1

1	FEDERAL TRADE COMMISSION
2
3
4	SECTION 2 HEARINGS
5	UNDERSTANDING SINGLE-FIRM BEHAVIOR:
6	REMEDIES
7
8
9
10	MARCH 28, 2007
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25

For The Record, Inc.

2

1	P R O C E E D I N G S
2	- - - - -
3	MS. KURSH: Good morning, everyone. Thank you
4	for joining us.
5	I'm Gail Kursh. I'm with the Legal Policy
6	Section of the Antitrust Division, and I would like to
7	welcome everyone this morning to the first of three
8	panels on remedies in Section 2 cases.
9	These panels are part of an ongoing series of
10	public hearings on single-firm conduct.
11	My co-moderator today is Dan Ducore, the
12	assistant director of the Compliance Division in the
13	FTC's Bureau of Competition.
14	The Department of Justice and the Federal Trade
15	Commission are jointly sponsoring these hearings to help
16	advance development of the law concerning Section 2 of
17	the Sherman Act.
18	We began these hearings last June and have
19	covered a wide range of single-firm conduct that may
20	raise antitrust issues, including predatory pricing and
21	predatory bidding, tying, refusals to deal, exclusive
22	dealing, bundled rebates and misleading and deceptive
23	practices, among other topics.
24	It seems fitting to us as we get toward the end
25	of these hearings that we now address remedies.

For The Record, Inc.

3

1	However, it would have been just as fitting for
2	us to have addressed remedies at the very outset of
3	these hearings.
4	While I expect our panelists today may disagree
5	on the effectiveness of past Section 2 remedies and
6	perhaps even have differing views on the appropriate
7	goals of Section 2 remedies, I hope that we can all
8	agree today that crafting appropriate remedies in
9	Section 2 cases is critically important and that
10	consideration of remedies should begin very early in an
11	investigation or litigation.
12	So on behalf of the division, I want to thank
13	our panelists for participating today and agreeing to
14	share their insights with us.
15	I will introduce each panelist in more detail
16	before he speaks. But in brief, our speakers in order
17	are Dave Heiner, vice president and deputy general
18	counsel for Microsoft; Robert Crandall, a senior fellow
19	at the Brookings Institute; Per Hellstrom, chief of Unit
20	C-3 of the Directorate General for Competition, the
21	European Commission; and Tad Lipsky, a partner at Latham
22	& Watkins and former Deputy Assistant Attorney General
23	for the Antitrust Division.
24	I also want to thank my colleagues at the FTC
25	and at the division for organizing these hearings. And

For The Record, Inc.

4

1	our panel this morning will go as follows. We will ask
2	each of the four panelists to speak for approximately 15
3	minutes. We will then take a short break.
4	The panelists will each be given a couple
5	minutes to respond to each other and then we will have a
6	moderated discussion that Dan and I will lead.
7	We will not be taking any questions from the
8	floor, and we intend to end today at 12 noon, take a
9	lunch break and begin the afternoon session at around
10	1:30.
11	Before introducing our first speaker, I will
12	turn things over to Dan and let him make an
13	introduction.
14	MR. DUCORE: Thanks, Gail. On behalf of the
15	Federal Trade Commission, I also want to thank our
16	panelists for agreeing to share their time and
17	especially their views with us this morning.
18	Briefly, the remedies issue is obviously from
19	the agency's point of view about more than simply money
20	damages. That is somebody else's issue.
21	But certainly more so than in an area like
22	merger enforcement, Sherman 2 cases present much more of
23	a one of a kind kind of concern when you are trying to
24	develop the remedy in the sense that you have to be very
25	careful that the particular remedy matches the

For The Record, Inc.

5

1	particular facts and the particular theory of harm in
2	your case.
3	I expect today we will hear a lot about the
4	critical thinking that must go into fashioning effective
5	remedies for particular problems. And Gail is certainly
6	correct -- and I have seen this in my own experience --
7	that you have to be thinking about remedies at the
8	earliest stages of your case and, for an enforcement
9	agency, at the earliest stages of your investigation.
10	As someone who thinks about remedies pretty much
11	full-time, I'm going to be particularly interested in
12	hearing about both the broad approaches but also about
13	some of the smaller issues, including things such as
14	let's say administrability and the pitfalls and dangers
15	that can face an agency as it maybe starts to go off the
16	cliff and become an industry regulator.
17	With those introductions, let me get started,
18	with the exception I have to make a couple of logistical
19	announcements.
20	First, if there is an alarm, please go down the
21	stairway and get out of the building and follow the
22	instructions of people. You will be actually going
23	across the street. Second, the closest restrooms, men's
24	out the door and to the left, women out the door and
25	past the elevators to the left.

For The Record, Inc.

6

1	And finally, especially for the panelists,
2	please turn off cell phones, electronic devices,
3	especially things like Blackberries. They can create
4	static on the microphones.
5	With that more mundane information, let me turn
6	it back to Gail to introduce the first speaker.
7	MS. KURSH: Thanks, Dan. David Heiner is vice
8	president and deputy general counsel at Microsoft
9	Corporation, where he heads up the legal department's
10	antitrust group.
11	Since joining Microsoft in 1994, Dave has played
12	a leading role in Microsoft's response to government
13	antitrust proceedings in the United States, Europe and
14	Asia.
15	He is the author of "Assessing Tying Claims in
16	the Context of Software Integration: A Suggested
17	Framework for Applying the Rule of Reason Analysis."
18	Dave.
19	MR. HEINER: Thank you, Gail and Dan, for the
20	invitation to speak here today, which I very much
21	appreciate.
22	This is a subject upon which I think it is fair
23	to say Microsoft has quite a bit of experience, working
24	largely with many people I see in the room.
25	At the outset, I thought it might be useful to

For The Record, Inc.

7

1	briefly recap the remedies to which Microsoft has been
2	subject over the past decade or so.
3	In 1994, a consent decree was put in place and a
4	nearly identical European Union undertaking were put in
5	place. These were mostly contractual in nature.
6	In 2002, a consent decree and associated
7	litigated final judgment were entered in the Section 2
8	case against Microsoft.
9	The Section 2 case was followed by a number of
10	competitor lawsuits. Hundreds of consumer class actions
11	were filed. Nearly all of these private cases have been
12	settled with payments and some conduct relief as well.
13	In March 2004, the European Commission issued
14	its decision against Microsoft. The Commission took a
15	different approach to the issues than did the U.S.
16	court.
17	In February 2006, the Korean Fair Trade
18	Commission issued its decision against Microsoft. The
19	KFTC took yet a third approach. The EC and KFTC
20	decisions are on appeal now.
21	As you might imagine, all of this generates
22	quite a bit of work within Microsoft and its law
23	department.
24	When I joined the company in 1994, I was the
25	first antitrust lawyer at the company. Today I lead a

For The Record, Inc.

