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	UNDERSTANDING SINGLE-FIRM BEHAVIOR:
9	CONDUCT AS RELATED TO COMPETITION
10	TUESDAY, MAY 8, 2007
11
12
13
14
15	HELD AT:
16	UNITED STATES FEDERAL TRADE COMMISSION
17	HEADQUARTERS BUILDING, ROOM 432
18	600 PENNSYLVANIA AVENUE, N.W.
19	WASHINGTON, D.C.
20	9:00 A.M. TO 1:00 P.M.
21
22
23
24	Reported and transcribed by:
25	Brenda Smonskey

For The Record, Inc.

2

1	MODERATORS:
2	DEBORAH PLATT MAJORAS
3	Chairman
4	Federal Trade Commission
5	and
6	TOM BARNETT
7	Assistant Attorney General
8	Antitrust Division, U.S. Department of Justice
9
10	PANELISTS:
11
12	Susan Creighton
13	Jeff Eisenach
14	Tim Muris
15	Bob Pitofsky
16	Doug Melamed
17	Jim Rill
18	Charles F. (Rick) Rule
19	Greg Sidak
20
21
22
23
24
25

For The Record, Inc.

3

1	C O N T E N T S
2
3	Introduction........................................................................................................................................................... 4
4
5	Moderated Discussion......................................................................................................................................... 13
6
7	Conclusion.........................................................................................................................................................162
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25

For The Record, Inc.

4

1	P R O C E E D I N G S
2	- - - - -
3	CHAIRMAN MAJORAS: Good morning, everyone.
4	Welcome to this final wrap-up panel of the
5	hearings that we, the FTC, together with the DOJ
6	Antitrust Division have been holding over the course
7	of almost the past year.
8	I'm delighted to be here today to moderate
9	this final session with my very good friend and
10	colleague, Tom Barnett, Assistant Attorney General
11	for the Antitrust Division.
12	So I thank you all for being here. I also
13	thank our panelists for taking the time away to be
14	with us this morning.
15	Before I get started, I should ask all of
16	you just as a courtesy that if you have anything on
17	that rings or otherwise makes noise, if you could
18	turn off at least that part of it. We would
19	appreciate it.
20	We ask that you not make comments, at least
21	not above your breath, during the session or yell
22	out questions from the audience, please.
23	I want to start this morning by thanking the
24	staff from the FTC and from the Department of
25	Justice Antitrust Division for their incredible work

For The Record, Inc.

5

1	over the course of the last year in putting together
2	27 Section 2 hearing sessions over the course of the
3	year.
4	These things have gotten to the point where
5	I think they go so well and so smoothly that you
6	forget how much work is going on behind the scenes.
7	But I see Pat here and Bill Cohen and Gail.
8	They can tell you all the work that has gone on
9	behind the scenes. We are truly grateful for their
10	contributions.
11	I also want to express my appreciation to
12	the 130 panelists we have had over the course of
13	these sessions. They have made an incredible
14	contribution to these hearings.
15	I wanted to convene the hearings because it
16	seemed to me that the debate over where we should be
17	drawing the permissible lines for conduct by firms
18	with market power needed something of a boost.
19	I was a little bit worried that it might be
20	getting stuck. It seemed like we were drawing
21	lines, to be sure, but we were drawing more like
22	battle lines around certain tests or certain
23	arguments.
24	And our hope was that through these hearings
25	we could identify or highlight areas certainly of

For The Record, Inc.

6

1	broad consensus in enforcement against single-firm
2	conduct and then also draw out the areas that
3	require further rigorous analysis and guidance.
4	So starting with the opening session on June
5	20th, we have held hearings on a wide range of
6	conduct, from predatory pricing to exclusive dealing
7	to bundled and loyalty rebates and the whole
8	spectrum, as well as sessions on monopoly power,
9	remedies, market definition.
10	We also held a session on empirical
11	research, during which we heard about the research
12	that exists on Section 2 areas as well as areas
13	where further research would be helpful.
14	We held a session on international
15	perspectives, where we heard from a number of
16	foreign competition agency officials as well as
17	practitioners and academics in the field.
18	We held a session on business history in
19	which we examined some of the more important
20	monopolization cases of the past century.
21	We had a session on business strategy so we
22	could learn more about what business schools are
23	teaching future business leaders and executives,
24	what they are teaching them and how that could
25	ultimately impact competition and conduct.

For The Record, Inc.

7

1	I had hoped, as you all know, from the very
2	beginning that we could get a fair amount of input
3	from the business community so we could actually
4	really think about certain types of conduct, why
5	folks are engaged in it.
6	And I was pleased that we were able to hold
7	two out of town hearings this time, get outside the
8	Beltway. We held a hearing in Berkeley, California
9	and Chicago, Illinois, which I was very pleased
10	about.
11	Through all this, we have endeavored to
12	select panelists that could provide a wide diversity
13	for us of viewpoints on these important topics.
14	So here we are. We are at the last
15	roundtable discussion. We held another almost last
16	roundtable discussion last week. So here we are
17	today.
18	We will ask our panelists to comment on a
19	wide range of issues.
20	We will not have speaker presentations
21	today. We will get directly into questions from our
22	panelists, which we thought would be a richer forum
23	to take advantage of the great wisdom and experience
24	of this distinguished panel.
25	With that, I will tell you -- I think you

For The Record, Inc.

8

1	probably know who they are, but I'm going to tell
2	you.
3	I will start with four of the panelists who
4	I will introduce. Tom will introduce the others.
5	I will introduce all the former FTC folks,
6	and Tom will introduce the former DOJ folks plus
7	one.
8	I was thinking what we might do is have them
9	duke it out. Maybe we can solve all the problems.
10	We have a new form of clearance agreement of some
11	sort.
12	So to my far right is Susan Creighton.
13	Susan is a partner at the Wilson Sonsini firm after
14	having served here as the director of the FTC's
15	Bureau of Competition, and it has been my great
16	pleasure to work with Susan.
17	Susan is quite well known in this area of
18	Section 2 law and in particular of late in the area
19	of cheap exclusion.
20	So we will look forward to her comments
21	today.
22	Jeff Eisenach is the chairman of Criterion
23	Economics and adjunct professor at the George Mason
24	School of Law.
25	He has served in senior policy positions at

For The Record, Inc.

9

1	the FTC and also at the Office of Management and
2	Budget. He was a cofounder of the Progress and
3	Freedom Foundation. And he is also someone willing
4	to play golf with me.
5	Tim Muris -- I can't introduce Doug because
6	he used to be at DOJ. Sorry, Doug. So did I.
7	Tim Muris will be here. We knew that he
8	would have to be a little bit late today. I will go
9	ahead and introduce him anyway.
10	He is a George Mason University Foundation
11	professor of law, of counsel at O'Melveny & Myers
12	and a co-chair of that firm's antitrust practice.
13	He also, of course, served as chairman of
14	the FTC until 2004. And in his previous life in the
15	'80s was director both of the Bureau of Competition
16	and the Bureau of Consumer Protection.
17	Tim will be here later this morning.
18	Finally, to Tom's left we have Bob Pitofsky,
19	the Joseph and Madeline Sheehy professor in
20	antitrust and trade regulation law at Georgetown
21	University Law Center, where he formerly served as
22	dean.
23	He is also counsel at Arnold & Porter and
24	formerly chairman of the FTC, prior to Tim Muris, of
25	course.

For The Record, Inc.

