Participant Presentations And Session Transcript For June 22

This document is available in two formats: this web page (for browsing content) and PDF (comparable to original document formatting). To view the PDF you will need Acrobat Reader, which may be downloaded from the Adobe site. For an official signed copy, please contact the Antitrust Documents Group.

Page 1


2 and




9 THURSDAY, JUNE 22, 2006






19 9:30 A.M. TO 4:00 P.M.

24 Reported and transcribed by:

25 Susanne Bergling, RMR-CLR

Page 2



3 Chief, Legal Policy Section

4 Antitrust Division, Department of Justice

5 and


7 Attorney

8 Bureau of Competition, Federal Trade Commission


11 Morning Session:

12 Patrick Bolton

13 Kenneth G. Elzinga

14 A. Douglas Melamed

15 Janusz Ordover

17 Afternoon Session:

18 Tim Brennan

19 John Kirkwood

20 Janet L. McDavid

21 Steven C. Salop

22 Frederick R. Warren-Boulton


Page 3

1 C O N T E N T S


4 Introduction

5 Presentations:

6      Kenneth Elzinga

7      Janusz Ordover

8      Patrick Bolton

9      A. Douglas Melamed

10 Moderated Discussion

11 Lunch Recess


14 Introduction

15 Presentations:

16      Tim Brennan

17      John Kirkwood

18      Janet L. McDavid

19      Steven C. Salop

20      Frederick R. Warren-Boulton

21 Moderated Discussion


Page 4

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

2 - - - - -

3 MR. POTTER: Thank you for coming, everybody.

4 This is the first substantive hearing on predatory

5 pricing from the Section 2 hearings. My name is Bob

6 Potter. I'm the Chief, Legal Policy Section, Antitrust

7 Division, Department of Justice, and I will be the lead

8 moderator for this morning's session. Sitting to my

9 left is Pat Schultheiss, an attorney with the Federal

10 Trade Commission's Bureau of Competition's Office of

11 Policy and Coordination. She will be the co-moderator

12 for this morning and the lead moderator this afternoon

13 on the buy-side predatory pricing.

14 Before we start, just a couple of housekeeping

15 things that I need to say. One, for the courtesy of the

16 audience and the panelists, please turn off any cell

17 phones, Blackberries or other devices that may make

18 noise during the hearing.

19 Second, the restrooms. The men's restroom is

20 out the double doors to the left, on your left. The

21 ladies restroom is out the double doors, past the

22 elevator bank, to the left, and I saw this morning that

23 neither of them had hot water, so, if you want hot

24 water, you're out of luck.

25 MS. SCHULTHEISS: There is no place in the

Page 5

1 building that has it right now.

2 MR. POTTER: Third, and perhaps most important,

3 in the unlikely event that there is an emergency in the

4 building, please calmly and quickly go out the doors to

5 your right and down the stairs. The Federal Trade

6 Commission has a policy of meeting in the Sculpture

7 Garden, which is on Constitution Avenue. If you don't

8 know where it is, just follow the line of people leaving

9 the building, and I am sure you will get there.

10 This morning, we are very grateful for having a

11 very distinguished panel to talk with us about predatory

12 pricing and Section 2. Our panelists are Ken Elzinga,

13 Professor Ken Elzinga of the University of Virginia;

14 Professor Janusz Ordover of New York University;

15 Professor Patrick Bolton of Columbia University; and

16 Doug Melamed of the law firm Wilmer Hale and former

17 Deputy Assistant Attorney General of the Antitrust

18 Division and Acting Assistant Attorney General of the

19 Antitrust Division.

20 The format for this morning is each of the

21 panelists will give a 10 to 15-minute presentation, then

22 we will have a short break, and then we will have sort

23 of a moderated round table discussion for the rest of

24 the time.

25 We want to thank the panelists. I'll introduce

Page 6

1 them each before their speech as opposed to giving

2 everybody's introduction right now, and for the first

3 instance, I will tell you that although I'll give you a

4 short description, a much longer and better description

5 is contained in the biographical information that we

6 have.

7 Our first speaker this morning is Professor Ken

8 Elzinga of the University of Virginia. Professor

9 Elzinga is the Robert C. Taylor Professor of Economics

10 at UVA. He has a long and distinguished teaching career

11 at UVA, having been a faculty member there, although I'm

12 sure it doesn't look like it, for over 40 years.

13 Even more importantly for today's purposes,

14 Professor Elzinga is a creative and prolific academic

15 writer, having authored more than 70 economic articles,

16 a number of which have focused on predatory pricing.

17 In addition, perhaps even more importantly,

18 Professor Elzinga has been an expert witness in some of

19 the most important predatory pricing cases in the

20 history of antitrust, including Brooke Group,

21 Matsushita, and most recently, Spirit Airlines.

22 With that, please join me in welcoming Professor

23 Elzinga.

24 DR. ELZINGA: Thank you, Bob. I am going to

25 speak from the table here if that's all right, and I

Page 7

1 have got 15 minutes, max, to talk about predatory

2 pricing. That's a big topic. So, hold on to your

3 seats.

4 As was mentioned, I was the economic expert for

5 the defendants in the last two Supreme Court cases on

6 predation, the first one being Matsushita -- that really

7 dates me for some people in this crowd -- and then

8 Brooke Group or what I still call Liggett v. Brown &

9 Williamson, and then also, as was mentioned, I was

10 involved more recently in a predatory pricing case,

11 Spirit Airlines v. Northwest. I did an economic

12 analysis for Spirit, a so-called low-cost carrier. This

13 case had a happy landing for Professor Ordover at the

14 district court level, it had a happy landing for me at

15 the circuit court level, and the final destination of

16 this case is still unknown, but I hope to make a few

17 remarks about it later.

18 When I first started speaking about this

19 subject, before a number of you in this room were even

20 born, there was not much economic analysis embedded in a

21 predatory pricing case. You basically answered two

22 questions. Were prices declining in the market -- not

23 necessarily below cost, mind you, just going down -- and

24 did the defendant generate documents with pugilistic or

25 militaristic metaphors? "We are going to cut off their

Page 8

1 air supply. We are going to squish them like a bug."

2 If I had to pick two events, I am just doing a

3 brief intellectual history here, if I had to pick two

4 events that changed all this, it would be the Court's

5 opinion in Matsushita with its famous line that

6 predatory pricing schemes are rarely tried and even more

7 rarely successful. That statement was based on the

8 Court's exegesis of research about predatory pricing in

9 the economics literature. Almost all of this research

10 suggested that predation would be a strategy that would

11 be difficult to pull off.

12 And the second event was the publication of an

13 article by Don Turner -- the first Assistant AG to

14 enlist an academic economist in the front office, that

15 should always be pointed out -- and Phil Areeda in the

16 Harvard Law Review. It's the most often cited article

17 in antitrust scholarship, led to the Areeda-Turner Test.

18 Now, for this audience, I don't need to review

19 that article or that test, but let me mention for the

20 record how powerful was the hidden economic logic in

21 this famous test by using an iconic product from

22 Matsushita, a 19-inch black and white portable TV set, a

23 consumer electronic products my students today cannot

24 even imagine.

25 Let's say -- and these numbers are not way

Page 9

1 off -- that this set was sold by Toshiba, one of the

2 defendants, to Sears for $95, and the average total cost

3 was $100, but the average variable cost was $90. So, we

4 have ATC equals 100, P equals 95, AVC equals 90. Almost

5 everyone at the time believed Toshiba was selling below

6 cost. After all, how could Toshiba survive with that

7 type of price-cost relationship? And it took an

8 instinct for economic reasoning or a recollection of a

9 price theory course to realize that such a price was

10 above the shut-down point, it was cash flow positive,

11 and that Toshiba was better off making the sale to Sears

12 than not making that sale, and the Areeda-Turner article

13 convinced a lot of people, including a lot of people in

14 this building and a building nearby, of something that

15 economists have known since Alfred Marshall, and that

16 is, in economics, what happens at the margin really does

17 matter.

18 What was missing from Areeda-Turner was a way of

19 thinking about the period of recoupment. They set the

20 stage for a more sophisticated -- I did not say highly

21 sophisticated -- but a more sophisticated economic

22 analysis that the Court adopts in Brooke Group. The

23 Court in Brooke Group recognized that even if a firm

24 charged a price below cost, whatever was the cost

25 benchmark, if the firm couldn't recover its losses, it

Page 10

1 was difficult to make a case for antitrust enforcement,

2 because the aspiring predator would simply shoot itself

3 in the foot if there was no recoupment, and this

4 economic logic behind plausible recoupment entailed two

5 analytical constructs.

6 The first one is real clear in Brooke Group and

7 the second one is not transparent. The first is the

8 recognition that predation is like a capital

9 expenditure. In Brooke Group, the Court cites a paper

10 by David Mills and me entitled "Investment in

11 Predation." Economists have always recognized that a

12 dollar invested today requires more than a dollar in

13 future products because of the time value of money, and

14 Brooke Group understood that and applied that logic to

15 predatory pricing, that losses from predation need to be

16 recouped and not just on dollar-for-dollar basis.

17 The second point follows from the first: Unless

18 entry and exit conditions are symmetrical, the

19 recoupment returns for the aspiring monopolist must be

20 enjoyed for a longer time period than the time frame in

21 which the aspiring monopolist shouldered the cost of the

22 predation strategy, and I could just do a footnote here

23 on Matsushita and how much the world has changed.

24 The plaintiffs in Matsushita thought they were

25 making a good case for their side by arguing that the

Page 11

1 Japanese charged prices below cost for years and years

2 and years, over a decade, not recognizing that the

3 longer is that time period, the more difficult it would

4 be -- indeed, I think mathematically impossible given

5 the power of compound interest -- to ever make up the

6 gains.

7 For those of you who are attorneys, and that

8 would be most of you in this room, I'll tell you what I

9 find to be a fascinating war story from Matsushita. I

10 did some back-of-the-envelope calculations as to what a

11 19-inch black and white TV set would have to sell for

12 under the plaintiff's argument that predation had gone

13 on for 15 years, that is, these sets had been sold below

14 cost for 15 years. What would a 19-inch black and white

15 TV set have to sell for? And I found it would be like

16 $800 into infinity.

17 Now, I don't know if this is one of the things

18 that economists talk about when we are not in the

19 presence of antitrust lawyers. The antitrust lawyers

20 thought, don't ever make that argument on the stand,

21 because the plaintiffs will say, well, even the

22 professor on the other side says the television sets

23 will sell for $800 a year into infinity because of this

24 case. And I said, no, that can't be. They can't sell

25 for that much. They sell for $100 now. They are not

Page 12

1 going to go up to 800, trust me.

2 But my point is, the predator wants the period

3 of recoupment to be long, not the period of predation to

4 be long. The financial rewards that a successful

5 predator is going to enjoy is the present value of the

6 sum of each period's future return once the target has

7 conceded the battle.

8 Now, remember, a business firm has some hurdle

9 rate or internal rate of return before it signs onto any

10 investment project. Signing on for a predatory pricing

11 strategy to an economist is no different. The higher is

12 the hurdle rate, the bigger and longer the monthly

13 returns have to be during the period of recoupment.

14 And Grant, if you could show my first slide,

15 please.

16 In my experience, if one plays with the math

17 that I have at the top, which shows the monthly

18 sacrifice and the hurdle rate and the time period versus

19 the monthly return, it's hard to look at past episodes

20 of predation and come up with examples where recoupment

21 is mathematically possible. To my mind, when I try to

22 teach my students just the basic economics of the

23 elementary price theory level class, the important

24 asymmetry for predation is the one in the little box at

25 the bottom, if you can see it on the slide, slow entry

Page 13

1 but quick exits by target firms.

2 Putting the math aside, putting even the

3 diagrams aside, if there is slow entry but quick exits

4 by target firms, then there's a possibility that

5 predation can be successful. There's got to be, in

6 other words, an economic asymmetry between exit and

7 entry conditions in the market, and think about what

8 that means. In most markets where entry is easy, exit

9 is easy. So, predation simply won't work in those

10 markets. And in like fashion, in markets where entry is

11 difficult, that helps an aspiring predator, exit will be

12 slow, and that is bad news for an aspiring predator.

13 So, what the successful predator needs is a market

14 setting where exit is quick, but entry or supply

15 expansion is slow.

16 Now, in the Spirit-Northwest case, one of the

17 factors persuading me that predatory pricing was

18 plausible or rational for Northwest was because the exit

19 of Spirit, that was the target LCC, the target low-cost

20 carrier, took place quickly, but re-entry and supply

21 expansion was difficult. Spirit Airlines pulled

22 capacity out of Detroit quickly when Northwest cut its

23 fares in the two markets that Spirit served, but Spirit

24 could not enter and expand rapidly during Northwest's

25 recoupment period, because Spirit faced an entry barrier

Page 14

1 in the form of access to gates at the Detroit Airport.

2 Now, I went into the Spirit Airline case as

3 someone from Missouri or Chicago, maybe either metaphor

4 fits, but I ended up concluding that Spirit was a victim

5 of predatory pricing by Northwest, and I'll just say as

6 an aside, this is a case in which Fred Kahn should have

7 testified and not myself. Professor Kahn knows more

8 about the economics of airlines than most any group of

9 economists combined, but he was unable to participate,

10 though he was convinced that predation took place, as I

11 slowly -- kicking and screaming -- came to conclude.

12 The pricing trends in the Spirit case are a

13 textbook example of what predatory pricing would look

14 like. If I could have the first slide, this shows the

15 prices in the Philadelphia area -- I think the first

16 one -- yes, in Philadelphia. There were two city pairs,

17 Detroit, Boston and Philadelphia, and you will see that

18 Northwest prices in both of these are high. Spirit

19 enters; Northwest prices fall dramatically. Spirit

20 exits; Northwest prices jump up. If you show the other

21 slide, you will see basically the same scenario.

22 Now, these price trends -- I want to stress

23 this -- they are merely suggestive. They are not

24 dispositive of predatory pricing. Once a pricing

25 scenario like this is observed, then there follows the

Page 15

1 mind-numbing exercise of comparing revenues with some

2 measure of variable costs, and this is a difficult task

3 in the best of circumstances. It is by no means simple

4 in the airline passenger industry. In the Spirit case,

5 this was a battle between Professor Ordover, Janusz, for

6 Northwest, and Dr. Dan Kaplan was the economist for

7 Spirit. There was also a recoupment analysis done by my

8 colleague David Mills.

9 Briefly, from my perspective, going back to the

10 little box on the bottom of my first slide, one key to

11 the success for Northwest was simply how quickly Spirit

12 exited and the duration of the recoupment period, and

13 that's consistent with the first slide that I presented.

14 I was going to show one more slide, but in the

15 interests of time, I am going to pass on that.

16 Let me conclude this way: Antitrust always has

17 surprises. That is one of the reasons I have enjoyed

18 being an antitrust economist all these years. Let me

19 close by mentioning the surprise for me in the Spirit

20 case.

21 At the last minute, Spirit's attorneys suggested

22 that a price below average total cost but above average

23 variable cost could be predatory, and the Circuit Court,

24 at the tail end of its opinion, seems to suggest that at

25 least in the market circumstances of this case,

Page 16

1 Northwest's conduct may have been predatory even if its

2 fare structure exceeds, as the Circuit Court put it, and

3 I'm quoting here, "an appropriate measure of average

4 variable costs."

5 Now, Spirit's attorneys were pleased with this

6 little present, I am sure. I can restrain my enthusiasm

7 for the way the Circuit Court closed out its opinion.

8 This might take us into a more European view of

9 predation under Article 82, where prices greater than

10 average variable costs might be construed as predatory

11 and where, as I understand that in Europe, there is a

12 continued interest in intent documents and there is no

13 recoupment requirement, again, as I understand it.

14 Like most economists, I can restrain my

15 enthusiasm for the misuse of intent documents. I hold

16 the opposite view here of what had been the conventional

17 view in antitrust. To me, pugilistic and militaristic

18 metaphors are a welcome signal, not of predation, but of

19 competition in a market that doesn't have a stodgy "live

20 and let live" oligopoly setting, and where you see those

21 documents, to me, the prima facie case is that

22 consumers, albeit not rivals, but consumers are the

23 beneficiaries of head-to-head competition and not

24 predation.

25 MR. POTTER: Thank you.

Page 17

1 DR. ELZINGA: Sure, thank you.

2 (Applause.)

3 MR. POTTER: Our second speaker today is

4 Professor Janusz Ordover. Professor Ordover is a

5 Professor of Economics and a former Director of the

6 Masters in Economics Program at New York University,

7 also Director of Competition Policy Associates in

8 Washington, D.C. I first met him when he was the Deputy

9 Assistant Attorney General for Economics in the

10 Antitrust Division.

11 While at the Antitrust Division, Janusz was a

12 member of the White House deregulation task force. He

13 guided economic analysis of antitrust enforcement and

14 acted as a major liaison between the Justice Department

15 and various regulatory agencies.

16 Professor Ordover has written extensively about

17 predatory pricing and has a great deal of experience as

18 an expert witness in predatory pricing cases. He was an

19 expert for the defendant in the Division's American

20 Airlines case, and he is, as Professor Elzinga said, an

21 expert in the Spirit versus Northwest case on

22 Northwest's behalf.

23 Professor Ordover, welcome.

24 (Applause.)

25 DR. ORDOVER: Well, while we're getting set up,

Page 18

1 thank you very much. It's always a pleasure to

2 participate in these kinds of hearings.

3 Predation, of course, marks a lot of my

4 antitrust life. The first time I unveiled my thinking

5 on the subject of predation was about 1980 at the FTC

6 hearings on predatory conduct, and at that time, I think

7 I was attacked -- Professor Willig and I were attacked

8 by Frank Easterbrook, David Scheffman and Mike Scherer,

9 so essentially from left to right, everybody thought we

10 were completely foolish, and Mike Scherer said it was

11 the worst antitrust paper ever written, unlike the

12 Areeda-Turner paper, obviously, has its own different

13 reputation.

14 And then, just a few years ago, it was my

15 misfortune to fly into Ken Elzinga, who has never seen

16 predation other than the case that I was involved in.

17 Something is wrong here. So, I don't know what's wrong,

18 but I guess maybe I will switch careers in my waning

19 years.

20 In any case, what I wanted to do today is to

21 quickly run through some of the ideas that I have been

22 toying with in the antitrust predation field for some

23 past 20-some odd years and perhaps follow up on some of

24 the comments that Ken made, although I will not try to

25 relitigate Spirit versus Northwest. This will have to

Page 19

1 await Northwest's exit out of bankruptcy. So, unless

2 they get into bankrupt predating, but you never know.

3 So, now we are in a holding pattern until somebody

4 coughs up some money and we can actually go back and

5 litigate the antitrust part of the airline life.

6 In any case, what I wanted to do was just go

7 through a few slides focusing essentially on some of the

8 issues that have been discussed over the years, and that

9 is how to analyze challenged conduct from

10 monopolization, particularly paying some attention to

11 predatory behavior.

12 I was going to simply jettison this whole talk

13 by simply saying one should have no price predation

14 cases, but I thought that would be too quick an exit, so

15 I have to torture you for a bit longer to convince you

16 maybe that we should think about it as a possible

17 solution to our woes in this antitrust patch, without,

18 at the same time, suggesting that we should throw out

19 all kinds of scrutiny of firms' conduct, which consists

20 of much more sophisticated pricing from other aspects of

21 what they do, behavior, what I would often call

22 competitive response package, which is I think a term I

23 coined for my testimony in U.S. V. American Airlines,

24 where actually American Airlines' behavior was not just

25 simply pricing but involved a lot of other things that

Page 20

1 the Government alleged were designed to, in fact, retain

2 or maintain or defend American Airlines' position at its

3 hub, Dallas-Fort Worth.

4 So, I'm always thinking about competitive

5 response packages as strategies designed either to exit

6 the rival or to prevent the rival from coming in or

7 possibly designed to contain the rival, and I think it's

8 the last category of strategies which I believe is of

9 great interest and perhaps should be given a little bit

10 more time than we often do.

11 But in any case, the question becomes, how

12 should the decision-maker delineate conduct that does

13 not harm competition by harming scarce rivals from

14 standard, day-to-day market interactions? And

15 economists have been pulling their hairs out since

16 Areeda-Turner, 1975 paper, so we are now in 31 years, 30

17 years of thinking about it, and there is no solution as

18 evidenced by the articles in the latest Antitrust Law

19 Journal, where everybody is still fighting over

20 important things but without actually coming to any

21 particular conclusion.

22 I have been associated over the years with

23 something called the sacrifice test, but I always

24 thought of sacrifice test actually as a version of the

25 welfare test. In other words, what attracted me and

Page 21

1 Professor Willig to thinking about the so-called profit

2 sacrifice approach to delineating procompetitive versus

3 anticompetitive conduct, or at least neutral from

4 anticompetitive conduct, was the notion that at least in

5 some well-defined range of circumstances, these two

6 tests ought to give a pretty close set of answers.

7 In other words, that one was not -- that is, the

8 sacrifice test -- was not somehow biased, setting aside

9 the difficulties of implementation, but it somehow was

10 not biased one way or the other against deterring what

11 would be anticompetitive conduct or what would not be

12 anticompetitive conduct, relating to too much conduct

13 that, in fact, would be harmful. We have been able to

14 show in a variety of circumstances -- in fact, these two

15 tests coincide for the very simple reason that a pursuit

16 of profit, which is the engine of market economies, in

17 fact, is a kind of behavior that generally or frequently

18 does, in fact, conduce to welfare maximization. Seeking

19 profit is a good thing, it is not a bad thing, and

20 therefore, it is not surprising that if you write down

21 your economic model correctly, or at least correctly for

22 the purposes at hand here, that in many circumstances,

23 these two tests will give you the same kind of answer.

24 So, there might be, however, a range of

25 circumstances in which these two tests fall apart by

Page 22

1 virtue of the fact that the basic condition under which

2 they do coincide is potentially difficult to meet, and

3 that condition is incomprehensively stated as the third

4 bullet on this slide, but the basic idea here is that if

5 the incumbent firm can effectively, without creating

6 additional distortions, extract profits by its pricing

7 strategies and other strategies, then any strategy that

8 actually lowers the profits relative to that extraction

9 ought to signal, at least as a first step to the

10 analysis, ought to signal that a firm may have some

11 other aims in pursuing that strategy, something that I

12 think Bernheim and Whinston have now been calling over

13 the years as trying to create market power in what's

14 called a noncoincident market, okay?

15 So, the action takes place in market A, assuming

16 we have it well defined, but the goal essentially turns

17 out to be gaining incremental power or preventing

18 erosion of power in some other market, which Schullman

19 called a noncoincident market, let's say, which could

20 be, in Areeda-Turner world, it could be the same market

21 but in the future day, okay? So, what's the meaning of

22 noncoincident market is actually a little loose, but

23 that's the term that at least Berheim and Whinston in

24 their fine unpublished monograph on exclusionary

25 behavior utilized as a view for analyzing this kind of

Page 23

1 situation.

2 So, it could be a setting in which the whole

3 thing works beautifully. An example is an inferior

4 source of supply, this is the second thing which I think

5 is quite ubiquitous, in which the incumbent firm is

6 faced with competition from another firm or a firm that

7 constrains its ability to exercise pricing freedom,

8 which provides an inferior product, and therefore,

9 enables the incumbent firm to earn supra-competitive

10 profits, at least profits higher than some rents, but

11 getting rid of that firm would, in fact, lift the

12 ceiling and therefore would enable the firm to raise the

13 price even higher.

14 The problem turns out to be that maybe exiting

15 that firm may be just very difficult; however, a

16 circumstance that we have analyzed, Willig and I, under

17 the rubric of systems competition, informs a view of the

18 circumstance in which actually disabling a component

19 that the other firm needs in order to be a full-fledged

20 parcel, bundle and bundle competitor with the incumbent

21 firm will, indeed, lift the ceiling and therefore enable

22 the firm to exercise incremental market power. So, the

23 idea that we have pursued, and the idea which I think is

24 actually fruitful, is that in many circumstances, the

25 goal of the competitive response package is not

Page 24

1 necessarily to kill or to weaken or to disable the

2 person that or the firm against whom this conduct is

3 being perpetrated, but rather, to try to lessen or

4 weaken some other kind of restraint which cooperates

5 complimentarily with the firm whose market presence is

6 being weakened.

7 I think if you look at these Microsoft cases,

8 some of which were discussed along the same lines, this

9 is a fruitful way of thinking about it, but you can

10 immediately see that the economics of the situation is

11 much more difficult than the one instance in the

12 Areeda-Turner case, which is drop the price below some

13 level of cost, you go perhaps profit-negative, assuming

14 you know how to calculate profits, you know how to

15 calculate revenues maybe, you know how to calculate

16 costs maybe, and you can compare the two and see what

17 happens, you are losing money, and as a result of which,

18 it is anticompetitive.

19 But in the situation like this, you don't have

20 to be losing money on anything unless you try to look at

21 the situation in a somewhat different way, which is

22 where the efficient component pricing rule tells you how

23 to look at that situation that I have just described.

24 The efficient component pricing rule for those of you

25 who are not regulatory freaks like myself is a rule that

Page 25

1 tells you what the price of a scarce bottleneck should

2 be if it does not involve any kind of profit sacrifice,

3 okay? So, ECPR is a way of thinking about pricing

4 access, pricing access to the component that is needed

5 in order for the firm to be a viable component system or

6 system competitor.

7 Another example along the same lines, which

8 again focuses on a complicated pricing strategy, not

9 simply dropping price below some measure of cost, was

10 discussed in Ortho v. Abbott. Actually, I worked for

11 Ortho in that case, and there the situation was, again,

12 packaged pricing of a very interesting kind, in part

13 interesting because the buyer insisted on firms offering

14 not only unbundled pricing, but also bundled pricing

15 with a different number of components put in. The buyer

16 needed to buy five tests. There was a regulatory

17 presence out there that required that every blood

18 screening used five tests at that time, I think now it's

19 six, and Abbott was the one that could offer five of

20 them, Ortho could only offer three.

21 Then the question was, could Ortho compete

22 against Abbott if it did not get the access to the

23 remaining two, either because the buyer could create the

24 bundle or because Ortho could buy the necessary input

25 and then resell the bundle? Again, in this case, it

Page 26

1 turns out that there is some discussion that potentially

2 Abbott was pricing the incremental two tests at levels

3 that were unprofitable, that violated some version of

4 what we called a second ago the efficient component

5 pricing rule.

6 What was very complicated in that case was, A,

7 that Ortho did not give me any cost data. So, I

8 couldn't say anything, whether it was true or not, but I

9 did derive the test on a napkin, so other than the

10 Laffer Curve that also was derived on a napkin, this is

11 probably the second most famous napkin in the history of

12 economics. But in any case, the point I'm making is

13 that in this case at least we had a way of dealing with

14 an issue, but we had no reason to explain why this was

15 going on, and I think that's a very important aspect of

16 any predation case, which is that the plaintiff makes a

17 clear connection between the conduct that is at issue

18 and the anticompetitive impact that is being challenged

19 as leading to this anticompetitive outcome down the

20 road.

21 Virgin versus BA, another complicated case that

22 pitted Bernheim against Schmalensee, actually a

23 beautiful battle -- I think it was Schmalensee -- of

24 battle in IO, in which, again, there was no simple

25 pricing strategy, but rather, a complex pricing strategy

Page 27

1 that Bernheim showed leads to an equilibrium in which

2 there's relatively cheap exclusion but in which no price

3 is technically below marginal cost, simply understood,

4 yet as we know, all of these tests that we have in front

5 of us do have some flavor of comparing something to

6 cost, and again, what Bernheim tried to demonstrate in

7 that case, that a simple comparison of price to

8 something like marginal cost may be a flawed way to go

9 if you put that pricing in a strategy that British

10 Airways allegedly developed in a broader context.

11 Quantity-forcing contracts, I think we will skip

12 that, only because we have to, A, rush, and B, we will

13 talk about it in the fall, so I am going to skip that

14 unless it comes up in questions.