8

1	group of about 30 professionals dedicated full-time to
2	antitrust counseling and compliance.
3	This group includes software developers and
4	business personnel as well as lawyers and paralegals.
5	All told, a few hundred people at Microsoft are
6	engaged in compliance work over the past few years.
7	I would like to begin with a suggestion on the
8	overall approach to fashioning relief.
9	I would suggest that it's probably better to
10	focus on creating or preserving opportunities for
11	competitors rather than limiting the defendant's ability
12	to deliver consumer value. This is the approach very
13	much taken by the U.S. consent decree.
14	The Court of Appeals had reversed and remanded
15	the Section 1 tying claim against Microsoft but affirmed
16	Section 2 liability relating to the manner in which
17	Internet Explorer had been integrated into Windows 98.
18	The decree that resulted did not require that
19	any functionality be removed from Windows. Rather,
20	every provision of the decree is directed at creating or
21	preserving opportunities for competitors, both as a
22	matter of product design and contractually.
23	The focus is upon ensuring that distribution
24	channels remain open. This is an approach that was
25	strongly approved by the Court of Appeals in 2004.

For The Record, Inc.

9

1	Today, new Windows PCs come loaded up with
2	software from Microsoft's competitors, such as Google,
3	Yahoo, AOL, Semantec, McAfee and many others.
4	Under this approach, consumers benefit from the
5	ability to choose either integrated solutions or
6	separate stand-alone software or, as is so often the
7	case, to use both.
8	The European Commission has taken a different
9	approach. The Commission ordered Microsoft to create
10	new versions of Windows from which media playback
11	functionality had been removed. These are called
12	Windows XPN and Windows Vista N.
13	They were built following extensive compliance
14	discussions with the European Commission staff. They
15	are available in every European language.
16	However, not a single PC manufacturer has chosen
17	to license these operating systems. These operating
18	systems sit on the shelf. Costs have been imposed, but
19	there is little apparent benefit for anyone.
20	I will return to another aspect of this in a
21	moment. For now, I would note only that the U.S.
22	approach seems far more effective at advancing antitrust
23	values.
24	This focus on creating opportunity tells us
25	something about the proper objectives of antitrust

For The Record, Inc.

10

1	remedies.
2	I would suggest that remedies should be put in
3	place in order to safeguard competitive opportunities
4	but not necessarily to engineer any particular market
5	outcome, such as a reduction in market share. This is
6	for the market to determine once any competitive
7	restraints have been removed.
8	Indeed, even if engineering market outcomes were
9	thought to be desirable theory, it is hard to see how
10	this could be accomplished in practice in most cases.
11	By its nature, a remedy will only govern the
12	conduct of the defendant, not other market participants.
13	Everyone else, competitors, developers of
14	complementary products and, most notably, consumers will
15	act according to their self interest.
16	This is particularly noteworthy in high-tech
17	industries where products often interconnect with each
18	other in different ways.
19	For example, both the U.S. and EU remedies
20	require Microsoft to make available certain technology
21	called communication protocols to its competitors for
22	use in their products.
23	About 30 firms have taken licenses to this
24	technology under the U.S. program and one firm to date
25	under the similar European program.

For The Record, Inc.

11

1	But whether firms choose to take a license and
2	what kinds of products they build with those licenses
3	is, of course, entirely up to them and outside the
4	control of either Microsoft or any antitrust agency.
5	This general point is relevant outside the
6	context of access remedies as well.
7	Internet Explorer continues to have very high
8	share, although declining. Should this be seen as a
9	shortcoming of the U.S. consent decree?
10	Well, the open source Firefox Web browser now
11	has about 14 percent share, up from zero just a few
12	years ago.
13	Given the safeguards set up by the consent
14	decree which apply on a worldwide basis, there is no
15	reason why Firefox couldn't have a much higher share if
16	that reflected consumer preferences.
17	In fact, Firefox's share is about 33 percent in
18	some major European countries, up from 20 percent just a
19	year ago.
20	This focus on competitive opportunity rather
21	than outcome of market shares is especially important, I
22	think, in government actions.
23	As the Court of Appeals explained in the
24	Microsoft case, liability can be established with little
25	or no proof of actual market impact from the conduct at

For The Record, Inc.

12

1	issue.
2	This is what the court termed in the Microsoft
3	case a rather endogenous test for causation.
4	In fact, the District Court found that there was
5	no proof that the success of Internet Explorer had been
6	due to unlawful conduct.
7	Where there is no proof of market impact in the
8	first place, it would seem especially inappropriate to
9	expect a remedy to bring about a particular market
10	outcome.
11	This brings me to my third observation.
12	Whatever the proper role of antitrust remedies may be in
13	the abstract, I think, as Gail and Dan said at the
14	outset, it is really quite important that they be fully
15	thought through before liability proceedings are
16	commenced.
17	This is true for at least two reasons.
18	First and most importantly, if it is hard to
19	devise an appropriate remedy, that may suggest that
20	there is no liability in the first place. At the very
21	least, it may suggest that the liability rules were not
22	sufficiently clear to provide any real guidance or
23	notice to the defendant of what would be termed unlawful
24	later.
25	Second, absent a clear view on the question of

For The Record, Inc.

13

1	remedy, it may be difficult or impossible to obtain
2	rapid relief through settlement.
3	These points are well illustrated I think by
4	Microsoft's experience in dealing with the Windows tying
5	issues through the years.
6	The addition of new functionality to Windows can
7	present competitive challenges for firms that wish to
8	offer comparable functionality separately.
9	Antitrust agencies around the world have focused
10	on that over the past 10 years.
11	At the same time and as the Court of Appeals
12	noted in the Microsoft case, such integration can lead
13	to important benefits for software developers, PC
14	manufacturers, in fact, to the entire PC ecosystem.
15	That's why functionality has been integrated
16	into new operating system products steadily over the
17	past 20 years or so and why in fact we see integration
18	of functions as quite a common function across many
19	product categories.
20	So one has the question how should these
21	competing considerations be addressed in a remedy?
22	In the U.S., the consent decree approach I
23	outlined earlier is now in place. But there were quite
24	a few bumps along the road to getting there, including
25	three rounds of failed settlement talks, one before

For The Record, Inc.

