Participant Presentations and Session Transcript for June 20

1

1	UNITED STATES FEDERAL TRADE COMMISSION
2	and
3	UNITED STATES DEPARTMENT OF JUSTICE
4
5
6
7	SHERMAN ACT SECTION 2 JOINT HEARING
8	WELCOME AND OVERVIEW OF HEARINGS
9	TUESDAY, JUNE 20, 2006
10
11
12
13
14	HELD AT:
15	600 PENNSYLVANIA AVENUE, N.W.
16	WASHINGTON, D.C.
17	UNITED STATES FEDERAL TRADE COMMISSION
18	HEADQUARTERS BUILDING, ROOM 432
19	2:00 P.M. to 4:00 P.M.
20
21
22
23
24	Reported and transcribed by:
25	Susanne Bergling, RMR-CLR

2

1	MODERATOR:
2	WILLIAM BLUMENTHAL
3	Federal Trade Commission
4
5	PANELISTS:
6
7	Deborah Platt Majoras
8	Thomas O. Barnett
9	Dennis Carlton
10	Herbert Hovenkamp
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25

3

1	C O N T E N T S
2
3	Introduction
4
5	Presentations:
6	Deborah Platt Majoras
7	Thomas O. Barnett
8	Herbert Hovenkamp
9	Dennis Carlton
10
11	Moderated Discussion
12
13	Adjournment
14
15
16
17
18
19
20
21
22
23
24
25

4

1	P R O C E E D I N G S
2	- - - - -
3	MR. BLUMENTHAL: Ladies and gentlemen, good
4	afternoon. I'm Bill Blumenthal from the FTC, and I'd
5	like to welcome you to the first of the joint Justice
6	Department Antitrust Division and Federal Trade
7	Commission hearings into Section 2 of the Sherman Act.
8	The purpose of these hearings is to explore how
9	best to identify anticompetitive exclusionary conduct
10	for purposes of antitrust enforcement. We are
11	envisioning a series of hearings that will kick off
12	today and will continue through December, probably two,
13	three, four hearings a month, with the exception of
14	August. After today's kick-off hearing, we are going to
15	have another hearing on Thursday of this week examining
16	predatory pricing, we will have a hearing in mid-July
17	examining refusals to deal, take a little bit of a
18	breather, and then resume in September with what would
19	then be a series of examinations.
20	The agencies are expecting to focus on legal
21	doctrine, on jurisprudence, economic research, and
22	business and consumer experience. We have a Federal
23	Register notice that is outstanding. We invite public
24	comment on a wide range of topics, and we hope that
25	those of you who are here, as well as many others, will

5

1	have an opportunity to submit comments on the topics
2	that we address. We are open to receiving those any
3	time through the final hearing, that is, through
4	December, although the earlier the better for our
5	purposes.
6	We are honored today to have a special panel to
7	kick off the hearings. They probably do not need much
8	introduction, so I am going to be very brief in offering
9	the introductions. In the order in which they will be
10	speaking, first we have Deborah Platt Majoras, Chairman
11	of the Federal Trade Commission. Thomas Barnett, the
12	Assistant Attorney General for Antitrust in the Justice
13	Department. Herb Hovenkamp, who is probably known to
14	most of you as -- as many things -- a professor of law
15	at the University of Iowa, but probably better known as
16	a co-author and a reviser of the leading treatise in the
17	antitrust field as well as a prolific author of many,
18	many other volumes, the most recent of which is recently
19	out, The Antitrust Enterprise: Principle and Execution,
20	available through Harvard University Press, with an
21	imprint of this year.
22	And finally Dennis Carlton, also known to many
23	of you in many capacities, most notably professor of
24	economics at the Graduate School of Business at the
25	University of Chicago, former president and still very

6

1	active in Lexicon, frequent expert witness, author of
2	many, many articles, author of I guess two of the
3	leading economics texts in the field. I'll leave it at
4	that. You all know Dennis Carlton.
5	I have several preliminary announcements that I
6	am going to make, one of which is legally mandated. The
7	first, which is not legally mandated but is a common
8	courtesy, if any of you have cell phones or
9	Blackberries, pagers, iPods, things of that nature,
10	please do as I'm doing right now and set it into silent
11	or manner mode, although if you do it like I just did,
12	you just dialed 7. Okay, there we go.
13	Second, I have been asked to let you know that
14	the men's room is immediately out these doors and to the
15	left. The ladies' room is out these doors, to the other
16	side of the elevator banks, and to the left.
17	And finally, the legally mandated announcement
18	is the one that says, as a safety tip for our visitors,
19	if the building alarms go off, please proceed calmly and
20	quickly as instructed. If we must leave the building,
21	take the stairway, which is to the right on the
22	Pennsylvania Avenue side, and after leaving the
23	building, please follow the stream of FTC people who are
24	practiced in this evolution. We will all go to the
25	Sculpture Garden, catty-cornered across the street at

7

1	7th and Constitution, and we will assemble there, where
2	noses will be counted.
3	With that, it gives me great pleasure to turn
4	the podium over to the Chairman of the Federal Trade
5	Commission, Debbie Majoras.
6	(Applause.)
7	CHAIRMAN MAJORAS: Good afternoon, everyone, and
8	thank you very much, Bill. Together with my good friend
9	and colleague, Assistant Attorney General Tom Barnett,
10	it is my great pleasure to welcome you to these hearings
11	in which we will be exploring conduct under Section 2,
12	and we are privileged to have two of our most
13	distinguished antitrust scholars here, Professors
14	Hovenkamp and Carlton.
15	Now, at the start of any new endeavor, it is
16	important to reflect on why we are undertaking it.
17	Beginning in 1990, the McKinsey Global Institute, led by
18	founding director William W. Louis, undertook a 12-year
19	study of the economic performance of 13 nations seeking
20	to understand globalization, and more fundamentally, the
21	disparities between rich and poor. The study showed
22	that levels of productivity made the difference between
23	rich and poor nations. What, though, made the
24	difference in the levels of productivity? The answer
25	they found was "undistorted competition in product

8

1	markets."
2	In his book in which he reports the results of
3	the study, Mr. Lewis says, "Most economic analysis ends
4	up attributing most of the differences in economic
5	performance to differences in labor and capital markets.
6	This conclusion is incorrect. Differences in
7	competition in product markets are much more important."
8	McKinsey also asked why the highly productive
9	United States has higher competitive intensity than
10	other nations. Mr. Lewis sums up the answer by saying
11	that, in the United States, "Consumer is king." More
12	specifically, he says, "[t]he United States adopted the
13	view that the purpose of an economy was to serve
14	consumers much earlier than any other society," and we
15	continue to "hold this view more strongly than almost
16	any other place." And he concludes that, in fact,
17	"Consumers are the only political force that can stand
18	up to producer interest, big government, and the
19	technocratic, political, business, and intellectual."
20	This is why we are here. The FTC and the
21	Antitrust Division have the responsibility to ensure
22	that competition in U.S. markets is free of distortion
23	and that consumers are protected not from markets but
24	through markets unburdened by anticompetitive conduct
25	and government-imposed restrictions. This work is

9

1	critical, indeed, to the well-being of the American
2	people. Over the past few decades, the United States
3	has substantially deregulated critical industries,
4	including transportation, telecommunication and energy,
5	to the substantial benefit of the economy and consumers.
6	As government regulators have given way to free markets,
7	much of the responsibility for protecting consumers
8	shifts to competition agencies and courts. While
9	competition is distorted when governments regulate or
10	intervene excessively, it also is true that private
11	actors can and do distort competition.
12	Breaking up cartels, preventing mergers that
13	will substantially reduce competition, and halting
14	conduct that goes beyond aggressive competition to
15	distorting it is vital to promoting vigorous competition
16	and maximizing consumer welfare, and we have developed a
17	great deal of consensus regarding appropriate antitrust
18	policy, I think, as it relates to cartels and to mergers
19	and other horizontal conduct, as a result of which our
20	enforcement has become more transparent and predictable,
21	which then, in turn, makes it easier for market
22	participants to make decisions about their own conduct.
23	Unilateral or "single-firm" conduct, however,
24	still vexes us. Even though we can find some
25	respectable measure of consensus around principles that

10

1	should apply, we find a range of opinions from
2	knowledgeable people about how to apply those principles
3	to enforcement in the market, and the question of the
4	proper test that our agencies should apply and that
5	courts should apply to conduct of the single firm with
6	market power now has dominated our antitrust debate for
7	several years.
8	We are not alone. Across the globe, over the
9	past quarter century, economic systems in which the
10	state owns the firms and central planners set out prices
11	and levels of output have given way to competition where
12	the forces of supply and demand determine prices and
13	allocate the resources, and we have worked hard to
14	promote the economic and political benefits of markets.
15	With attempts to introduce market economies have come
16	new competition authorities, today numbering around 100,
17	when only 15 years ago, we had just 20. And even
18	countries that for decades have had nearly total state
19	control over their economies, like China, are now
20	dedicating substantial resources to drafting competition
21	laws.
22	Currently, the issue of how to evaluate
23	unilateral conduct is the most heavily discussed and
24	debated area of competition policy in the international
25	arena. Just to give you a few examples, last week, FTC

11

1	and DOJ officials attended the EC's hearing to review
2	their policy under Article 82, which addresses conduct
3	by dominant firms. Officials from both agencies
4	recently held talks with our colleagues in Japan and
5	Mexico and Canada on the issue. We recently had panels
6	on it in the OECD. And since the International
7	Competition Network established a working group on
8	unilateral conduct in May, the FTC, which will co-chair
9	that group, has received expressions of interest from
10	more countries wanting to be involved than we have ever
11	had in any other working group in the ICN.
12	So, why the strong interest? Well, first, many
13	nations are facing the challenge of converting from
14	state-owned or supported monopolists to markets with
15	more than one participant, which is no small challenge,
16	as we ourselves have learned in trying to deregulate
17	certain markets like electricity. And, indeed, to
18	enforcers in those nations, it then becomes companies
19	with market power, not horizontal competitors, that are
20	the evil that must be attacked. Second, disagreement
21	among competition authorities about how to treat
22	unilateral conduct produces uncertainty in national and
23	international markets, which reduces the market
24	efficiency and imposes costs. And third, the analysis
25	of unilateral conduct in the identification of that

12

1	which is anticompetitive presents unique challenges that
2	are not present or at least are less present in the core
3	antitrust concern of conduct between competitors, and by
4	now, these unique challenges I think are familiar.
5	First and fundamentally -- and we discuss it all
6	the time, but that doesn't make it less difficult -- and
7	that is it is difficult to distinguish between
8	aggressive procompetitive unilateral conduct and
9	anticompetitive unilateral conduct. As the D.C. Circuit
10	said in the Microsoft case, "The challenge for an
11	antitrust court lies in stating a general rule for
12	distinguishing between exclusionary acts which reduce
13	social welfare and competitive acts which increase it,"
14	and this is tough, because as Judge Diane Wood wrote for
15	the Seventh Circuit, "distinguishing between legitimate
16	and unlawful unilateral conduct requires subtle economic
17	judgments about particular business practices." So,
18	while it's difficult, it must be done and it must be
19	done well.
20	Second, the process of distinguishing between
21	permissible and impermissible conduct must be relatively
22	consistent and transparent so that firms are able to
23	incorporate it into their decision-making. While there
24	are relatively few findings of Section 2 liability,
25	there nonetheless are a large number of different types