15 Just because I don't believe that true price

16 predation is an antitrust offense that is of great

17 interest, it does not mean that we as economists and you

18 as enforcers do not have plenty to focus on. I believe

19 that business strategies, these competitive response

20 packages, that have a strong commitment value, are

21 actually a more relevant focus than just simply pure

22 price predation, which creates all types of problems as

23 these papers in ALJ demonstrate.

24 Commitment to discount, which is Virgin versus

25 BA, commitment to product design, commitment to defend

Page 28

1 lucrative markets, which I call the "new era" tying

2 models, network economies, commitments to effectively

3 raising rival's cost of competing, are the types of

4 strategies that we are now slowly beginning to

5 understand with the help of very fancy economic models

6 and beautiful game theory.

7 The question that I think we will have to leave

8 for Patrick to help us answer is whether or not we can

9 actually fashion workable tests that will take into

10 account these kinds of complications that economists

11 have been focusing on.

12 Thank you.

13 (Applause.)

14 MR. POTTER: Thank you, Janusz.

15 Our next speaker is Professor Patrick Bolton.

16 Professor Bolton is the David Zalaznick Professor of

17 Business. He began as Assistant Professor at U-Cal

18 Berkeley, then moved to Harvard. Then he was the John

19 H. Scully Professor of Finance and Economics at

20 Princeton University.

21 Professor Bolton's research and areas of

22 interest are in contract theory and contracting issues

23 in corporate finance and industrial organization. One

24 of his particular areas of research is the impact of

25 strategic economic game models on predatory pricing

Page 29

1 theory.

2 Professor Bolton, welcome.

3 DR. BOLTON: Thank you, Bob. It's a pleasure

4 and an honor to be on this panel.

5 Unlike Professors Elzinga and Ordover, I have no

6 experience as an expert, haven't had that pleasure, and

7 if you want, I'm a new entrant. We will see whether

8 this will elicit predatory response from the economists.

9 So, my interest, as Robert just alluded to, my

10 interest in this topic came from reading the original

11 McGee article, which claimed that predation couldn't be

12 a rational economic strategy, and, you know, I read this

13 article again and again, and I just was not convinced,

14 and this led me later on to write a theory piece with

15 David Scharfstein where we outlined how predation could

16 be a rational strategy if it took the form of financial

17 predation, and I will say a little bit more about that

18 in my presentation.

19 And then later, I had the good fortune of

20 meeting with Joe Brodley, who introduced me to the new

21 developments in policy under Brooke Group and

22 highlighted some of the problems with the new policy and

23 also some of the new opportunities and challenges, and

24 that then led to our, in my view, very fruitful

25 collaboration on our article, which I will make the

Page 30

1 centerpiece of my brief presentation today.

2 So, I thought I would start by saying first, you

3 know, where are potential areas of agreement among

4 economists and legal scholars and where there are still

5 areas of disagreement. I would argue that this is

6 relatively easy, that we are all in agreement on the

7 general definition on predatory pricing. Namely, it's a

8 price reduction that is only profitable because of the

9 added market power the predator gains from eliminating,

10 disciplining or inhibiting the competitive conduct, and

11 to summarize what both Professors Elzinga and Ordover

12 said earlier, you can distinguish two phases in any

13 predatory pricing episode, a sacrifice phase and a

14 recoupment phase. As Professor Elzinga wrote elsewhere,

15 you can think of predation as an investment in market

16 power. So, I would say that there is general agreement

17 on this characterization.

18 Where there is more disagreement is on policy,

19 and, well, there had been long disagreements on basic

20 economic premise, whether predation is an economically

21 rational strategy and how prevalent predatory pricing

22 episodes are. My sense is that this is an area of

23 convergence, at least on the first bullet point. I

24 think nowadays it is more and more widely accepted that

25 predation can be an economically rational strategy.

Page 31

1 On the second bullet point, I think there are

2 still some areas of disagreement, but I would argue that

3 over time, things have moved in the direction of

4 thinking of predatory pricing as being more prevalent

5 than we thought before and also more likely to succeed

6 than we thought before, in part because our initial

7 beliefs were built on writing, McGee's writing,

8 suggesting that it couldn't be rational, and those

9 writings, I would argue, are now obsolete.

10 There are, however, still very sharp

11 disagreements on the legal standard. Some people argue

12 that we should have simple rules. Others have argued

13 that we should always err on the side of

14 under-deterrence to reduce the risk of false positives,

15 and the policy under Brooke Group is characterized as

16 both being simple and under-deterring. I would argue

17 against this.

18 Now, let me skip the description of Brooke

19 Group, because I imagine most of you are familiar with

20 it, so it involves both a cost test and a recoupment

21 test, and let me emphasize potential problems first with

22 the new policy, and namely, when we look at the facts on

23 what happened post-Brooke, what we find is that since

24 Brooke, plaintiffs have not prevailed in a single case,

25 and almost all cases have been decided by summary

Page 32

1 judgment, and it is only very recently that we are

2 seeing some action on predatory pricing, particularly in

3 the case of Spirit versus Northwest.

4 So, what are the problems with the present

5 policy? Well, first of all, and I think we will discuss

6 this later on in the question time, I would argue that

7 the basic problem with the present policy is that the

8 cost test is highly unreliable. Professor Elzinga

9 earlier qualified proving a cost test as a mind-numbing

10 exercise. I would fully agree with that. I would say

11 that when you go into the details of trying to prove a

12 cost test, you will lose track of the economics of the

13 problem and of the case, and in particular, a very

14 narrow interpretation of the cost test, price being

15 below average variable cost, is a very poor proxy for

16 measuring profit sacrifice, which is what we are trying

17 to go after.

18 Another problem with current policy, we have

19 never gone to a point where we had to ask about a

20 possible efficiency defense on the part of the

21 defendant. There has never been any talk of applying

22 the same rigorous recoupment criteria that the plaintiff

23 has to fulfill on the defendant in proving an efficiency

24 defense. I would argue we should go in that direction.

25 But just to emphasize, I think that the major

Page 33

1 problem with present policy is its failure to focus on

2 the main issues, and those are what is the predatory

3 strategy, what strategy drives alleged predation, first

4 of all, and second, what are the possible dynamic

5 efficiencies and how do you balance procompetitive and

6 predatory effects? And this is where our article takes

7 off and proposes an alternative approach, which I would

8 summarize as taking away some weight off the cost test

9 and emphasizing instead intent, bringing back intent,

10 but intent as structured by an economic analysis, and so

11 this is what in my short time I want to briefly go into.

12 So, specifically, we are thinking that any

13 approach based on intent should be based on strategic

14 analysis of predatory pricing, and in our article, we

15 emphasize at least two well-proven strategies, which are

16 financial market predation and reputation effect

17 predation. We also discuss test market predation. Of

18 course, as Professor Ordover highlighted, predation can

19 take many different and complex forms, and one should

20 not necessarily reduce one's self to just those few

21 strategies, and one should allow for any

22 well-articulated and rational strategy that might be

23 used. I might comment on that later on.

24 Anyway, so what we argue in our paper is that

25 this approach has two advantages. One is that it can

Page 34

1 reduce the risk of false positives, and second, that it

2 puts the spotlight back on what we are really trying to

3 determine, which is discriminate between procompetitive

4 and anticompetitive effects, and there we can use intent

5 as our guide, evidence of intent as a guide to possible

6 defense, and what I mean by intent is not what Professor

7 Elzinga has referred to as militaristic and pugilistic

8 language, but evidence of a deliberate effort to exclude

9 and evidence of pursuit of a predatory strategy.

10 So, in our article, we outline five legal

11 elements to a predatory pricing test. Let me enumerate

12 them first, and then I will go into some of them in more

13 detail. The first element, which is straightforward, is

14 there should be a facilitating market structure. The

15 second element is the scheme of predation and supporting

16 evidence. Third, probable recoupment. Fourth, price

17 below cost. And those four elements would constitute a

18 prima facie case of predatory pricing.

19 I have put the fourth element in brackets here

20 to emphasize the fact that we try to de-emphasize the

21 cost test, and we would agree with the appeals court

22 opinion in the Spirit Airlines case that predatory

23 pricing which is above some measure of average variable

24 cost but below average total cost, that kind of pricing

25 could be predatory. Then, however, we add, if you

Page 35

1 de-emphasize the cost test, we want to add as a safe

2 harbor the -- allow for an efficiency defense.

3 So, how do we prove those elements? Well, some

4 of them are straightforward, and I will not go into -ï½­

5 so, facilitating market structure is any evidence of

6 market power. The scheme of predation and supporting

7 evidence, I want to give you an example of how you go

8 about doing this. So, I will in particular take out of

9 our article the example we have on financial market

10 predation, and so under this element, what is important

11 is to establish that the conditions to implement a given

12 strategy are present and to provide direct or

13 circumstantial evidence showing that this strategy is

14 being implemented.

15 Recoupment, again, this is relatively

16 straightforward. You would want to show evidence of

17 exclusion and disciplining of rivals, and we stressed

18 the idea that second, that you should emphasize probable

19 recoupment instead of actual recoupment, because what

20 matters is whether at the time when this strategy was

21 being chosen, whether at that time, at the time of the

22 information the incumbent had at that time, whether it

23 made sense to implement such a strategy, and we know, as

24 in our own investments in finance, we know that at the

25 time when we make a decision of investment, we make an

Page 36

1 analysis using this kind of cash flows that suggests

2 that we have a positive net present value investment,

3 but that is no guarantee that when we actually undertake

4 the investment, it will end up being profitable. So, we

5 would emphasize probable recoupment, and in particular,

6 put a lot of weight on market structure that makes

7 recoupment likely in the future.

8 Let me also emphasize here the "or related" in

9 brackets, and this is a point that Professor Ordover

10 emphasized, that recoupment shouldn't just be seen in

11 the narrow market where predation takes place. It could

12 be obtained in a related -- I forget the term you

13 used -ï½­

14 DR. ORDOVER: Noncoincident.

15 DR. BOLTON: -- noncoincident market.

16 On price below cost, I do not have much to add

17 to what I have said already except that in the paper we

18 emphasize that a better measure than average variable

19 cost would be average avoidable cost, and a better

20 measure for long-run average cost would be long-run

21 average incremental cost. I do not want to go further

22 into this, because making fine distinctions about these

23 definitions could end up being a mind-numbing exercise,

24 and it just highlights the difficulty with applying the

25 cost test.

Page 37

1 So, what I would like to emphasize, though, is

2 that we would argue that failure to meet the cost test,

3 in particular, failure to establish pricing below

4 average variable cost, should not be grounds for a

5 dismissal on summary judgment and that, in fact, the way

6 to go would be to balance the cost test with an

7 efficiency defense. So, I would argue that if you are

8 able to show that there was pricing below average total

9 cost but above average variable cost but that there was

10 absolutely no efficiency defense, plausible efficiency

11 defense provided, that that would then make a strong

12 case for predatory pricing.

13 So, the efficiency defense, we spent a lot of

14 time in the paper on that, because one of the weaknesses

15 of the policy under Brooke Group and the Areeda-Turner

16 Test is that it really neglects looking at efficiencies,

17 and so we would argue that an efficiency defense does

18 provide safe harbor in itself for price competition that

19 benefits consumers, and we distinguish between defensive

20 defenses and market-expanding defenses and provide in

21 the paper an approach to proving those defenses. So,

22 defensive defenses, we mean by unilateral best response

23 mainly and minimizing losses from unexpected market

24 developments, and as for market-expanding defenses, we

25 really mean here promotional pricing, learning-by-doing,

Page 38

1 and network externalities.

2 So, let me move on perhaps in the few minutes

3 that I have left with an illustrative example. How do

4 you prove financial market predation in a particular

5 case? So, very briefly, the theory here, you know, what

6 is financial market predation, why does it work?

7 Well, the reason why it works is because in

8 corporate finance, there are imperfections -- and there

9 is enormous literature on this -- there are

10 imperfections in capital markets due to agency problems

11 in lending, and as I have argued and have written in my

12 paper with David Scharfstein, a predator can take

13 advantage of those imperfections and drive out an

14 entrant by basically drying up financing.

15 So, how do you go about proving financial market

16 predation? So, we distinguish five essential

17 preconditions. One, the prey's dependent on outside

18 financing. The prey's outside funding depends on its

19 cash flow. Three, predation will reduce the prey's cash

20 flow sufficient to threaten its continued viability.

21 All these are fairly straightforward. Four, the

22 predator knows of the prey's dependence on outside

23 funding or can be assumed to know based on easily

24 accessible facts or rational conjecture. And five, the

25 predator can finance predation internally or has

Page 39

1 substantially better access to external credit than the

2 prey.

3 I think in the Spirit Airlines case, I quickly

4 looked at it, most of these elements you would be able

5 to establish.

6 So, the example we have in the paper is about

7 entry into the cable TV market in Sacramento. This is a

8 case that predates Brooke, and here are the facts. So,

9 this is an entrant with outside financing amounting to

10 $6 million, entered in a small district in the

11 Sacramento area, the Arden District, serving 5000 homes,

12 and the entrant's intention was, of course, to reach a

13 bigger market share and expand gradually in the

14 Sacramento area. The incumbent Sacramento TV company

15 responded to this entrant with drastic price-cutting,

16 and after eight months, the entrant exited. So, how

17 would we prove a scheme of financial predation here?

18 Well, first of all, the dependence on outside

19 funding, what do we know? What are the facts here?

20 Well, first of all, the prey obtained funds through a

21 loan, and the entrant's owners were unwilling to commit

22 more capital than they had initially. Secondly, outside

23 financing depends on cash flow. Well, the incumbent

24 targeted its price reductions on the entrant's customers

25 and potential customers, and that obviously had the

Page 40

1 effect of reducing cash flow. Predation will reduce

2 cash flow and threaten viability. Again, that is easy

3 to establish in this case.

4 The predator knows of the prey's dependence on

5 outside funding. Well, here it turns out there is

6 evidence, intent evidence, memorandum from the

7 incumbent's files that speaks of sending a message to

8 the entrant's bankers. Well, that's relatively easy to

9 establish here. And then finally, the predator has

10 better access to credit than the prey. Again, that is

11 an easy proof in this particular case.

12 So, let me -- sorry for having stepped over my

13 time -- so, let me just quickly conclude with

14 highlighting one potential concern with our approach,

15 and that is something that Posner mentions in his second

16 edition of his antitrust book, and he argues that one

17 concern one might have with evidence of intent is that

18 it's really "a function of luck and of defendant's legal

19 sophistication." So, we would argue that this concern

20 is reduced if the plaintiff is also required to prove,

21 as we articulate in our article, all the other elements,

22 and if what you are required to establish is the

23 implementability of a rational predatory strategy.

24 So, let me end with that.

25 MR. POTTER: Thank you very much.

Page 41

1 (Applause.)

2 MR. POTTER: Our final speaker today is also the

3 only lawyer on the panel, although Doug is very used to

4 dealing with economists, so I am sure it will not be a

5 problem for him to follow them.

6 Doug is a partner at Wilmer Hale, and he is the

7 co-chair of the firm's Antitrust and Competition

8 Department. He has significant experience in a number

9 of government investigations, both government and

10 private litigation, substantial antitrust counseling,

11 and some of that counseling in investigatory work, in

12 litigation work, has involved predatory pricing.

13 From 1996 to 2001, Mr. Melamed served as the

14 Principal Deputy Assistant Attorney General in the

15 Antitrust Division and then as the Acting Assistant

16 Attorney General in the Antitrust Division. He's a

17 prolific writer, a frequent speaker, always has

18 interesting viewpoints that are well thought out. His

19 most recent -ï½­

20 MR. MELAMED: Don't raise the bar, please.

21 MR. POTTER: -- his most recent article, which

22 appears in the summer 2006 Antitrust Law Journal

23 provides a thought-provoking commentary on whether there

24 are unifying principles under Section 2.

25 Mr. Melamed, welcome.

Page 42

1 MR. MELAMED: Thank you.

2 Well, I am a lawyer, and much though I enjoy

3 listening to economists and talking to them, I am going

4 to be talking as a lawyer now and giving a lawyer's

5 perspective on some aspects of the predatory pricing

6 issue.

7 Let me start by saying, I think Brooke Group was

8 correctly decided, an important decision, it brought

9 needed rigor and order to predatory pricing law, but I

10 am concerned about what has happened to it in the life

11 of the law. There is a kind of -- I do not know if this

12 is the right word -- a kind of rarefaction of Brooke

13 Group that I think has done some mischief, and let me

14 tell you what I mean.

15 As everyone knows, Brooke Group has proven to be

16 a defendant friendly standard. As Professor Bolton

17 noted, no plaintiff has won a predatory pricing case

18 post-Brooke Group. Not surprisingly, therefore, when

19 price is an element of the allegedly unlawful strategy,

20 the defendant argues that the standard to be applied by

21 the Court should be Brooke Group, and, of course, they

22 are entitled to do that, because if that's the law, they

23 ought to make that argument, and certainly I have done

24 that myself.

25 But if it is not a straightforward price-cutting

Page 43

1 case, if it is a little complicated, the plaintiff says,

2 "No, no, no, this was different, bundled discounts,

3 aggressive buying, low prices conditioned on exclusivity

4 or other preferential treatment and so on." So, you

5 have a legal dispute. Does Brooke Group apply? Is this

6 the right category, predatory pricing, in which Brooke

7 Group applies, or does the conduct at issue belong in a

8 different category?

9 And there is a kind of a notion that there is an

10 apparent precision of Brooke Group, the price-cost test

11 and the recoupment test, that is uniquely valuable but

12 uniquely applicable to predatory pricing, and one

13 consequence of this is that when the Court decides in

14 this kind of stovepipe analysis that the conduct before

15 it really should not be considered predatory pricing,

16 too often, courts seem to find themselves in a kind of

17 "deer in the headlights, what do I do now" posture, and

18 the result is incoherent decisions like LePage's or

19 courts affirming nonsensical and meaningless jury

20 instructions like Weyerhaeuser and basically a casting

21 about in the way that Professor Elhauge had spoke of

22 Section 2 as a kind of incoherent mess.

23 I think this stovepipe or essentialist way of

24 looking at predatory pricing has created these kinds of

25 dichotomies as categorization, and it has inhibited the

Page 44

1 development of a more robust antitrust jurisprudence,

2 one that can help courts make reasoned decisions about

3 conduct that they do not think falls into a precise,

4 well-established category, whether it be exclusive

5 dealing or predatory pricing or whatever.

6 Put differently, instead of inducing from Brooke

7 Group principles of broader application in the kind of

8 common law tradition which antitrust has in other

9 contexts involved, the process seems to have separated

10 predatory pricing from other forms of exclusionary

11 conduct, and it's done so because there has been in what

12 I call this rarefaction a number of propositions about

13 predatory pricing that are taken for granted or thought

14 to be true or thought to be unique to predatory pricing,

15 and I want to express some skepticism about that. There

16 is a lot of these propositions I have in mind, four or

17 maybe three depending on my time, and I want to express

18 skepticism either that they are true or that they are

19 unique to predatory pricing or perhaps both.

20 So, proposition one, to apply the price-cost

21 test, we need to select some term of art from the

22 economists as our measure of cost, average variable cost

23 or something like that. Now, this is a big topic. I

24 will make just a couple observations.

25 Almost everyone seems to agree that some kind of

Page 45

1 incremental cost is the right measure, because we want

2 to know whether the allegedly predatory sales cost so

3 much that either the defendant must have intended some

4 predatory scheme or, at the very least, that the cost of

5 the sales exceeded the amount consumers were willing to

6 pay for them and therefore resulted in a welfare loss.

7 Areeda and Turner say, "Well, marginal cost is

8 the right test, but it's hard to prove, so let's use

9 average variable cost as a proxy," and now we have this

10 debate for 30 years, "Well, average variable cost really

11 isn't a good proxy, we should use average long-run

12 incremental cost or average total cost, may depend on

13 the circumstances," and you all probably read the

14 article, too, the discussion paper which went through

15 this discussion at great length.

16 Why are we even having this conversation? Why

17 are we debating these categories about technical

18 economic jargon that might have made sense in the

19 Areeda-Turner world in 1975, a simple static price

20 series model, and you can draw the ABC curve, the

21 marginal cost curve, and you can talk about these

22 metaphors, what's going on in the real word, but that

23 doesn't make any sense in the real world as I have

24 experienced it as a lawyer.

25 Areeda talked about additional increments of

Page 46

1 output. I have rarely had a client say to me, "I'm

2 thinking of pushing more widgets off my production line.

3 How low can I go in price?" That's not how the problem

4 comes up in the real world, and if it looks like that,

5 there's a lot more going on.

6 The kind of predatory pricing problems I've

7 counseled clients on in recent years are things like

8 this: Price offerings to early adopters in a de facto

9 standards war; prices in two-sided markets; decisions to

10 assign a plant or an airplane to one market or one

11 segment rather than another. In these situations, I

12 think these terms of art that economists have, they are

13 very valuable in their models and their heuristic

14 exercises, don't have much value, and even if they have

15 value to the economists, they don't have much value to

16 the lawyer and the client.

17 What I find is valuable is saying to the client,

18 when I'm talking about costs, "What are the costs you

19 are incurring to engage in the strategy at issue that

20 you wouldn't otherwise have incurred?" Clients

21 understand that question, and it's not always a trivial

22 question, but I think it's one they can answer. So, I

23 think avoidable costs -- and I don't mean that as some

24 technical term, I mean simply as the but for costs of

25 the allegedly unlawful conduct -- is the cost measure,

Page 47

1 okay?

2 Proposition two, price-cutting is beneficial to

3 consumers, so we should therefore have a standard that

4 errs in favor of avoiding false positives. Then Judge

5 Breyer, in the wonderful "bird in the hand" metaphor, I

6 think most famously perhaps articulated that.

7 Here is my concern: Sure, price-cutting is good

8 for consumers, no question about that. So are all sorts

9 of other things that companies do for consumers. In

10 fact, as I understand, from what the economists tell us,

11 that innovation does a lot more for welfare than

12 improving allocative efficiency by some price cuts and

13 supra-competitive down toward competitive levels. So,

14 why don't we -- and innovation, by the way, could be

15 inventing the PC or it could be coming up with an

16 improved method of distribution because of tying

17 arrangements or because of exclusive dealing. It could

18 be anything that improves the value of the product to

19 consumers.

20 In fact, cost of sale reductions could be

21 beneficial certainly to a total welfare sense and

22 ultimately to consumers as well. So, why do we single

23 out price-cutting, which I don't think has any unique

24 benefits to consumers?

25 Now, there is one thing about price-cutting that

Page 48

1 is different, and that is it's unambiguously in the

2 interests of consumers. A product improvement, you

3 know, the car with the air conditioner might look like

4 it's better, but maybe consumers would rather have

5 better mileage. So, there is some ambiguity about

6 whether other forms of conduct benefit consumers, but

7 why do we have a legal superstructure built on the

8 premise that pricing is unique?

9 At some point, if we do that simply because it's

10 easier to identify the consumer benefit, don't we begin

11 to look like the economists searching for the keys under

12 the light post? At the very least, when the defendant

13 is able to show that his conduct is benefiting

14 consumers, why treat predatory pricing any differently?

15 Proposition three, the recoupment requirement is

16 central to and a great contribution to predatory pricing

17 law. Let me be clear. I strongly believe there should

18 be something like a recoupment requirement at least in

19 the sense of a market power screen; that is to say, a

20 plaintiff ought to have to prove that the allegedly

21 predatory scheme will pay off for the defendant by

22 creating additional market power or preserving market

23 power that will guard against -- kind of belt and

24 suspenders -- a mistake in the application of the

25 price-cost test, and it will preserve antitrust

Page 49

1 violations for those cases where there is competitive

2 harm, and we won't worry about the others.

3 I think, in fact, there should be such a screen

4 in all cases of exclusionary conduct. The problem is, I

5 think in many quarters, including some of my

6 predecessors this morning, the recoupment test is

7 understood to mean that the plaintiff should prove,

8 should quantify, the defendant's investment in the

9 predatory strategy and then quantify his

10 supracompetitive returns during the recoupment period,

11 discount them by risk and uncertainty and time, and

12 conclude that the recoupment exceeds the investment.

13 Now, I think evidence of that sort, on that

14 issue, whether introduced by the plaintiff or the

15 defendant, should be relevant in a predatory pricing

16 case, because it certainly illuminates the likelihood

17 that what is going on here is some exclusionary conduct,

18 but I am very skeptical of the notion that that should

19 be an element of the offense. It clearly complicates

20 the proceedings, increases costs. It may be an

21 impossible burden for the plaintiff in a multi-market

22 reputation effect recoupment story.

23 If taken literally, you would have to go to a

24 profit-maximizing standard to figure out the defendant's

25 investment in the predatory strategy, because you

Page 50

1 wouldn't be asking simply what did it cost him below

2 cost, you would be asking how much in profits did he

3 sacrifice. It's not necessary in order to identify

4 anticompetitive conduct, because if we think we got the

5 price-cost test right and the guy is selling below cost,

6 you can actually, it seems to me, infer that he expects

7 to recoup. It's not needed, because the market power

8 screen will identify the cases of competitive harm. And

9 finally -- and this is a point that I don't know that

10 it's original to me, but I haven't seen it before -- I

11 think it is an illusion that we're measuring something

12 about the welfare effects of the conduct when we use a

13 recoupment screen.

14 The welfare question in predatory pricing is

15 whether the welfare gains, consumer or total, during the

16 rivalry period, the competitive period, are greater than

17 or less than the welfare costs, consumer or total,

18 during the recoupment period, but the recoupment test

19 doesn't measure either of those. The recoupment test

20 measures producer surplus in the competitive period

21 versus producer surplus in the recoupment period, and it

22 doesn't take a whole lot of imagination to think of

23 situations where the results could be different, where

24 you could have, for example, recoupment but no welfare

25 loss from an allegedly predatory strategy.

Page 51

1 So, proposition four, in applying the Brooke

2 Group price-cost test, the price of the product at issue

3 is the appropriate price to compare to cost. That in my

4 view is only partially correct. Obviously you look at

5 the price of or the revenues generated by the additional

6 sales attributable to the predatory conduct. You don't

7 look at the price of, of course, the inframarginal

8 units, the units that would have been sold anyhow,

9 because those units didn't exclude the rival or at least

10 they didn't exclude them by reason of the

11 anticompetitive conduct.

12 But that's not all there is to it. Suppose

13 we're in a two-sided market. Suppose you're cutting

14 price on circulation of the newspaper in order to

15 generate more readers and therefore more advertising

16 revenues. Surely you want to take into account the

17 incremental advertising revenues.

18 Suppose you have complimentary revenues. You

19 know, the Government didn't accuse Microsoft of

20 predatory pricing because the browser was free when

21 bundled with the operating system. Because it was a

22 plausible story that it increased revenues, we didn't -ï½­

23 increased revenues for the operating system. What about

24 revenues lost from inframarginal sales; that is to say,

25 the sales that the defendant would have made anyhow even

Page 52

1 if he had not engaged in the predatory scheme, but he

2 would have made them at a higher price if he hadn't cut

3 prices?

4 To me, another way of putting that question is,

5 are we concerned with the incremental revenues or the

6 revenues from incremental sales? The law chooses wisely

7 in my view the latter. It ignores the loss of

8 inframarginal revenues, I think -- I know Professor

9 Bolton may disagree with this -- I think the law wisely

10 ignores that, because if you want to go into those lost

11 inframarginal revenues, you have to have a profit

12 maximization test, you know, what would have been the

13 profit-maximizing outcome of the strategy, and that is

14 in most cases going to be virtually impossible it seems

15 to me for the Court to figure out and surely impossible

16 for the firm to figure out in real time when it's trying

17 to comply with the law.

18 As implied by my discussion a minute ago with

19 the recoupment test, it's not going to correlate with

20 the welfare trade-offs you are looking at, although it

21 may illuminate a little bit, but most important, it

22 seems to me, is that price cuts on the inframarginal

23 purchases, price cuts until they are below some measure

24 of cost for the incremental units, enhance welfare, and

25 they enhance efficiency, and we all know that story,

Page 53

1 right, going toward the competitive outcome, and you

2 reduce dead weight loss.

