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

1UNITED STATES FEDERAL TRADE COMMISSION

2and

3UNITED STATES DEPARTMENT OF JUSTICE

4

5

6

7SHERMAN ACT SECTION 2 JOINT HEARING

8 WELCOME AND OVERVIEW OF HEARINGS

9TUESDAY, JUNE 20, 2006

10

11

12

13

14 HELD AT:

15600 PENNSYLVANIA AVENUE, N.W.

16 WASHINGTON, D.C.

17UNITED STATES FEDERAL TRADE COMMISSION

18 HEADQUARTERS BUILDING, ROOM 432

192:00 P.M. to 4:00 P.M.

20

21

22

23

24Reported and transcribed by:

25Susanne Bergling, RMR-CLR

2

1MODERATOR:

2WILLIAM BLUMENTHAL

3Federal Trade Commission

4

5PANELISTS:

6

7Deborah Platt Majoras

8Thomas O. Barnett

9Dennis Carlton

10Herbert Hovenkamp

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

3

1C O N T E N T S

2

3Introduction

4

5Presentations:

6     Deborah Platt Majoras

7     Thomas O. Barnett

8     Herbert Hovenkamp

9     Dennis Carlton

10

11Moderated Discussion

12

13Adjournment

14

15

16

17

18

19

20

21

22

23

24

25

4

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

2-  -  -  -  -

3MR. BLUMENTHAL: Ladies and gentlemen, good

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

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

6Department Antitrust Division and Federal Trade

7Commission hearings into Section 2 of the Sherman Act.

8The purpose of these hearings is to explore how

9best to identify anticompetitive exclusionary conduct

10for purposes of antitrust enforcement. We are

11envisioning a series of hearings that will kick off

12today and will continue through December, probably two,

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

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

15have another hearing on Thursday of this week examining

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

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

18breather, and then resume in September with what would

19then be a series of examinations.

20The agencies are expecting to focus on legal

21doctrine, on jurisprudence, economic research, and

22business and consumer experience. We have a Federal

23Register notice that is outstanding. We invite public

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

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

5

1have an opportunity to submit comments on the topics

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

3time through the final hearing, that is, through

4December, although the earlier the better for our

5purposes.

6We are honored today to have a special panel to

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

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

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

10speaking, first we have Deborah Platt Majoras, Chairman

11of the Federal Trade Commission. Thomas Barnett, the

12Assistant Attorney General for Antitrust in the Justice

13Department. Herb Hovenkamp, who is probably known to

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

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

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

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

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

19out, The Antitrust Enterprise: Principle and Execution,

20available through Harvard University Press, with an

21imprint of this year.

22And finally Dennis Carlton, also known to many

23of you in many capacities, most notably professor of

24economics at the Graduate School of Business at the

25University of Chicago, former president and still very

6

1active in Lexicon, frequent expert witness, author of

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

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

4that. You all know Dennis Carlton.

5I have several preliminary announcements that I

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

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

8courtesy, if any of you have cell phones or

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

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

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

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

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

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

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

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

17And finally, the legally mandated announcement

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

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

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

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

22Pennsylvania Avenue side, and after leaving the

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

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

25Sculpture Garden, catty-cornered across the street at

7

17th and Constitution, and we will assemble there, where

2noses will be counted.

3With that, it gives me great pleasure to turn

4the podium over to the Chairman of the Federal Trade

5Commission, Debbie Majoras.

6(Applause.)

7CHAIRMAN MAJORAS: Good afternoon, everyone, and

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

9and colleague, Assistant Attorney General Tom Barnett,

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

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

12and we are privileged to have two of our most

13distinguished antitrust scholars here, Professors

14Hovenkamp and Carlton.

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

16important to reflect on why we are undertaking it.

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

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

19study of the economic performance of 13 nations seeking

20to understand globalization, and more fundamentally, the

21disparities between rich and poor. The study showed

22that levels of productivity made the difference between

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

24difference in the levels of productivity? The answer

25they found was "undistorted competition in product

8

1markets."

2In his book in which he reports the results of

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

4up attributing most of the differences in economic

5performance to differences in labor and capital markets.

6This conclusion is incorrect. Differences in

7competition in product markets are much more important."

8McKinsey also asked why the highly productive

9United States has higher competitive intensity than

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

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

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

13view that the purpose of an economy was to serve

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

15continue to "hold this view more strongly than almost

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

17"Consumers are the only political force that can stand

18up to producer interest, big government, and the

19technocratic, political, business, and intellectual."

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

21Antitrust Division have the responsibility to ensure

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

23and that consumers are protected not from markets but

24through markets unburdened by anticompetitive conduct

25and government-imposed restrictions. This work is

9

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

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

3has substantially deregulated critical industries,

4including transportation, telecommunication and energy,

5to the substantial benefit of the economy and consumers.

6As government regulators have given way to free markets,

7much of the responsibility for protecting consumers

8shifts to competition agencies and courts. While

9competition is distorted when governments regulate or

10intervene excessively, it also is true that private

11actors can and do distort competition.

12Breaking up cartels, preventing mergers that

13will substantially reduce competition, and halting

14conduct that goes beyond aggressive competition to

15distorting it is vital to promoting vigorous competition

16and maximizing consumer welfare, and we have developed a

17great deal of consensus regarding appropriate antitrust

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

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

20enforcement has become more transparent and predictable,

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

22participants to make decisions about their own conduct.

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

24still vexes us. Even though we can find some

25respectable measure of consensus around principles that

10

1should apply, we find a range of opinions from

2knowledgeable people about how to apply those principles

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

4proper test that our agencies should apply and that

5courts should apply to conduct of the single firm with

6market power now has dominated our antitrust debate for

7several years.

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

9past quarter century, economic systems in which the

10state owns the firms and central planners set out prices

11and levels of output have given way to competition where

12the forces of supply and demand determine prices and

13allocate the resources, and we have worked hard to

14promote the economic and political benefits of markets.

15With attempts to introduce market economies have come

16new competition authorities, today numbering around 100,

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

18countries that for decades have had nearly total state

19control over their economies, like China, are now

20dedicating substantial resources to drafting competition

21laws.

22Currently, the issue of how to evaluate

23unilateral conduct is the most heavily discussed and

24debated area of competition policy in the international

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

11

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

2their policy under Article 82, which addresses conduct

3by dominant firms. Officials from both agencies

4recently held talks with our colleagues in Japan and

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

6on it in the OECD. And since the International

7Competition Network established a working group on

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

9that group, has received expressions of interest from

10more countries wanting to be involved than we have ever

11had in any other working group in the ICN.

