UNITED STATES FEDERAL TRADE COMMISSION

Agency decision

Ask Donna

What actually matters in this document.

Text

UNILATERAL EFFECTS ANALYSIS

AND LITIGATION WORKSHOP

U.S. Federal Trade Commission

601 New Jersey Avenue, N.W.

Conference Center

Washington, D.C.

20001

Tuesday, February 12, 2008

9:00 a.m. to 5:00 p.m.

Reported by:

Susanne Bergling, RMR-CLR

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

PARTICIPANTS

Introductory Remarks:

Deborah Platt Majoras

Panel 1:

Andrew I. Gavil

Janet L. McDavid

Robert Willig

David P. Wales, Moderator

Panel 2:

Jonathan B. Baker

Kathryn M. Fenton

Richard G. Parker

Daniel M. Wall

Jeffrey Schmidt, Moderator

Panel 3:

Hon. Douglas Ginsburg

Hon. Diane Wood

Michael J. Bloom

Richard Liebeskind

William E. Kovacic, Moderator

Panel 4:

William Baer

Susan Creighton

Richard Rapp

Constance Robinson

J. Thomas Rosch, Moderator

Panel 5:

Orley Ashenfelter

Dennis Carlton

Carl Shapiro

Joseph Simons

Michael R. Baye, Moderator

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

CONTENTS

Introductory Remarks.............................

Deborah Platt Majoras............................

Panel 1..........................................

Andrew I. Gavil..................................

Janet L. McDavid.................................

Robert Willig....................................

Discussion.......................................

Panel 2..........................................

Jonathan B. Baker................................

Kathryn M. Fenton................................

Richard G. Parker................................

Daniel M. Wall...................................

Discussion....................................... 107

Panel 3.......................................... 129

Michael J. Bloom.................................130/157

Richard Liebeskind............................... 143

Discussion....................................... 167

Panel 4.......................................... 178

Discussion....................................... 181

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

Panel 5.......................................... 227

Orley Ashenfelter................................ 244

Dennis Carlton................................... 231

Carl Shapiro..................................... 237

Joseph Simons.................................... 254

Discussion....................................... 258

Conclusion....................................... 284

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

P R O C E E D I N G S

INTRODUCTION

MR. SCHMIDT:

Good morning.

I think we are

going to try to start the program.

Workshop on Unilateral Effects.

Director of the Bureau of Competition, and we are very

glad to have you here today.

about this program.

workshop is the brainchild of Chairman Majoras, and it

represents the best of the FTC in trying to better

understand some of the important competition policy

issues that we face.

Welcome to the FTC's

I am Jeff Schmidt, the

We are really excited

As some of you may know, this

I have the chore of doing a couple housekeeping

tasks here, so if you will indulge me as I go through

this to make sure that I have covered the requirements.

I think the -- let's see, the first thing is I have been

asked to remind you that the agenda today is a full one,

so that if you can try to be back in your seats by the

time lunch is over with and breaks are over with, we can

hopefully stay on schedule.

And I have also been asked to ask you to use the

side doors instead of the center doors, for reasons that

are not particularly clear to me.

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

Secondly, if you will turn off the ringers on

your cell phones, BlackBerries, pagers, and the like,

and I will do likewise when I get down from here.

And third, the restrooms are out the glass

doors, past the security desk, and then behind the

elevator bank to the left.

restrooms are located there.

Both the men's and women's

And then fourth, if you do leave the building

during the day, unfortunately, for those of you who are

not FTC employees, you will need to go through security

again.

couple extra minutes to do that.

So, if you can be sure to give yourselves a

And then finally, as a federal government

agency, we do practice certain safety measures.

Probably the most important thing for you to know is --

obviously you know the one exit that you came in through

-- if you need to leave the building in the event of an

emergency.

There will be FTC people who will also be obviously here

and are on site in the event that we have any problems,

but, of course, we are not anticipating that.

There is also an exit immediately behind us.

So, with that, I'd like to welcome the Chairman

of the Federal Trade Commission, Deborah Platt Majoras,

to open our workshop.

(Applause.)

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

OPENING REMARKS

CHAIRMAN MAJORAS:

Well, thank you very much,

everyone.

the morning in Washington, especially on election day.

It is always good to see a robust crowd in

I welcome you to this workshop at the FTC.

many of you know, the FTC has found that when we are

working through particular policy issues, we often find

it very valuable to bring in experts from the outside

who can then, in a public forum, communicate their views

and help us think through the issue.

discussions can take whatever form or length is required

for the issue.

Our public

Just last week, for example, we held a one-day

round table with DOJ to explore our Joint Technical

Assistance Program in the international arena.

about a year ago this week, we had a two-day forum on

the broadband access issue, which has been dubbed Net

Neutrality.

past 18 months, we and DOJ have hosted 29 sessions of

experts discussing the appropriate application of

Section 2 of the Sherman Act to business conduct.

Just

And then, as many of you know, over the

So, today, you have been good enough to join us

as we gather to discuss unilateral effects analysis in

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

merger review and in the litigation context, and I am

pleased to say that we have gathered really a highly

knowledgeable and thoughtful group of panelists, and I

am very grateful to all of you for agreeing to lend your

views.

Back in February of 2004, the FTC and DOJ held a

merger enforcement workshop, which focused on whether

the analytical framework set forth in the 1992

Guidelines, which, of course, had its roots in the 1982

Guidelines, was adequately serving the dual purposes of

leading to the correct decisions in horizontal merger

review and providing reasonably clear guidance to

businesses and their counselors.

The workshop participants generally agreed that,

in fact, the Guidelines framework was serving those

purposes.

a reworking of the Guidelines, but rather, the agencies'

commentary on the Horizontal Merger Guidelines, through

which we explained, by reference to specific cases,

including cases where we had closed the investigation,

how we have applied the Guidelines to actual mergers.

So, borne out of that workshop, then, was not

If you reviewed the section on unilateral

effects, it shows a large number of enforcement actions,

most of which resulted in consent decrees.

little doubt, I think, among antitrust practitioners

There can be

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

that unilateral effects is recognized as a central

antitrust concern, and that the Government has a record

of success in obtaining relief in these cases.

Of course, the record is not perfect.

litigated matters, both the FTC and DOJ have suffered

some losses in differentiated products cases under a

unilateral effects theory.

in the Whole Foods case, the district court did not

grant the preliminary injunction that the FTC sought,

and before that, DOJ lost the SunGard and Oracle

challenges.

cases in which a unilateral effects theory of harm has

been alleged, as in Staples, Swedish Match, and Libbey,

the courts' decisions have really not expressly

discussed the application of unilateral effects theory.

Most recently, for the FTC,

Even when the Government has prevailed in

Now, there may, of course, be no meaningful

pattern in these losses.

likely will lose some cases over time, as only the

toughest cases result in litigation; and try as we do,

we cannot determine with absolute precision on which

side of the line a close case will fall according to a

court.

if we believe that we have the evidence to support our

position that a merger is likely to be anticompetitive.

If we are doing our jobs, we

Still, we cannot shy away from the tough cases

Clearly, though, if you look at the cases and

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

particularly the losses, they do show, I think, what we

experience, which is that there are challenges in

proving a relevant market in which we allege that the

likely harm will arise out of the loss of competition

between two competitors that have served as next-best

substitutes to one another for a significant number of

customers.

Recall that, for example, in the Oracle case,

the Justice Department sought to bar Oracle's

acquisition of PeopleSoft.

incumbent manufacturers in a market defined as

enterprise resource planning system software that

handles human resources management and financial

management systems for customers that made minimum

purchases of $500,000.

of course, argued for a much broader market that

included not just those programs, but also other forms

of ERP programs, as well as non-ERP software solutions,

and would not have limited the market by size of

customer sales.

proposed market expanded the number of market

participants.

These were two of the three

By comparison, the defendants,

So, not surprisingly, defendants'

I am obviously simplifying in the interest of

time here, but there, the court found that DOJ failed to

prove its alleged product market, at least in part

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

because it was not consistent with business delineations

recognized within the industry.

presented testimony from numerous customers that they

might prefer defendants' products over some of the

alternatives, but, said the court, none testified about

how they would respond in actual purchases to a

post-merger SSNIP.

the Government to rely principally on qualitative

materials like market research reports and declarations

from customers and industry consultants.

The Government had

Lack of hard, quantitative data led

The defendants countered with examples of users

that had implemented alternatives to the defendants'

products.

Government had failed to define the alleged, narrow,

relevant market, which meant that the shares that you

then calculate to show concentration levels weren't

correct and that ultimately, the Government's estimates

of competitive effects, based on that market definition,

also had to be disregarded.

Ultimately, the court found that the

Then you go to the SunGard case.

The district

court there rejected DOJ's market definition in refusing

to bar SunGard from acquiring the assets of Comdisco.

These companies, as well as IBM, were in the business of

providing shared hot-site services which are backup

computer centers that you use in the event of a

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

disaster.

shared hot-site services for customers with mainframe

and midrange computer processing centers.

contended that there were a lot of alternatives to these

that customers could and did turn to to safeguard

themselves in the event of disasters.

The Government alleged a market that was

Defendants

Both sides offered customer testimony to support

their contentions, but there the court rejected the

customer testimony, finding that both sides were

engaging in cherry-picking sampling and that neither

side's witnesses were representative of all existing and

future customers.

relevant market that was neither the narrow market that

DOJ had alleged or the broader market that the

defendants had alleged.

market somewhere in between.

Ultimately, the court found a

In fact, the court found a

And finally, if you look at the Commission's

challenge to Whole Foods' acquisition of Wild Oats, the

court there rejected the contention that the relevant

market was the premium natural and organic supermarket.

There, the Government presented not only economic

evidence but evidence that was taken from the parties

themselves that, in fact, showed that the two were

uniquely close competitors.

Whole Foods and Wild Oats competed at a certain level

There was no doubt that

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

with other supermarkets, and we never denied that, but

staff presented evidence that the companies believed

that the other was a uniquely close competitor, and

thus, made decisions on that basis; and as the Whole

Foods CEO told his board in justifying the transaction,

that the acquisition would eliminate Wild Oats as a

platform for conventional supermarkets to get into the

organic market segment, and the entry through that

avenue would be only a threat to his market position.

And in addition, after paying a premium for stores,

Whole Foods made clear it had the intention to close

dozens of stores and to scrap plans to build new stores.

Of course, the district court did not see the

evidence there as we did and concluded that we were

wrong about what constituted the relevant market, and

that case is now on appeal.

Don't get me wrong.

The courts play an

absolutely critical role in U.S. merger enforcement.

Indeed, almost uniquely so if you look at our courts'

role in comparison with many courts around the world.

And after every litigated case, it is very important

that we carefully evaluate the courts' decisions, our

own analysis, and our evidentiary presentations.

You know, the fact that litigated cases happen

so infrequently -- indeed, the three cases litigated by

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

the FTC over the past year were virtually unprecedented

over the past couple of decades.

haven't litigated so many cases in a year.

makes it all the more important that we learn from each

and every court decision.

The agencies just

So, that

In addition, because most merger decisions are

not litigated, we have a great responsibility to ensure

that we are basing those decisions, most of which result

in consent decrees, on solid analysis which would be

supportable in the courts if litigation were necessary.

And if we lose, it is essential that we take a critical

look at our legal analysis and presentation to

determine, to the extent we can, how and why we were

unable to convince the court of our position.

In this regard, I am very proud of the

debriefing efforts that are being undertaken and have

been for the last six months within our agency among the

economists and the lawyers to think these things

through, and today's workshop is another step in our

process.

but given the human limitations on objectivity, we may

be so close to a case or an approach or a set of

strategies that our own introspective evaluation is just

simply not enough.

We can identify ways to improve internally,

The workshop combines a lot of our thinking,

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

covering many of the areas that we and others outside

have identified as worthy of discussion.

has market definition, which has been such an important

tool in analysis, become an end in unilateral effects

cases rather than a means to determine if the merged

entity will have the ability to exercise power?

is it because, as Professors Farrell and Shapiro argue

and probably will talk about today in a preliminary

draft paper, the Guidelines have shoehorned unilateral

effects analysis into the traditional market definition

concentration framework that has its roots in

coordinated effects analysis?

For example,

If so,

We will define markets in unilateral effects

cases in problematical ways in litigation, because given

the nature of the analysis of closeness of substitution,

they appear to judges to have been gerrymandered and not

always consistent with our views as consumers; and, of

course, we are all consumers, including judges.

ready to touch the third rail and discuss whether market

definition is necessary in a case in which we can

present direct evidence of competitive effects?

regard, are we just getting tripped up over our own

terminology and our step-by-step analysis, and should we

do a better job of explaining, as I tried in the

Evanston opinion, that in differentiated product

Are we

In that

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

unilateral effects cases, market definition and

competitive effects are simply two sides of the same

coin no matter how we label?

might argue, stick to traditional market definition and

concentration calculations because, while sometimes

imperfect, they provide important disciplines on legal

analysis?

the fact that a huge percentage of mergers we review

have to be analyzed within only 30 days or less,

necessitating that we have to have some tools to be able

to find the right answer quickly?

evidence and how we present it?

reject customer declarations, customer testimony,

parties' unvarnished statements about competition and

mergers in favor of litigation declarations and economic

evidence at different times, all of which, some of us

believe, at least at some points, to be very important

evidence in these cases.

Or should we, as some

Should our thoughts on this be influenced by

What about our

We have had judges

Are we moving toward a system where fancy

econometrics will win the day, much like we hear about

jurors who have seen so much CSI and Law & Order on TV

that they insist on fancy DNA or fingerprint evidence in

order to find guilt in a case?

noneconomic and economic evidence are most probative in

these cases, and how does our answer vary by factual

What types of

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

conditions, where we have dynamic versus static markets;

if we have industrial products cases versus retail

cases, direct to consumer?