13

1	of conduct that may raise competition concerns and would
2	fall under Section 2.
3	And third, while antitrust practitioners have
4	had substantial success devising remedies for joint
5	conduct, devising remedies for single-firm behavior
6	presents significant difficulties. As Professors Areeda
7	and Hovenkamp put it, "By contrast with concerted
8	conduct, unilateral behavior is difficult to evaluate or
9	remedy by any means short of governmental management of
10	the enterprise."
11	We have much to work with as we move forward
12	with these hearings. Already a number of experienced
13	experts have proposed the adoption of a single test for
14	evaluating nearly all types of potentially exclusionary
15	conduct. Some have argued for a test that focuses on
16	the impact of the conduct on consumer welfare. Others
17	support analyzing whether the conduct involves the
18	short-term sacrifice of profits. Others support a
19	no-economic-sense test, which asks whether the cost of
20	engaging in the exclusionary conduct makes sense only
21	because it serves to eliminate competition.
22	Judge Posner has written that the inquiry should
23	focus on whether the conduct excludes other equally
24	efficient rivals, and still other practitioners and
25	scholars oppose the adoption of any single unilateral

14

1	conduct test and instead favor consideration of
2	different tests for particular types of exclusionary
3	conduct. And then, of course, when you go out into the
4	world, you see that there are many other opinions on the
5	type of test or framework that should be used.
6	So, proponents of the various tests and
7	approaches already have done a very good job of laying
8	out, I think, the relative merits, and virtually all
9	have acknowledged that their preferred approach is
10	probably not perfect. At these hearings, I hope we can
11	tackle this issue by starting with the conduct itself.
12	The hearings will have panels that will focus on
13	specific types of conduct that at least to date we know
14	can implicate liability. We want the panels to discuss
15	the conduct from the market perspective, from the ground
16	up; that is, to examine why and when firms engage in
17	certain practices, how they do it, what effects it
18	produces for the firm, for other firms, both customers
19	and the competition, and for consumers. And we should
20	look at whether firms in competitive markets also engage
21	in the same conduct, and if so, examine why that is. We
22	want these discussions, to the extent possible, to
23	include knowledgeable businesspeople or at least their
24	advisers, and from these discussions, we then should
25	endeavor to develop sign posts for when the conduct may

15

1	harm competition and when it typically does not. From
2	the sign posts, we hopefully can draw some guiding
3	principles, and only then should we turn to examining
4	the current state of the law as it has been applied to
5	such conduct and then determining what workable rules
6	can be applied to the specific conduct at issue. That
7	way, we can then see, can we pull these together into a
8	single test or a broader set of rules? And even if we
9	don't produce a consensus on the universal test or
10	tests, I'm optimistic that we can identify relative
11	consensus on a number of principles and then on how to
12	approach at least some fraction, hopefully a significant
13	fraction, of single-firm conduct we encounter.
14	In these discussions, we need to be careful not
15	to permit labels or semantic differences to get in the
16	way. In some discussions I've heard on these issues, I
17	have been worried that people are actually talking past
18	one another. In addition, this debate must not become
19	so academic that even if it could be resolved, it might
20	not have much practical application in the marketplace.
21	Indeed, last week I was speaking with a long-time
22	antitrust practitioner about these hearings and about
23	the debate over a proper test, and he said, that while
24	he thought the Section 2 issues were very important,
25	nonetheless, the search for the "holy grail" test might

16

1	just be something in which only about 27 people have an
2	interest. So, we really want to be careful about that.
3	I do think we start with some substantial
4	consensus about core underlying principles and factors
5	that should underlie any evaluation of unilateral
6	conduct.
7	First, the only type of unilateral conduct that
8	should implicate the antitrust laws is conduct that
9	produces durable harm to competition, leading to higher
10	prices, reduced output, lower quality or lower rates of
11	innovation. As much as we may value the success of
12	particular companies, the health of the companies
13	themselves is not the concern of antitrust law.
14	Second, there is consensus that antitrust
15	standards that govern unilateral conduct must not in
16	themselves deter competition, efficiency, or innovation,
17	and this is what we mean when we constantly say that we
18	worry about false positives. Obviously pervasive and
19	aggressive competition in which firms consistently try
20	to better each other by providing higher quality goods
21	and services at lower cost is crucial to maximizing
22	consumer welfare. So, the antitrust laws should then
23	never condemn market power that is obtained through the
24	development of superior products and services,
25	regardless of how many competitors are driven from the

17

1	marketplace in the process, and that, of course, has
2	been accepted by the courts.
3	Third, there is consensus that the standards for
4	evaluating unilateral conduct must be clear and
5	practical to administer or as practical as they can be
6	to administer. The most analytically sound principles
7	will provide little value to us if firms can't interpret
8	them when they are making their business decisions.
9	And, of course, courts have to be able to interpret and
10	apply rules as well.
11	And while I want to emphasize that I am going to
12	use the hearings to continue developing my own thinking
13	on these issues, I do approach them, in addition to the
14	broad principles, with a number of other hypotheses.
15	First, any legal framework needs to avoid
16	second-guessing business judgments that were objectively
17	reasonable at the time they were made. An ex post facto
18	examination of the hypothetical effects of alternative
19	courses of conduct is likely to chill legitimate
20	business behavior. Second, to be practical, any legal
21	framework must be able to evaluate conduct that both
22	generates efficiencies and produces anticompetitive
23	exclusion. If we only had to worry about conduct for
24	which the effects are obvious, we probably would not be
25	here today. And third, any test or tests must account

18

1	for the fact that certain types of unilateral conduct
2	are significantly more likely to cause competitive harm
3	than others. For example, most would agree that
4	unilateral above-cost pricing at monopoly levels should
5	not be condemned under the antitrust laws. Similarly,
6	behavior that some commentators have termed "cheap
7	exclusion," such as the use of government processes to
8	unlawfully extend the life of a patent, is generally
9	viewed as unlawful exclusionary conduct. And this may
10	mean that there is no unitary test or that we simply
11	need a broad framework that can accommodate a spectrum
12	or perhaps a sliding scale for the levels of harm, and
13	proposals have been made for how we might think about
14	the distinctions that could be made, including Deputy
15	Bureau of Competition Director Ken Glazer's proposal
16	that we analyze conduct by distinguishing between
17	conduct that's coercive versus incentivising.
18	Now, in the Microsoft case, the D.C. Circuit
19	applied what I view as a sensible weighted balance
20	approach to Microsoft's conduct that's largely
21	consistent with the three principles I just discussed.
22	Some have criticized the framework used in Microsoft as
23	insufficiently structured or unfocused, and I understand
24	where that comes from, but I think if we look at how it
25	was actually applied, it may be a workable framework

19

1	that incorporates the principles on which we have wide
2	consensus. I mean, perhaps the same criticism about
3	being unstructured could be applied to the Section 1
4	rule of reason and, in fact, probably is at times, but
5	as applied to, for example, joint ventures, the
6	balancing has been weighted I think in the right
7	direction.
8	First, the Microsoft court did not attempt to
9	substitute ex post facto its judgment for the business
10	judgments that were made ex ante, or to determine what
11	actions might have been better overall for consumers.
12	For example, the Court did not base its findings on an
13	ex post analysis of the impact of Microsoft's conduct on
14	the prices charged to consumers.
15	The Microsoft court also demonstrated that to
16	evaluate whether certain types of unilateral conduct
17	violate the antitrust laws does require an examination
18	of both likely anticompetitive and procompetitive
19	effects. For example, the Court analyzed the legality
20	of a Microsoft license provision that prohibited OEMs
21	from modifying the initial boot sequence. Microsoft did
22	not dispute that that restriction limited competition
23	against IE. The Court nonetheless held that the
24	restriction was not a violation because it concluded
25	that preventing the Windows desktop from ever being seen

20

1	at all in the boot sequence was a substantial alteration
2	of Microsoft's copyrighted work that could produce harm
3	that outweighs the marginal anticompetitive effect of
4	the prohibition. The Court performed this same analysis
5	across two dozen types of conduct, examining both the
6	anticompetitive effects and procompetitive
7	justifications, taking care, though, to ensure that it
8	not chill procompetitive behavior.
9	And finally, the D.C. Circuit made clear that it
10	did not consider all types of unilateral conduct to
11	raise equal concerns under the antitrust laws. For
12	example, the Court stated that courts need to be very
13	skeptical about claims that a dominant firm's design
14	changes harm competition and, by implication, violate
15	the antitrust laws.
16	One final note about the hearings. I hope that
17	our latest panels, which we will hold on remedies, will
18	produce a productive discussion. It simply is not
19	possible to implement sound competition policy for
20	single-firm conduct without giving careful thought to
21	remedies. Despite their importance, though, I think the
22	issues relating to remedies have not received extensive
23	attention. Take the Microsoft case, for example, which
24	although it received and still receives a bit of
25	notoriety, I have been stricken by how few productive

21

1	discussions of the remedy and the D.C. Circuit decision
2	that accepted the DOJ remedy while rejecting other
3	remedies have actually occurred, and while that might
4	have stemmed from some of the market dissatisfaction
5	over that remedy, I think these hearings should give the
6	Section 2 remedy issue the prominence that it deserves
7	in our analysis. After all, if you have done these
8	cases, you know that devising and drafting remedial
9	provisions in monopolization cases can be more difficult
10	than determining whether a violation has even occurred.
11	At bottom, through these hearings and through
12	our work, we need to remember that antitrust is the
13	means, not the end. Rather, the end is undistorted
14	competition driven by "king" -- and I would say "and
15	queen" -- consumer, and the challenge is to keep
16	competition undistorted while not distorting it
17	ourselves in the process.
18	So, I thank you again for attending the opening
19	of these hearings, and we look forward to all of your
20	contributions. Thank you very much.
21	(Applause.)
22	MR. BLUMENTHAL: Thank you, Chairman.
23	General Barnett?
24	MR. BARNETT: I am going to attempt to be
25	somewhat high-tech here. We will see if it works. Ah.

