Participant Presentations And Session Transcript For June 20


This document is available in two formats: this web page (for browsing content) and PDF (comparable to original document formatting). To view the PDF you will need Acrobat Reader, which may be downloaded from the Adobe site. For an official signed copy, please contact the Antitrust Documents Group.



1

1 UNITED STATES FEDERAL TRADE COMMISSION

2 and

3 UNITED STATES DEPARTMENT OF JUSTICE

4
5
6
7 SHERMAN ACT SECTION 2 JOINT HEARING

8 WELCOME AND OVERVIEW OF HEARINGS

9 TUESDAY, JUNE 20, 2006

10
11
12
13
14 HELD AT:

15 600 PENNSYLVANIA AVENUE, N.W.

16 WASHINGTON, D.C.

17 UNITED STATES FEDERAL TRADE COMMISSION

18 HEADQUARTERS BUILDING, ROOM 432

19 2:00 P.M. to 4:00 P.M.

20
21
22
23
24 Reported and transcribed by:

25 Susanne Bergling, RMR-CLR

2

1 MODERATOR:

2 WILLIAM BLUMENTHAL

3 Federal Trade Commission

4
5 PANELISTS:

6
7 Deborah Platt Majoras

8 Thomas O. Barnett

9 Dennis Carlton

10 Herbert Hovenkamp

11
12
13
14
15
16
17
18
19
20
21
22
23
24
25

3

1 C O N T E N T S

2
3 Introduction

4
5 Presentations:

6      Deborah Platt Majoras

7      Thomas O. Barnett

8      Herbert Hovenkamp

9      Dennis Carlton

10
11 Moderated Discussion

12
13 Adjournment

14
15
16
17
18
19
20
21
22
23
24
25

4

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

2 -  -  -  -  -

3 MR. BLUMENTHAL: Ladies and gentlemen, good

4 afternoon. I'm Bill Blumenthal from the FTC, and I'd

5 like to welcome you to the first of the joint Justice

6 Department Antitrust Division and Federal Trade

7 Commission hearings into Section 2 of the Sherman Act.

8 The purpose of these hearings is to explore how

9 best to identify anticompetitive exclusionary conduct

10 for purposes of antitrust enforcement. We are

11 envisioning a series of hearings that will kick off

12 today and will continue through December, probably two,

13 three, four hearings a month, with the exception of

14 August. After today's kick-off hearing, we are going to

15 have another hearing on Thursday of this week examining

16 predatory pricing, we will have a hearing in mid-July

17 examining refusals to deal, take a little bit of a

18 breather, and then resume in September with what would

19 then be a series of examinations.

20 The agencies are expecting to focus on legal

21 doctrine, on jurisprudence, economic research, and

22 business and consumer experience. We have a Federal

23 Register notice that is outstanding. We invite public

24 comment on a wide range of topics, and we hope that

25 those of you who are here, as well as many others, will

5

1 have an opportunity to submit comments on the topics

2 that we address. We are open to receiving those any

3 time through the final hearing, that is, through

4 December, although the earlier the better for our

5 purposes.

6 We are honored today to have a special panel to

7 kick off the hearings. They probably do not need much

8 introduction, so I am going to be very brief in offering

9 the introductions. In the order in which they will be

10 speaking, first we have Deborah Platt Majoras, Chairman

11 of the Federal Trade Commission. Thomas Barnett, the

12 Assistant Attorney General for Antitrust in the Justice

13 Department. Herb Hovenkamp, who is probably known to

14 most of you as -- as many things -- a professor of law

15 at the University of Iowa, but probably better known as

16 a co-author and a reviser of the leading treatise in the

17 antitrust field as well as a prolific author of many,

18 many other volumes, the most recent of which is recently

19 out, The Antitrust Enterprise: Principle and Execution,

20 available through Harvard University Press, with an

21 imprint of this year.

22 And finally Dennis Carlton, also known to many

23 of you in many capacities, most notably professor of

24 economics at the Graduate School of Business at the

25 University of Chicago, former president and still very

6

1 active in Lexicon, frequent expert witness, author of

2 many, many articles, author of I guess two of the

3 leading economics texts in the field. I'll leave it at

4 that. You all know Dennis Carlton.

5 I have several preliminary announcements that I

6 am going to make, one of which is legally mandated. The

7 first, which is not legally mandated but is a common

8 courtesy, if any of you have cell phones or

9 Blackberries, pagers, iPods, things of that nature,

10 please do as I'm doing right now and set it into silent

11 or manner mode, although if you do it like I just did,

12 you just dialed 7. Okay, there we go.

13 Second, I have been asked to let you know that

14 the men's room is immediately out these doors and to the

15 left. The ladies' room is out these doors, to the other

16 side of the elevator banks, and to the left.

17 And finally, the legally mandated announcement

18 is the one that says, as a safety tip for our visitors,

19 if the building alarms go off, please proceed calmly and

20 quickly as instructed. If we must leave the building,

21 take the stairway, which is to the right on the

22 Pennsylvania Avenue side, and after leaving the

23 building, please follow the stream of FTC people who are

24 practiced in this evolution. We will all go to the

25 Sculpture Garden, catty-cornered across the street at

7

1 7th and Constitution, and we will assemble there, where

2 noses will be counted.

3 With that, it gives me great pleasure to turn

4 the podium over to the Chairman of the Federal Trade

5 Commission, Debbie Majoras.

6 (Applause.)

7 CHAIRMAN MAJORAS: Good afternoon, everyone, and

8 thank you very much, Bill. Together with my good friend

9 and colleague, Assistant Attorney General Tom Barnett,

10 it is my great pleasure to welcome you to these hearings

11 in which we will be exploring conduct under Section 2,

12 and we are privileged to have two of our most

13 distinguished antitrust scholars here, Professors

14 Hovenkamp and Carlton.

15 Now, at the start of any new endeavor, it is

16 important to reflect on why we are undertaking it.

17 Beginning in 1990, the McKinsey Global Institute, led by

18 founding director William W. Louis, undertook a 12-year

19 study of the economic performance of 13 nations seeking

20 to understand globalization, and more fundamentally, the

21 disparities between rich and poor. The study showed

22 that levels of productivity made the difference between

23 rich and poor nations. What, though, made the

24 difference in the levels of productivity? The answer

25 they found was "undistorted competition in product

8

1 markets."

2 In his book in which he reports the results of

3 the study, Mr. Lewis says, "Most economic analysis ends

4 up attributing most of the differences in economic

5 performance to differences in labor and capital markets.

6 This conclusion is incorrect. Differences in

7 competition in product markets are much more important."

8 McKinsey also asked why the highly productive

9 United States has higher competitive intensity than

10 other nations. Mr. Lewis sums up the answer by saying

11 that, in the United States, "Consumer is king." More

12 specifically, he says, "[t]he United States adopted the

13 view that the purpose of an economy was to serve

14 consumers much earlier than any other society," and we

15 continue to "hold this view more strongly than almost

16 any other place." And he concludes that, in fact,

17 "Consumers are the only political force that can stand

18 up to producer interest, big government, and the

19 technocratic, political, business, and intellectual."

20 This is why we are here. The FTC and the

21 Antitrust Division have the responsibility to ensure

22 that competition in U.S. markets is free of distortion

23 and that consumers are protected not from markets but

24 through markets unburdened by anticompetitive conduct

25 and government-imposed restrictions. This work is

9

1 critical, indeed, to the well-being of the American

2 people. Over the past few decades, the United States

3 has substantially deregulated critical industries,

4 including transportation, telecommunication and energy,

5 to the substantial benefit of the economy and consumers.

6 As government regulators have given way to free markets,

7 much of the responsibility for protecting consumers

8 shifts to competition agencies and courts. While

9 competition is distorted when governments regulate or

10 intervene excessively, it also is true that private

11 actors can and do distort competition.

12 Breaking up cartels, preventing mergers that

13 will substantially reduce competition, and halting

14 conduct that goes beyond aggressive competition to

15 distorting it is vital to promoting vigorous competition

16 and maximizing consumer welfare, and we have developed a

17 great deal of consensus regarding appropriate antitrust

18 policy, I think, as it relates to cartels and to mergers

19 and other horizontal conduct, as a result of which our

20 enforcement has become more transparent and predictable,

21 which then, in turn, makes it easier for market

22 participants to make decisions about their own conduct.

23 Unilateral or "single-firm" conduct, however,

24 still vexes us. Even though we can find some

25 respectable measure of consensus around principles that

10

1 should apply, we find a range of opinions from

2 knowledgeable people about how to apply those principles

3 to enforcement in the market, and the question of the

4 proper test that our agencies should apply and that

5 courts should apply to conduct of the single firm with

6 market power now has dominated our antitrust debate for

7 several years.

8 We are not alone. Across the globe, over the

9 past quarter century, economic systems in which the

10 state owns the firms and central planners set out prices

11 and levels of output have given way to competition where

12 the forces of supply and demand determine prices and

13 allocate the resources, and we have worked hard to

14 promote the economic and political benefits of markets.

15 With attempts to introduce market economies have come

16 new competition authorities, today numbering around 100,

17 when only 15 years ago, we had just 20. And even

18 countries that for decades have had nearly total state

19 control over their economies, like China, are now

20 dedicating substantial resources to drafting competition

21 laws.

22 Currently, the issue of how to evaluate

23 unilateral conduct is the most heavily discussed and

24 debated area of competition policy in the international

25 arena. Just to give you a few examples, last week, FTC

11

1 and DOJ officials attended the EC's hearing to review

2 their policy under Article 82, which addresses conduct

3 by dominant firms. Officials from both agencies

4 recently held talks with our colleagues in Japan and

5 Mexico and Canada on the issue. We recently had panels

6 on it in the OECD. And since the International

7 Competition Network established a working group on

8 unilateral conduct in May, the FTC, which will co-chair

9 that group, has received expressions of interest from

10 more countries wanting to be involved than we have ever

11 had in any other working group in the ICN.

12 So, why the strong interest? Well, first, many

13 nations are facing the challenge of converting from

14 state-owned or supported monopolists to markets with

15 more than one participant, which is no small challenge,

16 as we ourselves have learned in trying to deregulate

17 certain markets like electricity. And, indeed, to

18 enforcers in those nations, it then becomes companies

19 with market power, not horizontal competitors, that are

20 the evil that must be attacked. Second, disagreement

21 among competition authorities about how to treat

22 unilateral conduct produces uncertainty in national and

23 international markets, which reduces the market

24 efficiency and imposes costs. And third, the analysis

25 of unilateral conduct in the identification of that

12

1 which is anticompetitive presents unique challenges that

2 are not present or at least are less present in the core

3 antitrust concern of conduct between competitors, and by

4 now, these unique challenges I think are familiar.

5 First and fundamentally -- and we discuss it all

6 the time, but that doesn't make it less difficult -- and

7 that is it is difficult to distinguish between

8 aggressive procompetitive unilateral conduct and

9 anticompetitive unilateral conduct. As the D.C. Circuit

10 said in the Microsoft case, "The challenge for an

11 antitrust court lies in stating a general rule for

12 distinguishing between exclusionary acts which reduce

13 social welfare and competitive acts which increase it,"

14 and this is tough, because as Judge Diane Wood wrote for

15 the Seventh Circuit, "distinguishing between legitimate

16 and unlawful unilateral conduct requires subtle economic

17 judgments about particular business practices." So,

18 while it's difficult, it must be done and it must be

19 done well.