How do we handle new economic learning when we

go in to court?

not a static discipline, and we want to learn as the

economics develop.

litigation standpoint?

experts?

especially if the market definition -- and you heard

some of the ones that I mentioned in some of these

cases -- are just simply not intuitive to us as

consumers?

This is very important, because ours is

So, how do we handle that from a

How important are industry

And how can we best tell the story to a judge,

Now, later today, I am very excited that we are

going to have a mock closing argument over a

hypothetical ice cream merger, and as you will see from

the facts there, the Government in that hypothetical

case alleged that superpremium ice cream is a separate

market from other types, with the defense taking the

position that ice cream is ice cream.

the economics and facts are not necessarily completely

in alignment with what our intuition might be.

panel will provide us with really an exceptional

opportunity to hear how two experienced judges go about

weighing the often complex and contradictory testimony

As we will see,

So, this

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

in economics, which is typically presented in an

antitrust merger case.

So, with that, I would like to thank you all for

being here to discuss with us this important topic, and,

again, many thanks to our panelists who have agreed to

be here with us.

introduce to you, to begin the first panel, David Wales,

who's the Deputy Director of the Bureau of Competition.

I will stop now, and I would like to

(Applause.)

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

PANEL 1:

FOUNDATIONS OF UNILATERAL EFFECTS THEORIES:

CORE FEATURES, ECONOMIC BASES,

AND POTENTIAL GROUNDS FOR ATTACK

MR. WALES:

Great.

Thanks a lot, Debbie.

We are, to reiterate, very excited today about

our various panels, and I personally am very excited

about this panel.

participants and hopefully we will have some great

dialogue.

I think we have some great

The way we would like to kick it off is just to

talk about some of the foundations of unilateral

effects, some of its core features, economic bases, and

potential grounds for attack, and other general topics

to set up some of the additional discussions that we

will have.

The format is going to work this way:

Each of

the three -- now three -- panelists will have brief

presentations to talk about some of the issues they

think are important, that they want to convey, and then

what we would like to do is open it up to discussion,

hopefully get an active discussion as to some of these

issues and drill down a bit further on some of the key

points.

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

So, with that I am going to go ahead and give a

brief introduction of the panelists, and then I am going

to ask them to go ahead and start their presentations.

First off, we have, all the way down at the end,

Andrew Gavil.

University School of Law.

He has been a member of the

Howard faculty since 1989.

Prior to joining the

faculty, he practiced antitrust law and commercial

litigation with law firms in Chicago and Denver.

He is

the lead author of Antitrust Law in Perspective:

Cases,

Concepts and Problems in Competition Policy, and is

currently at work with the co-author, Professor Harry

First, on Microsoft and the Globalization of Competition

Policy:

received the Warren Rosmarin Award for Excellence in

Teaching and Service at the Law School and serves as a

faculty advisor to the Howard Law Journal.

Professor Gavil teaches law at Howard

A Study in Antitrust Institutions.

Next up we have Robert Willig.

In 2004, he

Professor Willig

teaches economics at Princeton University.

He's a

former supervisor of economics research at Bell

Laboratories.

of Policies Affecting Prices and Products, and

Contestable Markets and the Theory of Industry

Structure, and co-editor of The Handbook of Industrial

Organization and Can Privatization Deliver?

He is the co-author of Welfare Analysis

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

Infrastructure for Latin America, and numerous articles.

A fellow of the Econometric Society, he has served on

the editorial boards of the American Economic Review and

the Journal of Industrial Economics.

Antitrust Division in the U.S. Department of Justice as

Deputy Assistant Attorney General for Economics.

Finally we have Jan McDavid.

He served in the

She is a partner

at Hogan & Hartson here in D.C.

She focuses primarily

on antitrust and trade regulation litigation and

counseling.

Antitrust Section of the American Bar Association,

including Chair.

Council of the U.S. Chamber of Commerce, and has served

on antitrust task forces with the U.S. Department of

Defense.

and articles involving antitrust, including the

Antitrust Evidence Handbook, Mergers & Acquisitions, and

Antitrust & Trade Associations Practice Guide.

Ms. McDavid's recognition includes The Best of the Best

Competition and Antitrust Section; Legal Times of

Washington Top Antitrust Lawyers; The International

Who's Who of Business Lawyers; and Guide to the World's

Leading Competition Lawyers.

She has served in multiple positions of the

She also is a member of the Antitrust

She is the author or co-author of many books

We are thrilled to have each of you here today.

With that I think what we would like to do is

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

start off, Professor Gavil, with your presentation.

take it away.

PROFESSOR GAVIL:

MR. WALES:

PROFESSOR GAVIL:

So,

The slides?

Yes.

Good morning, everyone.

I am

delighted to be here, and I thank Chairman Majoras and

Andrew for inviting me to join you.

To start off our first panel, I was asked to see

if in about five or seven minutes I could sum up the

history of unilateral effects.

that.

So, I will try and do

I thought that in just a few slides I would talk

a little bit about the roots of unilateral effects

doctrine, both legal and economic, and how it fits into

the larger picture of merger analysis.

thinking about various phases we have gone through in

terms of merger enforcement analysis.

That got me

I start with a hypothesis, and it was really

late last night when I typed this, so maybe it should

have a question mark at the end.

my hypothesis, so I will pose it more so as a

question -- a possible hypothesis.

I am not sure this is

In a sense, unilateral effects is both the

oldest and the newest theory of anticompetitive harm for

mergers.

The underlying legal and economic theories are

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

neither novel, nor new.

theory has certainly been refined; it has been

elaborated.

will talk a little bit about that, which have clearly

been aided by technology and there is increased access

to data, which also, aided by technology, has been very

significant.

mind, and as Chairman Majoras already put it for us, is

why has the contemporary theory of unilateral effects

proven to be such a difficult sell in the courts?

What is newer?

Well, the

There are new empirical techniques, and we

But the question, of course, on everyone's

The basic larger idea of merger to monopoly, of

course, is original to the Sherman Act.

quotation from Hans B. Thorelli, Federal Antitrust

Policy:

Here is a

That "Sherman" -- talking here about John

Sherman -- "wanted the bill to cover the great

industrial trusts proper as well as mergers and other

tight combinations when of a monopolistic nature there

can be no doubt."

So the idea that we should prohibit mergers to

monopoly is a very old idea in antitrust.

It was

supposed to be covered by the Sherman Act.

In many of

the early merger cases that came out of the great merger

wave, Northern Securities, U.S. Steel, although of

varying success in terms of enforcement, the basic

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

theory was merger to monopoly, and the trusts themselves

were combines.

They were viewed as mergers to monopoly.

The 1950 amendments ushered in the non-

monopolistic merger period, somewhat in response to the

Columbia Steel case of 1948, although there are other

factors as well.

these merger challenges from the twenties to the

forties.

a different set of concerns.

altered the focus from a focus on merger to monopoly to

what we might call nonmonopolistic mergers.

The Government was losing a number of

Congress decided to step in.

They clearly had

They broadened out and

We might also call these the wilderness years,

as the anchor, even in early thinking about merger to

monopoly, was a little bit more clear than what happened

in this period.

"trend towards concentration," a concept which is

typified by cases like Brown Shoe, Von's, and Pabst, and

which we now teach against in casebooks, toward the

structural approach, and the general concerns it raised

about market shares that were obviously elevating.

was the idea of making predictions from market structure

that took form in the Philadelphia National Bank

presumption, and, of course, was reflected in the first

Merger Guidelines in 1968.

There was an evolution from emphasis on

Here

From 1968 to 1992, there was an effort to better

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

define what the true anticompetitive theory was, and in

a sense this period led to a commingling and a fusion of

two competing traditions.

tradition going back to the 19th Century in economics,

and the other was the structural presumption, which had

developed in some of the writings on industrial

organization economics in the 1950s.

One was the oligolopy

If you go back, as I did, looking at Stigler and

Posner and Bork and contrast them with Kaysen and Turner

(1959), you really see these two very different sets of

ideas competing for influence in terms of merger policy.

Their first offspring was the coordinated effects theory

in the 1982 Guidelines and the way the Guidelines are

structured.

Shapiro explore in their paper, I will mention that a

little later on.

that combine pieces of different theories I think is one

of the issues that is going to emerge today as

important.

are reflected in different pieces of the Guidelines, and

like a puzzle where the lines between the pieces are

still very defined, they do not always quite fit

together very well, and sometimes they can even work at

cross-purposes.

This is a point that Joe Farrell and Carl

The attempt to structure Guidelines

We have different intellectual thoughts that

From the mid-1980s to the present, there was

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

something of a reintegration of the pre-1950 and

post-1950 models.

mergers are reintegrated in the Guidelines.

Coordinated

and unilateral effects are both introduced.

Both, of

course, have roots in oligopoly theory, but both are

still tethered to the structural concepts in the

Guidelines.

Monopolistic and nonmonopolistic

For more sources on this history of unilateral

effects and its roots, I just cited a few of the

articles here on the slides, all of the authors being in

the room, Baker, Willig, and Denis, all go through some

of these issues of the intellectual roots of modern

unilateral theory.

Well, where do we go from here and what is the

discussion about today?

I think one issue that I wanted

to put out is, how do we relate developments in

unilateral effects to the larger context of modern

antitrust?

I wanted to put the idea out there.

unilateral effects parallels, in a sense, the tension

that now exists in Section 1 between actual effects and

the quick-look doctrine on the one hand and

circumstantial effects under the Sherman Act.

Coordinated cases tend still to be structural in some

sense, economic, and more sophisticated in others.

And this I am not quite sure I believe, but

Coordinated versus

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

But

to the degree they are relying on creating a

circumstantial, predictive case for coordinated effects,

they are more like the circumstantial approach to merger

analysis.

I tried to give a new name -- I don't know if it

will work or stick -- but unilateral effects is more

akin to "predicting actual effects" based on empirical

evidence, and in that sense, it really can be located in

the circle with cases like NCAA and Indiana Federation

and California Dental and Polygram, cases that try to,

as the Chairman was talking about earlier, try to look

at actual effects and market definition, market power,

as flip sides of an issue.

As the court said in NCAA and again in Indiana

Federation, traditional market power analysis involved

defining a relevant market, calculating market shares,

and predicting market power and consequence

anticompetitive effects from large and durable shares.

The Court has held, however, that doing so was just a

surrogate for actual anticompetitive effects.

have the actual anticompetitive effects, you shouldn't

need to do those things.

When you

The tension about that has arisen with respect

to such actual effects cases is similar to the tension

that exists now around unilateral effects.

Concerns

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

about the reliability of actual effects evidence have

also caused some push-back in non-merger areas.

productive step we could take would be to get merger

analysis, instead of in its own pigeonhole, relocated in

the larger picture of what is happening in antitrust.

So, one

The irony of precision -- last slide here -- why

are unilateral effects cases a tough sell in court?

economists, there is the appeal of empiricism.

very appealing.

quotation out of one of Jon Baker's articles:

For

They are

They -- based on data -- I pulled this

"[i]f the facts support a unilateral theory, it

is clear as a matter of economic logic why the

particular merger would likely lead to higher prices."

This reminded me a little bit of the language in

Polygram where the FTC talked about anticompetitive

effects being "intuitively obvious" based on economic

analysis.

decision-makers?

But what is the challenge for

Why the resistance?

Well, in a sense, the models can be more complex

than the traditional PNB presumption.

This is somewhat

ironic since the models were designed to yield a greater

degree of precision, a greater degree of understanding,

yet the models themselves are more complex.

presumption was by comparison easy, like per se rules,

like other burden-shifting devices.

The PNB

It did not require

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

a lot of understanding to say:

that's a lot!"

"40 plus 20 is 60.

Ooh,

Empirical evidence also may be confusing when

combined with traditional structural evidence.

appear highly dependent on assumptions, and, therefore,

subject to manipulation if the assumptions change.

can be a little bit more rigorous in theory than

practice.

And I think there is a larger issue, one that David

Meyer talked about in a speech last fall.

whether we like it or not, at something of a historical

moment in antitrust, where courts are proving very

skeptical about antitrust cases, and unilateral effects

has run into that skepticism as it tries to develop and

evolve in the courts.

It can

Sometimes the data do not match the theory.

We are,

Those are my opening comments, and I will turn

it back over to the panel.

MR. WALES:

Next we have Professor Willig with some brief

Great.

Thanks, Professor Gavil.

remarks.

PROFESSOR WILLIG:

Brief?

I face an interesting challenge.

I was asked to

cover the Merger Guidelines, a short overview to be

sure, unilateral effects therein, the history of

antitrust, and the economics of unilateral effects, and

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

I wasn't given five to seven; I was given three to five.

MR. WALES:

PROFESSOR WILLIG:

I lied.

can talk really fast.

With another cup of coffee, I

So, who's got the coffee for me?

Elements of the Guidelines in an historical

context:

a relevant market?

algorithms, those of us who love that kind of thing, but

the idea of a relevant market is so simple that I think

we should remember its basic concept all day long

throughout the discussions.

collection of the principal sources of competitive

discipline on the products of the merging firms,

especially the overlapping products of the merging

firms.

First and foremost, relevant market.

What is

I know we talk about all the

A relevant market is a

If you collect all the sources of competitive

discipline and you put them all under a single source of

control, then you should be seeing some elevation of

monopoly power, and hence, the hypothetical monopoly

test as the way to make sure that you have got all of

the principal sources of competitive discipline

identified and collected in the relevant market.

idea of it is simple.

just the way to make sure that you have actually got

market power there collected in these various sources of

The

The hypothetical monopoly test is

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

competitive discipline.

This was the idea of the 1982 Guidelines, along

with a way of counting concentration within a relevant

market.

away from the technocratics, the Herfindahls and the

like -- remember when that was a bizarre thing?

remember that.

feel like it was yesterday and I was already old when

these things happened.

dream of mine.

great, exactly the way he looked -- God knows when.

improvement, but no change.

(Laughter.)

PROFESSOR WILLIG:

The concentration question, again, taking it

I mean, I hate to be an historian and

That is sort of a dangerous

Never mind how Jon looks.