22

1	I want to thank, first off, the FTC for hosting
2	the first of these hearings on Section 2 and for their
3	help in organizing and planning and discussing the
4	issues that have brought us together. Both Chairman
5	Majoras, as well as the other FTC Commissioners and the
6	FTC staff, have all been extraordinarily helpful, and I
7	want to thank all of you as well as the Antitrust
8	Division staff who have worked very hard to prepare for
9	the session today and who will continue to work hard
10	over the months ahead.
11	I also want to thank Herb and Dennis for
12	agreeing to be with us. I feel very honored to have
13	such distinguished commentators, and feel that it is a
14	terrific start to these important hearings.
15	I am going to start with the same issue that
16	Debbie started with, which is why are we sponsoring
17	these hearings? And I can tell you up front, much of
18	what you are going to hear will echo some of what Debbie
19	said, but it was done unilaterally and without
20	consultation, and so I hope you will judge it at least
21	under the Section 2 standard, not under Section 1.
22	We do feel that it is important. As Debbie
23	said, unilateral conduct can harm consumer welfare. I
24	think there is a consensus that the focus of the
25	antitrust laws in the United States is to preserve and

23

1	promote or prevent harm to consumer welfare and that
2	unilateral conduct is an important element of that.
3	I also agree that this is the area of probably
4	the least consensus. I think there are large areas of
5	consensus within Section 2, but there are significant
6	areas where I think we have room for further
7	understanding.
8	These hearings, with the combination of legal,
9	economic, business and governmental/private
10	perspectives, provide us with a unique opportunity to
11	advance our understanding, and I believe that that will
12	help us to advance the development of the law. It can
13	provide helpful guidance to the courts, guidance to the
14	business community, and as Debbie quite eloquently put
15	it, to the international community that is now focused
16	on this issue.
17	There is a long tradition of the agencies
18	leading the development of competition law. I need only
19	point to Don Turner and the 1968 Merger Guidelines and
20	the formulation by Bill Baxter in 1982 to provide an
21	example of what has become the standard reference for
22	analyzing mergers, not only in U.S. courts, but really
23	around the world in many ways.
24	With respect to the international community,
25	again, I do want to both echo and underscore what Debbie

24

1	said. This is an issue that is at the forefront of
2	people's minds as we talk to officials on every
3	continent, and one example that sort of helped drive
4	this point home a bit, I was at a conference a while ago
5	in Budapest of Southeastern European former Soviet block
6	countries, and we were talking about a topic that the
7	Antitrust Division often talks about, which is the
8	importance of cartel enforcement, and one of the
9	officials approached me at a break and said, "I agree
10	with you, cartels are a terrible thing. I just wish
11	that our markets had enough participants so that they
12	could collude together. They don't have anyone to
13	collude with. So, we are focused on this dominant
14	former state-owned enterprise and how we can introduce
15	competition into this economy." It just drove home for
16	me, at least, the importance of this issue. It is
17	important here, but I think its importance abroad cannot
18	be over-emphasized.
19	The Supreme Court, to its credit, addressed the
20	issue of monopoly 96 years ago. That is when it decided
21	the Standard Oil case, and while we think of it as a
22	rule of reason case, it did talk about the three evils
23	of monopoly. It talked about first the power to fix
24	price and thereby injure the public; second, the power
25	of enabling a limitation on production; and third, the

25

1	danger of deterioration in quality of the monopolized
2	article, which it deemed was the inevitable result of
3	the monopolistic control of its production. Price
4	increases, output reductions, quality deterioration,
5	those are still the same three touchstones that we look
6	to that you heard Debbie talk about that go all the way
7	back to the Supreme Court's discussion of the issue in
8	1910.
9	As we have talked about it in the 96 years since
10	that decision, there has emerged I would say sort of a
11	dichotomy or two different views of monopoly. While we
12	would all agree that they can have their evils, and this
13	was articulated in part by John Hicks in 1935, who
14	talked about the evils of monopoly in the terms of a
15	quiet life. He talked about the fact that the
16	monopolist may not be out there trying to get the
17	highest price he absolutely can get, maximizing in the
18	short term the most profit that he or she can get, but
19	really, it is the lack of competitive zeal, the ability
20	to sit back and relax, to not have to research, develop,
21	to innovate at a frantic level. That is a major harm of
22	monopoly, and that is something on which we are very
23	focused in terms of preventing.
24	Now, at the same time, the Supreme Court just
25	last year articulated a different view of monopoly. In

26

1	the Trinko decision, the Court said, "The mere
2	possession of monopoly power and the concomitant
3	charging of monopoly prices is not only not unlawful, it
4	is an important element of the free market system. The
5	opportunity to charge monopoly prices, at least for a
6	short period, is what attracts business acumen in the
7	first place. It induces risk-taking that produces
8	innovation and economic growth."
9	All the way back in 1942, in Capitalism,
10	Socialism and Democracy, Joseph Schumpeter talked about
11	a similar process called creative destruction or the
12	gales of creative destruction, and I compliment my staff
13	who came up with the tornado there, but I have always,
14	since I read this in college, this -- be careful of the
15	gale behind you -- I have always liked this image,
16	because it talks about how the marketplace is a rough
17	place. It involves vigorous aggressive activity, people
18	fail, people are driven out of business, but it is
19	through that destructive process that you get creation.
20	Indeed, a similar image I was thinking about
21	recently, when somebody was talking to me about the
22	National Forest Service, I grew up watching the
23	commercials about Smokey the Bear and how forest fires
24	were such a terrible thing. How could we be against
25	forest fires? It turns out the National Park Service

27

1	realizes that preventing forest fires can be a bad
2	thing; that if you prevent them for too long, you create
3	much bigger, larger, hotter fires that cause more
4	permanent destruction to the ecosystem when they do
5	occur. Periodic smaller fires are actually a good and
6	healthy part of the process. That to me is another
7	illustration of this basic image. Competitive, creative
8	destruction in the marketplace is something that we want
9	to preserve and protect, not chill along the lines that
10	Debbie was talking about.
11	So, how do we reconcile these two views of a
12	monopoly, as a bad thing that causes sloth and
13	relaxation and a lack of competitive drive versus the
14	benefits of creative destruction, the opportunity to get
15	to a monopoly? Well, this somewhat conflicting view was
16	illustrated in a book written in 1964, and this was R.W.
17	Grant expressing some frustration about the treatment of
18	monopolies, and I will read this to you in a moment, but
19	the basic story here is of a man named Tom Smith who
20	invents a bread machine. It will produce terrific
21	bread, it will slice it, it will wrap it, all for less
22	than a penny a loaf, and as you can imagine, he very
23	shortly owns the market for bread in the United States
24	and is making large sums of money. He is ultimately,
25	however, brought low by the men of antitrust who bring

28

1	an antitrust case against him for making too much money
2	on the backs of consumers and driving everybody else out
3	of business, and he crafts a poem here to illustrate
4	this frustration.
5	"You're gouging on your prices
6	if you charge more than the rest
7	But it's unfair competition
8	if you think you can charge less!
9	A second point that we would make
10	to help avoid confusion:
11	Don't try to charge the same amount!
12	That would be collusion.
13	You must compete -- but not too much
14	for if you do, you see
15	then the market would be yours -ï½
16	and that would be monopoly!
17	It's very similar in many ways to the admonition
18	of Learned Hand in the Alcoa case who said that the
19	successful competitor, having been urged to compete,
20	must not be turned upon when he or she succeeds.
21	So, having expressed that frustration back in
22	the 1940s and 1960s, where are we today? One of our
23	esteemed both I would say academics and judicial members
24	of the antitrust community, Richard Posner, Judge
25	Posner, remarked just last year, "Antitrust policy

29

1	toward 'unilateral abuses of market power' is 'the
2	biggest substantive issue facing antitrust today.'"
3	And if I can, if you will excuse me, preempt
4	Herb possibly, last year Herb is quoted or wrote,
5	"Notwithstanding a century of litigation," 96 years
6	since the Standard Oil decision, "the scope and the
7	meaning of exclusionary conduct under the Sherman Act
8	remain poorly defined."
9	Now, there are areas where I think there are
10	relatively easy answers. Doug Melamed has written about
11	the concept of naked exclusionary practices. I mean, if
12	you blow up your competitor's factory, few of us would
13	find that to be defensible conduct. That's a fairly
14	easy case for not finding liability. I also think there
15	are some fairly easy candidates for safe harbor
16	provisions. If you engage in conduct that merely
17	reduces your cost of production, that seems to me
18	beneficial to consumer welfare.
19	The difficulty lies in cases, as Debbie
20	referenced, that have the potential for both beneficial
21	cost reductions, innovation, development, integration,
22	and at the same time potentially anticompetitive
23	exclusion. How do we deal with those situations?
24	Well, some relatively recent Supreme Court
25	decisions have shown progress in this direction. In the

30

1	Brooke Group case, which is, of course, a predatory
2	pricing case, it dealt specifically with the issue of
3	recoupment and holding that Liggett in that case had not
4	shown the opportunity or the ability to recoup, but the
5	case in my view, at least, stands for more than that and
6	discusses, for example, specifically how harm to a
7	competitor does not demonstrate harm to competition.
8	There was little doubt in that case that there were
9	discount programs aimed at and/or that had a harmful
10	effect on Liggett, but the Court was quite clear that as
11	long as that does not harm competition, that is not an
12	antitrust problem.
13	Second, the Court also talked about the
14	practical ability of a judicial tribunal to regulate a
15	problem and avoid chilling legitimate price cutting.
16	It's recognizing the limitations of the body that is
17	administering the law. I would expand that to include
18	the limitations of agencies as well as courts, but it's
19	certainly a relevant consideration, and recognizing that
20	aggressive price cutting can be beneficial for consumers
21	and we do not want to chill it. Thus, it created
22	effectively a safe harbor against predatory pricing
23	claims where the prices were above some appropriate
24	measure of cost.
25	And the Court expressly acknowledged in creating

31

1	the safe harbor that there was at least the theoretical
2	possibility that there could be harm to consumers, harm
3	to consumer welfare, from some above-cost pricing, but
4	recognizing it was likely to do more harm than good to
5	try to ferret out those individual cases.
6	More recently, in the Trinko decision, the Court
7	obviously had a somewhat more limited holding but
8	discussed on a broader basis some of these same similar
9	Section 2 issues. It underscored the need for
10	administrable rules, clear objective standards. It
11	talked about the fact that being able to craft a remedy
12	that is both clear and administrable by the Court is
13	very important, endorsing Professor Areeda, in that no
14	court should impose a duty to deal that it cannot
15	explain or adequately and reasonably supervise, and
16	implicitly, at least, that not all problems may have
17	antitrust solutions.
18	While I think there are many areas of consensus,
19	there are many areas where we have a lot to learn. As
20	Debbie indicated, our panels are going to focus on
21	different aspects of conduct. We will start on Thursday
22	with a panel discussing predatory pricing and predatory
23	buying. Brooke Group answered a lot of questions. It
24	did not answer, among other things, what is the
25	appropriate measure of cost? Is it marginal cost? Is

32

1	it average cost? Is it average avoidable cost? Is it
2	average total cost? There has been a lot of discussion
3	about that, and we are looking forward to hearing
4	people's views on that.
5	The question on predatory pricing or remedy, are
6	you going to enjoin lower pricing? Weyerhaeuser is a
7	case with which you may be familiar. There's a cert
8	petition pending before the Supreme Court right now. It
9	involves a predatory bidding situation. The Solicitor
10	General's Office has filed an amicus brief on this front
11	encouraging the Court to take cert and to examine it. I
12	view it, at least, as an opportunity for the Court to
13	reaffirm in the Section 2 context that clear and
14	objective standards are extraordinarily important.
15	There's a jury instruction at issue in this case that
16	talks about prices that are unfair or unnecessarily
17	high. This is an opportunity for the Court to make
18	clear that a jury's post hoc determination of what it
19	believes was unfair is not the appropriate criteria for
20	determining whether or not there should be Section 2
21	liability.
22	Refusals to deal will follow. Again, this
23	raises very significant issues. When, if ever, should a
24	firm be compelled to deal with a competitor? If you are
25	compelled to deal, under what terms and conditions?