14

1	Judge Posner in Chicago.
2	I think it is fair to say at least part of the
3	reason why those settlement talks failed is that there
4	was disagreement among the DOJ and the various states as
5	to what would be a suitable form of relief.
6	Absent a clear view on this, no agreement could
7	be reached, and the eventual remedy was delayed.
8	The history in Brussels is instructive as well.
9	In early 2004, Microsoft proposed a variety of remedies
10	to address the Commission's concerns regarding the
11	inclusion of media functionality in Windows.
12	The Commission case team devoted a great deal of
13	effort to defining and exploring those proposals, and
14	Microsoft is grateful for that.
15	Ultimately, however, the Commission determined
16	that a general remedy should be devised that would
17	address all future tying cases.
18	Given the range of possible fact patterns and
19	the benefits of integration, however, neither the
20	Commission nor Microsoft was able to articulate any such
21	rule that would govern future product design decisions
22	despite prodigious efforts by both sides.
23	As a result, settlement talks failed. The
24	Commission proceeded to impose the logical remedy for a
25	tying case, which was an order to untie. As a result,

For The Record, Inc.

15

1	PC manufacturers and consumers can now choose to get
2	Windows without media functionality.
3	As I have said, they have chosen the
4	full-feature version of Windows, as one might expect.
5	So the question becomes should it be unlawful
6	for a firm to fail to create a product for which there
7	is no appreciable consumer demand? Here consideration
8	of remedy may suggest that there was no unlawful tie in
9	the first place.
10	The same might be said about the package
11	discounting that was at issue in Lepages or the
12	selective discounting and output increases that were at
13	issue in the American Airlines case.
14	I would like to conclude with two final
15	observations of a practical nature.
16	First, in Microsoft's experience, it would seem
17	that the legal process is generally best suited to
18	contractual remedies.
19	Particular cases may call for other forms of
20	relief. But we should recognize that these come with
21	significant challenges for all concerned.
22	Contracts are good because they are within the
23	purview of lawyers. We understand contracts. We know
24	how to read them.
25	They are relatively easy to monitor, both for

For The Record, Inc.

16

1	the defendant and for the enforcement agency, and I
2	would note that essentially no issue of significance has
3	arisen through the years in Microsoft's compliance with
4	the contractual provisions of the U.S. consent decree.
5	Product design remedies are more difficult.
6	Here considerable technical expertise may be required in
7	order to devise and monitor a remedy.
8	Ultimately lawyers will remain responsible for
9	making compliance judgments regarding highly technical
10	matters, and this may be hard, even with expert
11	technical help.
12	In addition, agency lawyers will inevitably find
13	themselves drawn into the details of product design and
14	even the details of making engineering trade-offs which
15	are essential to the product design process.
16	To deal with these kinds of complexities, the
17	technical committee set up under the U.S. consent decree
18	now has more than 40 full-time employees.
19	Remedies that require sharing of complex
20	technical information are also quite challenging.
21	Technological complexity can quickly lead, I think it is
22	safe to say, to enforcement complexity.
23	Protocol licensing, for example, is just one of
24	eight major provisions of the U.S. consent decree, but
25	it takes up the lion's share of the compliance work,

For The Record, Inc.

17

1	both for Microsoft and for the agencies.
2	The EU protocol remedy introduces still greater
3	complexity. That is because it seeks to enable
4	fundamentally different computer operating systems with
5	different computer architectures to work together as if
6	they were one.
7	This is a computer science project, and even the
8	Commission itself has recently said that making this
9	work would require a massive development effort by third
10	parties, and that hasn't happened yet.
11	The result has been considerable frustration for
12	the Commission and for Microsoft.
13	This past summer, the Commission imposed fines
14	upon Microsoft of 280 million Euro for failing to
15	complete this project to the satisfaction of the
16	technical advisors set up under that decision.
17	Pricing is another challenge and likely will be
18	for any access case that involves information goods,
19	such as software.
20	The protocol technology that Microsoft has
21	developed was developed over the course of about 10
22	years. It is covered by 35 patents, and many more are
23	pending. It is covered by copyright and trade secret
24	law.
25	How is this to be valued? The answer is not

For The Record, Inc.

18

1	entirely obvious given the many ways that software is
2	monetized today and the varying business models that
3	people have.
4	Microsoft has suggested pricing that is
5	comparable to that which is in place under the U.S.
6	program where many firms have taken licenses. That
7	pricing is backed up by more than a thousand pages of
8	analysis and justification that the Commission
9	requested.
10	The Commission has taken issue with this
11	pricing, however, and is threatening to impose new fines
12	that could run to additional hundreds of millions of
13	Euros.
14	My final observation relates to globalization.
15	From Microsoft's perspective and I think it is
16	fair to say from that of other high-tech companies, it
17	is increasingly important that antitrust agencies
18	cooperate closely on remedies and show due respect for
19	principles of international comity.
20	For sound economic reasons, the Windows
21	operating system is essentially identical all over the
22	world. That uniformity is critical to the role that
23	Windows provides in enabling compatibility between
24	literally thousands of complementary software products
25	and hardware products.

For The Record, Inc.

19

1	And that is threatened today by the varying
2	approaches taken to the Windows tying issues in the
3	United States, in Europe and Korea, which I haven't had
4	time to go through this morning.
5	In the compulsory licensing area, I think it is
6	safe to say that the U.S. and foreign countries are
7	taking a different approach to compulsory licensing.
8	In the age of the Internet, once trade secrets
9	are revealed, they can never be recovered. Absent
10	greater deference to comity principles, we may well find
11	that the legal regime that imposes the most onerous
12	legal requirements de facto prevails on a worldwide
13	basis.
14	Again, thanks very much, Gail and Dan. I
15	appreciate the opportunity to speak here today.
16	(Applause.)
17	MS. KURSH: Thanks, Dave. Robert Crandall will
18	be next.
19	Robert is a senior fellow in economic studies at
20	the Brookings Institution. He has previously served as
21	acting deputy and assistant director at the Council on
22	Wage and Price Stability.
23	He has written extensively on antitrust policy,
24	with a particular emphasis on the telecommunications
25	sector and emerging issues in wireless and broadband

For The Record, Inc.