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

13nations are facing the challenge of converting from

14state-owned or supported monopolists to markets with

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

16as we ourselves have learned in trying to deregulate

17certain markets like electricity. And, indeed, to

18enforcers in those nations, it then becomes companies

19with market power, not horizontal competitors, that are

20the evil that must be attacked. Second, disagreement

21among competition authorities about how to treat

22unilateral conduct produces uncertainty in national and

23international markets, which reduces the market

24efficiency and imposes costs. And third, the analysis

25of unilateral conduct in the identification of that

12

1which is anticompetitive presents unique challenges that

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

3antitrust concern of conduct between competitors, and by

4now, these unique challenges I think are familiar.

5First and fundamentally -- and we discuss it all

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

7that is it is difficult to distinguish between

8aggressive procompetitive unilateral conduct and

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

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

11antitrust court lies in stating a general rule for

12distinguishing between exclusionary acts which reduce

13social welfare and competitive acts which increase it,"

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

15the Seventh Circuit, "distinguishing between legitimate

16and unlawful unilateral conduct requires subtle economic

17judgments about particular business practices." So,

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

19done well.

20Second, the process of distinguishing between

21permissible and impermissible conduct must be relatively

22consistent and transparent so that firms are able to

23incorporate it into their decision-making. While there

24are relatively few findings of Section 2 liability,

25there nonetheless are a large number of different types

13

1of conduct that may raise competition concerns and would

2fall under Section 2.

3And third, while antitrust practitioners have

4had substantial success devising remedies for joint

5conduct, devising remedies for single-firm behavior

6presents significant difficulties. As Professors Areeda

7and Hovenkamp put it, "By contrast with concerted

8conduct, unilateral behavior is difficult to evaluate or

9remedy by any means short of governmental management of

10the enterprise."

11We have much to work with as we move forward

12with these hearings. Already a number of experienced

13experts have proposed the adoption of a single test for

14evaluating nearly all types of potentially exclusionary

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

16the impact of the conduct on consumer welfare. Others

17support analyzing whether the conduct involves the

18short-term sacrifice of profits. Others support a

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

20engaging in the exclusionary conduct makes sense only

21because it serves to eliminate competition.

22Judge Posner has written that the inquiry should

23focus on whether the conduct excludes other equally

24efficient rivals, and still other practitioners and

25scholars oppose the adoption of any single unilateral

14

1conduct test and instead favor consideration of

2different tests for particular types of exclusionary

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

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

5type of test or framework that should be used.

6So, proponents of the various tests and

7approaches already have done a very good job of laying

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

9have acknowledged that their preferred approach is

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

11tackle this issue by starting with the conduct itself.

12The hearings will have panels that will focus on

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

14can implicate liability. We want the panels to discuss

15the conduct from the market perspective, from the ground

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

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

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

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

20look at whether firms in competitive markets also engage

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

22want these discussions, to the extent possible, to

23include knowledgeable businesspeople or at least their

24advisers, and from these discussions, we then should

25endeavor to develop sign posts for when the conduct may

15

1harm competition and when it typically does not. From

2the sign posts, we hopefully can draw some guiding

3principles, and only then should we turn to examining

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

5such conduct and then determining what workable rules

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

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

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

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

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

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

12approach at least some fraction, hopefully a significant

13fraction, of single-firm conduct we encounter.

14In these discussions, we need to be careful not

15to permit labels or semantic differences to get in the

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

17have been worried that people are actually talking past

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

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

20not have much practical application in the marketplace.

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

22antitrust practitioner about these hearings and about

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

24he thought the Section 2 issues were very important,

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

16

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

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

3I do think we start with some substantial

4consensus about core underlying principles and factors

5that should underlie any evaluation of unilateral

6conduct.

7First, the only type of unilateral conduct that

8should implicate the antitrust laws is conduct that

9produces durable harm to competition, leading to higher

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

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

12particular companies, the health of the companies

13themselves is not the concern of antitrust law.

14Second, there is consensus that antitrust

15standards that govern unilateral conduct must not in

16themselves deter competition, efficiency, or innovation,

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

18worry about false positives. Obviously pervasive and

19aggressive competition in which firms consistently try

20to better each other by providing higher quality goods

21and services at lower cost is crucial to maximizing

22consumer welfare. So, the antitrust laws should then

23never condemn market power that is obtained through the

24development of superior products and services,

25regardless of how many competitors are driven from the

17

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

2been accepted by the courts.

3Third, there is consensus that the standards for

4evaluating unilateral conduct must be clear and

5practical to administer or as practical as they can be

6to administer. The most analytically sound principles

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

8them when they are making their business decisions.

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

10apply rules as well.

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

12use the hearings to continue developing my own thinking

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

14broad principles, with a number of other hypotheses.

15First, any legal framework needs to avoid

16second-guessing business judgments that were objectively

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

18examination of the hypothetical effects of alternative

19courses of conduct is likely to chill legitimate

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

21framework must be able to evaluate conduct that both

22generates efficiencies and produces anticompetitive

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

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

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

18

1for the fact that certain types of unilateral conduct

2are significantly more likely to cause competitive harm

3than others. For example, most would agree that

4unilateral above-cost pricing at monopoly levels should

5not be condemned under the antitrust laws. Similarly,

6behavior that some commentators have termed "cheap

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

8unlawfully extend the life of a patent, is generally

9viewed as unlawful exclusionary conduct. And this may

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

11need a broad framework that can accommodate a spectrum

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

13proposals have been made for how we might think about

14the distinctions that could be made, including Deputy

15Bureau of Competition Director Ken Glazer's proposal

16that we analyze conduct by distinguishing between

17conduct that's coercive versus incentivising.

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

19applied what I view as a sensible weighted balance

20approach to Microsoft's conduct that's largely

21consistent with the three principles I just discussed.

22Some have criticized the framework used in Microsoft as

23insufficiently structured or unfocused, and I understand

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

25was actually applied, it may be a workable framework

19

1that incorporates the principles on which we have wide

2consensus. I mean, perhaps the same criticism about

3being unstructured could be applied to the Section 1

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

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

6balancing has been weighted I think in the right

7direction.

8First, the Microsoft court did not attempt to

9substitute ex post facto its judgment for the business

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

11actions might have been better overall for consumers.

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

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

14the prices charged to consumers.

15The Microsoft court also demonstrated that to

16evaluate whether certain types of unilateral conduct

17violate the antitrust laws does require an examination

18of both likely anticompetitive and procompetitive

19effects. For example, the Court analyzed the legality

20of a Microsoft license provision that prohibited OEMs

21from modifying the initial boot sequence. Microsoft did

22not dispute that that restriction limited competition

23against IE. The Court nonetheless held that the

24restriction was not a violation because it concluded

25that preventing the Windows desktop from ever being seen

20

1at all in the boot sequence was a substantial alteration

2of Microsoft's copyrighted work that could produce harm

3that outweighs the marginal anticompetitive effect of

4the prohibition. The Court performed this same analysis

5across two dozen types of conduct, examining both the

6anticompetitive effects and procompetitive

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

8not chill procompetitive behavior.