He looks

So, why do we count

concentration and change in concentration?

Well, a

relevant market is a place where a hypothetical

monopolist could or would exercise monopoly power.

change in concentration and the level asks, well, what

does the merger do to bring us to the status of that

hypothetical monopolist?

actually bring us to that hypothetical monopoly?

goes hand in glove with the idea of the relevant market.

The Herfindahl is a very clever way to measure

concentration.

collect share data and see how concentrated they are.

The

How close will the merger

It is nothing but an arithmetic way to

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

Now, everybody keeps saying today -- and I have

heard this before as well -- that the 1982 Guidelines

are all about collusion, coordinated effects, as we

would call it today.

there.

was being published as coming right out of a Cournot

model.

Waterson, and, in fact, Ordover and I were asked to

write a review of those '82 Guidelines.

consulting on the Division on them when they were being

written with Larry White, and in '83, Ordover and I

wrote, "Why do they keep using the word collusion in the

Guidelines?

models like Cournot with what we would call today

unilateral effects."

a lack of language, than a distortion of the ideas.

obviously did better a decade later by looking it in the

face, but to say that the '82 Guidelines were really

about collusion I think is a grave intellectual error if

we are doing history, and that was my assignment.

Hey, I was there; Larry White was

It turns out that the Herfindahl Index, by 1982,

You all remember this, economists Cowling and

I had done some

They are actually talking about oligopoly

I think it was more a mislabeling,

Now we move on to the current Guidelines --

hopefully still current -- and we have coordinated

effects, which we are not talking about today, and we

have unilateral effects, and I'd like to highlight three

different cases of unilateral effects that are squarely

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

in the Guidelines, and here, too, I am worried that we

are losing track about which one it is that we are

speaking of.

First of all, unilateral effects apply in the

Guidelines to the case of "homogeneous products,"

commodities in the common parlance.

this a market in which firms are distinguished by their

capacities rather than by the characteristics of their

products, because they are all basically the same;

hence, homogeneous products.

totally good sense in a market of homogeneous products.

The economics of it are very simple.

The Guidelines call

Unilateral effects make

The idea is that if a firm gets bigger in a

space of homogeneous products, then it has got a bigger

base of capacity on which to enjoy a price rise, and so

a big merger tends to enhance the incentives of the

newly merged firm to cut back on output so as to push

the price up, because now, it has got more capacity on

which to enjoy the positive profit effects of that price

rise.

Not elaborate, not fancy, not about merger

simulation models, although we have lots of analytics to

handle that if we want to, but it is not what we are

usually talking about on a day like today, but it is

still unilateral effects.

So, I think we need to

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

sharpen our language away from just unilateral effects

to unilateral effects in markets where the products are

importantly differentiated to separate out the case of

the commodities.

Second of all, within the category of

differentiated products, there is a main case really in

the Guidelines where the differentiated products are --

I am calling it today generally differentiated.

Baker and I and Paul Denis debated this stuff for much

of two years together.

products are ones that compete with others in the

relevant market, but kind of generally, without any

specific product-to-product relationships.

Think about cold remedies.

Jon

Generally differentiated

I mean, does anybody

really know what the subcategories are of cold remedies?

Everybody's got their favorites, and yet each cold

remedy basically competes with all the other ones.

Maybe a pharmacologist would know the difference, but we

consumers sure don't.

are all kind of mushed together in one big pot, no

specific competitive relationships.

Or midsize cars, you know, they

Well, in a market like that, it makes sense to

think that the share of a product is indicative of its

competitive significance as an alternative to whatever

your favorite product is; that shares really connote

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

competitive significance, because the competitive

relationships are general throughout the marketplace.

That is the lead case of differentiated products under

the Guidelines, and there, relevant market makes just as

good sense as it does for a homogeneous product industry

that collects all these products that interact

importantly; concentration makes sense as a measure of

significance, and off we go.

There is a lot of economics lying behind this.

The Logit model of demand handles this.

We all grew up

on the CES Utility model of monopolistic competition,

and in markets like that, this is exactly the kind of

interaction among the products.

differentiated products stuff.

This is really classic

What we are all getting confused about is the

third case where the competition among differentiated

products is not general; instead, it is local, and where

differentiation is local, market share is not indicative

of competitive significance as a matter of substitution

for any other product.

products, no.

Some products yes; other

Think about Toyota Camrys.

They are very

successful cars, and yet they are in no way interesting

substitutes for the BMW drivers in the crowd.

maybe an Audi with a low market share is a much closer

Instead,

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

source of substitution for the BMW than would be the

best-selling Camry.

are discernible; they are different; people recognize

them as such; and they drive the importance of different

substitution relationships.

of unilateral effects.

talking about the third one, and I think it would really

help to clarify that in our discussions.

So, here, product characteristics

So, three different kinds

Today, we are really only

When we have localized effects, we are going to

have small, narrow relevant markets.

You know,

Bimmer-oriented relevant markets instead of all cars or

all midsize cars, and what we are hearing is all judges

who I guess do not drive Bimmers find it a little bit

harder to understand.

A proposal I would make today -- and I am not

going to wait for the question, I just want to slip it

in -- the proposal is that we accept the idea that

markets can be narrow where competition is localized --

bite that bullet -- and accept the idea that sometimes

the best evidence for what constitutes the true, narrow

relevant market is not our normal kind of intuition

about, "Oh, a car is a car; a grocery store is a grocery

store; a stationery story is a stationery store," but we

allow ourselves, where appropriate and where the

evidence is there, to deduce market definition from

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

evidence about competitive effects; that after we look

at the Staples/Office Depot evidence, that where there

are only two superstores instead of three, prices are

higher, that teaches us that the office superstores are

not in the same relevant market as your corner

drugstore, which I would have thought intuitively, but

the evidence proves that is not true.

proves that, indeed, the relevant market is office

superstores.

sources of evidence, but the statistics that show that

are our best evidence for market definition.

The evidence

I wouldn't have known that through other

Why shouldn't we allow markets to be defined

using best evidence?

kinds of data, that would be our best evidence.

not that markets are irrelevant.

should be willing to test them and to prove them,

sometimes using the same kind of information that we use

for competitive effects, where we have such solid

evidence.

And in cases where we have those

It is

It is just that we

It is not wrong in Whole Foods for the judge to

be debating what the relevant market is -- all

supermarkets or just organically oriented ones.

very much the right question, and I think the judge was

on the right beam in trying to figure out what the best

source of persuasive evidence was.

That is

I don't know what

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

the right answer is either.

would have been all over the lot just like the judge

was.

hard question.

aware of those data, but, I mean, maybe you are right.

But I think the judge was grappling with the right

question, and why not allow competitive effects and

natural experiments to be part of the evidence that does

drive a determination of the relevant market, along with

competitive effects?

with that.

So, if I were the judge, I

I don't know if it was a wrong process.

Maybe the FTC knows better.

It is a

I am not

I think there is nothing wrong

I think there is a danger in eliminating the

idea of a relevant market, because not forcing ourselves

to actually enumerate, out loud, all the sources of

important competitive discipline creates the danger that

in our weaker moments, when we are not absolutely on our

game -- and I know mostly we are in this room, but

sometimes we are off our game -- when you are on the

other side of me, for example -- that under those

circumstances, you should be impelled by the process to

enumerate all of what you think are the important

sources of competitive discipline, and the process of

relevant market is the force that makes us do that.

Just saying, "Oh, it is obvious that these two products

are the closest substitutes, end of story," is a

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

dangerous way to lead our process as far as the law is

concerned.

Thank you.

MR. WALES:

We are now going to turn to Jan McDavid with her

Thank you, Professor.

opening statement.

Jan?

MS. McDAVID:

In recent years, as we have been talking about,

Thanks, David.

the agencies have increasingly relied on unilateral

effects theories.

economists in the room, can tell us whether the

techniques underlying these theories are appropriate and

debate which theory is appropriate in a particular case.

I am not an economist; I don't play one on television.

I hire people like Bobby for that.

Other panelists, and especially the

Instead, I'd like to discuss these issues from

the perspective of an antitrust practitioner who has to

explain them to business people who are making decisions

about potential transactions and who interact with the

staff of the agency about particular transactions.

Now, it has always seemed logical to me to

consider whether a merger that eliminates direct

competition between the merging parties substantially

reduces overall competition within the meaning of

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

Section 7.

very common sense notion that a merger is likely to have

more a harmful competitive effect if the merging parties

are particularly close competitors.

Unilateral effects analysis is based on the

The most obvious example, of course, is a merger

to monopoly in which there is no competition remaining

following a transaction.

transactions in which some rivals remain could produce

those competitive effects.

won't.

before the courts, is how do you distinguish between all

of these different formulations?

But it also seems logical that

In other circumstances, they

The question before us, before the agencies and

I have always found that the easiest way to

explain these concepts to business people is the next

best substitutes formulation, and so that is basically

what I have done.

Now, as a Colorado skier, I often use the Vail

case as the paradigm that I walk my clients through in

trying to have them understand competitive effects.

About ten years ago, Vail resorts, which operates both

Vail and Beaver Creek, proposed to acquire the Ralston

resort ski properties in Colorado.

thought Ralston only made dog food will be surprised to

know that they actually operated Breckinridge, Arapahoe

Basin, and Keystone, and did not do so especially well.

Those of you who

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

The Division concluded that there were two kinds

of skiers:

get on an airplane and fly somewhere to ski, and if

prices go up for us, we could go somewhere else.

could get on an airplane to Salt Lake rather than to

Denver if I wanted to go skiing.

what they called the front-range skiers, the folks who

get in their cars somewhere in the Denver metropolitan

area and drive about two-and-a-half hours to a ski area,

and they concluded that that was the market in which

they needed to analyze the effects of the proposed

Vail-Ralston transaction.

There were destination skiers, like me, who

And then there were

The competitive impact statement made it clear

that the Division was applying a unilateral effects

theory to the case.

deterred from increasing its prices at Vail and Beaver

Creek by the fact that skiers could go to Keystone

instead, if prices were to be increased at Vail and

Beaver Creek, or Breckinridge or Arapahoe Basin.

Vail also owned Keystone, Breckinridge, and A-Basin,

they would also pick up the revenues on the sales of

those tickets, and therefore, a price increase might

become profitable.

Before the merger, Vail was

But if

Based on an econometric analysis, using largely

survey data -- and that is a point I really do want to

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

come back to -- and data on margins, the Division

concluded that a price increase of a dollar per ticket

was likely in the event that Vail owned Vail, Beaver

Creek, and Keystone, because Keystone was the next best

substitute.

would fix this problem.

They also concluded that divesting A-Basin

Now, the antitrust agencies' ability to engage

in the type of analysis that they used in the Vail case

or in the other cases we have been talking about has

been made possible by the kinds of rich data sources

that are available, as well as computers.

involving branded food products, for example, IRI and

Nielsen data permit very elaborate econometric models in

which we can actually use transaction data to test these

propositions.

in branded food products are not available most of the

time, and even in branded food product transactions,

they actually focus on competition at the wrong level,

because they are focusing on the prices set by

retailers, not the prices set by the manufacturers of

the food products who are actually engaged in the

merger.

In cases

But the retail scanner data that we have

So, what substitutes for these kind of data are

available and how does the quality of the data affect

the quality of the analysis in which we are engaging?

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

It seems logical to me that differences in the quality

of the data are very likely to lead to differences in

the quality of the economic analysis that is being done

and that use of data that is not reliable may lead to

skewed and unreliable results.

An awful lot of the debate is also about the

kinds of assumptions that are being used, and if you

vary the assumptions, you vary the outcome.

possible, under the Guidelines and under the Commentary,

to find unilateral effects at even low market shares.

Many of us believed there was a 35 percent safe harbor

in the Guidelines, but the Commentary says there isn't.

Where is the right line?

effects predicts some kind of a price increase absent

some significant efficiencies.

We all know how reliable

the efficiency estimates are.

All of this can skew the

outcome in ways that may render the results at least

suspicious and make people skeptical.

It is very

Every model of unilateral

Now, I bring to this process the skepticism that

I also bring to the HHI analysis.

The HHIs lead to a

mathematical result which looks precise on its face, but

we all know that it varies entirely based on market

definition and market shares, neither of which are very

reliable, and then you just square it and add it up.

So, it all depends on where you start as to where you

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

end up.

For that reason, I rarely try to define markets

in the transactions I am working on.

almost immediately, on competitive effects analysis,

because that is where I have always thought the game was

going to be played.

are a very useful first screen for thinking about the

transactions into which we should start conducting that

kind of elaborate analysis, but they create an

artificial sense of precision where no real precision is

possible, and I am concerned that some of the same

things happen with respect to the kinds of unilateral

effects analyses that we have been undertaking.

I always zero in,

I have always thought that the HHIs

Let's go back to the Vail case as an example.

People who ski in Colorado who probably agree that

Keystone was the most likely next best substitute for

Vail and Beaver Creek, with Breckinridge being a close

second.

survey data would allow you to conclude that prices

would go up one dollar or we would be especially

skeptical that divesting Arapahoe Basin was going to fix

that problem.

I think we would have been very skeptical that

I have never skied at Arapahoe Basin.

It is way

too hard for me.

There are people there who sleep with

their dogs in their Volkswagen buses in the parking lot.

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

It is not a substitute for Vail.

an economic model that suggests that it is.

So why should we trust

So, I think where all of this takes me is that

we have to bring some common sense to these kinds of

analyses, and that is where I am concerned that the

agencies are running into resistance.

they've been doing appears to be gerrymandered or

jury-rigged and doesn't pass the common sense test.

When your judge is someone who's been sentencing drug

offenders in the morning and is handling unilateral

effects analysis in the afternoon, you have to be

conscious of the limitations of your audience.

don't do the math either.

Some of what

They

Judge Wood, who handled the cereals transaction,

brought Fred Kahn in to advise her as effectively her

law clerk when she tried that case, even though she was

a very experienced antitrust lawyer and very good at the

economics.

sorts of problems that we have to be conscious of.