3 So, it seems to me that we ought to ignore

4 inframarginal revenues. I didn't mention this earlier,

5 I meant to say it, but costs ought to include

6 opportunity costs, ought to include the cost of moving

7 allocating assets to the predatory scheme rather than

8 somewhere else. That's part of the avoidable cost it

9 seems to me.

10 Forgoing inframarginal revenues in my view

11 shouldn't be treated as an opportunity cost, at least

12 not for this purpose, because they are not a cost. They

13 don't involve the consumption of any resources. They

14 are simply a transfer payment actually from producer to

15 consumer, and I don't see why we should take that into

16 account in the calculation.

17 Okay, so what does all this come down to? It

18 comes down to, I think, predatory pricing law ought to

19 be looked at in a common sense way. Predatory pricing

20 law ought to be looked at straightforwardly as pricing

21 that is not efficient, that is to say, pricing whose

22 avoidable costs exceeds the revenues generated by the

23 sales in question, and thus, pricing that reduces

24 welfare during the rivalry period.

25 If it's efficient pricing, if it increases

Page 54

1 rivalry during the welfare period, the competitive

2 period, because consumers value the marginal units, or

3 the compliments that they generate, more than the

4 avoidable costs of those units, it seems to me we ought

5 to call this competition on the merits, and it ought to

6 be lawful.

7 Now, looked at this way, it seems to me,

8 predatory pricing isn't all that special. If we think

9 of it in this common sense way and simply ask where the

10 conduct is efficient in this sense, we have both in my

11 view a sound approach to predatory pricing and the

12 beginning of a more general theory of exclusionary

13 conduct that can avoid the pitfalls of the stovepipe

14 analysis to predatory pricing.

15 MR. POTTER: Thank you, Doug.

16 (Applause.)

17 MR. POTTER: Before we begin our round table

18 discussion, we will take a short maybe ten-minute break

19 to let people use the facilities and stretch, and if

20 they have to call their offices, call their offices.

21 MS. SCHULTHEISS: And coffee upstairs if they

22 need it, 7th floor, if you need coffee or water.

23 (A brief recess was taken.)

24 MR. POTTER: In deciding how to handle the round

25 table discussion, I thought maybe one of the effective

Page 55

1 ways of doing this would be to put up various

2 propositions on the screen and then ask for agreement or

3 disagreement among the panelists. If there's agreement,

4 fine, we have reached a consensus point, we can go on,

5 and we have solved the issue, and if there is

6 disagreement, we can debate the issue. The panelists

7 all have this in front of them, so they do not have to

8 turn around and look at the screen every three seconds.

9 Do you want to put the first one up?

10 MR. POTTER: I think Professor Bolton already

11 indicated there might be convergence around this point,

12 but there used to be economic literature saying that

13 predatory pricing was an irrational business strategy.

14 The proposition for the consideration of the panelists

15 is that predatory pricing can be a rational business

16 strategy. Is there anyone on the panel who disagrees

17 with this?

18 DR. ELZINGA: There is no disagreement. I would

19 like to correct one matter for the record, at least I

20 think this is a correction. Patrick indicated that his

21 reading of John McGee's classic article on predatory

22 pricing in the Standard Oil case or the lack thereof

23 indicated that predatory pricing was always irrational.

24 I think that's unfair to Professor McGee. That is not

25 my exegesis of the article.

Page 56

1 I think the position of McGee and the Chicago

2 School generally is that predatory pricing can be a

3 rational business strategy, it's just it's a very

4 unusual one, defined where it's successful, where it

5 works.

6 DR. ORDOVER: Well, I certainly agree with the

7 statement, with a couple of -- I don't know how many

8 caveats, but first -- five caveats -- one, two, three -ï½­

9 the first caveat is we have got to define what predation

10 means. Second, we have to figure out what the price is.

11 Third, we have got to figure out how to engage

12 rationality. Other than that, I think it's all fine.

13 Other than that, how was the performance, right? So,

14 this is exactly the way I see it.

15 I mean, this is surely a statement that has a

16 meaning as long as we can agree on the meaning of the

17 terms or words that go into the statement. None of

18 these things are relatively or clearly defined. We

19 already have different standards for predation. In the

20 airline case that Ken and I are in, pricing to whom is

21 an issue that -- average price on the aircraft? Is it

22 the price to the business passengers, the leisure

23 passengers? A huge amount of disagreement. Is it the

24 price of the incremental unit? Is the price averaged

25 out over the volume that is being sold?

Page 57

1 Anyway, what's rational? I guess

2 profit-maximizing, over what horizon, what discount

3 rates we are going to use? All of these things enter

4 into what we have been struggling with, which is to say

5 that we have something -- we understand a basic core set

6 of issues, but I think these remaining areas of

7 disagreement are really needed to breathe light -ï½­

8 MR. POTTER: And with the later slides, we will

9 get into those specific areas.

10 DR. ORDOVER: I haven't looked at them. Ex ante

11 assessment, huh?

12 MR. MELAMED: Just a comment provoked by what

13 Janusz said, there is always -- at least in my

14 experience in cases I have dealt with, I am not involved

15 in Spirit -- difficult questions about what are the

16 products you're talking about, what prices are you

17 talking about, is it the leisure passengers or whoever

18 it may be. There is, if there is discipline in the

19 overall case, however, some discipline on the parties on

20 that issue, if the plaintiff wants to argue that the

21 price is predatory because he found one passenger in

22 seat 14B where the price was below cost, he is probably

23 not going to be able to prove that he was driven out of

24 the market on account of the predatory price, and so if

25 the courts are rigorous in connecting the allegedly

Page 58

1 predatory activity with the requirement of proving a

2 causal connection with the creation or maintenance of

3 market power, some of the sort intellectual concerns

4 that Janusz has may become less practically important.

5 DR. ORDOVER: Actually, it was 15C that's at

6 issue.

7 MR. POTTER: Patrick?

8 DR. BOLTON: So, you know, I may well have read

9 too much into McGee's article. Having said that, I

10 think it does -- the legacy that's left is tremendous

11 skepticism, and what I wanted to say was that there has

12 been new scholarship started in the 1980s, rigorous

13 economic scholarship based on rigorous game theory

14 analysis showing exactly how predatory pricing strategy

15 could be rational, and I think what I want to say is

16 that where things have changed is that slowly, this

17 literature is being brought in, is being acknowledged,

18 and is being recognized, and so what I wanted to say is

19 that, if anything, today, we should be less skeptical

20 about the rationale for predatory pricing than we have

21 been and that the Supreme Court has been in its Brooke

22 decision and its Matsushita decision, which was based on

23 older writing which couldn't be articulated using the

24 tools of the modern game theory.

25 MR. POTTER: Okay, subject to Janusz's caveats,

Page 59

1 I will take that as agreement among the panel.

2 The second proposition, this is a quote from the

3 Supreme Court in 1986, two decades old now. "Predatory

4 pricing is 'rarely tried, and even more rarely

5 successful,'" was repeated in Brooke Group in '93. Does

6 the panel think that this is still a correct statement?

7 Doug?

8 MR. MELAMED: Well, I don't know. I will leave

9 it to the economists. The question is whether it means

10 anything. You know, murder may be rare, too, in some

11 statistical sense.

12 But I wanted to say something about that,

13 because I think in my own thinking, at least, until

14 yesterday when I was preparing for this, there was some

15 sloppiness, and maybe that's true of others, as well.

16 In Matsushita, interestingly, when I looked at it, that

17 was when the proposition was first set, it was used as a

18 factual proposition to aid the Court's assessment of the

19 evidence and to say is the predatory theory here

20 sufficiently plausible that we should let it go to the

21 jury?

22 It morphed into something else by the time of

23 Brooke Group. It morphed into the rationale for

24 defining predation a particular way. If it's used that

25 way, we have to be very careful about what we mean. If

Page 60

1 we mean pricing below cost is rarely tried and even more

2 rarely successful, it's rationally then used in

3 Matsushita, but it doesn't support Brooke Group, because

4 that would be like saying killing with an ice pick is

5 very rare, so let's define murder as consisting solely

6 of killing with an ice pick.

7 The question, if you want to justify or explore

8 the wisdom of defining predatory pricing as pricing

9 below cost, the question is, what about the conduct that

10 isn't deemed to be predatory pricing by that definition,

11 some kind of profit sacrifice at above cost levels, is

12 that rarely tried and rarely successful? And I'm just

13 not sure that there has been rigor in thinking about

14 what this statement means.

15 MR. POTTER: Building off of this slide, does

16 anybody have a view on whether predatory pricing is more

17 or less likely in certain industries because of the

18 characteristics of those industries?

19 DR. ELZINGA: Yes, I certainly do. I have a

20 belief that predatory pricing is more likely to occur

21 where the target firm will exit quickly and be unlike -ï½­

22 either the target firm or other capacity will be

23 unlikely to enter again, and just picking up on

24 something that Doug said, where you are trying to look

25 for some more simple benchmark, he suggested just

Page 61

1 focusing on where avoidable costs exceed the revenues of

2 the practice, well, that's a very helpful way of

3 thinking about predation.

4 I think it's just as powerful, maybe even more

5 illuminating, to focus on entry and exit conditions as a

6 kind of filter, and I am a little surprised that Doug

7 never mentioned focusing on exit and entry. That is

8 kind of the mirror image of what he is getting at, but I

9 think it is clearer and analytically more robust.

10 MR. POTTER: Janusz?

11 DR. ORDOVER: Well, I think certainly by the

12 basic principle of self-selection, you at least observe

13 an attempt to induce an exit in the industry in which

14 exit is likely to be relatively quick or not too

15 costly -- it will be not too costly to engage in such a

16 strategy and in which, as Bobby and I said, re-entry is

17 very difficult entry or re-entry is very difficult. If

18 re-entry is trivial, as it generally could be in the

19 airline industry, setting aside the question of

20 signaling predation, setting aside gate constraints and

21 those kind of things -- which were not present in

22 Detroit, just by the way.

23 I think that obviously nobody in his right mind

24 is going to try to exit somebody who has invested

25 hundreds of millions of dollars of sunk capital that is

Page 62

1 simply impossible to take out, but you can try very

2 aggressively and actively to prevent that person from

3 putting in another hundred million dollars of to-be-sunk

4 capital. So, you can try to accomplish something

5 different, but actually self-selection and rational

6 business behavior that we have all accepted as a premise

7 of what firms do, such as that you are not going to try

8 it when it is not likely to be successful, which is why

9 when we get to the recoupment phase of this whole thing,

10 we will probably have different views from what the

11 slides will ask us to say, but it is all part and parcel

12 of the same aspect of the analysis, which is to say, you

13 have to look at the entry and re-entry barriers and the

14 exit barriers or problems with trying to dislodge the

15 rival or problems with the ability to increase the entry

16 or impediment facing the incumbent. If you cannot

17 accomplish entry-enhancing creation of a barrier, then

18 you are not likely to go after that, because somebody is

19 going to come back sooner or later. How soon is

20 unpredictable.

21 DR. BOLTON: I have very little to add, just two

22 remarks. There used to be a time when economists

23 characterized the airline industry as a contestable

24 market. I just want to remark that we have come a long

25 way from that conclusion. Now we are I think defining

Page 63

1 the airline industry as particularly prone to predatory

2 pricing.

3 And on the rarely tried and even more rarely

4 successful, I want to be even more outrageous by saying

5 that, you know, nuclear bombs have been rarely tried,

6 but they have been very successful. We have to look at

7 the deterrent effect of episodic, very rare predatory

8 pricing. So, you know, you look back at predatory

9 pricing in the telecom industry at the beginning of the

10 century or in the tobacco industry, it was followed by

11 prolonged periods of lack of entry and oligopolistic

12 pricing with very high returns to the firms, which is

13 evidence that consumers were not getting the low prices

14 that they deserved.

15 MR. POTTER: Proposition three, because lower

16 prices immediately benefit consumers, we should be

17 extremely careful not to adopt legal rules that can

18 result in false positives; that is, condemn legitimate

19 price-cutting. This seems to be a fundamental basis of

20 Brooke Group, at least. Anybody have any agreement or

21 disagreement with this? First say agreement.

22 DR. ELZINGA: Agreement certainly for me.

23 DR. BOLTON: I beg to disagree on the following

24 grounds, not in principle, but on the basis of the

25 evidence. How concerned should we be about false

Page 64

1 positives today after a quarter century of systematic

2 rejection of predatory pricing allegations? How worried

3 should we be today that firms will be very cautious in

4 their pricing and will refrain from aggressive pricing

5 after this record?

6 I think in principle, we should be worried about

7 this, but I am not sure that with the past history of

8 predatory pricing enforcement that this is still a major

9 concern.

10 MR. POTTER: Ken, I think you wanted to comment.

11 DR. ELZINGA: Yes, let me comment at two

12 different levels.

13 First of all, there is no doubt, since

14 Matsushita, that the economists have taught us things

15 that we did not know at the time about models in which

16 predatory pricing can be successful for the predator

17 under conditions of certain financial asymmetries or

18 information asymmetries or information effects, but if

19 you look at some of the cases, the most recent, I think,

20 or if I'm mistaken, the most recent predatory pricing

21 case brought by the FTC, a long time ago, was the coffee

22 case, General Foods Coffee case, and the staff was

23 unsuccessful on that.

24 When we look at the record, did Maxwell House,

25 which had all the things that would fit nicely into this

Page 65

1 model of reputation effects, signaling and so on, where

2 you might think, boy, this looks like predatory pricing,

3 the way the game theorists would structure the world,

4 and people like Milgrom and Roberts have referred to

5 that case as illustrative of applying their models to

6 predatory pricing.

7 Well, that was a case in which Maxwell House was

8 trying to keep Folgers from moving east. They were

9 singularly unsuccessful. Folgers rolled out nationally,

10 and if you walk around a bit, you just don't see Maxwell

11 House of having visual evidence of being a monopolist in

12 the coffee industry today. You are much more apt to see

13 Starbucks than Maxwell House.

14 Matsushita, you think about the signaling

15 effects or the reputation effects that the Japanese had

16 and the popular culture at that time about being

17 price-aggressive. You look at the television industry

18 today -- now remember, this is a case the Japanese

19 won -- they have less than 40 percent of the television

20 business, total, all the companies combined. The

21 largest television producer in the world is in China.

22 Brooke Group, the idea there was the majors, led

23 by Brown & Williamson, would dial down the discount

24 segment. That was a term used over and over again in

25 Brooke Group. The discount segment would be dialed

Page 66

1 down. Everybody would be left buying a full revenue

2 cigarette if they were a smoker.

3 The discount segment continues to grow. It's

4 about 40 percent of the industry now. So, if all of

5 these cases had been decided differently using game

6 theoretic approach or a concern that Patrick expresses,

7 I think consumers would be worse off. I really do.

8 MR. POTTER: Janusz?

9 DR. ORDOVER: One comment. I think that there

10 is an issue that we may want to talk about a little bit

11 more, and that is to say, the rigor and the reviews of

12 the galaxy of predation models that are based on really

13 state-of-the-art game theory, and the question, what

14 follows from those in terms of public policy? To me,

15 that is the biggest problem that I have been totally

16 incapable of resolving in my own head, but in the end,

17 coming down on the proposition that while we cannot be

18 as perhaps lackadaisical about anticompetitive

19 exclusionary behavior as the Court in this famous quote

20 was, we still need to take some kind of tools that the

21 courts can use to say, yes, yes, I agree, things can

22 happen, and Milgrom and Roberts and Kreps-Wilson, they

23 all have shown all those things, and many others follow

24 and, you know, your lovely paper with Scharfstein on -ï½­

25 what, signal jamming or -- it was, signal jamming paper,

Page 67

1 which was the coffee case, and all those things are all

2 true.

3 And then we come back to the question, what to

4 do with that, how to translate it into something that a

5 businessperson, who has to be counseled, will be able to

6 understand in day-to-day operations, and how will the

7 Court be able to take these principles of game theory,

8 subgame perfect, Nash equilibria and all these things,

9 and translate it into some simple rules that, you know,

10 thou shall not do what? Thou shall not signal that you

11 are going to be a tough guy? You can't say that. You

12 have to be able to translate it into something. "Look,

13 you can write any memos you want, you can do anything

14 you want, but you cannot do X."

15 I think that it is absolutely essential that we

16 take these models and we translate them into principles

17 that are implementable by the business people, by the

18 lawyers and by the courts. Otherwise, we are nowhere,

19 and I think what we have been struggling with is trying

20 to come to articulation of some principles that are

21 actually understandable, and I think Doug went a long

22 way in proposing that we actually take the learning of

23 these models as implying we should not dismiss these

24 cases, but we should take the learning of these models

25 and figure out what they mean in terms of implementable

Page 68

1 rules by all the stakeholders, and that includes, of

2 course, consumers as well.

3 MR. POTTER: Doug, in your dealings with your

4 clients, without a rule that is under-inclusive by

5 protecting against false positives, is it your belief

6 that monopolists wouldn't price close to the line?

7 MR. MELAMED: Ah, I'm not sure I understood the

8 question. I think you are asking, should we worry about

9 over-deterrence?

10 MR. POTTER: Well, if we don't protect against

11 false positives, will the chilling effect of getting too

12 close to the line lead people with monopoly power not to

13 lower their prices to consumers because they're worried

14 about false liability?

15 MR. MELAMED: Sometimes. I do not know whether

16 the overall economy, with the relative magnitude, what

17 its effects are. I particularly agree with how Janusz

18 started. The signals you send to the business community

19 are much more important frankly than whether the cases

20 are right or wrong. If every case at the margin were

21 wrongly decided but we were generally setting a useful

22 set of standards, the law would be pretty good. So, the

23 question is the false positives versus the false

24 negatives.

25 Generally speaking, with the state of the law

Page 69

1 today, you have a slide later on, is it hard to counsel

2 your client? No, I say not to worry about it, because

3 the -- but actually -- actually -ï½­

4 DR. ORDOVER: Can we go home now?

5 MR. MELAMED: But actually, I say more than

6 that. First I say you may lose the characterization

7 issue, you may not be able to prove predatory pricing,

8 but then I say, "Wait a minute, there are certain

9 settings in which you could get hurt. Is your target

10 likely not only to withdraw from this market but, for

11 example, to go out of business and become bankrupt and

12 his only asset may be a lawsuit? How litigious is he?

13 Is this a part of some broader commercial strategy?"

14 So, there are situations I think even with the

15 law today totally in favor of the weight of false

16 positives where it probably does deter some

17 procompetitive pricing. Whether on balance at this rule

18 or at some other rule that harm is greater than the harm

19 of false negatives I'll leave to the economists.

20 MR. POTTER: All right.

21 Next one, establishing a reasonable prospect of

22 recoupment should be essential in any analysis of

23 predatory pricing. Is there anyone who disagrees with

24 this statement?

25 MR. MELAMED: Only to the extent I already said

Page 70

1 so.

2 MR. POTTER: Janusz?

3 DR. ORDOVER: Oh, I think the point I want to

4 make is that from my perspective, this recoupment

5 component is really part and parcel of a prior filter.

6 Now, you can try to do it at the later stage. My

7 preference is to ask the question whether the particular

8 markets, market or markets, in which this

9 anticompetitive conduct is alleged to be exclusionary,

10 anti-consumer, whatever characterizations you want to

11 attach, is acceptable to incremental exercise of market

12 power, and if the answer is no because, you know, you

13 get rid of this particular rival, but, you know, quick

14 as a bunny, somebody else is going to show up who may be

15 even more competitively advantaged rival, then there is

16 no need to somehow construct this potentially

17 complicated analytics.

18 As is clear from Ken Elzinga's net present value

19 calculation, it is a very, very difficult step, possibly

20 as difficult as the step of measuring revenues to costs,

21 which costs which revenues and so on and so forth. So,

22 I would say that as a filter, you certainly would want

23 to implement a step during which the parties will slug

24 it out, one saying, "Look, I get rid of you, there is

25 ten more coming. I get rid of you, that will carry no

Page 71

1 visible signal for the rest of the players that may be

2 sitting out on the outskirts and waiting what to do."

3 Or it could be that the firm which you are

4 trying to induce to exit or to restrain its expansion is

5 what I called in the first slide a scarce competitor,

6 and, in fact, there is something very special, very

7 particular about that rival which cannot be replicated,

8 and in that case, yes, you get to the point in which the

9 assessment of this later recoupment or the implications

10 of this strategy is critical, and if you cannot show -ï½­

11 you, the plaintiff -- that if you exit the marketplace

12 or if you get cut back in the marketplace, economic

13 welfare is going to be hurt in some way, then I think

14 you have gone very far in challenging the conduct at

15 issue.

16 MR. POTTER: Next slide -- oh, I'm sorry.

17 DR. ELZINGA: I was just going to say, I think I

18 am saying just the same thing that Janusz said but

19 perhaps in just a couple words. I do not think you need

20 to do a recoupment analysis for many predation

21 allegations, because entry conditions or prices and

22 costs will tell you you needn't take that extra step.

23 DR. BOLTON: Can I just add -ï½­

24 MR. POTTER: Sure, go ahead.

25 DR. BOLTON: -- one comment? So, I agree with

Page 72

1 Janusz that in principle, recoupment is important, it is

2 the right question to ask, but in terms of how do you

3 administer a recoupment test, I think the weight has to

4 be on what you call the reasonable prospect, and I think

5 a narrow reading of a recoupment test, as you criticized

6 earlier, I would criticize as well.

7 MR. POTTER: Okay, fair enough.

8 Next slide. Prices above some measure of cost,

9 and you can all pick your own measure of cost that you

10 think is the best cost, whatever it is, should not be

11 considered predatory. Is there anyone who disagrees

12 with this?

13 Patrick, do you want to say anything?

14 DR. BOLTON: So, from the -- well, we know that

15 a policy of -- after Brooke Group is that a price -- at

16 least a price above average total cost should not be

17 considered as predatory. I am happy to live with that,

18 although I am not sure that it is always a wise policy.

19 MR. POTTER: When it is -ï½­

20 DR. BOLTON: I would disagree, though, with the

21 statement that prices above average variable cost should

22 not be considered as predatory.

23 MR. POTTER: You just mentioned that you might

24 disagree in certain instances that even prices above

25 average total cost should not be predatory.

Page 73

1 DR. BOLTON: Could be predatory. In principle,

2 in theory, there are situations where prices above -ï½­

3 even price above total cost can be predatory.

4 MR. POTTER: Can you give an example of those?

5 DR. BOLTON: Well, an example of a large

6 incumbent with increasing returns, scaled technology,

7 facing a small entrant that has not been able to reach

8 minimum cost capacity, you could exclude that entrant by

9 pricing lower than monopoly price but still above your

10 average total cost and exclude the entrant.

11 DR. ORDOVER: Maybe I could just ask you a

12 question. Would you comment on the cost principles -- I

13 have been puzzled by them myself -- that follow from

14 these various game theory like models of, say,

15 Kreps-Wilson? They do not seem to give clear cost

16 benchmarks. Is that true? Is that your reading as

17 well?

18 DR. BOLTON: Yes, that is correct. They do not

19 give a clear reading on cost benchmarks, and I think

20 there is a whole group of economists who have been

21 working on predatory pricing who think that costs are a

22 very poor way of discriminating between anticompetitive

23 effects and procompetitive effects, that there are as

24 likely to be false positives as there are to be false

25 negatives. There are many situations where pricing is

Page 74

1 below even average variable cost, and it is efficient.

2 It is not predatory. So, a lot of economists feel it is

3 just a poor test.

4 MR. POTTER: Doug?

5 MR. MELAMED: Let me ask Patrick this question.

6 I understand the theory, even if I cannot understand the

7 game theory, of why an above cost, even above total

8 cost, but below profit -- monopoly profit-maximizing

9 test could be predatory in the sense that it could

10 exclude a rival and in the long run we are all going to

11 be worse off for it.

12 What I don't understand and I am interested in

13 your reaction to is how one turns that into a legal rule

14 that companies can comply with. I mean, how do you -ï½­

15 you know, if -- sure, if you imagine -- if you posit a

16 stable market on day one and then the entrant comes and

17 maybe you have a good historic benchmark and you can

18 say, "Gee, he's changed his pricing," but even then you

19 have to ask the question, "Well, what would the monopoly

20 profit maximizing price be with the new entrant?" Is

21 each company supposed to hire a game theorist and work

22 out the game and figure out -ï½­

23 DR. ORDOVER: Yes.

24 MR. MELAMED: -- what the price is? In other

25 words, how do we implement that test?

Page 75

1 MR. POTTER: Patrick is looking for future

2 employment.

3 DR. ORDOVER: We all are.

4 DR. BOLTON: Administerability is a serious

5 concern, I take that. I'm happy with a rule that

6 says -- I would not object to a rule that says price

7 above average total cost is per se legal as a way of

8 implementing an easily administrable rule.

9 As for determining whether it is procompetitive

10 or anticompetitive conduct, I think there -- while

11 business decisions are taken on average in a rational

12 way, and you have to get justifications for the kind of

13 policy you are implementing, these justifications of ten

14 find their form in written documents in the company,

15 whether it is emails or other board room records, and as

16 I emphasized in my presentation, this is evidence of

17 intent, which is extremely valuable. Intent here, that

18 kind of intent evidence, is a very good guide to the

19 kind of effects a policy can have, and there, I think we

20 can be on pretty firm ground, and we do not have to

21 do -- we do not have to hire a game theorist to do that

22 kind of analysis.

23 MR. MELAMED: Sometimes companies adapt to the

24 law, and if they are well counseled, they know how to

25 write pieces of paper that perhaps articulate an

Page 76

1 economic rationale rather than intent.

2 DR. ELZINGA: And in like fashion, Doug, I

3 suspect you have encountered clients who just aren't

4 aware that when they write things, they have to be

5 written with an eye towards antitrust enforcement, and

6 so you do find documents coming -- to have militaristic

7 or powerful metaphors that have nothing to do with

8 consumer welfare and may, in fact, represent exaggerated

9 views of the company's prowess and stature in the

10 marketplace.

11 MR. MELAMED: I find almost invariably they are

12 written by lunatic middle managers, but -ï½­

13 MR. POTTER: That wasn't your position when you

14 were Deputy Assistant Attorney General.

15 MR. MELAMED: Well, you know, you learn.

16 DR. ORDOVER: Or those documents are usually by

17 investment bankers who are pedaling a particular deal or

18 something like that, alleging that as a result of action

19 X or Y, the firm would be able to leverage its market

20 power from one market to another. If one were to take

21 these arguments -- take these documents seriously, that

22 would be the end of most of the Chicago Business School,

23 presumably, investment banking, but also, the ability of

24 business people to compete in the marketplace, because

25 this is what these guys are selling. So, you have to

Page 77

1 read their signal, which is the investment advice or

2 business advice which they proffer, as being an attempt

3 to market the product at above competitive price to the,

4 you know, willing or unwilling buyers, and I think that

5 that is why I am very worried about reading all kinds -ï½­

6 I mean, I have seen documents probably as good or as bad

7 as Doug or anyone, and I try to discount their value

8 because they are frequently misleading.

9 Now, this is not to say that people who run the

10 companies do not have an insight into the marketplaces

11 in which they are competing, but I think there are

12 limits of the kind of inferences you can really make

13 from those types of documents, especially when they are

14 also written by third parties with a very special agenda

15 of their own in my view.

16 MR. POTTER: Next slide, as long as we're on

17 costs, let's throw this out, a variety of cost tests.

18 The proposition for the panel is, average avoidable

19 cost, which for definition, cost per incremental unit

20 that does not have to be paid if the incremental units

21 are not produced, is the best cost measure to use if

22 forced to use the Brooke Group analysis. Is there any

23 disagreement with that?

24 Doug?

25 MR. MELAMED: Yes, I actually disagree with it

Page 78

1 as phrased. I think it would be more useful to use

2 avoidable cost compared -- and then add up the revenues,

3 because when you say average avoidable cost compared to

4 price, you are limiting yourself to the revenues from

5 the product in question, and you can't take account by

6 that formula, at least, of two-sided markets and

7 everything else.

8 DR. ORDOVER: And networks. I think the issue,

9 just to pitch it to the folks if you want to raise it or

10 discuss it further, I think in the American Airlines

11 case, there was a lot of debate as to what the right

12 benchmark of cost was in my view, and at least the

13 Government had I think proposed four, if I am not

14 mistaken -ï½­

15 MR. POTTER: Correct.

16 DR. ORDOVER: -- that may be three too many,

17 but, you know, we have offered at least one or two

18 ourselves.