20

1	competition.
2	Among the antitrust topics on which he has
3	written is the effectiveness or lack thereof, I guess,
4	of relief in government Section 2 cases.
5	Bob.
6	MR. CRANDALL: Thank you, Gail. It is a
7	pleasure to be here.
8	I haven't written that extensively in antitrust.
9	In fact, I spent most of my career looking at regulatory
10	activities that range as far as environmental policy and
11	fuel economy standards and more recently
12	telecommunications regulation, which, of course, is
13	related to competition policy.
14	I have not spent as much time as my colleagues
15	on this panel have, I'm sure, on the details of
16	antitrust, nor the details of Section 2 remedies.
17	My purpose today is to provoke, frankly, and for
18	that reason I'm somewhat disappointed we will not have
19	questions in the audience, though I'm not sure how many
20	economists are in the audience anyway.
21	You see the title of my presentation. I will
22	focus on the AT&T divestiture, not simply because that's
23	the one I know a little bit about, but because some work
24	which I have done and which Clifford Winston and I have
25	done and Ken Elzinga and I have done on Section 2 relief

For The Record, Inc.

21

1	using a case-by-case approach to this, which I think is
2	the only way to go about it.
3	Each one of these cases is sui generis. It is
4	hard to do a more general study. It suggests very
5	little effect it creates on the market, on competition
6	in the market, on output, on prices.
7	In fact, not because of shameless
8	self-promotion, but because I would like to provoke
9	people to read the articles and maybe prove me wrong, I
10	have listed the articles in this first slide.
11	But the one case that everybody comes to as the
12	example of success in Section 2 structural relief cases
13	particularly is the AT&T divestiture, which, of course,
14	was negotiated someplace on a ski slope in Utah in 1982
15	and was executed effective January 1, 1984 after about
16	10 years of litigation.
17	Indeed, at first I would have been a supporter
18	of that and perhaps anyone that was wishing to get into
19	a debate with me on this would find things I have said
20	in the past, 20 years ago, that I might have approved.
21	Maybe I was overly seduced by Bill Baxter, who was a
22	very persuasive guy and a very good fellow to boot.
23	But over time I have come to question whether in
24	fact even the AT&T case can be considered a success in
25	terms of relief from a Section 2 prosecution.

For The Record, Inc.

22

1	Now, most people in the room would know about
2	this case. We don't have to spend much time on it.
3	The principal outcome was a divestiture of the
4	Bell operating companies from the rest of AT&T, AT&T
5	keeping the manufacturing and long distance arms and a
6	large share of the research operations.
7	The near-term result -- and I will show in a
8	second -- is long distance service increased. Long
9	distance service output increased and U.S. long distance
10	rates fell.
11	I do these slides myself. That's why they look
12	so bad.
13	But was the increased long distance competition
14	due to vertical divestiture? This is a very different
15	world in 1982.
16	AT&T accounted for 80, 85 percent of the access
17	lines and almost the same percentage of total telephone
18	subscribers, including wireless.
19	AT&T's wireless service was not launched until
20	1983 in Chicago, its cellular service. They had a more
21	mundane wireless service prior to that.
22	This is the period in which I think it was
23	McKinsey was predicting there would be a demand for no
24	more than one million cell phones in the United States.
25	At this time, though, we had so-called universal

For The Record, Inc.

23

1	service pricing, which is really inverse Ramsey pricing,
2	for you economists in the room. I suppose antitrust
3	lawyers may understand the jargon.
4	This, in fact, invited entry into long distance
5	service. It invited the likes of Bill McGowan of
6	Microwave Communications Incorporated -- he said he
7	changed the name to MCI because he didn't want people to
8	think he was going to fry them -- to enter the long
9	distance service to figure a way to get access to AT&T's
10	service, particularly with very low prices.
11	Indeed, they battled that out for many years,
12	culminating in a private antitrust action and convincing
13	the Justice Department in '74 to file the Section 2
14	case.
15	Once again, the question was was the vertical
16	divestiture which resulted in this case necessary to
17	promote long distance competition? What are the
18	numbers?
19	Here are the numbers on real interstate long
20	distance rates and AT&T's average share of revenues
21	using the same access on the left-hand side.
22	And you see that starting in 1984, after the
23	divestiture, AT&T steadily lost market share and long
24	distance rates came down steadily.
25	This is taken to reflect success of the decree.

For The Record, Inc.

24

1	Of course, we don't know what the but for would look
2	like. And it may well be that without some action that
3	we wouldn't have had this result.
4	In fact, most other countries had to take action
5	themselves, and they took action only along one
6	dimension of the decree. No other country that I'm
7	aware of has actually required a divestiture of their
8	operating companies from manufacturing or long distance
9	service companies in their country.
10	Virtually all of them, however, at some time
11	after 1984, as late as 1998 in the EU, required access,
12	equal access to the incumbent local exchange company
13	switches for terminating or originating calls.
14	This obviously is regulated access, and in any
15	regulated access there is going to be an argument about
16	the price. But nobody engaged in vertical divestiture.
17	One could argue that what Offcom is doing in the
18	U.K. today is a very mild version of structural
19	separation with British Telecom. We will see how that
20	works out.
21	But no other country actually engaged in
22	vertical divestiture.
23	Now, if you look at what happened to the price
24	of long distance services, comparing the U.S. interstate
25	rate -- the intrastate rates didn't go down as fast

For The Record, Inc.

25

1	because the states controlled and used their regulatory
2	controls to keep those prices relatively high, the
3	inverse Ramsey pricing continues to stay low.
4	If you look at the U.S. interstate prices
5	against the average for Canada and wouldn't make much
6	difference which one you use for the EU. I use the
7	three-minute price here, and I think they also publish
8	10-minute prices in the annual monitoring reports that
9	the EU does on monitoring effects of their regulatory
10	program.
11	What you see here is prices came down even more
12	rapidly in Canada and the EU, much more quickly
13	subsequent to their liberalization than it did in the
14	United States subsequent to ours.
15	In fact, equal access to the switches was all
16	that was required. And the FCC in the United States had
17	not done this of its own volition prior to the bringing
18	of the AT&T case in 1974 or prior to the negotiation of
19	the consent decree, the divestiture with the equal
20	access provisions in it in 1982.
21	Now, in no small part long distance rates in the
22	United States fell because of declining access charges.
23	One of the things -- and you could take this as
24	a measure of the success of the decree. One of the
25	things that the decree did was to expose exactly how

For The Record, Inc.