20 Second, the process of distinguishing between

21 permissible and impermissible conduct must be relatively

22 consistent and transparent so that firms are able to

23 incorporate it into their decision-making. While there

24 are relatively few findings of Section 2 liability,

25 there nonetheless are a large number of different types

13

1 of conduct that may raise competition concerns and would

2 fall under Section 2.

3 And third, while antitrust practitioners have

4 had substantial success devising remedies for joint

5 conduct, devising remedies for single-firm behavior

6 presents significant difficulties. As Professors Areeda

7 and Hovenkamp put it, "By contrast with concerted

8 conduct, unilateral behavior is difficult to evaluate or

9 remedy by any means short of governmental management of

10 the enterprise."

11 We have much to work with as we move forward

12 with these hearings. Already a number of experienced

13 experts have proposed the adoption of a single test for

14 evaluating nearly all types of potentially exclusionary

15 conduct. Some have argued for a test that focuses on

16 the impact of the conduct on consumer welfare. Others

17 support analyzing whether the conduct involves the

18 short-term sacrifice of profits. Others support a

19 no-economic-sense test, which asks whether the cost of

20 engaging in the exclusionary conduct makes sense only

21 because it serves to eliminate competition.

22 Judge Posner has written that the inquiry should

23 focus on whether the conduct excludes other equally

24 efficient rivals, and still other practitioners and

25 scholars oppose the adoption of any single unilateral

14

1 conduct test and instead favor consideration of

2 different tests for particular types of exclusionary

3 conduct. And then, of course, when you go out into the

4 world, you see that there are many other opinions on the

5 type of test or framework that should be used.

6 So, proponents of the various tests and

7 approaches already have done a very good job of laying

8 out, I think, the relative merits, and virtually all

9 have acknowledged that their preferred approach is

10 probably not perfect. At these hearings, I hope we can

11 tackle this issue by starting with the conduct itself.

12 The hearings will have panels that will focus on

13 specific types of conduct that at least to date we know

14 can implicate liability. We want the panels to discuss

15 the conduct from the market perspective, from the ground

16 up; that is, to examine why and when firms engage in

17 certain practices, how they do it, what effects it

18 produces for the firm, for other firms, both customers

19 and the competition, and for consumers. And we should

20 look at whether firms in competitive markets also engage

21 in the same conduct, and if so, examine why that is. We

22 want these discussions, to the extent possible, to

23 include knowledgeable businesspeople or at least their

24 advisers, and from these discussions, we then should

25 endeavor to develop sign posts for when the conduct may

15

1 harm competition and when it typically does not. From

2 the sign posts, we hopefully can draw some guiding

3 principles, and only then should we turn to examining

4 the current state of the law as it has been applied to

5 such conduct and then determining what workable rules

6 can be applied to the specific conduct at issue. That

7 way, we can then see, can we pull these together into a

8 single test or a broader set of rules? And even if we

9 don't produce a consensus on the universal test or

10 tests, I'm optimistic that we can identify relative

11 consensus on a number of principles and then on how to

12 approach at least some fraction, hopefully a significant

13 fraction, of single-firm conduct we encounter.

14 In these discussions, we need to be careful not

15 to permit labels or semantic differences to get in the

16 way. In some discussions I've heard on these issues, I

17 have been worried that people are actually talking past

18 one another. In addition, this debate must not become

19 so academic that even if it could be resolved, it might

20 not have much practical application in the marketplace.

21 Indeed, last week I was speaking with a long-time

22 antitrust practitioner about these hearings and about

23 the debate over a proper test, and he said, that while

24 he thought the Section 2 issues were very important,

25 nonetheless, the search for the "holy grail" test might

16

1 just be something in which only about 27 people have an

2 interest. So, we really want to be careful about that.

3 I do think we start with some substantial

4 consensus about core underlying principles and factors

5 that should underlie any evaluation of unilateral

6 conduct.

7 First, the only type of unilateral conduct that

8 should implicate the antitrust laws is conduct that

9 produces durable harm to competition, leading to higher

10 prices, reduced output, lower quality or lower rates of

11 innovation. As much as we may value the success of

12 particular companies, the health of the companies

13 themselves is not the concern of antitrust law.

14 Second, there is consensus that antitrust

15 standards that govern unilateral conduct must not in

16 themselves deter competition, efficiency, or innovation,

17 and this is what we mean when we constantly say that we

18 worry about false positives. Obviously pervasive and

19 aggressive competition in which firms consistently try

20 to better each other by providing higher quality goods

21 and services at lower cost is crucial to maximizing

22 consumer welfare. So, the antitrust laws should then

23 never condemn market power that is obtained through the

24 development of superior products and services,

25 regardless of how many competitors are driven from the

17

1 marketplace in the process, and that, of course, has

2 been accepted by the courts.

3 Third, there is consensus that the standards for

4 evaluating unilateral conduct must be clear and

5 practical to administer or as practical as they can be

6 to administer. The most analytically sound principles

7 will provide little value to us if firms can't interpret

8 them when they are making their business decisions.

9 And, of course, courts have to be able to interpret and

10 apply rules as well.

11 And while I want to emphasize that I am going to

12 use the hearings to continue developing my own thinking

13 on these issues, I do approach them, in addition to the

14 broad principles, with a number of other hypotheses.

15 First, any legal framework needs to avoid

16 second-guessing business judgments that were objectively

17 reasonable at the time they were made. An ex post facto

18 examination of the hypothetical effects of alternative

19 courses of conduct is likely to chill legitimate

20 business behavior. Second, to be practical, any legal

21 framework must be able to evaluate conduct that both

22 generates efficiencies and produces anticompetitive

23 exclusion. If we only had to worry about conduct for

24 which the effects are obvious, we probably would not be

25 here today. And third, any test or tests must account

18

1 for the fact that certain types of unilateral conduct

2 are significantly more likely to cause competitive harm

3 than others. For example, most would agree that

4 unilateral above-cost pricing at monopoly levels should

5 not be condemned under the antitrust laws. Similarly,

6 behavior that some commentators have termed "cheap

7 exclusion," such as the use of government processes to

8 unlawfully extend the life of a patent, is generally

9 viewed as unlawful exclusionary conduct. And this may

10 mean that there is no unitary test or that we simply

11 need a broad framework that can accommodate a spectrum

12 or perhaps a sliding scale for the levels of harm, and

13 proposals have been made for how we might think about

14 the distinctions that could be made, including Deputy

15 Bureau of Competition Director Ken Glazer's proposal

16 that we analyze conduct by distinguishing between

17 conduct that's coercive versus incentivising.

18 Now, in the Microsoft case, the D.C. Circuit

19 applied what I view as a sensible weighted balance

20 approach to Microsoft's conduct that's largely

21 consistent with the three principles I just discussed.

22 Some have criticized the framework used in Microsoft as

23 insufficiently structured or unfocused, and I understand

24 where that comes from, but I think if we look at how it

25 was actually applied, it may be a workable framework

19

1 that incorporates the principles on which we have wide

2 consensus. I mean, perhaps the same criticism about

3 being unstructured could be applied to the Section 1

4 rule of reason and, in fact, probably is at times, but

5 as applied to, for example, joint ventures, the

6 balancing has been weighted I think in the right

7 direction.

8 First, the Microsoft court did not attempt to

9 substitute ex post facto its judgment for the business

10 judgments that were made ex ante, or to determine what

11 actions might have been better overall for consumers.

12 For example, the Court did not base its findings on an

13 ex post analysis of the impact of Microsoft's conduct on

14 the prices charged to consumers.

15 The Microsoft court also demonstrated that to

16 evaluate whether certain types of unilateral conduct

17 violate the antitrust laws does require an examination

18 of both likely anticompetitive and procompetitive

19 effects. For example, the Court analyzed the legality

20 of a Microsoft license provision that prohibited OEMs

21 from modifying the initial boot sequence. Microsoft did

22 not dispute that that restriction limited competition

23 against IE. The Court nonetheless held that the

24 restriction was not a violation because it concluded

25 that preventing the Windows desktop from ever being seen

20

1 at all in the boot sequence was a substantial alteration

2 of Microsoft's copyrighted work that could produce harm

3 that outweighs the marginal anticompetitive effect of

4 the prohibition. The Court performed this same analysis

5 across two dozen types of conduct, examining both the

6 anticompetitive effects and procompetitive

7 justifications, taking care, though, to ensure that it

8 not chill procompetitive behavior.

9 And finally, the D.C. Circuit made clear that it

10 did not consider all types of unilateral conduct to

11 raise equal concerns under the antitrust laws. For

12 example, the Court stated that courts need to be very

13 skeptical about claims that a dominant firm's design

14 changes harm competition and, by implication, violate

15 the antitrust laws.

16 One final note about the hearings. I hope that

17 our latest panels, which we will hold on remedies, will

18 produce a productive discussion. It simply is not

19 possible to implement sound competition policy for

20 single-firm conduct without giving careful thought to

21 remedies. Despite their importance, though, I think the

22 issues relating to remedies have not received extensive

23 attention. Take the Microsoft case, for example, which

24 although it received and still receives a bit of

25 notoriety, I have been stricken by how few productive

21

1 discussions of the remedy and the D.C. Circuit decision

2 that accepted the DOJ remedy while rejecting other

3 remedies have actually occurred, and while that might

4 have stemmed from some of the market dissatisfaction

5 over that remedy, I think these hearings should give the

6 Section 2 remedy issue the prominence that it deserves

7 in our analysis. After all, if you have done these

8 cases, you know that devising and drafting remedial

9 provisions in monopolization cases can be more difficult

10 than determining whether a violation has even occurred.

11 At bottom, through these hearings and through

12 our work, we need to remember that antitrust is the

13 means, not the end. Rather, the end is undistorted

14 competition driven by "king" -- and I would say "and

15 queen" -- consumer, and the challenge is to keep

16 competition undistorted while not distorting it

17 ourselves in the process.

18 So, I thank you again for attending the opening

19 of these hearings, and we look forward to all of your

20 contributions. Thank you very much.

21 (Applause.)

22 MR. BLUMENTHAL: Thank you, Chairman.

23 General Barnett?

24 MR. BARNETT: I am going to attempt to be

25 somewhat high-tech here. We will see if it works. Ah.

22

1 I want to thank, first off, the FTC for hosting

2 the first of these hearings on Section 2 and for their

3 help in organizing and planning and discussing the

4 issues that have brought us together. Both Chairman

5 Majoras, as well as the other FTC Commissioners and the

6 FTC staff, have all been extraordinarily helpful, and I

7 want to thank all of you as well as the Antitrust

8 Division staff who have worked very hard to prepare for

9 the session today and who will continue to work hard

10 over the months ahead.

11 I also want to thank Herb and Dennis for

12 agreeing to be with us. I feel very honored to have

13 such distinguished commentators, and feel that it is a

14 terrific start to these important hearings.

15 I am going to start with the same issue that

16 Debbie started with, which is why are we sponsoring

17 these hearings? And I can tell you up front, much of

18 what you are going to hear will echo some of what Debbie

19 said, but it was done unilaterally and without

20 consultation, and so I hope you will judge it at least

21 under the Section 2 standard, not under Section 1.