19 The same thing in Spirit, I think we have come

20 up with two different measures of cost, but they would

21 apply to different types of outputs, all passengers

22 rather than leisure versus business. So, there is a lot

23 of wiggle room as to what it is that this cost measure

24 is going to be applied to, and as Doug pointed out, it

25 is also key to figure out what is the measure of

Page 79

1 revenues against which these costs are to be assessed.

2 So, I think that this is not a bad -- again,

3 this is not a bad standard, but I think it is important

4 to understand what is avoidability here that is at

5 stake.

6 For example, in the American Airlines case, we

7 thought, at least I thought quite strongly, that the

8 right set of costs would be those that if -- the airline

9 would avoid if it were to exit or substantially cut back

10 on a particular route, and that actually includes a lot

11 of costs that would be avoided, because it would include

12 avoiding the aircraft costs which were at issue, those

13 would be significant, much more than many other costs,

14 and it could be cutting back at the hub, perhaps,

15 cutting back at the stations from which the airline

16 would exit.

17 So, these avoidable costs which we looked at at

18 the route level are typically the kind of costs business

19 people look at when they make business decisions in the

20 airline business, and I thought there was a good measure

21 of avoidable cost. It happens so that the increment of

22 output over which we are looking at was that of the

23 route as opposed to -- it could be a seat or it could be

24 an aircraft or it could be a flight or something or an

25 aircraft day, because these aircraft, they fly in

Page 80

1 strange ways around the globe, but there is all these

2 things that can be taken into account that could confuse

3 or could illuminate the matter.

4 MR. POTTER: I wanted to follow up on this slide

5 for a couple other questions. One, Doug, I think you

6 talked about this in your presentation, but I was

7 wondering if anybody had a view of whether opportunity

8 costs should be considered in viewing the cost test.


10 MR. POTTER: Doug, I thought you were a yes,

11 Doug. Does anybody else on the panel have any views on

12 that?

13 DR. ORDOVER: Well, as I said, I think in

14 American Airlines, we had an internal discussion of

15 what's the meaning -- what to do about the aircraft. I

16 mean, there is no denial that American Airlines brought

17 in additional flights. You could say, well, is there an

18 opportunity cost of that aircraft, and if there is, how

19 are you going to measure it? And one measure, which I

20 thought was the most easy to implement, would be to look

21 at the lease rates as opposed to trying to understand

22 what is another route that this aircraft could have been

23 flying or was this aircraft sitting somewhere in the

24 desert in Arizona and doing nothing? Maybe American has

25 a lot of aircraft like that, they could fly them at very

Page 81

1 low incremental capital cost, but I thought that the

2 most reasonable assumption would be to assume that the

3 airline uses its aircraft properly and it could actually

4 deploy an aircraft by leasing a new one, which is why an

5 18-month time horizon I thought would reflect the

6 leasing strategies and the fact that heavy and costly

7 equipment was deployed, and there is no way of avoiding

8 counting the assets that are not being used, which

9 aircraft were one of them, and there are others that I

10 thought were appropriately included as well. That was

11 not the view necessarily of all of my colleagues on that

12 case.

13 MR. POTTER: Ken?

14 DR. ELZINGA: I do not think you could trust any

15 economist who would say opportunity costs should not be

16 considered. I mean, opportunity cost is the main

17 analytical construct that we bring to the social

18 sciences, that the cost of something is the highest

19 valued opportunity forgone, but the problem is that in

20 antitrust, opportunity cost is a Promethean expression,

21 and it is very difficult to unpack it, and one of the

22 sobering things for me, who has worked in this area for

23 a while and tried to think about it, is how fragile some

24 of the cases are, some of them that I have been involved

25 in, some of them that I have studied, to the taxonomy of

Page 82

1 cost.

2 The case that Janusz and I were involved in,

3 what do you mean by the price of the product? Is it to

4 all passengers, or is it to a group of passengers who

5 are called leisure and price-sensitive and business and

6 insensitive to price? The case can pivot upon that

7 taxonomy.

8 In Brooke Group, after hundreds of hours of

9 thought, one of the things that distressed me is that

10 price above average variable cost pivoted upon -- it

11 could pivot upon how you counted layers of tobacco.

12 Tobacco is stored for years to age, and if you used a

13 LIFO method, it looks like it violated Areeda-Turner.

14 If you used LOFI, it looked like it was okay, and I

15 don't like living in a world where it pivots upon what

16 accounting standard you use.

17 One of the things economists supposedly also

18 bring to the table is to get people out of using

19 accounting data and to think in terms of opportunity

20 costs, but even trying to apply that standard is

21 problematic, and when you live in a world where there's

22 a predatory pricing allegation, you probably are in a

23 world where prices are close to costs, by whatever

24 measure, and so then you start to figure out, well, how

25 close and above or below, you inherently, if you are

Page 83

1 going to use a cost-based standard, have to deal with

2 accounting data, but you try to always put it through

3 the filter of opportunity costs.

4 DR. ORDOVER: Just as an anecdote, in the

5 American Airlines case, I lived through like three days

6 of deposition and a number of questions related to the

7 question of how we treated these carts that people put

8 their luggage on when they pick it up at the station

9 that was going to be exited potentially, so I finally in

10 desperation said, "Look, if the whole goddamn case turns

11 on how we treat these carts, then the Government

12 shouldn't be bringing a case like that." There has to

13 be something more to it than that, right?

14 And I think somewhere -- I mean, tobacco is

15 probably more important to cigarettes than the carts are

16 to the airline, but again, this is really demonstrating

17 very well the kind of deep-level accounting issues and

18 cost treatment of issues that can go whichever way

19 somebody wants them to go, and therefore, to say that

20 something is average avoidable cost is, again, the same

21 thing. It is average, it is avoidable, it is a cost,

22 but other than that...

23 MR. POTTER: If you are going to require

24 opportunity costs to be considered, how does that differ

25 from requiring the defendant to maximize his profit?

Page 84

1 Because one view of opportunity costs is he had an

2 opportunity to get more profit. How do you distinguish

3 the two, if at all?

4 Patrick?

5 DR. BOLTON: Yes, so, this is where Doug drew a

6 very clear line that you should not count lost revenues

7 and inframarginal sales as an opportunity cost. I would

8 say as a matter of theory, you should count that as an

9 opportunity cost. The real question is just one of

10 administerability.

11 The other point I would make in this respect is

12 that -ï½­

13 DR. ELZINGA: I am sorry, could you just explain

14 for the benefit of at least me why, why you would count

15 that, the inframarginal?

16 DR. BOLTON: Well, because the question we are

17 trying to answer is, what is driving the price

18 reduction? Is this a move by the incumbent that will

19 raise profits irrespective of its anticompetitive

20 effects or not? To be able to answer that question, we

21 need to understand the nature of the profit sacrifice,

22 the size of the profit sacrifice and what justifies it.

23 Is there an efficiency rationale or is there an

24 anticompetitive rationale? So, we cannot avoid it, and

25 that is what I was going to say with respect to

Page 85

1 recoupment.

2 Recoupment is the right question to ask. When

3 you try and make a recoupment analysis, you are

4 comparing profit sacrifice and return on investment.

5 So, it is inescapable, you have to look at lost revenue

6 and inframarginal sales when you do your recoupment

7 analysis.

8 Now, if you do it for your recoupment analysis,

9 I don't see why we should not take that into account for

10 the cost test, but I will leave that for -- I think I

11 see that as a practical problem and not a conceptual

12 problem.

13 MR. POTTER: Doug, do you want any extra time or

14 not?

15 MR. MELAMED: Well, the woman who is

16 transcribing this said I spoke so fast that maybe my

17 words were not caught before, but I will assume she got

18 them.

19 MR. POTTER: Okay. What about, how do we

20 distinguish in situations the appropriate costs when -ï½­

21 essentially price discrimination or nonlinear pricing?

22 Does anybody have any views on that?

23 Doug? No, that is not on the slide.

24 MR. MELAMED: What are the avoidable costs?

25 MR. POTTER: Fine, let's go to the next slide.

Page 86

1 Out-of-market reputation effects are so hard to

2 assess they should not be considered in an analysis, and

3 let me just give you another minute on this. I am

4 thinking of a scenario whereby -- let's take the airline

5 industry. An airline allegedly predates on, let's say,

6 Dallas-Wichita. The airline has a bunch of other

7 markets where it's in, and perhaps it's got monopoly

8 power in a number of those markets. Maybe it doesn't

9 know, maybe we don't know, who potential entrants will

10 be in those other markets, but when the management team

11 sits down to determine what they are going to do in

12 Wichita-Dallas, they sit there and say, "Well, let's

13 drive the guy out, because future people then won't

14 challenge us in our other markets." How does that, if

15 at all, get analyzed in determining recoupment?

16 Patrick?

17 DR. BOLTON: Yes. So, we elaborate on this

18 point in our article. So, reputation effects do come

19 into a recoupment analysis to the extent that reputation

20 effects may raise the barriers to re-entry into the

21 market, and they do raise them in the form of making

22 other competitors aware that should they enter this

23 market, the first thing they will be facing is a tough

24 price war.

25 How do you prove reputation effects? That is

Page 87

1 the harder question. But I do not think that that is

2 necessarily insurmountable. Again, there can be

3 analysis by the incumbent suggesting that this is a

4 profitable strategy, that reputation effects work, that

5 if we drive out this rival in this market, we will

6 benefit because there will not be other rivals or we

7 will be slowing down the growth of this rival.

8 So, this analysis is recognized by the

9 incumbent. If you find circumstantial evidence

10 suggesting that financiers think in those terms, that

11 they will raise the cost of funding of a new entrant

12 because they recognize that the market that the new

13 entrant is about to enter is one where there have been

14 past price war episodes, then I think this is all

15 evidence that this is a problem.

16 MR. POTTER: Ken?

17 DR. ELZINGA: This is a very strange statement,

18 and it could be easily misunderstood, but we have to

19 remember that the case for new entry and the enthusiasm

20 that we often have in antitrust for new entry can be

21 exaggerated. New entrants can inefficiently use

22 society's scarce resources. There are lots of

23 businesses that have no business entering an industry

24 because they use the resources inefficiently, and one of

25 the good things that keeps inefficient entrants out is

Page 88

1 the reputation of incumbent firms for being tough,

2 aggressive, low-cost competitors.

3 My concern -- and again, so much of the

4 difference, perhaps, between Patrick and myself is one

5 of administerability -- is once you start bringing in

6 reputation effects as a potential hammer for antitrust

7 plaintiffs, what is the consequence of that for all the

8 good things that reputations do of incumbent firms to

9 keep people, even for their own good, out of markets in

10 which they have no business competing because they will

11 not be efficient utilizers of society's scarce resources

12 in those settings?

13 DR. ORDOVER: I think that the reputation

14 effects are almost a cornerstone of the new game

15 theoretic model of anticompetitive behavior of the sort

16 that Patrick summarized in his talk and his paper, and

17 there is no question that firms act in ways that try to

18 convey signals to the outside world and to the inside

19 world, and I think I would agree with Ken, not going so

20 far as we should not find entrants to be all such great

21 participants -- I like entrants. I think they should -ï½­

22 you know, let them slug it out.

23 Let's not create a presumption that some of them

24 may be inefficient ones or some of them are efficient

25 ones. Who the hell knows? It is the crucible of the

Page 89

1 marketplace that ultimately will determine that.

2 But my issue is, again, I guess where we agree

3 is administerability, and then to say, yes, indeed, it

4 is plausible to postulate the reputation effects. We

5 have the economic models. What we don't know in real

6 life is how many of these new entrants do you have to

7 kill in the airline business before somebody finally

8 realizes, hey, I'm not coming in, and empirical work

9 shows that no matter how many of them you squish, they

10 always come back, and so you say, am I still in the

11 reputation-building way or am I in the recoupment phase

12 or how am I going to account for that other than to say,

13 look, you go ahead and do what you want to do, compete

14 as hard as you want, but you should not break the

15 following simple rules, whatever they might be, because

16 I cannot account for all the other additional

17 considerations.

18 However, I think it is appropriate to say that

19 these reputational effects that we are encountering in

20 economic theoretical literature, but also in some

21 empirical stuff. In fact, there is some beautiful work

22 by Canadians on the supermarkets in Canada, indeed,

23 indicating that some reputational effects that have been

24 established. I just don't see how I can translate that

25 into an administrable test for the courts and for

Page 90

1 counsel, because what can Doug say to somebody who says,

2 "Look, I think I want to kill three guys because I think

3 that will be enough," and he goes, "No, only kill two."

4 I mean, what? What do you say? I just don't know how

5 to do it. I wouldn't know how to do it.

6 MR. MELAMED: Look, if the client comes to me

7 and says, in effect, I want to cut my prices to below my

8 avoidable costs, I might say, "Why are you doing that?

9 You are going to run the risk of losing an antitrust

10 case." And if he says to me, "I'll beat the rap because

11 they will never hang me with the recoupment thing,"

12 because if my recoupment is going to be my reputation, I

13 might say, "That ought to be illegal."

14 That is to say, as Patrick says, if you can

15 prove a plausible recoupment story, a reputational

16 story, that, in fact, you are gaining market power

17 because you are gaining reputation and it is not just

18 the lawyers -- the plaintiff's lawyer's fantasy, then I

19 don't know why that's not enough to satisfy the market

20 power screen. In the Microsoft case, for example, which

21 I believe was rightly decided, the proof of competitive

22 effects was, you know, rather conjectural, but you had

23 conduct that unequivocally didn't do anybody any good

24 and you had a plausible theory of a market power entry

25 barrier story and the fact that Microsoft believed that.

Page 91

1 Why wouldn't that be enough just because we have price

2 here as opposed to the set of conduct that was at issue

3 in Microsoft?

4 DR. ORDOVER: I think that there are substantial

5 sunk costs of coming in, combined with the signaling, I

6 think you have a plausible story to tell, say, look, you

7 know, you are trying to convince these people that if

8 they come in, there is going to be an aggressive price,

9 and with the substantial sunk costs at issue, that might

10 be something that will take you over the edge, and they

11 say, I'll stay out of the relevant market, but it is the

12 combination of the informational aspects of behavior

13 coupled with the structural features, which is

14 substantial up-front costs, which you require and that

15 the market power screen really -- to say, ah, this

16 market is susceptible to anticompetitive behavior, it's

17 susceptible to recoupment or to price elevation if you

18 protect it.

19 So, this acts as a part of the analytical story

20 that is being told, but again -- and I am perfectly

21 happy to accept it. What I am trying to say, it has to

22 go hand in hand with another aspect of the proof, which

23 is to say that the informational aspects are conjoined

24 with the real exposure that the entrants will face if

25 things go wrong, so that when he comes in, it loses a

Page 92

1 lot of money if it stays, because it cannot exit, and

2 therefore, it is not willing to try. If exit and entry

3 are easy, then I don't believe there is much power to

4 that informational signal.

5 MR. MELAMED: I agree completely with that.

6 MR. POTTER: Next slide. I know at least one of

7 our panelists said in his presentation that he would

8 disagree with this, so, Patrick, I believe that is you,

9 but we will see what the others think. Meeting

10 competition should not be a defense to predatory

11 pricing. Is there anybody who agrees with that?

12 MR. MELAMED: Agrees that it should not be?

13 MR. POTTER: Should not be. Doug, one

14 agreement. Patrick, you disagree. Okay, Doug, why -ï½­

15 DR. ORDOVER: Wait, should not be -ï½­

16 MR. POTTER: Should not be a defense. So, you

17 have essentially a high-cost producer. A lower-cost

18 producer comes in, more efficient, at lower price. The

19 high-cost producer cuts his price, lowers cost to meet

20 competition. Should that be protected or not in a

21 predatory pricing case?

22 DR. BOLTON: So, on the meeting competition

23 defense, if meeting the competition is a best response,

24 then this should be a defense. So, in principle, this

25 is an admissible defense. Administerability, again,

Page 93

1 concerns are important here. For example, what do we

2 mean by meeting the competition? Is matching the price

3 of the entrant meeting the competition? Is that how we

4 define it? I would argue that's dangerous, because the

5 products may not be the same. If the incumbent's

6 product is higher quality than the entrant's, then

7 matching the price of the entrant is not meeting

8 competition. It's -ï½­

9 MR. POTTER: So, a jury is going to decide what

10 the quality-adjusted price is?

11 DR. BOLTON: (Nodding.)

12 MR. POTTER: Doug?

13 MR. MELAMED: I think Patrick and I might not

14 actually disagree but just use different words. He said

15 if this is the best response. If it's the best

16 response, then it would seem to me that the revenues

17 generated by the response are in excess of the avoidable

18 costs, in which case it passes the price-cost test, but

19 if that's not the case, if it fails that test, it's an

20 inefficient response. The fact that he's meeting

21 competition I don't think should make it a safe harbor.

22 MR. POTTER: On a more general basis, it's not

23 one of the slides, but what role should efficiencies

24 play as a defense to predatory pricing? I know,

25 Patrick, you think they should play a central role. Any

Page 94

1 of the other panelists have a view on that?

2 (No response.)

3 MR. POTTER: Seeing none, we will go on to the

4 next slide.

5 It is extremely difficult to craft an effective

6 injunctive remedy in predatory pricing cases, and I'm

7 thinking of the -- I think there was at least one TRO in

8 a case where the judge said for purposes of the TRO, the

9 company couldn't price above the price that it had set

10 on August 1st, you know -ï½­

11 DR. ORDOVER: Of any year?

12 MR. POTTER: Well, it was a particular year, you

13 know, subject to changes in raw material costs. You

14 know, is that the remedy that -ï½­

15 DR. ORDOVER: That was the Baumol Test, right,

16 you can cut the price, but you can't raise it for five

17 years?

18 MR. POTTER: What is the injunctive remedy? I

19 understand what the damage remedy is if it's a private

20 case. What is the effective injunctive remedy? Is

21 there any?

22 Doug, have you given that some thought?

23 MR. MELAMED: Yes, that question I have. I

24 think it is very difficult, very dicey. There may be a

25 circumstance in which it makes sense for a court to

Page 95

1 specify a price in that sense, I can't offhand think of

2 one, but I don't know why it's a terrible thing to

3 simply say, "I declare the conduct to be illegal, and I

4 order you to stop pricing below avoidable costs."

5 DR. ORDOVER: Having first defined it.

6 MR. MELAMED: Right, of course, assuming the law

7 has been decided so that we know what that means,

8 because I think the action in the government case, for

9 example, is to help the law evolve into sensible

10 principles, and then the deterrent effect might be

11 served by the damages rather than having government

12 regulate through injunctions.

13 MR. POTTER: Ken?

14 DR. ELZINGA: Well, probably like everybody on

15 around the table and everybody on the other side of the

16 table, I'm suspicious of having antitrust become a price

17 regulatory regime. It may be that in a genuine

18 predatory pricing case, as the Court has the authority,

19 that you could get at some other part of the structure

20 of the market that allows the predatory pricing to be a

21 viable marketing strategy. Patrick gives the example in

22 his article, which I found persuasive, of the Intel

23 Communications, whether the Court would have the

24 authority to get at the regulatory issue that allows the

25 financial asymmetries and the resource asymmetries to

Page 96

1 make the predation successful.

2 I don't know, and I'm picking this not to pick a

3 dispute with Janusz, but let's say for point of argument

4 that -- I genuinely mean that -- but let's say that gate

5 constraints are the one variable that might make

6 predation successful in the airline industry, and if you

7 can get away from gate restraints, then new entrants

8 could always come in and unravel any successful

9 predation scheme. I would much prefer in the setting

10 like that for a court to say, "Well, instead of trying

11 to monitor and manipulate prices of airline tickets to

12 make sure they're above some measure of cost, that we do

13 away with that particular structural constraint that

14 keeps the new entrant from being viable at such and such

15 an airport because they can't get gates."

16 If that were the case, perhaps that would be the

17 way to get at it. That would be more appealing to me

18 than having the Court monitor prices over time.

19 DR. ORDOVER: I cannot disagree on the gate

20 issues. It has been recognized in the airline business

21 that gates and slots are one of these assets that the

22 contestability literature perhaps forgot about when it

23 was first deployed in the airline business, capital and

24 wings, but I think that is a very sound prescription. I

25 just have, again, one little caveat.

Page 97

1 One of the reasons these gate problems often

2 arise is because the airlines actually could finance a

3 large part of the construction of the airport, and that

4 becomes an issue, who is going to -- unless the airport

5 is a public resource that is not paid for by the

6 airlines or by one airline investing in a -- that's in a

7 part of the airport, if it's actually paying for these

8 gates, then I think it becomes potentially expropriation

9 of what could be a costly investment, and I think we

10 will have to worry about the remedy from the standpoint

11 of investment incentives, in other words, opening up the

12 scarce asset.

13 I don't have to worry about that so much perhaps

14 in the airline industry, but other industries.

15 Obviously Microsoft raised the question and said, "Wait

16 a minute, we are investing -- the remedy is to open up

17 the API. Hey, we are spending a huge amount of money

18 innovating in this space, and now you are telling me to

19 open up." So, this is again, you know, perhaps even

20 more problematic than regulating price, to regulate

21 access. It is an equally complex or even perhaps more

22 complex issue.

23 MR. POTTER: Let's go to the last slide, and we

24 are running a little bit out of time, but I definitely

25 wanted to get this question in and get a response from

Page 98

1 each of the panelists.

2 The final slide is, if there was one thing you

3 could change with the current legal approach to

4 predatory pricing, what would it be? And since we

5 started with Ken, I think this time we will end with

6 Ken, in reverse order. Doug, why don't you take this

7 one first.

8 MR. MELAMED: I think I would just try to

9 demystify it a little bit and think of it simply as part

10 of a complicated set of strategies that companies use

11 that under some set of circumstances can be

12 anticompetitive.

13 MR. POTTER: Patrick?

14 DR. BOLTON: Yes, I would agree with that, and I

15 would also vote for de-emphasis of the cost test and

16 putting intent back as a possible way of proving

17 predatory pricing, and here, I think it would be helpful

18 to maybe articulate the guidelines on how one would -ï½­

19 what's a legitimate way of proving intent and perhaps,

20 you know, move in that direction.

21 MR. POTTER: Janusz?

22 DR. ORDOVER: I don't think I have a favorite.

23 I will just say that I will agree with Doug, that we

24 need to get clarity on what are the public policy or

25 economic principles that either underlie the tests that

Page 99

1 are being proposed, where does the -- you know, the cost

2 tests, where do they come from? There has to be -ï½­

3 going back, I think at this point we have enough

4 learning to try to go back to first principles and try

5 to understand what it is that we are trying to

6 accomplish, taking full account of the administerability

7 of whatever provisions are going to ultimately be

8 developed, but I think it would be foolish for us -- for

9 me, anyway -- to vote for the least favorite aspect of

10 what is out there at this point.

11 MR. POTTER: Ken?

12 DR. ELZINGA: Well, I certainly can't argue

13 against Janusz's call for clarity, but I think we are in

14 a pretty congenial equilibrium right now.

15 MR. POTTER: Good. I just want to point out

16 that over on the side, we have some of this afternoon's

17 panelists that were kind enough to be here this morning.

18 They are John Kirkwood, Tim Brennan and Rick

19 Warren-Boulton, and just right before we leave, I will

20 give each of you 30 seconds to say anything you wanted

21 to say about this morning's panel, or if you just want

22 to wait until this afternoon, feel free. Anyone want to

23 take a go?

24 DR. WARREN-BOULTON: I think I will wait for

25 this afternoon, but -ï½­

Page 100

1 MR. POTTER: You will all wait for this

2 afternoon?

3 Well, if you could join me in thanking the

4 panelists.

5 (Applause.)

6 MR. POTTER: That will end the morning session,

7 and the afternoon session begins at 1:30. Thank you

8 very much.

9 (Whereupon, at 11:56 a.m., a lunch recess was

10 taken.)


Page 101


2 (1:28 p.m.)

3 MS. SCHULTHEISS: Everybody ready? Okay, good

4 afternoon, and for those of you who were not here this

5 morning, my name is Pat Schultheiss. I'm an attorney

6 with the Federal Trade Commission's Bureau of

7 Competition in the Office of Policy and Coordination,

8 and I am the lead moderator for this afternoon. My

9 co-moderator is Bob Potter, who is the Chief of the

10 Legal Policy Section at the Federal -- at the Department

11 of Justice's Antitrust Division. I was going to put you

12 at the FTC again, sorry, Bob. I can't just get those

13 words out.

14 Before we start, a few housekeeping matters.

15 First, for everybody's benefit, please turn off your

16 cell phones, Blackberries, any other device that might

17 make noise during the session. We appreciate that.

18 Second, the restrooms are out to -- men's room

19 right to the left, and the women's room, past the

20 elevators and to your left. There are little signs out

21 there to guide you as well.

22 Third, in the very rare event that the alarms

23 happen to go off, please just calmly proceed down the

24 staircase and follow the zillion FTC attorneys and staff

25 that will be going towards 7th Street to the Sculpture

Page 102

1 Garden. Those are the preliminaries.

2 This afternoon's topic is dealing with predatory

3 pricing but looking at it from the buying or bidding

4 side. We have assembled a very distinguished panel who

5 will discuss I think, among other things, just how

6 common buy-side predatory pricing is in the real world,

7 if at all, whether it's common at all, what Section 2 of

8 the Sherman Act can and should be doing about predatory

9 buying or bidding, and I'm sure there will be many other

10 things. I know we will have a little bit of raising

11 rivals' costs from Professor Salop and others. So, with

12 that, let me introduce very briefly the panel, and then

13 I will introduce with a little bit of greater detail

14 each panelist right before they speak.

15 Our panelists this afternoon, in the order they

16 will be speaking, are Professor Jack Kirkwood from the

17 University of Seattle; Professor Tim Brennan from the

18 University of Maryland, Baltimore County; Professor

19 Steve Salop from Georgetown University; Rick

20 Warren-Boulton, a consultant with Microeconomic

21 Consulting and Research Associates; and Janet McDavid, a

22 partner with Hogan & Hartson. Each of the panelists

23 will give a 10 to 15-minute presentation. After that,

24 we will take a brief break and then have a panel

25 discussion for the remainder of our time. I would like

Page 103

1 to thank all the panelists for being here and for the

2 morning panelists who have decided to stay and enjoy the

3 rest of the discussion. We greatly appreciate the

4 willingness of all of the panelists to give their time,

5 not just here today, but also the time input into

6 preparing for this session.

7 Our first speaker today is going to be Jack

8 Kirkwood, as I said. Jack is a professor of law at the

9 Seattle University School of Law. Before joining the

10 Seattle University, Professor Kirkwood was an attorney

11 with the Federal Trade Commission. Before leaving the

12 rarified air of Washington, D.C., he was the director of

13 several policy offices here in the Premerger

14 Notification Program, but then decided to head out to

15 the Pacific Northwest, and he joined the Seattle

16 Regional Office, where he headed up numerous antitrust

17 investigations and cases.

18 Professor Kirkwood has edited two books and

19 published numerous articles, and he recently addressed

20 the topic of today's hearing in his article in the

21 Antitrust Law Journal, "Buyer Power and Exclusionary

22 Conduct: Should Brooke Group Set the Standards For

23 Buyer-Induced Price Discrimination and Predatory

24 Bidding?"

25 In addition, Jack Kirkwood is a consultant for

Page 104

1 the plaintiffs in the Weyerhaeuser appeal against

2 Ross-Simmons in the Ninth Circuit. I think we only have

3 one panelist who is not somehow or another involved in

4 the Weyerhaeuser case, but with that, Jack.

5 MR. KIRKWOOD: Thanks, Pat, and thank you to

6 both agencies for inviting me.

7 This panel's focus is especially significant.

8 It's not only an intellectual or antitrust policy

9 question, but it is a question of how should the Supreme

10 Court come out in a case that in my sense is they are

11 very likely to take, Ross-Simmons versus Weyerhaeuser,

12 and the central issue in that case, of course, will be

13 should Brooke Group apply. Should Brooke Group's

14 price-cost and recoupment tests apply to a practice that

15 has been called predatory overbuying or predatory

16 bidding.