26

1	much if you want to call it broadly subsidy was going on
2	between long distance service and other service, local
3	access to the telephone network in the regulatory
4	process at the Federal Communications Commission.
5	With very high access charges now having to be
6	levied to keep the rates at about the same level, the
7	FCC started the process of rebalancing rates, lowering
8	access charges and putting all those complicated charges
9	that you and I don't understand on the back of our
10	telephone bills, which are in fact designed to try to
11	shield from the public information about what's really
12	going on here.
13	But it made good policy sense to put these
14	nontraffic-sensitive charges as a fixed charge on your
15	telephone bill and to lower the traffic-sensitive
16	charges of long distance by doing so.
17	In fact, a great deal, as you can see, of the
18	decline in long distance rates occurred because of the
19	decline in access charges.
20	My friends who worked on the AT&T case at DOJ
21	and others who have been involved in this process over
22	the years who don't like my presentation will argue with
23	me that this could not have happened but for the case.
24	In fact, that is one of the benefits of the
25	case, I suppose. We could not have persuaded the FCC to

For The Record, Inc.

27

1	undertake both equal access and to rebalance rates but
2	for the divestiture.
3	I suppose that is a benefit. But again, the
4	problem here was not AT&T's monopolization activities,
5	unless you consider their lobbying activities of the FCC
6	as part of it, but, rather, the FCC's seeming
7	incompetence or reluctance to do the right thing to
8	maximize economic welfare for people using the telephone
9	network.
10	As a mea culpa, I was actually at the FCC
11	advising part-time one of the commissioners, Glen
12	Robinson, who is now professor of administrative law at
13	University of Virginia. So I guess I'm tied up in the
14	complicity in all that.
15	This is simply saying much the same thing, that
16	in fact what happened was as a result of the
17	divestiture, there was an exposure of the folly of the
18	universal service pricing policy, something which the
19	FCC addressed with great opposition from so-called
20	consumer groups, who claimed that millions of low-income
21	people would fall off the telephone network.
22	Of course, we know better than that because we
23	know the price elasticity in the demand for access to
24	telephone service is very, very low.
25	What about the costs of the decree? My own

For The Record, Inc.

28

1	estimation -- and I haven't seen anyone else attempt to
2	address this -- was that we lost about $5 billion of
3	output just in the transition from the old AT&T to the
4	new AT&T over 1984-85, in that period.
5	There are some estimates -- now, these were
6	funded by the Bell companies as attempting to get out of
7	the decree. There is an estimate by Paul Ruben, a
8	colleague at Emory, that the process of administering
9	the line of business restrictions in the decree totaled
10	about $1.4 billion over time.
11	Though what happened -- and this goes to Dave's
12	presentation on trying to provide technological
13	prescriptions and deal with changes in technology in an
14	antitrust decree. What happened was that the market
15	changed rather dramatically.
16	Something called the Internet came up and the
17	separation of interstate from local and intrastate
18	services in the decree became extremely problematic.
19	Not only that, but the information restriction which
20	eventually was abolished by the Court of Appeals also
21	was a problem at a time when obviously information and
22	transmission switching of signals were melding together.
23	Now, here is one of the more interesting -- I
24	mentioned earlier that the estimates, at least one
25	commercial estimate of what cellular technology was

For The Record, Inc.

29

1	going to do back in the 1980s was seriously wrong.
2	At the very time that the AT&T case was brought,
3	the FCC was deciding what to do about the so-called
4	cellular spectrum.
5	It took about another nine years for them
6	finally to have one of these licenses begin to --
7	through one of these licenses for service to become
8	available. There was a lengthy hassle over how not only
9	to allocate the spectrum but how to assign it and divvy
10	it up among players.
11	We know what they did. They decided to have
12	only two licenses -- why is not at all clear -- and to
13	give one of them to the incumbent wireline carrier on
14	the grounds that I suppose that wireless was
15	complementary and not likely to be competitive with
16	wireline service.
17	Obviously one's perspective on that would change
18	over time.
19	So it wasn't until 1983 that wireless service
20	began. This is the time when the consent decree was
21	just going into effect, after it had been negotiated.
22	And in the negotiation of the consent decree, the Bell
23	companies were allowed to keep one of the wireless
24	licenses.
25	In retrospect, wireless became the most serious

For The Record, Inc.

30

1	competitor for a long period of time. Cable and VOiP
2	may now take its place in the future.
3	And it was certainly a mistake to do that. But
4	the bigger mistake was only to assign two bands to
5	cellular service.
6	It wasn't until we ran a huge federal budget
7	deficit and the Congress decided we needed to raise
8	money through spectrum auctions that we began to get
9	more spectrum allocated to, more and more licenses
10	awarded for cellular service.
11	And, of course, starting about 1995, 1996, the
12	new PCS cellular licenses were bid on and began to
13	operate, and we went from two carriers to six national
14	carriers over a period of time through a contorted
15	process I won't bore you with right now, because the
16	stuff was licensed on a local market by local market
17	basis rather than national basis.
18	But the important message here is what drove
19	competition starting in the late '90s was wireless, and
20	particularly long distance competition in the '90s, and
21	now I would even argue competition for the local access.
22	What I show you here is a chart in which the top
23	red line shows what we would have expected interstate
24	terminating switched access minutes to look like given
25	what was happening to prices and GDP, and the dotted

For The Record, Inc.

31

1	line below, what actually happened to wireline
2	terminating and interstate access.
3	The gap that opens up there is primarily due to
4	wireless. That is, wireless began to take a very, very
5	large share.
6	Interestingly enough, this whole thing developed
7	because of the development of these national plans which
8	most of us have called from anywhere to anywhere. They
9	were introduced first by AT&T, still the largest long
10	distance player.
11	They cannibalized their own business with this,
12	because then ever other cellular company had to follow
13	in the next year. And today, of course, we have not
14	only a proliferation of these plans, but the plans also
15	allow zero per minute calling in nighttime and on
16	weekends.
17	This number, which goes through 2004, is
18	woefully out of date. I haven't tried to update it.
19	I would think that a very, very large share,
20	overwhelming majority of all interstate long distance
21	minutes now go over wireless. As I said, this may
22	change with VOiP.
23	The price of the decree -- and this is one of
24	the problems of any of these decrees -- is it is
25	difficult to get rid of it.