22 We do feel that it is important. As Debbie

23 said, unilateral conduct can harm consumer welfare. I

24 think there is a consensus that the focus of the

25 antitrust laws in the United States is to preserve and

23

1 promote or prevent harm to consumer welfare and that

2 unilateral conduct is an important element of that.

3 I also agree that this is the area of probably

4 the least consensus. I think there are large areas of

5 consensus within Section 2, but there are significant

6 areas where I think we have room for further

7 understanding.

8 These hearings, with the combination of legal,

9 economic, business and governmental/private

10 perspectives, provide us with a unique opportunity to

11 advance our understanding, and I believe that that will

12 help us to advance the development of the law. It can

13 provide helpful guidance to the courts, guidance to the

14 business community, and as Debbie quite eloquently put

15 it, to the international community that is now focused

16 on this issue.

17 There is a long tradition of the agencies

18 leading the development of competition law. I need only

19 point to Don Turner and the 1968 Merger Guidelines and

20 the formulation by Bill Baxter in 1982 to provide an

21 example of what has become the standard reference for

22 analyzing mergers, not only in U.S. courts, but really

23 around the world in many ways.

24 With respect to the international community,

25 again, I do want to both echo and underscore what Debbie

24

1 said. This is an issue that is at the forefront of

2 people's minds as we talk to officials on every

3 continent, and one example that sort of helped drive

4 this point home a bit, I was at a conference a while ago

5 in Budapest of Southeastern European former Soviet block

6 countries, and we were talking about a topic that the

7 Antitrust Division often talks about, which is the

8 importance of cartel enforcement, and one of the

9 officials approached me at a break and said, "I agree

10 with you, cartels are a terrible thing. I just wish

11 that our markets had enough participants so that they

12 could collude together. They don't have anyone to

13 collude with. So, we are focused on this dominant

14 former state-owned enterprise and how we can introduce

15 competition into this economy." It just drove home for

16 me, at least, the importance of this issue. It is

17 important here, but I think its importance abroad cannot

18 be over-emphasized.

19 The Supreme Court, to its credit, addressed the

20 issue of monopoly 96 years ago. That is when it decided

21 the Standard Oil case, and while we think of it as a

22 rule of reason case, it did talk about the three evils

23 of monopoly. It talked about first the power to fix

24 price and thereby injure the public; second, the power

25 of enabling a limitation on production; and third, the

25

1 danger of deterioration in quality of the monopolized

2 article, which it deemed was the inevitable result of

3 the monopolistic control of its production. Price

4 increases, output reductions, quality deterioration,

5 those are still the same three touchstones that we look

6 to that you heard Debbie talk about that go all the way

7 back to the Supreme Court's discussion of the issue in

8 1910.

9 As we have talked about it in the 96 years since

10 that decision, there has emerged I would say sort of a

11 dichotomy or two different views of monopoly. While we

12 would all agree that they can have their evils, and this

13 was articulated in part by John Hicks in 1935, who

14 talked about the evils of monopoly in the terms of a

15 quiet life. He talked about the fact that the

16 monopolist may not be out there trying to get the

17 highest price he absolutely can get, maximizing in the

18 short term the most profit that he or she can get, but

19 really, it is the lack of competitive zeal, the ability

20 to sit back and relax, to not have to research, develop,

21 to innovate at a frantic level. That is a major harm of

22 monopoly, and that is something on which we are very

23 focused in terms of preventing.

24 Now, at the same time, the Supreme Court just

25 last year articulated a different view of monopoly. In

26

1 the Trinko decision, the Court said, "The mere

2 possession of monopoly power and the concomitant

3 charging of monopoly prices is not only not unlawful, it

4 is an important element of the free market system. The

5 opportunity to charge monopoly prices, at least for a

6 short period, is what attracts business acumen in the

7 first place. It induces risk-taking that produces

8 innovation and economic growth."

9 All the way back in 1942, in Capitalism,

10 Socialism and Democracy, Joseph Schumpeter talked about

11 a similar process called creative destruction or the

12 gales of creative destruction, and I compliment my staff

13 who came up with the tornado there, but I have always,

14 since I read this in college, this -- be careful of the

15 gale behind you -- I have always liked this image,

16 because it talks about how the marketplace is a rough

17 place. It involves vigorous aggressive activity, people

18 fail, people are driven out of business, but it is

19 through that destructive process that you get creation.

20 Indeed, a similar image I was thinking about

21 recently, when somebody was talking to me about the

22 National Forest Service, I grew up watching the

23 commercials about Smokey the Bear and how forest fires

24 were such a terrible thing. How could we be against

25 forest fires? It turns out the National Park Service

27

1 realizes that preventing forest fires can be a bad

2 thing; that if you prevent them for too long, you create

3 much bigger, larger, hotter fires that cause more

4 permanent destruction to the ecosystem when they do

5 occur. Periodic smaller fires are actually a good and

6 healthy part of the process. That to me is another

7 illustration of this basic image. Competitive, creative

8 destruction in the marketplace is something that we want

9 to preserve and protect, not chill along the lines that

10 Debbie was talking about.

11 So, how do we reconcile these two views of a

12 monopoly, as a bad thing that causes sloth and

13 relaxation and a lack of competitive drive versus the

14 benefits of creative destruction, the opportunity to get

15 to a monopoly? Well, this somewhat conflicting view was

16 illustrated in a book written in 1964, and this was R.W.

17 Grant expressing some frustration about the treatment of

18 monopolies, and I will read this to you in a moment, but

19 the basic story here is of a man named Tom Smith who

20 invents a bread machine. It will produce terrific

21 bread, it will slice it, it will wrap it, all for less

22 than a penny a loaf, and as you can imagine, he very

23 shortly owns the market for bread in the United States

24 and is making large sums of money. He is ultimately,

25 however, brought low by the men of antitrust who bring

28

1 an antitrust case against him for making too much money

2 on the backs of consumers and driving everybody else out

3 of business, and he crafts a poem here to illustrate

4 this frustration.

5 "You're gouging on your prices

6 if you charge more than the rest

7 But it's unfair competition

8 if you think you can charge less!

9 A second point that we would make

10 to help avoid confusion:

11 Don't try to charge the same amount!

12 That would be collusion.

13 You must compete -- but not too much

14 for if you do, you see

15 then the market would be yours -ï½­

16 and that would be monopoly!

17 It's very similar in many ways to the admonition

18 of Learned Hand in the Alcoa case who said that the

19 successful competitor, having been urged to compete,

20 must not be turned upon when he or she succeeds.

21 So, having expressed that frustration back in

22 the 1940s and 1960s, where are we today? One of our

23 esteemed both I would say academics and judicial members

24 of the antitrust community, Richard Posner, Judge

25 Posner, remarked just last year, "Antitrust policy

29

1 toward 'unilateral abuses of market power' is 'the

2 biggest substantive issue facing antitrust today.'"

3 And if I can, if you will excuse me, preempt

4 Herb possibly, last year Herb is quoted or wrote,

5 "Notwithstanding a century of litigation," 96 years

6 since the Standard Oil decision, "the scope and the

7 meaning of exclusionary conduct under the Sherman Act

8 remain poorly defined."

9 Now, there are areas where I think there are

10 relatively easy answers. Doug Melamed has written about

11 the concept of naked exclusionary practices. I mean, if

12 you blow up your competitor's factory, few of us would

13 find that to be defensible conduct. That's a fairly

14 easy case for not finding liability. I also think there

15 are some fairly easy candidates for safe harbor

16 provisions. If you engage in conduct that merely

17 reduces your cost of production, that seems to me

18 beneficial to consumer welfare.

19 The difficulty lies in cases, as Debbie

20 referenced, that have the potential for both beneficial

21 cost reductions, innovation, development, integration,

22 and at the same time potentially anticompetitive

23 exclusion. How do we deal with those situations?

24 Well, some relatively recent Supreme Court

25 decisions have shown progress in this direction. In the

30

1 Brooke Group case, which is, of course, a predatory

2 pricing case, it dealt specifically with the issue of

3 recoupment and holding that Liggett in that case had not

4 shown the opportunity or the ability to recoup, but the

5 case in my view, at least, stands for more than that and

6 discusses, for example, specifically how harm to a

7 competitor does not demonstrate harm to competition.

8 There was little doubt in that case that there were

9 discount programs aimed at and/or that had a harmful

10 effect on Liggett, but the Court was quite clear that as

11 long as that does not harm competition, that is not an

12 antitrust problem.

13 Second, the Court also talked about the

14 practical ability of a judicial tribunal to regulate a

15 problem and avoid chilling legitimate price cutting.

16 It's recognizing the limitations of the body that is

17 administering the law. I would expand that to include

18 the limitations of agencies as well as courts, but it's

19 certainly a relevant consideration, and recognizing that

20 aggressive price cutting can be beneficial for consumers

21 and we do not want to chill it. Thus, it created

22 effectively a safe harbor against predatory pricing

23 claims where the prices were above some appropriate

24 measure of cost.

25 And the Court expressly acknowledged in creating

31

1 the safe harbor that there was at least the theoretical

2 possibility that there could be harm to consumers, harm

3 to consumer welfare, from some above-cost pricing, but

4 recognizing it was likely to do more harm than good to

5 try to ferret out those individual cases.

6 More recently, in the Trinko decision, the Court

7 obviously had a somewhat more limited holding but

8 discussed on a broader basis some of these same similar

9 Section 2 issues. It underscored the need for

10 administrable rules, clear objective standards. It

11 talked about the fact that being able to craft a remedy

12 that is both clear and administrable by the Court is

13 very important, endorsing Professor Areeda, in that no

14 court should impose a duty to deal that it cannot

15 explain or adequately and reasonably supervise, and

16 implicitly, at least, that not all problems may have

17 antitrust solutions.

18 While I think there are many areas of consensus,

19 there are many areas where we have a lot to learn. As

20 Debbie indicated, our panels are going to focus on

21 different aspects of conduct. We will start on Thursday

22 with a panel discussing predatory pricing and predatory

23 buying. Brooke Group answered a lot of questions. It

24 did not answer, among other things, what is the

25 appropriate measure of cost? Is it marginal cost? Is

32

1 it average cost? Is it average avoidable cost? Is it

2 average total cost? There has been a lot of discussion

3 about that, and we are looking forward to hearing

4 people's views on that.

5 The question on predatory pricing or remedy, are

6 you going to enjoin lower pricing? Weyerhaeuser is a

7 case with which you may be familiar. There's a cert

8 petition pending before the Supreme Court right now. It

9 involves a predatory bidding situation. The Solicitor

10 General's Office has filed an amicus brief on this front

11 encouraging the Court to take cert and to examine it. I

12 view it, at least, as an opportunity for the Court to

13 reaffirm in the Section 2 context that clear and

14 objective standards are extraordinarily important.

15 There's a jury instruction at issue in this case that

16 talks about prices that are unfair or unnecessarily

17 high. This is an opportunity for the Court to make

18 clear that a jury's post hoc determination of what it

19 believes was unfair is not the appropriate criteria for

20 determining whether or not there should be Section 2

21 liability.

22 Refusals to deal will follow. Again, this

23 raises very significant issues. When, if ever, should a

24 firm be compelled to deal with a competitor? If you are

25 compelled to deal, under what terms and conditions?

33

1 Will the Court be able to administer it? A range of

2 issues which we are, again, looking forward to hearing

3 the experts' views on it.