17 This is, as I've conceived it and as Steve and

18 others have looked at it, is in major respects the

19 mirror image of predatory pricing on the sell-side.

20 With predatory bidding, a dominant buyer bids up the

21 price of a critical input, forcing up the market price,

22 and in certain circumstances, making it impossible for

23 rival buyers to continue to compete, causing either

24 their exit or their inability to constrain the dominant

25 buyer's future exercise of power, and hence, the

Page 105

1 dominant buyer captures monopsony power that it wouldn't

2 otherwise have, and in some circumstances, this may lead

3 to a long-run harm to welfare, most directly to supplier

4 welfare but also possibly to consumer welfare as well.

5 Given these similarities, all of the Supreme

6 Court's stated rationales for applying Brooke Group

7 apply to predatory bidding as well, at least to some

8 degree. Furthermore, the only alternative test approved

9 by the Ninth Circuit was an exceptionally vague jury

10 instruction allowing the jury to find liability if it

11 found that Weyerhaeuser bid more than "necessary" in

12 order to prevent the plaintiff from buying at a "fair

13 price." That strikes me as too vague and many others as

14 well.

15 So, if that test is not acceptable, should we

16 resort to Brooke Group or try something in the middle?

17 And what I am going to suggest today is that a

18 middle-of-the-road test is more appropriate, a test that

19 has two parts. One, the plaintiff would have to show

20 harm to welfare, I will explain what that means, but the

21 defendant would get a complete defense if it can show

22 that it would pass the no-economic-sense test.

23 Why a middle-of-the-road test, why not Brooke

24 Group? In some significant respects, predatory bidding

25 is different from predatory pricing. There is, of

Page 106

1 course, one cosmetic difference. In predatory bidding,

2 the first thing that happens is input prices go up,

3 output prices to consumers do not go down, at least

4 initially. So, there is this key difference in terms of

5 what we normally think of as the central focus of

6 antitrust on providing low prices to consumers, but that

7 is essentially a cosmetic rather than an important

8 difference, because when a buyer bids up input prices

9 that benefit suppliers, it can be procompetitive, and

10 through an output effect I will describe, it can benefit

11 consumers as well. So, we should be concerned with

12 chilling procompetitive bidding for inputs just as we

13 are concerned about chilling procompetitive price cuts.

14 There is, though, a more significant difference.

15 Compared to predatory pricing cases, predatory bidding

16 cases have been brought less frequently, have been won

17 less frequently, and arguably, there have been no false

18 positives, no liability findings where it appeared that

19 the defendant had not, indeed, harmed welfare. That is

20 arguable, not hardly proven.

21 In the last two decades, since the mid-eighties,

22 there have only been two cases in which the plaintiffs

23 have won. Both have involved bidding up timber prices

24 in the Pacific Northwest, and if the Supreme Court takes

25 Weyerhaeuser, we are not even sure that the plaintiff

Page 107

1 will have won the second case. Why is it that these

2 cases seem to be rarer and that the proportion of

3 successful predation may be higher? Rick Warren-Boulton

4 will address that. I will defer to him on that. I have

5 suggestions, but he can cover it.

6 This track record suggests to me, at least, that

7 it is a little too early to apply a Brooke Group

8 price-cost safe harbor test to predatory bidding. To be

9 sure, the number of successful cases is too small to

10 produce a reliable conclusion that predatory bidding is

11 more dangerous than predatory pricing to welfare. We

12 are not there yet, yet the track record does suggest

13 that the danger of deterring procompetitive bidding is

14 less high than it is with predatory pricing, at least if

15 there were a stiff rule that applied to a plaintiff as I

16 will suggest.

17 This is consistent with my experience at the FTC

18 as head of the Planning Office, the Evaluation Office,

19 and then as a staff attorney in Seattle. I received

20 over the years many complaints about price-cutting but

21 never, ever, a complaint about bidding up input prices.

22 There are two other reasons to think we ought at

23 this point to choose a more flexible test rather than a

24 Brooke Group safe harbor. One is, there has been to

25 date much less scholarly or judicial analysis of the

Page 108

1 practice, though thanks to the agencies and thanks to

2 Steve, we are working on that. Even so, even so, the

3 pile of articles on predatory bidding does not compare

4 to the mountain on predatory pricing.

5 There is also a growing, though probably still

6 minority, view that the Brooke Group average variable

7 cost test, at least as interpreted that way, may not be

8 the right standard even for predatory pricing where the

9 concern with chilling procompetitive price competition

10 is greater; that at least it counsels against extended

11 Brooke Group predatory bidding.

12 What would an alternative test look like? My

13 suggestion, just a proposal, is to put a stiff burden on

14 the plaintiff and to give the defendant a complete

15 defense. The plaintiff's burden would be to show harm

16 to welfare. So, a plaintiff would have to prove the

17 elements of a welfare-reducing instance of predatory

18 bidding. So, they would have to show that, yes, the

19 price was bid up; yes, at least some significant rivals

20 were constrained in their ability to hold up, since we

21 are talking about bidding, hold up the alleged

22 predator's price; as a result, the predator got

23 monopsony power it would not otherwise have; and most

24 important of all, the plaintiff would have to show that

25 the long-run impact on welfare was negative. So, to

Page 109

1 pick up on one of Ken's key points this morning, the

2 plaintiff would have to show that it was relatively easy

3 to induce exit, but either re-entry or new entry would

4 be more difficult.

5 How should we measure welfare? I have

6 deliberately not put an adjective in front of it. It

7 seems to me there are two principal possibilities: One,

8 supplier welfare; two, consumer welfare. Steve is going

9 to talk about either of those measures, particularly

10 consumer welfare as opposed to total welfare. I will

11 skip that debate.

12 Between supplier welfare and consumer welfare,

13 it seems that both precedent and ease of measurement

14 favor supplier welfare. The cases that have looked at

15 monopsony abuses and at buyer cartels tend, on balance,

16 to focus on the impact on suppliers rather than on

17 consumers.

18 In addition, there can be instances of

19 substantial harm with little or no measurable effect on

20 consumers. So, if the plaintiff had to show some sort

21 of significant, discernible, provable effect on

22 consumers, that would be harder. So, I am tending to

23 favor a supplier welfare test, but you could use a

24 consumer welfare test at least for most cases. Why is

25 that? Because successful monopsonization is likely to

Page 110

1 harm consumers in two ways.

2 One, as many economists have pointed out, if the

3 predatory bidding produces a net increase in long-run

4 monopsony power, then there is likely to be a reduction

5 in output. The dominant buyer is likely, on balance,

6 over time, to buy less, and if it buys less, it is

7 likely to produce less, and that means there is likely

8 to be less final product on the market. And so, if the

9 demand curve is neither totally vertical nor totally

10 horizontal, if it is the normal downward sloping type,

11 then less output is going to put some upward pressure on

12 price.

13 So, the mere output effect will tend to harm

14 consumers, again, maybe not noticeably, but there is

15 that linkage, and that does not require market power.

16 That is, the dominant buyer does not have to have market

17 power in the final product market for this effect to

18 occur. It occurs through the output reduction caused by

19 the monopsonization.

20 There could be, though, a market power effect,

21 as Rick may emphasize in his talk. Suppose the relevant

22 market downstream was limited to the product whose input

23 price was bid up. Then, if the dominant buyer

24 eliminates its key rivals as buyers, it will also

25 eliminate them as sellers. So, it may gain both

Page 111

1 monopsony power upstream and monopoly power downstream,

2 and so then that would magnify the potential consumer

3 effect.

4 We could, therefore, use a consumer welfare

5 test, as Steve may suggest. There might have to be

6 exceptions, though, from such a test where predatory

7 bidding leads to monopsonization but consumers are

8 unaffected.

9 Whatever criterion is used for welfare, it seems

10 to me that a welfare test would provide substantial

11 protection to defendants. First, successful

12 monopsonization appears to be rare and appears to be

13 limited to certain markets, as Rick will suggest,

14 markets where there is inelastic supply, and that is not

15 commonly observed, typically in labor or natural

16 resource markets.

17 The power buyers that we all know or suspect,

18 the Wal-Marts, the Barnes & Nobles, the Costcos, they

19 don't induce lower prices by monopsonization. That is,

20 they do not go to their suppliers and say, "I am going

21 to cut back my output a little bit, and I expect,

22 because that will reduce your marginal cost, that you

23 will give me a lower price." Rather, they engage in

24 bargaining tactics, and at the risk of oversimplifying,

25 the way they obtain a lower price is, in essence, saying

Page 112

1 that I am going to increase my purchases over what they

2 would otherwise be if and only if you give me a lower

3 price. So, that is not monopsonization.

4 So, one, you have the limited set of cases, and

5 two, you have all of those elements that the plaintiff

6 has to prove. I will not repeat them again, but showing

7 net long run harm to welfare is not an easy task. This

8 is essentially a full rule of reason analysis, and as

9 you well know, private plaintiffs do not often prevail

10 in full rule of reason cases. That has been the record

11 under Section 1 and is likely to be the record under

12 Section 2 as well.

13 You might say, "Ah, but private plaintiffs

14 prevailed in LePage's and below in Weyerhaeuser," but

15 the difference is in neither of those cases did the

16 courts insist on a full rule of reason net welfare test.

17 Is such a test unworkable? It is certainly

18 reasonable to contend it is, but we do use it in

19 horizontal merger cases under Section 7 and in full rule

20 of reason cases under Section 1, and as Steve has

21 pointed out, we are not really balancing immeasurables

22 when we use this long-run welfare test. We are not

23 trying to decide what's more important, national

24 security or freedom of speech. We are asking whether

25 the long-run impact on our target group, let's say

Page 113

1 suppliers, is positive or negative.

2 Still, still, the inquiry would not be easy for

3 a defendant to predict its outcome, and the inquiry

4 would stretch over a longer time period than in the case

5 of a Section 7 matter certainly and probably the typical

6 Section 1 matter, because we are talking about a

7 long-run impact on welfare, and the key issue there, as

8 Ken has stressed this morning, is entry barriers, and

9 that is an uncertain and controversial topic. So, my

10 sense is that we should not rely solely on a welfare

11 test, that we should create an efficiency defense for

12 the defendant, and for that, I have borrowed from the

13 no-economic-sense test advocated by the Division.

14 It seems to me that if the defendant can show

15 that bidding up input prices was profitable, without

16 regard to any increase in monopsony power, that it

17 should have a complete defense. This would put the

18 burden on the party that best knows its own

19 profitability and would give it an out if it could

20 provide a good answer to the question, why did you do

21 this?

22 So, to conclude, let me give you just a simple

23 example. Suppose, as is my understanding of

24 Weyerhaeuser's theory, suppose the dominant buyer

25 improved its production process, lowering its marginal

Page 114

1 cost. Then the new profit-maximizing price, without

2 regard to monopsony power, might be an increase in

3 output. That would entail buying more input, selling

4 more of the final product, so there would be a margin

5 reduction from paying more for the input and getting a

6 little less for the final product, but if the lower

7 marginal cost more than compensated for that, without

8 figuring in any increase in margin due to monopsony

9 power, then the defendant would be excused.

10 I am happy to talk more about that, but my time

11 is up.

12 MS. SCHULTHEISS: Thank you, Jack.

13 Our next speaker is Tim Brennan, who is, as I

14 indicated, a Professor of Public Policy and Economics at

15 the University of Maryland, Baltimore County. Professor

16 Brennan also has been serving as the 2006 T.D. MacDonald

17 Chair in Industrial Economics at the Canadian

18 Competition Bureau.

19 Before joining the University of Maryland,

20 Professor Brennan held a number of positions focusing on

21 economics and antitrust, including staff economist at

22 the Antitrust Division, senior economist for industrial

23 organization and regulatory policy on the staff of the

24 White House Council of Economic Advisers, and a

25 consultant to the Bureau of Economics here at the

Page 115

1 Federal Trade Commission.

2 Professor Brennan's research areas related to

3 antitrust include regulatory economics, monopolization

4 law, exclusionary conduct, vertical integration, and the

5 competition-regulation interface. His articles have

6 appeared in numerous journals in economics, law, and

7 other fields.

8 Tim, would you care to start?

9 DR. BRENNAN: Thank you.

10 I am grateful to the Department of Justice and

11 to the FTC for the invitation to participate on this

12 predatory buying panel. It is a great honor for me to

13 be here. I am especially grateful because I have been

14 thinking for longer than I care to remember about how to

15 support Section 2 of the Sherman Act, and yet reconcile

16 it with less controversial, more accepted frameworks for

17 prosecuting cartels and horizontal mergers.

18 I will offer a suggestion along those lines

19 today. Although I believe that my suggestion will make

20 deserving exclusion cases easier to bring, some aspects

21 may be significantly different from established

22 jurisprudence. For that reason, I particularly

23 recognize the privilege of having a place at this

24 distinguished table.

25 Before proceeding, I need to say my statement

Page 116

1 today reflects solely my own opinions and does not

2 represent those of the Competition Bureau or any of its

3 staff.

4 For this complex topic, I offer a series of

5 recommendations.

6 Predation or exclusion? Pick one or the

7 other -- they are fundamentally different.

8 When first asked to participate in a panel on

9 "predatory buying," my response was to object to the

10 title. We should recognize that "monopolization"

11 entails two essentially different types of practices,

12 one that for shorthand could be called "predation," and

13 the other "exclusion." The most succinct distinction is

14 that predation cases involve doing too much of a good

15 thing to bring about a bad result later. There, the

16 understandable concern is with deterring energetic

17 competition -- not discouraging firms from charging low

18 prices, adding product features, and the like.

19 Exclusion cases, on the other hand, involve

20 doing a bad thing now. One way or another they come

21 down to acquiring control and effective market power

22 over a supplier or access to an input or service needed

23 to compete, what economists called complements. The

24 most explicit way to accomplish such control would be

25 through a series of exclusive contracts with the

Page 117

1 complement's suppliers. It may involve overbuying

2 inputs through explicit purchase or, as I'll suggest

3 below, bundling, rebates, or other forms of "leaving

4 money on the table." I call this practice "complement

5 market monopolization," or CMM.

6 The major problem with single-firm conduct law

7 is the failure to recognize the essential difference

8 between these two types of conduct, leading to the

9 counterproductive imposition of predation standards on

10 exclusion cases. Perhaps the failure arises from a

11 presumption that one statute -- Section 2 -- must imply

12 one principle. Perhaps it follows from the persistent

13 belief that Section 2 must be premised on harm to

14 rivals. Since competition also harms rivals, Section 2

15 law is thus driven by fear of over-deterrence. Instead,

16 exclusion cases should be recognized as different, where

17 we can apply horizontal tools and not predation screens

18 to the delineation and protection of complement markets.

19 Two, genuine predatory buying cases will be

20 rare; when they occur, validate necessary assumptions.

21 I would have changed the title of this panel to

22 "Exclusionary Buying," because the leading cases involve

23 creating of market power over complements. The recent

24 DOJ/FTC cert petition in Weyerhaeuser v. Ross-Simmons

25 illustrates an exception that proves the rule. The

Page 118

1 exception is unusual, in that the concern is not that a

2 timber processor would acquire so much control over a

3 relevant market in uncut trees to be able to raise their

4 effective price. Rather, according to the petition, the

5 allegation is that a mill would pay too much for trees

6 to drive out other buyers, with subsequent recoupment by

7 cutting prices paid for trees in the future.

8 I have little to say about which market power,

9 price-to-cost, and recoupment tests are appropriate for

10 preventing over-deterrence in these rare predatory

11 buying cases. I do suggest that courts demand not only

12 evidence appropriate for such tests. They should also

13 demand evidence that specific assumptions behind

14 strategic models are satisfied, i.e., that the alleged

15 predator either has a reputation for non-profit

16 maximizing behavior to protect, or benefits from

17 identified asymmetric failures in capital markets.

18 Theoretical possibility alone does not make a practice

19 harmful.

20 Three, for exclusion cases, the first and

21 crucial step is to delineate a complement market being

22 monopolized using the Horizontal Merger Guidelines

23 procedures.

24 Market power is often characterized not just as

25 the ability to raise price but also as "the ability to

Page 119

1 exclude." This is a mistake of imprecision. Ability to

2 raise the price of X depends upon entry barriers or

3 other impediments to competition, but those do not

4 depend upon the price of X. Higher X prices would, if

5 anything, encourage entry. Rather, the ability to

6 exclude depends upon control over the prices of Y, Z, W,

7 or something else needed to enter and produce X.

8 Delineation of that relevant complement market

9 should therefore be the first step in all exclusion

10 cases. Taking Dentsply as an example, the case rested

11 on the premise that the national distributors constitute

12 what in merger contexts we would regard as a relevant

13 market, in this case for the distribution of teeth to

14 dental labs. The Merger Guidelines provide the useful

15 framework for testing this premise. They ask whether

16 teeth manufacturers would turn to other distributors,

17 whether there would be entry into that distribution

18 market in response to a "small but significant

19 nontransitory increase in price" of using such dealers.

20 I do not know the facts of that case and thus

21 the answers, but the Merger Guidelines ask exactly the

22 right questions. Cases eventually turn to evidence of

23 entry or substitution into the complement market, but

24 they do not make such concerns central -- the best

25 indicator being the continued identification of a

Page 120

1 relevant market as that in which the alleged monopolizer

2 is already dominant, not that over inputs or services

3 competitors need to compete. Control over such a

4 complement market is not only sufficient to raise

5 competitive concerns; it is necessary for

6 anticompetitive exclusion.

7 Hence, plaintiffs should focus on identifying

8 that complement market and showing that the practices at

9 hand cover enough of it to raise the complement's price.

10 In effect, one should ask if one would be troubled if

11 the complement providers covered by the alleged

12 exclusionary practice merged. Unlike usual

13 characterizations of monopolization cases, this is one

14 we know how to answer -- use the Merger Guidelines. If

15 the answer is no, stop; if the answer is yes, go to the

16 next step.

17 Next, having delineated the relevant complement

18 market, the second step should be to establish the price

19 effect in that market.

20 Barriers to entry cannot be raised, and

21 competition impeded, by any more than the extent to

22 which the price of the complement can be raised.

23 Sometimes this higher price will be explicit, sometimes

24 it will be only an inferred higher price -- Professor

25 Carlton has usefully called it a "shadow price" -- if

Page 121

1 the exclusionary practice so ties up the complement

2 market that only higher priced substitutes, including

3 self-provision, are available.

4 Explicit exclusive dealing contracts offer one

5 such standard: Firms wanting to use those dealers would

6 have to cover the cost of breaching the contract. Other

7 alleged exclusionary practices, such as bundle discounts

8 or royalty rebates, may create a significant price

9 increase -- once one has established the first step.

10 Next, the standard for assessing the

11 exclusionary effect of a bundle or rebate is not whether

12 an incremental price is below incremental cost, but its

13 effect on the price of the complement.

14 Following the last point, one could ask whether

15 bundles, rebates, or other programs have to increase the

16 effective price of the complement as much as it would

17 explicit contracts. I have no reason to believe it

18 should. Were we to follow the Merger Guidelines, as we

19 should for complement market definition, we might only

20 need ask if the practice leads to a small but

21 significant nontransitory increase in price of the

22 complement.

23 This tells us that whether a bundle is

24 anticompetitive has nothing to do with a predation-like

25 test. It does not depend on whether the incremental

Page 122

1 price of adding a good to a bundle, or of supplying more

2 of a product given a discount, is less than some measure

3 of marginal or average variable cost. Rather, it

4 depends only on the extent to which such practices

5 create market power in order to raise the price others

6 must pay for the services provided by retailers,

7 distributors, or other complement providers getting the

8 discount.

9 Predation case screens -- profit sacrifice,

10 equally efficient competitor, and prior dominance -- do

11 not belong in exclusion cases.

12 Even for predation, we have heard today, some

13 commentators have noted that some or all of the screens

14 need not increase competition and consumer benefit.

15 Nevertheless, they may be appropriate to prevent

16 over-deterrence of competition through low prices or

17 added features. However, in exclusion cases,

18 controlling a monopoly share of complement markets is

19 not inherently procompetitive, and thus need not have

20 high bars for its protection.

21 The profit sacrifice or "no business sense"

22 test -- the two are equivalent if one assumes that

23 "business sense" means "maximize profits" -- substitutes

24 concern with intent and tactics for concern with

25 effects, as if whether someone had been murdered depends

Page 123

1 upon the price paid for the gun. Others have noted that

2 it creates an absolute efficiencies defense, in that a

3 penny of gain from a practice excuses untold

4 anticompetitive harms. As Rick Warren-Boulton has said,

5 the test is notably inappropriate when regulated

6 monopolists do the excluding.

7 Although I have criticized "raising rivals'

8 costs," mostly for its emphasis on "rivals," Steve

9 deserves enormous credit for pointing out long ago that

10 predatory sacrifice and recoupment is unnecessary to

11 point out the tactics that raise those costs. My

12 difference is that I would focus primarily on the

13 complement market.

14 Ironically, the test also forgets that once upon

15 a time, profit sacrifice implied previously unobserved

16 efficiency, not anticompetitive harm. We learned that

17 exclusive territories, exclusive dealing, tying, and

18 even resale price maintenance must generate efficiencies

19 because they reduce demand, making even monopolists

20 worse off otherwise. That realization gradually

21 reformed most vertical restraint law. Assuming now that

22 a profit sacrifice must be anticompetitive forgets

23 antitrust history and invites us to repeat mistakes that

24 have not been fully undone after nearly a century.

25 On equally efficient competitors, I point out

Page 124

1 what should be obvious: Inefficient competitors hold

2 down price. Complement market monopolization leading to

3 their exclusion can raise price and harm consumers.

4 Having gone after two sacred cows, I may as well

5 finish off the herd: The Grinnell prior possession of

6 monopoly test can also impede meritorious exclusion

7 cases. It distracts attention away from the complement

8 market, focusing instead on the characteristics of who

9 monopolized it. Prior dominance could even be a

10 defense, but once complement market monopolization is

11 shown, it should be up to a defendant to claim that it

12 has no consequence because of monopoly elsewhere in the

13 production chain.

14 Moreover, this test is counterproductive.

15 Proving the cost, demand, and entry barriers necessary

16 to establish prior dominance undercuts the argument that

17 the alleged exclusionary practice makes a difference.

18 Using Richard Posner's phrase, the monopoly should be

19 "fragile" at worst. An exclusion case will be strongest

20 if the sector would be competitive, but for the practice

21 under scrutiny.

22 Ask whether we would apply these standards to

23 mergers. Should all mergers be legal unless one could

24 show they would be unprofitable but for anticompetitive

25 harm? Should any merger, including to monopoly, be

Page 125

1 legal if a more efficient firm buys and eliminates a

2 less efficient competitor? Of course not. Even prior

3 dominance may make the incremental effect of a merger

4 less troubling. If these tests would gut merger law,

5 and if exclusion cases are akin to acquisitions in the

6 complement market, they do not belong on this side of

7 Section 2.

8 Consider share-based rather than "all or

9 nothing" remedies.

10 Analogy to mergers opens the door to more

11 creative remedies. Generally, either a practice is

12 okay, or it is not and should be stopped. We should

13 instead take a share-based approach. Exclusive dealing

14 contracts, bundles, or other alleged monopolizing

15 practices might have efficiency benefits. The problem

16 is not the practices per se, but their scale -- that

17 they pre-empt so much of the complement market to raise

18 its price significantly. Rather, defendants should be

19 allowed to retain the practice, but only over a

20 nondominant share of the complement market, 35 percent,

21 50 percent or some appropriate number. If the practice

22 is actually efficient, it will be kept. If it serves

23 only to exclude, this remedy would lead to its

24 discontinuance.

25 Last, focus on the creation of monopolies, not

Page 126

1 their maintenance.

2 About two years ago, I gave a talk at the FTC on

3 these ideas, entitled "Saving Section 2." As I began,

4 an economist there asked, "Why should anyone want to

5 save Section 2?" My answer may not have satisfied him,

6 but in short, it is that it can and should be saved.

7 Were all Section 2, single-firm conduct cases about

8 protecting a monopolist's rivals by drawing vague or

9 impossible lines between competing just enough and too

10 much, I might have agreed with the questioner. However,

11 exclusion cases are not about maintaining monopolies but

12 creating new ones. In focusing on complement market

13 monopolization, such cases can and should be no more

14 controversial than collusion and merger cases are today.

15 Thank you again for the privilege of allowing me

16 to share these observations. I hope I can clarify them

17 through responses to any questions you have here and as

18 they arise in the future. Thanks very much.

19 (Applause.)

20 MS. SCHULTHEISS: Our next speaker will be Steve

21 Salop, who is a Professor of Economics and Law at the

22 Georgetown University Law Center where Steve teaches

23 courses in antitrust law and economics and economic

24 reasoning for lawyers. Dr. Salop also has a consulting

25 practice at CRA International involving a variety of

Page 127

1 antitrust issues.

2 Before joining the Georgetown faculty, Steve

3 held positions at the Federal Trade Commission, and a

4 while back, the Civil Aeronautics Board and the Federal

5 Reserve Board. Professor Salop has written numerous

6 articles in various areas of antitrust economics and

7 law, many of which take a, quote unquote, "post-Chicago"

8 approach. Professor Salop recently published two

9 articles in the Antitrust Law Journal that concern

10 exclusionary behavior and monopoly power.

11 Of particular importance for today's hearing is

12 his article, "Anticompetitive Overbuying By Power

13 Buyers." It contrasts predatory versus raising rivals'

14 costs, overbuying behavior, and I will let him go

15 further into that.

16 And in addition, Professor Salop has also

17 consulted for Weyerhaeuser in its appeal and district

18 court decision. Is that correct?

19 DR. SALOP: Yes, that's correct.

20 MS. SCHULTHEISS: So, with that, I will hand it

21 over to you.

22 DR. SALOP: Okay, thank you.

23 I want to talk about these two types of

24 overbuying, predatory overbuying, what Jack called

25 predatory bidding, and raising rivals' costs overbuying,

Page 128

1 but to just set it up, I think a key place to begin is a

2 notion that Tim Brennan touched on, the fact that in

3 Section 2, there are really two distinct paradigms, and

4 I believe that the way individual people think about

5 Section 2 has a lot to do with which paradigm they have

6 in mind, you know, which one animates them, and so I

7 want to stress the difference between these two

8 paradigms.

9 One paradigm is the predatory pricing paradigm,

10 seller-side predatory pricing, and the other is the

11 raising rivals' costs or non-price predation paradigm.

12 Now, in my view, and a key element in what I am going to

13 talk about today, and, indeed, much of my work, is that

14 conduct that fits into the raising rivals' cost paradigm

15 raises much greater concerns than conduct that fits

16 within or that people characterize as the predatory

17 pricing paradigm.

18 Now, we all know the claims about why predatory

19 pricing is seldom attempted and rarely succeeds. In the

20 short run, the predator loses more money than the

21 victim. Secondly, it only works if the victim exits.

22 Otherwise, there is no -- they will never be able to

23 recoup. And third, consumers benefit from the lower

24 prices, and the harm to consumers is mere speculative

25 impact in the future. So, for all those reasons, it is

Page 129

1 argued that predatory pricing is unlikely to be tried,

2 it is unlikely to succeed, and it is unlikely to harm

3 consumers, and therefore, we should have a really light

4 hand in predatory pricing.

5 Taken as a paradigm, conduct that raises rivals'

6 costs raises much greater antitrust concerns; hence,

7 let's say it is more likely to succeed, more likely to

8 harm consumers. Why? Well, first, there is no need to

9 induce competitors to exit. If you raise competitors'

10 costs, variable costs, they will tend to raise price,

11 and the excluding firm will gain even if the competitors

12 do not exit. You would rather compete against a

13 high-cost competitor than a low-cost competitor.

14 Secondly, there is no necessity for short-run

15 profit sacrifice, chronological profit sacrifice, of the

16 sort there is in predatory pricing. If the rivals'

17 costs are increased, they will raise price immediately.

18 The predator, the excluding firm, will gain immediately.

19 So, there is no issue that the predator has to lose

20 money for a while and then only gain later.