10

1	We have a lot for which we are grateful to
2	Bob, but one I think is that Bob really
3	reinvigorated this concept of hearings at the FTC
4	during his tenure.
5	That, of course, is the tremendous legacy
6	that brings us here today. So thank you.
7	Now I would like to turn it over to Tom
8	Barnett.
9	MR. BARNETT: Thank you, Debbie.
10	I also would like to underscore my thanks to
11	the staff, who have worked very hard. And in some
12	sense it seems like yesterday, it was almost a year
13	ago when Debbie and I stood up, I think over in that
14	corner of the room, along with a few other people
15	and helped launch these hearings.
16	But to the staff I have a feeling that may
17	seem like about 10 years ago, given the number of
18	sessions and panelists and issues.
19	As we were working through the preparation
20	for the hearing today, one of the things that really
21	struck me is the range of issues and the depth of
22	thought that has gone into preparing each and every
23	one of these sessions.
24	I know it is a tremendous amount of time and
25	effort. But I also agree with Debbie that this is

For The Record, Inc.

11

1	an extraordinarily important topic.
2	I have long viewed this, along with I think
3	Judge Posner who said this as well, really to be the
4	most challenging area of antitrust enforcement in
5	many ways, because large dominant firms can impose
6	very significant costs in terms of consumer welfare.
7	It is also the most difficult area in which
8	to avoid making mistakes as a government enforcer,
9	both in terms of condemning conduct that actually
10	can be beneficial, and even if you find a problem,
11	in crafting remedies that will fix the problem
12	without doing more harm than good.
13	And while I do agree that there are many
14	areas of consensus at least within the United States
15	in this area -- and I think the hearings have done a
16	good job of highlighting some of those things -- I
17	also think there are some very important issues that
18	remain open.
19	And I'm optimistic with the wide range of
20	experience and talent that we have had, the benefit
21	of economists, lawyers, business people, academics,
22	and certainly with the degree of experience and
23	wisdom we have at the panel here today, I expect we
24	will have resolved all of this by 1:00 today.
25	With that, I do want to move toward the

For The Record, Inc.

12

1	discussion, which we have a lot to cover in a lot of
2	very -- it seems like a long time, but I have a
3	feeling it will go quickly.
4	So let me just move to the introductions.
5	I will start off with introducing Doug
6	Melamed, who is a partner and co-chair of
7	WilmerHale's -- do you say WilmerHale?
8	MR. MELAMED: I am supposed to.
9	MR. BARNETT: -- antitrust and competition
10	department and former Deputy Assistant Attorney of
11	the Department of Justice's Antitrust Division,
12	where he had a little bit of experience in some
13	Section 2 matters.
14	And then over to my left is Jim Rill, who
15	I'm sure everyone knows, who is a partner at Howrey
16	and the former Assistant Attorney General of the
17	Antitrust Division.
18	To his left is Rick Rule, who is a partner
19	at Cadwalader, Wickersham & Taft and also a former
20	Assistant Attorney General at the Antitrust
21	Division.
22	And down at the left is Greg Sidak, who is a
23	visiting professor of law at Georgetown University
24	Law Center and a founder of Criterion Economics.
25	He served as the deputy general counsel of

For The Record, Inc.

13

1	the FCC and senior counsel and economist to the
2	Council of Economic Advisors over in the executive
3	branch.
4	So welcome to everyone. And with that I say
5	why don't we get to it.
6	In terms of format, Debbie and I thought we
7	would basically play tag team in terms of who will
8	lead off each topic, with the idea, however, that
9	each of us will jump in as seems useful.
10	And we are going to start off with the first
11	topic being general standards and issues.
12	I will ask the very first question in the
13	broadest possible form, which is I would like to ask
14	which one or two issues -- and I would ask no more
15	than two to keep it short -- that you think are the
16	biggest problems or concerns facing antitrust
17	enforcement today in the area of Section 2 that we
18	should try to address in the report that comes out
19	of this.
20	To start off, why don't I ask Jim Rill to
21	jump in.
22	MR. RILL: Thank you, Tom.
23	Let me say it is an extraordinary honor to
24	be here on this panel of august personages and to be
25	invited to participate.

For The Record, Inc.

14

1	I think one issue stands out in a claim
2	being addressed in the report, and I emphasize
3	report, not necessarily guidelines, but an
4	analytical report -- hopefully with some sense of
5	conclusion and advocacy -- and that is the area of
6	bundled pricing and loyalty discounts.
7	The area has abounded in some confusion ever
8	since the LePage's-3M decision. There are several
9	court decisions on the way up that may add clarity
10	or possibly further confusion to the issue.
11	But trying to provide advice in that
12	particular area is daunting. I think that there are
13	a number of solutions out there, or at least
14	potential solutions out there as we get into more
15	the merits of the discussion today.
16	But I think those particular areas are ones
17	that really stand out above the others in looking
18	for a detailed analysis and what I would propose to
19	be a report, which I earnestly hope is forthcoming
20	as a results of these hearings.
21	MR. BARNETT: Thank you.
22	Bob, would you like to give us your
23	perspective?
24	MR. PITOFSKY: Thank you.
25	It is very similar to Jim.

For The Record, Inc.

15

1	We talked about whether we could reach
2	consensus. I suspect the best chance we have of
3	reaching consensus is on the issue of what is the
4	most pressing set of issues facing antitrust, and I
5	think it is defining exclusionary behavior under
6	Section 2.
7	I think it is a set of issues that is most
8	confusing, hard to predict, hard to counsel, hard
9	for judges to deal with.
10	Some people will hold out for the Robinson
11	Patman Act, but I don't quite think that is really
12	the toughest set of questions.
13	And as we will discuss today, what sort of
14	rule should we build on? Is it the balancing test
15	that was unanimously adopted by the Court of Appeals
16	in Microsoft and echoed I think in Aspen, or these
17	unitary tests. We all know the balancing test has
18	its flaws in terms of unpredictability and
19	difficulty in implementing in the context of a legal
20	proceeding.
21	But should we look for a unitary test, which
22	people understandably and with my admiration have
23	tried to come up with -- sacrifice of profits,
24	driving out a less efficient competitor and so
25	forth.

For The Record, Inc.

16

1	I will give away my bottom line right now.
2	I think the unitary tests, much as I admire the
3	creativity of them, don't work, do more harm than
4	good. And therefore, I would stick with the
5	balancing test.
6	But I think that's what a lot of our
7	discussion this morning should be directed toward.
8	MR. BARNETT: Doug?
9	MR. MELAMED: I think the most important
10	thing that can come out of these hearings would be
11	an explicit clarification or articulation of the
12	purpose of rules about exclusionary conduct.
13	I had occasion before coming today to look
14	through some of the summaries of the hearings that
15	you have held thus far. I haven't read all the
16	testimony. But I did look at the summaries.
17	I had the impression that it was like an
18	unbounded exercise for a public policy class at the
19	Kennedy School.
20	There are all sorts of people with all sorts
21	of views about how to address tying, exclusive
22	dealing, predatory pricing, whatever the topic is,
23	unstated often in the dialogue, and I think often
24	explaining the disagreements among the parties, were
25	differences in assumptions about the purpose of

For The Record, Inc.

17

1	antitrust.
2	Is it consumer welfare? Is it total
3	welfare? Is it dynamic analysis? Is it static
4	analysis? And so forth.
5	This problem doesn't arise in cases of
6	collusion, because in these cases, I think both the
7	normative and the analytical converge on the
8	understanding that the issue is, does the
9	arrangement increase or decrease the output of the
10	parties to the agreement.
11	In exclusion cases, we are often dealing
12	with a trade-off between the efficiency benefits to
13	the defendant and the exclusionary impact on rivals.
14	And I think we don't have a clear understanding of
15	what the antitrust objective is dealing with that
16	trade-off.
17	My own view is that none of the sort of
18	economic factors mentioned above is a sufficient
19	statement of the objectives. If you look at the
20	cases, and I think the cases are wise in this
21	regard, you see, of course, Trinko, saying that
22	monopoly profits can be a good thing.
23	More important, I think, you see some of the
24	earlier cases, Grinnell and ALCOA, cases that say in
25	effect quite explicitly that, if a monopolist gains

For The Record, Inc.