4 Loyalty discounts, another area that we will be

5 looking at. A couple of years ago, the United States

6 urged the Supreme Court not to take cert in the LePage's

7 case. That involved bundled discounts. That was not

8 because we necessarily agreed with the Third Circuit's

9 decision or analysis. Indeed, if you parse that

10 decision, I think it is very difficult to come up with a

11 clear standard of liability. There has been, in the

12 wake of LePage's, a flurry of attention by academics, by

13 legal scholars, on this issue of bundled discounts,

14 loyalty discounts, and we are looking and hoping to see

15 whether or not any consensus has developed on any of

16 these issues.

17 Should it be viewed as a predatory pricing

18 tactic, as exclusive dealing, as a tying tactic? Are

19 there safe harbors that can be developed even if we

20 cannot develop a single, clear answer for all cases?

21 Tying and exclusive dealing, Debbie mentioned that you

22 sometimes, when you see things in a competitive market,

23 that ought to make you question whether or not there are

24 benefits associated with it. Tying and exclusive

25 dealing can have anticompetitive effects. Look at our

34

1 Dentsply case as a recent example. By the same token,

2 we see these practices in competitive markets, and we

3 need to better understand what benefits there are and

4 when there are not.

5 Towards the end of the year, we expect to turn

6 toward some more general principles. Is there an

7 overarching standard for Section 2 cases and liability?

8 We all agree that consumer welfare is an appropriate

9 standard. Trying to operationalize that in a particular

10 case with particular conduct is more challenging, and

11 there is less agreement on that. Debbie outlined the

12 range of potential tests. The Antitrust Division in a

13 number of recent cases looked to the no-economic-sense

14 test. As I have talked with people about that, one

15 issue that I find is that people have different ideas of

16 what the test is. So, over and above discussing what

17 the appropriate test ought to be, there is some

18 confusion about what is meant in terms of what are you

19 going to look at and what the rules are. That may be

20 part of the semantic difference that Debbie was

21 referencing. Clarifying some of those things as well as

22 the underlying substantive issues I think can be

23 beneficial.

24 We may look at the issue of whether there are

25 different duties or different criteria for tying claims

35

1 under Section 3 of the Clayton Act versus Section 1 or

2 Section 2 of the Sherman Act.

3 Here, I have two reasons for putting this up.

4 As you can see, this associate is responding to a

5 request, "I'll be happy to give you innovative thinking.

6 What are your guidelines?" An example of having too

7 cabined an approach, too narrow a guidelines can be the

8 antithesis of innovative thinking, can restrain the

9 benefits that you may achieve through your innovation

10 and development. That is part of the creative

11 destruction that we want to encourage, not discourage,

12 as this cartoon suggests may be happening. So, I raise

13 that to say that while I am now going to talk about six

14 possible principles to inform our discussions, I do not

15 mean them to cabin or prevent a wide-ranging, open and

16 frank exchange of ideas.

17 So, first off, individual firms with market

18 monopoly power can act anticompetitively and harm

19 consumer welfare, and we should seek to identify and

20 prosecute such conduct. This is an important first

21 principle. If it were not true, we could just abolish

22 Section 2. That is not what we are here to do. We are

23 here to better focus and identify those instances where

24 there really is harm to consumer welfare.

25 Second, mere size, mere market share, does not

36

1 necessarily demonstrate competitive harm. It can

2 demonstrate superior acumen, effort, zeal, et cetera.

3 Third, injury to competitors does not

4 demonstrate competitive harm, a point that has been

5 talked about in a number of contexts.

6 Fourth, the need for clear, objective and

7 administrable rules, so that businesses, at the time

8 they are taking actions, can understand where the lines

9 are and can conform their behavior so they are not

10 deterred from engaging in procompetitive activity, so

11 that courts are not asked to do things that are beyond

12 their competence, and that agencies can do the same.

13 Fifth, avoid chilling procompetitive conduct,

14 and certainly an interrelated point, self-explanatory.

15 And finally, the remedy must promote

16 competition. A remedy that harms competition can be

17 worse than no remedy at all, an important point worthy

18 of bearing in mind.

19 Again, I want to thank the FTC, our panelists

20 for agreeing to kick off these hearings. We will

21 continue again on Thursday. We very much are interested

22 in a free, open and wide-ranging discussion of these

23 issues and are excited about the prospect.

24 With that, I will turn it over to Herb.

25 (Applause.)

37

1 DR. HOVENKAMP: Thank you. I am very grateful

2 and appreciative of being invited here, with particular

3 thanks to Chairperson Majoras and General Barnett for

4 extending this invitation.

5 In keeping with the thrust of this opening

6 meeting, which I believe is quite general, what I would

7 like to do is give kind of an overview of where I think

8 the fault lines and concerns in Section 2 lie. In the

9 future, future hearings, you are going to hear about

10 specific practices such as predatory pricing or refusals

11 to deal in considerable detail, and I am not going to do

12 that today. I am going to go through them rather

13 quickly and just point out where I think work needs to

14 be done and where the FTC and the Antitrust Division and

15 private litigants can use some clarification and

16 understanding.

17 I am going to divide my talk into three parts,

18 though the parts are not equal in size. First, a very

19 short one on market power or monopoly power, then a

20 rather long one on conduct issues, and then finally, a

21 much shorter one again on remedies.

22 With respect to power, the Merger Guidelines, in

23 particular the 1992 Merger Guidelines, the series of

24 guidelines that began with 1984, did a remarkable job of

25 rationalizing and simplifying the approach to market

38

1 delineation and assessment of the potential for

2 collusion or other types of anticompetitive behavior

3 that grow out of mergers. Some portions of the Merger

4 Guidelines market delineation sections are relevant to

5 Section 2 enforcement, but many are not, because the

6 question that one asks in a Section 2 case is

7 fundamentally different from the one that one asks in a

8 merger case.

9 In a merger case, we generally start out with

10 the presumption that a market is more or less

11 competitive, it may be oligopolistic or moderately

12 competitive prior to the merger, and what we really want

13 to know is whether the quality of competition is going

14 to deteriorate as a consequence of the merger. In

15 keeping with that, the SSNIP test, small but significant

16 nontransitory increase in price test, considers whether

17 a further increase in price would cause new entry or

18 other situations that would make this future price

19 increase unsustainable.

20 In a Section 2 case, by contrast, the opening

21 presumption is that the defendant or the firm under

22 examination is already a monopolist, is already charging

23 monopoly prices, and as a result, the SSNIP test is

24 really not the appropriate one in most circumstances,

25 although it certainly could be relevant in certain cases

39

1 like those involving an attempt to monopolize where the

2 defendant is not a monopolist at the time the conduct is

3 being assessed.

4 I do not have a solution to propose here. Those

5 of you who are familiar with this area know that this

6 involves something that in monopolization law we call

7 the Cellophane fallacy or the fallacy of inferring that

8 a firm lacks power because there is high

9 cross-elasticity of demand with the products of others

10 at current market prices, and, of course, if you

11 multiply that examination by asking what the response

12 would be to a yet further increase by a firm that is

13 already a monopolist, you might very well conclude that

14 the firm lacks this type of market power, because in

15 response to a yet further price increase, there would be

16 so much substitution away from the dominant firm's

17 product that the price increase would be unprofitable.

18 Well, if you took that approach, you would be

19 committing an error; namely, you would be ignoring the

20 fact that that firm is already a monopolist and

21 presumably already charging its profit-maximizing price.

22 So, I think one of the things that ought to be of

23 concern to the agencies as they go through these

24 hearings is to pay some special attention to the

25 formulation of usable presumptions that single firms can

40

1 use for assessing whether they have individual market

2 power and thus can be made liable to a Section 2

3 inquiry.

4 Let me just add to that, that that may involve

5 certain approaches that we have more or less given short

6 shrift to or rejected in the past. For example, it may

7 mean that we will not look at residual elasticity of

8 demand, which looks at the existing power that firms

9 have. We may have to look at things like price-cost

10 margins or rates of return. Some of these approaches

11 have been discredited in the past, but that does not

12 mean that they cannot be rehabilitated.

13 Okay, I want to spend a little more time on

14 monopolizing conduct. I am going to open by giving the

15 definition of monopolizing conduct from The Antitrust

16 Law Treatise that I am privileged to write, because it

17 is very general, has a number of flaws, but

18 nevertheless, I happen to like it for reasons I will

19 explain in a little while. The Antitrust Law Treatise

20 defines exclusionary conduct as conduct that is, number

21 one, reasonably capable of creating, enlarging or

22 prolonging monopoly power by impairing the opportunities

23 of rivals; and two, that either does not benefit

24 consumers at all or is unnecessary for the particular

25 consumer benefits that the acts produce; or three,

41

1 produces harms that are disproportionate to the

2 benefits; and finally, the assessment of the conduct

3 must be within the administrative capacity of the

4 antitrust tribunal.

5 Like I say, that test is very general. It is

6 not particularly helpful to assessing particular

7 instances of exclusionary conduct if it is the only

8 thing you have. You certainly would not want to give a

9 jury that test as an instruction and shut them up with

10 no further instruction and ask whether the defendant's

11 conduct was exclusionary, but the test was never

12 intended that way. It was, in fact, designed to be a

13 basic principle to be used in conjunction with specific

14 rules for specific types of antitrust cases, and it is

15 my view that that is fundamentally what Section 2

16 conduct jurisprudence needs to do.

17 I think there are very, very helpful general

18 tests. I like Greg Werden's no-economic-sense test. I

19 think there is much to be said for it. I think it

20 produces a few false negatives. Nevertheless, it's a

21 very, very good starting point. I like Judge Posner's

22 test that Chairperson Majoras mentioned in her talk,

23 which is conduct which under the circumstances is

24 capable of excluding an equally efficient rival. Once

25 again, I think it produces a few too many false

42

1 negatives, but they are good starting places.

2 However, none of them is a substitute for the

3 formulation of good technical rules covering individual

4 types of conduct; namely, pricing, abuses of the

5 intellectual property system, refusals to deal and so

6 on, okay?

7 In the few minutes I have, I cannot do any more

8 than scratch the surface, but I would like to give you

9 just a few observations about where we are in various

10 areas involving specific exclusionary practices and

11 where I think some of the problems lie.

12 With respect to predatory pricing, I believe

13 that both the Areeda-Turner test, as it was formulated

14 in 1975 and has later been incorporated into The

15 Antitrust Law Treatise, plus the elaboration of the

16 recoupment requirement in the Brooke Group case in 1993,

17 fundamentally set predatory pricing law on the right

18 track. I am a strong believer in the view that prices

19 must be below some measure of cost. Furthermore, they

20 must be below some measure of incremental cost; that is,

21 pricing is driven by concerns for variable costs, not

22 principally by fixed costs. That does not mean that

23 there are not a few problems.

24 One problem that I think needs to be assessed is

25 the problem of predatory pricing in oligopoly industries

43

1 by nondominant firms. That was, in fact, the facts of

2 Brooke Group. Strictly speaking, that may not be a

3 Section 2 issue. In fact, it may be an issue where the

4 Justice Department might reconsider its long-standing

5 opposition to bringing Robinson-Patman Act suits since

6 the late 1970s report on the Robinson-Patman Act and

7 create an exception for primary line enforcement given

8 the premise that with respect to primary line

9 enforcement, the principles that the Court follows are

10 basically the principles that are laid out in the

11 Sherman Act, and as a result, all of the overreaching

12 that applies to secondary line enforcement of the

13 Robinson-Patman Act need not apply here.

14 The problem with predatory pricing and oligopoly

15 is that victims have a different set of incentives than

16 they do in monopoly. Predatory pricing as a Section 2

17 problem involves predatory pricing designed to destroy a

18 rival. That is a very, very difficult thing to do. The

19 rival clearly has incentives to resist.