For The Record, Inc.

32

1	The price for getting rid of it after 10 or 12
2	years was the 1996 Telecommunications Act which -- and I
3	won't go into in great detail; we don't have time -- is
4	subject to its own folly and led to enormous battles
5	between entrants and particularly MCI and AT&T and the
6	regional Bell companies, led to an unbundling regime
7	which got more and more liberal as more and more of the
8	new entrants failed and ended up with a thing called a
9	uni-platform which means the entrant could use all of
10	the facilities of the Bell companies at discounted
11	rates, 50 to 60 percent off retail, through the
12	so-called unbundling process, a provision which was
13	eventually overturned by the Court of Appeals which said
14	it went too far.
15	Also, there was a line-sharing provision which
16	is still in existence throughout Europe and most other
17	countries of the world, Japan, Australia, but which also
18	was ruled as an unjustified extension of the unbundling
19	regime by the D.C. Circuit.
20	In fact, the great savior of folly in U.S.
21	telecommunications was Steven Williams of the U.S. Court
22	of Appeals, now retired or senior status.
23	What happened in the '96 act, we wasted at least
24	$50 billion of investment. Where the stuff went nobody
25	knows. I can't find it on eBay today.

For The Record, Inc.

33

1	MCI and AT&T were forced to enter the arms of
2	Verizon and SBC respectively, not because of the ending,
3	the D.C. Circuit opinions. They would have been forced
4	into it anyway because wireless was eating their lunch
5	so rapidly that their revenues were declining by 10 to
6	15 percent per year.
7	So, after 12 years of the AT&T decree and nine
8	years after the 1996 act, we reverted back to a
9	vertically integrated telecom sector.
10	It was not antitrust, although you could argue
11	that antitrust, certainly the equal access provision did
12	generate the nascent competition early on in long
13	distance services.
14	But we could have gotten there without antitrust
15	had the FCC been on the job or had they realized the
16	benefits of doing this. We led the way with the AT&T
17	decree, and then the rest of the world could follow with
18	their equal access provisions.
19	Today, the local bottleneck is largely
20	irrelevant. And, in fact, despite the rhetoric
21	surrounding it, the local telephone companies are in
22	deep trouble because they do not have a network which is
23	easily capable of delivering high-speed video on demand
24	and are, therefore, having to spend enormous amounts of
25	money to upgrade their networks to catch the cable

For The Record, Inc.

34

1	companies who very easily can offer voice
2	telecommunications services.
3	As a result, what has happened is with the
4	change in the regulatory regime, the incumbent local
5	carriers are now investing enormous sums in their
6	networks, far more, by the way, than the more regulated
7	EU carriers are investing in Europe or, for that matter,
8	more regulated carriers in Australia. Japan still is
9	investing a lot despite a heavily regulated system.
10	I don't know that I can give you general lessons
11	from all this. I think this is sui generis.
12	The AT&T decree may have worked in a narrow
13	sense in that it did introduce equal access into long
14	distance.
15	The cost of the vertical divestiture was
16	extremely high. Was it necessary? I think in
17	retrospect I can say probably not.
18	But I didn't have the foresight at the time to
19	say that. And it is easy enough to go back and be a
20	Monday morning quarterback.
21	But I think it is at least too facile to say
22	this is a decree that clearly was a success and one
23	which we ought to follow in other cases, although one
24	wonders what other industry would offer the same
25	opportunities for this type of vertical divestiture and

For The Record, Inc.

35

1	access.
2	But perhaps Dave has some ideas on that.
3	With that, let me just stop and say one of the
4	things that Cliff Winston and I really wanted to provoke
5	is economists looking at the impacts of antitrust
6	decrees, antitrust policy in general, more empirical
7	work.
8	Our conclusion in our paper wasn't that
9	antitrust is a failure. It was that we have no
10	empirical evidence that it is a success. And that is a
11	serious problem for a policy that's only been in place
12	for 117 years.
13	So we hope to provoke people into doing research
14	and either proving what we have done so far right or
15	wrong, as the case may be.
16	I thank you for your attention.
17	(Applause.)
18	MS. KURSH: Thank you, Bob.
19	I would like to now ask Per Hellstrom to come
20	up. Per is chief of the Unit C-3 at the Directorate
21	General for Competition, European Commission.
22	He is actively involved in the European
23	Commission's case against Microsoft, and we are very
24	grateful to him for traveling across the Atlantic to
25	share his perspectives based on his experiences with

For The Record, Inc.

36

1	Article 82.
2	MR. HELLSTROM: I would like to provide a
3	European perspective to this issue of remedies.
4	I don't really intend to go into detail in any
5	particular case. I certainly don't want this to turn
6	into another hearing on the Microsoft case.
7	I already defended that case once before the
8	court, and we are still awaiting the judgment in that
9	case.
10	But I could mention that as some of you may be
11	aware, the Commission is currently undertaking a review
12	of its policy under Article 82, which is our provision
13	for single-firm behavior.
14	And in addition to that, we are also reviewing
15	our policy as regards remedies, both under Article 81,
16	cartels, et cetera, and Article 82. And we are
17	preparing some internal guidance in this regard.
18	Just a brief overview of the legal framework in
19	Europe, which may be different than the U.S. framework.
20	We have Article 82 of the treaty, which states
21	that abuse of a dominant position shall be prohibited.
22	Now, this provision has direct effect and it can
23	be relied upon by private parties before national
24	courts, and it is the implementing regulation,
25	Regulation 1/2003, which provides the enforcement powers

For The Record, Inc.