18

1	his monopoly power by skill, foresight and industry,
2	that's okay.
3	Those cases embrace a normative proposition
4	that is very important to the fact that antitrust
5	has been supported by the political system in this
6	country for 120 years. That normative proposition
7	is that if the conduct is permissible, in some sense
8	defined without regard to its consequences, it's
9	okay.
10	So what we have to do on the conduct
11	element, exclusionary conduct, is to focus on the
12	quality of the conduct defined without regard to its
13	impact on consumer welfare or dynamic welfare or
14	whatever.
15	It happens, I believe, that if you do that,
16	you are adopting, at least if you do it the way I
17	would do it, what works out to be a very good proxy
18	in the real world, given the problems of
19	administrability and so forth, for achieving the
20	economic objectives.
21	In any event, I think you cannot focus just
22	on the economic objectives. You have to identify
23	clearly the normative objectives of exclusionary
24	conduct law.
25	CHAIRMAN MAJORAS: Anybody want to take that

For The Record, Inc.

19

1	on in terms of whether that is enough, whether
2	looking at the conduct of the defendant rather than
3	the impact on consumers or competitors is adequate?
4	MR. PITOFSKY: I already said I'm
5	uncomfortable with that. It puts the focus in the
6	wrong place.
7	My concern is not the behavior of the
8	monopolist, the defendant. I thought antitrust laws
9	were designed to advance and I think the bottom line
10	is, consumer welfare.
11	If you are looking for consumer welfare, I
12	think it is relevant but not dispositive to know
13	what the intent of the monopolist is and what the
14	nature of its conduct is.
15	But I want to pick up that just because the
16	monopolist behavior is efficient or involves a
17		sacrifice of profit doesn't answer the question. I
18	want to know how anticompetitive it is with respect
19	to consumers.
20	I thought at least in this country consumer
21	welfare and not total welfare -- maybe you can
22	challenge it in academia, but as far as the courts
23	are concerned -- consumer welfare is what it's
24	about.
25	MR. MELAMED: Can I make a brief comment in

For The Record, Inc.

20

1	response to Bob, just a question?
2	Bob, if a firm builds a better mousetrap and
3	as a result obtains enduring market power, and the
4	effect of the enduring market power is overall to
5	make consumers worse off than they would have been
6	if they never built the mousetrap, do you condemn
7	that conduct because --
8	MR. PITOFSKY: How do consumers come out
9	worse off in the face of a better mousetrap?
10	MR. MELAMED: My mousetrap is 5 percent
11	better than the incumbents', I drive the incumbents
12	all out of business; after they leave, I raise
13	prices 5 percent. It is easy to think of
14	hypotheticals where consumers are worse off.
15	MR. PITOFSKY: That's superior skill as far
16	as I'm concerned and I don't have any problem with
17	it. But it's not the typical case.
18	MR. BARNETT: I'm not sure we have so much
19	disagreement.
20	Rick, you want to jump in?
21	MR. RULE: Sure. I am for once to the left
22	of both Doug and Bob. And perhaps I wouldn't say it
23	is one of the few times, because I actually agree
24	with them a lot.
25	But I think I agree with Bob probably

For The Record, Inc.

21

1	wholeheartedly, I guess. I said this before.
2	I worry about the unitary approaches to
3	single-firm conduct. I think it creates a lot of
4	issues.
5	My own personal view is, as I said before, I
6	don't think the world would be a terrible place
7	without Section 2 of the Sherman Act, because I
8	think most of the conduct that is worthy of
9	condemnation can be attacked through various other
10	legal means.
11	So to me, I would say the biggest issue is
12	cabining Section 2 and focusing it.
13	The problem with the unitary standards is, I
14	think, they presume a sort of capability of
15	regulators and enforcers and courts to distinguish
16	efficient from inefficient conduct that just doesn't
17	exist.
18	I think that I have always been very
19	impressed by some of the writings of Judge
20	Easterbrook and particularly the limits of
21	antitrust.
22	And the fact is, if you look, I think,
23	historically at tests that put a burden on a
24	defendant to justify its conduct as efficient,
25	inevitably the courts find it very difficult to

For The Record, Inc.

22

1	agree or to see an efficiency.
2	So I think the focus really ought to be on
3	whether or not there is exclusion, foreclosure, or
4	whatever you want to say of competition.
5	I don't think that is a sufficient condition
6	to condemn something, but I think it is necessary.
7	It may be that the foreclosure, or the
8	exclusion is due to the fact that there is a better,
9	more desirable mousetrap, and that is an efficiency
10	defense, and I think there ought to be allowed an
11	efficiency defense.
12	But I think that an absolutely necessary
13	condition is market power on the part of the
14	individual and exclusion of competition.
15	The last point that I would make that I
16	think is often left unsaid in these sorts of
17	discussions but I think is very important, when you
18	are talking about going after unilateral conduct and
19	you don't have an agreement, you don't have all the
20	issues that I think, quite rightly, warrant
21	antitrust enforcement when you are talking about an
22	agreement. When you are talking about going after
23	unilateral conduct, you are essentially talking
24	about the government regulating behavior of
25	individuals, maybe companies. But it is unilateral

For The Record, Inc.

23

1	action.
2	And there, I think, we as a society, given
3	the way we are organized, should be very concerned
4	not only about the adverse economic effects, the
5	false positives, but also about the impact on
6	liberty, on creativity, and on all of the benefits,
7	not only to the economy, but also to our political
8	life that individual freedom and liberty bring.
9	CHAIRMAN MAJORAS: Susan, you were going to
10	make a comment before Rick.
11	MS. CREIGHTON: That's all right. I can
12	encompass it in my remarks, which was I have sort of
13	a 1 and 2A and B. Hopefully that is not breaking
14	the rules.
15	So the first point and I think actually
16	maybe directly in contrast to Doug, the first thing
17	I would love to see come out of the report is an
18	affirmation that the principle that I think
19	underlies the rule of reason both for Section 1 and
20	Section 2, which is consumer welfare as sort of the
21	touchstone for our analysis, should be really the
22	governing principle in terms of what we adopt for
23	specific rules for conduct under Section 2.
24	I think, like Bob, I'm not saying we can
25	come up with a single unifying test that would cover

For The Record, Inc.

24

1	all that type of conduct. But I believe that we
2	should be assessing the particular tests that we
3	adopt with respect to particular conduct in terms of
4	whether or not it does maximize consumer welfare and
5	is consistent with the rule of reason.
6	So I would use something like the Microsoft
7	test as sort of our default unless and until we can
8	conclude with respect to particular types of
9	behavior that there is another type of test that we
10	have in predatory pricing that more specifically
11	advances the balance of maximizing consumer welfare
12	for that particular type of conduct.
13	The second thing that I would like to see
14	come out of the report, and this may be a little bit
15	outside the direct question of the adoption of
16	substantive rules under Section 2, is I think that
17	there are two powerful ways in which our analysis of
18	Section 2 substantive standards gets distorted by
19	things that don't directly relate to the merits of
20	Section 2 liability, which is, first, the prospect
21	of treble damages in private litigation, and the
22	second is the question of the scope of privileges
23	and immunities.
24	I think just as in our analysis of patent
25	reform, I think many people in the antitrust

For The Record, Inc.