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

10did not consider all types of unilateral conduct to

11raise equal concerns under the antitrust laws. For

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

13skeptical about claims that a dominant firm's design

14changes harm competition and, by implication, violate

15the antitrust laws.

16One final note about the hearings. I hope that

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

18produce a productive discussion. It simply is not

19possible to implement sound competition policy for

20single-firm conduct without giving careful thought to

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

22issues relating to remedies have not received extensive

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

24although it received and still receives a bit of

25notoriety, I have been stricken by how few productive

21

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

2that accepted the DOJ remedy while rejecting other

3remedies have actually occurred, and while that might

4have stemmed from some of the market dissatisfaction

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

6Section 2 remedy issue the prominence that it deserves

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

8cases, you know that devising and drafting remedial

9provisions in monopolization cases can be more difficult

10than determining whether a violation has even occurred.

11At bottom, through these hearings and through

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

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

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

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

16competition undistorted while not distorting it

17ourselves in the process.

18So, I thank you again for attending the opening

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

20contributions. Thank you very much.

21(Applause.)

22MR. BLUMENTHAL: Thank you, Chairman.

23General Barnett?

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

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

22

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

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

3help in organizing and planning and discussing the

4issues that have brought us together. Both Chairman

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

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

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

8Division staff who have worked very hard to prepare for

9the session today and who will continue to work hard

10over the months ahead.

11I also want to thank Herb and Dennis for

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

13such distinguished commentators, and feel that it is a

14terrific start to these important hearings.

15I am going to start with the same issue that

16Debbie started with, which is why are we sponsoring

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

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

19said, but it was done unilaterally and without

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

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

22We do feel that it is important. As Debbie

23said, unilateral conduct can harm consumer welfare. I

24think there is a consensus that the focus of the

25antitrust laws in the United States is to preserve and

23

1promote or prevent harm to consumer welfare and that

2unilateral conduct is an important element of that.

3I also agree that this is the area of probably

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

5consensus within Section 2, but there are significant

6areas where I think we have room for further

7understanding.

8These hearings, with the combination of legal,

9economic, business and governmental/private

10perspectives, provide us with a unique opportunity to

11advance our understanding, and I believe that that will

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

13provide helpful guidance to the courts, guidance to the

14business community, and as Debbie quite eloquently put

15it, to the international community that is now focused

16on this issue.

17There is a long tradition of the agencies

18leading the development of competition law. I need only

19point to Don Turner and the 1968 Merger Guidelines and

20the formulation by Bill Baxter in 1982 to provide an

21example of what has become the standard reference for

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

23around the world in many ways.

24With respect to the international community,

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

24

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

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

3continent, and one example that sort of helped drive

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

5in Budapest of Southeastern European former Soviet block

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

7Antitrust Division often talks about, which is the

8importance of cartel enforcement, and one of the

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

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

11that our markets had enough participants so that they

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

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

14former state-owned enterprise and how we can introduce

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

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

17important here, but I think its importance abroad cannot

18be over-emphasized.

19The Supreme Court, to its credit, addressed the

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

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

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

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

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

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

25

1danger of deterioration in quality of the monopolized

2article, which it deemed was the inevitable result of

3the monopolistic control of its production. Price

4increases, output reductions, quality deterioration,

5those are still the same three touchstones that we look

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

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

81910.

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

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

11dichotomy or two different views of monopoly. While we

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

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

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

15quiet life. He talked about the fact that the

16monopolist may not be out there trying to get the

17highest price he absolutely can get, maximizing in the

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

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

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

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

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

23focused in terms of preventing.

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

25last year articulated a different view of monopoly. In

26

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

2possession of monopoly power and the concomitant

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

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

5opportunity to charge monopoly prices, at least for a

6short period, is what attracts business acumen in the

7first place. It induces risk-taking that produces

8innovation and economic growth."

9All the way back in 1942, in Capitalism,

10Socialism and Democracy, Joseph Schumpeter talked about

11a similar process called creative destruction or the

12gales of creative destruction, and I compliment my staff

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

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

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

16because it talks about how the marketplace is a rough

17place. It involves vigorous aggressive activity, people

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

19through that destructive process that you get creation.

20Indeed, a similar image I was thinking about

21recently, when somebody was talking to me about the

22National Forest Service, I grew up watching the

23commercials about Smokey the Bear and how forest fires

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

25forest fires? It turns out the National Park Service

27

1realizes that preventing forest fires can be a bad

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

3much bigger, larger, hotter fires that cause more

4permanent destruction to the ecosystem when they do

5occur. Periodic smaller fires are actually a good and

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

7illustration of this basic image. Competitive, creative

8destruction in the marketplace is something that we want

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

10Debbie was talking about.

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

12monopoly, as a bad thing that causes sloth and

13relaxation and a lack of competitive drive versus the

14benefits of creative destruction, the opportunity to get

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

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

17Grant expressing some frustration about the treatment of

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

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

20invents a bread machine. It will produce terrific

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

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

23shortly owns the market for bread in the United States

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

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

28

1an antitrust case against him for making too much money

2on the backs of consumers and driving everybody else out

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

4this frustration.

5"You're gouging on your prices

6if you charge more than the rest

7But it's unfair competition

8if you think you can charge less!

9A second point that we would make

10to help avoid confusion:

11Don't try to charge the same amount!

12That would be collusion.

13You must compete -- but not too much

14for if you do, you see

15then the market would be yours -

16and that would be monopoly!

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

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

19successful competitor, having been urged to compete,

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

21So, having expressed that frustration back in

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

23esteemed both I would say academics and judicial members

24of the antitrust community, Richard Posner, Judge

25Posner, remarked just last year, "Antitrust policy

29

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

2biggest substantive issue facing antitrust today.'"

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

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

5"Notwithstanding a century of litigation," 96 years

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

7meaning of exclusionary conduct under the Sherman Act

8remain poorly defined."

9Now, there are areas where I think there are

10relatively easy answers. Doug Melamed has written about

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

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

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

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

15are some fairly easy candidates for safe harbor

16provisions. If you engage in conduct that merely

17reduces your cost of production, that seems to me

18beneficial to consumer welfare.

19The difficulty lies in cases, as Debbie

20referenced, that have the potential for both beneficial

21cost reductions, innovation, development, integration,

22and at the same time potentially anticompetitive

23exclusion. How do we deal with those situations?

24Well, some relatively recent Supreme Court

25decisions have shown progress in this direction. In the

30

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

2pricing case, it dealt specifically with the issue of

3recoupment and holding that Liggett in that case had not

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

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

6discusses, for example, specifically how harm to a

7competitor does not demonstrate harm to competition.

8There was little doubt in that case that there were

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

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

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

12antitrust problem.

