A, whether or not the law is being
23	violated in those areas, and B, if it is being violated,
24	how can you do the very best job you can to try and
25	solve the problem seems to me to be a worth while

79

1	expenditure of not only government time, but also in
2	some cases in private litigation time, too.
3	Keep at it, I guess, is my final conclusion.
4	MR. HILLEBOE: Michael, and also I would ask you
5	to address your points of the speed and cost of
6	antitrust litigation are duly noted. If you have any
7	profound suggestions with respect to those or practical
8	suggestions or any other type of suggestions.
9	MR. ELIASBERG: Or those quick and speedy cases,
10	I was very interested in that.
11	MR. CUNNINGHAM: Right, profound thoughts
12	probably won't be forthcoming, but I will try and offer
13	a couple. I take a pretty simple approach as a business
14	person. I have a difficult problem, I keep working on
15	it and keep attacking it until I come up with a
16	solution.
17	I think, you know, serious examination of the
18	effects of the Microsoft remedies is worth while, but
19	there is assuredly deterrent value. One part of the
20	advice that I tell my client, which I didn't mention
21	before, is that I believe it assuredly moderates
22	behavior for us to have any participation and then for
23	the case to be brought at all.
24	Indeed, in the area of some of the protocols
25	that have been licensed that Bill referred to, I deeply

80

1	wonder whether Microsoft would have reached out to Red
2	Hat and requested our assistance and consultation in
3	producing a very, very simple protocol license that's
4	one page, we'll never know the cause/effect of both the
5	EU action and the U.S. action, but there's reason to
6	think that some of that may moderate behavior.
7	I think in the case of Bill's examination, also,
8	I would just comment that continuing to look at those
9	facts are important. For example, Bill pointed out that
10	there are other ways to interoperate. Other ways to
11	interoperate that are fundamentally disadvantaged is not
12	interoperation. It doesn't work.
13	The IT community, you know, competes on the
14	speed, efficiency, and look and feel of interoperation.
15	So, simply concluding that there may be other protocols
16	out there that may have issued since the decree, at
17	least some of them, may not be complete examination. I
18	should point out, Bill was kind enough to provide me a
19	draft of his entire paper, which I didn't have a chance
20	to look at before, so if it's addressed in the paper, my
21	apologies.
22	I think that, you know, these are terribly hard
23	problems to work on, and I just don't see where, without
24	learning and gaining experience in how to better address
25	conduct remedies, we're able to make effective inroads

81

1	into some of these fast-moving markets.
2	MR. HILLEBOE: Bill?
3	MR. PAGE: I just have a few kind of stray
4	comments. I was struck by Renata's point about focusing
5	on a remedy early, and I agree that that is really
6	critical, and I would suggest that particularly in a
7	case that ends in a consent decree, before litigation,
8	it's absolutely essential.
9	What I -- part of the problem I saw in the
10	Microsoft remedial issue was that the case lasted so
11	long that it was a moving target to think about the
12	remedy, you know, that at -- that by the time the case
13	was over, the remedy that people wanted was different
14	from the one they would have predicted early in the
15	litigation.
16	So, you know, particularly for cases that last
17	longer than just a couple of years, it's particularly
18	difficult to be sure the remedy from the outset and be
19	building a factual basis for it.
20	I think the point about avoiding mandatory kinds
21	of remedies as opposed to prohibitory remedies is a
22	valid one. I would just caution, though, that in the
23	Microsoft case, there was another mandatory remedy to
24	reveal the APIs that Microsoft uses to interact with its
25	middleware, between the Windows operating system and its

82

1	middleware, and that one seems not to have caused that
2	many problems. And I suspect that the reason for that
3	is that Microsoft's whole business is marketing APIs,
4	and documenting APIs. If they couldn't do that, they
5	wouldn't be in business.
6	So, that was a much more straightforward problem
7	than marketing protocols, their own proprietary
8	protocols, and I think that's, you know, perhaps that
9	explains some of the difficulties that have been found
10	in documenting that.
11	So, not all mandatory types of relief will
12	necessarily be as problematic as this one. On the issue
13	of the technical committee, I want to combine this with
14	the idea that the courts should retain jurisdiction, and
15	periodically review the experience in enforcement. The
16	technical committee I think is one institutional concern
17	that I have about the technical committee, certainly
18	they are quite expert. I know nothing about them
19	individually, but certainly no one would challenge their
20	technical capacity, but they were given a single task,
21	and that was to assure that the documentation is first
22	rate, flawless. And, you know, as Howard pointed out,
23	who cares how much Microsoft pays, to do that, and so
24	it's a very expensive process to meet that kind of
25	standard.