21 Similarly, there is no short-run consumer

22 benefit, and this one I think is very important. In

23 predatory pricing, the consumers inevitably benefit from

24 that lower price in the short run. In raising rivals'

25 costs, there is no inherent consumer benefit. You raise

Page 130

1 rivals' costs, they raise price or contract, and the

2 defendant raises price, so consumers are harmed

3 immediately. There is no such thing as naked predatory

4 pricing, you know, just all bad, the way there is with

5 naked price-fixing, but there is naked raising rivals'

6 costs. One could conceive of that burning down the

7 factory, so on and so forth, and such conduct that

8 actually shows up in certain cases.

9 So, for all those reasons, I think that you have

10 these two paradigms that are distinct, and I think most

11 of the time you should be thinking in terms of the

12 raising rivals' costs paradigm. I think it is a better

13 paradigm for Section 2. I think that the predatory

14 pricing is the exceptional paradigm, not the norm.

15 Well, now let's apply this to anticompetitive

16 overbuying. Now, there are -- by overbuying, I mean

17 conduct where the defendant goes into the input market

18 and bids up the price of the input. Usually if you bid

19 up the price of the input, you are generally almost

20 surely going to buy more than you would have otherwise.

21 So, it is often called overbuying. Indeed, in the

22 literature, it was initially referred to as overbuying

23 cases, and I guess sort of the classic case that

24 economists studied initially was the bauxite aspect of

25 the Alcoa case, where Alcoa was alleged to have

Page 131

1 overbought bauxite in order to raise the costs to its

2 aluminum rivals.

3 Well, there are two distinct overbuying

4 allegations that correspond analytically to the two

5 exclusion paradigms. There is predatory overbuying,

6 Jack talked about, and then there is raising rivals'

7 costs overbuying, and the difference between these two

8 paradigms is the goal of predatory bidding or predatory

9 overbuying is to gain monopsony power to the input

10 market, as Jack pointed out. The goal of raising

11 rivals' costs overbuying is to raise your rivals' costs

12 and then gain market power in the downstream output

13 market.

14 Okay, so if you think about Alcoa, they could

15 have overbought bauxite for -- one reason would be to

16 ultimately knock out the other purchasers of bauxite so

17 it could then be a bauxite monopsonist, and the

18 alternative would be that they did it in order to raise

19 the price of bauxite to its rivals so that they could

20 ultimately monopolize, raise the price, of aluminum. Of

21 course, in a given case, you could have both, but at the

22 same time, in a case, you could have one or the other.

23 Now, interestingly, in the Weyerhaeuser case and

24 the Ross-Simmons case, both allegations were made in the

25 complaint. It was alleged that Weyerhaeuser, by its

Page 132

1 conduct, would ultimately gain monopsony power in the

2 timber market, in the purchase of timber, and secondly,

3 it was argued that Ross-Simmons would gain market power

4 in a downstream alder wood, alder hardwood market, and

5 Ross-Simmons did not carry its burden on the raising

6 rivals' costs piece, and so the part that has gone up is

7 just the predatory overbuying piece.

8 Okay, well, how do I think we should evaluate

9 these two types of conduct? I want to separate them.

10 So, first, the predatory overbuying, as I said, it's

11 market power in the upstream market that's the goal, and

12 in general, I have in mind a four-step legal standard,

13 very close, very, very close, to what Jack Kirkwood

14 called for, and I will talk about it and then stress the

15 differences.

16 So, the four steps would be, you have got to

17 show buyer power, monopsony power, and artificially

18 inflated input purchasing, and it's really the latter,

19 you have to show that they bought more and that the

20 price went up. You have to show exit or permanent

21 capacity reduction of the input market competitors, that

22 should be, and then there has got to be some kind of

23 recoupment through buyer-side monopsony power in the

24 input market, and finally, and this is very important,

25 you have to show net consumer harm.

Page 133

1 Now, usually in these cases there will be a

2 short-run consumer benefit during the predatory period,

3 because when the defendant buys more of the input, it

4 will produce more output, and so the price of output

5 will go down, and then during a recoupment period, it

6 goes in the other direction, so consumers benefit in the

7 short run, harmed in the long run, and for that reason,

8 in order to show -- in order to gain -- show liability,

9 the plaintiff would have to show consumer harm on

10 balance.

11 Okay, now, let me go through the steps in a

12 little more detail. First of all, note, Jack Kirkwood

13 said, well, I'm not sure you need consumer harm, maybe

14 it's enough to have supplier harm, so that is one

15 difference between our standards to date. Jack actually

16 in his article had this last step, consumer harm. So,

17 he has broadened his position today.

18 Now, the first step, the issue here is the

19 question is, is the increased purchasing artificial or

20 is it competitively driven? Now, you know, there are a

21 lot of good reasons why a firm may increase its input

22 purchases in this year versus last year. For example,

23 maybe the demand for its output went up. It needs more

24 inputs to produce more output. Maybe it is not so much

25 its demand went up, maybe it decided to change its

Page 134

1 business strategy. Maybe it decided to decide to grow

2 its market share rather than go for a high price in the

3 output market. Or maybe it got a new production process

4 that is more efficient, and that leads it to want more

5 inputs in order to expand. Fourth, maybe something

6 happened in the input market. It used to have monopsony

7 power, and it finds it has lost that monopsony power or

8 has less of it, and so it wants to stop acting like a

9 monopsonist.

10 All four of those reasons would lead it to

11 increase its demand for inputs. As its demand for

12 inputs goes up, the price of the inputs would tend to

13 rise, and its purchases would tend to rise. So, for

14 these reasons, I mean, what would be most suspicious

15 would be if the defendant bought extra inputs and then

16 did not use them, just warehoused them, all right,

17 because that would suggest it was not buying more in

18 order to produce more output, but rather, did it in

19 order to raise price and drive its rivals out of

20 business. So, warehousing would be an issue.

21 Of course, the fact that it has inventories does

22 not necessarily mean it is warehousing. It could have

23 been an error. It could have just bought more thinking

24 it was going to need it and then it just did not need

25 it. So, one has to be careful there.

Page 135

1 Because of the concern, you know, that there are

2 all these legitimate reasons why you might want to

3 purchase more output, and the fact that there is this

4 inherent short-run consumer benefit, I am very worried

5 that there could be false positives, and for that

6 reason, I am willing to put on the Brooke Group style

7 test of output priced below cost, where it's really sort

8 of that marginal revenue product, or Rick went on to the

9 value of marginal product, being less than the input

10 price. Because here, like with predatory pricing, there

11 is inherent consumer benefit in the short run that may

12 or may not be offset by consumer harm in the long run,

13 and because there is that balance, it is a lot like

14 predatory pricing, and so the rule might be close to the

15 predatory pricing rule.

16 I also think that in the end, this is why the

17 Supreme Court will opt for a Brooke Group kind of test,

18 because it is so close to the reasoning in Brooke Group

19 that they are going to probably find it irresistible to

20 change the rule.

21 Now, it is possible that there is no consumer

22 benefit in the short run. If the demand for the output

23 was perfectly elastic, demand for the input was

24 perfectly elastic, in terms of price taker, then, when

25 it increases its purchases, it will not reduce the price

Page 136

1 that consumers pay for the outputs. There would be no

2 consumer benefit. But on the other hand, and under

3 those same circumstances, there would not be any

4 consumer harm. So, under those circumstances, the

5 plaintiff would lose anyway. So, I think the Brooke

6 Group, adding the price-cost test in Brooke Group, will

7 not cause any damage in this situation.

8 Now, when it comes to raising rivals' costs

9 overbuying, I feel a more interventional stance is

10 necessary. Again, I have got a four-step legal

11 standard. I do not have the Brooke Group piece, and, of

12 course, the analysis is somewhat different, because the

13 goal here is to gain market power in the downstream

14 market, not to gain market power in the upstream market.

15 So, you would still ask whether there were good reasons

16 for the firm to increase its demand for the input, but I

17 would not go so far as the Brooke Group test.

18 Here, you not only have warehousing could be a

19 concern, but also naked purchasing. Now, what I mean by

20 naked purchasing is suppose the firm does not even use

21 the input. It might buy up some of the input for the

22 sole purpose of making it more expensive for its rivals.

23 For example, in the Alcoa case, it was alleged that

24 Alcoa bought exclusive contracts to electricity from

25 utilities where it never had a plan. They just bought

Page 137

1 the exclusive so that other aluminum companies could not

2 buy electricity in those regions. So, that would be an

3 example of naked buying, there where an exclusive wasn't

4 overbuying, it was simply buying the exclusive.

5 You would never have naked overbuying for

6 predatory overbuying, because you can't get a

7 monopsony -- why would you want a monopsony? Why would

8 you want to knock rivals out and get a monopsony over

9 some input that you don't even use? So, you know, it

10 just does not compute.

11 So, assuming that you can show inflated input

12 purchasing, then you go through the standard type of

13 raising rivals' costs analysis to show consumer harm.

14 You have to show that rivals' costs were raised

15 materially, but that's not enough. That's just harm to

16 competitors. We know in Section 2 that is not enough.

17 You also have to show harm to competition, as you have

18 to show downstream market power, some power over price

19 downstream. And then, as before, you need to show net

20 consumer harm, because the overpurchasing could have

21 efficiency benefits. Maybe the firm is producing more

22 output and, you know, competing harder, that is a

23 benefit. So, you want to have that consumer harm

24 standard. As I said, it is more interventionist.

25 Now, by consumer harm, I really mean it. I mean

Page 138

1 true consumer harm. I do not mean supplier welfare, and

2 I do not mean what Robert Bork called the consumer

3 welfare standard, which was really a total welfare

4 standard. I think Bork was either deceiving people,

5 which is what Herb Hovenkamp has said, or maybe Bork was

6 just confused, and the reason I think it is possible he

7 was confused is that if you care about the total welfare

8 standard, then competitor injury is enough to carry the

9 day, because competitors are part of total welfare.

10 Indeed, it is easy to construct mainstream

11 simple examples in which conduct could harm competitors,

12 consumers could benefit with lower prices, but yet total

13 welfare could fall, and yet under the total welfare

14 standard, total welfare would go down, and so if you had

15 in mind a total welfare standard, that conduct would be

16 illegal. Now, I just cannot believe that Bork would

17 want to make such conduct illegal. So, I have chosen to

18 believe, out of respect, that he was confused rather

19 than attempting to create a consumer protection problem

20 or a court protection problem.

21 There are other reasons why I think the consumer

22 welfare standard is better. It is consistent with

23 precedent. It is what the agencies use. It is simple

24 to evaluate. Actually, the analysis I have done and

25 several other economists have done, it has shown that

Page 139

1 actually, even if you care about total welfare, for

2 several reasons, the consumer welfare standard actually

3 could lead to higher total welfare, and I will go into

4 that if that comes up later on.

5 Finally, I think the consumer welfare standard

6 supports innovation better than does the total welfare

7 standard. Again, I will go into that if there is time

8 later on.

9 So, thank you very much.

10 (Applause.)

11 MS. SCHULTHEISS: Okay, our next speaker is

12 Dr. Rick Warren-Boulton, a principal of MiCRA,

13 Microeconomic Consulting and Research Associates, an

14 economics consulting firm and research firm specializing

15 in antitrust litigation and regulatory matters. Before

16 joining MiCRA -ï½­

17 DR. WARREN-BOULTON: We like to call that

18 "MiCRA."

19 MS. SCHULTHEISS: I like to call it "MiCRA."

20 Rick Warren-Boulton served in a number of positions

21 involving economics and antitrust law. He was an

22 Associate Professor of Economics at Washington

23 University in St. Louis. He was the Chief Economist for

24 the Antitrust Division at the Department of Justice. He

25 was also a resident scholar at the American Enterprise

Page 140

1 Institute.

2 Rick has authored numerous publications -- you

3 laugh -ï½­

4 MR. KIRKWOOD: Are you still welcome there?

5 DR. WARREN-BOULTON: Yes, for lunch.

6 MS. SCHULTHEISS: Old friends, huh, Rick?

7 He has authored numerous publications, primarily

8 in the application of industrial organization economics

9 to antitrust and regulation. He's served as an expert

10 witness in numerous cases, including the Department of

11 Justice, U.S. versus AT&T, and for the FTC in our FTC

12 versus Staples and Office Depot merger. He also was an

13 expert with the Department of Justice in the United

14 States versus Microsoft.

15 In addition, Dr. Warren-Boulton was recently a

16 consultant in support of the Ross-Simmons side in its

17 effort to convince the Supreme Court to reject

18 Weyerhaeuser's petition for certiorari. Rick is yet

19 another one of our panelists who has some connection to

20 the Weyerhaeuser case.

21 Rick, with that...

22 DR. WARREN-BOULTON: Thank you. You know you

23 are getting old when you start talking about when you

24 were an expert witness in the AT&T case.

25 MS. SCHULTHEISS: And now they have all

Page 141

1 re-merged.

2 DR. WARREN-BOULTON: -- in 1981, and in terms of

3 my connection with this case, obviously I was -- I tried

4 to help Ross-Simmons in arguing to the DOJ and the FTC

5 that they should not file what they did, so it was a

6 complete failure, but I will just keep trying anyway,

7 because I might as well.

8 You know, the issue that I would like to talk

9 about here today is really a question as to whether or

10 not we need a test, or at the extreme, even a safe

11 harbor, and I think we need to distinguish between tests

12 and a safe harbor for cases of strictly what I would

13 call predatory overbidding to achieve monopsony power,

14 and I think what we are doing is, in the order of the

15 speakers here, we are going from general to specific.

16 You know, I think that the much more broader and

17 frankly more interesting question is exclusion, and I

18 think Tim is right when he says if you come away from

19 this with anything, it is, for heaven's sake, do not get

20 predatory pricing or predatory overbidding mixed up with

21 exclusion, and then I think Steve would say, do not get

22 it mixed up with raising rivals' costs, too, and I

23 agree, very much so, that those are two very important

24 questions.

25 I do not think that the analysis, nor the

Page 142

1 implications for policy, are the same for the three of

2 them, and I think what I am going to talk about today

3 is, in fact, the least interesting and least important

4 of the three, but that is what I thought my topic was,

5 and so that is what I want to talk about.

6 You know, I want to begin by distinguishing

7 between an economic definition of predation and a legal

8 test for predation, because coming up with a definition

9 for predation is extraordinarily easy. Any of us can

10 define predation. I was thinking about it before. I

11 think it is the reverse of pornography. We all know it.

12 We don't know it when we see it, but we know how to

13 define it.

14 You know, testing predation or coming up with a

15 test for predation is very difficult, because a test for

16 predation involves a balancing of the costs and

17 probability of false positives against the costs and

18 probability of a false negative, and so finding the

19 right tests or the right safe harbors for each

20 distinguishable situation is fundamentally an empirical

21 question, and it's a different question for every

22 situation that you can distinguish.

23 What I think for us at the moment to think about

24 it is is a couple of implications of that for coming up

25 with tests for predatory overbidding to achieve

Page 143

1 monopsony power. The first is, there is absolutely no

2 need to have the same rule for predatory pricing to

3 achieve monopoly power as predatory overbidding to

4 achieve monopsony power. You don't need to have the

5 same rule unless you can't tell the two situations

6 apart. But even a court, to be honest, it seems to me,

7 is unlikely to confuse predatory pricing with predatory

8 overbuying.

9 You know, when you have predatory pricing, you

10 look at the final product price, and it goes down, and

11 then it goes up, and when you look at predatory

12 overbidding, the initial price goes up, and then it goes

13 down. So, it seems to me that trying to order that you

14 need consistency between monopsony and monopoly

15 situations when what you are looking at is completely

16 different behavior seems to me to be unnecessary and

17 sort of pointless. It may not be the hobgoblin of

18 little minds, but there is absolutely no reason to say

19 that just because something is over here in the monopoly

20 world, that we should apply the same rule in a monopsony

21 world.

22 The second is, I think it is very easy to

23 distinguish between predatory overbidding and what I

24 think is going to be Janet's concern, which is between

25 predatory overbidding and either the exercise of

Page 144

1 monopsony power or, more importantly, the exercise of

2 bargaining power by power buyers. When you have

3 predatory overbuying, what happens is your input price

4 goes up, and then it goes down. That is what we are

5 looking at today. When you are looking at bargaining

6 power, in my experience, input prices go down and then

7 they stay down. You know, power buyers like Wal-Mart

8 pay lower prices, you know, from the get-go. Wal-Mart,

9 to my knowledge, has never been accused of paying too

10 high a price, even temporarily.

11 Once more, of course, is you can distinguish the

12 output effect, the classic monopsony problem, is you buy

13 too little. You know, Wal-Mart buys a lot. So, it

14 seems to me, you know, there should not be much concern

15 here about false positives when we are facing

16 allegations of predatory overbidding, and I hope that

17 Janet will disagree strenuously, and I am sure she will,

18 because otherwise, it will be no fun at all, but the

19 result of this, I think just as a sort of preamble, is I

20 find no urgent need for a safe harbor to deal with

21 allegations of predatory overbidding to gain monopsony

22 power. I do not think we should expect a flood of cases

23 that might be false positives, that if we do not

24 immediately nip this in the bud by allowing a

25 Brooke Group safe harbor for predatory overbidding, and

Page 145

1 I am certain Janet will disagree with that one.

2 The second point to me is there is no reason why

3 we can't proceed inductively, decide Ross-Simmons and

4 any of the other rare cases that arise on the merits,

5 and then try to generalize them. There are two

6 procedures here. Lawyers, in my experience, the idea is

7 that you have a whole bunch of cases, you try them on

8 the merits, and then you ask, "Gosh, is there some

9 common principle that's going on here?" It is very much

10 of an inductive reasoning process, and it works very

11 well.

12 I think the economist approach is highly

13 complementary. We go the reverse. We think

14 deductively. We think what are the first principles,

15 and we try to deduce, deductively, what the right

16 principles are, and, you know, with any luck, it is like

17 two people tunneling from opposite sides of the

18 mountain, you know, the lawyers going inductively and

19 the economists going deductively, and if all goes well,

20 we meet in the middle, and if we don't, we just keep

21 going until we get the other side, and we have two

22 tunnels.

23 But, you know, this strikes me as a situation in

24 which I think the lawyers -- I mean, look at how many

25 years we have been trying deductively to figure out what

Page 146

1 the right principle to use is on predatory pricing for,

2 you know, on the output side. We can't figure out the

3 damned thing. I mean, you know, look how far apart

4 those two gentlemen are, opposite sides of the table for

5 God's sake. They are never going to agree. So, I think

6 that the deductive approach really -- I mean, I think we

7 have plenty of time to get a few more Ross-Simmons cases

8 and then figure out how to go on.

9 But let us suppose -- you know, at this point, I

10 think I could sit down and say there is no need for me

11 to proceed further, but unfortunately, my panelists

12 won't let me do that, and anyway, I think I have another

13 six minutes. So, let's ask the question. Let's suppose

14 that you did do the unnecessary thing and you did ask,

15 well, let's ask, what kind of tests should we sort of

16 come up with for predatory overbidding to achieve

17 monopsony power? And the question I think you want to

18 ask is, what are the questions we want to ask?

19 And I think you really want to ask things like,

20 is monopsony different from monopoly? Are bidding

21 markets -- because remember, this was specifically a

22 bidding case -- are bidding markets different from, you

23 know, other markets? And I think that the first answer

24 is -- on both of them is yes. If you look at the

25 difference between monopsony and monopoly, I have a

Page 147

1 paper, which I am not posting on the web site until I

2 get a chance to sort of, you know, get all the

3 information, getting digs at Janet, so I will post it

4 later, but, you know, two of the main differences

5 between monopoly and monopsony it seems to me is

6 monopsony is much rarer, and it is easier to identify.

7 Monopsony is rarer for several reasons.

8 Basically supply inelasticity is much rarer than demand

9 elasticity. We all know this. Why is it? Well,

10 producers can substitute easier than consumers in

11 general, but most important, you know, in consumption,

12 diminishing returns, you know, the demand curve slopes

13 downwards, I mean, we all know that.

14 On the other hand, in production, the norm is

15 constant returns to scale, and so what we tend to get

16 is, you know, a situation in which -- I just -- you

17 know, I don't like -- I don't have a blackboard or a

18 white board, but if you were drawing a demand and supply

19 curve, you would feel perfectly normal drawing a

20 horizontal supply curve and a downward-sloping demand

21 curve, something which would look like this, but most of

22 you would think that what's somewhat less usual is to

23 have a completely horizontal demand curve, supply -- you

24 know, upward-sloping supply curve. I mean,

25 fundamentally, demand curves slope downwards. Supply

Page 148

1 curves usually are horizontal. So, the reason why we

2 get that is I think because there's a real difference.

3 There is diminishing returns of consumption and

4 relatively constant returns on production.

5 The second thing is I think monopsony is much

6 easier to identify, monopsony power, for several

7 reasons. One is that supply inelasticity is really only

8 observed in a few situations, and when I stop and I

9 think about what they are, they are most importantly, in

10 my experience, when a product is a by-product of some

11 other product or when it is an exhaustible natural

12 resource. Both of those are very unusual, and, you

13 know, both of them actually apply in this particular -ï½­

14 in the Ross-Simmons case. Alder turns out to be a

15 by-product or what they call a come-along product in the

16 industry, and it is an exhaustible product. So, you can

17 use the screen essentially of a by-product as a -- to

18 limit the scope of any decision.

19 By the way, there is a second reason, which I

20 always find kind of fun, which is why observing

21 monopsony is so much rarer than observing monopoly.

22 Monopsony is inefficient, and so nearly all monopsony

23 situations or potential monopsony situations are solved

24 by vertical integration, okay? If you are an aluminum

25 plant and you are right next to an electricity dam and

Page 149

1 you are the only buyer of electricity, you have

2 monopsony power towards that supplier, say, of

3 electricity, and exercising monopsony power can be

4 profitable. It is even more profitable, though, if you

5 buy the dam and get rid of the inefficiency.

6 So, vertical integration is profitable because

7 it gets rid of the inefficiency of monopsony, okay? But

8 vertical integration backwards is much, much easier than

9 vertical integration downwards and to the consumer, and

10 the nice test for this is if you ask where do we

11 actually see monopsony being exercised, in most cases,

12 it's situations where vertical integration is not

13 possible, and the classic example of that, at least

14 since the Civil War, is labor. You cannot buy people,

15 okay? And that is why the other day, you know, if you

16 were reading the New York Times, you would find that it

17 said a class action case, you know, on monopsony, who is

18 it? It is by a group of nurses who are being -- the

19 assertion is that they are being monopsonized. Now, if

20 the hospital could just buy the nurses, there would not

21 be a monopsony problem, okay? So, most monopsony cases

22 are in labor, because vertical integration does not

23 solve the problem.

24 All right, the second main difference that I

25 think we see between everything else we are talking

Page 150

1 about is that these are, of the cases that we are

2 talking about here, Ross-Simmons, is it is a bidding

3 market, and the question is, just like monopoly is

4 different from monopsony, bidding markets are different

5 from markets in which there simply is single price. The

6 classic problem with predatory pricing has always been,

7 and I think it was Steve who introduced this, that it is

8 inherently implausible, because the cost to the predator

9 with a high market share is so much greater than the

10 harm to the victim.

11 If you have a 90 percent market share and you

12 engage in predatory pricing, you bear 90 percent of the

13 costs, and the guy with the 10 percent share bears 10

14 percent of the costs, and that is so stupid, you know,

15 that it is fundamentally implausible. The guy does not

16 believe you are going to keep shooting yourself in the

17 foot over and over and over again. So, it is sort of

18 hard.

19 But, you know, when you can price-discriminate,

20 if the potential competitor can price-discriminate, if

21 prices are individually negotiated, then the cost to the

22 predator is no longer linked to the market share. There

23 is no longer any connection between the relative cost of

24 the predator and the predatee, you know, and then market

25 shares, and so you have a fundamentally different

Page 151

1 probability, you know, of a false positive here.

2 So, you know, my conclusion, looking at this, is

3 that if I am looking at the specific problem that we are

4 dealing with here, which is predatory overbidding to

5 achieve monopsony power, that if you have a combination

6 of three things, you have a very good probability of a

7 real case being there and a very low probability of

8 having a false positive. Those three things are a very,

9 very low supply elasticity of the input, which as I say

10 is almost always just when it is either a by-product or

11 a natural resource; secondly, the ability of the`

12 predator to price-discriminate, particularly in bidding

13 markets; and third, very strong barriers to re-entry.

14 If you have those three things, then I think you

15 have basically passed sort of the structural criteria

16 that you need to pass. You don't pass -- you know, I

17 think you have to pass all of those three, but if you

18 pass those three, then I think you have a very good

19 argument here that the predatory overbidding to gain

20 monopsony power is, in fact, a realistic effort.

21 Recoupment, what I am basically saying, is very

22 likely, and then you do not need to get into the kind of

23 cost tests that are in Brooke Group that I think most of

24 us have looked at and said, the problem with these cost

25 tests is that they generate a very -- particularly if

Page 152

1 you allow them as safe harbors, they allow a high rate

2 of false negatives. If you look at Ross-Simmons, it

3 fits all of those, alder is a by-product, prices are

4 individually negotiated, and the equipment that the

5 sawmills was using is actually specific to alder, okay?

6 If Ross-Simmons' equipment could have been used either

7 for alder or soft wood, then we would not be here. Then

8 basically, you know, Ross-Simmons could have been a

9 hit-and-run entrant. They could have come in, they

10 could have gone on, but the specificity of the equipment

11 means basically that there is a very high barrier to

12 re-entry.

13 If you look at those three criteria, I think it

14 merits what the Ninth Circuit sort of was saying. It

15 was saying that predatory overbidding for a resource

16 input in a highly inelastic supply, combined with a high

17 barrier to entry by the downstream firms, you know, that

18 is an extraordinarily narrow set of events, and if you

19 look at that and you combine that with a

20 no-economic-sense test, to understand the strategy and

21 say does it pass the economic sense, I do not think

22 there is any need for going into, you know, cost-based

23 tests. I think what we could basically do is accept the

24 Ninth Circuit decision, construe Ross-Simmons narrowly,

25 and we can all wait for the next 20 or 30 years to see

Page 153

1 if another case ever turns up.

2 You know, finally, in terms of the two gentlemen

3 over here, you know, I think we owe an enormous debt of

4 gratitude to what used to be called the Chicago School,

5 but I guess in this case more the University of Virginia

6 school, for 40 years of reducing false positives, and I

7 think that is yeoman's work at the academic level. I

8 would like to believe that the Reagan Administration and

9 Bill Baxter and the people who came in with him were at

10 least sort of implementing that, but I think one thing

11 we should realize is that getting rid of false negatives

12 is a complement to reducing false positives.

13 We all know on deterrence that optimal penalties

14 are a decreasing function of the probability of a false

15 positive, and for those of you who do not quite

16 immediately know what that means, what it means

17 basically is now that we have DNA testing, the case for

18 capital punishment is much stronger. That is the

19 easiest way to think of it.

20 So, what has happened is we have an accumulated

21 experience, we have many cases of alleged predation, we

22 have economists working on it, and I think we have

23 greatly reduced the level of false positives, and I

24 think the problem is, of course, is once you have

25 reduced the probability of false positives

Page 154

1 significantly, then it becomes efficient and desirable

2 to turn to the question of can we reduce the number of

3 false negatives. So, I think what we are seeing over

4 here, and I know Professor Elzinga is going to shoot me

5 for this, but I would say Professor Elzinga has made

6 Professor Bolton possible. I leave the two of you to

7 duke it out.

8 (Applause.)

9 MS. SCHULTHEISS: Our final speaker is Janet

10 McDavid. Janet McDavid is a partner in the Washington,

11 D.C. office of the law firm Hogan & Hartson, where she

12 focuses on antitrust and trade regulation, litigation

13 and counseling. She is widely recognized as a leading

14 authority in antitrust law, has been included in many

15 guides to top antitrust lawyers, and an author of many

16 books and articles on antitrust law.