13Second, the Court also talked about the

14practical ability of a judicial tribunal to regulate a

15problem and avoid chilling legitimate price cutting.

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

17administering the law. I would expand that to include

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

19certainly a relevant consideration, and recognizing that

20aggressive price cutting can be beneficial for consumers

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

22effectively a safe harbor against predatory pricing

23claims where the prices were above some appropriate

24measure of cost.

25And the Court expressly acknowledged in creating

31

1the safe harbor that there was at least the theoretical

2possibility that there could be harm to consumers, harm

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

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

5try to ferret out those individual cases.

6More recently, in the Trinko decision, the Court

7obviously had a somewhat more limited holding but

8discussed on a broader basis some of these same similar

9Section 2 issues. It underscored the need for

10administrable rules, clear objective standards. It

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

12that is both clear and administrable by the Court is

13very important, endorsing Professor Areeda, in that no

14court should impose a duty to deal that it cannot

15explain or adequately and reasonably supervise, and

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

17antitrust solutions.

18While I think there are many areas of consensus,

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

20Debbie indicated, our panels are going to focus on

21different aspects of conduct. We will start on Thursday

22with a panel discussing predatory pricing and predatory

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

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

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

32

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

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

3about that, and we are looking forward to hearing

4people's views on that.

5The question on predatory pricing or remedy, are

6you going to enjoin lower pricing? Weyerhaeuser is a

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

8petition pending before the Supreme Court right now. It

9involves a predatory bidding situation. The Solicitor

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

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

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

13reaffirm in the Section 2 context that clear and

14objective standards are extraordinarily important.

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

16talks about prices that are unfair or unnecessarily

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

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

19believes was unfair is not the appropriate criteria for

20determining whether or not there should be Section 2

21liability.

22Refusals to deal will follow. Again, this

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

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

25compelled to deal, under what terms and conditions?

33

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

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

3the experts' views on it.

4Loyalty discounts, another area that we will be

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

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

7case. That involved bundled discounts. That was not

8because we necessarily agreed with the Third Circuit's

9decision or analysis. Indeed, if you parse that

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

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

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

13legal scholars, on this issue of bundled discounts,

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

15whether or not any consensus has developed on any of

16these issues.

17Should it be viewed as a predatory pricing

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

19there safe harbors that can be developed even if we

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

21Tying and exclusive dealing, Debbie mentioned that you

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

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

24benefits associated with it. Tying and exclusive

25dealing can have anticompetitive effects. Look at our

34

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

2we see these practices in competitive markets, and we

3need to better understand what benefits there are and

4when there are not.

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

6toward some more general principles. Is there an

7overarching standard for Section 2 cases and liability?

8We all agree that consumer welfare is an appropriate

9standard. Trying to operationalize that in a particular

10case with particular conduct is more challenging, and

11there is less agreement on that. Debbie outlined the

12range of potential tests. The Antitrust Division in a

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

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

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

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

17the appropriate test ought to be, there is some

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

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

20part of the semantic difference that Debbie was

21referencing. Clarifying some of those things as well as

22the underlying substantive issues I think can be

23beneficial.

24We may look at the issue of whether there are

25different duties or different criteria for tying claims

35

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

2Section 2 of the Sherman Act.

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

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

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

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

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

8antithesis of innovative thinking, can restrain the

9benefits that you may achieve through your innovation

10and development. That is part of the creative

11destruction that we want to encourage, not discourage,

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

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

14possible principles to inform our discussions, I do not

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

16frank exchange of ideas.

17So, first off, individual firms with market

18monopoly power can act anticompetitively and harm

19consumer welfare, and we should seek to identify and

20prosecute such conduct. This is an important first

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

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

23here to better focus and identify those instances where

24there really is harm to consumer welfare.

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

36

1necessarily demonstrate competitive harm. It can

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

3Third, injury to competitors does not

4demonstrate competitive harm, a point that has been

5talked about in a number of contexts.

6Fourth, the need for clear, objective and

7administrable rules, so that businesses, at the time

8they are taking actions, can understand where the lines

9are and can conform their behavior so they are not

10deterred from engaging in procompetitive activity, so

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

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

13Fifth, avoid chilling procompetitive conduct,

14and certainly an interrelated point, self-explanatory.

15And finally, the remedy must promote

16competition. A remedy that harms competition can be

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

18of bearing in mind.

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

20for agreeing to kick off these hearings. We will

21continue again on Thursday. We very much are interested

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

23issues and are excited about the prospect.

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

25(Applause.)

37

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

2and appreciative of being invited here, with particular

3thanks to Chairperson Majoras and General Barnett for

4extending this invitation.

5In keeping with the thrust of this opening

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

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

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

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

10specific practices such as predatory pricing or refusals

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

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

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

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

15private litigants can use some clarification and

16understanding.

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

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

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

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

21much shorter one again on remedies.

22With respect to power, the Merger Guidelines, in

23particular the 1992 Merger Guidelines, the series of

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

25rationalizing and simplifying the approach to market

38

1delineation and assessment of the potential for

2collusion or other types of anticompetitive behavior

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

4Guidelines market delineation sections are relevant to

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

6question that one asks in a Section 2 case is

7fundamentally different from the one that one asks in a

8merger case.

9In a merger case, we generally start out with

10the presumption that a market is more or less

11competitive, it may be oligopolistic or moderately

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

13to know is whether the quality of competition is going

14to deteriorate as a consequence of the merger. In

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

16nontransitory increase in price test, considers whether

17a further increase in price would cause new entry or

18other situations that would make this future price

19increase unsustainable.

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

21presumption is that the defendant or the firm under

22examination is already a monopolist, is already charging

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

24really not the appropriate one in most circumstances,

25although it certainly could be relevant in certain cases

39

1like those involving an attempt to monopolize where the

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

3being assessed.

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

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

6involves something that in monopolization law we call

7the Cellophane fallacy or the fallacy of inferring that

8a firm lacks power because there is high

9cross-elasticity of demand with the products of others

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

11multiply that examination by asking what the response

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

13already a monopolist, you might very well conclude that

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

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

16so much substitution away from the dominant firm's

17product that the price increase would be unprofitable.

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

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

20fact that that firm is already a monopolist and

21presumably already charging its profit-maximizing price.

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

23concern to the agencies as they go through these

24hearings is to pay some special attention to the

25formulation of usable presumptions that single firms can

40

1use for assessing whether they have individual market

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

3inquiry.

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

5certain approaches that we have more or less given short

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

7mean that we will not look at residual elasticity of

8demand, which looks at the existing power that firms

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

10margins or rates of return. Some of these approaches

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

12mean that they cannot be rehabilitated.