25

1	community thought it is important not to remedy
2	problems with the patent system by adjusting
3	antitrust.
4	In the same way, I think it would be
5	important to try not to distort our analysis of
6	substantive antitrust analysis because of the fear
7	of treble damage liability, and if there is a
8	perspective that that is influencing or has a
9	powerful negative effect in terms of how Section 2
10	is being applied, that the agencies I would
11	encourage to address that head on as something that
12	Congress needs to address.
13	And in the same way, on sort of the opposite
14	side, I think that the ever-expanding scope of
15	privileges and immunities, the ability of people to
16	protect conduct that otherwise would be subject to
17	Section 2 is probably the single biggest deterrent
18	to the ability of the agencies effectively to
19	enforce against anticompetitive conduct.
20	That also would be an issue for the agencies
21	to identify for Congress and for the courts.
22	MR. BARNETT: Not hearing a lot of support
23	for a single unified test.
24	If I can turn to a slightly more specific
25	question, I guess, which is do you think that there

For The Record, Inc.

26

1	should be particular safe harbors, maybe conduct
2	specific or conduct-specific safe harbors under
3	Section 2, and if so, what are a couple of the areas
4	you would focus on?
5	I don't know if -- Greg or Jeff, you haven't
6	jumped in yet. If you want to tackle that one
7	initially.
8	MR. EISENACH: Let may say two things.
9	First of all, in my view, we have missed the
10	biggest issue in the room, and it is not in the
11	room, it is a couple thousand miles away across the
12	Atlantic and across the Pacific.
13	I agree with Jim, the LePage's decision
14	was -- what does Obi-Wan Kenobi say -- a powerful
15	disturbance in the force, and we all felt that
16	something bad had happened.
17	But that was a perturbance in a vastly more
18	settled pond than what we see going on around the
19	world.
20	I think reading the Article 82 Green paper
21	is in many ways an exercise in cognitive dissonance
22	for American antitrust professionals.
23	I guess if I were to suggest a number one
24	priority, both from a substantive perspective and
25	from the procedural perspective of venue shopping

For The Record, Inc.

27

1	and so forth, one of them has got to be trying to
2	continue the process of achieving convergence in the
3	major antitrust venues around the world. The EU is
4	not alone.
5	So I didn't want to let that go.
6	The second thing is that it seems to me that
7	the dichotomy between safe harbors and presumptions
8	on the one hand and a complete consumer welfare
9	approach on the other hand is a false one, and I
10	think it is captured in Doug's comment.
11	The question that Doug leaves me with is
12	what is the underlying analytical basis of the rules
13	that we do adopt? If it is not a consumer welfare
14	standard, then I don't know what it is.
15	I think our current safe harbors are quite
16	unsophisticated ones in many cases. I find it
17	inexplicable that 40 years after we began departing
18	from the structure conduct performance paradigm, we
19	are back at a point where the share of the number
20	one firm is somehow the proposed safe harbor in the
21	first step of a market power test.
22	I don't know what 75 percent or 50 percent
23	or 40 percent means out of context. And surely we
24	can state the safe harbors in more sophisticated
25	ways.

For The Record, Inc.

28

1	But it does not seem to me that there is any
2	necessary conflict between a safe harbor test or a
3	series of safe harbors or presumptions on one hand
4	and a consumer welfare analysis on the other hand.
5	Had Microsoft had some legitimate business
6	purposes for some of the conduct for which it was
7	found liable in the Court of Appeals ruling, it
8	might not have been found liable.
9	That's a good example, I think, of a
10	presumption for a safe harbor which very much is
11	within the context of the whole rule of reason
12	analysis.
13	CHAIRMAN MAJORAS: Can I just follow-up on
14	that for a second?
15	I would like to see what others think about
16	that.
17	When we look at what the Court of Appeals
18	did in Microsoft and we talk about it as a balancing
19	test, I have always looked at it as a weighted
20	balancing test.
21	I think we are right about this. If you
22	read, as the Court of Appeals went through every
23	allegation of conduct, any time Microsoft put up any
24	plausible business justification for it, that ruled
25	the day and that was the end of it.

For The Record, Inc.

29

1	It was just, I think when Microsoft said
2	"no, actually we didn't do those things," that then
3	the court said "oh, yes, you did, and because you
4	said you didn't, you didn't put forth a
5	justification, therefore you lose on that one."
6	It seemed to me the balancing test was
7	pretty weighted.
8	What do people think about that? Does that
9	make you feel better or worse about if the so-called
10	balancing test ended up sort of dominating in this
11	area going forward?
12	I know Doug is dying to weigh in.
13	MR. MELAMED: I think you are completely
14	right that the Microsoft Court never in fact
15	balanced.
16	In the two instances I believe it found that
17	there was a legitimate justification, and that was
18	the end of the analysis. Microsoft won.
19	In other instances, either because Microsoft
20	didn't advance a justification or the court rejected
21	it on the facts, Microsoft lost.
22	Let me comment on this idea of balancing
23	rule of reason in Section 2. It is a meaningless
24	concept. It is at best a throwback to the Chicago
25	Board of Trade case.

For The Record, Inc.

30

1	In collusion cases, we know that rule of
2	reason means, did the agreement increase or decrease
3	the outcome of the parties to the agreement.
4	There is no metric, no meaning to rule of
5	reason, where you have both benefits and harms and
6	you are trying to balance them or, in Hovenkamp's
7	terms, assess proportionality.
8	As to safe harbors, I agree with Rick.
9	There ought to be a safe harbor where the conduct
10	did not exclude rivals or create or maintain
11	monopoly power.
12	And on the other extreme, I think that cheap
13	exclusion and other forms of naked exclusion, in
14	which there is no efficiency you can condemn the
15	conduct if it excludes rivals and injures
16	competition, without more.
17	But to talk about rule of reason or
18	balancing as a solution to the problem where you
19	have both benefit and harm it seems to me is
20	nonsense. And I don't think any court does it.
21	My experience is that courts find either a
22	justification, in which case defendant wins, or no
23	justification, in which case plaintiff wins.
24	It seems to me talking about rule of reason
25	is an empty vessel that leads courts to do what the

For The Record, Inc.

31

1	LePage's court did, which is to say "I don't know
2	how to balance this, I don't know what to do with
3	this" and then come up with some crummy law because
4	it finds no guidance in the prior cases.
5	MR. PITOFSKY: The balancing test is the
6	baseline of all of antitrust.
7	The rule of reason compares procompetitive
8	justifications, anticompetitive effects.
9	Is there another way to get there without
10	examining the anticompetitive effects? That is true
11	of exclusive dealing, true of tying, true of
12	virtually everything regulated by antitrust, joint
13	ventures.
14	Merger is really a rule of reason analysis.
15	Why do you single out Section 2 of the Sherman Act
16	as an area where balancing is nonsense?
17	MR. MELAMED: Because I think of it as
18	collusion versus exclusion, not Section 1 versus
19	Section 2.
20	If you and I agree to a joint venture, we
21	can ask a simple question. Do the efficiencies
22	trump the market power? That is, does our output go
23	up or down?
24	If you exclude me from the market because
25	you have a more efficient exclusive dealing

For The Record, Inc.

32

1	agreement that enhances your ability to distribute
2	your product, you have the efficiency gains to you
3	and the exclusion to me and the consequences for my
4	customers.
5	I don't know of an algorithm that makes any
6	sense for weighing those two against each other.
7	MR. BARNETT: Rick.
8	MR. RULE: The only point I would make is
9	that, in this case, you are both right, I would say.
10	Bob's observation is sort of fundamentally
11	true about antitrust. Inherently in antitrust, you
12	are trying to balance harms to consumer welfare
13	against gains to consumer welfare.
14	I think Doug is right in the sense that it
15	becomes infinitely more difficult to make that
16	operational in a Section 2 context for a variety of
17	reasons.
18	So I agree with Doug that there is a need in
19	light of that to look for, if you will, operational
20	rules that incorporate that sort of insight of
21	balancing, but it is done in a way that courts can
22	actually manage.
23	You could argue that maybe they didn't do
24	such a great job in the Microsoft case. My
25	perspective is a little different than Debbie's, for

For The Record, Inc.