33

1	Will the Court be able to administer it? A range of
2	issues which we are, again, looking forward to hearing
3	the experts' views on it.
4	Loyalty discounts, another area that we will be
5	looking at. A couple of years ago, the United States
6	urged the Supreme Court not to take cert in the LePage's
7	case. That involved bundled discounts. That was not
8	because we necessarily agreed with the Third Circuit's
9	decision or analysis. Indeed, if you parse that
10	decision, I think it is very difficult to come up with a
11	clear standard of liability. There has been, in the
12	wake of LePage's, a flurry of attention by academics, by
13	legal scholars, on this issue of bundled discounts,
14	loyalty discounts, and we are looking and hoping to see
15	whether or not any consensus has developed on any of
16	these issues.
17	Should it be viewed as a predatory pricing
18	tactic, as exclusive dealing, as a tying tactic? Are
19	there safe harbors that can be developed even if we
20	cannot develop a single, clear answer for all cases?
21	Tying and exclusive dealing, Debbie mentioned that you
22	sometimes, when you see things in a competitive market,
23	that ought to make you question whether or not there are
24	benefits associated with it. Tying and exclusive
25	dealing can have anticompetitive effects. Look at our

34

1	Dentsply case as a recent example. By the same token,
2	we see these practices in competitive markets, and we
3	need to better understand what benefits there are and
4	when there are not.
5	Towards the end of the year, we expect to turn
6	toward some more general principles. Is there an
7	overarching standard for Section 2 cases and liability?
8	We all agree that consumer welfare is an appropriate
9	standard. Trying to operationalize that in a particular
10	case with particular conduct is more challenging, and
11	there is less agreement on that. Debbie outlined the
12	range of potential tests. The Antitrust Division in a
13	number of recent cases looked to the no-economic-sense
14	test. As I have talked with people about that, one
15	issue that I find is that people have different ideas of
16	what the test is. So, over and above discussing what
17	the appropriate test ought to be, there is some
18	confusion about what is meant in terms of what are you
19	going to look at and what the rules are. That may be
20	part of the semantic difference that Debbie was
21	referencing. Clarifying some of those things as well as
22	the underlying substantive issues I think can be
23	beneficial.
24	We may look at the issue of whether there are
25	different duties or different criteria for tying claims

35

1	under Section 3 of the Clayton Act versus Section 1 or
2	Section 2 of the Sherman Act.
3	Here, I have two reasons for putting this up.
4	As you can see, this associate is responding to a
5	request, "I'll be happy to give you innovative thinking.
6	What are your guidelines?" An example of having too
7	cabined an approach, too narrow a guidelines can be the
8	antithesis of innovative thinking, can restrain the
9	benefits that you may achieve through your innovation
10	and development. That is part of the creative
11	destruction that we want to encourage, not discourage,
12	as this cartoon suggests may be happening. So, I raise
13	that to say that while I am now going to talk about six
14	possible principles to inform our discussions, I do not
15	mean them to cabin or prevent a wide-ranging, open and
16	frank exchange of ideas.
17	So, first off, individual firms with market
18	monopoly power can act anticompetitively and harm
19	consumer welfare, and we should seek to identify and
20	prosecute such conduct. This is an important first
21	principle. If it were not true, we could just abolish
22	Section 2. That is not what we are here to do. We are
23	here to better focus and identify those instances where
24	there really is harm to consumer welfare.
25	Second, mere size, mere market share, does not

36

1	necessarily demonstrate competitive harm. It can
2	demonstrate superior acumen, effort, zeal, et cetera.
3	Third, injury to competitors does not
4	demonstrate competitive harm, a point that has been
5	talked about in a number of contexts.
6	Fourth, the need for clear, objective and
7	administrable rules, so that businesses, at the time
8	they are taking actions, can understand where the lines
9	are and can conform their behavior so they are not
10	deterred from engaging in procompetitive activity, so
11	that courts are not asked to do things that are beyond
12	their competence, and that agencies can do the same.
13	Fifth, avoid chilling procompetitive conduct,
14	and certainly an interrelated point, self-explanatory.
15	And finally, the remedy must promote
16	competition. A remedy that harms competition can be
17	worse than no remedy at all, an important point worthy
18	of bearing in mind.
19	Again, I want to thank the FTC, our panelists
20	for agreeing to kick off these hearings. We will
21	continue again on Thursday. We very much are interested
22	in a free, open and wide-ranging discussion of these
23	issues and are excited about the prospect.
24	With that, I will turn it over to Herb.
25	(Applause.)

37

1	DR. HOVENKAMP: Thank you. I am very grateful
2	and appreciative of being invited here, with particular
3	thanks to Chairperson Majoras and General Barnett for
4	extending this invitation.
5	In keeping with the thrust of this opening
6	meeting, which I believe is quite general, what I would
7	like to do is give kind of an overview of where I think
8	the fault lines and concerns in Section 2 lie. In the
9	future, future hearings, you are going to hear about
10	specific practices such as predatory pricing or refusals
11	to deal in considerable detail, and I am not going to do
12	that today. I am going to go through them rather
13	quickly and just point out where I think work needs to
14	be done and where the FTC and the Antitrust Division and
15	private litigants can use some clarification and
16	understanding.
17	I am going to divide my talk into three parts,
18	though the parts are not equal in size. First, a very
19	short one on market power or monopoly power, then a
20	rather long one on conduct issues, and then finally, a
21	much shorter one again on remedies.
22	With respect to power, the Merger Guidelines, in
23	particular the 1992 Merger Guidelines, the series of
24	guidelines that began with 1984, did a remarkable job of
25	rationalizing and simplifying the approach to market

38

1	delineation and assessment of the potential for
2	collusion or other types of anticompetitive behavior
3	that grow out of mergers. Some portions of the Merger
4	Guidelines market delineation sections are relevant to
5	Section 2 enforcement, but many are not, because the
6	question that one asks in a Section 2 case is
7	fundamentally different from the one that one asks in a
8	merger case.
9	In a merger case, we generally start out with
10	the presumption that a market is more or less
11	competitive, it may be oligopolistic or moderately
12	competitive prior to the merger, and what we really want
13	to know is whether the quality of competition is going
14	to deteriorate as a consequence of the merger. In
15	keeping with that, the SSNIP test, small but significant
16	nontransitory increase in price test, considers whether
17	a further increase in price would cause new entry or
18	other situations that would make this future price
19	increase unsustainable.
20	In a Section 2 case, by contrast, the opening
21	presumption is that the defendant or the firm under
22	examination is already a monopolist, is already charging
23	monopoly prices, and as a result, the SSNIP test is
24	really not the appropriate one in most circumstances,
25	although it certainly could be relevant in certain cases

39

1	like those involving an attempt to monopolize where the
2	defendant is not a monopolist at the time the conduct is
3	being assessed.
4	I do not have a solution to propose here. Those
5	of you who are familiar with this area know that this
6	involves something that in monopolization law we call
7	the Cellophane fallacy or the fallacy of inferring that
8	a firm lacks power because there is high
9	cross-elasticity of demand with the products of others
10	at current market prices, and, of course, if you
11	multiply that examination by asking what the response
12	would be to a yet further increase by a firm that is
13	already a monopolist, you might very well conclude that
14	the firm lacks this type of market power, because in
15	response to a yet further price increase, there would be
16	so much substitution away from the dominant firm's
17	product that the price increase would be unprofitable.
18	Well, if you took that approach, you would be
19	committing an error; namely, you would be ignoring the
20	fact that that firm is already a monopolist and
21	presumably already charging its profit-maximizing price.
22	So, I think one of the things that ought to be of
23	concern to the agencies as they go through these
24	hearings is to pay some special attention to the
25	formulation of usable presumptions that single firms can

40

1	use for assessing whether they have individual market
2	power and thus can be made liable to a Section 2
3	inquiry.
4	Let me just add to that, that that may involve
5	certain approaches that we have more or less given short
6	shrift to or rejected in the past. For example, it may
7	mean that we will not look at residual elasticity of
8	demand, which looks at the existing power that firms
9	have. We may have to look at things like price-cost
10	margins or rates of return. Some of these approaches
11	have been discredited in the past, but that does not
12	mean that they cannot be rehabilitated.
13	Okay, I want to spend a little more time on
14	monopolizing conduct. I am going to open by giving the
15	definition of monopolizing conduct from The Antitrust
16	Law Treatise that I am privileged to write, because it
17	is very general, has a number of flaws, but
18	nevertheless, I happen to like it for reasons I will
19	explain in a little while. The Antitrust Law Treatise
20	defines exclusionary conduct as conduct that is, number
21	one, reasonably capable of creating, enlarging or
22	prolonging monopoly power by impairing the opportunities
23	of rivals; and two, that either does not benefit
24	consumers at all or is unnecessary for the particular
25	consumer benefits that the acts produce; or three,

41

1	produces harms that are disproportionate to the
2	benefits; and finally, the assessment of the conduct
3	must be within the administrative capacity of the
4	antitrust tribunal.
5	Like I say, that test is very general. It is
6	not particularly helpful to assessing particular
7	instances of exclusionary conduct if it is the only
8	thing you have. You certainly would not want to give a
9	jury that test as an instruction and shut them up with
10	no further instruction and ask whether the defendant's
11	conduct was exclusionary, but the test was never
12	intended that way. It was, in fact, designed to be a
13	basic principle to be used in conjunction with specific
14	rules for specific types of antitrust cases, and it is
15	my view that that is fundamentally what Section 2
16	conduct jurisprudence needs to do.
17	I think there are very, very helpful general
18	tests. I like Greg Werden's no-economic-sense test. I
19	think there is much to be said for it. I think it
20	produces a few false negatives. Nevertheless, it's a
21	very, very good starting point. I like Judge Posner's
22	test that Chairperson Majoras mentioned in her talk,
23	which is conduct which under the circumstances is
24	capable of excluding an equally efficient rival. Once
25	again, I think it produces a few too many false

42

1	negatives, but they are good starting places.
2	However, none of them is a substitute for the
3	formulation of good technical rules covering individual
4	types of conduct; namely, pricing, abuses of the
5	intellectual property system, refusals to deal and so
6	on, okay?
7	In the few minutes I have, I cannot do any more
8	than scratch the surface, but I would like to give you
9	just a few observations about where we are in various
10	areas involving specific exclusionary practices and
11	where I think some of the problems lie.
12	With respect to predatory pricing, I believe
13	that both the Areeda-Turner test, as it was formulated
14	in 1975 and has later been incorporated into The
15	Antitrust Law Treatise, plus the elaboration of the
16	recoupment requirement in the Brooke Group case in 1993,
17	fundamentally set predatory pricing law on the right
18	track. I am a strong believer in the view that prices
19	must be below some measure of cost. Furthermore, they
20	must be below some measure of incremental cost; that is,
21	pricing is driven by concerns for variable costs, not
22	principally by fixed costs. That does not mean that
23	there are not a few problems.
24	One problem that I think needs to be assessed is
25	the problem of predatory pricing in oligopoly industries

43

1	by nondominant firms. That was, in fact, the facts of
2	Brooke Group. Strictly speaking, that may not be a
3	Section 2 issue. In fact, it may be an issue where the
4	Justice Department might reconsider its long-standing
5	opposition to bringing Robinson-Patman Act suits since
6	the late 1970s report on the Robinson-Patman Act and
7	create an exception for primary line enforcement given
8	the premise that with respect to primary line
9	enforcement, the principles that the Court follows are
10	basically the principles that are laid out in the
11	Sherman Act, and as a result, all of the overreaching
12	that applies to secondary line enforcement of the
13	Robinson-Patman Act need not apply here.
14	The problem with predatory pricing and oligopoly
15	is that victims have a different set of incentives than
16	they do in monopoly. Predatory pricing as a Section 2
17	problem involves predatory pricing designed to destroy a
18	rival. That is a very, very difficult thing to do. The
19	rival clearly has incentives to resist.
20	On the other hand, predatory pricing and
21	oligopoly frequently is used simply to enforce or bring
22	the oligopoly back into order so that the noncompliant
23	firm will once again raise its price to the oligopoly
24	levels; that is, the set of incentives that the target
25	of predatory pricing and oligopoly has are incentives to