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

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

15definition of monopolizing conduct from The Antitrust

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

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

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

19explain in a little while. The Antitrust Law Treatise

20defines exclusionary conduct as conduct that is, number

21one, reasonably capable of creating, enlarging or

22prolonging monopoly power by impairing the opportunities

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

24consumers at all or is unnecessary for the particular

25consumer benefits that the acts produce; or three,

41

1produces harms that are disproportionate to the

2benefits; and finally, the assessment of the conduct

3must be within the administrative capacity of the

4antitrust tribunal.

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

6not particularly helpful to assessing particular

7instances of exclusionary conduct if it is the only

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

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

10no further instruction and ask whether the defendant's

11conduct was exclusionary, but the test was never

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

13basic principle to be used in conjunction with specific

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

15my view that that is fundamentally what Section 2

16conduct jurisprudence needs to do.

17I think there are very, very helpful general

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

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

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

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

22test that Chairperson Majoras mentioned in her talk,

23which is conduct which under the circumstances is

24capable of excluding an equally efficient rival. Once

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

42

1negatives, but they are good starting places.

2However, none of them is a substitute for the

3formulation of good technical rules covering individual

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

5intellectual property system, refusals to deal and so

6on, okay?

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

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

9just a few observations about where we are in various

10areas involving specific exclusionary practices and

11where I think some of the problems lie.

12With respect to predatory pricing, I believe

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

14in 1975 and has later been incorporated into The

15Antitrust Law Treatise, plus the elaboration of the

16recoupment requirement in the Brooke Group case in 1993,

17fundamentally set predatory pricing law on the right

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

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

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

21pricing is driven by concerns for variable costs, not

22principally by fixed costs. That does not mean that

23there are not a few problems.

24One problem that I think needs to be assessed is

25the problem of predatory pricing in oligopoly industries

43

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

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

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

4Justice Department might reconsider its long-standing

5opposition to bringing Robinson-Patman Act suits since

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

7create an exception for primary line enforcement given

8the premise that with respect to primary line

9enforcement, the principles that the Court follows are

10basically the principles that are laid out in the

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

12that applies to secondary line enforcement of the

13Robinson-Patman Act need not apply here.

14The problem with predatory pricing and oligopoly

15is that victims have a different set of incentives than

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

17problem involves predatory pricing designed to destroy a

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

19rival clearly has incentives to resist.

20On the other hand, predatory pricing and

21oligopoly frequently is used simply to enforce or bring

22the oligopoly back into order so that the noncompliant

23firm will once again raise its price to the oligopoly

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

25of predatory pricing and oligopoly has are incentives to

44

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

2believe predatory pricing in oligopoly industries is

3fundamentally a more plausible strategy than strict

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

5given somewhat closer scrutiny.

6The other problem has to do with the measurement

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

8measure of cost is incremental cost, which can mean

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

10some kind of additional factor for depreciable long-term

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

12the Areeda-Turner formulation. The average variable

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

14very well in certain kinds of markets that have very

15high fixed cost components and particularly in markets

16that are characterized by a lot of intellectual property

17or certain kinds of public utility or transportation

18markets, such as the airline industry.

19I think Ken Elzinga's analysis in the Spirit

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

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

22predatory pricing cases in the airline industry. It is

23the one industry where predatory pricing claims seem

24plausible, and some attention needs to be paid to

25modifying or, if necessary, rejecting and adopting a

45

1different cost test for such industries.

2On the Weyerhaeuser case and predatory buying, I

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

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

5instruction that General Barnett described that

6permitted a jury to find simply that predatory buying

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

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

9appreciate how frequently such situations can come up;

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

11This is not going to be an idiosyncratic situation.

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

13Circuit's decision is permitted to stand.

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

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

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

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

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

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

20think an average variable cost test might work quite

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

22defendant's resale prices to below its costs.

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

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

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

46

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

2predatory buying is claimed is a relatively low

3proportion of the value of the finished product, because

4in that case, the variation in purchase might actually

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

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

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

8tiny proportion of the value of the finished product.

9In all cases, however, I believe that there

10should be a recoupment requirement equivalent to that in

11Brooke Group and that the Ninth Circuit erred not only

12in its failure to require a showing of prices below

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

14showing of recoupment.

15With respect to patents, there is too much to

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

17general observation. Mr. Barnett mentioned Joseph

18Schumpeter's Capitalism, Socialism and Democracy, this

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

20chapter on creative destruction that the amount of

21welfare contributed to the economy through innovation is

22far, far greater than the amount contributed by moderate

23movements from oligopoly to competitive industries. I

24mean, Schumpeter basically looked at the prior half

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

47

1industrial revolution, of the theorizing of economists

2like Edward Chamberlin and Joan Robinson, who were very

3upset about oligopoly and imperfections in the economy,

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

5American consumers were much, much impoverished compared

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

7could be further from the truth."

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

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

10things Schumpeter concluded is that they came from

11innovation. Schumpeter's premises were formalized and

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

131950s in which Solo himself concluded that as much as 80

14percent of economic gain comes from innovation rather

15than simple improvements in price-cost relationships.

16Now, neither Schumpeter nor Solo was talking

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

18of course, there is this enormous lingering question out

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

20facilitate the optimal amount of innovation or whether

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

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

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

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

25One thing, however, that that work suggests is

48

1that antitrust needs to be much more concerned with

2restraints on innovation. We have generally measured

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

4relationships, deviations from marginal cost pricing.

5Harm to innovation is always included kind of as an

6afterthought, but it has never been very well formalized

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

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

9determining when prices deviate from marginal costs and

10what the price elasticities facing firms are.

11Predicting the consequences of restraints on innovation

12is far more difficult, because innovation always takes

13us by surprise.

14We will never know, for example, what the

15consequences were of Microsoft's successful attempts to

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

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

18things that Bill Gates feared in commoditizing the

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

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

21rightfully skeptical when they turn away private

22plaintiffs who claim that the injury that they suffer is

23an injury caused by a lack of innovation.

24So, I believe this is one area where the

25Government should move into the fore, because they do

49

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

2causation in the strict private plaintiff sense. I

3think restraints on innovation are something that need

4far more development in Section 2 law than they have

5received in the past.

6With respect to vertical exclusion, I just have

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

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

9government enforcement to move away from Section 1 of

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

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

12enforcing laws against tying or tying-like practices and

13exclusive dealing, and I believe that is the correct

14movement. Fundamentally, tying and exclusive dealing

15ought to be regarded as dominant firm exclusionary

16practices. They are rarely anticompetitive at

17nondominant levels, and fundamentally, they do not

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

19Unlike resale price maintenance or Sylvania-style

20restraints, they are typically not the product of

21bargaining and traditional agreement between dealers and

22manufacturers.

23No, most tying and most exclusive dealing is

24imposed by manufacturers unilaterally on dealers. The

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

50

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

2exclusionary practice, number one. The agreement

3requirements really get in the way of appropriate

4analysis of tying and exclusive dealing in most

5situations. And finally, the market power requirement

6should be equivalent to those that we assess in

7monopolization cases.