83

1	On the other hand, I think at some point, the
2	court should come back and ask the question, is this
3	accomplishing as much as we could accomplish in other
4	ways? In other words, the economic question is always
5	compared to what? And particularly if we can
6	preemptively think about these issues before they come
7	up, but also, if we can think about them down the road,
8	perhaps as an opportunity for mid-course corrections
9	that could reduce costs and perhaps benefit the market
10	better.
11	Just finally, on the issue of whether high
12	technology markets require or it's more appropriate to
13	use remedies in them because of network effects, I would
14	only caution that the literature on network effects
15	doesn't exactly say that competition doesn't work in
16	these markets. It doesn't necessarily say that network
17	effects are bad, I mean, when you think about it,
18	network effects are simply economies of scale on the
19	demand side. In other words, they benefit consumers,
20	and so the concern that they are simply a barrier to
21	entry I think somewhat overstates the case.
22	Markets converge on a single standard for
23	reasons that are actually beneficial to consumers. It
24	doesn't necessarily follow, then, that government
25	intervention is necessary, and I would add to that the

84

1	issue of compatibility is also not so simple, because
2	markets characterized by network effects can sometimes
3	compete very effectively with totally incompatible
4	systems, as we observed in the video game console market
5	where, you know, it's a constant leapfrog competition of
6	totally incompatible systems of hardware and software.
7	And that is a very effective model for competition.
8	So, it doesn't necessarily follow that we should
9	be promoting interoperability in all circumstances.
10	MR. HILLEBOE: Marina?
11	MS. LAO: I actually only have a few comments.
12	I think the presentations today highlight the
13	difficulties involved. For instance, Bill's
14	presentation focused on the problems that I had tried to
15	shy away from, and that is there are major difficulties
16	in using and implementing forward-looking remedies.
17	And Michael's points, I think, drive home the
18	need, for more active government intervention, because I
19	think private Section 2 cases are extremely difficult to
20	prove, especially since proving anticompetitive effects
21	now often requires economic proof. When the violation
22	involves technology that hasn't fully emerged yet, it's
23	very difficult to show that there is actual
24	anticompetitive effect. I pretty much agree with most
25	of what Renata and Howard said.

85

1	MR. HILLEBOE: Okay, thank you.
2	Bill, just as a point of clarification, I think
3	you had indicated that Microsoft was licensing its
4	source code. Just to clarify that, I think you probably
5	mean it's licensing portions of its source code that are
6	associated with interoperability issues. Is that
7	correct?
8	MR. PAGE: It's allowing licensees of the
9	protocols access to the source code in order to help
10	them use the protocols.
11	MR. HILLEBOE: Right, but not the crown jewels,
12	so to speak?
13	MR. PAGE: No, they're not saying here's our
14	source code, you can use it, you know, for whatever
15	purpose, it's purely to assure -- there were some of the
16	licensees, or prospective licensee who said that they
17	really needed access to the source code, more than they
18	needed the specification of the protocols. And I'm not
19	enough of a geek to know why that would be, but this is
20	in response to that.
21	And interestingly, that is an important
22	concession, I would say, on Microsoft's part, because
23	that was one of the proposed remedial provisions that
24	the non-settling states wanted to have added to the
25	final judgment was to require Microsoft to disclose its

86

1	source code for these purposes, and the court refused to
2	order that.
3	And, so, in this limited sort of disclosure, I
4	think is an important concession.
5	MR. HILLEBOE: And several folks have talked
6	about technical committees, and I wanted to direct a
7	question to Renata about that, since she's had a lot of
8	experience with that. I was wondering, Renata, if you
9	can offer us some insights with respect to setting up
10	the technical committees, given that in a conduct
11	remedy, when you're talking about high-tech markets, and
12	given the lack of expertise of lawyers and the fact that
13	we're not engineers, and it seems almost inevitable that
14	you're going to have a technical committee, were there
15	things that you may have changed from the way you did
16	it? Also, are there any differences in the European
17	monitoring trustee? Is that a different situation? And
18	also your thoughts about having all the parties involved
19	in terms of determining who the trustee or the committee
20	should be, including the defendant?
21	MS. HESSE: I'm looking back at Patty Brink, who
22	spent a lot of time with me trying to figure out how to
23	construct the technical committee, and truthfully, it
24	was in terms of the formation of the company, it was
25	like starting a new business. So, we had to work

87

1	through all sorts of issues that you wouldn't ever
2	anticipate, and we certainly didn't anticipate when we
3	thought about the provision, including how do you set up
4	a company so that it doesn't have tax liability, how do
5	you hire employees, how are they paid, all of these
6	things that none of us really knew how to do, and we
7	spent a lot of time consulting with various people to
8	figure that out.
9	The more important pieces of it, though, I think
10	really had to do with the selection of the technical
11	committee members, and if you look at the comments and
12	the response to the comments to the consent decree,
13	there were a number of people who said, whoa, you know,
14	Microsoft gets to pick and gets a role in picking at
15	least one, so the DOJ and the states picked one,
16	Microsoft picked one, and those two people picked the
17	third, and, you know, that's just, you know, they're
18	going to put one of their own people on there, and what
19	good is that really going to do.
20	And I think the interesting thing that happened
21	was that we really did find three people who were not
22	just technical experts, but also had been business
23	people, so people who had started technical companies,
24	and who really knew how to -- not only run the business
25	that they had to run, but also what the business reality