44

1	rejoin, start making profits once again. As a result, I
2	believe predatory pricing in oligopoly industries is
3	fundamentally a more plausible strategy than strict
4	monopoly predatory pricing, and I think it needs to be
5	given somewhat closer scrutiny.
6	The other problem has to do with the measurement
7	of relevant costs. As I said before, I think the proper
8	measure of cost is incremental cost, which can mean
9	short-run marginal cost, short-run marginal cost with
10	some kind of additional factor for depreciable long-term
11	assets. It can mean average variable cost, as it was in
12	the Areeda-Turner formulation. The average variable
13	cost tests or the marginal cost tests simply don't work
14	very well in certain kinds of markets that have very
15	high fixed cost components and particularly in markets
16	that are characterized by a lot of intellectual property
17	or certain kinds of public utility or transportation
18	markets, such as the airline industry.
19	I think Ken Elzinga's analysis in the Spirit
20	Airlines case last year in the Sixth Circuit was a very
21	good first step, but the Government shouldn't be losing
22	predatory pricing cases in the airline industry. It is
23	the one industry where predatory pricing claims seem
24	plausible, and some attention needs to be paid to
25	modifying or, if necessary, rejecting and adopting a

45

1	different cost test for such industries.
2	On the Weyerhaeuser case and predatory buying, I
3	am one of the critics. I hope the Supreme Court sees
4	fit to follow the SG and grant cert. I think the
5	instruction that General Barnett described that
6	permitted a jury to find simply that predatory buying
7	occurs when the defendant pays too much or more than a
8	fair price is an atrocity. I think few people fully
9	appreciate how frequently such situations can come up;
10	that is, buying of inputs during times of scarcity.
11	This is not going to be an idiosyncratic situation.
12	This kind of case will come up a lot if the Ninth
13	Circuit's decision is permitted to stand.
14	Now, having said that, the question is what kind
15	of test to come up with. Well, in the Weyerhaeuser
16	case, where first of all the timber at issue accounted
17	for some 60 or 70 percent of the value of the finished
18	hardwood, and secondly, where at least according to the
19	jury, the hardwood was resold in a competitive market, I
20	think an average variable cost test might work quite
21	well; that is, buying is predatory if it forces the
22	defendant's resale prices to below its costs.
23	I am a little troubled by the use of an average
24	variable cost or marginal cost test, however, in a
25	situation where, number one, the defendant may sell in

46

1	an oligopoly, and number two, where the input on which
2	predatory buying is claimed is a relatively low
3	proportion of the value of the finished product, because
4	in that case, the variation in purchase might actually
5	fall within the margins that the firm charges, that it
6	is going to be too hard to detect predatory buying in
7	cases where the value of the purchased input is only a
8	tiny proportion of the value of the finished product.
9	In all cases, however, I believe that there
10	should be a recoupment requirement equivalent to that in
11	Brooke Group and that the Ninth Circuit erred not only
12	in its failure to require a showing of prices below
13	cost, but also in its failure to require a fairly strict
14	showing of recoupment.
15	With respect to patents, there is too much to
16	say and too little time. I just want to make one fairly
17	general observation. Mr. Barnett mentioned Joseph
18	Schumpeter's Capitalism, Socialism and Democracy, this
19	very, very important book in 1942 which opined in the
20	chapter on creative destruction that the amount of
21	welfare contributed to the economy through innovation is
22	far, far greater than the amount contributed by moderate
23	movements from oligopoly to competitive industries. I
24	mean, Schumpeter basically looked at the prior half
25	century or so of development as a result of the second

47

1	industrial revolution, of the theorizing of economists
2	like Edward Chamberlin and Joan Robinson, who were very
3	upset about oligopoly and imperfections in the economy,
4	and said, "You'd think to listen to these people that
5	American consumers were much, much impoverished compared
6	to their position in the 1870s, and, in fact, nothing
7	could be further from the truth."
8	Well, where do all those gains come from if we
9	are now in this oligopolistic era? And one of the
10	things Schumpeter concluded is that they came from
11	innovation. Schumpeter's premises were formalized and
12	given empirical support in Robert Solo's work in the
13	1950s in which Solo himself concluded that as much as 80
14	percent of economic gain comes from innovation rather
15	than simple improvements in price-cost relationships.
16	Now, neither Schumpeter nor Solo was talking
17	about IP law. They were talking about innovation, and,
18	of course, there is this enormous lingering question out
19	there of whether the IP laws we have are sufficient to
20	facilitate the optimal amount of innovation or whether
21	they, in fact, may hinder innovation. Fundamentally,
22	that is not antitrust's problem. The antitrust laws
23	need to accept the existing IP laws, warts and all, and
24	I personally believe there are a fair number of warts.
25	One thing, however, that that work suggests is

48

1	that antitrust needs to be much more concerned with
2	restraints on innovation. We have generally measured
3	harm in the antitrust laws by looking at price-cost
4	relationships, deviations from marginal cost pricing.
5	Harm to innovation is always included kind of as an
6	afterthought, but it has never been very well formalized
7	into our models of harm, and actually, there are pretty
8	good reasons for that. We have very good rules for
9	determining when prices deviate from marginal costs and
10	what the price elasticities facing firms are.
11	Predicting the consequences of restraints on innovation
12	is far more difficult, because innovation always takes
13	us by surprise.
14	We will never know, for example, what the
15	consequences were of Microsoft's successful attempts to
16	get Intel to stop developing a Java-enabled chip. How
17	good would it have been? Would it have done all the
18	things that Bill Gates feared in commoditizing the
19	platform market and so on? Those are very hard things
20	to predict, and for that reason, I believe courts are
21	rightfully skeptical when they turn away private
22	plaintiffs who claim that the injury that they suffer is
23	an injury caused by a lack of innovation.
24	So, I believe this is one area where the
25	Government should move into the fore, because they do

49

1	not need to prove damages, they do not need to prove
2	causation in the strict private plaintiff sense. I
3	think restraints on innovation are something that need
4	far more development in Section 2 law than they have
5	received in the past.
6	With respect to vertical exclusion, I just have
7	a couple of comments. First of all, there has been a
8	not so subtle move over the last four or five years in
9	government enforcement to move away from Section 1 of
10	the Sherman Act and Section 3 of the Clayton Act and
11	towards Section 2 of the Sherman Act as a device for
12	enforcing laws against tying or tying-like practices and
13	exclusive dealing, and I believe that is the correct
14	movement. Fundamentally, tying and exclusive dealing
15	ought to be regarded as dominant firm exclusionary
16	practices. They are rarely anticompetitive at
17	nondominant levels, and fundamentally, they do not
18	depend on agreement in any meaningful sense of the word.
19	Unlike resale price maintenance or Sylvania-style
20	restraints, they are typically not the product of
21	bargaining and traditional agreement between dealers and
22	manufacturers.
23	No, most tying and most exclusive dealing is
24	imposed by manufacturers unilaterally on dealers. The
25	dealers generally do not like it, but they accept it as

50

1	the price of a dealership. It ought to be treated as an
2	exclusionary practice, number one. The agreement
3	requirements really get in the way of appropriate
4	analysis of tying and exclusive dealing in most
5	situations. And finally, the market power requirement
6	should be equivalent to those that we assess in
7	monopolization cases.
8	So, I laud the increased scrutiny of tying and
9	exclusive dealing under Section 2 of the Sherman Act.
10	Microsoft included both, but the Government won on its
11	Section 2 tying claims. Dentsply, of course, the
12	exclusive dealing case that the Government won a year or
13	two ago, was a Section 2 case.
14	On bundled discount -- you are going to have a
15	big hearing on these, right? You are going to talk
16	about bundled discounts a lot? Are they predatory
17	pricing or are they tying? I think they are a little
18	bit of both, and I think the way to analyze them is by
19	asking two questions in two different stages.
20	The first question you ask is, are two goods
21	subject to a bundled discount bundled together? Well,
22	what does that mean? Well, it means that an equally
23	efficient firm that offered only one of them could not
24	match the bundled offer. How do you get there? Well,
25	as several papers have shown, you basically attribute

51

1	the entire discount to the product upon which exclusion
2	is claimed, and then you ask whether the price of that
3	product, subject to the full discount, has fallen below
4	a relevant measure of cost, whatever cost measure you
5	would use in a predatory pricing case, okay?
6	That gets you to bundling; that is, that
7	predatory pricing test gets you an answer to the
8	question, are the two firms -- are the two products
9	bundled together? And if the answer is that no equally
10	efficient firm that offered only one of the products can
11	match the price, then they are bundled together, but
12	that is only the beginning rather than the end of the
13	inquiry. Tying is explicit bundling of products
14	together, and yet most tying is perfectly legal. So,
15	once we have decided that two products are bundled
16	together, we have yet a further set of questions to ask
17	about whether there is foreclosure, whether the
18	foreclosure is justified under the circumstances by cost
19	reductions, improvements in consumer satisfaction,
20	quality control, in many instances price discrimination,
21	and so on.
22	Finally, on conduct, on refusals to deal, my
23	suggestion is that the Government simply get out of the
24	business of enforcing the law against simple refusals to
25	deal. Now, conditional refusals are something else.

52

1	Conditions usually mean exclusive dealing or tying.
2	Lots of things, including price fixing, can amount to
3	conditional refusals to deal, but if we are talking
4	about simple refusals to deal in the Trinko or Aspen
5	sense, I think the administrative problems are so
6	horrific, the disincentives created to competitive
7	behavior are so substantial, that the best thing that
8	the Government can do is stay away, and, in fact, that
9	is pretty much what they have been doing, even going so
10	far as to support the defendants in the Trinko case.
11	Okay, then let me turn finally and very briefly
12	to the subject of remedies. Both General Barnett and
13	Debbie Majoras spoke at some length about the importance
14	of remedy. I simply want to underscore what they said.
15	In fact, I would go a little bit further and say that
16	every Section 2 action that the Government brings ought
17	to begin with an exit strategy, right? We have talked
18	about Iraq, we have talked about exit strategies, and
19	now we have discovered that whatever exit strategy we
20	have, we probably could have had a better one, and the
21	same thing applies to Section 2.
22	Section 2 has no moral content. The only
23	purpose in bringing these cases is to make the economy
24	work better, and if you do not have a clear picture of
25	the kind of remedy you want when you go in, then you

53

1	really have to wonder whether it is worth bringing the
2	action to begin with.
3	For a long, long period of our history,
4	beginning with Standard Oil and through the 1960s, the
5	preferred remedies were structural or mandatory breakup
6	of firms. For relatively good reasons, those kinds of
7	remedies have fallen into some disrepute. Many of them
8	were very, very poorly designed. For example, the
9	remedy in the United Machinery case, which may have
10	ruined the firm, although there was some good evidence
11	that USM's technology in its Beverly, Massachusetts
12	plant was pretty obsolete already to begin with, or
13	remedies like the one in Grinnell, which didn't really
14	break up the monopoly at all, but just divided up the
15	market into a whole bunch of little monopolies.
16	Today, we operate in a regime in which
17	structural remedies in Section 2 cases appear to be
18	disfavored; conduct remedies are preferred.
19	Unfortunately, I think the record that we are developing
20	with respect to conduct remedies is not much better than
21	the record we developed with respect to structural
22	remedies in the 1960s and earlier. I think the verdict
23	is still out on the Microsoft remedy, largely because of
24	the two-year extension, but once that time period has
25	run, we have to look back and say, "Well, exactly what