8So, I laud the increased scrutiny of tying and

9exclusive dealing under Section 2 of the Sherman Act.

10Microsoft included both, but the Government won on its

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

12exclusive dealing case that the Government won a year or

13two ago, was a Section 2 case.

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

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

16about bundled discounts a lot? Are they predatory

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

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

19asking two questions in two different stages.

20The first question you ask is, are two goods

21subject to a bundled discount bundled together? Well,

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

23efficient firm that offered only one of them could not

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

25as several papers have shown, you basically attribute

51

1the entire discount to the product upon which exclusion

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

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

4a relevant measure of cost, whatever cost measure you

5would use in a predatory pricing case, okay?

6That gets you to bundling; that is, that

7predatory pricing test gets you an answer to the

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

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

10efficient firm that offered only one of the products can

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

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

13inquiry. Tying is explicit bundling of products

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

15once we have decided that two products are bundled

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

17about whether there is foreclosure, whether the

18foreclosure is justified under the circumstances by cost

19reductions, improvements in consumer satisfaction,

20quality control, in many instances price discrimination,

21and so on.

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

23suggestion is that the Government simply get out of the

24business of enforcing the law against simple refusals to

25deal. Now, conditional refusals are something else.

52

1Conditions usually mean exclusive dealing or tying.

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

3conditional refusals to deal, but if we are talking

4about simple refusals to deal in the Trinko or Aspen

5sense, I think the administrative problems are so

6horrific, the disincentives created to competitive

7behavior are so substantial, that the best thing that

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

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

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

11Okay, then let me turn finally and very briefly

12to the subject of remedies. Both General Barnett and

13Debbie Majoras spoke at some length about the importance

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

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

16every Section 2 action that the Government brings ought

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

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

19now we have discovered that whatever exit strategy we

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

21same thing applies to Section 2.

22Section 2 has no moral content. The only

23purpose in bringing these cases is to make the economy

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

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

53

1really have to wonder whether it is worth bringing the

2action to begin with.

3For a long, long period of our history,

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

5preferred remedies were structural or mandatory breakup

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

7remedies have fallen into some disrepute. Many of them

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

9remedy in the United Machinery case, which may have

10ruined the firm, although there was some good evidence

11that USM's technology in its Beverly, Massachusetts

12plant was pretty obsolete already to begin with, or

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

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

15market into a whole bunch of little monopolies.

16Today, we operate in a regime in which

17structural remedies in Section 2 cases appear to be

18disfavored; conduct remedies are preferred.

19Unfortunately, I think the record that we are developing

20with respect to conduct remedies is not much better than

21the record we developed with respect to structural

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

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

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

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

54

1did we accomplish here in terms of making this market

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

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

4I would like to see some more serious attention

5be paid once again to structural remedies or at least

6modified structural remedies, things like compulsory

7licensing, as mechanisms for restoring competition.

8Compulsory licensing is a dirty word in the United

9States when we are talking about general forcing of

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

11are talking about here. We are talking about proven

12antitrust violators who are frequently forced to give up

13their plants and other kinds of hardware. Compulsory

14licensing under that set of circumstances is a perfectly

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

16more seriously than we have in the past.

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

18will turn the floor over to Professor Carlton. Thank

19you.

20(Applause.)

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

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

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

24such distinguished panelists.

25Exclusionary conduct is an important policy

55

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

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

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

4adjudicate them in the context of Section 2 claims,

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

6handle cartl claims. The real problem is that

7competition harms rivals just like exclusionary

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

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

10what should be the optimal policy for exclusionary

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

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

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

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

15change their business behavior in contemplation of legal

16rules. That means that although it is definitely

17informative for economists and lawyers to study the

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

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

20that is not really a study of antitrust policy.

21To appropriately do a study of antitrust policy,

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

23were adjudicated differently, that is, there was a

24different policy, or perhaps you have to look at

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

56

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

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

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

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

5has on the behavior of nonlitigants.

6Since this is an introductory panel, one of the

7advantages you get is you can yell out these hard

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

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

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

11just saying here that this is actually a quite important

12problem and definitely deserves some research time.

13What I will talk about today and on which I

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

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

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

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

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

19Now, it is important to distinguish a tool from

20an objective. The objective is to maximize consumer

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

22consumer plus producer surplus or just consumer surplus,

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

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

25The tools you choose are used to identify acts

57

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

2entails errors. You are going to have false negatives

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

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

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

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

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

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

9the objectives sometimes, so there definitely will be

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

11should be used.

12From an economic point of view, the question of

13whether a particular act harmed consumers is a very

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

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

16could go about trying to answer it with economic tools

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

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

19if there is both an efficiency effect and an

20exclusionary effect on rivals, but sometimes it will not

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

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

23FTC, with their staffs of knowledgeable economists and

24attorneys, should be focusing on answering the direct

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

58

1of possible tools, though obviously they need,

2especially if they are going to think about litigation,

3they need to worry how judges and juries will use

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

5litigated.

6Now, matters change as the inquiry shifts away

7from the government agencies to courts with judges and

8juries who may have less economic sophistication, and

9their simple but imperfect rules may be better than an

10unstructured inquiry, but simple rules should not be

11viewed as anything but simple rules, crude guides that

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

13dangerous to use a simple rule that could ultimately

14subvert the goal of maximizing welfare.

15Now, let me turn to first the profit sacrifice

16test and then to market definition. The profit

17sacrifice test, or a close relative, the

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

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

20fine in the hands of some of its accomplished

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

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

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

24exclusionary act, we are going to reach agreement nine

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

59

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

2be misused.

3What I worry about is when there is someone

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

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

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

7raises all the danger signs associated with possibly

8confusing competition with exclusionary conduct. Let me

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

10reasons.

11First, all strategic behavior -- and every

12business school teaches this -- all strategic behavior

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

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

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

16of a competitive battle. Investments in advertising,

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

18mistakenly and easily be condemned by jurors convinced

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

20accommodating, less exclusionary. That "but for"

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

22the exclusionary effect -- that hypothetical thought

23experiment is actually quite a difficult one to

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

25to consumer welfare that you can necessarily get around

60

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

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

3conduct, what hypothetical world will you be proposing,

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

5rate associated with application of that rule.

6The second reason I am worried about it is

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

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

9Coase, one of my colleagues at the University of

10Chicago, always fond of explaining how little economists

11know about business and that businessmen really know a

12lot about business, and just because an economist cannot

13understand an action or a jury cannot understand an

14action, he said why should that create antitrust

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

16serious concern.