88

1	of the various technical issues that they were advising
2	on.
3	And as it turned out, they really formed a
4	whole, and they worked a lot with Craig Hunt, who is the
5	nonsettling states group, who is sitting out in the
6	audience, also. And they have, you know, coalesced as
7	an entity unto themselves and the Microsoft appointee
8	plays no different role in -- the Microsoft selected
9	person plays no different role than any of the other
10	members. And I think that has been really a tremendous
11	success.
12	I think the things that one would go back and
13	look at again are the provisions in section 4 of the
14	final judgment, which is the technical committee one,
15	that relate to what the technical committee can say
16	publicly and do publicly. And this is always -- and
17	that's a big difference between the monitor trustee in
18	Europe, and the technical committee in the U.S.
19	In the U.S., the technical committee is not
20	allowed to make public statements without prior approval
21	of anybody, and their work product can't go directly to
22	the court. In terms of a compliance or enforcement
23	effort. And I think there were good, reasonable reasons
24	to do that, and I think in the end that's probably the
25	right way to do it, but in Europe, that's not how

89

1	they've done it. And so their monitoring trustee
2	actually will testify at hearings about whether or not
3	Microsoft is in compliance with the final judgment.
4	And those are two very different roles, and I
5	think it's important to think about when you're
6	constructing something like this, which of those two
7	roles you want the person to play. I think having them
8	play both roles is pretty dicy.
9	MR. HILLEBOE: And I know Bill from his comments
10	expressed some skepticism about having a technical
11	committee and having another regulatory body. I was
12	wondering what the other speakers thought about having a
13	technical committee, and if they don't like that idea,
14	if they have some suggested alternatives to that.
15	Howard, do you have any thoughts about that?
16	MR. SHELANSKI: I mean, I think technical
17	committees for the reasons that Bill outlined are likely
18	to be extremely tricky, and so the only thing I have to
19	add is probably what others have said.
20	I think a technical committee should be reserved
21	for circumstances in which we have a pretty clear idea
22	of what needs to be accomplished, a pretty clear idea of
23	the market demand for that outcome.
24	MR. HILLEBOE: Michael, do you have some
25	thoughts about that?

90

1	MR. CUNNINGHAM: Yeah, I personally think that
2	at least if there's going to be a conduct remedy, not
3	having a technical committee would be a fatal flaw. The
4	technology is simply too complex, too subtle and too
5	fast moving to not have, you know, that advice.
6	But turning back to some of Bill's observations,
7	the fact that the technical committee had a thousand
8	comments when they sought to implement the protocols,
9	might suggest a massive failure to comply. And, you
10	know, the fact that the technical committee ran into
11	difficulties, maybe because it's difficult, which is
12	partly true, may be difficult because people were not
13	trying to comply in good faith. I don't know.
14	MR. HILLEBOE: And Bill, did you have some
15	alternatives to having this regulatory body?
16	MR. PAGE: Just on this one last point, before I
17	answer that, most of the status reports do indicate that
18	the technical committee, or the plaintiffs, were not
19	really questioning Microsoft's effort. I mean, there
20	are occasionally comments where they're disturbed by
21	this or they're disturbed by that, but in general, the
22	tone is one of this is a huge job, and we're having
23	problems accomplishing it and we're both trying in good
24	faith to do it. That's in general what I thought from
25	these reports.

91

1	And I should just say that the reports are
2	pitched at a certain level so that there's only so much
3	understanding you can get from them. And maybe if they
4	were any more technical, I wouldn't understand them at
5	all, but I'm a little bit like a denizen of Plato's
6	caves seeing the reflections of reality on the wall and
7	the reality is really outside of the cave and I can't
8	really tell for sure everything that's going on.
9	But to some degree, that is the position of the
10	court, and as Renata said, the technical committee is
11	sealed off from the court, which means that its
12	observations need to be mediated by the lawyers, who I
13	suspect probably don't understand the technical issues
14	much better than I do, and I think that's a problem.
15	I mean, we have this technical body that does
16	understand the issues from a technical point of view,
17	but their antitrust significance has to be mediated by
18	people who essentially don't. And I think that's a --
19	that's a difficulty that perhaps wouldn't be the case if
20	we had a more conventional administrative agency where
21	expertise were, you know, the problems of addressing
22	expertise and using it in decision-making were more
23	formally, you know, implemented.
24	MR. HILLEBOE: Marina, do you have any thoughts
25	on this?