54

1	did we accomplish here in terms of making this market
2	more competitive? And to the extent the market is more
3	competitive, to what extent was it owed to the remedy?"
4	I would like to see some more serious attention
5	be paid once again to structural remedies or at least
6	modified structural remedies, things like compulsory
7	licensing, as mechanisms for restoring competition.
8	Compulsory licensing is a dirty word in the United
9	States when we are talking about general forcing of
10	firms to share their patents, but that is not what we
11	are talking about here. We are talking about proven
12	antitrust violators who are frequently forced to give up
13	their plants and other kinds of hardware. Compulsory
14	licensing under that set of circumstances is a perfectly
15	viable remedy, and I would like to see us use it much
16	more seriously than we have in the past.
17	I am afraid I have gone over my time, and so I
18	will turn the floor over to Professor Carlton. Thank
19	you.
20	(Applause.)
21	DR. CARLTON: Okay, thank you. It is a pleasure
22	to be here, and I express my appreciation to the
23	organizers for inviting me, and I am honored to sit with
24	such distinguished panelists.
25	Exclusionary conduct is an important policy

55

1	topic. There is a great amount of debate as to what is
2	an exclusionary act and how to deal with it. There is
3	much less consensus on what bad acts are and how to
4	adjudicate them in the context of Section 2 claims,
5	than, for example, what cartel behavior is or how to
6	handle cartl claims. The real problem is that
7	competition harms rivals just like exclusionary
8	behavior, and it is sometimes easy to confuse the two.
9	As a general matter, it is very hard to study
10	what should be the optimal policy for exclusionary
11	conduct. The reason, one reason, is that the biggest
12	effect of any antitrust policy is likely to be, not on
13	litigants in litigated cases, but rather, on firms that
14	are not involved in litigation at all but are forced to
15	change their business behavior in contemplation of legal
16	rules. That means that although it is definitely
17	informative for economists and lawyers to study the
18	outcome of individual cases and you can learn a lot -ï½
19	did the court get it right, did they get it wrong -ï½
20	that is not really a study of antitrust policy.
21	To appropriately do a study of antitrust policy,
22	you have to look at either times when the antitrust laws
23	were adjudicated differently, that is, there was a
24	different policy, or perhaps you have to look at
25	different countries, and that is hard to do. Looking at

56

1	litigated cases will give you a biased view. They are
2	very self-selected cases and really will not allow you
3	to focus on what may well be the biggest costs of an
4	antitrust policy, and that is, the chilling effect it
5	has on the behavior of nonlitigants.
6	Since this is an introductory panel, one of the
7	advantages you get is you can yell out these hard
8	questions, and then when they are not answered by the
9	end, you can say, "See, you didn't answer my question,"
10	so I am not suggesting I have a simple answer. I am
11	just saying here that this is actually a quite important
12	problem and definitely deserves some research time.
13	What I will talk about today and on which I
14	think I do have some answers are two topics. One has to
15	do with the profit sacrifice test, or its variant, the
16	no-economic-sense test, and the other has to do with
17	market definition. Each of those is used or proposed to
18	be used as a tool to uncover Section 2 violations.
19	Now, it is important to distinguish a tool from
20	an objective. The objective is to maximize consumer
21	welfare, and whether when I say consumer welfare, I mean
22	consumer plus producer surplus or just consumer surplus,
23	that is a debate we can hold for another time. It does
24	not matter to the remarks I am going to make.
25	The tools you choose are used to identify acts

57

1	that harm welfare. Now, the use of tools necessarily
2	entails errors. You are going to have false negatives
3	and false positives. It is going to happen because,
4	one, you do not always have perfect information. Even
5	if you are the smartest economist, you make errors, and
6	if you are a juror who knows no economics, you might
7	make errors, and second, the tools are not quite the
8	same as the objective. So, the tools will differ from
9	the objectives sometimes, so there definitely will be
10	errors, and the real question is whether and when tools
11	should be used.
12	From an economic point of view, the question of
13	whether a particular act harmed consumers is a very
14	well-posed question that I could, for example, assign to
15	a Ph.D. student writing his thesis, and that person
16	could go about trying to answer it with economic tools
17	and econometric tools. Now, it is true it sometimes may
18	be hard for that person to reach an answer, especially
19	if there is both an efficiency effect and an
20	exclusionary effect on rivals, but sometimes it will not
21	be so hard, okay, but it is a well-posed question.
22	Now, from a policy perspective, the DOJ and the
23	FTC, with their staffs of knowledgeable economists and
24	attorneys, should be focusing on answering the direct
25	question, is there harm to welfare, and not on the use

58

1	of possible tools, though obviously they need,
2	especially if they are going to think about litigation,
3	they need to worry how judges and juries will use
4	whatever tools they are told to use when a case is
5	litigated.
6	Now, matters change as the inquiry shifts away
7	from the government agencies to courts with judges and
8	juries who may have less economic sophistication, and
9	their simple but imperfect rules may be better than an
10	unstructured inquiry, but simple rules should not be
11	viewed as anything but simple rules, crude guides that
12	sometimes work for some acts, but not all. It can be
13	dangerous to use a simple rule that could ultimately
14	subvert the goal of maximizing welfare.
15	Now, let me turn to first the profit sacrifice
16	test and then to market definition. The profit
17	sacrifice test, or a close relative, the
18	no-economic-sense test, asks would the act make sense
19	but for its exclusionary effect? And that test may work
20	fine in the hands of some of its accomplished
21	proponents, and in particular, I have in mind Greg
22	Werden, who is sitting here in the audience. My hunch
23	is if I sit down with Greg and we are talking about an
24	exclusionary act, we are going to reach agreement nine
25	times out of ten, and he is aware of the limitations of

59

1	the test and would carve out exceptions so it would not
2	be misused.
3	What I worry about is when there is someone
4	implementing the test who is not as smart as Greg.
5	Suppose they have a judge or a jury who is not an
6	economist? What worries me about the test is that it
7	raises all the danger signs associated with possibly
8	confusing competition with exclusionary conduct. Let me
9	try and explain why, and I will give you two or three
10	reasons.
11	First, all strategic behavior -- and every
12	business school teaches this -- all strategic behavior
13	is designed to improve one firm's position relative to
14	the other one. It is relative position that matters,
15	and that is what is going to often determine the outcome
16	of a competitive battle. Investments in advertising,
17	investments in R&D, price discounts, all of these could
18	mistakenly and easily be condemned by jurors convinced
19	that the firm engaged in the act could have been more
20	accommodating, less exclusionary. That "but for"
21	standard -- that is, would the act make sense but for
22	the exclusionary effect -- that hypothetical thought
23	experiment is actually quite a difficult one to
24	implement. That does not mean if you are measuring harm
25	to consumer welfare that you can necessarily get around

60

1	it. Maybe it is unavoidable sometimes, but I am worried
2	when you are postulating the "but for" the exclusionary
3	conduct, what hypothetical world will you be proposing,
4	and what worries me is there could be a very high error
5	rate associated with application of that rule.
6	The second reason I am worried about it is
7	because of the way it is formulated: Does it make
8	economic sense to engage in this behavior? Now, Ronald
9	Coase, one of my colleagues at the University of
10	Chicago, always fond of explaining how little economists
11	know about business and that businessmen really know a
12	lot about business, and just because an economist cannot
13	understand an action or a jury cannot understand an
14	action, he said why should that create antitrust
15	liability for the poor firm? And that I think is a
16	serious concern.
17	The profit sacrifice test may sometimes work and
18	may be appropriate for some actions, but I think it is
19	dangerous to enshrine it as a general proposition. It
20	strikes me as much more appropriate to devise tests and
21	screens that fit particular exclusionary acts,
22	especially because under Section 2 we have a range of
23	such widely different behavior that people have attacked
24	as exclusionary, so many different acts fall under
25	Section 2, and with each act, I would be concerned as to

61

1	how the legal treatment of the act is creating special
2	risks for chilling competitive behavior, and that is
3	going to differ from act to act to act.
4	So, for example, to pick up on what Herb was
5	saying, let us suppose you are looking at an industry
6	that is undergoing rapid technological change, and there
7	is an exclusionary claim that the way the product was
8	designed is a problem. Well, you have to worry in that
9	instance whether you are depriving consumers of a new
10	product if you attack the firm for its product design,
11	and that can lead to large losses. So, in that
12	situation, I might want to give more weight to the
13	firm's efficiency claim for fear of causing a large dead
14	weight loss than in other situations.
15	Let me give you a second example. Let's talk
16	about the Areeda-Turner test for predatory pricing.
17	Basically that's a test that says if your price is below
18	some measure of cost, unless it is, I am not going to
19	worry. So, the implicit idea is that if you see price
20	below some measure of cost, that is a big enough
21	deviation from what we usually think of as profit
22	maximization that there is something fishy, and the
23	reason they chose price below a cost as the predation
24	standard rather than above-cost pricing is because they
25	were very worried about chilling competition that drives

62

1	price down. So, they were specifically worried in the
2	context of predatory pricing of erroneous behavior, of
3	erroneous condemnation, which then would chill
4	competition, very specialized to the predatory act,
5	okay? Not a general principle.
6	If you misread it, what the Areeda-Turner test
7	said, you could misread it as follows: You could say,
8	ah, price below cost, the firm is not profit-maximizing.
9	If the firm is not profit-maximizing, there is something
10	fishy going on, that's a violation. Now, if you read it
11	that way, which would be an incorrect way to read it,
12	that is saying that any deviation from profit-maximizing
13	behavior would be an antitrust violation, and that would
14	make me nervous. That is the concern I have about a
15	profit sacrifice test, because if you are not maximizing
16	profit, your failure to maximize profit is a sacrifice
17	of profit, and I do not want get into the situation in
18	which I claim that firms must be maximizing profit as I
19	see it or it is an antitrust violation.
20	Now, you could remedy that. You could say, all
21	right, I will not say you have to be exactly, I will
22	give you a margin of error, but that is really my
23	concern with the test, not that it is not in some
24	situations useful, but that I would not want to enshrine
25	it as a general principle.

