Page 1
1 | UNITED STATES FEDERAL TRADE COMMISSION
|
2 | and
|
3 | UNITED STATES DEPARTMENT OF JUSTICE
|
4 |
|
5 |
|
6 |
|
7 | SHERMAN ACT SECTION 2 JOINT HEARING
|
8 | PREDATORY PRICING
|
9 | THURSDAY, JUNE 22, 2006
|
10 |
|
11 |
|
12 |
|
13 |
|
14 | HELD AT:
|
15 | 600 PENNSYLVANIA AVENUE, N.W.
|
16 | WASHINGTON, D.C.
|
17 | UNITED STATES FEDERAL TRADE COMMISSION
|
18 | HEADQUARTERS BUILDING, ROOM 432
|
19 | 9:30 A.M. TO 4:00 P.M.
|
20 |
|
21 |
|
22 |
|
23 |
|
24 | Reported and transcribed by:
|
25 | Susanne Bergling, RMR-CLR |
Page 2
1 | MODERATORS:
|
2 | ROBERT POTTER
|
3 | Chief, Legal Policy Section
|
4 | Antitrust Division, Department of Justice
|
5 | and
|
6 | PATRICIA SCHULTHEISS
|
7 | Attorney
|
8 | Bureau of Competition, Federal Trade Commission
|
9 |
|
10 | PANELISTS:
|
11 | Morning Session:
|
12 | Patrick Bolton
|
13 | Kenneth G. Elzinga
|
14 | A. Douglas Melamed
|
15 | Janusz Ordover
|
16 |
|
17 | Afternoon Session:
|
18 | Tim Brennan
|
19 | John Kirkwood
|
20 | Janet L. McDavid
|
21 | Steven C. Salop
|
22 | Frederick R. Warren-Boulton
|
23 |
|
24 |
|
25 | |
Page 3
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 -
|
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 -
|
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 -
|
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 -
|
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 -
|
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 -
|
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 -
|
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 -
|
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 -
|
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 -
|
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 -
|
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 -
|
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 -
|
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 -
|
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 -
|
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 -
|
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 -
|
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 -
|
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.
|
9 | MR. MELAMED: Yes.
|
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 -
|
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 -
|
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 -
|
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 -
|
15 | DR. ORDOVER: Wait, should not be -
|
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 -
|
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 -
|
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 -
|
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 -
|
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 -
|
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 - |
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.)
|
11 |
|
12 |
|
13 |
|
14 |
|
15 |
|
16 |
|
17 |
|
18 |
|
19 |
|
20 |
|
21 |
|
22 |
|
23 |
|
24 |
|
25 | |
Page 101
1 | AFTERNOON SESSION
|
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 -
|
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 -
|
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 -
|
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 -
|
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.
|
13 | MS. SCHULTHEISS: Tim?
|
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 -
|
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.
|
22 | MS. SCHULTHEISS: Janet?
|
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.
|
17 | MS. SCHULTHEISS: Tim?
|
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 -
|
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.
|
17 | MS. SCHULTHEISS: Okay.
|
18 | DR. WARREN-BOULTON: I think I am probably
|
19 | agreeing with everybody. First of all, I agree that -
|
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.
|
2 | MS. SCHULTHEISS: Janet?
|
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 -
|
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 -
|
16 | MS. SCHULTHEISS: Right.
|
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 -
|
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 -
|
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 -
|
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 -
|
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 -
|
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 -
|
14 | MS. SCHULTHEISS: Well, that is horizontal,
|
15 | though.
|
16 | DR. SALOP: But that is where antitrust starts.
|
17 | MS. SCHULTHEISS: Right.
|
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.
|
3 | MS. SCHULTHEISS: Okay, Tim?
|
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 -
|
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 -
|
24 | you can have different rules. There is no -
|
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 -
|
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 -
|
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.
|
25 | MS. SCHULTHEISS: Jack? |
Page 192
1 | DR. WARREN-BOULTON: Total welfare -- I mean -
|
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
1 | DR. WARREN-BOULTON: Well -
|
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 -
|
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 - |
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 -
|
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 -
|
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 - |
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.
|
22 | DR. WARREN-BOULTON: I agree.
|
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.
|
9 | MS. SCHULTHEISS: Jack?
|
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?
|
11 | MR. KIRKWOOD: Yes.
|
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 -
|
16 | MS. SCHULTHEISS: Deterrence.
|
17 | MR. KIRKWOOD: -- private action.
|
18 | MS. SCHULTHEISS: Rick?
|
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 -
|
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 -
|
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 -
|
12 | DR. WARREN-BOULTON: I agree.
|
13 | MS. SCHULTHEISS: Does anybody know?
|
14 | MS. McDAVID: I don't remember who got them.
|
15 | DR. WARREN-BOULTON: Yes.
|
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 -
|
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 -
|
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 -
|
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 -
|
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.)
|
6 |
|
7 |
|
8 |
|
9 |
|
10 |
|
11 |
|
12 |
|
13 |
|
14 |
|
15 |
|
16 |
|
17 |
|
18 |
|
19 |
|
20 |
|
21 |
|
22 |
|
23 |
|
24 |
|
25 | |
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
|
2 | DOCKET/FILE NUMBER: P062106
|
3 | CASE TITLE: SECTION 2 HEARING, PREDATORY PRICING
|
4 | DATE: JUNE 22, 2006
|
5 |
|
6 | I HEREBY CERTIFY that the transcript contained
|
7 | herein is a full and accurate transcript of the notes
|
8 | taken by me at the hearing on the above cause before the
|
9 | FEDERAL TRADE COMMISSION to the best of my knowledge and
|
10 | belief.
|
11 |
|
12 | DATED: 7/9/06
|
13 |
|
14 |
|
15 |
|
16 | SUSANNE BERGLING, RMR-CLR
|
17 |
|
18 | C E R T I F I C A T I O N O F P R O O F R E A D E R
|
19 |
|
20 | I HEREBY CERTIFY that I proofread the transcript
|
21 | for accuracy in spelling, hyphenation, punctuation and
|
22 | format.
|
23 |
|
24 |
|
25 | DIANE QUADE |
|