92

1	(No response.)
2	MR. HILLEBOE: Okay. You know, one of the
3	outstanding features of these types of markets that we
4	look for are the presence of network effects, and some
5	people have discussed this, but I think it's important
6	to cover this. Is there a consensus with respect to in
7	markets where you have network effects, are those
8	markets that tend toward monopoly or toward a
9	winner-take-all or winner-take-most equilibrium, or some
10	people have suggested that, or is that overly simplistic
11	or is that a capricious argument. What are your
12	thoughts on that, Howard?
13	MR. SHELANSKI: Well, first let me say that I
14	think that the markets that are truly likely to tip to
15	monopoly are few. I think it's a fairly circumstance
16	where a network market will precipitously tip to
17	monopoly, but it can happen.
18	Not all cases where network market tips to
19	monopoly yield bad outcomes. First of all, those
20	monopolies can be unstable. There's a fair amount of
21	research that actually shows that network markets
22	flip-flop more frequently under some conditions than is
23	good for consumers. Because they're stuck with legacy
24	technologies that don't migrate forward to the product
25	of new innovator.

93

1	So, I think that just because something is a
2	network market doesn't mean that we need to worry about
3	some kind of tragedy of tipping. But it -- it can
4	happen. And then where it does happen, I think that the
5	remedial problem is really a challenging one. The
6	structural remedy can break up network effects,
7	interoperability remedies can lead to the need for
8	behavioral oversight, but also, we want to be careful, I
9	think one of the commentators, it might have been Bill,
10	pointed out, we don't necessarily want to mandate
11	interoperability, even when recommending a network
12	market, because new standards come into the market that
13	could improve things for people and you don't want to
14	eliminate the incentive to try to create the new network
15	standard.
16	So, I think network monopolies can arise, one
17	should not presume that they are too easily going to tip
18	to monopoly, even though their demand side of positive
19	externalities. We've seen cases where multiple systems
20	exist, and where they do exist, I think the remedy needs
21	to be thought about very carefully. Structural remedies
22	can be risky, interoperability is not always worth
23	mandating.
24	So, in those markets, it would seem the simplest
25	and baseline remedy would be if there is some kind of

94

1	conduct that is clearly putting impediments in the paths
2	of an innovator, enjoin that conduct, whether you go
3	farther and engage in structural relief or mandate to
4	interoperability should be undertaken with extreme
5	caution.
6	MR. HILLEBOE: Renata, did you want to comment?
7	MS. HESSE: I guess I think that the presence of
8	network effects in a market does at least open up the
9	door for the suggestion that the market may be more
10	susceptible to a monopoly -- to monopoly power being
11	exercised, or existing. I also think that network
12	effects can benefit consumers in many ways. So, there's
13	a hard balance there, because you don't -- you honestly
14	don't want to do something that will then take away the
15	benefit of the network effect that the consumer derives.
16	But I think they tend to raise barriers to entry,
17	whether or not those are long-standing and durable
18	barriers is I think the really big question, and if they
19	are, how you fix them.
20	MR. HILLEBOE: And Michael is somebody who is
21	out in those markets every day. What's your view?
22	MR. CUNNINGHAM: I'm not sure I can provide a
23	broad across the industry, certainly the network effects
24	in the markets we participate in is a very, very
25	profound -- has very profound effects on competition.

95

1	So, I also can recognize that there are consumer
2	benefits to it and I agree with Howard's comments that
3	it probably presents some special challenges in
4	structuring a remedy and that certainly structural
5	remedies could present some real issues.
6	MR. HILLEBOE: And Michael, precisely how do you
7	think they affect competition if they present a barrier
8	to entry? Is that essentially what you said?
9	MR. CUNNINGHAM: Yeah, they present a barrier to
10	entry. I think they also, because they present a
11	barrier to entry, they permit, you know, migration into
12	adjacent markets.
13	MR. HILLEBOE: And Bill?
14	MR. PAGE: One of the observations that was made
15	fairly early in the effort to integrate antitrust and
16	network effects, and I think it was Mark Rome who stated
17	it, one of the observations that had been made was that
18	when you're in this period of standards competition, in
19	between two incompatible standards and it's not entirely
20	clear which is going to become the dominant standard,
21	there's a huge incentive for firms to engage in
22	practices that don't look rationale. Penetration
23	pricing, giving stuff away for free, and so forth, and
24	part of the difficulty is that if you look down that
25	list of things that they have the incentive to do, a lot

96

1	of them look like antitrust violations. You know, it's
2	just rational to engage in practices that can look like
3	antitrust violations, and what they are is standards
4	competition, they're exactly what the literature would
5	predict as standards competition.
6	So, that is a serious dilemma for applying the
7	antitrust laws in these markets. On the other hand, you
8	know, one of the -- one of the supposed paradoxes in the
9	Microsoft case was, you know, who cares who the
10	Microsoft or Java, for example, wins, or Netscape/Java,
11	or Netscape alone, because all you'll have is just the
12	new monster. And who cares? You know, you'll just wind
13	up with one firm dominating the market and you'll have a
14	monopoly and so what.
15	And I think there's a very good answer to that,
16	that actually came up in the oral argument in the
17	Microsoft case, and that I take that the Court of
18	Appeals accepted, because they didn't even discuss it in
19	their opinion, and that is that you don't want a biased
20	choice. In other words, it does matter who wins.
21	You're going to have a monopolist, it does matter which
22	is the monopolist, and the network effects, the
23	literature would suggest, that in some circumstances,
24	network effects can exclude even a product that's better
25	setting aside the network advantage.