And that is, I think, an illustration of the

So, I would like to use the unilateral effects

analysis as part of a holistic analysis of all of the

evidence.

results with the more traditional models, considering

the company's strategic planning documents; who do they

think are their most significant rivals; what do the

I have always thought we get to pretty good

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

customers say; what is the evidence about entry; is one

of the companies failing; is one of the company's

ability to compete on an ongoing basis impaired in the

future.

the evidence consistent?

place?

Does this tell us an overall story?

Is all of

Does it take you to the same

If that is the case, I think you can be

reasonably confident about the kind of decision you are

reaching.

skeptical, and the agencies will encounter a skeptical

audience in a federal judge.

lessons are things we have to keep in mind as we do

these sorts of analyses.

MR. WALES:

We will kick things off a little bit.

If it does not, then the agencies should be

I think those kinds of

Thanks, Jan.

I thought

I would ask some questions and hopefully get the

dialogue going.

It seems that there is not a lot of dispute that

unilateral effects is a valid theory and one that we

think should be applied in the appropriate cases,

especially in differentiated product merger cases, but

the reality is it has been a tough sell to judges, and I

guess the question is, what do we take from that?

are the reasons why we think that judges are having a

hard time?

Is it the fact that perhaps unilateral

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

What

effects is not a sound theory?

Is it more practical in

the sense that there are assumptions, intuitive

problems?

think the problems are?

PROFESSOR GAVIL:

Are the Guidelines to blame?

What do you

Well, the Guidelines are a

product of a long history and tradition, and again, I

would say that you need to look at it in the larger

context of antitrust.

relevant markets and market definition and market shares

and assumptions that you draw from that, connections

between that and the possibilities of anticompetitive

effects, for a long time.

going to be an easy process, and the evidence is going

to have to be especially compelling.

We have been thinking about

So, shaking that loose is not

I think if something does differentiate Staples,

it is that the evidence was especially compelling.

is difficult from the outside to evaluate how compelling

the evidence is in cases still pending, like Whole

Foods, where we just don't know all of the evidence that

was introduced.

And I think a second part of it is Bobby's

comment that maybe we shouldn't be trying to persuade

anyone to totally let go of that structural tradition.

I combine that with Jan's comment -- this has been true

in nonmerger cases -- when the two kinds of evidence are

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

pointing in the same direction, you are going to have

the strongest case.

Now, that means a lot of work maybe, but when

the direct and circumstantial evidence in non- merger

cases is pointing towards market power, those cases are

pretty hard to rebut.

combination of thoughts here that lead to that

conclusion.

MS. McDAVID:

So, maybe there is this sort of

I think one of the things about

Staples we should remember is that although we had very

complicated economic analysis by Professor Ashenfelter,

there was also some really simple stuff.

higher where there was one firm and prices were higher

where there were two firms than they were when there

were three.

people who don't do the math.

Prices were

That was a pretty simple paradigm for even

PROFESSOR WILLIG:

It seems to me that the basic

thought behind differentiated products or local

competitive effects, the basic thought is totally

intuitive.

breakfast table test at home, which is to say that,

look, it turns out that when my favorite car is being

priced by the marketing people, the first thing they

look to is this closely competing car, and maybe we

actually have evidence from the companies of that or

I mean, it passes my dinner table, even my

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

maybe we can readily deduce that, but as an expert, that

would be my opening line if I am talking to my family or

to a common lay judge, is to say, look, what is keeping

prices where they are today is largely and importantly

competition with this other product, and guess what,

after the merger, that product will be in the same

executive suite, the margin will be just going into the

same pocket as the margin on the BMW, my favorite car,

and so that source of price competition will be gone.

Now, Your Honor, believe me, I have looked at

other possible sources of competition, and there are

other ones, but they are just nowhere near as important

to the pricing of the BMW as that Audi car, and now Audi

and BMW are threatening to merge.

a broader relevant market, I have tabulated all the

other possible sources of competition, and they do have

some effect, but not nearly as important as the effect

that would be lost because of this merger.

about that?

Jan?

MS. McDAVID:

So, I have looked at

What is hard

No, I think that is pretty simple,

Bobby.

By the way, I have always thought that the

Division's case in Oracle made a great deal of sense.

The problem was that the market, as defined, was not

really a product the company sold.

It, therefore,

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

looked jury-rigged, and I think that is just part of the

problem.

It failed the common sense test.

MR. WALES:

What about one of the -- I guess in

the merger commentaries it talks about the fact that you

can have both quantitative and qualitative evidence that

may be probative of the closeness of substitution of the

various products and, of course, the potential

competitive effect.

Is it the case now that you must have

quantitative evidence, despite the fact that the

commentaries talk about how you can have either

quantitative or qualitative information, like business

documents?

judge was more focused on the quantitative as opposed to

the qualitative evidence, where there was some pretty

good qualitative evidence in the business documents.

Obviously in Whole Foods, it seemed like the

MS. McDAVID:

We have to do both.

The reality

is when we are proposing a transaction, we have to do

both.

the matters that I handle before the agencies, I

encourage my economists to share all of their data, all

of their analyses, almost sit in a room with the agency

economist and be as cooperative as possible.

get to the right kinds of outcomes.

what we did in the cruise lines case, and many people

There is no alternative, and, you know, in all of

We will

That is absolutely

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

hold that out as the model.

agency's going to do it.

destruction circumstance.

PROFESSOR WILLIG:

We have to do it.

The

It is a mutually assured

I mean, to me, the

quantification, aside from our satisfaction in using

professional standards as economists, but the

substantive question that has to be addressed -- and

this brings us back to relevant market, I think -- is

suppose that we can all agree, intuitively, that B is

the closest substitute for A, and A would be the sellers

are threatening to merge, but that really is not the end

of the story, nor is it even the end of the story to say

how closely substitutable A and B are, because in many,

many local or bigger markets, there is a C, D, and E

lurking behind A and B.

Those of you who know Princeton, if you get off

Route 1 to make a right turn to come to the campus down

Washington Road, there is a little traffic circle, and

on that traffic circle there is two gas stations, and

they are head-to-head competitors.

literally head to head on the traffic circle.

always use this in class.

stations merge?

they are close substitutes, so wouldn't you bust the

merger right away?

I mean, they are

So, I

What if those two gas

What do you say, class?

You can see

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

So, anybody who says yes never makes it to the

midterm as far as I am concerned.

Half a mile down Route 1, there are five other gas

stations.

merge, we would lose that head-to-head competition, but

it would not be a substantial or it might not be a

substantial change in the state of competition, because

there is all these other gas stations just a half a mile

down the road.

But you know what?

Now, it is true if those two gas stations

This is what scares me about getting rid of

relevant market when it comes to localized competition

among differentiated products.

say, right away, "No, no, we have got to stop that

merger," without asking what else is there right behind

that pair of closest substitutes?

question that the relevant market forces us to answer,

to pick it up, saying, "Well, yeah, there are other

sources of competition, but you know what, they are not

nearly as important."

Half of my class will

And that is the

But we need some quantification to get us to the

ability to conclude whether or not those other gas

stations are closely enough competitive to these two

that are head-on to see whether their merger will

significantly tend to raise price, or whether, instead,

C, D, and E will provide ample competitive discipline to

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

stop there from being a significant price increase

because of the merger.

quantification is necessary.

That means some kind of

When I tell you half a mile, you know the

answer, but when we are talking about cold remedies or

supermarkets of different kinds, we have no ready such

quantification, and now we are into a real debate that

is frustrating a lot of people.

MS. McDAVID:

I do not think it matters what you

call it -- or whether you focus on relevant market or

market shares, what you have to determine are what are

the -- as Bobby put it -- the sources of competitive

discipline post-transaction on the merging parties?

you are going to have to identify them and talk about

how significant they are.

MR. WALES:

And

It seems that judges have had a hard

time, though, in terms of applying the Guidelines and

understanding the difference between identifying that

localized competition that we think matters in terms of

the unique constraint on the merging, differentiated

products, and defining a broader market that might

contain more distant competitive constraints.

need to rethink how the Guidelines work in

differentiated product cases?

MS. McDAVID:

Do we

Well, the Commentary made an

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

effort to do that, but we are regularly reminded that

the Guidelines are not law.

reminded us of that in his Whole Foods opinion.

Guidelines are sources of explanation and an

extraordinarily useful framework for us to use before

the agencies, but fundamentally, they are not going to

bind a court.

think is what you really need.

I think Judge Friedman

So, the

Some explanation, in whatever format, I

MR. WALES:

Would anyone support amending the

Guidelines?

PROFESSOR GAVIL:

The Guidelines have become

kind of a two-edged sword I think for the agencies.

Yes, formally, they are not law.

all state -- not only the Merger Guidelines, but all of

the enforcement agency guidelines -- that they are not

intended to establish a litigation format; they do not

specify burdens of proof.

agencies use them in courts, the degree to which parties

use them and hold the agencies to them, means that they

have become very influential documents in court.

are looked to as demarking lines for burden-shifting

when you look at the steps of the Guidelines.

Guidelines, on their face, would seem to suggest that

you always start by defining a relevant market and

calculating market shares.

Yes, formally, they

But the degree to which the

They

And the

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

So, when you say change the Guidelines, ask

should we change the Guidelines, well, to what end and

with what consequences?

difficult challenge for the agencies to articulate

enforcement standards to two communities.

articulating to the business community their intentions

with respect to enforcement efforts, but then when they

go to court, in part, given the Supreme Court's absence

from mergers for so long, when they go to court, they

are kind of trying to use the cases that are available,

that are the best cases.

Guidelines as if it were law, as if it were their own

law.

I think it has become a

They are

Yet they have to live with the

So, it is a challenging question, what to do

with the Guidelines, and can you fix the problem in

court by changing the Guidelines, by further developing

the theories?

said, when those first '82 Guidelines came out with HHIs

and SSNIP, you know, there was giggling in the room at

the ABA meeting -- "what could this be and what court

would ever do this?"

changed.

Maybe.

Coming back to something Bobby

And with time, that has clearly

So, maybe part of the answer is that changing

the Guidelines could change things, but it may not

change things in the next case or it may take some time

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

until those ideas filter through and gain the confidence

of lawyers and judges as well.

PROFESSOR WILLIG:

Well, let me ask, just to

pose my own question, if you were to think with me that

the best way to go is to deliver the message that the

way to determine relevant market is through best

evidence, which sometimes may be consumer survey -- God

help us -- sometimes through your own stomach as a

consumer, but sometimes through real consideration of

marketing data or natural experiments, like in office

products case, get the message out that we do need to

determine relevant markets, but we can sometimes do it

backwards.

we would do for competitive effects but use that as the

source of best evidence for relevant market.

Sometimes we can do the same analysis that

What is the best way to get that message out?

Is it a revision of the Guidelines?

it next time there is a document that talks about best

practices, that that becomes a prominent example?

in court explicitly that way?

courts better than I, what is the best way to deliver a

message of that kind?

MR. WALES:

Is it a speech?

Do it

Those of you who know

One additional point, is the 35

percent threshold in the Guidelines.

We have seen some

courts reject that, actually in Oracle, there were some

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

pretty negative comments on it; other courts were

willing to accept it as another proxy in the attempt to

measure the closeness of substitution between the

merging products.

The Merger Commentaries talk about it as merely

a screen and not a safe harbor.

place in antitrust cases?

that something we should consider changing?

MS. McDAVID:

Does it still have a

Should we be using it?

Well, if you go back to my common

sense notion, when the agencies challenge a transaction

where the market shares are below 35 percent, it

suggests that there are a number of rivals that really

matter out there.

lot of skepticism about a challenge under those

circumstances.

compelling case about why the other 65 percent is not

sufficient to constrain the exercise of market power in

that circumstance.

I think that you are going to find a

You are going to have to have a pretty

PROFESSOR WILLIG:

I think it is a form of

prosecutorial discipline, because it does force the

agency to articulate a narrow enough relevant market to

get past the 35 percent threshold and to confess that,

indeed, we are talking about localized competition, that

is the theory of the case.

is articulated, that is what is driving the bringing of

No matter how explicitly it

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

the case -- maybe appropriately, there is no doubt about

it -- but then the relevant market has to be articulated

as a narrow one, and then the 35 percent threshold will

be met easily.

The question is, will the court find that narrow

market to be credible?

be credible.

court is weighing in from a lay point of view.

And if not, maybe it shouldn't

It really is a matter of judgment, and the

MS. McDAVID:

Think back to the Grinnell case

where the Court talked about the market definition as a

red-haired, green-eyed man with the limp.

that the kind of thing you want to argue to a judge who

is going to be viewing this through his or her prism,

which may or may not include an economics background?

PROFESSOR WILLIG:

I mean, is

Or maybe the judge will like

to sleep in a van with the dogs and go skiing.

MS. McDAVID:

Exactly.

PROFESSOR GAVIL:

One thought just to add here

is I think safe harbors are important.

And I think that

not all market definition is going to be rocket science.

And the challenge is, if you have got a market

definition that does require more data, that is one that

is a little bit more complex, stating safe harbors can

suggest a false level of certainty -- using a safe

harbor that is based on a numerical threshold suggests a

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

degree of precision that may not be there with more

ambiguous markets.

terms of the less rocket science market definition, so

transactions can be identified that just are not going

to be on the table.

do not know, but the concept of having some easily

discernible area of safe behavior is an important one in

enforcement.

of antitrust enforcement.

But it does give some guidance in

Whether that is the right number, I

We talk about it again in all other areas

MS. McDAVID:

The cruise lines case is an

interesting example of market definition, because the

Commission's statement defined a market limited to

cruise lines, but then it became really clear that in a

competitive effects analysis, the exercise of market

power would be constrained by other vacation choices.

Therefore, we focused on competitive effects, which is

where I think the game really needs to be played.

MR. WALES:

Okay, put your agency hats on.