33

1	perhaps obvious reasons.
2	I think a lot of the company's
3	justifications were given the back of the hand,
4	frankly.
5	But I do believe that -- and I think this is
6	pretty consistent in Section 2 -- there is this
7	tendency, although it is a very difficult hurdle for
8	defendants to get over, but if defendants can show
9	that their conduct has a legitimate justification
10	for it, it typically is a good defense to a Section
11	2 claim, regardless of its impact.
12	I think that is probably an appropriate way
13	to approach it. Maybe Doug agrees with that.
14	The concern I have always had with a lot of
15	these tests is that at the end of the day, you have
16	to conclude that the conduct actually does exclude
17	somebody.
18	One of the reasons that you look at the
19	number one firm's market power, I would say, is a
20	legal reason. Section 2 talks about monopolization,
21	for better or worse.
22	That concept, other than a firm's market
23	power and its position relative to its competitors,
24	is meaningless. You have to give some meaning to
25	the law. That is what the law is.

For The Record, Inc.

34

1	That's the single basis for attacking
2	unilateral behavior.
3	MR. PITOFSKY: The sentence was there are a
4	number of reasons why the rule of reason works in
5	many areas of antitrust but not Section 2.
6	I would be curious as to what those other
7	reasons are.
8	MR. RULE: If I said that, I'm not sure --
9	I think the concept of reasonableness is the
10	appropriate way to approach it.
11	The question of what the rule looks like in
12	Section 2 is more difficult.
13	One, it is more difficult because, unlike
14	Section 1 where you have an obvious target which is
15	an agreement that is in some way explicit between
16	two parties and you can look at it, in Section 2,
17	the conduct is not that explicit. It tends to be
18	implicit. It is something a company has done
19	unilaterally.
20	It is also very difficult to extricate it
21	from all the other competitive conduct that a
22	company engages in and evaluate it that way.
23	You have the fact that intent evidence, in
24	my opinion, is completely worthless in this area,
25	because you can't distinguish intent evidence that

For The Record, Inc.

35

1	shows a desire to be vigorously procompetitive or
2	vigorously anticompetitive.
3	You also have the fact that -- and this was
4	really Doug's point, which was perhaps his principal
5	point -- unlike Section 1, where you can look and
6	say, "okay, gee, we have an agreement and what does
7	it do to market power, does it create it, is it an
8	exercise of market power?"
9	In Section 2, it is always indirect. First
10	off, we don't condemn a company unilaterally from
11	exercising market power.
12	One of the things that's interesting about
13	Trinko is the point the court makes that, rather
14	than condemning a monopolist for charging monopoly
15	price, we actually want him to do that because
16	that's his reward if he has gotten it through luck,
17	skill or foresight in doing it.
18	So instead, in a monopolization case, what
19	you are looking at is some sort of indirect impact
20	because there is an adverse effect on a competitor,
21	which you then have to translate into some impact on
22	consumer welfare.
23	Then you have to compare it with the
24	procompetitive benefits. That's very difficult.
25	That goes sort of to Doug's point.

For The Record, Inc.

36

1	There is no algorithm for making that
2	comparison that I'm aware of from economists.
3	Instead, you have to try to develop rules,
4	whether they are safe harbors, whether they are sort
5	of general market power screens or something,
6	because I think saying that you are going to
7	directly measure and balance the procompetitive and
8	anticompetitive effects is probably fooling yourself
9	and the courts because it is not really possible.
10	Instead, you have to come up with rules that
11	are directed to trying to make that balance but
12	probably in some kind of gross fashion.
13	CHAIRMAN MAJORAS: I have a question about
14	the safe harbor concept.
15	Before I do, Greg, you have been so patient
16	down there. Is there anything you want to add on
17	any of these topics?
18	MR. SIDAK: I was going to go off in a
19	completely different direction.
20	Okay. I think that one of the big questions
21	that Section 2 poses is whether the jurisprudence in
22	this area is robust with respect to alternative
23	objective functions of the firm, alternative revenue
24	models, alternative production technologies.
25	By that, I mean suppose you change the

For The Record, Inc.

37

1	assumption that a firm is a profit maximizer. Does
2	our existing jurisprudence on predatory pricing, for
3	example, give us much guidance?
4	It is not such a hypothetical question. For
5	example, the U.S. Postal Service is now subject to
6	antitrust -- it has had its antitrust immunity
7	lifted with respect to products that are not within
8	the statutory monopoly.
9	The last time I checked, the U.S. Postal
10	Service was not a profit maximizer.
11	With respect to revenue models, implicit in
12	a lot of the discussion we have had so far is that
13	we are talking about product markets that are pretty
14	easy to get our arms around, relatively mature
15	products.
16	What if we are talking about some of the
17	kinds of products and services that are at the
18	intersection of the Internet, telecommunications,
19	financial services and the like, where you have
20	multisited markets, you have multiproduct firms.
21	We can all agree that consumer welfare is
22	what we are trying to maximize. But which
23	consumers?
24	A given business practice may result in some
25	service being given away for free to one set of

For The Record, Inc.

38

1	consumers. And that clearly benefits them. But is
2	there an adverse effect on some other set of
3	consumers?
4	So I think the consumer welfare objective is
5	just the beginning of the analysis.
6	When we are looking at some of these more
7	complex markets with multiple sides or firms that
8	are multiproduct firms, in which they may be
9	subsidizing a particular product in order to
10	stimulate the network effects and then with respect
11	to the production technology point, I think that
12	antitrust jurisprudence, compared to the traditional
13	law and economics of sector-specific regulation is
14	not very agile with respect to multiproduct firms.
15	I think this is one place where the
16	Europeans actually have shown some greater skill
17	than American courts.
18	In a case like the Deutsche predatory
19	pricing case in the EC, where they explicitly
20	recognized the multiproduct nature of the firm and
21	had to calibrate the predatory pricing rule to
22	reflect the fact that there were multiple products
23	involved.
24	So they used Jerry Fowell-Haber's
25	combinatorial cost test to try to establish what the

For The Record, Inc.

39

1	appropriate price floor was for the particular
2	service in question that was allegedly being priced
3	below its cost.
4	So I think that the robustness of Section 2
5	jurisprudence across these different economic
6	dimensions is an important issue.
7	The other really big thing -- and I will
8	stop here -- is remedies and evaluation of the
9	efficacy of enforcement and of particular remedies.
10	We don't have much of a tradition. I'm not
11	sure we have much of a tool kit for knowing whether
12	we are systematically improving or reducing consumer
13	welfare over the long haul.
14	Much of the discussion about whether one
15	kind of rule is better than a different kind of rule
16	is really a question of are we minimizing the sum of
17	type 1 and type 2 errors under one approach rather
18	than another.
19	I don't know how we can possibly answer that
20	question unless we have some sort of time series to
21	look at.
22	Lawyers, that's not their stock in trade to
23	do that sort of thing. It is a very difficult task
24	to undertake.
25	CHAIRMAN MAJORAS: I agree with you on

For The Record, Inc.

40

1	remedies. I'm looking forward to discussing that
2	further with you.
3	I know Jim Rill was going to make a comment.
4	MR. RILL: I would just as soon follow-up if
5	you are going to start on safe harbors. If you want
6	to lead that off.
7	CHAIRMAN MAJORAS: Yes, I will. What is
8	interesting is, Jeff, I understand your point about,
9	for heaven's sakes, when you talk about safe
10	harbors, aren't you really talking about a market
11	share of safe harbor, and then aren't we going
12	backward, not forward, in terms of structural
13	analysis.
14	I heard what Doug said in agreeing with Rick
15	on what the safe harbor ought to be. That requires
16	some real analysis to get there.
17	A safe harbor not based on structural
18	presumptions might help you if you are actually in
19	court because it gives you a better chance of
20	winning.
21	How does that help lawyers who are
22	counseling their clients and trying to keep them out
23	of there initially?
24	What kind of a safe harbor can we have that
25	is truly meaningful and keeps people out of the

For The Record, Inc.