97

1	So, you know, I'm not sure exactly where to come
2	down on it. Mark had a few suggestions, in his article
3	that was in Connecticut, and I don't remember the name
4	of it, but he had a few suggestions on how to, for
5	example, distinguish conventional with the sort of the
6	predicted penetration pricing from genuine predatory
7	pricing and how that might be adapted to network
8	markets.
9	MR. HILLEBOE: Marina, do you have any thoughts
10	on that?
11	MS. LAO: I think it's true that network effects
12	can be very efficient, and the example that I'm thinking
13	of is not a high-tech one, but is real estate
14	multi-listing. No one would say that the network
15	effects there are not efficient, and agree that in
16	remedies where network effects are efficient, we have to
17	be very sure -- we have to be very careful not to take
18	away the efficiencies.
19	So, for instance, in the real estate
20	multi-listing situation, perhaps you could force the
21	network to open itself up to competitors, but not try to
22	introduce a competing network.
23	MR. HILLEBOE: And moving on to sort of --
24	MR. CUNNINGHAM: Just one final thought.
25	MR. HILLEBOE: Sure.

98

1	MR. CUNNINGHAM: Just on the idea of preserving
2	innovation through standards competition, perhaps
3	apropos my principal comments, innovation also occurs
4	through open collaboration about open standards and
5	there's ample evidence about that. So, I think it's a
6	factor, but I don't think it's the only factor that
7	needs to be considered in that circumstance.
8	MR. HILLEBOE: Moving on to kind of a nuts and
9	bolts issue, Renata suggested that given the speed of
10	change in these markets, that perhaps a shorter consent
11	decree might be appropriate. Is that something that as
12	an antitrust enforcement agency we should be thinking
13	about?
14	Howard?
15	MR. SHELANSKI: Maybe I'm too optimistic about
16	the ability to advise consent decrees, I should know
17	better, I think I litigated waiver number 917 on the NIT
18	decree, but I'm not sure that I would shorten the decree
19	for the following reason, and I mean, I defer to you who
20	implement these daily to know better, but it would seem
21	to me that if it was easier to repeal and modify a
22	decree than to re-authorize one or to negotiate a new
23	one, I might put one in place for a longer period of
24	time and back off if it becomes moot and then go in the
25	other direction. That's an enforcement question I'm not

99

1	qualified to answer.
2	MR. ELIASBERG: If I could follow up on that one
3	with Howard. Howard, there were allusions to some sort
4	of a review process, in which the court or somehow or
5	another would open up the decree, not to see to
6	necessarily compliance with the decree, but with the
7	effectiveness of the decree. How would you factor that
8	into this whole question of term of decree?
9	MR. SHELANSKI: Well, I think it's a great idea,
10	and I would favor a review provisions, or, you know,
11	eventual sunset provisions in the absence of review.
12	But review, you know, review is very difficult. You
13	know, I'm not sure the second and third triennial
14	reviews under the AT&T decree ever occurred, and so --
15	and then the question of, well, what gives cause, what
16	gives cause to open them up, but having them there in a
17	decree so that someone can go get a mandamus and seek
18	relief.
19	MR. HILLEBOE: Do any other speakers have any
20	thoughts about that?
21	Yes, Bill?
22	MR. PAGE: I think in principle, I like short
23	decrees. On the other hand, it's a bit of a catch-22
24	when you're talking about the compulsory licensing
25	provisions, because how do you market to firms the idea

100

1	of building on, say, Microsoft's proprietary base, if
2	the license is going to expire in a few years? I mean,
3	how -- that seems to be like a contradictory -- I mean,
4	not that firms would ever necessarily want to be
5	building on Microsoft's proprietary protocols, in many
6	instances, they might choose not to do that even if they
7	were thought to be perpetual licenses, but I would be
8	concerned that at some point, the government is going to
9	leave the picture and Microsoft is going to yank my
10	protocols under the basis of my whole business.
11	So, you know, I guess it depends -- to my way of
12	thinking, it would depend on the nature of the remedy.
13	If it's a prohibitory remedy to remove specific
14	impediments, that would make sense for that to just be a
15	short-term one. But if there is a legitimate need for a
16	forward-looking remedy, then I think, you know, five
17	years is probably not enough, and certainly it hasn't
18	been enough in the protocol licensing provision.
19	MR. SHELANSKI: Can I just follow up really
20	quickly on that?
21	MR. HILLEBOE: Of course.
22	MR. SHELANSKI: I think Bill makes a good point,
23	I think the nature of the conduct really in some sense
24	has to derive what the length of the decree is. For
25	example, suppose somebody gets a network monopoly by

