Understanding Single-Firm Behavior: Remedies

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



2 and





10 THURSDAY, MARCH 29, 2007







1 A P P E A R A N C E S



9 Douglas Hilleboe, Federal Trade Commission

10 Ed Eliasberg, U.S. Department of Justice


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



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.


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


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


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


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


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


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


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,


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


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


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


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.


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


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


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.


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.


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


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


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


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


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?


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.


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


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


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.


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


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


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


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


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


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


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.


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


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


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.


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.


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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


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,


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.


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.


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


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


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


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


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


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.


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


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


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


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


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


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


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.


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


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


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


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


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


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


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


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


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


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


1 into some of these fast-moving markets.


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


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.


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


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.


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


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


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


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


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?


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.


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?


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.


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


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.


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


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.


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.


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


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


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


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


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


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


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.


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


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


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


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


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


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


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


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


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.


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?


23 MR. ELIASBERG: Anyone else have a rationale for

24 that?

25 (No response.)


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.)



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



5 DATE: March 29, 2007

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.

13                DATED: 4/3/07

15                SALLY JO BOWLING

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

19 I HEREBY CERTIFY that I proofread the transcript

20 for accuracy in spelling, hyphenation, punctuation and

21 format.

24                SARA J. VANCE


Updated June 25, 2015