20 On the other hand, predatory pricing and

21 oligopoly frequently is used simply to enforce or bring

22 the oligopoly back into order so that the noncompliant

23 firm will once again raise its price to the oligopoly

24 levels; that is, the set of incentives that the target

25 of predatory pricing and oligopoly has are incentives to

44

1 rejoin, start making profits once again. As a result, I

2 believe predatory pricing in oligopoly industries is

3 fundamentally a more plausible strategy than strict

4 monopoly predatory pricing, and I think it needs to be

5 given somewhat closer scrutiny.

6 The other problem has to do with the measurement

7 of relevant costs. As I said before, I think the proper

8 measure of cost is incremental cost, which can mean

9 short-run marginal cost, short-run marginal cost with

10 some kind of additional factor for depreciable long-term

11 assets. It can mean average variable cost, as it was in

12 the Areeda-Turner formulation. The average variable

13 cost tests or the marginal cost tests simply don't work

14 very well in certain kinds of markets that have very

15 high fixed cost components and particularly in markets

16 that are characterized by a lot of intellectual property

17 or certain kinds of public utility or transportation

18 markets, such as the airline industry.

19 I think Ken Elzinga's analysis in the Spirit

20 Airlines case last year in the Sixth Circuit was a very

21 good first step, but the Government shouldn't be losing

22 predatory pricing cases in the airline industry. It is

23 the one industry where predatory pricing claims seem

24 plausible, and some attention needs to be paid to

25 modifying or, if necessary, rejecting and adopting a

45

1 different cost test for such industries.

2 On the Weyerhaeuser case and predatory buying, I

3 am one of the critics. I hope the Supreme Court sees

4 fit to follow the SG and grant cert. I think the

5 instruction that General Barnett described that

6 permitted a jury to find simply that predatory buying

7 occurs when the defendant pays too much or more than a

8 fair price is an atrocity. I think few people fully

9 appreciate how frequently such situations can come up;

10 that is, buying of inputs during times of scarcity.

11 This is not going to be an idiosyncratic situation.

12 This kind of case will come up a lot if the Ninth

13 Circuit's decision is permitted to stand.

14 Now, having said that, the question is what kind

15 of test to come up with. Well, in the Weyerhaeuser

16 case, where first of all the timber at issue accounted

17 for some 60 or 70 percent of the value of the finished

18 hardwood, and secondly, where at least according to the

19 jury, the hardwood was resold in a competitive market, I

20 think an average variable cost test might work quite

21 well; that is, buying is predatory if it forces the

22 defendant's resale prices to below its costs.

23 I am a little troubled by the use of an average

24 variable cost or marginal cost test, however, in a

25 situation where, number one, the defendant may sell in

46

1 an oligopoly, and number two, where the input on which

2 predatory buying is claimed is a relatively low

3 proportion of the value of the finished product, because

4 in that case, the variation in purchase might actually

5 fall within the margins that the firm charges, that it

6 is going to be too hard to detect predatory buying in

7 cases where the value of the purchased input is only a

8 tiny proportion of the value of the finished product.

9 In all cases, however, I believe that there

10 should be a recoupment requirement equivalent to that in

11 Brooke Group and that the Ninth Circuit erred not only

12 in its failure to require a showing of prices below

13 cost, but also in its failure to require a fairly strict

14 showing of recoupment.

15 With respect to patents, there is too much to

16 say and too little time. I just want to make one fairly

17 general observation. Mr. Barnett mentioned Joseph

18 Schumpeter's Capitalism, Socialism and Democracy, this

19 very, very important book in 1942 which opined in the

20 chapter on creative destruction that the amount of

21 welfare contributed to the economy through innovation is

22 far, far greater than the amount contributed by moderate

23 movements from oligopoly to competitive industries. I

24 mean, Schumpeter basically looked at the prior half

25 century or so of development as a result of the second

47

1 industrial revolution, of the theorizing of economists

2 like Edward Chamberlin and Joan Robinson, who were very

3 upset about oligopoly and imperfections in the economy,

4 and said, "You'd think to listen to these people that

5 American consumers were much, much impoverished compared

6 to their position in the 1870s, and, in fact, nothing

7 could be further from the truth."

8 Well, where do all those gains come from if we

9 are now in this oligopolistic era? And one of the

10 things Schumpeter concluded is that they came from

11 innovation. Schumpeter's premises were formalized and

12 given empirical support in Robert Solo's work in the

13 1950s in which Solo himself concluded that as much as 80

14 percent of economic gain comes from innovation rather

15 than simple improvements in price-cost relationships.

16 Now, neither Schumpeter nor Solo was talking

17 about IP law. They were talking about innovation, and,

18 of course, there is this enormous lingering question out

19 there of whether the IP laws we have are sufficient to

20 facilitate the optimal amount of innovation or whether

21 they, in fact, may hinder innovation. Fundamentally,

22 that is not antitrust's problem. The antitrust laws

23 need to accept the existing IP laws, warts and all, and

24 I personally believe there are a fair number of warts.

25 One thing, however, that that work suggests is

48

1 that antitrust needs to be much more concerned with

2 restraints on innovation. We have generally measured

3 harm in the antitrust laws by looking at price-cost

4 relationships, deviations from marginal cost pricing.

5 Harm to innovation is always included kind of as an

6 afterthought, but it has never been very well formalized

7 into our models of harm, and actually, there are pretty

8 good reasons for that. We have very good rules for

9 determining when prices deviate from marginal costs and

10 what the price elasticities facing firms are.

11 Predicting the consequences of restraints on innovation

12 is far more difficult, because innovation always takes

13 us by surprise.

14 We will never know, for example, what the

15 consequences were of Microsoft's successful attempts to

16 get Intel to stop developing a Java-enabled chip. How

17 good would it have been? Would it have done all the

18 things that Bill Gates feared in commoditizing the

19 platform market and so on? Those are very hard things

20 to predict, and for that reason, I believe courts are

21 rightfully skeptical when they turn away private

22 plaintiffs who claim that the injury that they suffer is

23 an injury caused by a lack of innovation.

24 So, I believe this is one area where the

25 Government should move into the fore, because they do

49

1 not need to prove damages, they do not need to prove

2 causation in the strict private plaintiff sense. I

3 think restraints on innovation are something that need

4 far more development in Section 2 law than they have

5 received in the past.

6 With respect to vertical exclusion, I just have

7 a couple of comments. First of all, there has been a

8 not so subtle move over the last four or five years in

9 government enforcement to move away from Section 1 of

10 the Sherman Act and Section 3 of the Clayton Act and

11 towards Section 2 of the Sherman Act as a device for

12 enforcing laws against tying or tying-like practices and

13 exclusive dealing, and I believe that is the correct

14 movement. Fundamentally, tying and exclusive dealing

15 ought to be regarded as dominant firm exclusionary

16 practices. They are rarely anticompetitive at

17 nondominant levels, and fundamentally, they do not

18 depend on agreement in any meaningful sense of the word.

19 Unlike resale price maintenance or Sylvania-style

20 restraints, they are typically not the product of

21 bargaining and traditional agreement between dealers and

22 manufacturers.

23 No, most tying and most exclusive dealing is

24 imposed by manufacturers unilaterally on dealers. The

25 dealers generally do not like it, but they accept it as

50

1 the price of a dealership. It ought to be treated as an

2 exclusionary practice, number one. The agreement

3 requirements really get in the way of appropriate

4 analysis of tying and exclusive dealing in most

5 situations. And finally, the market power requirement

6 should be equivalent to those that we assess in

7 monopolization cases.

8 So, I laud the increased scrutiny of tying and

9 exclusive dealing under Section 2 of the Sherman Act.

10 Microsoft included both, but the Government won on its

11 Section 2 tying claims. Dentsply, of course, the

12 exclusive dealing case that the Government won a year or

13 two ago, was a Section 2 case.

14 On bundled discount -- you are going to have a

15 big hearing on these, right? You are going to talk

16 about bundled discounts a lot? Are they predatory

17 pricing or are they tying? I think they are a little

18 bit of both, and I think the way to analyze them is by

19 asking two questions in two different stages.

20 The first question you ask is, are two goods

21 subject to a bundled discount bundled together? Well,

22 what does that mean? Well, it means that an equally

23 efficient firm that offered only one of them could not

24 match the bundled offer. How do you get there? Well,

25 as several papers have shown, you basically attribute

51

1 the entire discount to the product upon which exclusion

2 is claimed, and then you ask whether the price of that

3 product, subject to the full discount, has fallen below

4 a relevant measure of cost, whatever cost measure you

5 would use in a predatory pricing case, okay?

6 That gets you to bundling; that is, that

7 predatory pricing test gets you an answer to the

8 question, are the two firms -- are the two products

9 bundled together? And if the answer is that no equally

10 efficient firm that offered only one of the products can

11 match the price, then they are bundled together, but

12 that is only the beginning rather than the end of the

13 inquiry. Tying is explicit bundling of products

14 together, and yet most tying is perfectly legal. So,

15 once we have decided that two products are bundled

16 together, we have yet a further set of questions to ask

17 about whether there is foreclosure, whether the

18 foreclosure is justified under the circumstances by cost

19 reductions, improvements in consumer satisfaction,

20 quality control, in many instances price discrimination,

21 and so on.

22 Finally, on conduct, on refusals to deal, my

23 suggestion is that the Government simply get out of the

24 business of enforcing the law against simple refusals to

25 deal. Now, conditional refusals are something else.

52

1 Conditions usually mean exclusive dealing or tying.

2 Lots of things, including price fixing, can amount to

3 conditional refusals to deal, but if we are talking

4 about simple refusals to deal in the Trinko or Aspen

5 sense, I think the administrative problems are so

6 horrific, the disincentives created to competitive

7 behavior are so substantial, that the best thing that

8 the Government can do is stay away, and, in fact, that

9 is pretty much what they have been doing, even going so

10 far as to support the defendants in the Trinko case.

11 Okay, then let me turn finally and very briefly

12 to the subject of remedies. Both General Barnett and

13 Debbie Majoras spoke at some length about the importance

14 of remedy. I simply want to underscore what they said.

15 In fact, I would go a little bit further and say that

16 every Section 2 action that the Government brings ought

17 to begin with an exit strategy, right? We have talked

18 about Iraq, we have talked about exit strategies, and

19 now we have discovered that whatever exit strategy we

20 have, we probably could have had a better one, and the

21 same thing applies to Section 2.

22 Section 2 has no moral content. The only

23 purpose in bringing these cases is to make the economy

24 work better, and if you do not have a clear picture of

25 the kind of remedy you want when you go in, then you

53

1 really have to wonder whether it is worth bringing the

2 action to begin with.

3 For a long, long period of our history,

4 beginning with Standard Oil and through the 1960s, the

5 preferred remedies were structural or mandatory breakup

6 of firms. For relatively good reasons, those kinds of

7 remedies have fallen into some disrepute. Many of them

8 were very, very poorly designed. For example, the

9 remedy in the United Machinery case, which may have

10 ruined the firm, although there was some good evidence

11 that USM's technology in its Beverly, Massachusetts

12 plant was pretty obsolete already to begin with, or

13 remedies like the one in Grinnell, which didn't really

14 break up the monopoly at all, but just divided up the

15 market into a whole bunch of little monopolies.