You

are back at the agencies.

the agency be looking for in terms of good unilateral

effects cases?

circumstances you think necessary, perhaps even

including some of the most recent cases -- were they

ones we should have brought?

agency be focusing on?

What types of matters should

What are the specific factual

Which ones should the

Obviously merger to monopoly is

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

the easiest, but I do not think anyone's going to say

that is all we should be looking at.

MS. McDAVID:

I would go back to circumstances

in which the evidence aligns, where the economic

evidence is consistent with the parties' internal

strategic planning documents.

strategic planning documents as a first screen.

particularly focus on one another, that may be an

indication of next best substitutes, and, therefore, a

transaction should be subject to additional analysis.

But I'd use a combination of all of the evidence and be

sure it points in the same direction.

PROFESSOR WILLIG:

You can almost use their

If they

Yeah, Jan, we have both seen

an awful lot of collections of business documents where

a company is very fond of naming one competitor over and

over again strategically and where the sum total of the

competitive forces from all the others, on analysis,

turns out to be every bit as important.

MS. McDAVID:

I said first screen.

PROFESSOR WILLIG:

MS. McDAVID:

PROFESSOR WILLIG:

MS. McDAVID:

Yeah.

First screen.

But caution to that.

Of course.

It has got to be the

whole collection of all evidence, not just the strategic

planning documents, but including the views of the

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

customers, evidence of recent entry, the competitive

problems the particular firms face, the whole array of

evidence.

PROFESSOR GAVIL:

I think we have come to a

point where there is something of a paradox that makes

the question hard to answer.

need to bring the best case; the Government needs a win.

It is easy to say that.

to say that, well, all the evidence ought to be pointing

in the same direction.

It is easy to say they

And it is relatively easy, too,

Here is the reason I think it is somewhat

paradoxical.

The blatant merger to monopoly, like the

blatant cartel, is not going to happen, presumably, very

often.

going to be harder cases.

to be represented by people like Jan, who are making the

best possible arguments with the best possible

economists about why a particular transaction should be

permitted.

the general skepticism of the courts about antitrust

now, means there are not going to be any easy cases.

is going to be hard to choose the best case.

The cases that are going to be presented are

The merging firms are going

So, I think, in a sense, that, combined with

It's not to say that people do not still propose

extreme things and that that may come along and you may

get lucky and have a fish in the barrel to shoot, but I

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

think that we are more likely to be facing complex fact

patterns, complex economics, and close calls, and it may

have more to do, in terms of winning, with the luck of

the draw in which judge you get and how that judge

reacts to the package of evidence than all that much

that the agency can do or the parties can do.

going to be tough cases.

of areas of antitrust.

PROFESSOR WILLIG:

Those are

That is where we are in a lot

And, of course, don't forget

that how tough the cases are is, in a way, a testament

to the remaining credibility of the agencies, because

the cases that would be easy do not get to court.

the ones that are left to go to court are the really

hard ones, inevitably, and that is still true, despite

the somewhat checkered record of the agencies in courts

lately, and that is a testament to the lasting view of

this marketplace of the skills and the abilities of the

agencies.

So,

So, look on the bright side.

MR. WALES:

I think there has been a lot of talk

lately about the general skepticism about antitrust.

That skepticism is something that we feel more generally

in terms of talking to judges and others.

How do we deal with that?

How do we reduce that

skepticism and somehow renew the interest in strong

antitrust enforcement?

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

MS. McDAVID:

It is a forensic exercise.

It's

got to be.

educated about the concepts that underlie some of this.

The Antitrust Bar tries to do a good bit of that, and we

do supply copies of Antitrust Law Developments.

And I think the bench is becoming better

PROFESSOR GAVIL:

The only thing I would add

here is, again, I think context is important.

to get narrowly focused on our little corner of the

world in antitrust.

antitrust cases.

expensive process.

our antitrust case -- I am working on a symposium at

Howard on the history of Conley and Twombly -- and

Conley, in 1957, 50 years ago, was a civil rights case.

The five lawyers working on the case were all

African-American.

the nut of getting at intent to discriminate by a union

that was complicit in employer discrimination, and in

that context, at that moment in time, the court said,

"lower the pleading barrier, these cases have to go

forward."

civil litigation for 50 years.

We tend

Judges are not skeptical just about

Litigation has become a costly and

Twombly, which we think of as

They were basically trying to crack

That became the standard that we used in all

And then if you had to imagine what would be the

antithesis of that case, Twombly was potentially the

antithesis of that case -- a nationwide class action

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

involving potentially hundreds of millions of consumers

against all of the leading telecommunications companies,

and the court recoiled from Conley in that case.

Now, partly, that is a challenge of using the

same procedural standards in every kind of case that we

do, but what does that mean?

litigation system today with over a quarter of a million

cases filed each year in the federal courts.

of cases; a lot of them are complex; habeus can be just

as complex for a judge as antitrust; and there is

generally resistance to litigation.

looking outside antitrust is helpful in locating

ourselves in the larger world of federal court

litigation.

PROFESSOR WILLIG:

It means that we have a

It's a lot

So, again, I think

Do you think the public who

forms these troubling views, including the judges,

distinguishes adequately enough between cases brought by

the United States, by the FTC, and cases brought by the

adventuresome private bar?

I mean, maybe some of the bad rap that antitrust

has is because of the activist plaintiff's bar.

could be.

variable in their superficial and end validity than are

the cases brought by the agencies.

PROFESSOR GAVIL:

I think on average those cases are far more

Bobby, I think it is a good

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

point.

extent the agencies have fed the fires of hostility to

private actions, the courts' hostility to antitrust is

coming back and constraining the agencies as well.

One of the ironies, though, is that to the

But yes, clearly, if you look at the Supreme

Court decisions of the last two terms, there is a lot of

anti-private action rhetoric going on, and some of it

was coming from the government agencies that were

encouraging that view, and it came back to bite them in

a case like Credit Suisse, for example.

MS. McDAVID:

truth in that.

Trinko.

Certainly it was driving Twombly and

MR. WALES:

today.

Okay, I'd like to thank our panel

We had an excellent discussion.

(Applause.)

MR. WALES:

break.

very much.

I think there is a good bit of

The plan is to take a 15-minute

So, let's be back at 10:35, if we could.

(A brief recess was taken.)

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

Thanks

PANEL 2:

THE ROLE OF MARKET DEFINITION IN

UNILATERAL EFFECTS ANALYSIS AND

IN THE LITIGATION OF UNILATERAL EFFECTS CASES

MR. SCHMIDT:

The next panel is going to focus

on the role of market definition in unilateral effects

analysis.

panel that it is difficult to separate these panel

discussions so that they do not overlap at all, but our

focus is going to be on the requirement or the lack of

requirement to prove a relevant product market and the

various implications of that.

I think you have already seen from the first

We have a terrific panel to focus on that issue

with us today, and let me just take a minute to go

through the introductions, and then we will start right

in.

To my far left, Jon Baker.

Jon is a Professor

of Law at American University's Washington College of

Law, where he teaches courses primarily in the areas of

antitrust and economic regulation.

senior consultant with CRA International.

experience includes being the Director of the Bureau of

Economics -- we won't hold that against him -- at the

Federal Trade Commission, Senior Economist -- sorry,

Professor Baker is a

His previous

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

Mike, wherever Mike is -- Senior Economist at the

President's Council of Economic Advisors, Special

Assistant to the Deputy Assistant Attorney General in

the Antitrust Division, and Assistant Professor at

Dartmouth's School of Business Administration.

sure you know, Jon is co-author of an antitrust case

book and past editorial chair of the Antitrust Law

Journal and a past member of the Council of the ABA

Antitrust Section, and in 2004, he received American

University's Faculty Award for Outstanding Scholarship,

Research, and Other Professional Accomplishments, and in

1998, he received the FTC's Award for Distinguished

Service.

To my immediate left is Kathy Fenton.

As I am

Kathy is

a partner at Jones Day.

She's practiced antitrust law

for more than 25 years.

She is currently the Chair of

the Antitrust Section of the ABA and has served in

numerous positions, including editorial chair, of the

Antitrust Law Journal.

professional service committee and served as chair of

the ethics subcommittee.

on issues of professional responsibility, conflicts of

interest, and legal ethics, including serving as an

instructor on legal ethics for the D.C. Bar's new

admittees course.

She is a member of Jones Day's

She has written and lectured

Her recognitions include Who's Who in

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

American Law, The Best Lawyers in America, 2007.

previously served as an Attorney Advisor to the Chairman

of the FTC and was a law clerk here in the District of

Columbia, the District Court.

She

To my far right is Dan Wall, partner at Latham &

Watkins.

Competition Practice Group.

has been active in the Antitrust Section of the ABA,

also.

of the Antitrust magazine; was chair of both the

Computer Industry Committee and Sports and Entertainment

Industry Committee; organized and chaired The Stanford

Conference on Antitrust in the Technology Economy.

has also authored numerous articles on application of

economic theory to antitrust issues and on high

technology antitrust.

lawyer in the Antitrust Division of the U.S. Department

of Justice, and his recognitions include Chambers USA,

America's Leading Business Lawyers, The Best Lawyers in

America, Legal Media Group's Expert Guide to Competition

and Antitrust Lawyers, and Global Competition Review's

GCR 100.

Dan is Chair of Latham's Global Antitrust and

Throughout his career, Dan

Dan was a founder and served four years as editor

He began his career as a trial

Then to my immediate right is Rich Parker, a

partner at O'Melveny & Myers.

Rich is Co-Chair of that

firm's Antitrust/Competition Practice.

He returned to

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

O'Melveny in February 2001 after serving three years at

the FTC, as first Senior Deputy Director and then

Director of the Bureau of Competition.

recognized as a Leading Lawyer in Antitrust by the Legal

Times; named by the Global Competition Review as one of

the best antitrust defense lawyers in the United States;

and recognized as a leading antitrust practitioner by

Global Competition Review, Chambers Global, Chambers

USA, and Super Lawyers Magazine, and probably others.

He received the Distinguished Service Award also from

the FTC.

Rich has been

So, with that, I think we are going to try to

follow the same format that the first panel used, which

is to ask each of the panelists to give a short

presentation, and then we will go right into questions

and hopefully have a lively discussion.

going to start with Jon.

PROFESSOR BAKER:

I think we are

Good morning, everyone.

I am

delighted to have been asked to be here, and I see some

old friends.

for future reference, Bobby and Andy, I prefer to be

discussed for my ideas, not for how I look, okay?

It is also very nice to be discussed, but

My assignment is to talk about -- is to be a law

professor and to talk about the -- I can't help it, I

will be an economist, too -- talk about the pros and

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

cons of using market definition in unilateral effects

cases to set up the panel.

into three categories, so I am going to talk about legal

arguments, economic arguments, and litigation tactic

pros and cons.

The arguments neatly divide

So, on the legal side, we have to start with the

words of the statute, of Clayton Act Section 7, which

objects to acquisitions that substantially lessen

competition, and now I will quote, "in any line of

commerce or in any activity affecting commerce in any

section of the country," and that language, that

statutory language, arguably, makes proof of a market an

element of the offense.

On the other hand, if the Government can prove

harm to competition directly, there has to be some

market within which competition takes place, and, why

isn't that inference good enough to satisfy the statute?

I once wrote an article where I called that kind of

approach a res ipsa loquitur market definition.

words of the statute is one legal issue.

Another legal issue is the Oracle decision.

Judge Walker held that the Government must prove that

the merger must -- in a unilateral effects case, that

the merger must -- would create a monopoly or near

monopoly.

Monopoly is almost always demonstrated by

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

high market share, so the Government essentially has to

define a market to satisfy this element of what Judge

Walker sees as part of the offense.

that Judge Walker's holding in that decision is based on

a clear error in economic reasoning.

believe that other courts will follow it.

commonly used horizontal differentiation model that

Judge Walker seems to have in mind, unilateral effects

can arise in mergers that involve firms that are not the

largest in the market and that do not create a dominant

firm, just as a matter of economics.

legal pros and cons.

The con here is

So, I don't

Even in the

So, that is the

Now, economic pros and cons of defining a

market.

I think here I am going to start with the cons

and not the pros.

among sellers of differentiated products does not turn

on market shares.

as arising because the merger lets the firm recapture

profits that previously it would have lost were it to

have raised price, and so it now has, after the merger,

an incentive to raise price.

The economics of unilateral effects

You can think of unilateral effects

That is one intuition.

Another way of thinking about unilateral effects

is that they arise because the merger allows the firm to

remove the competitive response of an important rival,

and that makes the initial firm's residual demand less

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

elastic.

incentive to raise its price.

about unilateral effects among sellers of differentiated

products, the market shares do not directly matter to

the economic analysis.

Again, you can see how that would give it an

Either way you think

Now the other side of the story.

The market

shares would be a good indicator of pressure to raise

price if the diversion ratios or the demand elasticities

are related to them.

second choices are distributed similarly to customer

first choices, which is what Bobby was getting at this

morning when he talked about generally differentiated

products.

That could occur if the customer's

Also, high market shares likely indicate that

the diversion ratios are so high or that they are high

enough that they will generate some sort of unilateral

effects, unless the merging firms' products appeal to

very different groups of customers.

a 50 percent market share merges with a firm with a 20

percent market share, the two would have to be in very

different niches in order to not have a unilateral

effects problem.

burden.

So, if a firm with

The high shares almost shift the

Also on the pro side of using market definition

in the economics category, if the way you collect the

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

evidence relies on econometric evidence of diversion

ratios or demand elasticities, then, some sort of an at

least informal market definition is required to specify

the list of potential rivals that you have to include in

order to avoid bias in your analysis.

So, if you leave out an important rival when you

conduct the estimation, then the elasticity estimates

are most likely biased in the direction of overstating

the unilateral effects.

Bobby was also getting at this morning when he talked

about collecting the important sources of competitive

discipline.

in this context, as biasing the estimate of unilateral

effects because you left out the others down the road,

in Bobby's theory.