17The profit sacrifice test may sometimes work and

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

19dangerous to enshrine it as a general proposition. It

20strikes me as much more appropriate to devise tests and

21screens that fit particular exclusionary acts,

22especially because under Section 2 we have a range of

23such widely different behavior that people have attacked

24as exclusionary, so many different acts fall under

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

61

1how the legal treatment of the act is creating special

2risks for chilling competitive behavior, and that is

3going to differ from act to act to act.

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

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

6that is undergoing rapid technological change, and there

7is an exclusionary claim that the way the product was

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

9instance whether you are depriving consumers of a new

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

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

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

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

14weight loss than in other situations.

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

16about the Areeda-Turner test for predatory pricing.

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

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

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

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

21deviation from what we usually think of as profit

22maximization that there is something fishy, and the

23reason they chose price below a cost as the predation

24standard rather than above-cost pricing is because they

25were very worried about chilling competition that drives

62

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

2context of predatory pricing of erroneous behavior, of

3erroneous condemnation, which then would chill

4competition, very specialized to the predatory act,

5okay? Not a general principle.

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

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

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

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

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

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

12that is saying that any deviation from profit-maximizing

13behavior would be an antitrust violation, and that would

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

15profit sacrifice test, because if you are not maximizing

16profit, your failure to maximize profit is a sacrifice

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

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

19see it or it is an antitrust violation.

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

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

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

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

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

25it as a general principle.

63

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

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

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

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

5least it confirmed in my view that someone else thinks

6it deserves still more thought.

7But for the antitrust laws, industrial

8organization economists would not pay all that much

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

10fair to say that the reason it receives so much

11attention is precisely because it has been used as a

12screen or a requirement under our antitrust laws, and

13that is what gives it prominence.

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

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

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

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

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

19apply it in the nonmerger context that at least

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

21number of cases is the application of the Merger

22Guidelines, and it goes something like this.

23Define a market so that a hypothetical

24monopolist of those products can raise price 5 percent

25above the competitive level, the competitive price. In

64

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

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

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

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

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

6competitive price is. If I knew what the competitive

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

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

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

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

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

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

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

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

15really directly answer the question, because the

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

17you do?

18I think there are several alternatives. Herb

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

20completely satisfying. One alternative is to ignore the

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

22competitive price, I am going to use the existing

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

24the well-known what is called Cellophane fallacy in

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

65

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

2because you have already raised the price above the

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

4its problems.

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

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

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

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

9replaces one question with the other, what is marginal

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

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

12you could try and sift through accounting information,

13you could econometrically try and estimate a cost curve.

14It's difficult, okay?

15Moreover, suppose you do find prices above

16marginal cost. You have to face the realistic

17possibility that most markets are not perfectly

18competitive. In most markets, price will not equal

19marginal cost; therefore, and what you presumably must

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

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

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

23actually never been articulated that I have seen in a

24quantitative way.

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

66

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

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

3rate of return a firm is earning above the competitive

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

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

6competitive rate? This can be a difficult accounting

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

8economic exercise using accounting data. You would have

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

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

11likely to create a lot of controversy.

12Finally, you could estimate the demand curve

13facing the firm, and the benefit of estimating the

14demand curve is you could determine the elasticity and

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

16I think gives you useful information about certain types

17of competition. It raises issues over what time period

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

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

20question you are trying to answer is really a very

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

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

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

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

25the panel discussion. What is the distinction between

67

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

2does it mean to have individual market power in an

3oligopoly?

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

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

6ultimate goal of maximizing consumer welfare. You

7should create rules to fit the act, paying special

8attention to how the rules applied to this particular

9act will chill competition.

10The profit sacrifice test worries me a bit for

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

12sometimes but certainly not regard it as a general

13principle. With regard to market definition, I still

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

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

16it helps structure the analysis, but there is an

17inherent lack of precision in its application in Section

182 cases that might be worrisome, and there may well be

19cases where its use could be misleading.

20Thank you.

21(Applause.)

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

23for very, very interesting perspectives. We have about

2420 minutes for panel discussion, and because of airline

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

68

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

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

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

4the thoughts from either of the two professors.

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

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

7were very thoughtful presentations, as we expected, and

8I very much appreciate that.

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

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

11looking at restraints on innovation, and I was

12wondering, Herb, if you had any particular hypothetical

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

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

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

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

17about promoting innovation and how important that is to

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

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

20and explaining to courts, it is very difficult for

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

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

23at it.

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

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

69

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

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

3about this tension between innovation and competition.

4It is hard to defend restraints on innovation when they

5are defined properly, and the fact is that the

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

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

8can think of three situations.

9One is the air pollution cartel cases of the

101970s. I suspect some of you have not been around long

11enough to remember those, but those were challenges by

12farmers, class action, brought by agricultural groups to

13an alleged agreement among the major automobile

14producers for restraining the development of engines

15that produce less pollution, and what they wanted to

16collect as damages was harm to their crops that accrued

17as a result.

18Well, just stating it tells you why that

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

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

21determine what the impact of these more fuel-efficient

22engines would have been, how the farmers might have

23evaded it. It would be totally impossible for private

24plaintiffs to collect damages in such a case.

25Another one is the series of cases, a fairly

70

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

2against the cigarette companies in the late 1990s

3alleging various cartels to refrain from developing

4healthier cigarettes. To the best of my knowledge,

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

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

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

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

9ability to prove injury.

10Then, of course, there is the conclusion in

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

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

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

14spoken multiple-processor languages, something that Bill

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

16complaint allegation has actually been included in a few

17of the indirect purchaser cases against Microsoft, and

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

19ever collected a dollar for those failures of

20innovation.

21So, one of the problems is that restraints on

22innovation are very likely extraordinarily harmful if

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

24from innovation, then restraints on innovation could, in

25fact, be very harmful.

71

1Number two, because of the nature of innovation,

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

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

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

5damages actions under Section 4 of the Clayton Act.

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

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

8attention paid.

9CHAIRMAN MAJORAS: Well, it does help, because

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

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

12innovation incentives generally. I mean, this is the

13classic case that I was just discussing earlier with

14some folks in the consumer protection arena about

15pharmaceutical companies and their incentives to

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

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

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

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

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

21pharmaceutical prices?" But because the harm to

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

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

24account.

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

72

1reactions. First, I am extraordinarily pleased with the

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

3frame some interesting issues and pose some hard

4questions, but in some instances, also provided

5suggested answers or even policy directions, which is a

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

7appreciate hearing it.

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

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

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

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

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

13more or less aggressive in intervening in the sense that

14is there a greater risk of deterring innovation if you

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

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

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

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

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

20quoted Capitalism, Socialism and Democracy, the Joseph

21Schumpeter discussion of monopolies, and I have been

22taken with this gales of creative destruction imagery.

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

24and/or justifies monopolies as good in and of

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

73

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

2with everything that is in that book in terms of

3monopoly, but I do agree with the importance of

4innovation and not losing sight of that.