63

1	Let me now turn to a market definition, which I
2	was going to say I am sure you have heard a lot about
3	and probably do not want to hear more about, and I was
4	glad, though, that Herb did talk about it, because at
5	least it confirmed in my view that someone else thinks
6	it deserves still more thought.
7	But for the antitrust laws, industrial
8	organization economists would not pay all that much
9	attention as to how you define a market. I think it is
10	fair to say that the reason it receives so much
11	attention is precisely because it has been used as a
12	screen or a requirement under our antitrust laws, and
13	that is what gives it prominence.
14	Now, my own view is it is a very good but crude
15	tool, and it has this intuition behind it that if there
16	are lots of rivals, do not worry, and I think that is a
17	very good common sense rule, but it is hard to apply in
18	the non-merger context, and there have been attempts to
19	apply it in the nonmerger context that at least
20	sometimes strike me as odd. So, what I have seen in a
21	number of cases is the application of the Merger
22	Guidelines, and it goes something like this.
23	Define a market so that a hypothetical
24	monopolist of those products can raise price 5 percent
25	above the competitive level, the competitive price. In

64

1	a merger case, it would be the current price, but since
2	this is not a merger case, we say above the competitive
3	price. Well, that is a very well-posed question, and in
4	order to answer it, I have to say, "Okay, well, what is
5	that competitive price?" Well, I do not know what the
6	competitive price is. If I knew what the competitive
7	price is, I could look and see, is the current price
8	above the competitive price? And if it was, I would
9	say, yes, it is above. I would not then have to define
10	a market, take market shares and say, "Ah, you know,
11	based on all this analysis, you know, 10 is above 5 when
12	I started, and still, the market shares are so high now,
13	I conclude 10 is above 5." So, there is a circularity
14	to it that I find a little troubling. So, you cannot
15	really directly answer the question, because the
16	competitive price is not available. So, then, what do
17	you do?
18	I think there are several alternatives. Herb
19	mentioned some of them, and none of them I would say are
20	completely satisfying. One alternative is to ignore the
21	problem and simply say, "I am not going to use the
22	competitive price, I am going to use the existing
23	price," and Herb sort of indicated that this leads to
24	the well-known what is called Cellophane fallacy in
25	which it is possible that you will not find any market

65

1	power at the current price, but you do have market power
2	because you have already raised the price above the
3	competitive level. That is one thing you can do, with
4	its problems.
5	The second thing you can do is you could say,
6	"Okay, let me ask the following: Is price above
7	marginal cost where marginal cost I will use as my proxy
8	for the competitive price?" Of course, that kind of
9	replaces one question with the other, what is marginal
10	cost? Well, maybe you can go out and try and measure
11	marginal cost. If you had access to firm information,
12	you could try and sift through accounting information,
13	you could econometrically try and estimate a cost curve.
14	It's difficult, okay?
15	Moreover, suppose you do find prices above
16	marginal cost. You have to face the realistic
17	possibility that most markets are not perfectly
18	competitive. In most markets, price will not equal
19	marginal cost; therefore, and what you presumably must
20	mean, is that price deviates a lot from marginal cost if
21	you are worried about such a deviation. The amount of
22	deviation, the deviating a lot from marginal cost, has
23	actually never been articulated that I have seen in a
24	quantitative way.
25	Okay, suppose you do not like marginal cost. Is

66

1	there something else you can do? Well, I think there
2	is. Another thing you can do is you can ask, "Is the
3	rate of return a firm is earning above the competitive
4	rate?" We know what the competitive rate of return is.
5	Is the rate of return the firm is earning above the
6	competitive rate? This can be a difficult accounting
7	exercise, or better put, this can be a difficult
8	economic exercise using accounting data. You would have
9	to ask over what period of time, how does my answer
10	change depending on risk, not easy to do necessarily,
11	likely to create a lot of controversy.
12	Finally, you could estimate the demand curve
13	facing the firm, and the benefit of estimating the
14	demand curve is you could determine the elasticity and
15	the cross-elasticities that the firm is facing, and that
16	I think gives you useful information about certain types
17	of competition. It raises issues over what time period
18	you estimate the demand curve, but all of those strike
19	me as things you might be forced to do because the
20	question you are trying to answer is really a very
21	difficult one and does not admit a very simple answer.
22	Since time is running a little short, and I know
23	we want to leave time for discussion, I will just
24	mention two topics, and maybe we can come back to it in
25	the panel discussion. What is the distinction between

67

1	market power and monopoly power? And then second, what
2	does it mean to have individual market power in an
3	oligopoly?
4	Well, let me just conclude, screens can help but
5	can also become a danger if one loses sight of the
6	ultimate goal of maximizing consumer welfare. You
7	should create rules to fit the act, paying special
8	attention to how the rules applied to this particular
9	act will chill competition.
10	The profit sacrifice test worries me a bit for
11	the reasons I have explained, and I would use it
12	sometimes but certainly not regard it as a general
13	principle. With regard to market definition, I still
14	think it is a useful discipline, especially when you get
15	to court, for judges and juries to go through, because
16	it helps structure the analysis, but there is an
17	inherent lack of precision in its application in Section
18	2 cases that might be worrisome, and there may well be
19	cases where its use could be misleading.
20	Thank you.
21	(Applause.)
22	MR. BLUMENTHAL: Well, thank you to the panel
23	for very, very interesting perspectives. We have about
24	20 minutes for panel discussion, and because of airline
25	schedules, we are going to have to end promptly at the

68

1	appointed hour of 4:00, but I guess let me turn to the
2	first two speakers, who are agency speakers, and ask
3	whether you have any comments you would care to offer on
4	the thoughts from either of the two professors.
5	CHAIRMAN MAJORAS: Wow, I have a lot that I
6	could comment on. I thank both of them, because those
7	were very thoughtful presentations, as we expected, and
8	I very much appreciate that.
9	One thing that I was curious about is Herb's
10	advice that the agencies should step in a bit more in
11	looking at restraints on innovation, and I was
12	wondering, Herb, if you had any particular hypothetical
13	or context involved or if you can be even more critical
14	of what we have not done in the past, if that is easier
15	for you, in terms of where we could be looking for such
16	a thing, because it is true that we constantly talk
17	about promoting innovation and how important that is to
18	our work and how important it is that we not inhibit it,
19	but I find not only in enforcement but in forming policy
20	and explaining to courts, it is very difficult for
21	people to get their arms around it, because it is such
22	an amorphous concept, and I am not sure we are very good
23	at it.
24	DR. HOVENKAMP: It is an amorphous concept. In
25	part, though -- and I think this is an important

69

1	principle -- it is the one place where the IP laws and
2	the antitrust laws tell the same story. We always talk
3	about this tension between innovation and competition.
4	It is hard to defend restraints on innovation when they
5	are defined properly, and the fact is that the
6	enforcement record has not given us cause for much hope.
7	I can think of -- right now, off the top of my head, I
8	can think of three situations.
9	One is the air pollution cartel cases of the
10	1970s. I suspect some of you have not been around long
11	enough to remember those, but those were challenges by
12	farmers, class action, brought by agricultural groups to
13	an alleged agreement among the major automobile
14	producers for restraining the development of engines
15	that produce less pollution, and what they wanted to
16	collect as damages was harm to their crops that accrued
17	as a result.
18	Well, just stating it tells you why that
19	antitrust case is not going to go anywhere, right? You
20	would have to determine what was lost, you would have to
21	determine what the impact of these more fuel-efficient
22	engines would have been, how the farmers might have
23	evaded it. It would be totally impossible for private
24	plaintiffs to collect damages in such a case.
25	Another one is the series of cases, a fairly

70

1	large number of cases, that made it to the circuits
2	against the cigarette companies in the late 1990s
3	alleging various cartels to refrain from developing
4	healthier cigarettes. To the best of my knowledge,
5	every single one of those cases was dismissed, at least
6	the ones that went to the circuits, were dismissed on
7	grounds of standing, causation, provable injury, maybe a
8	couple of other reasons, but they all had to do with
9	ability to prove injury.
10	Then, of course, there is the conclusion in
11	Judge Jackson's Microsoft opinions, affirmed by the D.C.
12	Circuit, that Microsoft's attempts to force Intel not to
13	develop a Java-enabled chip, a chip that could have
14	spoken multiple-processor languages, something that Bill
15	Gates feared at the time, was unlawful, and that
16	complaint allegation has actually been included in a few
17	of the indirect purchaser cases against Microsoft, and
18	to the best of my knowledge, not a single plaintiff has
19	ever collected a dollar for those failures of
20	innovation.
21	So, one of the problems is that restraints on
22	innovation are very likely extraordinarily harmful if
23	they occur. If so much of our economic growth comes
24	from innovation, then restraints on innovation could, in
25	fact, be very harmful.

71

1	Number two, because of the nature of innovation,
2	because you can never predict where it will go, how much
3	value it will produce, how much people would be injured
4	by such a restraint, it is not good grist for private
5	damages actions under Section 4 of the Clayton Act.
6	I do not know if that is a satisfactory answer,
7	but it is just an area where I would like to see more
8	attention paid.
9	CHAIRMAN MAJORAS: Well, it does help, because
10	the other difficulty we have, not only in antitrust but
11	in forming just good economic policy, is how you balance
12	innovation incentives generally. I mean, this is the
13	classic case that I was just discussing earlier with
14	some folks in the consumer protection arena about
15	pharmaceutical companies and their incentives to
16	innovate and how this group of folks just came back from
17	Europe and said, "Okay, now we need to push in the
18	United States what other countries are doing," and I
19	said, "Well, what is going to happen to the incentives
20	to innovate if every country in the world put caps on
21	pharmaceutical prices?" But because the harm to
22	innovation from that is so hard to measure in the policy
23	arena, people do not necessarily, I think, take it into
24	account.
25	MR. BARNETT: Well, I guess a couple of

72

1	reactions. First, I am extraordinarily pleased with the
2	contributions that the two of you made. They not only
3	frame some interesting issues and pose some hard
4	questions, but in some instances, also provided
5	suggested answers or even policy directions, which is a
6	great way to start off the hearings, and we very much
7	appreciate hearing it.
8	I think, Herb, it is interesting to hear your
9	views on refusals to deal, whether we should be involved
10	in that or not, and I guess I was going to ask you, you
11	have touched on a little bit the issue of innovation,
12	and given its importance, does that mean we should be
13	more or less aggressive in intervening in the sense that
14	is there a greater risk of deterring innovation if you
15	get involved in some of these areas? I do not know the
16	answer to that. I just raise it as a thought.
17	Let me, if I can, pause and ask a -- well, I
18	will ask you one question, though, before I get to that.
19	I have always been of slightly two minds. I mean, I
20	quoted Capitalism, Socialism and Democracy, the Joseph
21	Schumpeter discussion of monopolies, and I have been
22	taken with this gales of creative destruction imagery.
23	Now, there is a lot in there that sort of apologizes for
24	and/or justifies monopolies as good in and of
25	themselves, and I guess some of hearing your remarks and

73

1	all caused me to want to clarify that I do not agree
2	with everything that is in that book in terms of
3	monopoly, but I do agree with the importance of
4	innovation and not losing sight of that.
5	But talking about this, you just made a comment
6	about remedies and a distinction, I think, between
7	damages and the Government. I would be interested to
8	hear more of your views on that point, if you think that
9	would be a good idea, to draw such a distinction.
10	DR. HOVENKAMP: The distinction between?
11	MR. BARNETT: Different remedies for
12	governmental ver -- well, for I would say maybe
13	injunctive versus damages type relief.
14	DR. HOVENKAMP: It has always been my opinion,
15	which the Supreme Court has rejected in California
16	versus American Stores, that structural remedies should
17	be the prerogative of the Government. That is not the
18	law, let's make sure everybody's clear about that. In
19	California versus American Stores, where the plaintiff
20	was the Federal Trade Commission, so it's a private -ï½
21	I'm sorry, was the State -- was the Attorney General of
22	the State of California, so it was a private party in
23	this rather unusual way we treat states attorneys
24	general in antitrust, but I believe structural remedies
25	ought to be something that only the Government ought to