This is something that I think

The gas station example could be understood

The third area where I want to talk about pros

and cons of defining markets and proving unilateral

effects cases has to do with litigation tactics.

the pros and cons depend on whether the Government would

define a narrow market or a broad market or not one at

all.

market.

through superstores rather than all office supplies, or

superpremium ice cream rather than ice cream, the kind

of things that we talk about in our professional world.

Here,

Let's suppose the Government defines a narrow

Here we have in mind, office supplies sold

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

The pro of defining a narrow market is that it

allows the Government to portray the case as a merger to

monopoly or near monopoly, and it also creates a causal

inference of unilateral effects when the market shares

are high, as with the 50 percent firm merging with a 20

percent firm, as we said before.

On the other hand, a narrow market may not be

persuasive if it looks gerrymandered.

particular problem if some of Bobby's Audi drivers would

go to BMW and some would go to Lexus.

is harder for him to sell his Audi/BMW market to a

court, particularly if more of the Audi customers would

go to Lexus than to BMW.

That could be a

It may be that it

Also, this approach potentially focuses

attention on the wrong issue.

primary attention to the extent of buyer substitution to

the third firms, the rivals outside the market, rather

than to the extent of the buyer substitution between the

merging firms, which is the source of the unilateral

effects.

substitution between the merging firms, but you are busy

worrying about, in market definition, the substitution

to the third firms.

That is, it directs your

The first thing you want to know is the

Now, let's suppose the Government defines a

broad market.

The pro here is that the market may seem

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

more intuitive, like Jan suggested this morning, unless

gerrymandered in its appearance.

allows the competitive effects case to take primary

place in telling the competitive effects story in

litigation for the Government and focus attention on the

way that the merger lets the firm recapture lost profits

or alter the competitive response of an important rival,

consistent with the economic theory.

on the theory, the economic theory.

The broad market

You are focusing

On the other hand, if you define a broad market,

you may essentially admit that a large number of firms

are rivals to the merging firms, that merging firms'

shares are small and that competitive effects are not

uniform, because they are concentrated in a small part

of the market.

trying the case, and they make the competitive effects

look small.

Government embroiled in this question of whether there

is a 35 percent safe harbor for unilateral effects or

not in the Merger Guidelines that was alluded to in the

last panel.

All those things are bad optics for

And there is also the danger of getting the

The final litigation choice would be not to

define a market at all.

Again, the benefit of that is

it focuses the case on the way the merger lets the firm

recapture the lost profits or removes the competitive

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

response of an important rival, consistent with the

economic theory.

avoid litigation problems with defining a broad market

when market shares are low, but the con is that may be

illusory, because the defendant would presumably define

a broad market, and so the Government may not actually

avoid the problems arising from defining a broad market.

It would seem the Government could

So, there you have it, an even-handed view of

pros and cons of proving markets in unilateral effects

cases.

MR. SCHMIDT:

Thanks, Jon.

Kathy?

MS. FENTON:

I was asked to share some thoughts on the legal

Thank you, Jeff.

need to prove market definition in unilateral effects

cases, and as Jon Baker already indicated, the reason we

are having this discussion goes back to the basic

language of Section 7, the requirement to show effects

"in any line of commerce in any section of the country,"

a mandate that some -- you may call them a strict

constructionist -- have identified as being the source

for any obligation to prove markets as part of your

affirmative showing of a Section 7 violation.

But I think the more interesting issue to focus

on in this area is the fact that much of the current

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

debate can be directly traced to the lack of recent

and/or relevant Supreme Court opinions on this subject.

I am sure there is a great analogy to children's

literature that is possible here, whether it is Rip Van

Winkle or The Sleeping Princesses, but your last

substantive merger case goes back to 1975, and the last

time the court spoke on this issue was a year earlier,

in 1974, in the Marine Bancorp case, where it set forth

a fairly traditional three-part analysis that says:

"The analysis of likely competitive effects from

a merger requires determinations of, one, a line of

commerce, a product market in which to assess the

transaction; two, the section of the country or

geographic market in which to assess the transaction;

and three, the transaction's probable effects on

competition in the relevant product and geographic

market."

Now, judges, tending to be relatively

conventional creatures, look at that language and see,

not surprisingly, a mandate to define a relevant market.

The silence on the subject for the ensuing years from

the Supreme Court has simply added to the proliferation

of approaches we see at the district court.

those approaches have been responding to other

developments occurring at the Supreme Court level

Some of

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

outside of Section 7, outside of the merger context, in

areas involving either collusion or monopoly claims,

because you have a whole series of cases, some of which

were briefly touched on by the opening panel, NCAA, Cal.

Dental, Polygram, and perhaps, most dramatically,

Indiana Federation of Dentists, that seem to eliminate

the need for formal market definition if there is actual

proof of anticompetitive effects.

And I think the quote from Indiana Dentists

probably captures this line of development outside the

merger area most dramatically, because there the Supreme

Court said:

"Since the purpose of the inquiries into market

definition and market power is to determine whether an

arrangement has the potential for genuine adverse

effects on competition, 'proof of actual detrimental

effects, such as a reduction of output,' can obviate the

need for inquiry into market power, which is but a

'surrogate for detrimental effects.'"

Needless to say, that precedent from the Supreme

Court has surfaced in numerous briefs, often by the

private plaintiffs or government agencies prosecuting a

unilateral effects merger, seeking to argue that the

formalities of market definition are not essential as an

element of proof, and the argument in that regard, I

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

think, is perhaps very nicely captured in a recent

article by Katz & Shelanski in the Antitrust Law

Journal, called "Mergers and Innovation," that takes a

slight detour through unilateral effects analysis and

says:

"If the formalities of market definition can be

skipped in favor of direct analysis of harm in

monopolization and collusion cases, there is no reason

why the same should not hold true for merger analysis

where the issue, likely competitive harm, is similar."

They go on to recognize that merger analysis has

some limitations.

prospective and predictive than other kinds of antitrust

cases where the conduct at issue frequently has been

ongoing for some time," but this simply means that

direct effects may be easier to show in nonmerger cases

and not that direct evidence of market power should not

have the same priority in merger cases where such

evidence is available.

They say it is "more often

I would suggest that economists probably have

more flexibility than district court judges in offering

that alternative as a way of resolving these cases, but

the debate continues, and as you look at the recent

district court opinions involving unilateral effects,

you know, Oracle, Whole Foods, Arch Coal, you really

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

could go down the litany, you see judges struggling with

this question of what is their obligation to formally

make findings of fact and conclusions of law on the

relevant market question, and they tend to engage in

activities that could be characterized as a market

definition exercise without necessarily acknowledging

their obligation to do so.

can identify for resolving this question is the

possibility of further Supreme Court statements on this

question.

And I think the only hope I

Now, in the world post Hart-Scott-Rodino

notification, that is going to be a difficult

proposition, just because most mergers that are

challenged by a government enforcement agency do not

hold together long enough to ever reach the point of

Supreme Court review, but I think there is one possible

candidate on the horizon that I offer for your

consideration.

definition not with respect to a product market but a

geographic market, and the case, of course, is the

Commission decision in Evanston, which is still

awaiting, as far as I know -- and I will bow to more

superior information sources -- a determination by the

parties to file an appeal with one of the U.S. circuit

courts.

It poses the question of role of market

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

But in that case, the Commission opinion dealing

with a post-closing challenge to a hospital merger

concluded:

"It is not necessary to define the relevant

geographic market, because it is possible to show,

through direct evidence, that the merger enabled the

merged parties to exercise market power unilaterally."

Thus, the Commission concluded, because the merger

enabled the parties to raise prices by a substantial

amount, at least equal to a SSNIP, through a unilateral

exercise of market power, the geographic area alleged by

the FTC to constitute a relevant market constituted a

well-defined antitrust geographic market under Section

Now, if that issue were preserved through the

appellate process, we certainly have the prospect of a

court of appeals chiming in on the need for relevant

market definition and, as I said, a possibility for

Supreme Court review since a concluded merger, a

divestiture challenge essentially, is sufficiently high

stakes that the parties might be incented to take that

step.

But in the absence of that, I think we are going

to continue to see a struggle at the district court

level as they look back to precedents, and it is not

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

just Marine Bancorp.

It is Philadelphia National Bank,

it is DuPont, that all contain the language about

defining relevant markets, as well as what I would

suggest are some practical limitations imposed by the

Merger Guidelines themselves and the Merger Guidelines

structure, because there, the five-part organization

embodied in the Guidelines has, in a sense, provided a

road map for a lot of subsequent district court

analysis.

You start with market definition and

concentration; you consider potential adverse effects;

you do an entry analysis; you consider efficiencies; you

deal with failing or exiting assets.

sounds like a mandate for relevant market definition,

and as a result, to borrow Andy's phrase from the

initial panel, it is probably a very hard sell for the

courts to try and avoid or escape that exercise, and in

particular, this combines with a number of other

practical aspects, including judicial skepticism of

economic analysis.

That, again,

And I was reminded in preparing for this

exercise of a fascinating quote from Ken Auletta's book,

World War 3.0, which, of course, is on the Microsoft

case, but he had, you might recall, conducted fairly

extensive interviews as part of the process for that

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

book.

One of the people he interviewed was Judge Hogan

of the district court here in the District of Columbia,

who some might view as one of the godfathers or patron

saints of unilateral effects analysis since he is the

author of the opinion not just in Staples, but also

Swedish Match a few years earlier.

They somehow got off the topic of Microsoft in

the discussion for Auletta's book and started talking

about the Staples/Office Depot case, and Auletta reports

in his book:

"When Judge Hogan presided over the Government's

antitrust action to block the proposed merger of Staples

and Office Depot, Hogan reported, 'We had a lot of

economic evidence, we had a lot of documentary evidence,

although in that case, the economic evidence that the

Government had was not at all convincing to me.

the internal company documents were more convincing.

That is why I stopped the merger.'"

I think

And that reality, I think, is something that you

are going to see reflected in perhaps less overt fashion

in many of the judicial decisions dealing with that

question.

MR. SCHMIDT:

Rich?

MR. PARKER:

Thanks, Kathy.

I am supposed to give the

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

government perspective on trying these cases, and as you

know, I am now playing on the other team, so it would

really be unfair if you quote this stuff back to me when

I am sitting next to a client.

trying to convince you to go away.

down as a rule.

When I'm down here

So, let's get that

What I want to talk about is how to put a case

like this together.

We have people who understand the

law and economics better than I do.

hear that from me.

and trying cases is an art, and everybody has a

different style, but here is the way I think about it.

You do not need to

So, here is my own personal view,

I was privileged, my first job out of law

school, to clerk for Judge William Matthew Byrne,

Junior, in Los Angeles, who passed away a year ago, who

was one of the best trial lawyers in Southern California

before he went on the bench.

And was a great trial judge and was a great teacher.

And I remember, when I was down there, we had this

really boring patent case.

dry than listen to this testimony in this chemical

patent case, but that was my job and my co-clerk's.

He won a lot of big cases.

I would rather watch paint

And the trial ended, and we went back to

chambers, and the judge said, "Well, "Justice West of

the Pecos" says that the plaintiffs ought to win here."

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

I said just looked at him.

He said, "By that I mean,

common sense, logic, my gut sense of what is fair and

reasonable," and then he went through and told a story

about what happened here, which is exactly the way

counsel probably should have tried the case, and said,

"Now, that is what my opinion ought to say, and you tell

me if we can get to a plaintiff victory under the case

law, and if we cannot, then we better have a meeting

and, figure something else out."

Pecos" has always been in the back of my mind.

stopped being a mentor to me, and that is the way I view

these cases.

"Justice West of the

He never

In my opinion, the Government ought to try these

cases with effects, and I do not think what I am saying

is anything inconsistent with what was said in the first

panel.

have an advantage in being the Government, and the

advantage is inherent judicial conservatism.

market that is working.

coming in with their fancy economists saying, "Well, we

are going to change this structure radically, but don't

worry, our efficiencies are going to do this, that, and

the other thing."

skepticism with a judge or with most judges about

radical changes in a functioning market, and you are

You start with effects.

Remember this.

You

You have a

And now you have these guys

And so I think you have an inherent

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

trying to stop that from happening.

advantage.

That is an

So, you play on that, and you build it by

showing what is going to happen that is bad here.

are people going to get hurt?

and others said this morning, there is an inherent

dinner table logic to unilateral effects.

not care about Bimmers and Audis, but Whoppers and Big

Macs or something like that they do.

competition from other burgers and maybe from Taco Bell,

but those two are unique competitors, and they look at

each other when they price their products, and if one

buys the other, that constraint is gone.

logic that makes a lot of "Justice West of the Pecos"

sense.

How

And as Dr. Willig said

Judges may

Sure there is

That is a

In my opinion, the most important support for

that case is the company's business documents.

they look at?

board?

Do they look at this fringe or do they look at

tacos?

Do they look at whatever?

each other?

build on that.

What do

What do they look at when they go to the

Or do they look at

That is the number one point.

And you

And the second thing you build on are customers.

Customers.

The Government cannot try, effectively, a

case without strong customer support, and by "customer

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

support," I don't mean just, "I hate the merger."

mean, "I have dealt with these people day-in and

day-out, for year after year, and I play them off each

other, and this, that, and the other thing, and I have

detailed knowledge, and in my opinion, I have benefited

from that competition, and let's not let it go away."

You cannot put on the stand a lot of people who

simply don't like the merger because they don't like the

merger but do not have any real experience in dealing

with the entity being purchased.

Arch Coal, where at least -- and this is Monday morning

quarterbacking -- but at least some of the witnesses in

that case had that problem.

Now, relevant market.

I am going back to

You have to prove a

relevant market.

Every case says that.

You can't

pretend like they do not say that, including your

favorite cases, starting with Chicago Bridge, your

latest victory, Swedish Match, every one of them, Baker

Hughes, Staples, Drug Wholesalers, you name it, they all

say it.

You have to do that.

But I suggest that the first tactic is to back

into the market from the effects.