41

1	legal system from the very beginning?
2	MR. EISENACH: Just very briefly. There are
3	others wiser than me on this.
4	First, I'm not opposed in any way to a 75
5	percent safe harbor or a 70 percent safe harbor. It
6	is better than a 50 percent safe harbor.
7	My point really went to the notion that
8	surely we can do better than share of the top firm
9	as a metric. That surely can't be the best we can
10	do.
11	But the second point would be that, again, I
12	think that the metrics can become more robust and
13	more sophisticated without becoming less useful.
14	Also, do we have it upside down when we look
15	at market shares first and entry second? I think we
16	do.
17	CHAIRMAN MAJORAS: Interesting. Jim?
18	MR. RILL: I think history has embedded us
19	with the notion of at least a market share test for
20	a safe harbor, at least as a starting point, only as
21	a starting point.
22	The International Competition Network
23	recently surveyed, as part of its single firm
24	conduct working group, the question of whether or
25	not -- first of all, I think something like 70

For The Record, Inc.

42

1	percent agreed that consumer welfare of the 35
2	nations that responded to the questionnaire, that
3	consumer welfare was the appropriate underlying
4	fundamental principle of monopolization Section 2,
5	Article 82 and related enforcement technology
6	techniques. But very little probing beyond that as
7	to what consumer welfare meant.
8	I think I have to say that Bob is a little
9	bit simplistic on this notion, and I think there is
10	a lot more latitude, but that is another issue.
11	I think that is a starting point. Again,
12	any number, about 70, 80, 90 percent of respondents
13	to the questionnaire would use a safe harbor
14	threshold of some level of market share, market
15	power, if you will.
16	Now, some of those safe harbors are rather
17	low. I think Japan is around 10 percent, which
18	doesn't give me a lot of comfort. 70 percent sounds
19	reasonable to me, maybe a little higher.
20	But I think we can get beyond that. I think
21	there is enough -- a lawyer quite clearly can
22	demonstrate, an economist can demonstrate that there
23	is a rich body of law in the United States stemming
24	from the law of predatory pricing which can bring
25	into the notion of consumer welfare certain

For The Record, Inc.

43

1	operational tests, if you will, that can be safe
2	harbors applicable not only in the predatory pricing
3	area but with some further depth analysis into areas
4	that go beyond single firm predatory pricing, in
5	fact, in all pricing areas, bundled pricing, loyalty
6	discounts and maybe developing into the areas of
7	coercive tying, one wants to think about not
8	contractual tie but price-related tie.
9	I think a thought given to that kind of an
10	operational safe harbor approach is not inconsistent
11	either with the unilateral or unitary test.
12	It doesn't seem inconsistent with a consumer
13	welfare analysis stemming from some of the
14	literature, at least in the Trinko decision and more
15	recently in the Weyerhaeuser decision, where the
16	Supreme Court provided that kind of approach to a
17	safe harbor from a legal operational basis and would
18	provide significantly greater clarity to those of us
19	who are trying to counsel companies and to
20	enforcement agencies as they move to the next stage.
21	MR. BARNETT: I think Jim's comments
22	actually began to quite conveniently and
23	appropriately blend into our next topic, having to
24	do with a definition of what is monopoly power and
25	by your reference to defining that through market

For The Record, Inc.

44

1	shares.
2	Debbie, I think you are going to lead off on
3	that.
4	CHAIRMAN MAJORAS: I will.
5	Doug, you have been dying to jump in on this
6	issue. I think it relates. If you want to go
7	first.
8	MR. MELAMED: I will be very brief.
9	Debbie, I was very glad that you asked the
10	safe harbor question in terms of the impact on
11	counseling rather than just the impact on
12	litigators, because the impact of antitrust rules in
13	litigation, it seems to me is much less important
14	than the impact of those rules on the millions of
15	decisions that businesses make every day that don't
16	reach the courts, that is, on the guidance that
17	antitrust law gives to the business community.
18	From my experience in counseling, market
19	share-type screens are of limited value because
20	market share depends on market definition, and it is
21	a binary concept and we are often sitting there,
22	saying well, gidgets might be in the market with
23	widgets, but they might not be and who knows.
24	In my experience, much more useful to the
25	client are guidelines and safe harbors that focus on

For The Record, Inc.

45

1	the nature of the defendant's conduct, things like
2	is the price below your avoidable cost, does it make
3	business sense, are you sacrificing a profit,
4	whatever it may be.
5	Even rules of that type I think are bad
6	rules are useful for counseling -- rules such as:
7	Is the exclusive dealing contract for a duration of
8	a year or less?
9	Those things that enable the defendant to
10	look at his conduct are much more valuable as safe
11	harbors than those that require him to analyze the
12	market.
13	CHAIRMAN MAJORAS: Okay.
14	Susan, as we look at the concept of monopoly
15	power and we typically begin the analysis with that
16	in a Section 2 context as well as in a Section 1
17	context, I should say -- welcome, Tim.
18	MR. MURIS: Thanks.
19	CHAIRMAN MAJORAS: As we look at this, do
20	you think it is useful for us to establish a sort of
21	conclusive presumption on market share?
22	We have had a couple comments here that the
23	market share screens are really not that useful and
24	you have to do so much analysis anyway in order to
25	define the market that it is not that useful.

For The Record, Inc.

46

1	You have certainly been on the enforcement
2	side. What do you think about those kinds of safe
3	harbors?
4	MS. CREIGHTON: I think both Professor
5	Elhauge and also maybe Tom Krattenmaker and
6	Professors Lande and Salgo have written a couple of
7	articles talking about how market power -- not
8	market power -- the percentage of the market that
9	you control actually can be helpful as direct
10	evidence regarding how profitable is it likely to be
11	to you and both your incentives and your ability to
12	enter into some kind of exclusionary conduct.
13	So it can be direct evidence and quite
14	important in that way.
15	I do get concerned about using, at least in
16	attempt cases, as a screen, because I think if you
17	looked at Unocal or Rambus, for example, without
18	getting into the -- sort of any standard-setting
19	case, the person may have had no market share at all
20	in whatever the relevant market was.
21	That does not necessarily dictate how
22	likely -- what the market share would have been or
23	their market power would have been if the
24	exclusionary conduct was successful.
25	So I would be concerned about saying it is

For The Record, Inc.

47

1	always required as a preliminary step before you get
2	to the question of -- one of the advantages that I
3	think or one of the things that American law
4	emphasizes which maybe the Europeans don't as much
5	is I think for them, they really do focus on market
6	share dominance, and then they have very strict
7	definitions of if you are one of those folks, what
8	can you do.
9	In the course of that, they really lose
10	sight of the question of the causation and whether
11	or not the conduct is conduct that we are concerned
12	about in terms of increasing barriers to entry or
13	otherwise increasing somebody's market power in a
14	way we would be concerned about.
15	I would be concerned also about using a
16	market power screen in the first instance to make
17	sure we don't lose sight of that important
18	additional causation requirement.
19	I think that could be a danger.
20	CHAIRMAN MAJORAS: On the question of
21	durability, I know that in prior panels the
22	panelists really agreed that we need to look at
23	market power and whether it is both substantial and
24	durable.
25	Susan, you certainly but I think everybody

For The Record, Inc.