16 Today, we operate in a regime in which

17 structural remedies in Section 2 cases appear to be

18 disfavored; conduct remedies are preferred.

19 Unfortunately, I think the record that we are developing

20 with respect to conduct remedies is not much better than

21 the record we developed with respect to structural

22 remedies in the 1960s and earlier. I think the verdict

23 is still out on the Microsoft remedy, largely because of

24 the two-year extension, but once that time period has

25 run, we have to look back and say, "Well, exactly what

54

1 did we accomplish here in terms of making this market

2 more competitive? And to the extent the market is more

3 competitive, to what extent was it owed to the remedy?"

4 I would like to see some more serious attention

5 be paid once again to structural remedies or at least

6 modified structural remedies, things like compulsory

7 licensing, as mechanisms for restoring competition.

8 Compulsory licensing is a dirty word in the United

9 States when we are talking about general forcing of

10 firms to share their patents, but that is not what we

11 are talking about here. We are talking about proven

12 antitrust violators who are frequently forced to give up

13 their plants and other kinds of hardware. Compulsory

14 licensing under that set of circumstances is a perfectly

15 viable remedy, and I would like to see us use it much

16 more seriously than we have in the past.

17 I am afraid I have gone over my time, and so I

18 will turn the floor over to Professor Carlton. Thank

19 you.

20 (Applause.)

21 DR. CARLTON: Okay, thank you. It is a pleasure

22 to be here, and I express my appreciation to the

23 organizers for inviting me, and I am honored to sit with

24 such distinguished panelists.

25 Exclusionary conduct is an important policy

55

1 topic. There is a great amount of debate as to what is

2 an exclusionary act and how to deal with it. There is

3 much less consensus on what bad acts are and how to

4 adjudicate them in the context of Section 2 claims,

5 than, for example, what cartel behavior is or how to

6 handle cartl claims. The real problem is that

7 competition harms rivals just like exclusionary

8 behavior, and it is sometimes easy to confuse the two.

9 As a general matter, it is very hard to study

10 what should be the optimal policy for exclusionary

11 conduct. The reason, one reason, is that the biggest

12 effect of any antitrust policy is likely to be, not on

13 litigants in litigated cases, but rather, on firms that

14 are not involved in litigation at all but are forced to

15 change their business behavior in contemplation of legal

16 rules. That means that although it is definitely

17 informative for economists and lawyers to study the

18 outcome of individual cases and you can learn a lot -ï½­

19 did the court get it right, did they get it wrong -ï½­

20 that is not really a study of antitrust policy.

21 To appropriately do a study of antitrust policy,

22 you have to look at either times when the antitrust laws

23 were adjudicated differently, that is, there was a

24 different policy, or perhaps you have to look at

25 different countries, and that is hard to do. Looking at

56

1 litigated cases will give you a biased view. They are

2 very self-selected cases and really will not allow you

3 to focus on what may well be the biggest costs of an

4 antitrust policy, and that is, the chilling effect it

5 has on the behavior of nonlitigants.

6 Since this is an introductory panel, one of the

7 advantages you get is you can yell out these hard

8 questions, and then when they are not answered by the

9 end, you can say, "See, you didn't answer my question,"

10 so I am not suggesting I have a simple answer. I am

11 just saying here that this is actually a quite important

12 problem and definitely deserves some research time.

13 What I will talk about today and on which I

14 think I do have some answers are two topics. One has to

15 do with the profit sacrifice test, or its variant, the

16 no-economic-sense test, and the other has to do with

17 market definition. Each of those is used or proposed to

18 be used as a tool to uncover Section 2 violations.

19 Now, it is important to distinguish a tool from

20 an objective. The objective is to maximize consumer

21 welfare, and whether when I say consumer welfare, I mean

22 consumer plus producer surplus or just consumer surplus,

23 that is a debate we can hold for another time. It does

24 not matter to the remarks I am going to make.

25 The tools you choose are used to identify acts

57

1 that harm welfare. Now, the use of tools necessarily

2 entails errors. You are going to have false negatives

3 and false positives. It is going to happen because,

4 one, you do not always have perfect information. Even

5 if you are the smartest economist, you make errors, and

6 if you are a juror who knows no economics, you might

7 make errors, and second, the tools are not quite the

8 same as the objective. So, the tools will differ from

9 the objectives sometimes, so there definitely will be

10 errors, and the real question is whether and when tools

11 should be used.

12 From an economic point of view, the question of

13 whether a particular act harmed consumers is a very

14 well-posed question that I could, for example, assign to

15 a Ph.D. student writing his thesis, and that person

16 could go about trying to answer it with economic tools

17 and econometric tools. Now, it is true it sometimes may

18 be hard for that person to reach an answer, especially

19 if there is both an efficiency effect and an

20 exclusionary effect on rivals, but sometimes it will not

21 be so hard, okay, but it is a well-posed question.

22 Now, from a policy perspective, the DOJ and the

23 FTC, with their staffs of knowledgeable economists and

24 attorneys, should be focusing on answering the direct

25 question, is there harm to welfare, and not on the use

58

1 of possible tools, though obviously they need,

2 especially if they are going to think about litigation,

3 they need to worry how judges and juries will use

4 whatever tools they are told to use when a case is

5 litigated.

6 Now, matters change as the inquiry shifts away

7 from the government agencies to courts with judges and

8 juries who may have less economic sophistication, and

9 their simple but imperfect rules may be better than an

10 unstructured inquiry, but simple rules should not be

11 viewed as anything but simple rules, crude guides that

12 sometimes work for some acts, but not all. It can be

13 dangerous to use a simple rule that could ultimately

14 subvert the goal of maximizing welfare.

15 Now, let me turn to first the profit sacrifice

16 test and then to market definition. The profit

17 sacrifice test, or a close relative, the

18 no-economic-sense test, asks would the act make sense

19 but for its exclusionary effect? And that test may work

20 fine in the hands of some of its accomplished

21 proponents, and in particular, I have in mind Greg

22 Werden, who is sitting here in the audience. My hunch

23 is if I sit down with Greg and we are talking about an

24 exclusionary act, we are going to reach agreement nine

25 times out of ten, and he is aware of the limitations of

59

1 the test and would carve out exceptions so it would not

2 be misused.

3 What I worry about is when there is someone

4 implementing the test who is not as smart as Greg.

5 Suppose they have a judge or a jury who is not an

6 economist? What worries me about the test is that it

7 raises all the danger signs associated with possibly

8 confusing competition with exclusionary conduct. Let me

9 try and explain why, and I will give you two or three

10 reasons.

11 First, all strategic behavior -- and every

12 business school teaches this -- all strategic behavior

13 is designed to improve one firm's position relative to

14 the other one. It is relative position that matters,

15 and that is what is going to often determine the outcome

16 of a competitive battle. Investments in advertising,

17 investments in R&D, price discounts, all of these could

18 mistakenly and easily be condemned by jurors convinced

19 that the firm engaged in the act could have been more

20 accommodating, less exclusionary. That "but for"

21 standard -- that is, would the act make sense but for

22 the exclusionary effect -- that hypothetical thought

23 experiment is actually quite a difficult one to

24 implement. That does not mean if you are measuring harm

25 to consumer welfare that you can necessarily get around

60

1 it. Maybe it is unavoidable sometimes, but I am worried

2 when you are postulating the "but for" the exclusionary

3 conduct, what hypothetical world will you be proposing,

4 and what worries me is there could be a very high error

5 rate associated with application of that rule.

6 The second reason I am worried about it is

7 because of the way it is formulated: Does it make

8 economic sense to engage in this behavior? Now, Ronald

9 Coase, one of my colleagues at the University of

10 Chicago, always fond of explaining how little economists

11 know about business and that businessmen really know a

12 lot about business, and just because an economist cannot

13 understand an action or a jury cannot understand an

14 action, he said why should that create antitrust

15 liability for the poor firm? And that I think is a

16 serious concern.

17 The profit sacrifice test may sometimes work and

18 may be appropriate for some actions, but I think it is

19 dangerous to enshrine it as a general proposition. It

20 strikes me as much more appropriate to devise tests and

21 screens that fit particular exclusionary acts,

22 especially because under Section 2 we have a range of

23 such widely different behavior that people have attacked

24 as exclusionary, so many different acts fall under

25 Section 2, and with each act, I would be concerned as to

61

1 how the legal treatment of the act is creating special

2 risks for chilling competitive behavior, and that is

3 going to differ from act to act to act.

4 So, for example, to pick up on what Herb was

5 saying, let us suppose you are looking at an industry

6 that is undergoing rapid technological change, and there

7 is an exclusionary claim that the way the product was

8 designed is a problem. Well, you have to worry in that

9 instance whether you are depriving consumers of a new

10 product if you attack the firm for its product design,

11 and that can lead to large losses. So, in that

12 situation, I might want to give more weight to the

13 firm's efficiency claim for fear of causing a large dead

14 weight loss than in other situations.

15 Let me give you a second example. Let's talk

16 about the Areeda-Turner test for predatory pricing.

17 Basically that's a test that says if your price is below

18 some measure of cost, unless it is, I am not going to

19 worry. So, the implicit idea is that if you see price

20 below some measure of cost, that is a big enough

21 deviation from what we usually think of as profit

22 maximization that there is something fishy, and the

23 reason they chose price below a cost as the predation

24 standard rather than above-cost pricing is because they

25 were very worried about chilling competition that drives

62

1 price down. So, they were specifically worried in the

2 context of predatory pricing of erroneous behavior, of

3 erroneous condemnation, which then would chill

4 competition, very specialized to the predatory act,

5 okay? Not a general principle.

6 If you misread it, what the Areeda-Turner test

7 said, you could misread it as follows: You could say,

8 ah, price below cost, the firm is not profit-maximizing.

9 If the firm is not profit-maximizing, there is something

10 fishy going on, that's a violation. Now, if you read it

11 that way, which would be an incorrect way to read it,

12 that is saying that any deviation from profit-maximizing

13 behavior would be an antitrust violation, and that would

14 make me nervous. That is the concern I have about a

15 profit sacrifice test, because if you are not maximizing

16 profit, your failure to maximize profit is a sacrifice

17 of profit, and I do not want get into the situation in

18 which I claim that firms must be maximizing profit as I

19 see it or it is an antitrust violation.

20 Now, you could remedy that. You could say, all

21 right, I will not say you have to be exactly, I will

22 give you a margin of error, but that is really my

23 concern with the test, not that it is not in some

24 situations useful, but that I would not want to enshrine

25 it as a general principle.

63

1 Let me now turn to a market definition, which I

2 was going to say I am sure you have heard a lot about

3 and probably do not want to hear more about, and I was

4 glad, though, that Herb did talk about it, because at

5 least it confirmed in my view that someone else thinks

6 it deserves still more thought.

7 But for the antitrust laws, industrial

8 organization economists would not pay all that much

9 attention as to how you define a market. I think it is

10 fair to say that the reason it receives so much

11 attention is precisely because it has been used as a

12 screen or a requirement under our antitrust laws, and

13 that is what gives it prominence.