101

1	penetration pricing, and now they get zero, and then
2	they undertake some type of conduct later once they have
3	their monopoly that prevents subsequent innovators by
4	doing the same thing, by exclusive dealing or something
5	else like that. I'm not sure that you want a short
6	decree there, because it's quite clear that the conduct
7	will always be harmful, and so I think tying it to the
8	conduct, there might not be a systematic answer.
9	MR. ELIASBERG: Actually, if I can follow up
10	with Renata, I think Renata you initially raised this
11	point. What are your thoughts on how to determine if a
12	shorter decree is appropriate, and also just how long
13	that shorter decree ought to be.
14	MS. HESSE: That's asking me impossible
15	questions. I actually agree with both Bill and Howard
16	that what kind of conduct it is that you're talking
17	about is going to be an important input into that
18	determination. It's clear that the five years was not
19	enough, for the section of the consent decree, or that
20	at least both Microsoft and all the plaintiffs came to
21	the conclusion that they needed more time.
22	So, and then there was a lot of work done, which
23	I think if you, you know, scour the status reports,
24	you'll see they're done to make sure that this problem
25	that Bill talked about, which was why would I invest in

102

1	this to begin with if it's going to get yanked out from
2	under me in the end, to see that the terms of the
3	licenses were flexible enough so that hopefully people
4	felt comfortable with that.
5	I think that the kinds of things to think about
6	when you're trying to decide whether or not a shorter or
7	longer decree makes sense have to do with both the way
8	in which the market changes, how quickly you think the
9	market is going to change, whether or not that matters
10	for the ultimate success of the remedy, whether or not
11	you think that there's a sort of simple one-shot
12	solution to the problem, and that if somebody can -- if
13	the particular conduct, if stopped for a period of time
14	will result in new entry, or in a lowering of a barrier
15	to entry that will be sufficient in a short period of
16	time to overcome the prospect of the network effect.
17	I think in most technology markets, despite the
18	fact that they move fast, this issue that Bill raised
19	about there being an underpinning in the monopolist's
20	technology that may be an important part of alleviating
21	the anticompetitive or the harm from the anticompetitive
22	conduct, would tend to suggest that shorter decrees
23	actually are not warranted in most cases.
24	On the other hand, you know, I think both of the
25	agencies have gone away from the idea of doing perpetual

103

1	decrees, ten years is generally the standard. So,
2	you're talking about the difference between five and ten
3	years, and it's hard to know precisely in what cases it
4	makes sense to do one or the other I guess.
5	MR. HILLEBOE: I thought Howard made an
6	interesting point, and it's something that we touched on
7	yesterday, but we kind of had a truncated discussion on
8	it, and that is I think there's a recognition frequently
9	in a case you see perceived liability, but you recognize
10	that it's going to be very difficult to come up with a
11	remedy. And the question what is the value of
12	proceeding and prosecuting that type of a case, and the
13	possible goals might be for deterrence, as Howard
14	suggested, or for establishing a precedent, or for
15	making it easier to bring a subsequent case.
16	I know Howard's view on that, but what do the
17	other speakers think about that? Renata, do you have
18	any thoughts about that? Or do you want to punt that
19	one?
20	MS. HESSE: How about this, why don't we start
21	down there, so Marina can go first.
22	MR. HILLEBOE: Marina?
23	MS. LAO: I believe that we should proceed if
24	the violation is egregious and if the consumer harm is
25	substantial, but where it is not substantial, and where

104

1	the act is borderline, then if we don't have a clear
2	remedy that is workable, then perhaps we should back
3	off.
4	MR. HILLEBOE: So, sort of a sliding scale in
5	your analysis?
6	MS. LAO: Sliding scale.
7	MR. HILLEBOE: Bill?
8	MR. PAGE: I would suggest that one remedy is
9	collateral estoppel, and that, you know, there are
10	plaintiffs who will not bring a case for the reasons
11	that we've just heard, that because it's simply
12	impossible to go up against the monopolist in
13	litigation, for practical terms. Just because an
14	injunctive remedy is not issued, does not necessarily
15	mean that there is not a remedial benefit, because there
16	can be follow-on litigation. I mean, the most recent
17	estimate I saw of the damages or the settlement amounts
18	in the Microsoft litigation was approaching nine billion
19	dollars. Even for Microsoft, nine billion, that will
20	get your attention.
21	So, I suspect that even establish -- and if the
22	case were brought with an eye for collateral estoppel, I
23	think there's every reason to bring a case.
24	MR. HILLEBOE: Michael?
25	MR. CUNNINGHAM: It's certainly consistent with