48

1	today now does so much work in dynamic industries
2	and technology industries in which even if you have
3	market power, it might be quite fleeting. There may
4	not be a durability.
5	Does that make it even less the case today,
6	that we should be looking first at market share
7	screens as a way to at least start to get into the
8	analysis?
9	Bob?
10	MR. PITOFSKY: I think you put it just right
11	toward the end of your remarks.
12	Marketshare is the ramp that leads you into
13	the analysis. The problem is sometimes judges and
14	lawyers think the ball game is over because of the
15	way in which the market has been defined. We
16	shouldn't do that.
17	When you get to the end of the analysis and
18	you look at conduct and barriers and all that, you
19	go back and see if your market share analysis is
20	correct in light of all these factors.
21	Of course, substantiality and durability are
22	critical. If you have market power, but it only
23	survives for a year and then is displaced by some
24	other product that is not really market power.
25	We know the barrier to entry is important.

For The Record, Inc.

49

1	This applies to high-tech. I have always been an
2	admirer of Andrew Groves' book "Only the Paranoid
3	Survive."
4	The whole idea of Learned Hand that market
5	power is a narcotic and competition is a stimulant,
6	you can't say that about these big high-tech
7	companies. They are extremely aggressive in their
8	innovation, and that's a factor that has to be taken
9	into account.
10	But unless you start with market power, I
11	don't know where else you start. It gets you going,
12	because some things, some behavior engaged in by a
13	company with 10 percent of the market is legal and
14	is illegal if the firm has 90 percent of the market
15	is illegal.
16	You have to address that question at an
17	early point. I skipped over the safe harbor.
18	Let me just say that first of all, I'm not
19	comfortable with safe harbors. I like rebuttable
20	presumptions because there are too many quirky
21	situations.
22	Somebody has 40 percent of the market but
23	everybody else has one percent each. So I think
24	that presumption of a safe harbor is rebuttable.
25	Secondly, the safe harbor is going to vary

For The Record, Inc.

50

1	according to the behavior you are dealing with. We
2	have safe harbors for exclusive dealing.
3	We have safe harbors for tie-in sales in
4	terms of the market power of the seller instituting
5	that program, 30, 40, 50 percent and so forth.
6	When you get to lying to the Patent Office,
7	I don't think there is a safe harbor. I don't think
8	there should be a safe harbor.
9	So I think safe harbors, of course, are
10	useful to people who are advising firms about what
11	they can and cannot do, but they should vary
12	according to the nature of the conduct.
13	MR. BARNETT: What if you lie to the Patent
14	Office and get a patent that actually confers no
15	market power, what do you mean there is no safe
16	harbor? Have you violated Section 2 then?
17	MR. PITOFSKY: If you lie to the Patent
18	Office? You are talking about Walker Process
19	insisting on defining the relevant market in order
20	to make out a violation for lying to the Patent
21	Office?
22	MR. BARNETT: The statement was if you lie
23	to the Patent Office, there should be no safe
24	harbor.
25	I'm just wondering what that means in terms

For The Record, Inc.

51

1	of if you get a patent, I think most of us agree
2	that doesn't necessarily give you market power if
3	you end up with a patent which does not give you
4	market power.
5	Have you violated Section 2 or not?
6	MR. PITOFSKY: Fair enough. My answer is
7	there are no redeeming virtues to lying to the
8	Patent Office, none whatsoever.
9	MR. BARNETT: I understand. But if I can
10	perhaps -- I thought it was a yes or no question.
11	MR. PITOFSKY: Okay. Here's my answer to
12	that.
13	CHAIRMAN MAJORAS: You are back in Congress.
14	MR. PITOFSKY: Horizontal price fixing may
15	confer no market power. We declare it illegal.
16	I think lying to the Patent Office is the
17	same thing.
18	MR. BARNETT: Fair enough.
19	CHAIRMAN MAJORAS: We have talked about --
20	Bob, you and some others have said if we don't start
21	with market share, where do we start. We have
22	started there for very long time.
23	But Jeff Eisenach said why don't we think
24	about entry first. I think that's what you said,
25	Jeff.

For The Record, Inc.

52

1	Anybody have any -- Greg?
2	MR. SIDAK: I haven't heard anybody utter
3	the words price elasticity. That's what I care
4	about. I don't care about market shares or entry.
5	If I can directly observe the price
6	elasticity of demand, I can make an inference about
7	whether it is profitable or not profitable to raise
8	price.
9	Let me give you a hypothetical example.
10	Suppose some high-tech industry, a firm has 40
11	percent of the market, casually defined.
12	It raises the price by 10 percent, and its
13	competitors over the same period of time lose market
14	share.
15	Would we infer that there is not a problem
16	because the market share is only 40 percent and that
17	is way below Judge Hand's ALCOA threshold or would
18	we look at a price increase or loss of competitor
19	market share and say that is a more direct set of
20	facts that elucidates what the price elasticity of
21	demand is?
22	CHAIRMAN MAJORAS: Rick Rule, could you
23	counsel a client on that basis?
24	MR. RULE: On price elasticities?
25	CHAIRMAN MAJORAS: Not you personally. I

For The Record, Inc.

53

1	have no doubt you could.
2	MR. RULE: I have generally not had to do
3	that, fortunately.
4	But there are always difficulties, and you
5	have to exercise care when you are counseling a
6	client. But frankly, I have always found the market
7	share requirements of Section 2 to be helpful in
8	terms of advising clients.
9	There are edge cases where it can be a
10	little difficult, and you can tell the client, "gee,
11	I know you don't think you have a monopoly and that
12	you are in a very competitive world, but there are
13	ways in which a court could find the opposite, so
14	you have to exercise some care."
15	But for a lot of companies, given the nature
16	of the industries they are in and what they are
17	doing, it is pretty clear that they don't have
18	market power, and you can worry about other parts of
19	the antitrust laws.
20	I will say that it is probably more
21	difficult as technology has moved along and as the
22	economy has gotten somewhat more dynamic and
23	complex, particularly for information industries.
24	It becomes a little more difficult to use the market
25	power and monopoly power market share screen that

For The Record, Inc.

54

1	traditionally we have used.
2	And I guess for that reason, when you are
3	counseling clients, you kind of have to have in the
4	back of your mind that there could be a way to
5	define the market that would suggest they do have
6	monopoly power.
7	So then you go directly to conduct. And in
8	those industries, particularly, conduct safe harbors
9	would probably be very helpful.
10	So to some extent, I think conduct safe
11	harbors are appropriate there. I will also say,
12	interestingly, in information industries, you rarely
13	get that concerned, at least I do, about pricing
14	issues. Because if you think about it, if they are
15	information industries, generally marginal cost will
16	be pretty low and you will recognize that predatory
17	pricing issues are not that problematic.
18	Generally, I think market share screens have
19	worked. They are more complex today, but they have
20	some value in counsel.
21	MR. BARNETT: Related to that, if I could
22	follow-up with Greg for a little bit, the economists
23	generally tell me that if you think about perfect
24	competition, the way you deal with that is you graph
25	that and it would be a perfectly horizontal demand

For The Record, Inc.

55

1	curve, and if you have downward sloping demand
2	curve, you have some degree of market power.
3	And if you measure that directly, it is
4	probably true that the vast majority of firms in the
5	United States have a somewhat downward-sloping
6	demand curve.
7	Does that mean they all have market power
8	and we should just move on from there? Or should we
9	try to deal with that in some meaningful sense to
10	help in part from a counseling perspective?
11	MR. SIDAK: Of course, they may have
12	differentiated products that explain the downward
13	slope of their firm demand curves.
14	The slope of the demand curve, of course,
15	doesn't tell you whether the firm is earning
16	monopoly rent or just quasi, a risk-adjusted return
17	on investment in innovative activities, for example.
18	So I don't think that the downward-sloping
19	demand curve itself is a cause for antitrust
20	intervention.
21	In terms of the market share, market power
22	filter that we have been discussing, I think it is
23	possible to directly infer something about the price
24	elasticity of demand for a firm even in the absence
25	of market shares if you have certain evidence.