At least in Judge

Hogan's court, you can do that.

could be that that is where the market came from in

Staples.

It is plain as plain

It is equally plain that that is where the

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

market came from in Swedish Match.

fight the last war, and this is a long time ago, and

Rick Liebeskind and I and Jon Baker were heavily

involved in Drug Wholesalers, and Judge Sporkin

believed, at the end of the day, that hospitals and

independent pharmacies could not protect themselves

against the merging parties, and that is how we ended up

both with effects and with the market.

from effects.

And generals always

You back into it

You try effects -- remember, things are working

great.

to happen if you change it.

say.

economists say.

competition or their efficiencies?

consumers' money on their efficiency study or whatever

other study they may have.

They want to change it.

Here is what is going

This is what the customers

This is what the documents say.

This is what the

What are you going to trust, existing

Don't bet the

All right, the government has run into some

trouble in some cases, and I wasn't in these cases, in,

say, Oracle and in Whole Foods, so I don't know every --

you know, Dan will talk about Oracle, and we are lucky

to have him here to talk about that perspective, but I

suspect that what happened in both cases is that the

government didn't prove effects, and everything got

bollixed up on market, but frankly, at the end of the

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

day, I will bet if you psychoanalyze the judge, you did

not prove effects in Whole Foods and that is how the

market ended up so broad.

And by the way, I want to compliment Paul Denis,

who I see back here, on that case, because my litigation

instinct on Whole Foods is that it looks like the

evidence was very strong in that case, and I am not sure

what happened.

I was not in the courtroom.

In Oracle, and Dan will go into this more, it

looks like the judge didn't believe the customers.

The

customers have to have real knowledge about the market,

and I think, by the way, that is what happened in Arch

Coal as well.

some of the customers really knew what they were talking

about, and it is clear in Oracle that that is what

happened.

what happened in those cases, is that you didn't prove

effects.

I do not think the judge thought that

So, those are the -- my best projection as to

Now, let's assume you are in the next case, and

you have a situation where you have a unilateral effect,

where you have something like the Whole Foods case,

where you have a problem in that intuitive logic may

suggest that Safeway ought to be in the market, and I

was driving in the car with my wife, who said, "How can

they bring that case, because Safeway has organic food?"

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

That was a problem that you faced in that case.

Here is what I do:

My colleague, Tim Muris, who

is an antitrust purist, would probably throw something

at me if I said this, but how about a submarket?

not analytically the greatest concept in the world, but

after all, this is about winning and you are a law

enforcement agency.

win, and submarkets are all over the case law,

undeniable.

are in all these cases, including the cases I just cited

to.

It is

Law enforcement agencies have to

It is not just Brown Shoe, but submarkets

It is there.

Number two, credibility is the key.

That is

what you have got in front of a judge, is credibility.

So, another alternative is to say, "You know, I will

tell you -- I will give them their supermarket

market" -- and again, I am doing Monday morning

quarterbacking here, but I am speaking hypothetically.

"I will give them their market.

Safeway, Giant, Food Lion, and everything else, and,

Your Honor, in most cases, we rely on the Philadelphia

National Bank presumption, but, you know, I do not need

any presumption.

need it, because I have got hard and fast evidence that

will show you that in 22 markets, 15 markets, or

whatever it is, what drives price are these two, and if

I will give them

I don't want a presumption.

I don't

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

you let this merger go through, those prices are going

up.

that, but I am also going to prove effects to you and I

do not need Philadelphia."

case, I would take that -- I would take that step.

I will give them their market.

I will give them

And I would -- in the right

Those are my thoughts, and I hope these most

certainly have been helpful to you, and I know it is

tough to lose these cases, it is very tough, because

anybody who tries cases who loses them, it is not a good

thing.

admirable for this agency to get all these people in

here and to look at what they've done and to be

self-critical and try to come up with some new concepts

and some ideas, and I really commend you for doing that.

The key point here is that I think it is very

I will turn it over to you, Dan.

MR. SCHMIDT:

Dan?

MR. WALL:

Thanks, Rich.

Good morning.

Let me pull something

up here.

So, thank you for the introduction, but we all

really know why I am here, and it is because of Oracle,

which Rich did mention, and that is okay, you know, he

got --

MR. PARKER:

MR. WALL:

I mentioned it, Dan.

Yeah.

You know, you have got to have

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

the villain in order to have an interesting plot, and I

will gladly be the villain here and give you some

defense perspectives.

In keeping with Commission policy, I will have

to ask all of my competitors to leave the room at this

point, just because I am going to be talking about some

strategy points in here, but I think that the issues

that are raised by this really are profound in the arena

that is much more my home than the law and economics as

well, which is the arena of trial, and it is a different

environment than any FTC or ABA Antitrust Section

conference.

It is a trial that is conducted before someone

who rarely is particularly expert.

In the Oracle case,

we actually had someone who had practiced antitrust law

professionally.

That is definitely the exception rather

than the rule.

And it is an arena in which somebody is

used to resolving contested facts in a wide variety of

cases based upon that kind of "West of the Pecos"

intuition that Rich was talking about, and if you do not

try your case, if you do not build your case with that

always in mind and with a firm understanding of what

people like me are going to do to try to deconstruct

your case and to break it down in the particular dynamic

of a trial, then I think that the odds of winning in

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

these cases go way down.

I am going to draw a lot on the Oracle

experience here, you know, because I had a lot of trial

materials that I could pull into my presentation and

demonstrate some of these things, but it is just -- you

know, it is just one case.

But I will say this, that even though I know very well

that the agencies all say that they'd bring the Oracle

case again if they had a chance, and if I were the head

of the Antitrust Division, I'd say that probably about

any case I lost, so I respect that.

This will always be true.

I will tell you that I felt very strongly, and

Commissioner Tom Rosch, who was my partner at the time

and tried that case with me, felt very strongly, before

that trial began, that we were going to win that case,

because the case that the Department of Justice had put

together was not sustainable in the arena of trial.

was going to get cut down by trial dynamics.

case is not resilient in the arena of trial, through

trial -- the dynamics, it doesn't matter how good it is,

because that is the arena that counts at the end of the

day.

If your

So, a few observations, and this is all about

the idea of do you use market definition or not or do

you put on a case without it.

The first one, don't --

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

don't think about it.

Guidelines, which we will use viciously against you,

this is a recipe for disaster, okay?

already heard some of the reasons, but it comes from the

fact that, as Kathy discussed, this just -- market

definition as an essential element of the analysis just

couldn't be more entrenched in the case law.

Under current case law and the

And you have

I bet you that on a dare for a beer, I could

cite you a hundred cases that in mergers and

monopolization and other market power kinds of offenses

say that this is a threshold requirement, and yes, there

is this little thread out there that talks about the

ability to prove effects, and I fear that as a defense

lawyer in a monopolization case in which the conduct has

occurred and the effects might be presently observable,

and I might fear that in a post-merger challenge, like

Evanston, where you have some ability to look at what's

happened.

But honestly, I don't fear that very much -- I

don't fear it very much at all in a typical merger case

where the analysis is prospective, because I know that,

by definition, the plaintiff, the Government, is not

going to have tangible prove of adverse effects.

are only going to have some documents that maybe they

can make a prediction from, and I can fight the

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

They

prediction game based upon market structure and market

definition arguments, and I will probably win that most

times.

The second point, you know, the Merger

Guidelines are your own worst enemy about this.

want to pursue cases in which the unilateral effects

market definition is not part of the equation, amend the

Guidelines.

an imperative, because what we do is we use the

Guidelines against you to impeach you, to say to the

judge, "Look, they are not even following their own

Guidelines."

position, and some of you will someday when you are in

our position.

argument; it is a "gotcha."

going to be able to run from the Merger Guidelines.

you know, it is been a long time since the Merger

Guidelines came out.

I think that would be an essential step for you to have

any credible program of trying to bring unilateral

effects cases without market definition.

Not a suggestion.

If you

I am telling you it is

You would do it, too, if you were in our

It is a natural argument; it is a great

You know, you are never

So,

Maybe it is time to revise them.

You know, the third point, there is this -- it

is not just that you have all this case law that says

that you have to have a defined relevant market.

is another body of case law that questions whether you

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

There

can show the substantial adverse effect on competition

if it is only on just some piece of the relevant market,

and we thought we were going to get into this in Oracle,

and then there was some change in DOJ strategy, and so

we didn't really have to do it as much, but having

looked into this, we were in a position to make a pretty

good argument that the effect had to be generalized or

that it at least had to -- you know, that there was some

quantitative sort of threshold that the percentage of

the consumers in the relevant market that would be

affected, and so that you couldn't just make an argument

that was about a very, very small group of consumers.

You know, I think that unilateral effects has a

tremendous danger of taking the economics too far.

You

know, in Oracle, which was based largely on this sort of

auction bidding theory, the Department of Justice's

position, taking it from its expert reports, at face

value, was that the adverse competitive effect would

only -- that only about 20 percent of the customers were

vulnerable to suffering this effect.

a big number in absolute terms, but query whether an

adverse effect that only hits one in five customers in

the market would survive as a matter of law.

Now, 20 percent is

But that's not really as far as this goes.

know, Carl Shapiro and Joe Farrell just published a very

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

You

provocative new article on this in which they have an

analysis that is basically -- that is driven by

diversion ratios and gross margins, and they have a

statement in there that you could show a unilateral

price elevation in an industry with high gross margins

where the diversion ratio between the firms is as low as

5 or 10 percent, and, you know, I have no doubt that

Carl's math is right, but I have got to tell you, bring

it on.

I mean, if you are going to bring a case and you

are going to try to say that this merger should be

stopped essentially because there are high gross margins

and one in ten losses of the merging parties are to each

other, I am going to come back with a very powerful

argument that that is just too de minimis, insubstantial

an effect to meet the substantiality requirements of

Section 7.

about doing this, and I think that that market

definition is what judges find as an intuitive governor

on this thing, on this whole process, of saying, "Show

me an effect that is substantial in a market."

So, I think you have got to be very careful

I want to -- this is a slide -- this was

actually from my opening statement in the Oracle case,

and it -- I bring this up just to -- just to show you

how cynical and mean we really are on the defense side,

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

because this is a -- I mean, this is what we do if a

plaintiff has a flakey market definition or if they are

running from market definition.

pretty credible theory that DOJ had developed during the

Hart-Scott-Rodino process, which was actually before I

got involved, that said that in these procurements for

these software systems, that essentially every bid was

akin to a relevant market, and then the Government

decided not to bring that case, to make that their

argument, when they filed it, saying that actually they

were bringing a "traditional case."

There was actually a

And I have no doubt that the reason was is

because they knew that they were going to get attacked

by us for having come in with a novel theory that

nullified the importance of market definition.

brought it up to make that point, you know, we brought

it up, and it is because there is nothing more valuable

to us than trying to convince the court that the

Government is cheating, because the Government comes in

with a tremendous reputation and sort of a presumption

of being right, and we have got to crack that.

this instance, you know, we will bring it up.

So, we

So, in

So, what I am telling you is there is no running

from market definition.

You are going to have to build

your cases around traditional markets, and you are not

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

going to -- you can't leave us any room to argue that

you are doing something else.

My second point about the approach of not having

market definition is to say good-bye to Philadelphia

National Bank, okay?

sharp and a little bit critical of the Government, but

the fact of the matter is that one of the reasons you

get yourselves into this mess on market definition is

you want your Philadelphia National Bank presumption,

and you are willing to do whatever it takes to get it,

okay?

Now, this may sound a little bit

Well, I would tell you that I do not actually

believe that the Philadelphia National Bank presumption

should apply to a unilateral effects case, because it

actually came out of the structure-conduct paradigm for

coordinated effects, and the Supreme Court has really

never addressed it in a unilateral effects context.

the thing is, what the Government is doing is they want

to make this estimate up here, which is from the

Government's brief in Oracle, where they say:

But

"Plaintiffs establish a prima facie case of a

Section 7 violation by demonstrating 'that the merger

would produce 'a firm controlling an undue percentage

share of the relevant market,'" et cetera, all very

familiar, tactically I get it, I understand it, but you

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

100

are actually taking a big risk if you gerrymander the

market in some way to get that when, if your economics

effect -- proof is very strong, you probably do not need

it to begin with.

So, what is the alternative?

Well, you actually

end up with the Whole Foods briefs that the Commission

has just filed, which contain exactly one reference to

Philadelphia National Bank and do not try to win the

case and leave the defendant in an essentially

unwinnable position through the presumption, but rather,

cut to the effects.

have to live in if you eschewed market definition.

This is the world that you would

Now, my third point is don't kid yourself that

the alternatives to market definition are practical or

persuasive, because they usually aren't, and this goes

to the point that a couple others have already made

about just the relative persuasiveness of different

kinds of proof.

court, rather than in university seminars,

persuasiveness is about intuition to the layperson, to

common sense, to very simple things like that.

And remember, you know, in district

And the thing that you have got to understand is

that the intuition that we rely on is the intuition that

mergers of firms that face a lot of competition won't

harm anybody.

That is a strong intuition, okay?

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

That

101

is a very simple intuition.

of drive the half mile, fool, you know, get the gas down

the street.

in any merger case, regardless of the theory, that the

merging parties have a lot of competition, I am feeling

pretty good about it.

economists, but if I have shown that we have got a lot

of competition, we are feeling pretty good about it.

That is Bobby's intuition

Everything will be fine.

And if we show,

You can come in with your

Now, in contrast, I mean, the economics of

unilateral effects are really, really complicated and

difficult to understand.

this, I see visually, because he recognizes that what I

have done is I have put up here on the slide what he

calls a simple, practical test for identifying

unilateral effects in his recent article, and, you know,

I won't go into it, because I am sure he'll be

discussing it, but, you know, it is got math, it has got

those things where you have to use the different font to

bring it down below the line, and it has got Greek in

it, you know, and my point is that regardless of how

good that is, I can do a pretty good job of making the

judge not think about it, okay?