37

1	to the European Commission to enforce Article 82 and
2	impose remedies.
3	Remedies are not mentioned in Article 82 itself.
4	In addition to that, we have the case law of the
5	Community Courts which, of course, has dealt with the
6	issue of remedies in some cases, and there is certain
7	decisional practice of the European Commission, the
8	Microsoft decision being one of those.
9	I believe, therefore, that one must separate the
10	issues of the finding of an abuse and the imposition of
11	a remedy, at least in our legal system.
12	Having said that, from the point of view of an
13	enforcement authority, I do share the view that it is
14	important to think about remedies early on in an
15	investigation.
16	But for the purposes of the discussion on
17	remedies, I think also we must assume that we have
18	already a valid finding of an abuse, for example, a
19	refusal to deal, tying, excessive pricing. And certain
20	aspects that could in theory be relevant for the
21	imposition of a remedy, such as the specific character
22	of the market, efficiencies, incentives to innovate,
23	et cetera, may in fact already have been taken into
24	account in the finding of the abuse.
25	Now, regulation 1/2003, that is the implementing

For The Record, Inc.

38

1	regulation that gives powers to the Commission.
2	The context with regard to remedies are Article
3	7, which gives the power for us to take prohibition
4	decisions and impose mandatory remedies; Article 9,
5	which provides for commitment decisions. That is
6	voluntary remedies where it is up to the parties to
7	propose adequate remedies. There is no finding of an
8	abuse.
9	And these decisions are only possible where the
10	Commission does not intend to impose a fine.
11	And then there is also provision for interim
12	measures in cases of urgency in Article 8.
13	I will focus today only on the first one,
14	Article 7, prohibition decisions, whereby the Commission
15	is entitled, where it finds an infringement of either
16	Article 81 or 82, to require the undertaking concerned
17	to bring such an infringement to end. For this purpose,
18	it may impose on them any behavioral or structural
19	remedies which are proportionate to the infringement
20	committed and necessary to bring the infringement
21	effectively to an end.
22	Structural remedies can only be imposed either
23	where there is no equally effective behavioral remedy or
24	where any equally effective behavioral remedy would be
25	more burdensome for the undertaking concerned than the

For The Record, Inc.

39

1	structural remedy.
2	In other words, two types of remedies are
3	possible, behavioral, structural.
4	As the wording indicates, the principles of
5	necessity and proportionality applies. And the aim, as
6	stated, is to bring the infringement effectively to an
7	end.
8	In other words, the Commission has the power to
9	require a company to restore the market conditions
10	absent the infringement and to impose remedies that are
11	necessary to that effect.
12	But, of course, details of any such measures can
13	only be decided on a case-by-case basis.
14	In addition to this, Recital 12 of the
15	regulation provides that with regard to structural
16	remedies, "changes in the structure of an undertaking as
17	it existed before the infringement was committed would
18	only be proportionate where there is a substantial risk
19	of a lasting or repeated infringement that derives from
20	the very structure of the undertaking."
21	Now, if we look at how this framework is applied
22	in practice, I believe that the standard scenario is to
23	have a cease and desist order plus fines.
24	In our terminology, fines are not really
25	remedies, but cease and desist orders are.

For The Record, Inc.

40

1	And by cease and desist orders, I mean an order
2	for the company to bring the abusive behavior to an end
3	and refrain from repeating such act and conduct as well
4	as any act or conduct having the same or equivalent
5	object or effect.
6	This is usually the standard phrase in an
7	Article 82 decision.
8	But a remedy, as we speak about it here today,
9	is an elaboration, then, sometimes an expansion of a
10	cease and desist order, either prescribing a certain
11	action or prohibiting a certain action, leaving the firm
12	discretion on what precisely to implement.
13	Now, how to design a remedy. In theory,
14	remedies or commitments should be effective,
15	proportionate/necessary, clear and precise, cost
16	efficient, transparent and consistent.
17	Of course, in practice, this is quite a
18	challenge.
19	And as mentioned, evidently there is an inherent
20	link between the nature of the infringement and the
21	remedies available to the Commission, and any assessment
22	of the effectiveness and necessity of the remedy must be
23	based on the facts and circumstances of each individual
24	case.
25	But here are some possible criterias, a

For The Record, Inc.

41

1	nonexhaustive list on how to assess the effectiveness of
2	a remedy, questions such as does the remedy lower
3	barriers to entry, is it likely to increase consumer
4	welfare, can it be practically implemented, monitored
5	and enforced and how quickly can the remedy restore
6	competition.
7	One question that has been raised is whether one
8	could foresee a two-step approach with regard to
9	remedies. That is, if the initial remedies imposed are
10	ineffective for one reason or another, could stricter
11	remedies be imposed.
12	Here there may be a difference in our respective
13	legal frameworks. In Europe, in order for us to impose
14	a new remedy if the initial remedy does not work, we
15	would have to respect the procedural rights of the
16	parties, and we would normally have to issue a so-called
17	statement of objections outlining the reasons why a new
18	remedy is required. And that, of course, is a procedure
19	that we know takes time.
20	Of course, this is true for the initial remedy.
21	We must provide sufficient notice in a statement of
22	objection of the remedies foreseen. It would be
23	possible to have some options, some alternatives and
24	allow the company concerned to comment on these.
25	With regard to behavioral remedies, a possible

For The Record, Inc.

42

1	definition of behavioral remedy, "a behavioral remedy is
2	a measure that obliges the concerned undertaking to act
3	in a certain way or to omit certain anticompetitive
4	conduct."
5	Compliance with behavioral remedies usually has
6	to be monitored and enforced. One can classify these
7	types of remedies according to the type of infringement,
8	antiforeclosure remedies, anticollusion remedies or
9	antiexploitation remedies.
10	I will not go into further detail on these now.
11	Common to most behavioral remedies is that they
12	do not change the incentive of the firms to engage in
13	anticompetitive behavior. As a consequence, compliance
14	has to be monitored to avoid circumvention.
15	Monitoring raises various questions as to who
16	should monitor and how. Should it be the European
17	Commission, some sector-specific regulator, competitors,
18	customers, trustees, national courts, or could one
19	resort to some arbitration mechanism, and how should all
20	this be organized.
21	I believe as regards monitoring, the U.S. is
22	probably more advanced in this regard than we are in
23	Europe. And we are currently looking into ways to
24	improve our effectiveness in this regard.
25	A structural remedy is a measure that

For The Record, Inc.