For The Record, Inc.

56

1	So in other words, I don't think you should
2	necessarily back away and say, well, this is way
3	below Judge Hand's threshold in ALCOA, there is no
4	way this could be a monopoly problem. It might be a
5	monopoly problem.
6	CHAIRMAN MAJORAS: Tim, you had a comment.
7	MR. MURIS: I thought Tom's point was quite
8	perceptive. It is not just differentiated products.
9	If you walk on the Mall, any hot dog vendor
10	who raises his price won't lose all his sales. That
11	means the demand is a downward-sloping curve. The
12	reason is transaction costs more than anything else;
13	in a world of positive costs, just about everybody
14	has a downward-sloping demand curve.
15	This fact has profound implications for
16	antitrust economics. Ben Klein has written the best
17	about this in his analysis of the Kodak case and
18	other articles.
19	It means that it is difficult to have simple
20	uses of Lerner indexes and downward sloping demand
21	as measures of anything meaningful.
22	CHAIRMAN MAJORAS: Any comment? No?
23	MR. SIDAK: A common problem when you start
24	looking at industries that are subject to some kind
25	of public service regulation, of course, is that

For The Record, Inc.

57

1	they may be compelled to sell products at low prices
2	or even below costs.
3	So the Lerner index actually has its
4	causation reversed. They have a high market share
5	because they are compelled to charge low margins or
6	negative margin.
7	I agree with Tim that the Lerner index is
8	uninformative and potentially misleading in
9	situations where you have significant economies of
10	scale.
11	MR. BARNETT: Jim, I will turn to you for
12	our next topic to lead off, because that is bundled
13	discounts. You have already revealed a particular
14	interest in that area.
15	We recently had a report issued by the
16	Antitrust Modernization Commission that addressed
17	this topic and set forth a three-part test to
18	determine whether or not there is a violation of
19	Section 2 from bundled discounts.
20	Just briefly, the first prong is allocating
21	all of the discounts to the competitive product --
22	sometimes referred to as the Ortho test -- second,
23	whether or not the defendant -- whether it is below
24	cost under that measure. Second, whether or not the
25	defendant is likely to recoup those losses. And

For The Record, Inc.

58

1	thirdly, whether the bundled or rebate program has
2	had or is likely to have an adverse effect on
3	competition.
4	Aside from the fact that the third prong
5	seems to sort of ask the ultimate question there,
6	the question is is this appropriate standard, is it
7	appropriate as a safe harbor but perhaps not the
8	standard or is it just something we should be
9	looking in a different direction?
10	MR. RULE: First of all, I think the AMC is
11	looking at it only when it relates to conduct by
12	someone who is judged to be a monopolist.
13	Moving on from that to the operational test,
14	I have some difficulty with let's call it the Ortho
15	or AMC allocation formula, both from an operational
16	and from, I think, an analytical standpoint.
17	From an operational standpoint, the
18	allocation itself of the totality of the discount
19	across to the single let's call it target product
20	creates something of a daunting task, and there is a
21	margin or opportunity for error there that I think
22	is quite substantial.
23	Secondly, from an analytical standpoint, I
24	think maybe it is operational as well, it raises the
25	problem of double counting or multiple penalties.

For The Record, Inc.

59

1	Just to take a hypothetical industry, if
2	someone sues on Post-its, and someone else sues on
3	sponges, and someone else sues on tape, and there
4	are three cases going on at the same time, does one
5	allocate the totality of the package discount to
6	each of those products, and in what position does
7	that put the defendant in?
8	However, I think I prefer that there is a
9	solid operational test to safe harbor. The
10	proposition that I think is embraced in Tim's
11	statement to the AMC is that the allocation of total
12	cost to total bundles would be a better way of
13	looking at a test that might suggest illegality.
14	On the other hand, it is possible that the
15	allocation test or the Ortho or AMC formula of
16	allocation, would be appropriate as a safe harbor.
17	This is the position taken in the brief of
18	several law professors recently filed in the Ninth
19	Circuit in the Peace Health case, Professor Crane
20	and others.
21	Recognizing that the difficulties that we
22	have suggested with that test as a presumption of
23	illegality, it might serve a purpose at least of a
24	safe harbor if practicable.
25	So far as recoupment is concerned, I think

For The Record, Inc.

60

1	in a pricing case, recoupment should be an element
2	of the offense and should be considered as part of
3	the potential safe harbor.
4	MR. BARNETT: Bob?
5	MR. PITOFSKY: Let me second what Jim has
6	just said and then let people take shots at it.
7	First of all, back to Doug's excellent point
8	in opening this whole discussion.
9	What is this all about? What are we
10	quarreling about here? It seems to me the point of
11	bundled discounts is it gives consumers a break. We
12	ought to not be too aggressive in deterring it.
13	We should not overdeter it, it, but be
14	careful in this area. Second, it seems to me to be
15	more sensible, as the Aveeda-Turner Treatise
16	originally said about this question when it first
17	came up, you want to allocate the discounts product
18	by product rather than put all of the discounts to
19	one product.
20	There is a serious danger that will drive
21	the price of that product below whatever predatory
22	pricing turns out to be.
23	So I share Jim's view. I think LePage's was
24	wrong, and if the court gets to overturn it and come
25	up with a more sensible rule, the better off we all

For The Record, Inc.

61

1	are.
2	MR. BARNETT: Tim?
3	MR. MURIS: Obviously, anything is better
4	than 3M, than turning it over to the jury.
5	The AMC deserves credit for trying to devise
6	a test. But there are serious theoretical,
7	empirical, and practical problems.
8	As Dennis Carlton said in the AMC report,
9	the bundled discounts can be used for procompetitive
10	reasons. For example, price discrimination can be
11	anticompetitive or procompetitive. It is difficult
12	to separate pro from anti and we need to be careful
13	for that reason.
14	The second theoretical problem is the
15	premise of the AMC allocation is to protect "equally
16	efficient competitors." The problem -- and there is
17	a nice footnote in the government's LePage's brief
18	about this -- is that someone who sells you one
19	thing that you want can't be as efficient as someone
20	who sells you two things that you want.
21	So the AMC's premise is a problem.
22	Moreover, empirically we know almost nothing that
23	tells us that there are anticompetitive problems
24	from bundling. Vernon Smith and I have put together
25	a paper that summarizes the work of his group, which

For The Record, Inc.

62

1	spent a lot of time using experimental economics to
2	take the theories of anticompetitive bundling and
3	show they actually hurt consumers.
4	Well, it was almost impossible to do. They
5	did find some ambiguous cases. Yet, if you do
6	anything to those ambiguous cases, bundling becomes
7	efficient. Thus, if the monopolist lacks a 100
8	percent share, if there are any efficiencies, like
9	transaction cost savings, and if you don't have very
10	strange-looking demand curves, bundling becomes
11	efficient. Obviously, experimental economics has
12	its limits, but it is certainly superior to simple
13	theoretical arguments.
14	There is also a tremendous practical
15	problem. Greg has done a lot of useful work in
16	valuing regulatory agencies, and there is some older
17	and good literature about allocating joint and
18	common costs. If you start trying to do this across
19	the products in a bundle, it is completely arbitrary
20	in terms of allocating these costs to some products
21	and not to others.
22	Finally, I do agree we need a safe harbor.
23	The Brooke Group allocation, the more general
24	allocation that Jim and Bob are discussing is the
25	one that I would support.