14 Now, my own view is it is a very good but crude

15 tool, and it has this intuition behind it that if there

16 are lots of rivals, do not worry, and I think that is a

17 very good common sense rule, but it is hard to apply in

18 the non-merger context, and there have been attempts to

19 apply it in the nonmerger context that at least

20 sometimes strike me as odd. So, what I have seen in a

21 number of cases is the application of the Merger

22 Guidelines, and it goes something like this.

23 Define a market so that a hypothetical

24 monopolist of those products can raise price 5 percent

25 above the competitive level, the competitive price. In

64

1 a merger case, it would be the current price, but since

2 this is not a merger case, we say above the competitive

3 price. Well, that is a very well-posed question, and in

4 order to answer it, I have to say, "Okay, well, what is

5 that competitive price?" Well, I do not know what the

6 competitive price is. If I knew what the competitive

7 price is, I could look and see, is the current price

8 above the competitive price? And if it was, I would

9 say, yes, it is above. I would not then have to define

10 a market, take market shares and say, "Ah, you know,

11 based on all this analysis, you know, 10 is above 5 when

12 I started, and still, the market shares are so high now,

13 I conclude 10 is above 5." So, there is a circularity

14 to it that I find a little troubling. So, you cannot

15 really directly answer the question, because the

16 competitive price is not available. So, then, what do

17 you do?

18 I think there are several alternatives. Herb

19 mentioned some of them, and none of them I would say are

20 completely satisfying. One alternative is to ignore the

21 problem and simply say, "I am not going to use the

22 competitive price, I am going to use the existing

23 price," and Herb sort of indicated that this leads to

24 the well-known what is called Cellophane fallacy in

25 which it is possible that you will not find any market

65

1 power at the current price, but you do have market power

2 because you have already raised the price above the

3 competitive level. That is one thing you can do, with

4 its problems.

5 The second thing you can do is you could say,

6 "Okay, let me ask the following: Is price above

7 marginal cost where marginal cost I will use as my proxy

8 for the competitive price?" Of course, that kind of

9 replaces one question with the other, what is marginal

10 cost? Well, maybe you can go out and try and measure

11 marginal cost. If you had access to firm information,

12 you could try and sift through accounting information,

13 you could econometrically try and estimate a cost curve.

14 It's difficult, okay?

15 Moreover, suppose you do find prices above

16 marginal cost. You have to face the realistic

17 possibility that most markets are not perfectly

18 competitive. In most markets, price will not equal

19 marginal cost; therefore, and what you presumably must

20 mean, is that price deviates a lot from marginal cost if

21 you are worried about such a deviation. The amount of

22 deviation, the deviating a lot from marginal cost, has

23 actually never been articulated that I have seen in a

24 quantitative way.

25 Okay, suppose you do not like marginal cost. Is

66

1 there something else you can do? Well, I think there

2 is. Another thing you can do is you can ask, "Is the

3 rate of return a firm is earning above the competitive

4 rate?" We know what the competitive rate of return is.

5 Is the rate of return the firm is earning above the

6 competitive rate? This can be a difficult accounting

7 exercise, or better put, this can be a difficult

8 economic exercise using accounting data. You would have

9 to ask over what period of time, how does my answer

10 change depending on risk, not easy to do necessarily,

11 likely to create a lot of controversy.

12 Finally, you could estimate the demand curve

13 facing the firm, and the benefit of estimating the

14 demand curve is you could determine the elasticity and

15 the cross-elasticities that the firm is facing, and that

16 I think gives you useful information about certain types

17 of competition. It raises issues over what time period

18 you estimate the demand curve, but all of those strike

19 me as things you might be forced to do because the

20 question you are trying to answer is really a very

21 difficult one and does not admit a very simple answer.

22 Since time is running a little short, and I know

23 we want to leave time for discussion, I will just

24 mention two topics, and maybe we can come back to it in

25 the panel discussion. What is the distinction between

67

1 market power and monopoly power? And then second, what

2 does it mean to have individual market power in an

3 oligopoly?

4 Well, let me just conclude, screens can help but

5 can also become a danger if one loses sight of the

6 ultimate goal of maximizing consumer welfare. You

7 should create rules to fit the act, paying special

8 attention to how the rules applied to this particular

9 act will chill competition.

10 The profit sacrifice test worries me a bit for

11 the reasons I have explained, and I would use it

12 sometimes but certainly not regard it as a general

13 principle. With regard to market definition, I still

14 think it is a useful discipline, especially when you get

15 to court, for judges and juries to go through, because

16 it helps structure the analysis, but there is an

17 inherent lack of precision in its application in Section

18 2 cases that might be worrisome, and there may well be

19 cases where its use could be misleading.

20 Thank you.

21 (Applause.)

22 MR. BLUMENTHAL: Well, thank you to the panel

23 for very, very interesting perspectives. We have about

24 20 minutes for panel discussion, and because of airline

25 schedules, we are going to have to end promptly at the

68

1 appointed hour of 4:00, but I guess let me turn to the

2 first two speakers, who are agency speakers, and ask

3 whether you have any comments you would care to offer on

4 the thoughts from either of the two professors.

5 CHAIRMAN MAJORAS: Wow, I have a lot that I

6 could comment on. I thank both of them, because those

7 were very thoughtful presentations, as we expected, and

8 I very much appreciate that.

9 One thing that I was curious about is Herb's

10 advice that the agencies should step in a bit more in

11 looking at restraints on innovation, and I was

12 wondering, Herb, if you had any particular hypothetical

13 or context involved or if you can be even more critical

14 of what we have not done in the past, if that is easier

15 for you, in terms of where we could be looking for such

16 a thing, because it is true that we constantly talk

17 about promoting innovation and how important that is to

18 our work and how important it is that we not inhibit it,

19 but I find not only in enforcement but in forming policy

20 and explaining to courts, it is very difficult for

21 people to get their arms around it, because it is such

22 an amorphous concept, and I am not sure we are very good

23 at it.

24 DR. HOVENKAMP: It is an amorphous concept. In

25 part, though -- and I think this is an important

69

1 principle -- it is the one place where the IP laws and

2 the antitrust laws tell the same story. We always talk

3 about this tension between innovation and competition.

4 It is hard to defend restraints on innovation when they

5 are defined properly, and the fact is that the

6 enforcement record has not given us cause for much hope.

7 I can think of -- right now, off the top of my head, I

8 can think of three situations.

9 One is the air pollution cartel cases of the

10 1970s. I suspect some of you have not been around long

11 enough to remember those, but those were challenges by

12 farmers, class action, brought by agricultural groups to

13 an alleged agreement among the major automobile

14 producers for restraining the development of engines

15 that produce less pollution, and what they wanted to

16 collect as damages was harm to their crops that accrued

17 as a result.

18 Well, just stating it tells you why that

19 antitrust case is not going to go anywhere, right? You

20 would have to determine what was lost, you would have to

21 determine what the impact of these more fuel-efficient

22 engines would have been, how the farmers might have

23 evaded it. It would be totally impossible for private

24 plaintiffs to collect damages in such a case.

25 Another one is the series of cases, a fairly

70

1 large number of cases, that made it to the circuits

2 against the cigarette companies in the late 1990s

3 alleging various cartels to refrain from developing

4 healthier cigarettes. To the best of my knowledge,

5 every single one of those cases was dismissed, at least

6 the ones that went to the circuits, were dismissed on

7 grounds of standing, causation, provable injury, maybe a

8 couple of other reasons, but they all had to do with

9 ability to prove injury.

10 Then, of course, there is the conclusion in

11 Judge Jackson's Microsoft opinions, affirmed by the D.C.

12 Circuit, that Microsoft's attempts to force Intel not to

13 develop a Java-enabled chip, a chip that could have

14 spoken multiple-processor languages, something that Bill

15 Gates feared at the time, was unlawful, and that

16 complaint allegation has actually been included in a few

17 of the indirect purchaser cases against Microsoft, and

18 to the best of my knowledge, not a single plaintiff has

19 ever collected a dollar for those failures of

20 innovation.

21 So, one of the problems is that restraints on

22 innovation are very likely extraordinarily harmful if

23 they occur. If so much of our economic growth comes

24 from innovation, then restraints on innovation could, in

25 fact, be very harmful.

71

1 Number two, because of the nature of innovation,

2 because you can never predict where it will go, how much

3 value it will produce, how much people would be injured

4 by such a restraint, it is not good grist for private

5 damages actions under Section 4 of the Clayton Act.

6 I do not know if that is a satisfactory answer,

7 but it is just an area where I would like to see more

8 attention paid.

9 CHAIRMAN MAJORAS: Well, it does help, because

10 the other difficulty we have, not only in antitrust but

11 in forming just good economic policy, is how you balance

12 innovation incentives generally. I mean, this is the

13 classic case that I was just discussing earlier with

14 some folks in the consumer protection arena about

15 pharmaceutical companies and their incentives to

16 innovate and how this group of folks just came back from

17 Europe and said, "Okay, now we need to push in the

18 United States what other countries are doing," and I

19 said, "Well, what is going to happen to the incentives

20 to innovate if every country in the world put caps on

21 pharmaceutical prices?" But because the harm to

22 innovation from that is so hard to measure in the policy

23 arena, people do not necessarily, I think, take it into

24 account.

25 MR. BARNETT: Well, I guess a couple of

72

1 reactions. First, I am extraordinarily pleased with the

2 contributions that the two of you made. They not only

3 frame some interesting issues and pose some hard

4 questions, but in some instances, also provided

5 suggested answers or even policy directions, which is a

6 great way to start off the hearings, and we very much

7 appreciate hearing it.

8 I think, Herb, it is interesting to hear your

9 views on refusals to deal, whether we should be involved

10 in that or not, and I guess I was going to ask you, you

11 have touched on a little bit the issue of innovation,

12 and given its importance, does that mean we should be

13 more or less aggressive in intervening in the sense that

14 is there a greater risk of deterring innovation if you

15 get involved in some of these areas? I do not know the

16 answer to that. I just raise it as a thought.

17 Let me, if I can, pause and ask a -- well, I

18 will ask you one question, though, before I get to that.

19 I have always been of slightly two minds. I mean, I

20 quoted Capitalism, Socialism and Democracy, the Joseph

21 Schumpeter discussion of monopolies, and I have been

22 taken with this gales of creative destruction imagery.

23 Now, there is a lot in there that sort of apologizes for

24 and/or justifies monopolies as good in and of

25 themselves, and I guess some of hearing your remarks and

73

1 all caused me to want to clarify that I do not agree

2 with everything that is in that book in terms of

3 monopoly, but I do agree with the importance of

4 innovation and not losing sight of that.

5 But talking about this, you just made a comment

6 about remedies and a distinction, I think, between

7 damages and the Government. I would be interested to

8 hear more of your views on that point, if you think that

9 would be a good idea, to draw such a distinction.

10 DR. HOVENKAMP: The distinction between?

11 MR. BARNETT: Different remedies for

12 governmental ver -- well, for I would say maybe

13 injunctive versus damages type relief.

14 DR. HOVENKAMP: It has always been my opinion,

15 which the Supreme Court has rejected in California

16 versus American Stores, that structural remedies should

17 be the prerogative of the Government. That is not the

18 law, let's make sure everybody's clear about that. In

19 California versus American Stores, where the plaintiff

20 was the Federal Trade Commission, so it's a private -ï½­

21 I'm sorry, was the State -- was the Attorney General of

22 the State of California, so it was a private party in

23 this rather unusual way we treat states attorneys

24 general in antitrust, but I believe structural remedies

25 ought to be something that only the Government ought to

74

1 do, because they have such extraordinary spillovers.