105

1	my visceral reaction and my advice to clients, to my
2	client, that it has a deterrent effect for typically
3	even more egregious behavior. I do think there are some
4	potential evidences that the deterrent effect is real.
5	I think in addition to the complaints that Howard laid
6	out when dealing with complicated problems the
7	experience of competition authorities in learning how to
8	deal with them and getting more sophisticated in dealing
9	with them is not a value that should be discarded value.
10	MS. HESSE: Actually, I think I said this
11	earlier, I actually agree with the notion of the
12	deterrent effect of taking action, even if you're not
13	100 percent sure that you can figure out a way to solve
14	the problem perfectly, or even reasonably well, and I
15	think there are a lot of people who would say, even
16	people who will say both, that the Microsoft decree has
17	been a failure, and has done nothing, and at the same
18	time say that it was a case that was worth bringing.
19	So, and I tend to -- I'm not taking a position
20	on whether it was a failure or not, but I agree that
21	even if you assume it was a failure, that the case
22	itself, both demonstrated that these were markets that
23	the government was capable of dealing with, that they
24	were capable of litigating against a huge company and
25	winning, and that, you know, nobody was, you know, above

106

1	the law. And that's an important point to make.
2	MR. HILLEBOE: Bill, I just have a question for
3	you. We talked yesterday about various goals in terms
4	of antitrust remedies, and you spent a great deal of
5	time talking about Microsoft. How would you
6	characterize, what's your opinion of what the goal was
7	for the government at the time they entered into that
8	remedy based upon reading from Charles James articles or
9	whatever, and do you think the goal was achieved?
10	MR. PAGE: You mean the consent decree?
11	MR. HILLEBOE: The 2002 consent.
12	MR. PAGE: Well, they're in a position where the
13	Court of Appeals had really given them not too much
14	choice. The thought of pursuing any type of structural
15	relief was impossible at that stage. So, at that point,
16	some sort of -- some sort of conduct was all that you
17	were going to get, and I suspect that -- well, perhaps
18	I'm not the best one to -- I'm certainly not going to
19	sort of assume what the goals were, but as I said
20	earlier, I think that by and large, the terms of the
21	consent decree and the parallel relief in the states'
22	remedy are closely tied to the theory of liability in
23	the government case.
24	Now, certainly the grandest standard by which we
25	would judge that would be does it restore the platform

107

1	threat? You know, does it create some sort of rival
2	platform that would threaten Microsoft, and by that
3	standard, you would have to say that it hasn't done
4	that. On the other hand, I think there are other ways
5	of evaluating the decree. I mean, one of the provisions
6	of the decree is to make sure -- there's an internal --
7	there are two, actual, internal Microsoft compliance
8	officers, and, you know, if you go back and listen to --
9	if you go back and read Judge Jackson's comments about
10	Microsoft, it's almost he said they were like, you know,
11	young punks or organized crime or, you know, defiant
12	organization, criminal enterprise, whatever, and I don't
13	think anyone -- well, I'm not sure that anyone would
14	necessarily say that that's the case now.
15	I think at least, you know, there is a huge --
16	in fact, there is one of the status reports describes
17	the Microsoft compliance program, I think they said
18	something like -- well, they've conducted these
19	antitrust compliance seminars worldwide, 15,000
20	employees have taken them, you know, all the executives
21	are schooled in the requirements of the consent decree
22	and the antitrust laws, it may all be window dressing,
23	but I suspect that there is a difference in attitude at
24	Microsoft because of this case.
25	MR. HILLEBOE: Any of the other speakers want to

108

1	comment on that?
2	MR. ELIASBERG: Yeah, a question I wanted to
3	touch base, actually, and start with you, Renata, you
4	indicated or suggested that there could be some
5	disruption to structural relief, indeed, sometimes it
6	can be cleaner and so forth. But we seem to have some
7	language from the Court of Appeals suggesting that we
8	should be extremely reluctant about thinking about
9	structural relief and indeed it should be the last
10	resort.
11	What thoughts do you have about just how
12	advisable is it for us to be thinking about structural
13	relief right out of the box with respect to such a
14	matter?
15	MS. HESSE: I think I read the Court of Appeals'
16	decision to be -- and this actually was something Bill
17	was talking about, also, to be focusing on the question
18	of causation and the importance of establishing
19	causation if you're then going to go and impose a
20	structural remedy. And that -- I think that is a very
21	important question.
22	I think the Court of Appeals' attitude toward
23	structural relief probably supports some of the things
24	that I said, which is that imposing it occasionally in a
25	Section 2 case or demonstrating that you're capable of