Carl has already reacted to

Carl may remember this story from a case we

worked on together, and everybody has heard of this

case, it is the trial of the Eastman Kodak and Image

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

102

Technical Services case, where we were up against Max

Blecher, one of the best plaintiff's lawyers in the

United States, and his expert, the plaintiff's expert,

is Jeffrey MacKie-Mason, and he's being put on the

stand, and the first question that the plaintiff's

lawyer asks his own expert is, "Dr. MacKie-Mason, isn't

it true that if you ask two economists the same

question, you get three answers?"

the economic testimony, because we were coming on with

Carl Shapiro and Janusz Ordover, and we had a lot to

say, and he didn't want the jury to care about it, and

so with his own expert, his first question is nullifying

the value of the economic testimony.

unilateral effects stuff is very, very complicated, and

it is something that you take a great risk as to whether

you are ever going to be able to get the judge to

understand and want to apply this.

Now, there is other cases.

He started nullifying

Well, this

I mean, I mentioned

Staples, and this is actually an exhibit from Staples,

which Jan McDavid was essentially referring to earlier,

and this -- you know, this was the evidence that they

had, and in -- and, you know, this is the mother lode

here.

were substantially higher in markets in which there was

Staples only and that the only real significant thing

This was realtime proof that the Staples prices

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

103

that brought their prices down was competition from

their merger partner.

stuff.

intuitive unilateral effects case.

I mean, that was really good

If you have that, you are going to make an

Let me contrast that with the merger simulation

in Oracle.

essentially an auction model that Preston McAfee came up

with.

these Logit models, which ironically demands market

shares in order to run the model.

function from market share.

use it as an alternative to market shares, but it was a

model in which assumptions about market shares were then

coupled with an assumption about how much surplus

sellers were currently capturing from their customers.

You know, that was so ivory tower-ish and so unreal and

so untethered to actual data that I don't think it ever

had a chance, but because it was also grounded in market

shares, it was DOA as soon as the market definition

shifted at all.

The merger simulation in Oracle was

It had no real world data on it.

It was one of

It implies a demand

So, first of all, you can't

You know, Jonathan and Carl wrote an article

criticizing Judge Walker's decision in which they make

the point that he was unfair to this model in demanding

more real world data, because they say that in their

experience, that real world data on prices, costs, and

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

104

output are invariably imperfect for a variety of

reasons.

responses.

You know, I can't help but offer a couple

First of all, it is not actually a valid

criticism of Judge Walker in Oracle, because Professor

McAfee had no data.

was running a market share-driven model, not a

data-driven model.

It was not an imperfect data.

But second, I'm sorry, but pervasive data

problems are a reason not to rely on merger simulations.

They don't -- they don't excuse it.

you are actually adding risk to your case, not cutting

it back.

If it's bad data,

So, fourth and finally, and I really -- I say

this with great sincerity, is that you have got to stop

taking the amount of trial risk that you are by arguing

for markets that are narrower than they have to be.

you believe in your competitive effects case, argue it

within a defensible market, and by that I mean a market

that is not going to get cut to ribbons.

Look, we know it is not working, okay?

We all

know it is not working, and that is having a market

definition that allows people like me to just gather up

the evidence that inevitably will be there of

competition from the firms that you have eliminated from

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

105

the market.

These were just a couple of slides, I could have

done a zillion of these, and I could take them from any

other case, but they were just some of the slides that

we used to identify firms that in Oracle the Government

said were not in the relevant market, and then we just

went to call reports and invoices and discovery

documents and all sorts of stuff, and we created long,

long, long lists of procurements in which these

customers who were not in the relevant market were, in

fact, competing with the merging firms or SAP, the third

firm, in the market.

And when we do that, there is nothing you can do

to stop us from having great days in court.

You can't,

because we have that evidence, and we can walk up to a

witness and say, "Are you saying that you don't compete

with Lawson?

Are you?"

And first the guy looks like a deer in the

headlights for a minute, and then he says something

like, "Well, we don't see them very often."

Then I will say, "Isn't it a fact you saw them

at Safeway?"

"I don't remember."

"Let me show you the document.

Isn't it a fact

you saw them at Food Lion?"

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

106

"I don't remember."

"Let me show you the document."

This is shooting fish in a barrel.

easy.

talent to do that, because you have got the documents

right in front of you, you know?

it will probably, you know, reduce the -- change the

slope of my demand curve by saying that, but it is

not that difficult to gather that stuff up, and you have

got to anticipate that.

and plan for it and don't let me do it.

bring your case by conceding me those people, do it.

You take away all my good stuff.

what you want to do.

Honestly, it really is.

This is so

It takes very little

I shouldn't say that,

You have got to anticipate that

And if you can

I mean, that's really

And that leads kind of to my sort of final point

here, which is, you know, if you believe in the

unilateral effects model, do it.

is -- you know, this is -- this is another quote --

sorry to keep picking on Jon and Carl, but this is a

positive one here.

an economic matter, unilateral effects don't turn on

market definition.

regardless of whether the case is framed as a merger

generating high concentration within a narrow market or

is the loss of direct competition between the merging

I mean -- now, this

They make the point here that, "As

The economic analysis is the same

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

107

firms within a broader market."

Okay, do you believe it?

If you believe it, do

the latter.

Don't let me make market definition the

linchpin of the case.

lose that case in the district court, you might have to

appeal it, and you might have to establish good law, but

that's how you are going to get to a place where this

unilateral effects theory is more powerful, and it has

the foundation that you are going to need to go forward

and win your cases.

Take it away from me.

You might

Thanks.

MR. SCHMIDT:

Jon, do you have any response to any of that?

Thanks, Dan.

assume you are in almost complete agreement.

PROFESSOR BAKER:

That was terrific, Dan and

Rich.

I think I have to switch now from being the

even-handed law professor to actually take a point of

view here.

Dan wants to put the agency in a box.

He says,

"If you define a narrow market, I am going to say it's

gerrymandered to evade market definition and avoid

recognizing the plain fact of competition from Lawson

and whoever all these other guys are, so you are going

to lose."

I am going to explain to the court that you are talking

Then he says, "If you define a broad market,

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

108

about effects that are only in 20 percent of this broad

market.

don't meet the substantiality test of Section 7."

Therefore, Dan says, "I am going to win either way.

Don't bring these cases."

that was the implication --

MR. WALL:

They're too small, they're de minimis, they

He didn't quite say that, but

Clearly I would never say that.

Give

me a break.

PROFESSOR BAKER:

That is not a happy box to be in, so let's see

Only against Dan's clients.

what we can do to kind of get ourselves out of it.

Now, Rich says, you basically have two choices.

You take the broad market or the narrow market, and work

with it.

you define.

were getting at -- and Bobby, too, earlier in the

conversation.

unilateral effects to the judge?

But the important question isn't what market

That it is really what both Dan and Rich

It is what is intuitive in explaining

What Dan wants to do, either way, in the box

that he puts you in, is to be able to say, "There are

lots of rivals, so the merger partner can't be an

important competitive constraint."

the box for Dan.

Government is that your eye isn't on the ball.

to say, "Wait a minute, the key issue here is that the

That is the point of

And the answer to that for the

You have

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

109

merging firm didn't price higher before because of

competition from the merger partner."

Sure, there was some competitive constraint from

all the other rivals, but what you are losing with the

merger is an important competitive constraint that will

make a difference.

customers also, like Mercedes-Benz and Lexus, but look

at their documents.

they care about, BMW, and when you look at the diversion

ratios and the margins that our expert, Dr. Shapiro, has

computed, they show you the same thing.

of getting out of the box by changing the focus from who

all these other rivals are to the fact that there is a

competitive constraint from the merger partner, which is

the essence of the unilateral effects case in the first

place.

Yes, I concede that, what, Audi

When they are pricing, they also --

It's a matter

Whether you articulate it as a submarket or, in

the economic analysis in the broader market, that's the

story that the Government needs to tell.

MR. WALL:

Look, the box exists.

This is the problem.

I didn't

create it.

The box exists.

What

you have now is choices for what is the optimal strategy

in a world of boxes.

is -- in a trial dynamic, that it is a good idea to

fight any issue, any issue at all, where there is going

You know, I don't think that it

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

110

to be a lot of evidence on the other person's side, and

they are going to be able to marshal it up and bash you

with it day after day.

And we make strategic retreats

all the time in trials.

We make strategic retreats.

And, you know, I do think that there are going

to be cases in which the -- while the box is there,

there is a very credible way of going, of saying,

"Sure" -- I mean, just take Oracle.

companies compete, no doubt about it.

would never -- far be it from us, for the Government, to

suggest that they don't compete, but we still believe

that we can establish that the rivalry between the

merging firms has substantial effects that are distinct

from the rest of the rivalry in the market."

the approach that I am saying that I think would

probably be more effective.

MR. PARKER:

"Sure, these

We don't -- we

And that's

I think Dan and I are in total

agreement on that, and as I have said, to go into a case

and simply say I am not relying on Philly Bank, I don't

need it, don't need a presumption, because I have got

the goods on these folks, I don't need it, I think that

can be extremely effective and would certainly mesh well

within the current case law.

MS. FENTON:

Yes, but, Dan and Rich, doesn't

that necessarily get you pretty close to an analysis

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

111

that will focus on, because of the uniqueness you just

emphasized, the disturber in the marketplace, the

maverick, that you sort of go down that line of analysis

as a necessary consequence of the approach you're

advocating?

MR. WALL:

necessarily be a maverick.

one possibility here, that the merger is taking on a

maverick or something like that, but, you know, just in

the standard differentiated product model, you know,

spatial competition or something like that, there's

nothing -- it's completely coherent to say that I am

going to draw the big circle around a bunch of

competitors, but that in this particular, you know,

sector of that circle, by the way, which is $100 million

of commerce a year, so it's a lot that you -- you know,

you shouldn't just be indifferent to it, that most of

the competitive interaction is between these two brands.

To me, that is a perfectly coherent case that I

personally would not muck up by trying to say that they

didn't have competition from the rest of the people in

the box.

Well, I mean, it doesn't have to

MS. FENTON:

It could be, I guess that's

But you almost seem to be

suggesting that the district court judge will know it

when he sees it.

I'm wondering what's the criteria that

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

112

you would offer him for identifying that particular

unique competition.

MR. WALL:

Oh, I offer nothing special other

than the unilateral effects analysis as it is

articulated in the Guidelines.

mean, from everything I have heard and read, there

appears to be no one who can actually explain where the

35 percent threshold comes from in the Guidelines.

got put in there somewhere along the way and without a

specific economic rationale.

I just would not -- I

The real intuition is that if a large group of

customers find the merging firms to be their next best

substitutes, that you could have a problem that won't be

addressed by other firms.

that theoretically.

I'd have no problem putting on a case under that theory.

MR. PARKER:

I don't have a problem with

It makes perfect sense to me, and

And it turns on what the company's

documents say, as I said, and it turns on what the

customers say, importantly.

I think the 35 percent threshold, by the way, is

a lose-lose situation for the Government.

If you do

find effects below 35 percent, then, you know, Dan

quotes the 35 percent against you, and if you are in 55

or 60 percent, which the Government usually is, it

doesn't matter.

So, I don't see -- I think the

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

113

Guidelines' 35 percent not only has a -- has no real

rationale that I've ever seen, but more importantly,

from your point of view, and since I am taking the

government position, I think it's bad for the

Government.

MR. WALL:

Again, there is a comment I want to

make about Judge Walker's opinion in Oracle and what he

was saying about this notion that you have to have a

monopoly or something like that.

line that people are talking about.

This is actually the

He says:

"In a unilateral effects case, a plaintiff is

attempting to prove that the merging parties could

unilaterally increase prices.

must demonstrate that the merging parties would enjoy a

post-merger monopoly or dominant position at least in a

localized competition space."

Accordingly, a plaintiff

As a participant in that battle, I would urge

you to consider that the emphasis is on the last clause,

the "at least in a localized competition space."

certainly weren't arguing that a unilateral effects case

required a merger to monopoly, never made that argument;

never said anything close to that argument.

said is that the concept required that there be some

identifiable space -- you know, group of customers -- in

which there were not good substitutes to the merging

What we

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

114

parties.

That's not terribly far off than what the

Guidelines say themselves.

We were contesting factually whether that

existed in the case, not to get too much into the

details.

identifiable space like that in which SAP, which is far

and away the largest business applications provider, was

not a good substitute for Oracle or PeopleSoft.

contesting that.

were saying that factually.

The Government was saying that there was an

We were

We said that that didn't exist.

And I believe that what Judge Walker was saying

there -- and I know, you know, it has been

interpreted -- and frankly, not unreasonably given the

language he used -- to say something grander -- but what

I think what he was saying is that you at least have got

to demonstrate that there is that space where there is

this -- some kind of dominance by the merging parties.

I wouldn't -- you know, I wouldn't read it as being a

whole lot more than that.

He does go on to worry about whether this is a

backdoor way of creating submarkets, and that's a

legitimate worry.

lot of people have raised that, whether unilateral

effects is a backdoor way of getting into submarkets,

but rather than decrying this as setting up a standard

He's not the first to raise that.

For The Record, Inc.

(301) 870-8025 - www.ftrinc.net - (800) 921-5555

115

which is impossible to meet, if I were litigating on

behalf of the Government, I would argue to reconcile it

with the Guidelines rather than create a conflict.

PROFESSOR BAKER:

May I add something on that?

MR. SCHMIDT:

PROFESSOR BAKER:

Oracle opinion in front of me.

is another place -- a second place in the opinion where

he doesn't use that localized competition language,

where he says something that sounds a lot stronger about

the merger to monopoly.

maybe it's a different point, but on the same general

issue -- that comes up when I hear, you know, "throw out

the Merger Guidelines" or "revise

This text is long and has been trimmed here. Open the source document for the complete record.

This is a copy of a public record, reproduced as it was published. It is not legal advice, and it may not be the version a court would rely on. Check the official source before you cite it.