74

1	do, because they have such extraordinary spillovers.
2	I mean, we have some confidence that the federal
3	antitrust agencies, because of their diversity, the
4	diversity of the industries that they represent, and
5	because of the high quality of the people that run them,
6	they are not captured industries, by and large, that we
7	may not have that level of confidence about
8	industry-specific agencies.
9	The one thing we know about private plaintiffs
10	is that they are always captured, right? Private
11	lawyers serve their clients. Their clients are
12	completely self-interested, and as a result, they do not
13	take overall effects into account, and as a result, my
14	own view is that structural relief, such as divestiture,
15	should simply be denied to private plaintiffs.
16	Is that an answer to your question?
17	MR. BARNETT: I just was interested in your
18	perspectives, so thanks.
19	MR. BLUMENTHAL: Herb, Dennis had spoken after
20	you did, so I guess I would ask whether you had any
21	rejoinder to any of Dennis' comments.
22	DR. HOVENKAMP: No, and I mean, Dennis knows a
23	million times more about this market definition stuff
24	than I do, and so I am very elated that he actually
25	agreed with me about most of it. I think you can

75

1	develop a few presumptive rules for dealing with market
2	power issues. For example, where the defendant, the
3	firm under investigation, faces competition from other
4	firms that use the same technology and apparently have
5	the same cost structure, then I think the inference of
6	competition, competitive pricing, is higher than in a
7	case like Cellophane. I think what makes the Cellophane
8	case so extraordinary is that the defendants, DuPont,
9	plus Sylvania as its licensee, produced Cellophane, but
10	the stuff that the Government ended up throwing into the
11	market was brown wrapping paper, tin foil, glassine -ï½
12	I'm not even sure I know what glassine is -- but these
13	were things that used different raw materials, they
14	almost certainly used different technologies for
15	producing them, and simply to conclude that such things
16	are in the same market simply because of high observed
17	cross-elasticity of demand is a very serious error.
18	It's an error courts continue to make.
19	There are several cases, for example, that have
20	concluded that rental videos, films shown in movie
21	theaters, and films shown on television are all in one
22	relevant market simply because they make the obvious
23	observation that if you look at any particular person,
24	sometimes she goes and rents a video, sometimes she goes
25	to a theater, sometimes she watches a movie on

76

1	television, without asking the question whether any one
2	of those technologies is sufficient to hold the other
3	ones to cost.
4	I would rather hear Dennis talk about this.
5	DR. CARLTON: I agree, but I will talk about
6	something about innovation, if that is all right,
7	because I think that is an important topic.
8	It is absolutely right, our standard of living
9	has increased because of technological change. The
10	question about what the implication of that is for
11	antitrust I think is easy to state. It is very hard to
12	implement. Part of the reason it is hard to implement
13	is because what we know about innovation and the market
14	structure that induces innovation is a lot less than
15	what we know about concentration and the effect of
16	concentration on raising price.
17	The evidence, I think most economists would say,
18	that competition is good for innovation. On the other
19	hand, if you pushed them and said, "Do you think four to
20	three makes a big difference when firms are innovating,"
21	you know, it is pretty hard to find evidence, I think,
22	and moreover, innovating in what? The idea of, for
23	example, an innovation market I think is not
24	particularly helpful except maybe in rare exceptions
25	like drugs where you can redefine it to be not an

77

1	innovation market, but rather, a future product market.
2	In drugs, you can look at the pipeline, and you know
3	what is coming out. So, it is not really an innovation
4	market as much as it is a prediction of future products.
5	It is much harder when you are looking at just
6	firms doing R&D in a general area to say, "What are they
7	doing R&D on? Is it even on similar products?" And if
8	you ever go back and do an experiment, like if you look
9	at major product introductions, who did them, sometimes
10	the people doing them are often outside the industry,
11	and no one predicted they would come in. So, I think
12	the difficulty in this area, when antitrust tries to
13	deal with it, is the difficulty that you have -- that
14	anyone would have -- in predicting who is going to be
15	the innovator. That is why I think it is very hard.
16	The other reason why I think it is very hard is
17	innovation markets inherently involve intellectual
18	property. People want to protect their intellectual
19	property. There are a lot -- if you look at, you know,
20	just -- there are some very interesting articles on the
21	Internet, who has exclusivity rights on certain types of
22	information that is generated. Restrictions, vertical
23	restrictions on intellectual property, who can use them,
24	who cannot use them, are very widespread, and therefore,
25	if you do have a big success, that will complicate some

78

1	antitrust enforcer's life, because they will say, "Well,
2	look, a big hit in the vertical restrictions, I am a
3	little worried," but you have got to go back and step
4	back and say, "Well, the whole idea of innovating was to
5	get the returns from this intellectual property, and
6	they would not have done it if they could not put the
7	restrictions on it."
8	So, I think it is a very difficult area. I
9	think it is made more difficult by our intellectual
10	property laws, and the problem that I really see arising
11	is that because so many patents are given for products
12	that may not be all that valuable, you create the
13	problem that the property right in a truly valuable
14	innovation is not that valuable, because now you have to
15	get cross-licenses from lots of other people, and that
16	is I think creating a lot of the antitrust problems, and
17	maybe the best way to fix it is not just through
18	antitrust but through reforming IP.
19	CHAIRMAN MAJORAS: As long as we are on the
20	subject of innovation, in the two minutes we have left,
21	does either of you have an opinion on the recent
22	phenomenon that we are seeing in the patent area known
23	as patent trolls, people who go out and buy up what some
24	people might term useless patents -- I don't know if you
25	can go that far -- and some of those people doing it, in

79

1	fact, as I understand it are plaintiff's lawyers, and
2	then wait until somebody creates an invention, and that
3	bundle of patents might look, you know, like it could
4	have been infringed, maybe it is, maybe it isn't, and
5	then pounce and say, "Okay, now you owe me really high
6	royalties."
7	And there are obviously many different
8	variations on this theme, you know, those that have zero
9	interest in ever making anything or innovating I think
10	are the ones that people refer to as patent trolls. Do
11	you have any comment on whether that is injuring
12	competition or something that any of us should be
13	worried about?
14	DR. HOVENKAMP: Well, my reaction to that is
15	that whether it's good or bad, it is fundamentally not
16	an antitrust problem. I mean, in order to be an
17	antitrust problem, it either has to monopolize the
18	market or create a dangerous probability of doing so or
19	else it has to be an agreement that restrains trade, and
20	simply surprising people with a patent and claiming high
21	royalties in and of itself does not do that.
22	Now, that does not mean if it were not used in
23	conjunction with some other practice, it couldn't sum up
24	to an antitrust violation.
25	DR. CARLTON: I would just add, I basically

80

1	agree with what Herb said. I think that it is a problem
2	for reform of the patent laws, and there are at least
3	two thoughts I have. The first is that one of the
4	things that gives a troll great power is the ability to
5	get an exclusion right, and you might want to encourage
6	the courts to give the surprised infringer time to
7	invent around the patent if that is possible, and that
8	can get rid of sometimes a lot of the pressure.
9	The other question you can ask is, are they ever
10	intending to use this themselves and implement it, and
11	that may influence how you treat them.
12	Having said that, you know, an economist always
13	has two sides to a story. A patent -- you may call them
14	a patent troll, but the flip side is I'm an inventor,
15	maybe I want to give my invention to someone else to
16	figure out, you know, all the licensing problems. So, I
17	mean, there is a flip side.
18	CHAIRMAN MAJORAS: Yes.
19	MR. BARNETT: Can I just ask quickly if based on
20	your remarks, Dennis, if you would be willing to get
21	into a locked room with Greg Werden for a session with a
22	tape recorder and we could solve nine out of every ten
23	Section 2 problems.
24	DR. CARLTON: I would be delighted.
25	MR. BLUMENTHAL: Dennis, if I could I guess ask

81

1	one last question in the two minutes we have, you had
2	thrown out a teaser at the end about the distinction
3	between market power and monopoly power, an issue that
4	was of some interest to the staff, although we had not
5	quite posed it, and I might add, it was something I
6	recalled George Stigler having denied as a distinction
7	in the Data General case about 20 years ago, although
8	the Ninth Circuit disagreed with that view. So, perhaps
9	there is a distinction after all.
10	Did you have in mind something more than just
11	quantum? Is it a character of the difference or -ï½
12	DR. CARLTON: Yes, actually, it is in my
13	textbook, so I know it very well. One definition of
14	market power could simply be that price is in excess of
15	marginal cost. That is a logical definition. Whether
16	you want to use it in a court, you know, I think that is
17	a different question. You might want to define it by
18	how much before you say it is something that triggers
19	some action, but you can distinguish market power from
20	monopoly power by saying monopoly power is not just
21	price in excess of marginal cost, but also, profits in
22	excess of the competitive rate of return. So, that is
23	at least a logical distinction.
24	I should say in the third edition of my book, I
25	dropped that paragraph, because no one seemed to pay

82

1	attention to it. Now, I change a lot of things when I
2	revise my book, and most people never comment, but on
3	this, I did get some comments, and people said, "We
4	thought it was useful," even though I had never seen
5	anyone -- I was unaware anyone had cited it, and I think
6	it is useful, so I have put it back in. Now, so, that
7	is an answer.
8	Now, whether that should be what courts use when
9	they are deciding on an exclusionary act is a slightly
10	different question, because I think courts really should
11	be asking a slightly different question, which is, let
12	us suppose you are talking about exclusive dealing. I
13	don't really care about what the rate of return is, I
14	don't think. I think what a court is saying is if you
15	are engaged in an act and there are other
16	characteristics of the market that make that act have a
17	greater effect than if you didn't, say, have as large a
18	market share, then I am going to be worried about that,
19	and that really does not have to do with rates of
20	return.
21	So, you know, I can see why courts could be
22	asking are the characteristics of the market that make
23	the exclusionary act under discussion much more a
24	competitive concern than not, whether that -- I don't
25	know that that necessarily corresponds to the

83

1	distinction I gave you. The distinction I gave you does
2	strike me as a sensible economic distinction.
3	MR. BLUMENTHAL: Well, thank you, and with that,
4	we are out of time. Let me ask you to join me in
5	thanking the panel for a great presentation.
6	(Applause.)
7	MR. BLUMENTHAL: We look forward to receiving
8	your comments in the months ahead. We hope to see you
9	Thursday and in the hearings to come. Thank you all.
10	(Whereupon, at 4:00 p.m., the hearing was
11	concluded.)
12
13
14
15
16
17
18
19
20
21
22
23
24
25

84

1	C E R T I F I C A T I O N O F R E P O R T E R
2	DOCKET/FILE NUMBER: P062106
3	CASE TITLE: SECTION 2 HEARING
4	DATE: JUNE 20, 2006
5
6	I HEREBY CERTIFY that the transcript contained
7	herein is a full and accurate transcript of the notes
8	taken by me at the hearing on the above cause before the
9	FEDERAL TRADE COMMISSION to the best of my knowledge and
10	belief.
11
12	DATED: 7/5/06
13
14
15
16	SUSANNE BERGLING, RMR-CLR
17
18	C E R T I F I C A T I O N O F P R O O F R E A D E R
19
20	I HEREBY CERTIFY that I proofread the transcript
21	for accuracy in spelling, hyphenation, punctuation and
22	format.
23
24
25	DIANE QUADE

Updated January 2, 2024