17 Ms. McDavid was previously the Chair of the

18 Section of Antitrust Law of the American Bar Association

19 and also a chair of its committee on Section 2 of the

20 Sherman Act. As a member of the Antitrust Council to

21 the U.S. Chamber of Commerce, Ms. McDavid also brings to

22 us a business perspective.

23 In addition, of particular relevance, again, to

24 today's hearing, Ms. McDavid was a co-author of the

25 brief for the Business Round Table and National

Page 155

1 Association of Manufacturers in support of

2 Weyerhaeuser's petition for certiorari, and Ms. McDavid

3 was also the lead author of a recent article in the

4 National Law Journal entitled "Predatory Purchasing?"

5 And at this, I give Janet the job of cleanup and

6 responding to some of the comments that have been made

7 thus far.

8 MS. McDAVID: I don't think my views will be

9 quite as extreme as Rick may have suggested. My

10 perspective here is that of a practicing lawyer who has

11 to try to advise clients on where the line is between

12 conduct that might be unlawful under the antitrust laws

13 and conduct in which they can engage with relatively low

14 risk, and Rick has suggested that greater clarity in

15 this area really is not necessary, we should just allow

16 the law to develop for a while and see whether

17 inductively or deductively we can establish some rules,

18 because the cases are rare.

19 I am not seeing dozens of these cases in my

20 practice, but I am increasingly seeing a lot of

21 monopsony questions coming up from my clients, and so I

22 do not think these cases are likely to be as rare as

23 perhaps has been hypothesized. I am hoping that these

24 hearings or the Weyerhaeuser case in the Supreme Court,

25 whichever comes out first, can provide an answer with

Page 156

1 respect to the predatory purchasing conundrum that can

2 reasonably be applied by businesspeople and their

3 lawyers.

4 Remember, the purchasing department of even a

5 Fortune 100 company does not include these guys. They

6 would not know a downward sloping demand curve if it hit

7 them in the face. And so the notion that that is a

8 standard that can be applied by a businessperson in a

9 purchasing department is just unrealistic.

10 It is actually quite funny, I think, that Rick

11 referred to Justice Stewart's definition of pornography,

12 because that is also the analogy that I had here. I

13 think the -ï½­

14 DR. WARREN-BOULTON: Like minds.

15 MS. McDAVID: -- Ninth Circuit standard in

16 Weyerhaeuser provides much less guidance than Justice

17 Stewart's standard with respect to pornography. There

18 is no basis here for advising a client, and this is a

19 serious mistake, because ultimately, businesspeople have

20 to understand the rules. Justice Breyer, when he was

21 still on the First Circuit, said that antitrust rules

22 must be clear enough for lawyers to explain them to

23 clients and be administratively workable, and that in

24 formulating antitrust liability standards, the courts

25 must consider what advice the lawyer is going to give.

Page 157

1 Chairman Majoras made essentially the same point

2 when she opened these hearings on Tuesday. The process

3 of distinguishing between the permissible and the

4 impermissible must be relatively consistent and

5 transparent so firms can incorporate it into their

6 decision-making.

7 So, where does that leave us with the Ninth

8 Circuit? We have a standard of liability that allowed a

9 jury, using 20/20 hindsight, to determine whether

10 Weyerhaeuser paid a higher price than necessary to

11 prevent Ross-Simmons from obtaining its inputs at a fair

12 price and that Weyerhaeuser may have purchased more logs

13 than necessary. How is a business person supposed to

14 apply that standard? It is entirely vague, open-ended,

15 and subjective. It gives us no way to draw the lines,

16 and a jury exercising its 20/20 hindsight can come out

17 in a completely different place than a perfectly

18 rational businessperson did.

19 I think it is significant that Weyerhaeuser

20 never lost any money on any of this. The division was

21 operating profitably the whole time. I reread the Ninth

22 Circuit decision this morning, and the Court talks about

23 the need for recoupment in order to render the strategy

24 profitable, but it does not actually impose that

25 requirement as part of the test. So, the Court gave lip

Page 158

1 service to recoupment without ever requiring it as part

2 of the standard.

3 So, one of the effects of all of this is to prop

4 up less efficient firms, such as Ross-Simmons, that are

5 unable to operate profitably at the same input prices

6 that a more efficient rival can afford to pay. Now,

7 firms compete for inputs just as they compete for sales,

8 especially if the inputs are scarce, and there is

9 language in the Ninth Circuit's decision that suggests

10 that the inputs were becoming more scarce. They

11 certainly were not more plentiful. So, we have a

12 circumstance in which firms are going to be deterred

13 from aggressive buying by the threat of liability for

14 treble damages as a consequence of a standard that they

15 cannot understand and cannot apply.

16 How is a firm supposed to know if a jury will

17 later determine whether it bought more than it needed,

18 whether it paid a price that was too high, whether it

19 paid a price that was not fair? If it buys too much,

20 there are lots of reasons that that could have happened,

21 as Steve was explaining earlier. It might have decided

22 to stockpile inventory to preclude future shortages or

23 to hedge against a future price increase. That happens.

24 We had it happen, for example, during the

25 Hurricane Katrina circumstance. Oil companies had

Page 159

1 inventories. They were able to sell it off during a

2 time of shortage. They might have overestimated their

3 needs. Businesses make mistakes. They might have

4 assumed that demand was going to increase, and it did

5 not. Again, businesses make mistakes. They may have

6 planned that sales were going to grow and guessed wrong,

7 or they may have chosen to deal with the predictable

8 supplier who has a reliable source and paid a slight

9 premium to do so.

10 Any one of these could, in hindsight, become the

11 basis for liability by a jury that is exercising a

12 standard based on fairness. Under no other circumstance

13 in the antitrust laws, except perhaps the regrettable

14 Robinson-Patman Act, do we consider a fairness standard.

15 It is simply not administrable. So, people like me will

16 have a hard time telling our clients what they should

17 do, and the businesspeople on the ground will have a

18 very hard time knowing what they should do and what they

19 can do safely and fairly.

20 Now, I will recommend to my clients, as I do in

21 all circumstances, that they document the reasons they

22 are doing things if they face a risk. I always tell

23 them that I would prefer to create our own legislative

24 history, that in the event the inevitable happens, I

25 would like there to be evidence in their files that

Page 160

1 explains, without us having to go back and explain it

2 again, why they did what they did and that there was a

3 rational business reason for what they are doing. I

4 will also ask them, can you make money if you purchase

5 at X price? Now, the unfortunate thing is, while I

6 think that is a perfectly rational question to be

7 asking, it is not relevant to the analysis in the Ninth

8 Circuit, whether or not they were going to be able to

9 make money if they paid X price for their inputs.

10 I do not understand why the Brooke Group

11 standards should not be at least relevant to the

12 analysis here. These are standards that business people

13 actually can understand. Very few of my clients have

14 ever known what their average variable cost is, but they

15 do know, if we talk about it in kind of gross terms, if

16 you pay X or if you charge X, will you make money or

17 will you be operating at a loss?

18 They also can understand the notion of whether

19 or not it will be profitable in the longer term. These

20 are standards businesspeople can understand, and the

21 courts have got to give them some kind of guidance here,

22 and if the courts fail to do so, I hope that these

23 hearings will, because without it, the business people

24 are going to be left largely rudderless.

25 MS. SCHULTHEISS: Thank you, Janet.

Page 161

1 (Applause.)

2 MS. SCHULTHEISS: I would like to thank all of

3 our panelists, and we are going to take a brief

4 ten-minute break, after which we will reconvene and have

5 a directed discussion with some questions. Thank you.

6 (A brief recess was taken.)

7 MS. SCHULTHEISS: Okay, let's get started with

8 the second part of our panel, which is the discussion

9 period, and before I start asking some specific

10 questions of the panelists, I would like to see if I can

11 get some consensus from the panelists on some of the

12 terms we are using and whether we are understanding them

13 correctly.

14 As I am hearing the presentations, I hear that

15 you are talking about two different things. You are

16 talking about predatory conduct that affects the input

17 suppliers and the price of the input supplies, and

18 another area is the predation or predatory conduct that

19 is directed at affecting your competitors in the output

20 market. Is that correct?

21 Let's just start with Jack and let's just go

22 through the panelists and see what you think. Is that a

23 correct statement?

24 MR. KIRKWOOD: I am not sure I would put it

25 exactly that way. I think that Steve's distinction is a

Page 162

1 good one, that if you as the dominant firm could raise

2 the input costs of a rival, and the question is whether

3 your ultimate goal is monopsony power or, instead,

4 downstream market power, and the first I have called

5 predatory bidding, as did the Ninth Circuit, Steve has

6 called it predatory overbuying. The second category, it

7 could come from predatory bidding, as I explained, but

8 the second category is more typically described as

9 raising rivals' costs, or as Tim would say, exclusion,

10 and maybe Tim could address the question of to what

11 extent is exclusion different from raising rivals'

12 costs.


14 DR. BRENNAN: Well, I agree with what Jack said.

15 Again, I agree with the distinction as he put it or, you

16 know, that Steve made. I would not call both predatory,

17 because as soon as you do that, then you start using

18 predatory tests in both contexts, and I think for

19 reasons that I have argued and Steve has suggested as

20 well and others, that I think those are inappropriate.

21 The main reason I make the distinction, I mean,

22 I suppose there is a bit of a long story here, but just

23 to keep it short, the two are -- well, first, when you

24 say that the harm from something involves hurting

25 rivals, then I think you invite people to go back into

Page 163

1 predation land, and I would rather call it something

2 else, and the reason I would call it the something else,

3 when I am calling it complement market monopolization,

4 is that in order to exclude people, in order to raise

5 their costs, whatever it is, you have to create and

6 exercise market power over something that they need, and

7 because antitrust authorities have ways of thinking

8 about that sort of thing, that is what we do with

9 mergers, that is what we do with cartels, you could in

10 some sense take an enormous part of -- as Steve, I

11 think, aptly put it -- the great lion's share of Section

12 2 cases and get them out from under the controversy that

13 attaches to them because of the view that what Section 2

14 is about is in some sense about competing so hard that

15 people get hurt too much.

16 MS. SCHULTHEISS: Steve, would you like to

17 address that?

18 DR. SALOP: Yes, sure, I agree with Jack that I

19 got it right.

20 DR. BRENNAN: I figured it was relatively safe.

21 DR. SALOP: I agree with the first two sentences

22 of what Tim said. I think when you use the term

23 "predatory," it is a loaded term. I never really liked

24 the term "nonprice predation" for that reason. That is

25 why I like calling it exclusion or raising rivals'

Page 164

1 costs.

2 Now, nonprice predation sometimes involves

3 prices, like here, prices are involved, and when you

4 call it predation, that tends to plug into Brooke Group,

5 and as I said, I think there are two vastly different

6 paradigms.

7 MS. SCHULTHEISS: So, in terms of what we are

8 speaking about here today, Rick focused on the narrow

9 with the predatory overbidding or overbuying that is in

10 the Weyerhaeuser case, but then there is a separate area

11 of conduct that involves some type of exclusionary

12 conduct, what you call raising rivals' costs, what Tim

13 might call exclusion, and looking at the complementary

14 market.

15 DR. SALOP: I think what all three of us were

16 taking issue with in your question was the word

17 "predatory." Had you just said "conduct that gives

18 market power in the input market" and "conduct that

19 gives market power in the output market," we all would

20 have -ï½­

21 MS. SCHULTHEISS: Do you all agree with that

22 one?

23 DR. BRENNAN: More or less.

24 MS. SCHULTHEISS: What about you, Rick?

25 DR. WARREN-BOULTON: First of all, yes, I think

Page 165

1 they are enormously different. I think that the first

2 is monopsony and the second is monopoly, and I think it

3 is incumbent upon a plaintiff to say which one -- which

4 world he's in. I agree that the false positive/false

5 negative trade-off is enormously different between the

6 two. I don't see how people can really -- or people

7 should not confuse them. I don't think they are

8 particularly relevant. I think all three of those

9 situations should be handled, you know, quite

10 separately, because they have different structure

11 requirements.

12 The final question, would I distinguish them in

13 terms of predation/nonpredation, no, I would distinguish

14 them I think in terms of, you know, simple pricing

15 versus, you know, more complex exclusion, you know,

16 price or nonprice. I mean, most of the really

17 interesting problems here really are exclusion done in

18 interesting and complicated and strategic ways and how

19 do we handle that. But again, I think at the very

20 beginning, it is really incumbent on people to say which

21 box they are in in any case.


23 MS. McDAVID: I think that Steve and Jack got it

24 right from my perspective.

25 MS. SCHULTHEISS: Okay. In terms of the

Page 166

1 antitrust agencies, would you agree, then, that we

2 should be more concerned with the latter category, that

3 being, if you want to call it, either the exclusionary

4 conduct or raising rivals' costs, versus the pure

5 predatory buying towards, you know, monopsony over the

6 input market? And I would wonder -- I want to find out

7 if you agree with that.

8 MR. KIRKWOOD: I would be inclined to agree with

9 it, yes. I have not done the kind of research that

10 Steve has done into the frequency of raising rivals'

11 costs problems, but from what I have read of what he and

12 others have written, it seems to be a more common

13 problem. One of the key elements of the analysis of

14 predatory bidding we have done is that it seems to be

15 quite rare. So, yes, in terms of where you would target

16 your enforcement resources, sure, surely.


18 DR. BRENNAN: Yes, mostly for the reasons Steve

19 said before, that the predation parts of things, whether

20 it is buying or selling, are going to be rare for the

21 reasons he outlined and I do not need to repeat, and so

22 for that reason my expectation would be to worry about

23 things where people do not have to sacrifice a lot and

24 take a lot of risks in order to get some speculative

25 benefit down the road.

Page 167

1 MS. SCHULTHEISS: Steve, I take it you agree.

2 DR. SALOP: Yes, I agree with what they said,

3 but they didn't answer your question, which is, well,

4 what the agency should not be doing, and frankly -ï½­

5 MS. SCHULTHEISS: I do not know if it is

6 necessarily what we should or should not, but our

7 primary focus. Should we be worried about one more than

8 the other?

9 DR. SALOP: The way you phrased it is something

10 that I think is all too apparent in this administration,

11 which is you are spending an awful lot of time deciding

12 what you should not do, and you are not doing a heck of

13 a lot, and so I think that in the exclusionary conduct

14 area, the agencies ought to get involved and start

15 looking for exclusionary conduct cases rather than

16 protecting monopolists.


18 DR. WARREN-BOULTON: I think I am probably

19 agreeing with everybody. First of all, I agree that -ï½­

20 I mean, the whole point is that, you know, predatory

21 overbidding to get monopsony power, as I said, is

22 incredibly rare. So, the answer to the first question

23 is yes, and oddly enough, I would agree with Steve that

24 I think exclusion is a real problem, and so if you are

25 going to spend more money on something, I would have you

Page 168

1 spend it on exclusion.


3 MS. McDAVID: I would urge the agencies actually

4 to play a role in the Section 2 enforcement area

5 generally, because too often these cases arise in the

6 context of disputes between business rivals, each of

7 whom brings baggage, whereas the agencies are trying to

8 do the right thing, and a case that an agency brings has

9 been vetted carefully as opposed to just being the

10 pissing contest between somebody who may have been

11 forced out of business and thinks that, of course, it

12 was unfair and something must have been wrong, it

13 couldn't have been an inefficient firm, it couldn't have

14 been incompetent. The agencies, I think, bring an

15 important balancing rule to the enforcement in the

16 Section 2 area generally and in this area as well.

17 MS. SCHULTHEISS: Let me ask you another

18 question, Janet, because you were talking about in terms

19 of counseling clients and what have you. Would you

20 agree, then, with Rick and most of the other panelists,

21 I think, that the overbidding or overbuying that we see

22 in the Weyerhaeuser case is an unusual situation?

23 MS. McDAVID: I have not seen a lot of this. I

24 think it probably is an unusual situation, but the

25 plaintiff's bar is extremely entrepreneurial, and I do

Page 169

1 not think we should assume that it will continue to be

2 an unusual circumstance.

3 MS. SCHULTHEISS: So, you -ï½­

4 MS. McDAVID: And virtually every major firm in

5 this country sues in the Ninth Circuit. So, unless the

6 law gets clarified, I think we will see these cases

7 because they will get syndicated.

8 MS. SCHULTHEISS: Steve, did you want to say

9 something in response to that?

10 DR. SALOP: Yes, I just wanted to -- I just

11 jotted down some overbuying cases before, and so I know

12 we see sort of the classic, you know, Ross-Simmons cases

13 being rare, but one, as I said before, Ross-Simmons

14 alleged raising rivals' costs as well as -- as well

15 as -ï½­


17 DR. SALOP: -- predatory bidding, and in terms

18 of the history of antitrust, you know, among sort of the

19 cases I teach and run across, we have got not just the

20 timber cases, we have got the tobacco case was an

21 overbuying case, the beef case, the Monfort case that -ï½­

22 I mean, the beef case was a conspiracy in overbuying,

23 the Monfort case was about overbuying, went to the

24 Supreme Court, Socony-Vacuum was about overbuying, Alcoa

25 had a piece that was involved in overbuying, and in

Page 170

1 terms of what is going on right now, stuff the agency

2 has in its shop, all the stuff about shelf space is all

3 about overbuying, too, you know, could be thought of as

4 an overbuying issue. There could be overbuying of

5 patents. Alcoa, you know, allegedly tied up some

6 patents, so it made it harder for people in -- not sort

7 of classic predatory bidding, but, you know, you could

8 certainly have overbuying of patents, too.

9 So, I do not think it is a -- I think this idea

10 that it is just exhaustible resources that are

11 by-products, I know that covers the old molybdenum case,

12 one of my favorite cases when I was at the FTC, but I

13 think it is rather broader than that.

14 DR. WARREN-BOULTON: Can I just make a -- I

15 mean, I have not gone back and looked at the cases, but

16 I would question how many cases are there -ï½­

17 MS. SCHULTHEISS: Can you speak into the

18 microphone?

19 DR. WARREN-BOULTON: I'm sorry, how many cases

20 are there in overbidding to gain monopsony power. Steve

21 cites a tobacco case. I do not think that was to gain

22 monopsony power for tobacco growers. I thought that was

23 to exclude other, you know, cigarette producers. The

24 beef case, I do not think they ever managed to -- it was

25 ever shown that they ever managed to raise the price of

Page 171

1 anything. The others I cannot remember, but, I mean, I

2 guess the question is, are there any cases other than

3 Reid, which was I don't know how long ago, and

4 Ross-Simmons that are predatory overbidding to achieve

5 monopsony power? I mean, I do not want to say sui

6 generis, because I do not know how to pronounce it

7 correctly, but it has got to be something like that.

8 Sui?

9 MS. McDAVID: Sui.

10 DR. WARREN-BOULTON: Sue me, sue you.

11 MS. SCHULTHEISS: All right, Jack would like to

12 say something.

13 MR. KIRKWOOD: Yes, Steve is certainly right

14 that there have been a number of overbuying allegations,

15 but if you look narrowly at how many of these cases are

16 predatory bidding cases, American Tobacco, as Rick and I

17 discussed last night, that seems fairly clearly a

18 raising rivals' costs case as opposed to a predatory

19 bidding case, because the kind of tobacco the major

20 producers bid up was a tobacco they did not use at all

21 and continued not to use it. So, they were not trying

22 to get monopsony power in it, because they were not

23 using it as an input.

24 And in beef, though there was a predatory

25 bidding allegation, as Rick suggests, they were not able

Page 172

1 to raise the price at all. And in Socony, that was

2 behavior designed to facilitate a cartel. Monfort had a

3 predatory bidding concern, but it was a concern raised

4 as part of an effort to enjoin a merger. So, there was

5 not any actual evidence of predatory bidding, and the

6 Court analyzed it as a predatory pricing case.

7 MS. SCHULTHEISS: Steve, and then I want to get

8 Janet's reaction.

9 DR. SALOP: Of course, the Supreme Court

10 misanalyzed Monfort, but putting that aside, I mean, I

11 agree, a lot of these are not predatory overbuying

12 cases, they are other things, but the Ross-Simmons jury

13 instruction does not require you to show anything other

14 than they paid more than necessary or bought more than

15 necessary or paid more than a fair price. So, all of

16 them would be swept in if we go inductively and just

17 allow the Ross-Simmons jury instruction to stand and

18 screw things up for another 15 years until Ken has to

19 get involved and clean up the area.

20 MS. McDAVID: That is precisely my concern, is

21 that if this is the law in the Ninth Circuit, virtually

22 anything can be brought as one of these cases, and they

23 probably will not be tried to judgment in the end,

24 because the defendants cannot take the risk of trying

25 those cases to judgment with the treble damages burden.

Page 173

1 MS. SCHULTHEISS: I am going to let Tim have the

2 last word on this particular point, and then I would

3 like to move on.

4 DR. BRENNAN: I think that the reasons that Rick

5 in particular elaborated on why these cases would be

6 rare in terms of the circumstances of the market,

7 exhaustible resources, the bidding market and stuff, I

8 think those are all important to keep in mind. I do not

9 think, though -- one should be careful about making the

10 distinction, one, between buying and selling, because

11 that can be just an arbitrary function on the nature of

12 the market.

13 Let me give you an example, which is pipelines.

14 I don't know as much -- I don't know as much now about

15 how the pipeline sector works as I used to, but

16 pipelines are kind of funny in the following sense: The

17 way oil pipelines worked is that they sold oil pipeline

18 delivery services to oil wells basically. The way the

19 gas pipeline industry worked was that they bought gas

20 from gas wells and then resold it at the end of the

21 pipeline. If the oil pipeline was exercising market

22 power against the oil wells, it would involve monopoly,

23 that they were raising the price of pipeline services.

24 If they were -- if the gas pipeline was exercising

25 market power against the gas wells, it would be

Page 174

1 monopsony. They would be driving down the price they

2 paid for the natural gas.

3 The triangles, the welfare losses, the dead

4 weight stuff, all that stuff would be exactly the same

5 in both cases, at least qualitatively it would be the

6 same, but one's monopoly and one's monopsony, and so

7 it's purely in some sense a -- I don't know, I won't say

8 it's arbitrary or artificial, but it is the way those

9 markets just happen to work. So, one needs to be

10 careful about making a distinction in that way, that one

11 can turn monopoly into monopsony in some sectors without

12 a great deal of difficulty.

13 DR. WARREN-BOULTON: Yes, I -ï½­

14 MS. SCHULTHEISS: Rick is going to have to get

15 the last word I see.

16 DR. WARREN-BOULTON: -- I very much agree. You

17 can be regarded as buying a service or selling, and it

18 can be based on form. I think one implication that I

19 think is kind of interesting is that in terms of a

20 welfare loss, you know, the idea that somehow welfare

21 losses are only important if they happen downstream

22 rather than upstream strikes me as kind of bizarre.

23 MS. SCHULTHEISS: Let's get to that in a second,

24 because I want to get to that issue.

25 DR. WARREN-BOULTON: But that is one of the

Page 175

1 reasons why, because we can characterize it as upstream

2 or downstream sort of at will.

3 MS. SCHULTHEISS: Okay. Does everybody agree

4 that the jury instruction in Weyerhaeuser was just

5 wrong?

6 MS. McDAVID: Yes.

7 MS. SCHULTHEISS: Is there anybody that

8 disagrees with that?

9 DR. WARREN-BOULTON: You know, I am an

10 economist, I am not an attorney. If I was giving a set

11 of jury instructions to a group of economists, it would

12 be quite different from those instructions.

13 MS. McDAVID: How about if they were grocers and

14 guys who pump gas, because that is who your jury is made

15 up of.

16 DR. WARREN-BOULTON: So if I had a set of

17 instructions that said something like if you were a

18 group of economists and I wanted to have a recoupment,

19 how would I say that in a way that a group of grocers

20 would understand? You know, maybe this is the best way

21 the Court could think of to explain it to a group of

22 grocers. I'm not sure it would have been a better

23 result if we would have explained it in sort of

24 mathematical formulas.

25 MS. McDAVID: It probably would have been worse.

Page 176

1 DR. WARREN-BOULTON: The question is, what does

2 the audience grasp?

3 DR. SALOP: You think saying they bought more

4 than necessary is a good proxy for recoupment? Is that

5 what you just said?

6 DR. WARREN-BOULTON: No, paying more than

7 necessary, it seems to me to be to a -- I don't know, an

8 ordinary person's idea of was there a profit sacrifice

9 here, all right, which is at least one element.

10 MS. SCHULTHEISS: But in terms of precedent, how

11 can that possibly -- you are just agreeing with the

12 statement I just said, then. You would not have that

13 kind of an objection to the standard that was set out in

14 the jury instruction.

15 DR. WARREN-BOULTON: You know, I am saying

16 anybody here, I think anybody here could probably today

17 come up with a better set of jury instructions than

18 that. How bad it is, I do not know.

19 MS. SCHULTHEISS: I do want -ï½­

20 DR. SALOP: Do you think a reasonable jury could

21 reach the right conclusion if they tried to apply that

22 jury instruction?

23 DR. WARREN-BOULTON: You know, actually, I have

24 been on a jury, and I have to tell you that my personal

25 experience is that what a jury decides has little, if

Page 177

1 nothing, to do with the jury instructions. I was on a

2 cocaine bust jury in the District of Columbia, and we

3 convicted him of carrying a machete. We did that

4 because two people in the room said under no conditions

5 would I convict a male in the District of Columbia

6 carrying cocaine. We decided that he should go to jail

7 for three months, and we cast around to try to think of

8 what kind of thing would get him to jail for three

9 months. So, we convicted him of carrying a machete,

10 so...

11 MS. SCHULTHEISS: All right, I get the point.

12 There have been some different standards put

13 out, and I would like to get some input, because it

14 appears -- I mean, we have talked about consumer

15 welfare, we have talked about total welfare, no economic

16 sense, and they have come into play in different ways,

17 and I would like to start with you, Janet.

18 Which of the standards do you think makes most

19 sense in this type of a case, in the case where you are

20 dealing with either overbidding or some kind of

21 overbuying situation?

22 MS. McDAVID: Well, I am taken with Steve's

23 standard, which is effectively a rule of reason analysis

24 applied to the sort of circumstances we have here and

25 which incorporates as part of the analysis the Brooke

Page 178

1 Group test, because I think it gives us some grounding.

2 MS. SCHULTHEISS: In dealing with the Brooke

3 Group test, though, then, do you have to have power then

4 in the output market in order to get that recoupment?

5 MS. McDAVID: You have to be able to make it up

6 somewhere, and exactly where you are going to make it up

7 might vary based on the particular circumstance and as

8 to whether it's a raising rivals' cost case or a

9 predation sort of case.

10 MS. SCHULTHEISS: But if you are recouping

11 solely on the input side, on the supply-side, then you

12 really do not necessarily have the consumer welfare

13 harm.

14 MS. McDAVID: You may not.

15 MS. SCHULTHEISS: And in that case, would you

16 find that there would not be a violation or there should

17 not be an antitrust violation found?

18 MS. McDAVID: I am actually not terribly

19 troubled by also applying the antitrust laws and

20 allowing the victim to be the supplier, just as we do in

21 the cases Gail used to investigate when she headed the

22 health care shop where we looked at large insurance

23 companies and whether they were going to acquire

24 sufficient power over doctors and hospitals. I think it

25 is a reasonable inquiry.

Page 179

1 MS. SCHULTHEISS: So, the consumer welfare test

2 should not necessarily be the end all and be all, okay.

3 MS. McDAVID: Not the exclusive. It should be

4 the major concern, but it would not exclude entirely

5 concern to suppliers.

6 MS. SCHULTHEISS: Okay, Rick?

7 DR. WARREN-BOULTON: Yes, I just categorically

8 disagree strenuously. I mean, I just do not see any

9 difference between a dollar in producer surplus that

10 goes to some guy who is growing timber and a dollar to a

11 guy who is buying a unit of lumber, nor do the Merger

12 Guidelines. I mean, I think the economics literature is

13 absolutely unambiguous on this. The welfare losses from

14 monopsony are, you know, the same as the welfare losses

15 from monopoly, consumer surplus.

16 I mean, I think Roger Noll wrote a very nice

17 article right before everybody else's article that

18 basically walked through that, you know, and as Tim

19 points out, is that, you know, trying to find some

20 distinction as to whether or not it is upstream versus

21 downstream is completely arbitrary. It certainly cannot

22 be a distinction that says, "Gosh, if it is a loss by

23 producer, it does not count."