2 I mean, we have some confidence that the federal

3 antitrust agencies, because of their diversity, the

4 diversity of the industries that they represent, and

5 because of the high quality of the people that run them,

6 they are not captured industries, by and large, that we

7 may not have that level of confidence about

8 industry-specific agencies.

9 The one thing we know about private plaintiffs

10 is that they are always captured, right? Private

11 lawyers serve their clients. Their clients are

12 completely self-interested, and as a result, they do not

13 take overall effects into account, and as a result, my

14 own view is that structural relief, such as divestiture,

15 should simply be denied to private plaintiffs.

16 Is that an answer to your question?

17 MR. BARNETT: I just was interested in your

18 perspectives, so thanks.

19 MR. BLUMENTHAL: Herb, Dennis had spoken after

20 you did, so I guess I would ask whether you had any

21 rejoinder to any of Dennis' comments.

22 DR. HOVENKAMP: No, and I mean, Dennis knows a

23 million times more about this market definition stuff

24 than I do, and so I am very elated that he actually

25 agreed with me about most of it. I think you can

75

1 develop a few presumptive rules for dealing with market

2 power issues. For example, where the defendant, the

3 firm under investigation, faces competition from other

4 firms that use the same technology and apparently have

5 the same cost structure, then I think the inference of

6 competition, competitive pricing, is higher than in a

7 case like Cellophane. I think what makes the Cellophane

8 case so extraordinary is that the defendants, DuPont,

9 plus Sylvania as its licensee, produced Cellophane, but

10 the stuff that the Government ended up throwing into the

11 market was brown wrapping paper, tin foil, glassine -ï½­

12 I'm not even sure I know what glassine is -- but these

13 were things that used different raw materials, they

14 almost certainly used different technologies for

15 producing them, and simply to conclude that such things

16 are in the same market simply because of high observed

17 cross-elasticity of demand is a very serious error.

18 It's an error courts continue to make.

19 There are several cases, for example, that have

20 concluded that rental videos, films shown in movie

21 theaters, and films shown on television are all in one

22 relevant market simply because they make the obvious

23 observation that if you look at any particular person,

24 sometimes she goes and rents a video, sometimes she goes

25 to a theater, sometimes she watches a movie on

76

1 television, without asking the question whether any one

2 of those technologies is sufficient to hold the other

3 ones to cost.

4 I would rather hear Dennis talk about this.

5 DR. CARLTON: I agree, but I will talk about

6 something about innovation, if that is all right,

7 because I think that is an important topic.

8 It is absolutely right, our standard of living

9 has increased because of technological change. The

10 question about what the implication of that is for

11 antitrust I think is easy to state. It is very hard to

12 implement. Part of the reason it is hard to implement

13 is because what we know about innovation and the market

14 structure that induces innovation is a lot less than

15 what we know about concentration and the effect of

16 concentration on raising price.

17 The evidence, I think most economists would say,

18 that competition is good for innovation. On the other

19 hand, if you pushed them and said, "Do you think four to

20 three makes a big difference when firms are innovating,"

21 you know, it is pretty hard to find evidence, I think,

22 and moreover, innovating in what? The idea of, for

23 example, an innovation market I think is not

24 particularly helpful except maybe in rare exceptions

25 like drugs where you can redefine it to be not an

77

1 innovation market, but rather, a future product market.

2 In drugs, you can look at the pipeline, and you know

3 what is coming out. So, it is not really an innovation

4 market as much as it is a prediction of future products.

5 It is much harder when you are looking at just

6 firms doing R&D in a general area to say, "What are they

7 doing R&D on? Is it even on similar products?" And if

8 you ever go back and do an experiment, like if you look

9 at major product introductions, who did them, sometimes

10 the people doing them are often outside the industry,

11 and no one predicted they would come in. So, I think

12 the difficulty in this area, when antitrust tries to

13 deal with it, is the difficulty that you have -- that

14 anyone would have -- in predicting who is going to be

15 the innovator. That is why I think it is very hard.

16 The other reason why I think it is very hard is

17 innovation markets inherently involve intellectual

18 property. People want to protect their intellectual

19 property. There are a lot -- if you look at, you know,

20 just -- there are some very interesting articles on the

21 Internet, who has exclusivity rights on certain types of

22 information that is generated. Restrictions, vertical

23 restrictions on intellectual property, who can use them,

24 who cannot use them, are very widespread, and therefore,

25 if you do have a big success, that will complicate some

78

1 antitrust enforcer's life, because they will say, "Well,

2 look, a big hit in the vertical restrictions, I am a

3 little worried," but you have got to go back and step

4 back and say, "Well, the whole idea of innovating was to

5 get the returns from this intellectual property, and

6 they would not have done it if they could not put the

7 restrictions on it."

8 So, I think it is a very difficult area. I

9 think it is made more difficult by our intellectual

10 property laws, and the problem that I really see arising

11 is that because so many patents are given for products

12 that may not be all that valuable, you create the

13 problem that the property right in a truly valuable

14 innovation is not that valuable, because now you have to

15 get cross-licenses from lots of other people, and that

16 is I think creating a lot of the antitrust problems, and

17 maybe the best way to fix it is not just through

18 antitrust but through reforming IP.

19 CHAIRMAN MAJORAS: As long as we are on the

20 subject of innovation, in the two minutes we have left,

21 does either of you have an opinion on the recent

22 phenomenon that we are seeing in the patent area known

23 as patent trolls, people who go out and buy up what some

24 people might term useless patents -- I don't know if you

25 can go that far -- and some of those people doing it, in

79

1 fact, as I understand it are plaintiff's lawyers, and

2 then wait until somebody creates an invention, and that

3 bundle of patents might look, you know, like it could

4 have been infringed, maybe it is, maybe it isn't, and

5 then pounce and say, "Okay, now you owe me really high

6 royalties."

7 And there are obviously many different

8 variations on this theme, you know, those that have zero

9 interest in ever making anything or innovating I think

10 are the ones that people refer to as patent trolls. Do

11 you have any comment on whether that is injuring

12 competition or something that any of us should be

13 worried about?

14 DR. HOVENKAMP: Well, my reaction to that is

15 that whether it's good or bad, it is fundamentally not

16 an antitrust problem. I mean, in order to be an

17 antitrust problem, it either has to monopolize the

18 market or create a dangerous probability of doing so or

19 else it has to be an agreement that restrains trade, and

20 simply surprising people with a patent and claiming high

21 royalties in and of itself does not do that.

22 Now, that does not mean if it were not used in

23 conjunction with some other practice, it couldn't sum up

24 to an antitrust violation.

25 DR. CARLTON: I would just add, I basically

80

1 agree with what Herb said. I think that it is a problem

2 for reform of the patent laws, and there are at least

3 two thoughts I have. The first is that one of the

4 things that gives a troll great power is the ability to

5 get an exclusion right, and you might want to encourage

6 the courts to give the surprised infringer time to

7 invent around the patent if that is possible, and that

8 can get rid of sometimes a lot of the pressure.

9 The other question you can ask is, are they ever

10 intending to use this themselves and implement it, and

11 that may influence how you treat them.

12 Having said that, you know, an economist always

13 has two sides to a story. A patent -- you may call them

14 a patent troll, but the flip side is I'm an inventor,

15 maybe I want to give my invention to someone else to

16 figure out, you know, all the licensing problems. So, I

17 mean, there is a flip side.

18 CHAIRMAN MAJORAS: Yes.

19 MR. BARNETT: Can I just ask quickly if based on

20 your remarks, Dennis, if you would be willing to get

21 into a locked room with Greg Werden for a session with a

22 tape recorder and we could solve nine out of every ten

23 Section 2 problems.

24 DR. CARLTON: I would be delighted.

25 MR. BLUMENTHAL: Dennis, if I could I guess ask

81

1 one last question in the two minutes we have, you had

2 thrown out a teaser at the end about the distinction

3 between market power and monopoly power, an issue that

4 was of some interest to the staff, although we had not

5 quite posed it, and I might add, it was something I

6 recalled George Stigler having denied as a distinction

7 in the Data General case about 20 years ago, although

8 the Ninth Circuit disagreed with that view. So, perhaps

9 there is a distinction after all.

10 Did you have in mind something more than just

11 quantum? Is it a character of the difference or -ï½­

12 DR. CARLTON: Yes, actually, it is in my

13 textbook, so I know it very well. One definition of

14 market power could simply be that price is in excess of

15 marginal cost. That is a logical definition. Whether

16 you want to use it in a court, you know, I think that is

17 a different question. You might want to define it by

18 how much before you say it is something that triggers

19 some action, but you can distinguish market power from

20 monopoly power by saying monopoly power is not just

21 price in excess of marginal cost, but also, profits in

22 excess of the competitive rate of return. So, that is

23 at least a logical distinction.

24 I should say in the third edition of my book, I

25 dropped that paragraph, because no one seemed to pay

82

1 attention to it. Now, I change a lot of things when I

2 revise my book, and most people never comment, but on

3 this, I did get some comments, and people said, "We

4 thought it was useful," even though I had never seen

5 anyone -- I was unaware anyone had cited it, and I think

6 it is useful, so I have put it back in. Now, so, that

7 is an answer.

8 Now, whether that should be what courts use when

9 they are deciding on an exclusionary act is a slightly

10 different question, because I think courts really should

11 be asking a slightly different question, which is, let

12 us suppose you are talking about exclusive dealing. I

13 don't really care about what the rate of return is, I

14 don't think. I think what a court is saying is if you

15 are engaged in an act and there are other

16 characteristics of the market that make that act have a

17 greater effect than if you didn't, say, have as large a

18 market share, then I am going to be worried about that,

19 and that really does not have to do with rates of

20 return.

21 So, you know, I can see why courts could be

22 asking are the characteristics of the market that make

23 the exclusionary act under discussion much more a

24 competitive concern than not, whether that -- I don't

25 know that that necessarily corresponds to the

83

1 distinction I gave you. The distinction I gave you does

2 strike me as a sensible economic distinction.

3 MR. BLUMENTHAL: Well, thank you, and with that,

4 we are out of time. Let me ask you to join me in

5 thanking the panel for a great presentation.

6 (Applause.)

7 MR. BLUMENTHAL: We look forward to receiving

8 your comments in the months ahead. We hope to see you

9 Thursday and in the hearings to come. Thank you all.

10 (Whereupon, at 4:00 p.m., the hearing was

11 concluded.)

12
13
14
15
16
17
18
19
20
21
22
23
24
25

84

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

2 DOCKET/FILE NUMBER: P062106

3 CASE TITLE: SECTION 2 HEARING

4 DATE: JUNE 20, 2006

5
6 I HEREBY CERTIFY that the transcript contained

7 herein is a full and accurate transcript of the notes

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

9 FEDERAL TRADE COMMISSION to the best of my knowledge and

10 belief.

11
12 DATED: 7/5/06

13
14
15
16 SUSANNE BERGLING, RMR-CLR

17
18 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
20 I HEREBY CERTIFY that I proofread the transcript

21 for accuracy in spelling, hyphenation, punctuation and

22 format.

23
24
25 DIANE QUADE

Updated June 25, 2015