109

1	doing that may have a greater deterrent effect, and that
2	people perceive that remedy, rightly or wrongly, to be a
3	more Draconian one than a behavioral remedy.
4	But the question of causation, I think, is
5	really an interesting one, because it does get to this
6	question of how do you know what the competitive
7	conditions of the marketplace would look like without
8	the bad exclusionary conduct? And nobody knows, really.
9	Nobody knows whether another platform effect would have
10	emerged. And so I think it's hard to say looking at at
11	least in the Microsoft context, looking at the
12	marketplace today, whether or not the decree has been a
13	booming success or, you know, an abject failure, if --
14	because you really don't know what would have happened.
15	And I think the record was -- had some information about
16	it, but I don't think anybody really knew whether
17	Netscape, in fact, was really a viable platform threat.
18	We knew that Microsoft was worried about it and thought
19	that it was.
20	So, I think I certainly wouldn't out of the box
21	say, it's not worth even spending your time thinking
22	about, because I think these cases are -- they're not
23	only hard to put together and then try, but they're very
24	difficult, and you should leave open all of your options
25	in terms of thinking about how to resolve, how to remedy

110

1	a problem that you've seen and I think that, you know, a
2	structural remedy would certainly be appropriate in the
3	right cases.
4	MR. ELIASBERG: Howard, did you have something
5	you wanted to add?
6	MR. SHELANSKI: Well, my tongue-in-cheek remark
7	earlier about the cost to Microsoft aside, I don't
8	believe any of us believe that the government should be
9	in the business of just creating costs for firms. So,
10	we need to be darn sure of the curative potential for --
11	I think for any remedy, and I think with a structural
12	remedy, I read the Court of Appeals, too, of being as
13	insisting on a tight causal link, and I would rephrase
14	that slightly as a strong curative likelihood of success
15	for the competitive harms.
16	And I think you want to be darn sure of that in
17	a structural setting, because especially in a high-tech
18	industry, I think the unintended consequences of
19	structural relief could be many.
20	MR. ELIASBERG: Something I also wanted to just
21	cover with the panelists, just to be sure we canvassed
22	all the views, Marina floated the notion of I'll
23	describe it as lowering rivals' costs as a strategy with
24	respect to shaping -- creating -- formulating relief. I
25	was curious if any other panelists had a reaction one

111

1	way or the other about the advisability or not of such
2	imposition. You can either volunteer or I'll just go
3	ahead and call on you.
4	MR. PAGE: Well, I would say that it's
5	appropriate if it's in response to actions that
6	anticompetitively raised rivals' costs. I don't know
7	that because a violation has been found that all
8	methods, and I don't want to characterize you saying
9	this, but all methods of lowering rivals' costs have
10	been appropriate.
11	So, again, lowering rivals' costs is certainly a
12	legitimate goal, if the causal link to the
13	anticompetitive conduct is established.
14	MS. LAO: I really see that as a conduit to
15	promoting consumer welfare, and not to benefit
16	competitors for the sake of benefitting the competitors.
17	MR. SHELANSKI: As a veteran of the unbundling
18	wars in Telecom, I twitch a little bit when I hear
19	lowering rivals' costs, and I think the one thing that
20	would give me pause is I would say maybe, if the cost
21	you're lowering is one that the defendant is being asked
22	to lower through the remedy is a cost that the defendant
23	created, and I think that that would be a tie that even
24	before thinking about it I would want to see there,
25	because otherwise, I think there's really great danger

112

1	for the agency to become an ongoing regulatory authority
2	as opposed to someone recommending particular
3	anticompetitive conduct.
4	MR. ELIASBERG: One more question.
5	MR. HILLEBOE: Sure.
6	MR. ELIASBERG: Actually, this one, Michael, is
7	to you. In your presentation, you made a comment about
8	situations where steps may be taken by an incumbent to
9	change structure of its product so that it could not be
10	transferability or used by a subsequent -- front by a
11	rival or something of that nature. In a case like that,
12	assuming for the moment that there was liability found,
13	found for that alteration or change in the product
14	design, what would be the type of relief you would think
15	would be -- what would be the remedy that you would
16	think would be the appropriate remedy in a situation
17	like that?
18	MR. CUNNINGHAM: In our industry, I guess with a
19	strong network effects, some interoperability remedy
20	would seem to be the one that you would need. Yeah.
21	MR. ELIASBERG: Nothing else comes to mind?
22	MR. CUNNINGHAM: No.
23	MR. ELIASBERG: Anyone else have a rationale for
24	that?
25	(No response.)

113

1	MR. HILLEBOE: Well, I note that it's close to
2	12:30. So, I just want to say on behalf of the FTC, and
3	my colleagues at DOJ, I wanted to say thank you very
4	much to these speakers, an excellent presentation, and I
5	want to remind and thank everyone for coming and remind
6	everyone that we have a final wrap-up in the coming
7	weeks. Thank you.
8	(Applause.)
9	(Whereupon, at 12:28 p.m., the hearing was
10	adjourned.)
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25

114

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 HEARINGS
5	DATE: March 29, 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
13	DATED: 4/3/07
14
15	SALLY JO BOWLING
16
17	C E R T I F I C A T I O N O F P R O O F R E A D E R
18
19	I HEREBY CERTIFY that I proofread the transcript
20	for accuracy in spelling, hyphenation, punctuation and
21	format.
22
23
24	SARA J. VANCE
25

Updated January 2, 2024