5But talking about this, you just made a comment

6about remedies and a distinction, I think, between

7damages and the Government. I would be interested to

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

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

10DR. HOVENKAMP: The distinction between?

11MR. BARNETT: Different remedies for

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

13injunctive versus damages type relief.

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

15which the Supreme Court has rejected in California

16versus American Stores, that structural remedies should

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

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

19California versus American Stores, where the plaintiff

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

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

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

23this rather unusual way we treat states attorneys

24general in antitrust, but I believe structural remedies

25ought to be something that only the Government ought to

74

1do, because they have such extraordinary spillovers.

2I mean, we have some confidence that the federal

3antitrust agencies, because of their diversity, the

4diversity of the industries that they represent, and

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

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

7may not have that level of confidence about

8industry-specific agencies.

9The one thing we know about private plaintiffs

10is that they are always captured, right? Private

11lawyers serve their clients. Their clients are

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

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

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

15should simply be denied to private plaintiffs.

16Is that an answer to your question?

17MR. BARNETT: I just was interested in your

18perspectives, so thanks.

19MR. BLUMENTHAL: Herb, Dennis had spoken after

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

21rejoinder to any of Dennis' comments.

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

23million times more about this market definition stuff

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

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

75

1develop a few presumptive rules for dealing with market

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

3firm under investigation, faces competition from other

4firms that use the same technology and apparently have

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

6competition, competitive pricing, is higher than in a

7case like Cellophane. I think what makes the Cellophane

8case so extraordinary is that the defendants, DuPont,

9plus Sylvania as its licensee, produced Cellophane, but

10the stuff that the Government ended up throwing into the

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

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

13were things that used different raw materials, they

14almost certainly used different technologies for

15producing them, and simply to conclude that such things

16are in the same market simply because of high observed

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

18It's an error courts continue to make.

19There are several cases, for example, that have

20concluded that rental videos, films shown in movie

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

22relevant market simply because they make the obvious

23observation that if you look at any particular person,

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

25to a theater, sometimes she watches a movie on

76

1television, without asking the question whether any one

2of those technologies is sufficient to hold the other

3ones to cost.

4I would rather hear Dennis talk about this.

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

6something about innovation, if that is all right,

7because I think that is an important topic.

8It is absolutely right, our standard of living

9has increased because of technological change. The

10question about what the implication of that is for

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

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

13is because what we know about innovation and the market

14structure that induces innovation is a lot less than

15what we know about concentration and the effect of

16concentration on raising price.

17The evidence, I think most economists would say,

18that competition is good for innovation. On the other

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

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

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

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

23example, an innovation market I think is not

24particularly helpful except maybe in rare exceptions

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

77

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

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

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

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

5It is much harder when you are looking at just

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

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

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

9at major product introductions, who did them, sometimes

10the people doing them are often outside the industry,

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

12the difficulty in this area, when antitrust tries to

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

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

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

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

17innovation markets inherently involve intellectual

18property. People want to protect their intellectual

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

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

21Internet, who has exclusivity rights on certain types of

22information that is generated. Restrictions, vertical

23restrictions on intellectual property, who can use them,

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

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

78

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

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

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

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

5get the returns from this intellectual property, and

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

7restrictions on it."

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

9think it is made more difficult by our intellectual

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

11is that because so many patents are given for products

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

13problem that the property right in a truly valuable

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

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

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

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

18antitrust but through reforming IP.

19CHAIRMAN MAJORAS: As long as we are on the

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

21does either of you have an opinion on the recent

22phenomenon that we are seeing in the patent area known

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

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

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

79

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

2then wait until somebody creates an invention, and that

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

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

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

6royalties."

7And there are obviously many different

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

9interest in ever making anything or innovating I think

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

11you have any comment on whether that is injuring

12competition or something that any of us should be

13worried about?

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

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

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

17antitrust problem, it either has to monopolize the

18market or create a dangerous probability of doing so or

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

20simply surprising people with a patent and claiming high

21royalties in and of itself does not do that.

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

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

24to an antitrust violation.

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

80

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

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

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

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

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

6the courts to give the surprised infringer time to

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

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

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

10intending to use this themselves and implement it, and

11that may influence how you treat them.

12Having said that, you know, an economist always

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

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

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

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

17mean, there is a flip side.

18CHAIRMAN MAJORAS: Yes.

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

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

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

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

23Section 2 problems.

24DR. CARLTON: I would be delighted.

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

81

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

2thrown out a teaser at the end about the distinction

3between market power and monopoly power, an issue that

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

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

6recalled George Stigler having denied as a distinction

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

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

9there is a distinction after all.

10Did you have in mind something more than just

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

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

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

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

15marginal cost. That is a logical definition. Whether

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

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

18how much before you say it is something that triggers

19some action, but you can distinguish market power from

20monopoly power by saying monopoly power is not just

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

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

23at least a logical distinction.

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

25dropped that paragraph, because no one seemed to pay

82

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

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

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

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

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

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

7is an answer.

8Now, whether that should be what courts use when

9they are deciding on an exclusionary act is a slightly

10different question, because I think courts really should

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

12us suppose you are talking about exclusive dealing. I

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

14don't think. I think what a court is saying is if you

15are engaged in an act and there are other

16characteristics of the market that make that act have a

17greater effect than if you didn't, say, have as large a

18market share, then I am going to be worried about that,

19and that really does not have to do with rates of

20return.

21So, you know, I can see why courts could be

22asking are the characteristics of the market that make

23the exclusionary act under discussion much more a

24competitive concern than not, whether that -- I don't

25know that that necessarily corresponds to the

83

1distinction I gave you. The distinction I gave you does

2strike me as a sensible economic distinction.

3MR. BLUMENTHAL: Well, thank you, and with that,

4we are out of time. Let me ask you to join me in

5thanking the panel for a great presentation.

6(Applause.)

7MR. BLUMENTHAL: We look forward to receiving

8your comments in the months ahead. We hope to see you

9Thursday and in the hearings to come. Thank you all.

10(Whereupon, at 4:00 p.m., the hearing was

11concluded.)

12

13

14

15

16

17

18

19

20

21

22

23

24

25

84

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

2DOCKET/FILE NUMBER: P062106

3CASE TITLE: SECTION 2 HEARING

4DATE: JUNE 20, 2006

5

6I HEREBY CERTIFY that the transcript contained

7herein is a full and accurate transcript of the notes

8taken by me at the hearing on the above cause before the

9FEDERAL TRADE COMMISSION to the best of my knowledge and

10belief.

11

12DATED: 7/5/06

13

14

15

16SUSANNE BERGLING, RMR-CLR

17

18C 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

20I HEREBY CERTIFY that I proofread the transcript

21for accuracy in spelling, hyphenation, punctuation and

22format.

23

24

25DIANE QUADE