43

1	effectively changes the structure of the market by a
2	transfer of property rights regarding tangible or
3	intangible assets, including the transfer of an entire
4	business unit that does not lead to any ongoing
5	relationships between the former and future owner.
6	After its completion, "a structural remedy
7	should not require any further monitoring."
8	So structural remedies would normally involve
9	the transfer of property rights, some form of
10	divestiture. There should not be any ongoing links.
11	There should be a one-off measure, a clean
12	break, and this remedy should remove incentives and/or
13	the means of a firm to infringe competition law.
14	It may be necessary to have some sort of
15	behavioral flanking measures. Monitoring and
16	enforcement should only be necessary until divestiture
17	is completed.
18	That would be an advantage compared to
19	behavioral remedies. However, structural remedies have
20	rarely been used in Europe under Article 82.
21	However, for the future, the Commission would
22	not hesitate to impose structural remedies when
23	necessary and appropriate. In fact, we could even be
24	obliged to do so, although, of course, again it would
25	depend on the circumstances of each case.

For The Record, Inc.

44

1	I would just like to conclude with a quote from
2	Mr. Charles A. James, the former Assistant Attorney
3	General at DOJ.
4	He has stated in an article that "an antitrust
5	remedy must stop the offending conduct, prevent its
6	recurrence and restore competition. Preventing
7	recurrence must involve proactive steps to address
8	conduct of similar nature. Restoration requires
9	prospective relief to create lost competition and may
10	involve actions to disadvantage the antitrust offender
11	and/or favor its rivals."
12	I believe the Commission would fully subscribe
13	to this statement, although I should add that the
14	Assistant Attorney General also emphasized that the
15	relief, however, must have its foundation in the
16	offending conduct.
17	So in the end, it all comes back to the inherent
18	link between the remedy and the infringement identified.
19	Thank you for your attention.
20	(Applause.)
21	MS. KURSH: Thank you, Per.
22	Tad Lipsky is a partner at Latham & Watkins and
23	a former Deputy Assistant Attorney General at the
24	Antitrust Division.
25	While at the division, he organized and

For The Record, Inc.

45

1	supervised preparation of the merger guidelines and the
2	Antitrust Division's view of the United States versus
3	IBM, among many important antitrust cases.
4	His career has spanned virtually every facet of
5	antitrust law, and he has served in both public and
6	private practice, both here and abroad.
7	Welcome, Tad.
8	MR. LIPSKY: Thanks, Gail and Dan. Your careers
9	have spanned almost every aspect of antitrust law too.
10	I must say you have the organization of these
11	hearings down to an art and science.
12	It is really a great pleasure to be able to
13	focus just on the substance and you are taking care of
14	all the rest.
15	So congratulations. This has been a fascinating
16	set of presentations this morning, and, indeed, the
17	whole record of the hearings has been very interesting.
18	I enjoyed it very much. I am sure it will end
19	up being a very signal contribution to a lot of
20	subjects.
21	The remedies in some respects is really the
22	whole debate.
23	Ultimately every antitrust case comes down to
24	what is the problem and what do you want to do about it.
25	If you don't have the answer to the remedy, you really

For The Record, Inc.

46

1	aren't out of the starting gate.
2	It is interesting when Bill Baxter came to
3	Washington, he had a specific plan for some things he
4	wanted to address. Actually, the AT&T case was not high
5	on his list.
6	His list was to begin an amicus program to
7	articulate to the courts in antitrust cases some
8	economic errors and omissions that he thought were
9	endemic in the precedent.
10	It is interesting we are seeing sort of the
11	final element of that play out just this week with the
12	Legion case.
13	Just about every landmark of judicial ignorance
14	that Bill had identified has now fallen, when you look
15	at the Monsanto and Associated General Contractors and
16	NCAA versus Board of Regents and Copperweld.
17	This is really getting down to the last part of
18	that program.
19	And then he wanted to rewrite what were then the
20	effective merger guidelines, the 1968 sort of Warren
21	court, Lyndon B. Johnson version of merger guidelines.
22	Finally, the third element on his list was what
23	he had seen -- he wanted to do something about judgments
24	and decrees and the way relief was handled in the
25	division, and that meant not only cleaning out a lot of

For The Record, Inc.

47

1	old decrees but, believe it or not, when Baxter came to
2	town in 1981, it was not yet the consistent practice,
3	although it was beginning to be more consistent to have
4	sunset provisions in judgment decrees and in consent
5	decrees entered by the Antitrust Division.
6	I think both at the Commission and the division
7	that is now pretty much uniformly the practice.
8	He abolished the judgment enforcement section
9	because he thought it was very pernicious to have a
10	separate judgment enforcement section which discouraged
11	connecting the theory of remedy to the theory of relief
12	sought in a case.
13	Bill had many memorable phrases, but his way of
14	summing up this problem was to say of the division
15	litigators, he said "Everybody likes to catch them, but
16	nobody wants to clean them," by which he meant if you
17	weren't willing to clean the fish, then you probably
18	shouldn't be fishing to catch it either.
19	My presentation is really in two parts. One is
20	talking about essential facilities and mandatory access,
21	because that is such a hot part of the remedies debate
22	in the context certainly of Section 2 cases,
23	monopolization cases.
24	But it is really the way of illustrating what I
25	think is a fundamental point that is sometimes lost in

For The Record, Inc.

48

1	debating the specifics of particular cases and I think
2	needs to be emphasized.
3	Perhaps not a Baxterian phrase, but my phrase to
4	capture the issue is no sense pretending.
5	If your image of the way an industry should work
6	in a modern capitalist competitive economy is that there
7	should be a number of competitors vying for advantage to
8	supply products and services that meet demand, there are
9	some industries where you are not going to have
10	multiple -- by virtue of the cost structure or some
11	other almost element of the technology or the market,
12	you are not going to have multiple competitors.
13	This is where the essential facility doctrine
14	really starts, from an implicit recognition that if you
15	have something that meets the essential facility
16	definition and it also is something that other
17	competitors cannot practically duplicate -- is I think
18	the phrase from the seminar cases -- what you have is a
19	classic declining cost industry where you simply are not
20	going to be able to structure it and expect optimal
21	results on a competitive basis.
22	You will have to consider the viability of
23	regulatory alternatives, price limits in the framework
24	of utility regulation or some other kind of public
25	intervention, and that puts you kind of in the space

For The Record, Inc.

49

1	where you have essentially got a fundamental departure I
2	think from the antitrust vision of the way an industry
3	is supposed to operate and you need to consider whether
4	you can even attack the problem with an antitrust-like
5	remedy, be it vertical divestiture or whatever, or
6	whether you need a regulatory scheme.