24 In that case, what you say is all input

25 monopolies are legal? I mean, you do not care if it is

Page 180

1 an input monopoly because, you know, the thing is bought

2 by a firm who does not necessarily pass it on? I mean,

3 the idea of restricting this to consumer surplus to me

4 is bizarre. I think basically you have got an

5 externality test, basically says I have got the bad guy

6 who is doing something, and the question is, is he

7 hurting the rest of society? And the rest of society

8 are the people he sells to and the people he buys from,

9 and those people stand, you know, on the same footing.

10 So, you know, I think what we used to call a

11 consumer welfare test when we were trying to read the

12 names back in the good old days is the sum of consumer

13 surplus and producer surplus. I think what most of us

14 say is you ignore the profits of the monopolist, but

15 that is it. Everybody else is in.

16 MS. SCHULTHEISS: Steve, is that what you mean

17 when you say consumer welfare or consumer harm?

18 DR. SALOP: No, I thought my slides were very

19 clear on that, and I have got actually a paper that I

20 submitted to the Antitrust Modernization Commission on

21 this issue. I think by consumer welfare I mean true

22 consumer welfare. I think that people who want to say

23 suppliers, losses to supplier welfare should be enough,

24 should consider whether they think harm to competitors

25 should be enough to carry an antitrust violation, and

Page 181

1 the Supreme Court has made it very clear that harm to

2 competitors is not enough, and I think the same thing

3 should be true with harm to suppliers.

4 The tricky part of this -- and I am sorry, this

5 is kind of a long answer -- the tricky part is that it

6 is quite clear to make an agreement -ï½­

7 MS. SCHULTHEISS: Could you speak more into the

8 mic? I'm sorry.

9 DR. SALOP: I'm sorry.

10 It's quite clear that naked agreements among

11 competitors, buyer-side competitors, to fix prices that

12 they make to inputs is illegal and should be per se

13 illegal, and -ï½­

14 MS. SCHULTHEISS: Well, that is horizontal,

15 though.

16 DR. SALOP: But that is where antitrust starts.


18 DR. SALOP: So, I think it is worth thinking

19 about that case, because a lot of people that I have

20 talked to when I say it should be about consumer harm,

21 not about supplier harm, they pretty quickly think about

22 the buyer cartel cases, and the way I would distinguish

23 is even in the buyer cartel side in the following way:

24 Suppose you have an agreement among competitors

25 to jointly set -- I will not use the loaded term

Page 182

1 "fix" -- to jointly set the price that they pay for

2 inputs. I think if that is naked, it is quite clear

3 that is per se illegal and should be per se illegal, but

4 to the extent that they have a justification, a

5 procompetitive, i.e., pro-consumer welfare justification

6 for that, then it ought to -- and I think it does -- go

7 into the rule of reason, just like in VMI.

8 If a group of sellers jointly sets a price, it

9 goes into a rule of reason if they have a procompetitive

10 justification for their actions, and then when you get

11 into the rule of reason, I think it makes sense that

12 consumer welfare rules, that it is not enough -- once

13 they show that consumer benefit, then the burden would

14 go to the plaintiff to prove that consumers are harmed.

15 It is not enough to show suppliers are harmed, and it is

16 not, I do not think, a balancing between the losses to

17 the input suppliers versus the effect on consumers and

18 versus the gains to the buyers that engage in the joint

19 price setting.

20 So, you know, I think antitrust is a consumer

21 welfare prescription, it is about consumer welfare, and

22 we should stick with that. We should stick with that

23 here, particularly in a situation where -- you know,

24 with this overbuying, where in a standard case,

25 consumers gain in the short run from it, just as they

Page 183

1 gain in the short run during the predatory period from

2 predatory pricing.


4 DR. BRENNAN: A few very quick observations.

5 First, I just happened to go to the AMC deliberations on

6 mergers last week, and the only thing -- almost the only

7 thing they argued about was the welfare standard, and

8 the only arguments in favor of the consumer welfare

9 standard that were given were essentially critical

10 rhetoric, that the basis of public support for antitrust

11 is if people believe it is about consumers rather than

12 about the economy as a whole, but there was not a

13 substantive argument offered in its favor. So, I do go

14 with that.

15 As far as the harm to competitors being unduly

16 counted goes, I suspect that in these cases, that is

17 balanced out by profits to the perpetrator or gains to

18 the consumers or someplace, that that is all going to be

19 just a transfer, and you still end up with the assorted

20 dead weight losses versus efficiencies that we are

21 familiar with.

22 As far as what sort of tests to use, more

23 specifically, on what I would call the exclusion cases,

24 I basically view those as essentially horizontal, taking

25 up an ever larger share of this complement market

Page 184

1 through exclusive dealing contracts or whatever it might

2 be and that we have horizontal tools for looking at

3 that. So, that is what I would use there.

4 On the predatory buying or predation cases

5 generally, I do not know enough to know at what point

6 one hits that balance between type I and type II error,

7 but that to me is what it is about, I think rather than

8 attempting to get each case exactly right, and I am just

9 going to leave it at that.

10 MS. SCHULTHEISS: I am going to let Steve make a

11 quick response and then Jack.

12 DR. SALOP: I am not surprised that the AMC was

13 confused during the hearings on the welfare standard.

14 Rick Rule testified that Bork used the Williamson

15 Diagram, he called it consumer welfare, as Rick said,

16 that was the old days when we were trying to confuse

17 people. The Supreme Court cited the Bork book.

18 Therefore, what the Supreme Court meant by consumer

19 welfare is total welfare, and that is the law, you know,

20 so I can certainly see why the AMC would get confused.

21 With respect to what the law ought to be, you

22 know, I think it is a complicated argument, but to Tim I

23 would say the following: It is not just a transfer.

24 Suppose you have entry into an industry by a relatively

25 high-cost entrant, enters the market, prices begin to

Page 185

1 come down, benefiting consumers, and then the

2 monopolist, which has got much lower costs, kills the

3 entrant, prices go back up, suppose demand is relatively

4 inelastic? Well, in that situation, the killing of the

5 entrant would raise total welfare, and I thought you

6 said in your statement that standards like that were no

7 good, but regardless of what you said, it is quite clear

8 that that is conduct that I think ought to be illegal.

9 MS. SCHULTHEISS: I want to get to Jack now.

10 MR. KIRKWOOD: Sure, sure. We are looking at

11 various welfare standards now, which is inevitably a

12 somewhat complex topic. One choice is between total

13 welfare and what Rick called third-party welfare, so let

14 me just talk about total welfare versus consumer

15 welfare.

16 That is ultimately a value choice. You can

17 think of the famous Williamsonian case where the merger

18 lowers cost but raises price. My value judgment is that

19 antitrust ought to stop that. There are arguments pro

20 and con, but the legislative history seems to reflect

21 that judgment, and every court that has ever faced that

22 issue has come out the same way.

23 On the more difficult and more judgmental choice

24 of supplier welfare versus consumer welfare, I agree

25 with much of what Steve said and originally wrote my

Page 186

1 article using a consumer welfare test, even in monopsony

2 cases, because generally there is a link between the

3 adverse impact on suppliers and the adverse impact on

4 consumers, but people raised two issues with me.

5 One, what about the case law? The case law

6 generally favors supplier welfare in a monopsony or

7 cartel case, and two, what about those instances, of

8 which Steve has described, where there is an adverse

9 impact on suppliers, suppliers are exploited, just like

10 in that merger, prices to them are lower, but no impact

11 whatsoever on consumers? Does that allow the practice?

12 And it does not seem it should.

13 DR. WARREN-BOULTON: Can I just make an example,

14 and I am responding really to Steve, to his statement

15 that somehow once we start talking about producer

16 welfare, we have to take into account the welfare of

17 competitors, and I do not think that is true. Let me

18 give you an example.

19 Suppose that I am a Kansas wheat farmer, and

20 what I do is I burn down all my neighbor's fields in the

21 county. Now, do we have a harm here? Yes. The

22 question is, is it an antitrust harm?

23 Now, you know, it is unlikely to me that I am

24 going to burn down all my neighbor's wheat farms because

25 I think that as a result, you know, there will be a

Page 187

1 shortage of wheat and I will be able to raise the price

2 of wheat. So, I do not have any consumer harm at all.

3 Let's suppose that there is no input problem at all.

4 Then I typically have harm. I do not have an antitrust

5 harm. I still presumably go to jail for arson, right?

6 So, I have harm, but it is not an antitrust harm.

7 Antitrust harm gets triggered when I have a

8 market impact, and it could be one of two things. It

9 could be that I really burn down enough wheat to

10 actually raise the Chicago price of wheat, unlikely, but

11 if somebody came to me and said, you know, what the guy

12 did is he burned down all his neighbor's wheat farms,

13 and why did he do it? He said because there is a local

14 labor market for workers, and if I burn down and put out

15 of business all my local -- you know, who are with me in

16 the local labor market, I will be able to reduce the

17 prices I have to pay my workers.

18 Now, that is monopsony. Now, do I think that is

19 an antitrust violation? Yes, I would say that that is

20 an antitrust violation. I do not care whether it is

21 raising the price of wheat or reducing the wage rate of

22 farmers, nor do I think, in particular, that somehow

23 that one, you know, on some ethical standard, they are

24 any different. In fact, wheat bread consumers are

25 probably richer than farm laborers. So, there is some

Page 188

1 distinction. I do not think you have to say that

2 somehow by bringing in producer surplus, we are somehow

3 worrying about competitors.

4 DR. SALOP: Well, you said something very

5 different. You said you would be willing to find an

6 antitrust violation where there is no consumer harm.

7 That is different from saying that you are adopting the

8 total welfare standard as the overarching standard to

9 govern antitrust cases, because if you were adopting the

10 total welfare standard to govern antitrust cases, then

11 the simple business tort of burning down your neighbor's

12 fields would lower total welfare, and it could support

13 an antitrust violation.

14 Certainly if you and your neighbor burned down

15 everybody else's fields, so you would not get into all

16 this complexity -ï½­

17 MS. SCHULTHEISS: Let's deal with a single firm,

18 which is what we are really focusing on.

19 DR. SALOP: Actually, we are not. We are

20 focusing on what the welfare standard should be.

21 MS. SCHULTHEISS: Right, but in connection with

22 Section 2.

23 DR. SALOP: Well, why would you have a different

24 welfare standard for Section 1 and Section 2? Why would

25 you want to make -- why would you want to gerrymander

Page 189

1 antitrust and make it incoherent? You know, Tim pointed

2 out this notion of the pipeline, you know, if you have

3 different rules governing how -- if the rules are very

4 different according to whether the pipeline does a

5 tolling agreement where they sell services or whether

6 they buy the gas and resell it at the other end, you are

7 going to have incoherent antitrust. If you are going to

8 have dramatically different rules for tying, exclusive

9 dealing, vertical mergers, where one's per se illegal,

10 one's rule of reason, and one's virtually per se legal,

11 well, since lawyers can characterize conduct pretty

12 easily as any one of those three boxes, you are going to

13 end up with very incoherent antitrust, which we did

14 until the eighties. So, I think saying this is Section

15 2, not Section 1, that is a recipe for disaster.

16 MS. SCHULTHEISS: Janet, did you want to respond

17 to that?

18 MS. McDAVID: No.

19 MS. SCHULTHEISS: Okay. Rick?

20 DR. WARREN-BOULTON: Well, the broad question

21 is, gee, do we have to have the same rules in every

22 situation? My answer is no. If you can distinguish

23 between situations, then you can have the same rules -ï½­

24 you can have different rules. There is no -ï½­

25 MS. SCHULTHEISS: But would the test be the

Page 190

1 same?

2 DR. WARREN-BOULTON: Sure. I mean, I think the

3 test for naked price-fixing is different than the

4 test -- you know, for the safe harbor for naked

5 price-fixing is very small. I think the safe harbor

6 for, you know, predatory pricing should be quite large.

7 MS. SCHULTHEISS: No, I mean in terms of the

8 consumer welfare versus total welfare, that test should

9 be the same across all of the violations or different?

10 Steve is arguing I think that it should be the same

11 welfare test regardless of the violation.

12 DR. WARREN-BOULTON: Oh, I see, yes. You can

13 argue what is the best welfare, but somehow suppliers

14 are upstream or downstream, welfare does not change

15 depending on -ï½­

16 MS. SCHULTHEISS: So, you agree with Steve,

17 then, that whatever welfare test -- whatever welfare

18 test is chosen should apply across all of the various

19 violations you're looking at?

20 DR. WARREN-BOULTON: As long as he chooses my

21 welfare test, then -ï½­

22 MS. SCHULTHEISS: Janet, would you agree with

23 that? Yes, no?

24 MS. McDAVID: I think we should probably move

25 on.

Page 191

1 MS. SCHULTHEISS: Well, I think the test is

2 important, though, and I think whether you apply the

3 same test or not is an important issue.

4 DR. WARREN-BOULTON: It has a huge effect on the

5 case in question.

6 MS. SCHULTHEISS: And I really would want to

7 know whether you think we should be using a consumer

8 welfare test or a total welfare test regardless of the

9 type of violation you are looking at.

10 MS. McDAVID: Well, I think I started by saying

11 much as Rick did, that the suppliers are entitled to a

12 competitive market just as buyers are, and so I would

13 not exclude a remedy for suppliers by using a test that

14 focused entirely on consumers.

15 MS. SCHULTHEISS: Exclusively.

16 DR. SALOP: Suppose two firms get together and

17 exchange a technology, a labor-saving technology, that

18 enables them to produce the same amount of output with

19 less labor, and as a result of that agreement, the firms

20 demand less labor and the wage rate goes down or some

21 other input, if you will, and suppliers -- here the

22 suppliers of labor are harmed. Do they have standing to

23 bring an antitrust case against that agreement?

24 DR. BRENNAN: No, because total welfare went up.


Page 192

1 DR. WARREN-BOULTON: Total welfare -- I mean -ï½­

2 MS. SCHULTHEISS: Let's do this in order. Jack

3 looked like he wanted to respond. Let me let Jack

4 respond first.

5 MR. KIRKWOOD: Yes, that was a point I didn't

6 comment on before, but I do agree with Steve there. If

7 you do have a case in which there is supplier harm but

8 consumer benefit, then I would go with the consumer

9 welfare standard. I think that does make antitrust more

10 coherent.

11 DR. BRENNAN: Total welfare went up in that

12 standard, so I would stick with it.

13 DR. SALOP: Yes. Now, how do you know that

14 total welfare went up? Did you do that calculation?

15 DR. BRENNAN: It is a -- I guess it is a

16 presumption for me, kind of 101, yes.

17 MS. SCHULTHEISS: I mean, I am going to give

18 Rick and Janet one more chance, and then we will get off

19 this.

20 DR. WARREN-BOULTON: How about we get a group of

21 consumers that get together a group of monopsonized

22 consumers, is this somehow a good thing? I don't think

23 so.

24 DR. SALOP: That is the cartel case decided by

25 Judge Breyer at the time, and he said it was okay.

Page 193


2 MS. McDAVID: I think in this -- in the

3 circumstance that Steve has posited, we have got an

4 integrated joint venture that will be evaluated under

5 the rule of reason, there is an efficiency and a

6 business rationale, and under the rule of reason, you

7 probably do not conclude it is illegal.

8 MS. SCHULTHEISS: Well, I want to get to

9 something else that is completely off of this topic for

10 a few minutes, because I want to make sure we have time

11 for it, and that is related to relief and remedies.

12 How do you deal with relief and remedies and in

13 particular in the overbidding situation? Should a court

14 enjoin the defendant's pricing? How do you deal with

15 that issue in the context of this type of conduct?

16 MS. McDAVID: Well, in the private party

17 litigation, it is going to be damages.

18 MS. SCHULTHEISS: I mean in a government case.

19 MS. McDAVID: In a government case, probably -ï½­

20 well, then you need to have a standard, and a standard

21 of fairness does not allow a remedy.

22 MS. SCHULTHEISS: But even if you have the

23 Brooke Group standard, what is the Government's remedy

24 in that situation on the predatory bidding, you know,

25 raise your prices or lower your prices? I mean -ï½­

Page 194

1 MS. McDAVID: It is pretty unlikely to me that

2 this is going to be a government case, although I think

3 there is a role for the Government to play in Section 2

4 enforcement, an important role. These cases are going

5 to come up in disputes among rivals.

6 MS. SCHULTHEISS: So, you are thinking the

7 overbidding context, it is not the type of case that the

8 Government should be getting into?

9 MS. McDAVID: It is the type of case that I

10 doubt the Government will get into. It is the type -ï½­

11 the Government typically intervenes in cases where there

12 is kind of a broader principle to be generated. This is

13 typically an intrafirm dispute or interfirm dispute, and

14 they rarely get into those circumstances.

15 MS. SCHULTHEISS: Well, when we are looking at

16 what test to apply, though, you know, for example, you

17 know, we did file an amicus brief in this case,

18 obviously concerned with the test and the jury

19 instructions from the Ninth Circuit, if it had been a

20 government case, and I mean this sincerely, what kind of

21 relief could the Government get in any kind of an

22 overbuying monopsony type case like this? Does anybody

23 have any -ï½­

24 DR. WARREN-BOULTON: I am about to agree with

25 Janet. I think there are situations in which we rely

Page 195

1 primarily on deterrents -- there are situations in which

2 deterrents, ex post penalties, are simply much, much

3 less expensive than ex ante penalties. I mean, you

4 think of how would you have hypothetically solved

5 Weyerhaeuser. Well, I suppose to me the answer is you

6 deter basically through either private litigation or a

7 fine. How could you prevent this situation? Would you

8 have to divest? You would have to sort of break up

9 Weyerhaeuser.

10 It seems to me that facing an alternative,

11 looking at the costs of preventing bad behavior through

12 structural means, that is to say, you know, forcing

13 Weyerhaeuser to sell off sawmills versus simply having

14 an ex post, you know, you will face damages if you, you

15 know, behave badly. A behavioral remedy is probably

16 better than a structural remedy, and if it is

17 behavioral, it seems to me I agree with Janet, it is

18 really something -- it is for private litigation.

19 Steve?

20 MS. SCHULTHEISS: Would you agree with that,

21 DR. SALOP: No, I thought -- I think there is an

22 answer. First of all, what the Government usually does

23 is it gets an injunction. You tell them not to violate

24 the standard anymore, and once you know what your

25 standard is -ï½­

Page 196

1 MS. SCHULTHEISS: We have to have a standard,

2 yes.

3 DR. SALOP: -- which hopefully you will know

4 once you bring the case, then you tell them, don't

5 violate the standard.

6 Secondly, I am not sure why Rick was so negative

7 with respect to structural relief. If what Weyerhaeuser

8 did was it knocked its rivals out of business and

9 thereby got a monopsony, then the way to jump-start

10 buy-side competition is to make Weyerhaeuser divest some

11 of its mills to re-establish that competition, or

12 perhaps if Weyerhaeuser simply has one big mill, you

13 would make Weyerhaeuser subsidize the entry of the small

14 sawmills that it knocked out of business. So, there is

15 a potential structural remedy there, or, of course,

16 maybe this would be one of the places where the FTC

17 should seek disgorgement.

18 MS. McDAVID: You see, that is precisely the

19 remedy for private damages. That's what private damages

20 are designed to achieve, is what Steve has just

21 described.


23 DR. BRENNAN: I agree with Steve, and I think

24 Ken said it earlier today, about if you have a

25 structural remedy, that that would be a better thing to

Page 197

1 do. I think the instance was slots, you know, making

2 that more competitive or something for airports.

3 The one thing that came up in the morning

4 discussion and also here that I honestly do not

5 understand is that somehow that seeking damages is

6 different and that this question only comes up with the

7 Government seeking injunctive relief, because if you

8 have a world where damages are collectible, then people

9 out there have to know what to do to avoid having to pay

10 damages, and the threat of damages in predation cases

11 is, I think, going to involve some kind of, in effect,

12 price regulation that says, you know, if you are bidding

13 now, if you set your price -- if you pay -- and that's

14 in essence the whole problem with the jury instruction,

15 right, that if you pay this much, it's okay, but if you

16 go beyond that, then you are paying too much, and then

17 you are liable to damages, and it is the threat of

18 damages that basically you are going to force people to

19 say, how high a price can I pay and I can't go above

20 that.

21 So, there is something I think inevitable about

22 this kind of law apart from the Government's specific

23 remedies that says that some prices are okay and some

24 prices aren't okay.

25 DR. SALOP: But that is not a lot different than

Page 198

1 saying you can have exclusive dealing arrangements with

2 six supermarkets but not eight. You talked earlier

3 about the share remedy, so why is one less administrable

4 than the other?

5 DR. BRENNAN: I mean, I think that is a good

6 question. I mean, it's -- you know, we are more

7 comfortable with merger law than we are with price

8 regulation, and maybe we shouldn't be.


10 MR. KIRKWOOD: Yes, Pat, your question is

11 excellent, because it forces all of us who are thinking

12 about what standards should be ideal to think about them

13 in the concrete situation where a court might enter an

14 injunction that incorporates the standard, and so if I

15 think a particular standard is appropriate, that would

16 incline me to believe that an injunction that wrote it

17 down would be appropriate, too.

18 I am a little reluctant, for the reason Ken

19 suggested, antitrust normally does not get into direct

20 price regulation, and you would think in the kind of

21 structured rule of reason that Steve and I have been

22 advocating, though with some differences, that there

23 would be more flexibility in the way a court would

24 interpret such a standard in an actual case than the way

25 a court might interpret an order that was written down

Page 199

1 where someone was seeking contempt sanctions. So, I

2 would be a little reluctant to enter an order here and

3 would want to pay attention to the possibility of

4 structural relief.

5 Weyerhaeuser had six different mills in the

6 area, so it is not inconceivable, and it acquired those

7 mills rather than building them internally.

8 MS. SCHULTHEISS: So, you would envision that it

9 would be possible for the Government to seek relief

10 other than just an injunction?


12 MS. SCHULTHEISS: Putting aside damages, you

13 know, the DOJ going for treble damages?

14 MR. KIRKWOOD: I mean, the preference, as Janet

15 suggested and Rick, is for -ï½­

16 MS. SCHULTHEISS: Deterrence.

17 MR. KIRKWOOD: -- private action.


19 DR. WARREN-BOULTON: One little problem, and I

20 think it was echoed by somebody this morning, who said

21 that, you know, the harm to competitors may correlate

22 but is not a very good measure of the harm to either

23 consumers or to producers. In the Ross-Simmons case,

24 the person who is suing for damages is Ross-Simmons,

25 another sawmill. I mean, the whole theory of this is

Page 200

1 the person who should be suing for damages is the timber

2 owner. I mean, they are the guys who supposedly have

3 been harmed by this, who are being monopsonized.

4 MS. McDAVID: They weren't sure, they were

5 getting a little extra.

6 DR. WARREN-BOULTON: The question is, why do we

7 have the competitor suing rather than the timberer, and

8 the answer is because the competitor gets hurt first,

9 right? Now, we could, of course, simply have said

10 forget it, let's wait, right? But having the competitor

11 sued is a sort of an ex ante, prophylactic way to

12 prevent presumably the monopsony harm to the person you

13 are really worried about, you know, who probably never

14 turned up in this litigation, which was the people who

15 were actually sawing the timber. They are the guys who

16 were supposedly monopsonized.

17 MS. SCHULTHEISS: I am going to ask Ken and

18 Patrick, since they have spent the afternoon with us and

19 have clearly given a lot of thought to these issues, and

20 we talked about it extensively this morning on the

21 sell-side, whether you have any response to the remedies

22 on this side of it, of the issue.

23 DR. ELZINGA: I just have a couple reactions.

24 First of all, I must confess -ï½­

25 MS. SCHULTHEISS: Is there a mic near you?

Page 201

1 MS. McDAVID: First of all, you notice they are

2 sitting closer together.

3 MR. KIRKWOOD: And further away from us.

4 DR. ELZINGA: But you will notice it is I who

5 moved, and I did that as a symbolic gesture in response

6 to what Rick said, and also I moved to my left, which is

7 uncharacteristic for me, as my students would know.

8 Two things: First of all, this has been

9 extremely helpful to me. I didn't realize this whole

10 topic on the buy-side was such a big issue, and that is

11 a reflection perhaps of my being out of a particular

12 loop, but this was a great way to learn about it.

13 I also have rarely seen a group of antitrust

14 experts basically in as much agreement as this group has

15 had. I mean, you can talk about, well, is it consumer

16 welfare or total welfare, and everybody can get into

17 snipping about that, but I thought there was remarkable

18 consensus among the panelists on the topic. So, thanks.

19 Patrick, you are going to have to move down

20 here.

21 MS. SCHULTHEISS: Patrick?

22 DR. WARREN-BOULTON: And now for something

23 completely different.

24 DR. BOLTON: Ken pretty much stole my thunder.

25 I am very much in agreement. Just on the last point

Page 202

1 about damages, it occurred to me, some of the issues

2 were brought up similar to the case, the

3 Sotheby's-Christie's -ï½­

4 MS. SCHULTHEISS: The auction case?

5 DR. BOLTON: -- cartel, yes, where there was an

6 issue who was harmed and how do you -- was it the -- was

7 it the buyers of art or was it the sellers and who

8 should be the recipient of the damages, and in that

9 case, I forget how it was decided in the end, but there

10 was reason to believe that the wrong party was

11 collecting the damages. Maybe someone else will -ï½­


13 MS. SCHULTHEISS: Does anybody know?

14 MS. McDAVID: I don't remember who got them.


16 MS. SCHULTHEISS: But it was a damages issue.

17 DR. WARREN-BOULTON: Well, the question was, is

18 if you read the auction -- it is pretty much like the

19 Social Security question, you know, if you pose -ï½­

20 Social Security taxes, who pays it, is it labor or is it

21 the firm? If you ask anybody on the street, they say,

22 oh, I pay half of it and my employer pays half of it.

23 If you ask an economist, and he will say 95 percent is

24 paid for by the workers, but the incidence of a tax

25 bears very little resemblance to the accounting and

Page 203

1 collection of that tax, and I think in the same way, it

2 is extraordinarily difficult to find out who actually

3 pays for antitrust violations by simply looking at the

4 accounting incidence.

5 DR. SALOP: Yes, I -ï½­

6 MS. SCHULTHEISS: We have two minutes.

7 DR. SALOP: Okay, so I will just make a law

8 professor remark. There is Illinois Brick and there is

9 Hanover Shoe, and that says the direct purchasers get to

10 sue the direct purchasers or sellers of the art. They

11 get the money. There is no pass-on defense by the

12 cartel, and the -- you know, the Supreme Court had a

13 reason for that, and so the -- I would think the proper

14 people to get the money under the current law are the

15 sellers of the art, not the buyers.

16 MS. SCHULTHEISS: Okay. Well, I think with

17 that -ï½­

18 DR. ELZINGA: I worked for the judge on that

19 case. Most of it went to the lawyers.

20 MS. SCHULTHEISS: Well, on that note -ï½­

21 MS. McDAVID: The entrepreneurial plaintiff's

22 bar.

23 DR. WARREN-BOULTON: Well, at least some of it

24 went to you.

25 MS. SCHULTHEISS: On that note, we will wrap

Page 204

1 this session up, and I ask you to join me in giving a

2 hand to our panelists. Thank you very much.

3 (Applause.)

4 (Whereupon, at 3:58 p.m., the hearing was

5 concluded.)


Page 205

1 C E R T I F I C A T I O N   O F   R E P O R T E R



4 DATE: JUNE 22, 2006

6 I HEREBY CERTIFY that the transcript contained

7 herein is a full and accurate transcript of the notes

8 taken by me at the hearing on the above cause before the

9 FEDERAL TRADE COMMISSION to the best of my knowledge and

10 belief.

12 DATED: 7/9/06


18 C E R T I F I C A T I O N   O F   P R O O F R E A D E R

20 I HEREBY CERTIFY that I proofread the transcript

21 for accuracy in spelling, hyphenation, punctuation and

22 format.


Updated June 25, 2015